127425110 Liquid Filtration

FILTRATION

by Nicholas P. Cheremisinoff, Ph.D.

Environmental Policy and Technology Project

United States Agency for International Development

Boston Oxford Johannesburg Melbourne New Delhi Singapore

LIQUID

Copyright © 1998 by Butterworth-Heinemann

A member of the Reed Elsevier group

All rights reserved.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted inany form or by any means, electronic, mechanical, photocopying, recording, or otherwise,without the prior written permission of the publisher.

Recognizing the importance of preserving what has been written, Butterworth-Heinemannprints its books on acid-free paper whenever possible.

Butterworth-Heinemann supports the efforts of American Forests and theGlobal ReLeaf program in its campaign for the betterment of trees, forests,and our environment.

ISBN: 0-7506-7047-9

The publisher offers special discounts on bulk orders of this book.For information, please contact:Manager of Special SalesButterworth-Heinemann225 Wildwood AvenueWoburn,MA01801-2041Tel: 781-904-2500Fax: 781-904-2620

For information on all Butterworth-Heinemann publications available, contact ourWorld Wide Web home page at: http://www.bh.com

10 9 8 7 6 5 4 3 2 1

Printed in the United States of America

DEDICATION

This volume is dedicated to the memory of Paul N. Cheremisinoff, M.S., P.E., whoauthored more than 300 technical books over his career as a chemical engineer andwas among the pioneers of pollution control and prevention.

This page intentionally left blank

CONTENTS

Preface ix

About the Author xi

Chapter 1 An Introduction to Liquid Filtration 1

Introduction 1The Porous Media 2The Filter Media 9Liquid Filtration Classification 10The Formation of Filter Cake 11Typical Industrial Filtration Conditions 12Washing and Dewatering Operations 12General Considerations for Process Engineers 13The Objectives of Filtration 14Preparation Stages for Filtration 15Equipment Selection Methodology 16Nomenclature 18

Chapter 2 Filter Media and Use of Filter Aids 19

Introduction 19Flexible Filter Media 20Rigid Filter Media 34Filter Media Selection Criteria 43Introduction to the Use of Filter Aids 47Examples of Filter Aids 50Filter Aid Selection 51Suggested Readings 57Nomenclature 58

Chapter 3 Cake Filtration and Filter Media Filtration 59

Introduction 59Dynamics of Cake Filtration 60Constant-Rate Filtration 70

Contents

Variable-Rate and -Pressure Filtration 72Constant-Pressure and -Rate Filtration 75Filter-Medium Filtration Formulas 75Constant-Pressure-Drop Filtration 75Filtration Mechanisms 81Constant Rate Filtration 83Suggested Readings 86Nomenclature 87

Chapter 4 Industrial Filtration Equipment 88

Introduction 88Rotary Drum Filters 89Cocurrent Filters 91Cross Mode Filters 98Cartridge Filters 103Diaphragm Filters 110High Pressure, Thin Cake Filters 115Thickeners 117Solids Washing 120Centrifugal Filtration 120Screw Presses 123Ultrafiltration 124Reverse Osmosis 134Closure 141

Chapter 5 Application of Filtration to Wastewater Treatment 142

Introduction 142Granular Media Filtration 142Bed Regeneration 148Flocculation Filtration 149Slow Sand Filtration 151Rapid Sand Filtration 153Chemical Mixing, Flocculation, and Solids Contact Processes 155Suggested Readings 162

Chapter 6 Advanced Membrane Technology for Wastewater Treatment 163

IntroductionOverview of Technology Case StudyCase Study SpecificsTechnology ApplicationMechanisms of Membrane SeparationsTreatment of Hazardous Wastes

Contents vii

Features of the Hyperfiltration System 173Process Economics 184Detailed Process and Technology Description 193Summary of Case Study Analytical Results 202Closure 210

Chapter 7 Sludge Dewatering Operations 211

Introduction 211Overview of Dewatering Technologies 212Use of Drying Beds 217Use of Vacuum Filtration 219Use of Pressure Filtration 222Use of Centrifugation 223Alternative Mechanical Dewatering Techniques 226Suggested Readings 227

Chapter 8 Industrial Wastewater Sources 229

Introduction 229Paper and Allied Products Industry Wastes 230Dairy Products Industry Wastes 232Textile Industry Wastes 237Pharmaceutical Industry Wastes 240Leather Tanning and Finishing Industry Wastes 243Petroleum Refining Industry Wastes 246Food and Meat Packing Industry Wastes 251Beverages Industry Wastes 254Plastics and Synthetic Materials Industry Wastes 258Blast Furnaces, Steel Works, and Rolling and Finishing Wastes 261Organic Chemicals Industry Wastes 265Metal Finishing Industry Wastes 268Closure 271Suggested Readings 271

Chapter 9 Filtration Equipment and Process Flow Sheets 272

Introduction 272Index to Equipment and Flow Sheet Diagrams 272

Index 316


PREFACE

This volume has been written as an introductory reference and working guide to thesubject of liquid filtration engineering. The book is designed to acquaint the newcomerto industry practices, and general design and operating methodology for filtrationprocesses. Emphasis is given to pollution control applications, however thetechnologies and equipment described herein are equally applicable to productrecovery and product purification applications.

The information presented in this volume is based largely on the author's collectednotes and lectures over the past 15 years. The volume is not intended for researchesor equipment developers, but rather for process engineers, plant engineers, andtechnicians who require basic knowledge of this important unit operation. Much of thedesign methodology and working equations presented have been tested on pilot plantstudies and applied to commercial and semi-commercial operations with success,however, neither the author nor publisher provide written or implied endorsementsthat these procedures will work in any or all cases. As with any piece of equipmentor process, the designer must consult with specific vendors, suppliers andmanufacturers, and further, should field test or at a minimum, conduct pilot tests toensure performance in the intended application. Filtration equipment, operationconditions, and the use of filtration aids are highly dependent upon the properties ofthe suspension being filtered. Furthermore, overall process constraints and economicscan have major impacts on the selection of equipment, their operating modes andcharacteristics, and efficiency.

The author wishes to extend a heartfelt gratitude to Butterworth — Heineinann fortheir fine production of this volume, and to members of the United StatesEnvironmental Protection Agency for their advise and consultation on some of thematerials presented herein.

Nicholas P. Cheremisinoff, Ph.D.


ABOUT THE AUTHOR

Nicholas P. Cheremisinoff is Director of the Industrial Waste Management Programin Ukraine, which is supported by the United States Agency for InternationalDevelopment, Washington D.C. He has nearly twenty years of applied research andindustry experience in the petrochemicals, oil and gas, rubber, and steel industries,and is considered a leading authority on waste management and process design. Dr.Cheremisinoff provides technical consulting services to both private industry andgovernment agencies and has worked extensively in Republics of the former SovietUnion, South America, Korea, the United States, and Western Europe, He is theauthor, co-author, or editor of over 100 engineering reference books dealing withwaste technologies and process designs, including the multivolume Encyclopedia ofFluid Mechanics by Gulf Publishing Company. Dr. Cheremisinoff received his B.S.,M.S. and Ph.D. degrees in chemical engineering from Clarkson College ofTechnology.


AN INTRODUCTION TO LIQUIDFILTRATION

introduction

In the simplest of terms, filtration is a unit operation that is designed to separatesuspended particles from a fluid media by passing the solution through a porousmembrane or medium. As the fluid or suspension is forced through the voids or poresof the filter medium, the solid particles are retained on the medium's surface or, insome cases, on the walls of the pores, while the fluid, which is referred to as thefiltrate, passes through.

The flow of fluids through a porous medium is of interest not only to the unitoperation of filtration, but to other processes, such as adsorption, chromatography,operations involving the flow of suspensions through packed columns, ion exchange,and various reactor engineering applications. In petroleum engineering applications,interest lies in the displacement of oil with gas, water and miscible solvents (includingsolutions of surface-active agents), and in reservoir flow problems. In hydrology,interest is in the movement of trace pollutants in water systems, the recovery of waterfor drinking and irrigation, and saltwater encroachment into freshwater reservoirs.In soil physics, interest lies in the movement of water, nutrients and pollutants intoplants. In biophysics, the subject of flow through porous media touches upon lifeprocesses such as the flow of fluids in the lungs and the kidney.

The physical parameters that relate the porous material to the hydrodynamics of floware porosity, permeability, tortuosity and connectivity. This chapter discusses thefundamentals of flow through porous media and relates these principles to theindustrial operations of filtration. As indicated in the preface of this volume, thesubject of filtration is discussed from a process engineering viewpoint, and inparticular from that of the chemical engineer. Filtration has a long history in thechemical engineering field both from the standpoint of the production of high purityproducts, as well as a technology extensively used in pollution control and prevention.

1

2 Liquid Filtration

The Porous Media

A porous medium may be described as a solid containing many holes and tortuouspassages. The number of holes or pores is sufficiently great that a volume average isneeded to estimate pertinent properties. Pores that occupy a definite fraction of thebulk volume constitute a complex network of voids. The manner in which holes orpores are embedded, the extent of their interconnection, and their location, size andshape characterize the porous medium. The term porosity refers to the fraction of themedium that contains voids. When a fluid is passed over the medium, the fraction ofthe medium (i.e., the pores) that contributes to the flow is referred to as the effectiveporosity.

There are many materials that can be classified as porous media, however, not all ofthem are of interest to the subject of filtration. In general, porous media are classifiedas either unconsolidated and consolidated and/or as ordered and random. Examplesof unconsolidated media are sand, glass beads, catalyst pellets, column packings, soil,gravel and packing such as charcoal. Examples of consolidated media are most of thenaturally occurring rocks, such as sandstones and limestones. Materials such asconcrete, cement, bricks, paper and cloth are manmade consolidated media. Orderedmedia are regular packings of various types of materials, such as spheres, columnpackings and wood. Random media have no particular correlating factor.

Porous media can be further categorized in terms of geometrical or structuralproperties as they relate to the matrix that affects flow and in terms of the flowproperties that describe the matrix from the standpoint of the contained fluid.Geometrical or structural properties are best represented by average properties, fromwhich these average structural properties are related to flow properties.

A microscopic description characterizes the structure of the pores. The objective ofpore-structure analysis is to provide a description that relates to the macroscopic orbulk flow properties. The major bulk properties that need to be correlated with poredescription or characterization are porosity, permeability, tortuosity and connectivity.In studying different samples of the same medium, it becomes apparent that thenumber of pore sizes, shapes, orientations and interconnections are enormous. Dueto this complexity, pore-structure description is most often a statistical distribution ofapparent pore sizes. This distribution is apparent because to convert measurements topore sizes one must resort to models that provide average or model pore sizes. Acommon approach to defining a characteristic pore size distribution is to model theporous medium as a bundle of straight cylindrical capillaries. The diameters of themodel capillaries are defined on the basis of a convenient distribution function.

Pore structure for unconsolidated media is inferred from a particle size distribution,the geometry of the particles and the packing arrangement of particles. The theory ofpacking is well established for symmetrical geometries such as spheres. Informationon particle size, geometry and the theory of packing allows relationships between poresize distributions and particle size distributions to be established.

An Introduction to Liquid Filtration 3

A macroscopic description is based on average or bulk properties at sizes much largerthan a single pore. In characterizing a porous medium macroscopically, one must dealwith the scale of description. The scale used depends on the manner and size in whichwe wish to model the porous medium. A simplified, but sometimes accurate, approachis to assume the medium to be ideal; meaning homogeneous, uniform and isotropic.

The term reservoir description is applied to characterizing a homogeneous system asopposed to heterogeneous. A reservoir description defines the reservoir at a levelwhere a property changes sufficiently so that more than a single average must be usedto model the flow. In this sense, a reservoir composed of a section of coarse graveland a section of fine sand, where these two materials are separated and havesignificantly different permeabilities, is heterogeneous in nature. Defining dimensions,locating areas and establishing average properties of the gravel and sand constitutesa reservoir description, and is a satisfactory approach for reservoir-level typeproblems. Unfortunately, to study the mechanisms of flow, the effects of nonidealmedia require more specific definitions.

Any discussion of flow through porous media inevitably touches upon Darcy's lawwhich is a relationship between the volumetric flowrate of a fluid flowing linearlythrough a porous medium and the energy loss of the fluid in motion.

Darcy's law is expressed as:

Q = ---- (1)A/z

where

A/z = Az + —— + constantP

The parameter, K, is a proportionality constant that is known as the hydraulicconductivity.

The relation is usually considered valid for creeping flow where the Reynolds number,as defined for a porous medium, is less than one. The Reynolds number in openconduit flow is the ratio of inertial to viscous forces and is defined in terms of acharacteristic length perpendicular to flow for the system. Using four times thehydraulic radius to replace the length perpendicular to flow and correcting the velocitywith porosity yields a Reynolds number in the form:

D v pRe = __JLj£L (3)

Darcy's law is considered valid where Re < 1.

4 Liquid Filtration

The hydraulic conductivity K depends on the properties of the fluid and on the porestructure of the medium. The hydraulic conductivity is temperature-dependent, sincethe properties of the fluid (density and viscosity) are temperature-dependent.Hydraulic conductivity can be written more specifically in terms of the intrinsicpermeability and the properties of the fluid.

K = (4)

where k is the intrinsic permeability of the porous medium and is a function only ofthe pore structure. The intrinsic permeability is not temperature-dependent. Indifferential form, Darcy's equation is:

Q k dpJ~ = /7 = - f /C\

A " J P)A j. ax

The minus sign results from the definition of Ap, which is equal to p2 - p l5 a negativequantity. The term q is the seepage velocity and is equivalent to the velocity ofapproach vro, which is also used in the definition of the Reynolds number.

Permeability is normally determined using linear flow in the incompressible orcompressible form, depending on whether a liquid or gas is used as the flowing fluid.The volumetric flowrate Q (or Qm) is determined at several pressure drops. Q (or Qm)is plotted versus the average pressure pm. The slope of this line will yield the fluidconductivity K or, if the fluid density and viscosity are known, it provides the intrinsicpermeability k. For gases, the fluid conductivity depends on pressure, so that

K = K\ 1+-

where b depends on the fluid and the porous medium. Under such circumstances astraight line results (as with a liquid), but it does not pass through the origin; insteadit has a slope of bK and intercept K. The explanation for this phenomenon is that gasesdo not always stick to the walls of the porous medium. This slippage shows up as anapparent dependence of the permeability on pressure.

Heterogeneity, nonuniformity and anisotropy must be defined in the volume-averagesense. These terms may be defined at the level of Darcy's law in terms ofpermeability. Permeability is more sensitive to conductance, mixing and capillarypressure than to porosity.

Heterogeneity, nonuniformity and anisotropy are defined as follows. On amacroscopic basis, they imply averaging over elemental volumes of radius e about apoint in the media, where e is sufficiently large that Darcy's law can be applied forappropriate Reynolds numbers. In other words, volumes are large relative to that of


a single pore. Further, e is the minimum radius that satisfies such a condition. If e istoo large, certain nonidealities may be obscured by burying their effects far within theelemental volume.

Heterogeneity, nonuniformity and anisotropy are based on the probability densitydistribution of permeability of random macroscopic elemental volumes selected fromthe medium, where the permeability is expressed by the one-dimensional form ofDarcy's law.

As noted earlier, the principal properties of nonideal porous media that establish thenature of the fluid flow are porosity, permeability, tortuosity and connectivity. In amacroscopic sense, porosity characterizes the effective pore volume of the medium.It is directly related to the size of the pores relative to the matrix. When porosity issubstituted, the details of the structure are lost.

Permeability is the conductance of the medium and has direct relevance to Darcy'slaw. Permeability is related to the pore size distribution, since the distribution of thesizes of entrances, exits and lengths of the pore walls constitutes the primaryresistance to flow. This parameter reflects the conductance of a given pore structure.Permeability and porosity are related; if the porosity is zero the permeability is zero.Although a correlation between these two parameters may exist, permeability cannotbe predicted from porosity alone, since additional parameters that contain moreinformation about the pore structure are needed. These additional parameters aretortuosity and connectivity. Tortuosity is defined as the relative average length of aflow path (i.e., the average length of the flow paths to the length of the medium). Itis a macroscopic measure of both the sinuosity of the flow path and the variation inpore size along the flow path. Both porosity and tortuosity correlate with permeability,but neither can be used alone to predict permeability.

Connectivity defines the arrangement and number of pore connections. For monosizepores, connectivity is the average number of pores per junction. The term representsa macroscopic measure of the number of pores at a junction. Connectivity correlateswith permeability, but cannot be used alone to predict permeability except in certainlimiting cases.

Difficulties in conceptual simplifications result from replacing the real porous mediumwith macroscopic parameters that are averages and that relate to some idealized modelof the medium. Tortuosity and connectivity are different features of the pore structureand are useful to interpret macroscopic flow properties, such as permeability, capillarypressure and dispersion.

Porous media is typically characterized as an ensemble of channels of various crosssections of the same length. The Navier-Stokes equations for all channels passing across section normal to the flow can be solved to give:

6 Liquid Filtration

Where parameter c is known as the Kozeny constant, which is essentially a shapefactor that is assigned different values depending on the configuration of the capillary(c = 0.5 for a circular capillary). S is the specific surface area of the channels. Forother than circular capillaries, a shape factor is included:

~> ckr" - — (8)l '

The specific surface for cylindrical pores is:

„ _ n2nrL _ 2Â J7 ~ -nnr L r

and

^2 2(j)

Wk (10)s2 ->-> A ~

Replacing 2/81/4 with shape parameter c and SA with a specific surface, the well knownKozeny equation is obtained.

Tortuosity T is basically a correction factor applied to the Kozeny equation to accountfor the fact that in a real medium the pores are not straight (i.e., the length of the mostprobable flow path is longer than the overall length of the porous medium):

*2 •

To determine the average porosity of a homogeneous but nonuniform medium, thecorrect mean of the distribution of porosity must be evaluated. The porosities ofnatural and artificial media usually are normally distributed. The average porosity ofa heterogeneous nonuniform medium is the volume-weighted average of the numberaverage:

m

E

E",


The average nonuniform permeability is spatially dependent. For a homogeneous butnonuniform medium, the average permeability is the correct mean (first moment) ofthe permeability distribution function. Permeability for a nonuniform medium isusually skewed. Most data for nonuniform permeability show permeability to bedistributed log-normally. The correct average for a homogeneous, nonuniformpermeability, assuming it is distributed log-normally, is the geometric mean, definedas:

<*> - n*,1 = 1

(14)

For flow in heterogeneous media, the average permeability depends on thearrangement and geometry of the nonuniform elements, each of which has a different,average permeability. Figure 1 conceptually illustrates nonuniform elements, wherethe elements are parallel to the flow.

Figure 1. Flow through parallel nonuniform elements of porous media.

Since flow is through parallel elements of different constant area, Darcy's law foreach element, assuming the overall length of each element is equal, is:

(15)

The flowrate through the entire system of elements is Q=QS+Q2+.Combining these expressions we obtain:

(16a)

Liquid Filtration

or

(I6b)

This means that the average permeability for this heterogeneous medium is thearea-weighted average of the average permeability of each of the elements. If thepermeability of each element is log-normally distributed, these are the geometricmeans. Reservoirs and soils are usually composed of heterogeneities that arenonuniform layers, so that only the thickness of the layers varies. This means that\{kp)} simplifies to:

h.(k.) + /L<JL> + . . .

If all the layers have the same thickness, then

h

£*,«*», = * <18)

where n is the number of layers.

Permeability is a volume-averaged property for a finite but small volume of a medium.Anisotropy in natural or manmade packed media may result from particle (or grain)orientation, bedding of different sizes of particles or layering of media of differentpermeability. A dilemma arises when considering whether to treat a directional effectas anisotropy or as an oriented heterogeneity.

In an oriented porous medium, the resistance to flow differs depending on thedirection. Thus, if there is a pressure gradient between two points and a particularfluid particle is followed, unless the pressure gradient is parallel to oriented flowpaths, the fluid particle will not travel from the original point to the point which onewould expect. Instead, the particle will drift.

Tortuosity and connectivity are difficult to relate to the nonuniformity and anisotropyof a medium. Attempts to predict permeability from a pore structure model requireinformation on tortuosity and connectivity.

From an industrial viewpoint, the objective of the unit operation of filtration is theseparation of suspended solid particles from a process fluid stream which isaccomplished by passing the suspension through a porous medium that is referred toas a filter medium. In forcing the fluid through the voids of the filter medium, fluidalone flows, but the solid particles are retained on the surface and in the medium'spores. The fluid discharging from the medium is called the filtrate. The operation may


be performed with either incompressible fluids (liquids) or slightly to highlycompressible fluids (gases). The physical mechanisms controlling filtration, althoughsimilar, vary with the degree of fluid compressibility. Although there are markedsimilarities in the particle capture mechanisms between the two fluid types, designmethodologies for filtration equipment vary markedly. This reference volumeconcentrates only on process liquid handling (i.e., incompressible fluid processing).

The Filter Media

The filter medium represents the heart of any filtration device. Ideally, solids arecollected on the feed side of the plate while filtrate is forced through the plate andcarried away on the leeward side. A filter medium is, by nature, inhomogeneous, withpores nonuniform in size, irregular in geometry and unevenly distributed over thesurface. Since flow through the medium takes place through the pores only, themicro-rate of liquid flow may result in large differences over the filter surface. Thisimplies that the top layers of the generated filter cake are inhomogeneous and,furthermore, are established based on the structure and properties of the filtermedium. Since the number of pore passages in the cake is large in comparison to thenumber in the filter medium, the cake's primary structure depends strongly on thestructure of the initial layers. This means that the cake and filter medium influenceeach other.

Pores with passages extending all the way through the filter medium are capable ofcapturing solid particles that are smaller than the narrowest cross section of thepassage. This is generally attributed to particle bridging or, in some cases, physicaladsorption.

Depending on the particular filtration technique and intended application, differentfilter media are employed. Examples of common media are sand, diatomite, coal,cotton or wool fabrics, metallic wire cloth, porous plates of quartz, chamotte, sinteredglass, metal powder, and powdered ebonite. The average pore size and configuration(including tortuosity and connectivity) are established from the size and form ofindividual elements from which the medium is manufactured. On the average, poresizes are greater for larger medium elements. In addition, pore configuration tends tobe more uniform with more uniform medium elements. The fabrication method of thefilter medium also affects average pore size and form. For example, porecharacteristics are altered when fibrous media are first pressed. Pore characteristicsalso depend on the properties of fibers in woven fabrics, as well as on the exactmethods of sintering glass and metal powders. Some filter media, such as cloths(especially fibrous layers), undergo considerable compression when subjected totypical pressures employed in industrial filtration operations. Other filter media, suchas ceramic, sintered plates of glass and metal powders, are stable under the sameoperating conditions. In addition, pore characteristics are greatly influenced by theseparation process occurring within the pore passages, as this leads to a decrease ineffective pore size and consequently an increase in flow resistance. This results fromparticle penetration into the pores of the filter medium.

10 Liquid Filtration

The separation of solid particles from a liquid via filtration is a complicated process.For practical reasons filter medium openings should be larger than the average sizeof the particles to be filtered. The filter medium chosen should be capable of retainingsolids by adsorption. Furthermore, interparticle cohesive forces should be largeenough to induce particle flocculation around the pore openings.

Liquid Filtration Classification

There are two major types of filtration: "cake" and "filter-medium" filtration. In theformer, solid particulates generate a cake on the surface of the filter medium. Infilter-medium filtration (also referred to as clarification), solid particulates becomeentrapped within the complex pore structure of the filter medium. The filter mediumfor the latter case consists of cartridges or granular media. Examples of granularmaterials are sand or anthracite coal.

Process engineers who specify filtration equipment for an intended application mustfirst account for the parameters governing the application and then select the filtrationequipment best suited for the job. There are two important parameters that must beconsidered, namely the method to be used for forcing liquid through the medium, andthe material that will constitute the filter medium.

When the resistance opposing fluid flow is small, gravity force effects fluid transportthrough a porous filter medium. Such a device is simply called a gravity filter. Ifgravity is insufficient to instigate flow, the pressure of the atmosphere is allowed toact on one side of the filtering medium, while a negative or suction pressure is appliedon the discharge side. This type of filtering device is referred to as a vacuum filter.The application of vacuum filters is typically limited to 15 psi pressure. If greaterforce is required, a positive pressure in excess of atmospheric can be applied to thesuspension by a pump. This motive force may be in the form of compressed airintroduced in a montejus, or the suspension may be directly forced through a pumpacting against the filter medium (as in the case of a filter press), or centrifugal forcemay be used to drive the suspension through a filter medium as is done in screencentrifuges.

Filtration is a hydrodynamic process hi which the fluid's volumetric rate is directlyproportional to the existing pressure gradient across the filter medium, and inverselyproportional to the flow resistance imposed by the connectivity, tortuosity and size ofthe medium's pores, and generated filter cake. The pressure gradient constitutes thedriving force responsible for the flow of fluid.

Regardless of how the pressure gradient is generated, the driving force increasesproportionally. However, in most cases, the rate of filtration increases more slowlythan the rate at which the pressure gradient rises. The explanation for this phenomenonis that as the gradient rises, the pores of filter medium and cake are compressed andconsequently the resistance to flow increases. For highly compressible cakes, bothdriving force and resistance increase nearly proportionally and any rise in the pressuredrop has a minor effect on the filtration rate.


The Formation of Filter Cake

Filtration operations are capable of handling suspensions of varying characteristicsranging from granular, incompressible, free-filtering materials to slimes and colloidalsuspensions in which the cakes are incompressible. These latter materials tend tocontaminate or foul the filter medium. The interaction between the particles insuspension and the filter medium determines to a large extent the specific mechanismsresponsible for filtration.

In practice cake filtration is used more often than filter-medium filtration. Uponachieving a certain thickness, the cake is removed from the medium by variousmechanical devices or by reversing the flow of filtrate. To prevent the formation ofmuddy filtrate at the beginning of the subsequent filtration cycle, a thin layer ofresidual particles is sometimes deposited onto the filter medium. For the same reason,the filtration cycle is initiated with a low, but gradually increasing pressure gradientat an approximately constant flowrate. The process is then operated at a constantpressure gradient while experiencing a gradual decrease in process rate.

The structure of the cake formed and, consequently, its resistance to liquid flowdepends on the properties of the solid particles and the liquid phase suspension, as wellas on the conditions of filtration. Cake structure is first established by hydrodynamicfactors (cake porosity, mean particle size, size distribution, and particle specificsurface area and sphericity). It is also strongly influenced by some factors that canconditionally be denoted as physicochemical. These factors are:

1. the rate of coagulation or peptization of solid particles,2. the presence of tar and colloidal impurities clogging the pores,3. the influence of electrokinetic potentials at the interphase in the presence of

ions, which decreases the effective pore cross section, and4. the presence of solvate shells on the solid particles (this action is manifested

at particle contact during cake formation).

Due to the combining effects of hydrodynamic and physicochemical factors, the studyof cake structure and resistance is extremely complex, and any mathematicaldescription based on theoretical considerations is at best only descriptive.

The influence of physicochemical factors is closely related to surface phenomena atthe solid-liquid boundary. It is especially manifested by the presence of small particlesin the suspension. Large particle sizes result in an increase in the relative influence ofhydrodynamic factors, while smaller sizes contribute to a more dramatic influencefrom physicochemical factors. No reliable methods exist to predict when the influenceof physicochemical factors may be neglected. However, as a general rule, for roughevaluations their influence may be assumed to be most pronounced in the particle sizerange of 15-20 /mi.


Typical Industrial Filtration Conditions

Two significant operating parameters influence the process of filtration: the pressuredifferential across the filtering plate, and the temperature of the suspension. Mostcakes may be considered compressible and, in general, their rate of compressibilityincreases with decreasing particle size. The temperature of the suspension influencesthe liquid-phase viscosity, which subsequently affects the ability of the filtrate to flowthrough the pores of the cake and the filter medium.

In addition, the filtration process can be affected by particle inhomogeneity and theability of the particles to undergo deformation when subjected to pressure and settlingcharacteristics due to the influence of gravity. Particle size inhomogeneity influencesthe geometry of the cake structure not only at the moment of its formation, but alsoduring the filtration process. During filtration, small particles retained on the outerlayers of the cake are often entrained by the liquid flow and transported to layerscloser to the filter medium, or even into the pores themselves. This results in anincrease in the resistances across the filter medium and the cake that is formed,

Particles that undergo deformation when subjected to transient or high pressures areusually responsible for the phenomenon known as pore clogging. Fortunately, whatnature has sometimes neglected in the filterability of suspensions, man can correctthrough the addition of coagulating and peptizing agents. These are additives whichcan drastically alter the cake properties and, subsequently lower flow resistance andultimately increase the filtration rate and the efficiency of separation. Filter aids maybe used to prevent the penetration of fine particles into the pores of a filter plate whenprocessing low concentration suspensions. Filter aids build up a porous, permeable,rigid lattice structure that retains solid particles on the filter medium surface, whilepermitting liquid to pass through. They are often employed as precoats with theprimary aim of protecting the filter medium. They may also be mixed with asuspension of diatomaceous silica type earth (>90% silica content). Cellulose andasbestos fiber pulps were typically employed for many years as well.

The discussions of the basic features of filtration given thus far illustrate that the unitoperation involves some rather complicated hydrodynamics that depend strongly onthe physical properties of both fluid and particles, as well as interaction with acomplex porous medium. The process is essentially influenced by two different groupsof factors, which can be broadly lumped into macro- and micro-properties. Macro-factors are related to variables such as the area of a filter medium, pressuredifferences, cake thickness and the viscosity of the liquid phase. Such parameters arereadily measured. Micro-factors include the influences of the size and configurationof pores in the cake and filter medium, the thickness of the electrical double layer onthe surface of solid particles, and other properties.

Washing and Dewatering Operations

When objectionable (i.e., contaminated or polluted), or valuable suspension liquorsare present, it becomes necessary to wash the filter cake to effect clean separation of


solids from the mother liquor or to recover the mother liquor from the solids. Theoperation known as de-watering involves forcing a clean fluid through the cake torecover residual liquid retained in the pores, directly after filtering or washing. If thefluid is gas, then liquid is displaced from the pores. Also, by preheating the gas, thehydrodynarnic process is aided by diffusional drying.

Dewatering is a complex process on a microscale, because it involves thehydrodynamics of two-phase flow. Although washing and dewatering are performedon a cake with an initially well defined pore structure, the flows become greatlydistorted and complex due to changing cake characteristics. The cake structureundergoes compression and disintegration during both operations, thus resulting in adramatic alteration of the pore structure.

General Considerations for Process Engineers

In specifying and designing filtration equipment, primary attention is given to optionsthat will minimize high cake resistance. This resistance is responsible for losses infiltration capacity. One option for achieving a required filtration capacity is the use ofa large number of filter modules. Increasing the physical size of equipment is feasibleonly within certain limitations as dictated by design considerations, allowableoperating conditions, and economic constraints.

A more flexible option from an operational viewpoint is the implementation ofprocess-oriented enhancements that intensify particle separation. This can be achievedby two different methods. In the first method, the suspension to be separated ispretreated to obtain a cake with minimal resistance. This involves the addition of filteraids, flocculants or electrolytes to the suspension.

In the second method, the period during which suspensions are formed provides theopportunity to alter suspension properties or conditions that are more favorable tolow-resistance cakes. For example, employing pure initial substances or performinga prefiltration operation under milder conditions tends to minimize the formation oftar and colloids. Similar results may be achieved through temperature control, bylimiting the duration of certain operations immediately before filtering such ascrystallization, or by controlling the rates and sequence of adding reagents.

Filtration equipment selection is often complex and sometimes confusing because of(1) the tremendous variations in suspension properties; (2) the sensitivities ofsuspension and cake properties to different process conditions; and (3) the variety offiltering equipment available. Generalities in selection criteria are, therefore, few;however, there are some guidelines applicable to certain classes of filtrationapplications. One example is the choice of a filter whose flow orientation is in thesame direction as gravity when handling polydispersed suspensions. Such anarrangement is more favorable than an upflow design, since larger particles will tendto settle first on the filter medium, thus preventing pores from clogging within themedium structure.


A further recommendation, depending on the application, is not to increase thepressure difference for the purpose of increasing the filtration rate. The cake may, forexample, be highly compressible; thus, increased pressure would result in significantincreases in the specific cake resistance. We may generalize the selection process tothe extent of applying three rules to all filtration problems:

1. The objectives of a filtration operation should be defined;2. Physical and/or chemical pretreatment options should be evaluated for the

intended application based on their availability, cost, ease of implementa-tion and ability to provide optimum filterability; and

3. Final filtration equipment selection should be based on the ability to meetall objectives of the application within economic constraints.

The Objectives of Filtration

The objectives for performing filtration usually fall into one of the followingcategories:

1. clarification for liquor purification,2. separation for solids recovery,3. separation for both liquid and solids recovery, and/or4. separation aimed at facilitating or improving other plant operations.

Clarification involves the removal of relatively small amounts of suspended solidsfrom suspension (typically below 0.15% concentration). A first approach toconsidering any clarification option is to define the required degree of purification.That is, the maximum allowable percentage of solids in the filtrate must beestablished. Compared with other filter devices, clarifying filters are of lesserimportance to pure chemical process work. They are primarily employed in beveragemanufacturing and water polishing operations, pharmaceutical filtration, fuel/lubricating oil clarification, electroplating solution conditioning, and dry-cleaningsolvent recovery. They are also heavily employed in fiber spinning and film extrusion.

In filtration for solids recovery, the concentration of solids suspension must be highenough to allow the formation of a sufficiently thick cake for discharge in the form ofa solid mass before the rate of flow is materially reduced. However, solidsconcentration alone is not the only criterion for adequate cake formation. Forexample, an 0.5% suspension of paper pulp may be readily cake-forming whereas a10% concentration of certain chemicals may require thickening to produce adischargeable cake.

Filtration for both solids and liquid recovery differs from filtration for solids recoveryalone in the cake building, washing and drying stages. If the filtrate is a valuableliquor, maximum washing is necessary to prevent its loss; but if it is valueless, excesswash liquor can be applied without regard to quality.


Finally, filtration can be applied to facilitate other plant operations. Like other unitoperations, filtration has the most immediate relationship to those operationsimmediately preceding and following it. Ahead of filtration, the step is often one ofpreparation. These prefiltration steps could include thickening, coagulating, heating,conditioning, pH adjustment or the handling of an unstable flow that must not bebroken by rapid pumping or agitation before filtration. Such preparation stages areused to obtain more filterable material. This allows a continuous operating mode,smaller filter areas or both. Figure 2 schematically summarizes the prefiltration andfinal processing steps.

CHEMICALS

SOURCE —I

RECYCLE TO

PLANT OR

DISCHARGE

SETTLING

TANKF I L T E R

TO

PROCESS

' F I L T E R

B A C K W A S H

CLARIFIER-

SUROE TANK

DEWATERINQSOLIDS

DISPOSAL

CHEMICAL

RECOVERY

Figure 2. Summary of prefiltration and final processing steps in a filtering operation.

Filtration may also serve as the preparatory step for the operation following it. Thelatter stages may be dry ing or incineration of solids, concentration or direct use of thefiltrate. Filtration equipment must be selected on the basis of their ability to deliverthe best feed material to the next step. Dry, thin, porous, flaky cakes are best suitedfor drying where grinding operations are not employed. In such cases, the cake willnot ball up, and quick drying can be achieved. A clear, concentrated filtrate often aidsdownstream treatment, whereby the filter can be operated to increase the efficiencyof the downstream equipment without affecting its own efficiency.

Preparation Stages for Filtration

A number of preparation steps alluded to earlier assist in achieving optimumfilterability. The major ones are briefly described below.

Use of Precoat and Filter Aids

Where particles of a colloidal nature are encountered in liquor clarification, a precoatand or filter aid are often required to prevent deposited particles from being carried


by strearnflow impact into the pores of the filter medium (or filter cake afterformation), thus reducing capacity.

A precoat serves only as a protective covering over the filter medium to prevent theparticles from reaching the pores, while the filter aid added to the influent assists inparticle separation and cake formation. Filter aids serve as obstructions, interveningbetween the particles to prevent their compacting, and producing, under the pressurevelocity impact, a more or less impervious layer on the filter medium, or if a precoatis used, on it.

In some instances, precoats are used, not because of danger to filter cloth clogging,but to permit the use of a coarser filter medium such as metallic cloths. This canextend operating life or improve corrosion resistance.

Coagulation

This is another means of dealing with colloidal or semicolloidal particles. It appliesparticularly to clarification in water and sewage filtration and in the filtration of veryfine solids. While flocculation often can be accomplished by agitation, the use ofchemical additives results in alteration of the physical structure of the suspended solidsto the extent of losing their colloidal nature and becoming more or less crystalline.This is usually accompanied by agglomeration. Clarification by settling may follow,if the specific gravity of the particles is sufficient to provide reasonably quicksupernatant clarity. Direct filtration may be applied if the filter area is not excessiveor if complete supernatant clarity is needed.

Temperature Control

Temperature has a direct impact on viscosity, which in turn affects the flowrate. It isan important factor in filtration, since lower viscosity leads to liquor penetration intosmaller voids and in shorter times. Occasionally, temperature plays a role in alteringthe particle form or composition, and this in turn affects the clarification rate.

The Control of pH

Proper pH control can result in clarification that might otherwise not be feasible, sincean increase in alkalinity or acidity may change soft, slimy solids into firm,free-filtering ones. In some cases precoats are employed, not because of the dangerof filter cloth clogging, but to allow the use of a coarser filter medium, such asmetallic cloth.

Equipment Selection Methodology

Equipment selection is seldom based on rigorous equations or elaborate mathematicalmodels. Where equations are used, they function as a directional guide in evaluatingdata or process arrangements. Projected results are derived most reliably from actual


plant operational data and experience where duplication is desired; from standards setup where there are few variations from plant to plant, so that results can be anticipatedwith an acceptable degree of confidence (as in municipal water filtration); or frompilot or laboratory tests of the actual material to be handled. Pilot plant runs aretypically designed for short durations and to closely duplicate actual operations.

Proper selection of equipment may be based on experiments performed in themanufacturer's laboratory, although this is not always feasible. Sometimes thematerial to be handled cannot readily be shipped; its physical or chemical conditionschange during the time lag between shipping and testing, or special conditions mustbe maintained during filtration that cannot be readily duplicated, such as refrigeration,solvent washing and inert gas use. A filter manufacturer's laboratory has theadvantage of having numerous types of filters and apparatus available withexperienced filtration engineers to evaluate results during and after test runs.

The use of pilot-plant filter assemblies is both common and a classical approach todesign methodology development. These combine the filter with pumps, receivers,mixers, etc., in a single compact unit and may be rented at a nominal fee from filtermanufacturers, who supply operating instructions and sometimes an operator.Preliminary tests are often run at the filter manufacturer's laboratory. Rough testsindicate what filter type to try in the pilot plant.

Comparative calculations of specific capacities of different filters or their specificfilter areas should be made as part of the evaluation. Such calculations may beperformed on the basis of experimental data obtained without using basic filtrationequations. In designing a new filtration unit after equipment selection, calculationsshould be made to determine the specific capacity or specific filtration area. Basicfiltration equations may be used for this purpose, with preliminary experimentalconstants evaluated. These constants contain information on the specific cakeresistance and the resistance of the filter medium.

The basic equations of filtration cannot always be used without introducingcorresponding corrections. This arises from the fact that these equations describe thefiltration process partially for ideal conditions when the influence of distorting factorsis eliminated. Among these factors are the instability of the cake resistance duringoperation and the variable resistance of the filter medium, as well as the settlingcharacteristics of solids. In these relationships, it is necessary to use statisticallyaveraged values of both resistances and to introduce corrections to account for particlesettling and other factors. In selecting filtration methods and evaluating constants inthe process equations, the principles of similarity modeling are relied on heavily.

Within the subject of filtration, a distinction is made between micro- andmacromodeling. The first one is related to modeling cake formation. The cake isassumed to have a well defined structure, in which the hydrodynamic andphysicochemical processes take place. Macromodeling presents few difficulties,because the models are process-oriented (i.e., they are specific to the particularoperation or specific equipment). If distorting side effects are not important, thefiltration process may be designed according to existing empirical correlations. In


practice, filtration, washing and dewatering often deviate substantially from theory.This occurs because of the distorting influences of filter features and the unaccountedfor properties of the suspension and cake.

Existing statistical methods permit prediction of macroscopic results of the processeswithout complete description of the microscopic phenomena. They are helpful inestablishing the hydrodynamic relations of liquid flow through porous bodies, dieevaluation of filtration quality with pore clogging, description of particle distributionsand in obtaining geometrical parameters of random layers of solid particles.

Nomenclature

A = area (m)b = parameter in slip flow expression for K (sec2-m/kg)c = shape factor, known as Kozeny constantD = diameter (m)Dp = particle diameter (m)g = acceleration due to gravity (m/sec )h = hydraulic head (m)k = intrinsic permeability (k )K = hydraulic conductivity (m/sec)L = characteristic macroscopic length (m)n = number of pore layersp = pressure (kg/sec -m)q = seepage velocity (m/sec)Q = volumetric flowrate (m3/sec)Qm = volumetric flowrate at average pressure pm (m3/sec)r = radiusRe = Reynolds numberS = specific surface (m2)V = volume (m)vm = velocity of approach (m/sec)x = coordinate (m)z = coordinate (in direction of gravity) (m)

Greek Symbol

fj. - viscosity (kg/m-sec)p = density (kg/m3)T = tortuosity4> = porosity

FILTER MEDIA AND USE OF FILTERAIDS

Introduction

In conventional filter-medium filtration practices, the filter medium may be describedas the workhorse of the process. Proper selection is often the most importantconsideration for assuring efficient suspension separation. A good filter mediumshould have the following characteristics:

The ability to retain a wide size distribution of solid particles from the suspension,Offer minimum hydraulic resistance to the filtrate flow,Allow easy discharge of cake,High resistance to chemical attack,Resist swelling when in contact with filtrate and washing liquid,Display good heat-resistance within the temperature ranges of filtration,Have sufficient strength to withstand filtering pressure and mechanical wear,Capable of avoiding wedging of particles into its pores.

There are many filter media from which to choose from; however, the optimum typeoften depends on the properties of the suspension and specific process conditions.Filter media may be classified into several groups, however the two most commonclasses are the surface-type and depth-media-type.

Surface-type filter media are distinguished by the fact that the solid particles ofsuspension on separation are mostly retained on the medium's surface. That is,particles do not penetrate into the pores. Common examples of this type of media arefilter paper, filter cloths, and wire mesh.

Depth-type filter media are largely used for liquid clarification. They are characterizedby the fact that the solid particles penetrate into the pores where they are retained. Thepores of such media are considerably larger than the particles of suspension. Thesuspension's concentration is generally not high enough to promote particle bridging

19


inside the pores. Particles are retained on the walls of the pores by adsorption, settlingand sticking. As a rule, depth-type filter media cannot retain all suspended particles,and their retention capacity is typically between 90-99%. Sand and filter aids, forexample, fall into this category.

Some filter media may act as either surface-type or depth-type, depending on the poresize and suspension properties (e.g., particle size, solids concentration and suspensionviscosity).

It is also common practice to classify filter media by their materials of construction.Examples are cotton, wool, linen, glass fiber, porous carbon, metals and rayons. Sucha classification is convenient for selection purposes, especially when resistance toaggressive suspensions is a consideration. We may also classify media according tostructure, with typical classes being rigid, flexible and semi-rigid or combinationmedia.

Filtration aids are employed to enhance filtration characteristics, particularly for hard-to-filter suspensions. These are normally applied as an admix to the suspensions. Therole of the filter aid is to built up a porous, permeable and rigid lattice structure thatassists in retaining solid particles while allowing liquid to flow through.

This chapter provides a working knowledge of the use and selection of filter aids.Further discussions are given in subsequent chapters.

Flexible Filter Media

Flexible nonmetallic materials have been widely used as filter media for many years.They are available in the form of fabrics or as preformed unwoven materials, but alsoin the form of perforated plates.

Fabric filter media are characterized by the characteristics of mesh count, meshopening, yarn size and the type of weave. The mesh count or thread count of a fabricis the number of threads per inch. Thread counts in both warp and weft directions arethe same, and are indicated by a single number. Warp threads run lengthwise in afabric and are parallel to the selvage edge. Weft or filling threads run across the widthof a fabric at right angles to the warp. Figure 1 illustrates the important constructionparameters that characterize a fiber-based fabric. Note that the space between threadsis the mesh opening. It is measured in units of micrometers or inches. Different yarnsizes are normally specified as a measurement of diameter in micrometers or mils(thousandths of an inch). Yarn sizes in the warp and weft directions are normally thesame, and are indicated by a single number.

Fabrics are available in differing mesh openings, and varying thread diameters. Thethread diameter affects the amount of open area in a particular cloth, which in turndetermines the filtration flowrate or throughput.

Filter Media and Use of Filter Aids 21

WEFT

MESH COUNT

HTHREADD I A M E T E R

Figure 1. Construction parameters that determine the characteristics of a fiber-based fabric.

A plain weave is the most basic weave, with a weft thread alternately going over onewarp thread and then under one warp thread. A twill weave produces a diagonal ortwill line across the fabric face. These diagonals are caused by moving the yarnintersections one weft thread higher on successive warp yams. A twill weave isdesignated 2/1, 2/2, or 3/1 depending on how many weft threads the warp threads goover and under. A satin weave has a smooth surface caused by carrying the warp (orthe weft) on the fabric surface over many weft (or warp) yarns. Intersections betweenwarp and weft are kept to a minimum, just sufficient to hold the fabric firmly togetherand still provide a smooth fabric surface. The percentage of open area in a textile filterindicates the proportion of total fabric area that is open, and can be determined by thefollowing relationship:

% open area(mesh opening)2

(mesh opening + thread diameter}'x 100 (1)

The following are some examples of different types of common flexible filter media,

Glass Cloths

Glass cloths are manufactured from glass yarns. They have high thermal resistance,high corrosion resistance and high tensile strength, and are easily handled; thecomposition and diameter of the fibers can be altered as desired. The disadvantages


of glass cloth are the lack of flexibility of individual fibers, causing splits andfractures, and its low resistance to abrasion. However, backing glass cloth with a leadplate, rubber mats or other rigid materials provides for longevity. Backing with cottonor rubber provides about 50% greater life than in cases where no backing is used,

Cotton Cloths

Cotton filter cloths are among the most widely used filter media. They have a limitedtendency to swell in liquids and are used for the separation of neutral suspensions attemperatures up to 100°C, as well as suspensions containing acids up to 3% or alkalieswith concentrations up to 10% at 15-20°C. Hydrochloric acid at 90-100°C destroyscotton fabric in about 1 hour, even at concentrations as low as 1.5%. Nitric acid hasthe same effect at concentrations of 2.5%, and sulfuric acid at 5%. Phosphoric acid(70%) destroys the cloth in about six days. Water and water solutions of aluminumsulfate cause cotton fabrics to undergo shrinkage.

Woven cotton filter cloths comprise ducks, twills, chain weaves, canton flannel andunbleached muslins. Cotton duck is a fabric weave that is a plain cloth with equal-thickness threads and texture in the "over one and under one" of the warp and woof.The twill weave is over two and under two with the next filling splitting the warpstrands and giving a diagonal rib at 45° if the number of warp and filling threads areequal. Canton flannel is a twill weave in which one surface has been brushed up togive a nap finish. A muslin cloth is a very thin duck weave, which is unbleached forfiltering. In chain weave one filling goes over two warp threads and under two, thenext reversing this; the third is a true twill sequence, and the next repeats the cycle.

A duck may be preferable to a twill of higher porosity, because the hard surface ofthe duck permits freer cake discharge. Under high increasing pressure a strong,durable cloth (duck) is required, since the first resistance is small as compared withthat during cake building. Certain types of filters, such as drum filters, cannot standuneven shrinkage and, in some cases, cloths must be preshrunk to ensure fitting duringthe life of the cloth.

Nitro-filter (nitrated cotton cloth) cloths are about the same thickness and texture asordinary cotton filtration cloths, but are distinguished by a harder surface. It isclaimed that the cake is easily detached and that clogging is rare. Their tensile strengthis 70-80% of that of the specially manufactured cotton cloths from which they areprepared. They are resistant to the corrosive action of sulfuric, nitric, mixed nitrationand hydrochloric acids. They are recommended for filtering sulfuric acid solutions to40% and at temperatures as high as 90°C, with the advantage of removing finelydivided amorphous particles, which would quickly clog most ceramic media. Nitro-filter cloths are composed of cellulose nitrate, which is an ester of cellulose. Anychemical compound that will saponify the ester will destroy the cloth. Caustic soda orpotash in strengths of 2% at 70°C or over; alkali sulfides, polysuifides andsulfohydrates; or mixtures of ethyl alcohol and ether, ethyl, amyl and butyl acetates,pyridine, ferrous sulfates, and other reducing agents are detrimental to the cloth.


Cellulose nitrate is inflammable and explosive when dry, but when soaked in water itis considered entirely safe if reasonable care is taken in handling. For this reason itis colored red and packed in special containers. Users are cautioned to keep the clothswet and to handle them carefully.

Wool Cloths

Wool cloths can be used to handle acid solutions with concentrations up to 5-6%.Wool cloth has a life comparable to that of cotton in neutral liquors. Wool is wovenin the duck-like square cloth weave, or with a raised nap; or it may be formed as afelt. Originally the smooth cloth weave was used for filtering electrolytic slimes andsimilar slurries. The hairlike fibers, as in cotton cloth, ensure good filtrate clarity.Long-nap wool cloth has found wide application in sewage sludge dewatering and incases where only ferric chloride is used for conditioning. The wool has a long life andit does not clog easily. Wool cloths are sold by weight, usually ranging 10-22 oz/yd2

with the majority at 12 oz/yd2. The clarity through wool cloths is considerably lessthan through cotton cloths.

Paper Pulp and Fiber Cloths

Paper pulp and fiber cloths are excellent materials for precoats and filter aids. Paperpulp gives a high rate of flow, is easily discharged and shows little tendency to clog.

Paper pulp's disadvantage lies in its preparation. Soda or sulfate pulp, most commonlyused, must be disintegrated and kept in suspension by agitation before precoating. Thisrequires considerable auxiliary equipment. Diatomaceous earths, while they should bekept in suspension, are very easy to handle and do not undergo disintegration.

Paper pulp compressed into pads is used in pressure filters for beverage clarification.After becoming dirty, as evidenced by decrease in the rate of flow, the paper may berepulped, water-washed and reformed into pads. Although this involves considerablework, excellent clarity and high flowrates are obtained. The impurities do not forma cake as such, but penetrate into the pad and can only be removed by repulping andwashing the pad.

Pads of a mixture of paper pulp and asbestos fiber are used hi bacteriologicalfiltrations. In sheet form it is employed in the laboratory for all kinds of filtration.Filter papers are made in many grades of porosity for use in porcelain and glassfunnels. Industrially, paper in the form of sheets is used directly or as a precoat infilter presses.

Used directly in lubricating clarification in a "blotter press", it acts much the samemanner as the paper pads, but is much thinner and is not reused. As a precoat, paperprotects the filter medium from slimy fines; it may be peeled off and discarded afterclogging, leaving the medium underneath clean.


Rubber Media

Rubber media appear as porous, flexible rubber sheets and microporous hard rubbersheets. Commercial rubber media have 1100-6400 holes/in.2 with pore diameters of0,012-0.004 in. They are manufactured out of soft rubber, hard rubber, flexible hardrubber and soft neoprene.

The medium is prepared on a master form, consisting of a heavy fabric belt, surfacedon one side with a layer of rubber filled with small round pits uniformly spaced. Thesepits are 0.020 in. deep, and the number per unit area and their surface diameterdetermine the porosity of the sheet. A thin layer of latex is fed to the moving belt bya spreader bar so that the latex completely covers the pits, yet does not run into them.This process traps air in each pit. The application of heat to the under-surface of theblanket by a steam plate causes the air to expand, blowing little bubbles in the film oflatex. When the bubbles burst, small holes are left, corresponding to the pits. Theblown rubber film, after drying, is cooled and the process repeated until the desiredthickness of sheet is obtained. The sheet is then stripped off of the master blanket andvulcanized,

Approximately 95% of the pits are reproduced as holes in the rubber sheet. The holesare not exactly cylindrical in shape but are reinforced by slight constrictions whichcontribute to strength and tear resistance. This type is referred to as "plain," and canbe made with fabric backing on one or both sides to control stretching characteristics.If the unvulcanized material is first stretched, and then vulcanized while stretched, itis called "expanded." Resulting holes are oval and have a higher porosity (sometimesup to 30%). Special compounds have been formulated for resistance to specificchemicals under high concentrations at elevated temperatures, such as 25% sulfuricacid at 180° F.

The smooth surface allows the removal of thinner cakes than is possible with cottonor wool fabrics. Rubber does not show progressive binding and it can be readilycleaned and used in temperatures up to 180°F. On the other hand, because a clearfiltrate is difficult to obtain when filtering finely divided solids, a precoat oftenbecomes necessary.

Synthetic Fiber Cloths

Cloths from synthetic fibers are superior to many of the natural cloths thus farconsidered. They do not swell as do natural fibers, are inert in many acid, alkaline andsolvent solutions and are resistant to various fungus and bacterial growths (the degreedepending on the particular fiber and use). Several synthetic fibers resist relativelyhigh temperatures, and have a smooth surface for easy cleaning and good solidsdischarge. Some of the most widely used synthetic filter media are nylon, Saran,Dacron, Dynel, Vinyon, Orion, and Acrilan. Table 1 compares the physical propertiesof several synthetic fiber filter media.


Tightly woven, monofilament (single-strand) yarns consist of small-diameterfilaments. They tend to lose their tensile strength, because their small diametersreduce their permeability; thus multifilament yarns are normally used. Monofilamentyarns in loose weaves provide high flowrates, good solids discharge, easy washing andhigh resistance to blinding, but the turbidity of the filtrate is high and recirculation isusually necessary, initially at least. Table 2 provides additional information on varioussynthetic filter fabrics.

Flexible Metallic Media

Flexible metallic media are especially suitable for handling corrosive liquors and forhigh-temperature filtration. They have good durability and are inert to physicalchanges. Metallic media are fabricated in the form of screens, wire windings, orwoven fabrics of steel, copper, bronze, nickel and different alloys.

Perforated sheets and screens are used for coarse separation, as supports for filtercloths or as filter aids. Metallic cloths are characterized by the method of wire weavesas well as by the size and form of holes and by the wire thickness. Metallic cloths maybe manufactured with more than 50,000 holes/cm2 and with hole sizes less than 20/am.

Table 1. Properties of woven filter cloth fibers.

FibersAcrilanAsbestosCottonDacronDyne!GlassNyionOrionSaranTeflonWool

AcidsGoodPooi-PoorFairGoodHighFail-GoodGoodHighFair

AlkaliesGoodPooi-FairFairGoodFairGoodFairGoodHighPoor

SolventsGoodPooi-GoodFail-GoodFail-GoodGoodGoodHighFail-

FiberTensile

StrengthHighLowHighHighFail-HighHighHighHighFail-Low

TemperatureLimit

275750300350200600300275240180300

MetaHic/Nonmetallic Cloth

Combination metallic and nonmetallic cloths consist of metallic wires and weak clothor asbestos threads. There are some difficulties in weaving when attempting tomaintain uniformity between wires and the cloth, and considerable dissatisfaction hasbeen experienced with such construction. While cotton weaves well with the asbestos,the cotton fibers destroy the fabric's resistance to heat and corrosion. Its use is,therefore, quite limited, despite its resistance to high temperatures, acids and mildew.


Cotton cloths are sometimes treated with metallic salts (copper sulfate) to improvetheir corrosion-resistant qualities. Such cloths are in the usual cotton filter clothgrades, and while they are not equivalent to metallic cloths, the treatment doesmaterially prolong the life of the cotton fiber.

Non woven Media

Nonwoven media are fabricated in the form of belts or sheets from cotton, wool,synthetic and asbestos fibers or their mixtures, as well as from paper mass. They maybe used in filters of different designs, for example, in filter presses, filters withhorizontal discs and rotary drum vacuum filters for liquid clarification. Most of theseapplications handle low suspension concentrations; examples are milk, beverages,lacquers and lubricating oils. Individual fibers in nonwoven media are usuallyconnected among them as a result of mechanical treatment. A less common approachis the addition of binding substances. Sometimes the media are protected from bothsides by loosely woven cloth. Nonwoven media of various materials and weights, andin several grades of retentiveness per unit weight can be formed, in either absorbentor nonabsorbent material. These filter media retain less dispersed particles (more than100 jum) on their surface, or close to it, and more dispersed particles within the depthsof the media.

Nonwoven filter media are mostly used for filter medium filtration with pore clogging.Because of the relatively low cost of this medium, it is often replaced after poreclogging. In some cases, non woven media are used for cake filtration. In this case,cake removal is so difficult that it must be removed altogether from the filter medium.Nonwoven filter media can be prepared so that pore sizes decrease in the directionfrom the surface of the filter media contacting suspension to the surface contacting thesupporting device. This decreases the hydraulic resistance of filtration and providesretention of relatively large particles of suspension over the outer layer of thenonwoven medium. Nonwoven filter media of synthetic, mechanically pressed fibersare manufactured by puncturing the fiber layer with needles (about 160punctures/cm2), and subsequent high temperature treatment with liquid which causesfiber contraction. Such filter media are distinguished by sufficient mechanical strengthand low hydraulic resistance, as well as uniform fiber distribution. Filter media fromfibers connected by a blinder are manufactured by pressing at 70N/cm2and 150°C.These media have sufficient mechanical strength, low porosity and are corrosionresistant.

Filter media may be manufactured by lining a very thin layer of heat-resistant metal(e.g., nickel 360) over a fiber surface of inorganic or organic material. Such filtermedia may withstand temperatures of 200°C and higher.

Of the flexible filter media described, the synthetic fabrics are perhaps the mostwidely relied on in industrial applications. Each filtration process must meet certainrequirements in relation to flowrate, clarity of filtrate, moisture of filter cake, cakerelease and nonbinding characteristics. The ability of a filter fabric to help meet thesecriteria, and to resist chemical and physical attack depend on such characteristics as

Thread Dim. , Mesh Weight Threadslia., Warp x Weft opening Air Permeability Thickness

Style No. Weave (oz/yd*) Warp x Weft (Pm) ( I 4 (fP/min) bd Nylon 6,6.6

Warp and Weft Monofilament 11 1-020 Plain 4.5 22 x 22 305 x 305 850 X 850 NAb 570 11 1-1 10 Piain 4.6 50 X 50 200 x 200 300 x 300 NA 350 11 1-150 Plain 3.2 62 X 62 150 X 150 250 X 250 NA 270 1 1 1- 160 Plain 2.4 107 X 76 100 x 100 250 X 230 NA 220 1 1 1- 170 Plain 4.6 29 X 29 250 X 250 600 X 600 NA 450 111-180 Plain 2.3 66 X 66 130 X 130 210 x 210 NA 270 111-190 Twill 5.5 38 X 38 250 X 250 420 X 420 NA 5 30 11 1-206 Ptain 5.3 183 X 43 150 X 150 170-210 310 11 1-220 Plain 2.9 80 X 80 125 X 125 175 X 175 NA 240 1 11-230 Plain 5.7 40 X 40 250 X 250 420 X 420 NA 450 1 1 1-056 Satin 7.2 109 x 42 205 X 300 350-400 450

I053 Plain 2.1 147 X 97 150-200 160 1093 Twill 3.5 297 X 122 75- 100 250 1103 Twill 3.7 297 X 135 40-70 250 1123 Twill 3.2 195 X 140 15-25 190 1153 Satin 3.2 236 X 99 15-25 210 1193 Satin 5.3 300 x 99 45-70 300 1203 Satin 6.8 152 X 76 20-30 390

Warp and Weft Monofilament

1212 Satin 6.5 178 x 97 50-80 710 1283 Plain 1.8 112 x 97 170-220 130

N 00

Style No. Air Permeability whlin)

1338 Len0 5.6 7 x 5 NA 630 3 1353 Plain 1363 Plain 1393 Twill 122-053 Twill 122-073 Oxford

1233 Satin Warp Monofilament, Weit Spun

Warp and Weft Monofilament and Metal Spun

9165 Twill Nylon 1 1

Warp and Weft Monofilarnent 1656 Satin 1666 Satin 1686 Satin 11 1-096 Satin

Nomex Warp Multifilament. Weft Spun

1513 Plain Polyester

Warp and Weft Moxicifilament 1713 Plain 1716 Plain

3.2 11.5 9.7 4.7

15.5

5 .O

4.1

8.2 9.0 7. I 9.5

7.7

4. I 4.1

64 X 48 69 x 28

236 x 53 80 X 117 72 x 21

320 x 71

297 x 132

99 x 53 111 x 53 107 x <91 99 x 53

107 X 66

335 x 84 350 x 79

180 x 290 180 x 290 180 x 180 205 x 300

50-100 1-3 5- 10 50-80 0.5-2

60- 100

25-40

200-300 125-200 125-200 150-300

50-80

80- 120 10-20

255 660 560 410 720

410

300

520 530 395 490

5 10

190 1x0

Table 2 Continued.

Style No.

17339656 B9813 B9884 B31 1-010 HSC

311-020Warp and Weft Multifilament

18131853189318%1933194319531956197319769881322-013322-033322-036322-043

322-070

Weave

PlainSatinSatinSatinPlainPlain

PlainLenoTwillTwillSatinTwillPlainPlainPlainPlainPlainTwillPlainPlainBrokenTwillPlain

„. . . . ,„. , .. ... Thread Diam., MeshWeight Threads/in., Warp x '/ / M\ «r L Warp x Weft Opening(oz/yd2) Weft * *~—-&

8.88.3

11.812.76.96.9

10.310.04.14.13.26.5

10.010.03.53.52.73.67.17.15.3

11.5

175 x 36104 X 53 200 X 300112 x 53 200 x 300145 x 53 200 X 25038 X 38 250 X 250 420 x 42038 X 38 250 X 250 420 x 420

43 x 307 X 5

295 x 112295 X 112290 X 99257 x 79

36 X 3036 X 3094 X 8191 x 84

152 X 81219 x 14781 X 5681 X 56

216 x 100

52 x 32

Air Permeability(tf/nun)

300-400300-400200-25080-150

NANA

1-3NA

40-755-10

35-5510-202-41-2

20-405-1540-605-103-71-3

30-50

1-2.5

Thickness

380450510440400400

500800250200190330420380195140270180300280250

600

Table 2 Continued.

Style No, Weave

322-073 Plain322-123 Twill

Warp Multifilament, Weft Spun323-013 Twill

Warp and Weft Spun1823 Twill1873 Plain1923 Plain333-020 Plain

AcrylicWarp and Weft Multifilament

2013 Plain2023 Twill

Warp Multifilament, Weft Spun2033 Twill

Warp and Weft Spun2063 Twill9856 Twill

Poly vi ny Ichloride2973 Twill

PolypropyleneWarp und Weft Monofllament

2773 Satin2776 Satin

„, . , , „,, , ,, ,,, Thread Diain., Mesh ,. „ ,,_Weight Threads/in., Warp x „ . Air Permeability/ ijo\ «7 «* Warp x Weft Opening f£,3. . ,(oz/yd ) Weft , , ; , (ft /mm)

(taa) u^m)

13.05.3

13.2

10.910.65.69.6

7.14.4

6.5

9.47.1

6.2

9.79.7

54 X 36216 X 100

76 X 53

51 x 2553 X 3064 x 5347 X 29

81 x 33102 X 69

102 x 61

69 X 4151 X 64

66 X 51

86 x 38 300 x 25086 x 38 320 X 290

0.5-1.540-60

4-6

30-407-15

70-10020-30

3-715-20

25-40

10-2040-60

60-90

70-10045-90

Thickness(fan)

580270

580

770610360570

480250

480

700580

325

620680

Air Permeability Thickness Thread Diam., Mesh

(Icm) (run) Weight Threads/in.,Wxp x Warp weft (oz/yd2) (ftJ/min) (/lm)

ojx?nng Weft Style No. Weave

2793 Twill 5 11-013 Satin 51 1-015 Satin 51 1-016 Satin 511-050 Plain 511-066 Basket 511-070 HS Twill 511-086 Plain 511-090 Plain 511-110 HS Twill 511-126 Twill 511-136 Satin 511-150 Plain

Warp Monofilament, Weft Multifilament

512-013 Satin 512-015 Satin 512-016 Satin

513-016 Satin

2723 Plain 2726 Plain 2843 Twill

Warp Monofilament, Weft Spun

Warp and Weft Multifilament

10.9 8.0 8.2 8.2 3.5 5.9 7.2 5.3 3.7 5.4 8.8 6.5 3.5

9.1 9.9 9.7

11.4

6.5 6.5

10.0

61 X 25 8 6 x 4 6 89 x 46 89 x 46 50 X 50

110 x 45 28 X 28

108 x 32 30 X 30 38 x 33 76 X 28

110 X 60 20 x 20

86 X 30 86 X 30 86 x 30

93 X 36

33 x 20 33 x 20

241 X 43

300 x 500 250 X 250 250 X 250 250 X 250 200 x 200 200 x 200 375 x 375 200 x 200 250 X 250 305 X 305 305 X 305 200 x 200 305 X 305

300-400 150-300 40-60

100-150 300 x 300 NA

50-100 530 x 530 NA

100-150 600 X 600 NA

NA 100-200 20-300

950 x 950 NA

100-150 10-30 40-80

5-10

5- 10 3-8

10-15

700 550 500 580 370 450 650 320 490 600 590 450 510

740 5 10 600

680

470 425 640

Table 2 Continued.

Style No.

285328369870522-013522-053522-060522-063522-070

Weave

TwillTwillTwillPlainTwillPlainPlainTwill

„, . . . T,, . ,. „, Thread Diara.. MeshWeight Threads/in., Warp x „, „. ' _t ,M\ «? L Warp x Weft Opening(oz/yd*) Weft v. . F^«~6

(Mm) (fan)

16.816.819.810.018.814.315.18.3

58 X 3058 X 3069 X 3043 x 2470 X 3076 X 2177 x 2158 x 39

Air Permeability(fp/min)

4-82-41-23-64-91-2.5

0.5-220-40

Thickness(jan)

120011001250690

1160960950630

j5*CL,

1teaB

Warp Multifiiament, Weft Spun2873523-040523-043523-053

TwillTwillTwillTwill

20.610.311.018.6

175 33 X70 X 3871 x 4072 X 30

5-104-73-52-5

1500840800

1200Warp and Weft Spun

Polyethylene

2783283328939811533-013533-033533-070

TwillTwillTwillTwillPlainTwillTwill

7.710.613.919.211.715.814.0

53 x 3353 X 3061 x 2353 X 2350 X 2358 X 2648 X 31

50-808-136-8

10-151-23-64-6

800100012001550850

12001050

Warp and Weft Monofilament2473 Satin 5.3 155 x 58 150 x 150 25 x 300 200-300 425

Table 2 Continued.

Style No.

248324862503411-010411-013411-015

Warp Monofilament, WeftMultifilament

2573Saran

Warp and Weft Monofllament233323362363611-010611-020611-040

Weave

SatinSatinLenoPlainPlainPlain

Satin

SatinSatinPlainTwillPlainPlain

Weight Threads/in., Warp x(oz/yd2) Weft

8.38.35.37.29.0

10.1

6.3

11.211.27.76.67.56.5

89896

105111112

150

666615503020

X

X

X

X

X

X

X

X

XX

X

X

X

48485424450

36

434315503020

Thread Diam.,Warp x Weft

(MH»)250 x280 x500 x220 X220 x220 x

250 X280 X320 x200 X250 x305 x

250270500220220220

250370380200250305

MeshOpening

(Mm)

40 x15 X

4200 X

65 x25 x

1050 x300 x600 x950 x

3001903000

3401801050300600950

Air Permeability(ft'/min)

250-350100-150

NA200-300100-17550-80

130-200

600-700200-300

NANANANA

Thickness(MH»)

665500

1500580650650

540

700370900500490550

Courtesy of Industrial Fabrics Corp., Minneapolis, MN.NA - not available.HS - high shrink.


fiber type, yarn size, thread count, type of weave, fabric finish and yarn type(monofilament, multifilament or spun).

Monofilament yarns consist of a single, continuous filament with a relatively smoothsurface. The different sizes are specified as a measurement of the diameter in mils orin micrometers. Multifilament yarns are made from many fine filaments extrudedsimultaneously. The different sizes are specified by a measurement of weight knownas the denier. These yarns are generally used for filter fabrics which require a smoothsurface and relatively tight weave. Spun yarns are made from filaments which arechopped in short lengths and then spun or twisted together. Spun yarns are made intofilter fabrics with a hairy, dense surface very suitable for filtration of very fineparticles.

It is necessary to select the type of fiber that will offer the most resistance tobreakdown normally caused by chemical, temperature and mechanical conditions ofthe filter process. Tables 3 through 5 can serve as rough guides to proper mediaselection. Table 6 provides linear conversion units between mesh size, inches andmicrometers.

Rigid Filter Media

Fixed Rigid Media

Fixed rigid media are available in the forms of disks, pads and cartridges. They arecomposed of firm, rigid particles set in permanent contact with one another. Themedia formed have excellent void uniformity, resistance to wear and ease in handlingas piece units. Depending on the particle size forming the filter media, temperature,pressure and time for caking, it is possible to manufacture media with differentporosities. The higher the pore uniformity, the more uniform the shape of theparticles. These media are distinguished by long life, high corrosion resistance andeasy cake removal. However, the particles that penetrate inside the pores are verydifficult to extract.

Metallic Media

Metallic filter media are widely used throughout the chemical and process industriesin the form of perforated or slotted plates of steel, bronze or other materials. Thesedesigns provide for easy removal of coarse particles and for supporting loose rigidmedia.

Powdered metal is a porous medium. The physical characteristics, chemical composi-tion, structure, porosity, strength, ductility, shape and size can be varied to meetspecial requirements. The porosity ranges up to 50% void by volume, tensile strengthup to 10,000 psi, varying inversely with porosity, and ductility of 3-5 % in tension, andhigher in compression.

Table 3, Physical properties and chemical resistances of fibers.

Type of Fiber

CottonSilk

Wool

Glass

Steel Fibers (Brunsmet * )Polyamides

PA 6 (Perlon ® )

PA 6.6 (Nylon)

PA 11 (Rislan*)

PA 12 (Vestamid « )

PA Nomex ®

Polyester

Polyacrylonitrile

SpecificGravity

1.51.37

1.3

2.54

7.9

1.14

1.14

1.04

1.02

1.38

1.38

1.15

Moisture ,, .., .. , MoistureAbsorption at „ÂO*^/JCOOIT\ Expansion•4U Moo r;, ,„,

65%RH l '

7-8.5 459-9.5 40

13-15 42

0 0

0 0

4.5 10-14

4.5 10-14

1.2

0.95

5

0.4 3-5

1-1.5 4.5-6

IroningTemperature

200140-160

140-160

300

300

120

150

120

230

150

160

MaximumWorking

Temperature

80-10090

80

250-300

500

100

100

100

90Stab. 120200-300

140-160

145

Melting Point

(°C)b

D>200B= 200-400D=170B= 200-400D=130S=815M=845-1150M= 1440- 1455

8 = 170M=2158=235M = 250-2558 = 160M= 183-1868 = 168M=175M=375D = 3718=230-249M=2568=235-250D-300

Resistance c

to Lightand

Weather

8C

C

R

R

C

C

C

C

C

G

R

Table 3 Continued,

Type of Fiber

Polyvinylchloride

Polyvinylidenechloride(Saran *)

PolyolefinsPolyethylene

High-Pressure

Low-Pressure

Polypropylene

Polytetrafluoroethylene (Teflon*)

SpecificGravity

1.38

1.7

0.92

0.95

0.9

2.1

MoistureAbsorption at200C(68°F),

65%RH

0.1 0

<0.1 0

<0.1 0

<0.1 0

<0.1 0

0 0

Moisture IroningExpansion Temperature

(%) <°C)

60-80

80

70

80

100

250

MaHnsujnWorking

Temperature(°C)

60

70

60

70

90

200

Melting Point

(°C)b

S=70D=180-1908 = 115-138M= 150- 170

8 = 107-110M= 110- 120S = 115M=124-1388 = 150-155M=163-1758=327D>275

Resistance c

to Lightand

Weather

R

G

U

U

C

R

Liquid F

iltration

Resistance Against

Type of Fiber InsectProof

Cotton Medium

Silk MediumWool BadGlass GoodSteel Fibers (Brunsmet * ) Good

Resistanceto Aging

Low

LowLowGoodGood

Acid AlkaliChlorocarbonicHydride

Unstable Low resistance, Resistantswelling

Low resistance UnstableLow resistance UnstableLow resistance UnstableLow resistance Resistant

ResistantResistantResistantResistant

Ketone

Resistant


Phenol

Resistant


Benzene

Resistant


Table 3 Continued. Resistance Against

Type of Fiber Insect Resistance CMorocarbonic Proof toAgiog Acid Alkali Hydride Ketone Phenol Benzene

PA 6 (Perlon @ ) Good Good Unstable Resistant PA 6.6 (Nylon) Good Good Unstable Resistant PA 11 (Rislan ) Good Good Low resistance Resistant PA 12(Vestamid") Good Good Low resistance Resistant PA Nomex @ Good Good Low resistance Resistant

Polyester Good Good Resistant Low resistance Polyacrylonitrile Good Good Resistant Low resistance Polyvinylchloride Good Good Resistant Resistant Polyvinylidenechloride Good Good Resistant Resistant except

Polyolefns (Saran? NH,,OH

Polyethylene High-pressure Good Good Resistant Resistant Low-Pressure Good Good Resistant Resistant

Polypropylene Good Good Resistant Resistant Polytetrafluoroethylene Good Good Resistant Resistant

(Teflon 4,

B Data compiled from different industry lirerature source& . b B = burning point; D = disintegration, M = melting point. S = softening point. c R = recommended; G = good; S = satisfactory. C = conditional: U = unsatisfacrory.

Resistant Resistant Resistant Swelling Resistant Resistant Resistant Resistant Resistant

Swelling Swelling Resistant Resistant

Resistant Resistant Resistant Swelling Resistant Resistant Resistant Unstable Resistant

Resisutiit Resistant Resistant Resistant

Unstable Unstable Unstable Unstable Unstable Unstable Resistant Unstable Unstable

Resistant Resistant Resistant Swelling Resistant Resistant Resistant Unstable Resistant

Resistant Resistant Resistant Resistant

Resistant Resistant Resistant


Table 4. Physical properties and chemical resistances of polyester fibers used for belt filters.

Specific Gravity 1.38Moisture Regain

At 65%RH and 68°F(20°C)(%) 0.4Water Retention Power (%) 3-5

Tensile StrengthcN/dtex 7-9.5Wet in % of dry 95-100

Elongation at Break% 10-20Wet in % of dry 100-105

Ultraviolet Light Resistance RResistance to Fungus, Rot, and Mildew RResistance to Dry Heat

Continuous°F 302°C 150

Short-Term Exposureop 392

°C 200Chemical Resistance to

Acids Cc

Acetic Acid Concentration RSulfuric Acid 20% RNitric Acid 10% CHydrochloric Acid 25% CAlkalies CSaturated Sodium Carbonate RChlorine Bleach Concentration RCaustic Soda 25% Ud

Ammonia Concentration UPotassium Permanganate 50% RFormaldehyde Concentration RChlorinated Hydrocarbons RBenzene RPhenol CKetones, Acetone R

a Average properties reported as based on typical industry sources,b R = recommendedc C = conditionald U = unsatisfactory

Table 5. Standard fibers and micrometer ratuigsfor bag filters.Available Micrometer Ratings

Construction

Felts PolyesterPolypropylene

Multifilament Meshes PolyesterNylon (heavy)

Monofilament Meshes NylonPolypropylene

1 3 5 10 15

X X X X XX X X X

25 50 75

X X XX X

X

X X

100 125 150 175

XXX X

X X

X

X X

200 250 300

X

X X X

X X XX

400 600 800

X XX

X X XX

Compatibility and Temperature Limits for Standard Bag Materials

Compatibility with

Fiber

PolyesterPolypropyleneNylon

Organic Solvents

ExcellentExcellentExcellent

Animal, Vegetable,and Petro Oils


Microorganisms


Alkalies

GoodExcellentGood

OrganicAcidsGoodExcellentFair

Oxidizing MineralAgents AcidsGood GoodGood GoodPoor Poor

TemperatureLimits (°F)

300225325

Bag Capacities and Dimensions

Bag Dimensions

Bag Size No.121 (inner)2 (inner)

Fits Rosedale Model No. Bag Surface Area815830815830

2.04.41.63.6

(ft2) Bag Volume (gal)

2.14.61.73.8

Length (in.)16.532.014.530.0

Diameter (in.)

775.755.75


Table 6. Comparative particle sizes.

U.S. Mesh33»/2

45678101214161820253035404550607080100120140170200230270325400

in.

0.2650.2230.1870.1570.1320.1110.09370.07870.06610.05550.04690.03940.03310.02800.02320.01970.01650.01380.01170.00980.00830.00700.00590.00490.00410.00350.00290.00240.00210.00170.0015

IMK

673056604760400033602830238020001680141011901000841707595500420354297250210177149125105887463534437

Powdered metal cannot be readily ground or machined. It is made in discs, sheets,cones or special shapes for filtering fuel oil, refrigerants, solvents, etc.

The smooth surface associated with a perforated plate permits brush cleaning orscrubbing in addition to the naturally easier discharge from such surfaces. The hardmetallic material has a long life, not being subject to abrasion or flexing. The size ofparticles filtered on such plates must be relatively large. Normally, plates are confinedto free-filtering materials where there is little danger of clogging.

Metallic filter media may be used either for cake filtration or depth filtration, i.e.,pore clogging. Regeneration of media may be achieved by dissolving solid particlesinside the pores or by back thrust of filtrate.


Ceramic Media

Ceramic filter media are manufactured from crushed and screened quartz or chamotte,which is then thoroughly mixed with a binder (for example, with silicate glass) andsintered. Quartz media are resistant to concentrated mineral acids but not resistant tolow-concentration alkalies or neutral water solutions of salts. Chamotte media areresistant to dilute and concentrated mineral acids and water solutions of their salts, buthave poor resistance to alkali liquids.

The rough surface of ceramic filter media promotes adsorption of particles andbridging. Sintering of chamotte with a binder results in large blocks from which filtermedia of any desired shape can be obtained. Using synthetic polymers as binders,ceramic filter media that do not contain plugged pores are obtained.

Diatomaceous Media

Diatomaceous media are available in various shapes. Their relatively uniform particlesize establishes high efficiency in retaining solid particles of sizes less than I/mi aswell as certain types of bacteria. Media in the form of plates and cartridges aremanufactured by sintering a mixture of diatomite with a binder.

Coal Media

Coal media are manufactured by mixing a fraction of crushed coke with an anthracenefraction of coal tar and subsequent forming under pressure, drying and heating in thepresence of a reducing flame. These media of high mechanical strength are good foruse in acids and alkalies.

Ebonite Media

Ebonite media are manufactured from partially vulcanized rubber, which is crushed,pressed and vulcanized. These media are resistant to acids, salt solutions and alkalies.They may be used for filtration at temperatures ranging from -10 to 4-110°C.

Foam Plastic Media

Foam plastic media are manufactured from poly vinyl chloride, polyurethane,polyethylene, polypropylene and the other polymer materials. The foam plastic mediaare economical.

Loose Rigid Media

Filter media may also be composed of particles that are rigid in structure, but areapplied in bulk loose form. That is, individual particles merely contact each other.This form has the advantage of being cheap and easy to keep clean by rearrangementof the particles. When the proper size and shape of particles are selected, the sectionof passage may be regulated over extremely wide limits. The disadvantages of a rigid


medium in simple contact are that it can be used conveniently only in a horizontalposition and that it does not allow removal of thick deposits or surface cleaning exceptby backwashing, without disturbing the filter bed.

Coal and Coke

Coal (hard) and coke are used in water filtration, primarily for the removal of coarsesuspensions, care being taken to prevent them from scouring or washing away,because of their relative lightness and fine division. Coal is principally composed ofcarbon, and is inert to acids and alkalies. Its irregular shapes are advantageous attimes over silica sand. Though inert to acids, sand is affected by alkalies, and itsspherical particle shape allows deeper solids penetration and quicker clogging thandoes coal. With the lighter weight of coal (normally 50 lb/ft3, compared with 100 lb/ft3

for sand), a greater surface area is exposed for solids entrapment.

Charcoal

Charcoal, whether animal or vegetable, when used as a filter medium, is required toperform the dual services of decoloring or adsorbing and filtering. The char filtersused in the sugar industry are largely decoloring agents and the activated carbons usedin water clarification are for deodorizing and removal of taste. There are many typesof charcoal in use as filter media, ranging from ordinary wood char to speciallyprepared carbons.

Diatomaceous Earth

Diatomaceous earths may resemble the forms of the charcoals. The earths areprimarily filter aids, precoats or adsorbents, the function of the filter medium beingsecondary. Fuller's earth and clays are used for decoloring applications; diatomaceousearths are used for clarification.

Precipitates and Salts

Precipitates or salts are used when corrosive liquor must be filtered, and where thereis no available medium of sufficient fineness that is corrosion-resistant and will notcontaminate the cake. In these cases, precipitates or salts are used on porous supports.In the filtration of caustic liquors, ordinary salt (sodium chloride) is used as the filtermedium in the form of a precoat over metallic cloth. This procedure has the advantagethat the salt medium will not be detrimental to either the cake or the filtrate ifinadvertently mixed with it.

Sand and Gravel

Sand and gravel are the most widely used of the rigid media simple contact. Most ofthe sand used this way is for the clarification of water for drinking or industrial uses.Washed, screened silica sand is sold in standard grades for this work and is used indepths ranging from a few inches to several feet, depending on the type of filter and


clarification requirements. Heavy, irregular grains, such as magnetite, give high ratesof flow and low penetration by the solid particles, and are easily cleaned. They are,however, considerably more expensive than silica sand, so their use is limited. Sandbeds are often gravel-supported, but gravel alone is seldom used as the filter medium,

Crushed Stone or Brick

Crashed stone or brick is used for coarse filtration of particularly corrosive liquors.Their use, however, is extremely limited and they are not considered important filtermedia.

Filter Media Selection Criteria

Due to the wide variety of filter media, filter designs, suspension properties, condi-tions for separation and cost, selection of the optimum filter medium is complex. Filtermedia selection should be guided by the following rule: a filter medium mustincorporate a maximum size of pores while at the same time providing a sufficientlypure filtrate. Fulfilment of this rule invokes difficulties because the increase ordecrease in pore size acts in opposite ways on the filtration rate and solids retentioncapacity.

The difficulty becomes accentuated by several other requirements that cannot beachieved through the selection of a single filter medium. Therefore, selection is oftenreduced to determining the most reasonable compromise between different, mutuallycontradictory requirements as applied to the filter medium at a specified set offiltration conditions. Because of this, some problems should be solved before finalmedium selection. For example, should attempts be made to increase filtration rate orfiltrate purity? Is cost or medium life more important? In some cases a relatively moreexpensive filter medium, such as a synthetic cloth, is only suitable under certainfiltration conditions, which practically eliminates any cost consideration in theselection process.

Thus, the choice may only be made after consideration of all requirements. It is,however, not practical to analyze and compare each requirement with the hope oflogically deducing the best choice. There is, unfortunately, no generalized formula forselection that is independent of the details of the intended application. Each cakerequires study of the specific considerations, which are determined by the details orthe separation process.

One can to outline a general approach for medium selection along with a test sequenceapplicable to a large group of filter media of the same type. There are three methodsof filter media tests: laboratory- or bench-scale pilot-unit, and plant tests. Thelaboratory-scale test is especially rapid and economical, but the results obtained areoften not entirely reliable and should only be considered preliminary. Pilot-unit testsprovide results that approach plant data. The most reliable results are often obtainedfrom plant trials.


Different filter media, regardless of the specific application, are distinguished by anumber of properties. The principal properties of interest are the permeability of themedium relative to a pure liquid, its retention capacity relative to solid particles ofknown size and the pore size distribution. These properties are examined in alaboratory environment and are critical for comparing different filter media.

The permeability relative to a pure liquid, usually water, may be determined with thehelp of different devices that operate on the principle of measurement of filtratevolume obtained over a definite time interval at known pressure drop and filtrationarea. The permeability is usually expressed in terms of the hydraulic resistance of thefilter medium. This value is found from:

,, ApSV = ——£ T >)

When the cake thickness is 0, we may write the equation as:

Rf - ft (3)/ y W

Note that:

Ap=Apt-Apf (4)

where Apt= pressure difference accounting for the hydrostatic pressure of a liquidcolumn at its flow through the filter medium, supporting structure anddevice channels

Apf = same pressure drop when the flow of liquid is through the supportingstructure and device channels

Analytical determination of the hydraulic resistance of the medium is difficult.However, for the simplest filter medium structures, certain empirical relationships areavailable to estimate hydraulic resistance. The relationship of hydraulic resistance ofa cloth of monofilament fiber versus fiber diameter and cloth porosity can be basedon a fixed-bed model.

In evaluation and selection of a filter medium, one should account for the fact thathydraulic resistance increases gradually with time. In particular, the relationshipbetween cloth resistance and the number of filter cycles is defined by:

* = RineKN (5)

The retentivity relative to solid particles (e.g., spherical particles of polystyrene ofdefinite size) is found from experiments determining the amount of these particles inthe suspension to be filtered before and after the filter media. The retentivity K is


determined as follows: where g', g" = amounts of solid particles in liquid samplebefore and after the medium, respectively.

The pore sizes distribution, as well as the average pore size, is determined by the"bubble" method. The filter medium to be investigated is located over a supportingdevice under a liquid surface that completely wets the medium material. Air isintroduced to the lower surface of the medium. Its pressure is gradually increased,resulting in the formation of single chains of bubbles. This corresponds to air passagesthrough the largest-diameter pores. As pressure is increased, the number of bubblechains increases due to air passing through the smaller pores. In many cases a criticalpressure is achieved where the liquid begins to "boil." This means that the filtermedium under investigation is characterized by sufficiently uniform pores. If there isno "boiling," the filter medium has pores of widely different sizes. The pore sizethrough which air passes is calculated from known relations. For those pores whosecross section may be assumed close to a triangle, the determining size should be thediameter of a circle that may be inscribed inside the triangle.

For orientation in cloth selection for a given process, the following information isessential: filtration objectives (obtaining cake, filtrate or both), and complete data (ifpossible) on the properties of solid particles (size, shape and density), liquid (acid,alkali or neutral, temperature, viscosity, and density), suspension (ratio of solids toliquid, particle aggregation and viscosity), and cake (specific resistance,compressibility, crystalline, friable, plastic, sticky or slimy). Also, the requiredcapacity must be known as well as what constitutes the driving force for the process(e.g., gravity force, vacuum or pressure). Based on such information, an appropriatecloth that is resistant to chemical, thermal and mechanical aggression may be selected.In selecting a cloth based on specific mechanical properties, the process driving forceand filter type must be accounted for. The filter design may determine one or moreof the following characteristics of the filter cloth: tensile strength, stability in bending,stability in abrasion, and/or ability of taking the form of a filter-supporting structure.Tensile strength is important, for example, in belt filters. Bending stability isimportant in applications of metallic woven cloths or synthetic monofilament cloths.If the cloth is subjected to abrasion, then glass cloth cannot be used even though it hasgood tensile strength.

From the viewpoint of accommodation to the filter-supporting structure, some clothscannot be used, even though the filtering characteristics are excellent. For rotary drumfilters, for example, the cloth is pressed onto the drum by the "caulking" method,which uses cords that pass over the drum. In this case, the closely woven clothsmanufactured from monofilament polyethylene or polypropylene fiber are lessdesirable than more flexible cloths of polyfilament fibers or staple cloths.


Depending on the type of filter device, additional requirements may be made of thecloth. For example, in a plate-and-frame press, the sealing properties of cloths arevery important. In this case, synthetic cloths are more applicable staple cloths,followed by polyfilament and monofilament cloths. In leaf filters operating undervacuum and pressure, the cloth is pulled up onto rigid frames. Since the size of a clothchanges when in contact with the suspension, it should be pretreated to minimizeshrinkage.

In selecting cloths made from synthetic materials, one must account for the fact thatstaple cloths provide a good retentivity of solid particles due to the short hairs on theirsurface. However, cake removal is often difficult from these cloths - more than fromcloths of polyfilament and, especially, monofilament fibers. The type of fiber weaveand pore size determine the degree of retentivity and permeability. The objective ofthe process, and the properties of particles, suspension and cake should be accountedfor. The cloth selected in this manner should be confirmed or corrected by laboratorytests. Such tests can be performed on a single filter. These tests, however, provide noinformation on progressive pore plugging and cloth wear. However, they do provideindications of expected filtrate pureness, capacity and final cake wetness.

A single-plate filter consists of a hollow flat plate, one side of which is covered bycloth. The unit is connected to a vacuum source and submerged into the suspension(filtration), then suspended in air to remove filtrate, or irrigated by a dispersed liquid(washing). The filter cloth is directed downward or upward or located vertically,depending on the type of filter that is being modeled in the study.

The following is a recommended sequence of tests that can assist in cloth selection forcontinuous vacuum filters.

If the cycle consists of only two operations (filtration and dewatering), tests should beconducted to determine the suspension weight concentration after 60 sec of filtrationand 120 sec of dewatering. The cake thickness should be measured and the cakeshould be removed to determine the weight of wet cake and the amount of liquid in it.The weight of filtrate and its purity are also determined. If the cake is poorly removedby the device, it is advisable to increase the dewatering time, yacuum or both. If thecake is poorly removed after an operating regime change, it should be tested withanother cloth. If the cake is removed satisfactorily, filtration time should be decreasedunder increased or decreased vacuum. Note that compressible cakes sometimes plugpores faster at higher vacuum. After the filtration test for a certain filter cycle (whichis based on the type of the filter being modeled), the suspension's properties shouldbe examined. Based on the assumed cycle, a new filtration test should be conductedand the characteristics of the process noted. Capacity (N/m2-hr), filtration rate (nrVm2-hr) and cake wetness can then be evaluated. Also, if possible, the air rate anddewatering time should be computed. The results of the first two or three tests shouldnot be taken into consideration because they cannot exactly characterize the propertiesof the cloth. A minimum of four or five tests is generally needed to achievereproducible results of the filtration rate and cake wetness to within 3-5 %.

Filter Media and Use of Filter Aids 4 7

When the cycle consists of filtration, washing and dewatering, the tests are consideredprincipally in the same manner. The economic aspects of cloth selection should beconsidered after complete determination of cloth characteristics.

Introduction to the Use of Filter Aids

Filter aids and/or flocculants are employed to improve the filtration characteristics ofhard-to-filter suspensions. A filter aid is a finely divided solid material, consisting ofhard, strong particles that are, en masse, incompressible. The most common filter aidsapplied as an admix to the suspension, are diatomaceous earth, expanded perlite,Solkafloc, fly ash or carbon. Filter aids build up a porous, permeable and rigid latticestructure that retains solid particles and allows the liquid to pass through. Thesematerials are applied in small quantities in clarification or in cases where compressiblesolids have the potential to foul the filter medium.

Filter aids may be applied in one of two ways. The first method involves the use ofa precoat filter aid, which can be applied as a thin layer over the filter before thesuspension is pumped to the apparatus. A precoat prevents fine suspension particlesfrom becoming so entangled in the filter medium that its resistance becomes excessive.Further, it facilitates the removal of cake at the end of the filtration cycle.

The second application method involves incorporation of a certain amount of thematerial with the suspension before introducing it to the filter. The addition of filteraids increases the porosity of the sludge, decreases its compressibility, and reduces theresistance of the cake.

In some cases the filter aid displays an adsorption action, which results in particleseparation of sizes down to 0.1 /mi. The adsorption ability of certain filter aids, suchas bleached earth and activated charcoals, is manifest by a decoloring of thesuspension's liquid phase. This practice is widely used for treating fats and oils.

Filter aids are also used in processing sugar, beer, wine, gelatin, antibiotics, glycerol,solvents, synthetic tars, caustic, sulfur and uranium salts, for water cleaning, and inpreparing galvanic solutions. The properties of these additives are determined by thecharacteristics of their individual components. For any filter aid, size distribution andthe optimal dosage are of great importance. Too low a dosage results in poor clarity;too great a dosage will result in the formation of very thick cakes. In general, a goodfilter aid should form a cake having high porosity (typically 0.85-0.9), low surfacearea and good particle-size distribution. An acceptable filter aid should have a muchlower filtration resistance than the material with which it is being mixed. It shouldreduce the filtration resistance by 67-75 % with the addition of no more than 25 % byweight of filter aid as a fraction of the total solids.

The addition of only a small amount of filter aid (e.g., 5% of the sludge solids) canactually cause an increase in the filtration resistance. When the amount of filter aid is


so small that the particles do not interact, they form a coherent structure, andresistance may be affected adversely.

filter Aid Requirements

Filter aids are assessed in terms of the rate of filtration and clarity of filtrate. Finelydispersed filter aids are capable of producing clear filtrate; however, they contributesignificantly to the specific resistance of the medium. As such, their application mustbe made in small doses. Filter aids comprised of coarse particles contributeconsiderably less specific resistance; consequently, a high filtration rate can beachieved with their use. Their disadvantage is that a "muddy" filtrate is produced.

The optimum filter aid should have maximum pore size and ensure a prespecifiedfiltrate clarity. Desirable properties/characteristics for the optimum filter aid include:

• The additive should provide a thin layer of solids having high porosity (0.85-0.90) over the factormedium's external surface. Suspension particles will ideally form a layered cake over the filter aid cakelayer. The high porosity or the filter aid layer will ensure a high filtration rate. Porosity is notdetermined by pore size alone. High porosity is still possible with small size pores,

• Filter aids should have low specific surface, since hydraulic resistance results from frictional lossesincurred as liquid flows past particle surfaces. Specific surface is inversely proportional to particle size.The rate of particle dispersity and the subsequent difference in specific surface determines diedeviations in filter aid quality from one material to another. For example, most of the diatomite specieshave approximately the same porosity; however, the coarser materials experience a smaller hydraulicresistance and have much less specific surface than the finer particle sizes.

• Filter aids should have a narrow fractional composition. Fine particles increase the hydraulic resistanceof the filter aid, whereas, coarse particles exhibit poor separation. Desired particle size distributionsare normally prepared by air classification, in which the finer size fractions are removed.

• In applications where the filter aid layer is to be formed on open-weave synthetic fabric or wire screens,wider size distributions may have to be prepared during operation. Filter aids should therefore have theflexibility to be doped with amounts of coarser sizes. This provides rapid particle bridging and settlingor the filter aid layer. For example, diatomite having an average particle size of 8 /urn may be readilyapplied to a screen with a mesh size of 175 ̂ m by simply adding a small quantity of filter aid with sizesthat are on the same order, but less in size than the mesh openings. Particle sizes typically around 100jum will readily form bridges over the screen openings and prevent the loss of filter aid in this example.

• The filter aid should be chemically inert to the liquid phase of the suspension and not decompose ordisintegrate in it.

The ability of an admix to be retained on the filter medium depends on both thesuspension's concentration and the filtration rate during this initial precoat stage. Thesame relationships for porosity and the specific resistance of the cake as functions ofsuspension concentration and filtration rate presented in Chapter 3 apply equally tofilter aid applications.

Applications

Filter aids are added in amounts needed for a suspension to acquire desirable filteringproperties and to prepare a homogeneous suspension before actual filtration. Filteraids increase the concentration of solids in the feed suspension. This promotes particle


bridging and creates a rigid lattice structure for the cake. In addition, they decreasethe floe deformation tendency. Irregular or angular particles tend to have betterbridging characteristics than spherical particles. As a rule, the weight of aid added tothe suspension should equal the particle weight in suspension. Typical filter aidadditions are in the range of 0.01-4% by weight of suspension; however, the exactamount can only be determined from experiments. Excess amounts of filter aid willdecrease the filter rate. Operations based on the addition of admixes to theirsuspensions may be described by the general equations of filtration with cakeformation. A plot of filtration time versus filtrate volume on rectangular coordinatesresults in a nearly parabolic curve passing through the origin. The same plot onlogarithmic coordinates, assuming that the medium resistance may be neglected,results in a straight line. This convenient linear relationship allows results obtainedfrom short-time filtration tests to be extrapolated to long-term operating performance(i.e., for several hours of operation). This reduces the need to make frequent, lengthytests and saves time in the filter selection process.

In precoating, the prime objective is to prevent the filter medium from fouling. Thevolume of initial precoat normally applied should be 25-50 times greater than thatnecessary to fill the filter and connecting lines. This amounts to about 5-10 lb/100 ft2

of filter area, which typically results in a V16- to !/g-in, precoat layer over the outersurface of the filter medium. An exception to this rule is in the precoating ofcontinuous rotary drum filters where a 2- to 4-in. cake is deposited before filtration.The recommended application method is to mix the precoat material with clear liquor(which may consist of a portion of the filtrate). This mixture should be recycled untilall the precoat has been deposited onto the filter medium. The unfiltered liquor followsthrough immediately without draining off excess filter aid liquor. This operationcontinues until a predetermined head loss develops, when the filter is shut down forcleaning and a new cycle.

In precoating, regardless of whether the objective is to prevent filter medium cloggingor to hold back fines from passing through the medium to contaminate the filtrate, themechanical function of the precoat is to behave as the actual filter medium. Since itis composed of incompressible, irregularly shaped particles, a high-porosity layer isformed within itself, unless it is impregnated during operation with foreigncompressible materials. Ideally, a uniform layer of precoat should be formed on thesurface of the filter medium. However, a nonuniform layer of precoat often occursdue to uneven medium resistance or fluctuations in the feed rate of filter aidsuspension. Cracks can form on the precoat layer that will allow suspension particlesto penetrate into the medium. To prevent cracking, the filter aid may be applied as acompact layer. On a rotating drum filter, for example, this may be accomplished byapplying a low concentration of filter aid (2-4%) at the maximum drum rpm. In otherfilter systems, maintaining a low pressure difference during the initial stages ofprecoating and then gradually increasing it with increasing layer thickness until thestart of filtration will help to minimize cake cracking. Also, with some filter aids (suchas diatomite or perlite), the addition of small amounts of fibrous material will producea more compact precoat cake. At low suspension concentrations (typically 0.01%),filter aids serve as a medium under conditions of gradual pore blocking. In this casethe amount of precoat is 10-25 N/m2 of the medium and its thickness is typically 3-10


mm. In such cases, the filter aid chosen should have sufficient pore size to allowsuspension particle penetration and retention within the precoat layer.

Examples of Filter Aids

The following are examples of some of the most common filter aids.

Diatomite

The most important filter aids from a volume standpoint are the diatomaceous silicatype (90% or better silica). These are manufactured from the siliceous fossil remainsof tiny marine plants known as diatoms.

Diatomaceous filter aids are available in various grades. This is possible because thenatural product can be modified by calcining and processing, and because filter aidsin different size ranges and size distributions have different properties. The filter aidsmay be classified on the basis of cake permeability to water and water flowrate. Finergrades are the slower-filtering products; however, they provide better clarificationthan do faster-filtering grades. Thus a fast-filtering aid may not provide the requiredclarification. However, by changing the physical character of the impurities (e.g., byproper coagulation), the same clarity may be obtained by using the fast-filteringgrades.

Calcinated diatomaceous additives are characterized by their high retention ability withrelatively low hydraulic resistance. Calcining dramatically affects the physical andchemical properties of diatomite, making it heat-resistant and practically insoluble instrong acids.

Perlite

Perlite is glasslike volcanic rock, called "volcanic glass," consisting of small particleswith cracks that retain 2-4% water and gas. Natural perlite is transformed to a filteraid by heating it to its melting temperature (about 1000°C), where it acquires plasticproperties and expands due to the emission of steam and gas. Under these conditionsits volume increases by a factor of 20. Beads of the material containing a large numberof cells are formed. The processed material is then crushed and classified to providedifferent grades.

The porosity of perlite is 0.85-0.9 and its volumetric weight is 500-1000 N/m3.Compared to diatomite, perlite has a smaller specific weight and comparable filterapplications typically require 30% less additive. Perlite is used for filtering glucosesolutions, sugar, pharmaceutical substances, natural oils, petroleum products,industrial waters and beverages. The principal advantage of perlite over diatomite isits relative purity. There is a danger that diatomite may foul filtering liquids withdissolved salts and colloidal clays.


Cellulose

Cellulose fiber is applied to cover metallic cloths. The fibers form a highlycompressed cake with good permeability for liquids, but a smaller retention ability forsolid particles than that of diatomite or perlite. The use of cellulose is recommendedonly in cakes where its specific properties are required. These properties include alack of ashes and good resistance to alkalies. The cost of cellulose is higher than thoseof diatomite and perlite.

Sawdust

This filter aid may be employed in cases where the suspension particles consist of avaluable product that may be roasted. For example, titanium dioxide can bemanufactured by calcining a mixture of sawdust and metal titanium acid. The mixtureis obtained as a filter cake after separating the corresponding suspension with a layerof filter aid.

Charcoal

Charcoal is not only employed in activated form for decoloring and adsorbingdissolved admixtures, but also in its unactivated form as a filter aid. It can be used insuspensions consisting of aggressive liquids (e.g., strong acids and alkalies). As withsawdust, it can be used to separate solids that may be roasted. On combustion, thecharcoal leaves a residue of roughly 2% ash. Particles of charcoal are porous and formcakes of high density but that have a lesser retention ability than does diatomite.

Fly Ash

This material has a number of industrial filtering applications, but primarily is appliedto dewatering sewage sludge. The precoat is built up to 2 in. thick from a 60% solidslurry. On untreated sludges, filtration rates of 25 Ib/ft2-hr or higher are obtainable.This rate can be more than doubled with treated sludges. The sludge is reduced froma liquid to a semidry state. Fly ash may also be used as a precoat in the treatment ofpapermill sludge.

Filter Aid Selection

Filter aid selection should be based on laboratory tests. Guidelines for selection mayonly be applied in the broadest sense, since there is almost an infinite number ofcombinations of filter media, filter aids and suspensions that will produce varyingdegrees of separation. The hydrodynamics of any filtration process are highlycomplex. Filtration is essentially a multiphase system in which interaction takes placebetween solids from the suspension, filter aid, filter medium, and a liquid phase.Experiments are needed in most applications not only in proper filter aid selection butin defining the method of their administration. Some general guidelines can be appliedto such evaluations:


• The filter aid must have the minimum hydraulic resistance and provide the desired rate of separation,

• An insufficient amount of filter aid leads to a reduction in filtrate quality. Excess amounts result inlosses in filtration rate.

• It is necessary to account for the method of application and characteristics of filter aids.

For precoat applications to rotary filters, the relative usefulness of various filter aidsrequires experimental data on the effects on the following variables:

1. type and grade of filter aid used as the precoat,2. drum speed,3. knife advance rate,4. vacuum pressure,5. filter medium,6. temperature of feed,7. filter medium submergence,8. manner of precoat application,9. sharpness of knife, and10. knife angle and bevel.

Tests should be conducted under a constant set of operating conditions such aspressure and temperature.

Such tests will reveal that the manner in which filter aid is removed depends on itsproperties and the drum speed. In some cases the filter aid will be removed as flakes,while other tests will show that the remaining portion of the cake will acquire a roughsurface. High drum speed and filter aid dispersity are known to enhance the qualityof the precoat layer.

For a low rate of knife advance the resistance of the filter aid decreases in spite of thedecrease in cake thickness. This is explained by the fact that for a single rotation ofthe drum, the particles of the suspension penetrate into the pores of filter at depthsgreater than 1.5 mm, and blockage occurs. Rotary drum filler systems are character-ized not only by rate, but by the ratio of filtrate volume to the weight of filter aid cut.Some generalizations for this system are:

1. The greater the permeability of the filter aid, the greater the amounts of filtrate collected.

2. A gradual decrease in filtration rate can be observed; however, at some point during the operationa large reduction in rate can also be observed.

3. In the operating regime characterized by a sharp drop in the filtration rate, the permeability ofthe filter aid is not zero, but has a definite value.

4. Two operating periods lead to a reduction in permeability. During the first period, filtrationprimarily proceeds with pore blocking on the outer surface of the filter aid. During the secondperiod, the filter aid pores undergo blocking within the filter aid cake.

Figure 2 illustrates the operation of a precoat filter with cake in place. Point Arepresents the entrance of the cake into the unfiltered suspension. Point B representsthe point of emergence and C is the point at which the cake, together with itsaccumulated suspension particles, reaches minimum permeability. Maximum filtration

Filter Media and Use of Filter Aids

Figure 2. Cross-sectional view of a precoat filter.

economy can be achieved by proper combination of submergence depth and rate ofrotation so that points B and C coincide. If minimum permeability is not achieved beforepoint C reaches location B, the drum is rotating too rapidly or the submergence is toolow, resulting in a portion of the filter aid capacity being wasted. Conversely, ifminimum permeability is reached before C reaches B, either the submergence levelcould be lowered or the drums rotated at a faster rate. Choice of decreasing thesubmergence level or increasing the rate of rotation would be dictated by the timerequirement for the drainage of accumulated filtrate from the filter cake. In accordancewith this thinking, if points B and C coincide with 50% submergence and 10 rpm drumspeed, a reduction in drum speed to 0.67 rpm would move point C back to Cls a furtherreduction in drum speed to 0.5 rpm would locate it at C2. If the blinding medium haszero effective permeability, filtrate collection under these circumstances should havea linear relationship with respect to drum speed. However, if the blinding medium hasappreciable permeability, additional filtrate will be collected in proportion to thatpermeability through the distances BC, and BC2 for 0.67- and 0.5-rpm speeds,respectively.

A graphical method is available for determining filter aid efficiency. Figure 3 showsa plot of filtrate rate versus cake thickness. Curve AB corresponds to the filtration ofclear liquid through a gradually decreasing layer of filter aid when the resistance ofthe filter medium is negligible. The curve is based on the basic law of filtration usinga special permeability for the filter aid. The line EF corresponds to suspensionseparation through a gradually decreasing layer of filter aid and filter medium ofknown resistance. Areas ABCD and EFCD are proportional to the filtrate volumes inan ideal and actual process, respectively. Since the areas under the curves areproportional to total output of the cycle, the observed output (CDEF) is much lowerthan that theoretically obtainable with clear liquid (ABCD). If we arbitrarily call thefilter aid 100% rate-efficient when operating with solids-free liquid under idealconditions, then the rate efficiency with the slurry involved may be calculated asfollows:


O 3

O-IILit,Oin*•<K

Figure 3. Relationship between blinding losses and theoretical efficiency.

area CDEFarea ABCD

xlOO =CDEF

ABCDxlOO

The expression for curve AB in terms of a and b is:

e = o 4 (8)

Hence, area ABCD may be computed by integration of the curve's formula. Analternative approach is to draw the equivalent line AB1, so that the cross-hatched areasare equal, from whence the rate efficiency is:

r, =r AB'CD

The following approximate procedure may also be used. Note in Figure 3 mat thecenter of the cake is 1.25 in. Areas AB'CD and CDEF can be computed from theproduct of the intercept of the diagonal line AB1 or EF and base CD.


MOxCD MO

The location of line AB can be determined for any filter aid from its permeability.

During the period when the knife performs a full cut, the filtration rate is at amaximum. During the inactive periods of the knife, the filtration rate is reduced to alevel representative of the permeability of the blinding layer. In Figure 3, this resultsin a filtration rate curve approximately parallel to the theoretical curve and displaceddownward. As shown, these curves are nearly linear. This supports the assumptionthat the resistance of the filter aid cake does not significantly restrict filtrate flowthrough the filter medium. Note that the total area ABCD theoretically represents100% filter aid rate efficiency and the area under observed curve EF is theperformance attained. Then, the area above the observed curve denotes the loss of rateefficiency due to blinding.

Blinding and penetration are rarely separate effects. Figure 4 illustrates a graphicevaluation of the rate efficiency together with applicable permeability corrections forthe blinding medium. The total area under theoretical curve AB1 represents 100% rateefficiency. That area under the actual line EF represents the filtration rate efficiencyof the filter aid alone. The middle region denotes the loss of filter aid capacity due toblinding (i.e., pore blocking). The top triangular region denotes the loss of capacityof the filter aid due to penetration of solids into the cake beyond the cutting depth ofthe knife.

The average filtration rate and filtrate volume produced per unit weight of filter aidspent may be determined from Figure 4. The combined influence of both factors onthe filtration efficiency is different depending on actual operating conditions.

The penetration of fine solids through the filter aid results in larger pores andconsequently in smaller specific resistances, which are less applicable than those withpores of smaller size where the fine particles do not penetrate. This can be explainedby the fact that the specific resistance of the first filter aid where particles penetratedbecomes higher than that of the second filter aid.

The efficiency of a rotary drum filter is expressed by the product of the averagefiltration rate and the ratio of filtrate volume to weight of filter aid spent. Efficiencyreaches a maximum at different rates of knife motion and varies according to theproperties of the specific filter aid. An increase in the thickness of a cake layer cutresults in a ratio of filtrate volume to weight of filter aid spent. One should accountfor the volume weight of different filter aids since this parameter varies considerablyeven among different grades of the same material.


To maintain an acceptable filtration rate, the following general guidelines apply:

1. If suspension particles are observed to intensely penetrate the filter aid cake, anothergrade with better solids retention properties should be selected.

2, If blinding of the outer surface of the filter aid is observed, then the rate of rotationshould be increased.

2.0 1.5 1.0 0.5 SeptumC A K E THICKNESS, L

Figure 4. Graphical evaluation of filter aid efficiency and blinding and penetration losses,

After preliminary selection of the filter aid type, the optimum amount of addition toa suspension and the preferred rate of filtration must be established. Final selectionshould be based on comparison of the suspension separation efficiency for differentfilter aids and filter aid grades.

Some general industry guidelines for filter aid precoating as applied to batch filterswith a flat filter medium operating under pressure are given below. Examples ofspecific filter types that fall into this category are filter presses and leaf filters.

It is generally recommended that a suspension of filter aid be prepared in a pure liquidby agitation before filtration. Delivery to the filter medium can usually be done witha centrifugal pump. The filtrate can be recycled to the filter aid suspension feed untilthe finished precoat is formed. Depending on the ratio of particle size of the filter aidto the pore diameters of the medium, as well as the rate of suspension feed and itsconcentration, two processes arise on the filter. First, the medium retains only someparticles and the remainder pass through with filtrate. Gradually, the medium's


retention ability increases and the amount of solid particles in the filtrate decreases.A short time after starting the process, the retention ability of the medium becomesso pronounced (for example, due to bridging) that the pure filtrate passes through,

In both cases, the suspension concentration in the filter body decreases gradually. Thesuspension of filter aid is usually delivered on the filter not filled with liquid;however, in some cases the filter is filled with pure liquid during the precoat stage,in such cases, the filter and suspension entering the unit is diluted. Its concentrationfirst increases, reaching a maximum value, then as the particles settle onto the filter,it decreases.

In most cases we may assume that the filtrate volume is equal to the suspension volume(the concentration of filter aid in suspension is typically 0.1-0.5%). The cake may beassumed incompressible. Its specific resistance is independent of concentration. Also,within the filter body, the suspension can be assumed to be perfectly mixed, withfiltration at a constant rate.

The following expressions describe the case where the suspension of filter aid isdelivered to a filter not filled with liquid. For filter aid particles completely retainedby the medium:

C = C0e-»

For the case where the retention ability of the medium increases gradually:

nlkcm A — T (12)V" v«"r>

where C = variable concentration of suspension in the filter body (mVm3)C0 = initial concentration of suspension in the mixer (mVm3)n = V/VS = degree of circulation of suspension in the systemV = filtrate volume (m3)Vs = initial volume of suspension in the mixer (m3)K= constantA,= Q/VS = coefficient of dissolution (sec"1)Q = amount of filtered liquid (m3/sec)

Suggested Readings

1. Masschelein, W.S., Unit Process in Drinking Water Treatment, MarcelDecker, Inc., NJ, 1992

2. Veshilind, P. A, Treatment and Disposal of Wastewater Sludges, Ann ArborScience, Ann Arbor, MI, 1975.


3. Noyes, R., Handbook of Pollution Control Processes, Noyes Publications,NJ, 1991

4. Cheremisinoff, N.P. and D.S. Azbel, Liquid Filtration for Process andPollution Control, SciTech Publishers, Inc, NJ.1989

Nomenclature

a = addition of filter aid (kg)b = coefficientc = concentration of filtration-impeding impurities (kg/m3)c() = initial concentration of suspension in the filter body

(kg/m3 or m3/m3)g', g" = amount of solid particles in liquid before and after the medium,

respectively (kg/m )K, K = emperical constants or retentivityL = cake thickness (in.)N = degree of circulation of suspension in the systemp = pressure (N/m3)Q = volumetric flowrate (m3/sec)r0 = specific cake resistance (m"1)R = total filtration resistance (m/sec)V = filtrate volume (m3)

Greek Symbols

rjr = filtration rate efficiency (%)1 = coefficient of dissolution (sec"1)jU = viscosity (P)i = time (sec)

Introduction

Cake filtration is the most common form of filtering employed by the chemical andprocess industries. This manifests into handling the permeability of a bed of porousmaterial, the schematic of which is shown in Figure 1, With high-solids-concentrationsuspensions, even relatively small particles (in comparison to the pore size) will notpass through the medium, but tend to remain on the filter surface, forming "bridges"over individual openings in the filter material.

Filtrate flows through the filter medium and cake because of an applied pressure, themagnitude of which is proportional to the filtration resistance. This resistance resultsfrom the frictional drag on the liquid as it passes through the filter and cake.

Hydrostatic pressure varies from a maximum at the point where liquid enters the cake,to zero where liquid is expelled from the medium; consequently, at any point in thecake the two are complementary. That is, the sum of the hydrostatic and compressionpressures on the solids always equals the total hydrostatic pressure at the face of thecake. Thus, the compression pressure acting on the solids varies from zero at the faceof the cake to a maximum at the filter medium.

When solid particles undergo separation from the mother suspension, they arecaptured both on the surface of the filter medium and within the inner pore passages.The penetration of solid particles into the filter medium increases the flow resistanceuntil the filtration cycle can no longer continue at economical throughput rates, atwhich time the medium itself must be replaced.

This chapter provides a summary of standard calculation methods for assessing cakeformation, behavior, and the overall efficiency of the filter-medium filtration process.

3

CAKE FILTRATION AND FILTER MEDIAFILTRATION


Dynamics of Cake Filtration

When the space above the suspension is subjected to a source of compressed gas (e.g.,air) or the space under the filter plate is connected to a vacuum source, filtration isaccomplished under a constant pressure differential (the pressure in the receivers isconstant). In this case, the rate of the process decreases due to an increase in the cakethickness and, consequently, flow resistance. A similar filtration process results froma pressure difference due to the hydrostatic pressure of a suspension layer of constantthickness located over the filter medium.

If the suspension is fed to the filter with a reciprocating pump at constant capacity,filtration is performed under constant flowrate. In this case, the pressure differentialincreases due to an increase in the cake resistance. If the suspension is fed by acentrifugal pump, its capacity decreases with an increase in cake resistance, andfiltration is performed at variable pressure differentials and flowrates.

The most favorable filtration operation with cake formation is a process whereby noclogging of the filter medium occurs. Such a process is observed at sufficiently highconcentrations of solid particles in suspension. From a practical standpoint thisconcentration may conditionally be assumed to be in excess of 1 % by volume.

Figure 1. Operating scheme of a filtration process; 1-filter; 2-filter medium; 3-suspension;4-filtrate; 5-cake.

To prevent pore clogging in the filter medium when handling relatively low solidsconcentrations (e.g., 0.1-1% by volume), general practice is to increase the solidsconcentration in thickeners before the suspension is fed to the filter.

Filtration is frequently accompanied by hindered or free gravitational settling of solidparticles. The relative directions of action between gravity force and filtrate motion

Cake Filtration and Filter Media Filtration 61

may be concurrent, countercurrent or crosscurrent, depending on the orientation of thefilter plate, as well as the sludge location above or below the filter plate. The differentorientations of gravity force and filtrate motion with their corresponding distributionof cake, suspension, filtrate and clear liquid are illustrated in Figure 2. Particlesedimentation complicates the filtration process and influences the controllingmechanisms. Furthermore, these influences vary depending on the relative directionsof gravity force and filtrate motion. If the suspension is above the filter medium(Figure 2A), particle settling leads to more rapid cake formation with a clear filtrate,which can be evacuated from the filter by decanting. If the suspension is under thefilter medium (Figure 2B), particle settling will prevent cake formation, and it isnecessary to mix the suspension to maintain homogeneity.

r\.:*

2 3

( A ) ( B )

Figure 2. Direction of gravity force action and filtrate motion in filters: A-cocurrent; B-countercurrent; C-crosscurrent; solid arrow-direction of gravity force action; dashed arrow-direction of filtrate motion; 1 -filter plate; 2-cake; 3-sludge; 4-filtrate; 5-clear liquid.

When the cake structure is composed of particles that are readily deformed or becomerearranged under pressure, the resulting cake is characterized as being compressible.Those that are not readily deformed are referred to as semicompressible, and thosethat deform only slightly are considered incompressible. Porosity (defined as the ratioof pore volume to the volume of cake) does not decrease with increasing pressuredrop. The porosity of a compressible cake decreases under pressure, and its hydraulicresistance to the flow of the liquid phase increases with an increase in the pressuredifferential across the filter media.

Cakes containing particles of inorganic substances with sizes in excess of 100 jum maybe considered incompressible, for all practical purposes. Examples of incompressiblecake-forming materials are sand and crystals of carbonates of calcium and sodium.The cakes containing particles of metal hydroxides, such as ferric hydroxide, cuprichydroxide, aluminum hydroxide, and sediments consisting of easy deformingaggregates, which are formed from primary fine crystals, are usually compressible.


At the completion of cake formation, treatment of the cake depends on the specificfiltration objectives. For example, the cake itself may have no value, whereas thefiltrate may. Depending on the disposal method and the properties of the particulates,the cake may be discarded in a dry form, or as a slurry. In both cases, the cake isusually subjected to washing, either immediately after its formation, or after a periodof drying. In some cases, a second washing is required, followed by a drying periodwhere all possible filtrate must be removed from the cake; or where wet discharge isfollowed by disposal: or where repulping and a second filtration occurs; or where drycake disposal is preferable. Similar treatment options are employed in cases where thecake is valuable and all contaminating liquors must be removed, or where both cakeand filtrate are valuable. In the latter, cake-forming filtration is employed, withoutwashing, to dewater cakes where a valueless, noncontaminating liquor forms theresidual suspension in the cake.

To understand the dynamics of the filtration process, a conceptual analysis is appliedin two parts. The first half considers the mechanism of flow within the cake, while thesecond examines the external conditions imposed on the cake and pumping system,which brings the results of the analysis of internal flow in accordance with theexternally imposed conditions throughout.

The characteristics of the pump relate the applied pressure on the cake to the flowrateat the exit face of the filter medium. The cake resistance determines the pressure drop.During filtration, liquid flows through the porous filter cake in the direction ofdecreasing hydraulic pressure gradient. The porosity (e) is at a minimum at the pointof contact between the cake and filter plate (i.e., where x = 0) and at a maximum atthe cake surface (x = L) where sludge enters. A schematic definition of this systemis illustrated in Figure 3.

SUBSCRIPT

SLURRY

Figure 3. Important parameters in cake formation.

The drag that is imposed on each particle is transmitted to adjacent particles.Therefore, the net solid compressive pressure increases as the filter plate isapproached, resulting in a decrease in porosity. Referring to Figure 4A, it may be


assumed that particles are in contact at one point only on their surface, and that liquidcompletely surrounds each particle. Hence, the liquid pressure acts uniformly in adirection along a plane perpendicular to the direction of flow. As the liquid flows pasteach particle, the integral of the normal component of force leads to form drag, andthe integration of the tangential components results in frictional drag. If the particlesare non-spherical, we may still assume single-point contacts between adjacent particlesas shown in Figure 4B.

Now consider flow through a cake (Figure 4C) with the membrane located at adistance x from the filter plate. Neglecting all forces in the cake other than thosecreated by drag and hydraulic pressure, a force balance from x to L gives:

Fs + ApL = Ap (1)

The applied pressure p is a function of time but not of distance x. Fs is the cumulativedrag on the particles, increasing in the direction from x = L to x = 0. Since singlepoint contact is assumed, the hydraulic pressure pL is effectively over the entire crosssection (A) of the cake; for example, against the fictitious membrane shown in Figure4B. Dividing Equation 1 by A and denoting the compressive drag pressure by ps =F/A, we obtain:

MEDIUM

( B J

SURFACE

SLURRY

(C )

Figure 4. Frictional drag on particles in compressible cakes.

The term ps is a fictitious pressure, because the cross-sectional area A is not equal toeither the surface area of the particles nor the actual contact areas In actual cakes,there is a small area of contact Ac whereby the pressure exerted on the solids may bedefined as FS/AC.

Taking differentials with respect to x, in the interior of the cake, we obtain:

dp, + dp, = 0 (3)


This expression implies that drag pressure increases and hydraulic pressure decreasesas fluid moves from the cake's outer surface toward the filter plate.

From Darcy's law, the hydraulic pressure gradient is linear through the cake if theporosity (e) and specific resistance (a) are constant. The cake may then be consideredincompressible. This is illustrated by the straight line obtained from a plot of flowrateper unit filter area versus pressure drop shown in Figure 5. The variations in porosityand specific resistance are accompanied by varying degrees of compressibility, alsoshown in Figure 5.

As noted in Chapter 1, filtration is primarily an application of fluid mechanics; thatis, filtrate flow is induced through a porous filter cake and filter medium. The rate ofthe filtration process is directly proportional to the driving force and inverselyproportional to the resistance.

INCOMPRESSIBLE

NORMALLYCOMPRESSIBLE

i-n-flw / ^^ Affe(T / ^X^ n ft ,-",..%.

XLUh I ' HIGHLY

COMPRESSI8LE

qct

PRESSURE DROP, Ap c

Figure 5. Flowrate/area versus pressure drop across the cake.

Because pore sizes in the cake and filter medium are small, and the liquid velocitythrough the pores is low, the filtrate flow may be considered laminar: hence,Poiseuille's law is applicable. Filtration rate is directly proportional to the differencein pressure and inversely proportional to the fluid viscosity and to the hydraulicresistance of the cake and filter medium. Because the pressure and hydraulicresistances of the cake and filter medium change with time, the variable rate offiltration may be expressed as:

Adi (4)


where V = volume of filtrate (m3)A = filtration area (m2)T = time of filtration (sec)

Assuming laminar flow through the filter channels, the basic equation of filtration asobtained from a force balance is:

1 dV Apu = - ^

A JT p(Re + Rf) IJ'

where Ap = pressure difference (N/m2)fji, = viscosity of filtrate (N-sec/m2)Rc = filter cake resistance (m"1)Rf = initial filter resistance (resistance of filter plate and

filter channels) (m4)u = filtration rate (m/sec), i.e., filtrate flow through cake

and filter platedV/dt = filtration rate (nvVsec), i.e., filtrate flow rate

Filter cake resistance (Rc) is the resistance to filtrate flow per unit area offiltration. Rc increases with increasing cake thickness during filtration. At anyinstant, RC depends on the mass of solids deposited on the filter plate as a result of thepassage of V (m3) filtrate. Rf may be assumed a constant. To determine therelationship between volume and residence time T, Equation 5 must be integrated,which means that Rc must be expressed in terms of V.

We denote the ratio of cake volume to filtrate volume as XQ. Hence, the cake volumeis x0V. An alternative expression for the cake volume is hc:A; where hc is the cakeheight in meters. Consequently:

xQV = hcA (6)

Hence, the thickness of the cake, uniformly distributed over the filter plate, is:

Vfa — y- /TV

c °A (1)

The filter cake resistance may be expressed as:

*c = VoT (8)

where r0= specific volumetric cake resistance (m"2).


As follows from Equation 8, r0 characterizes the resistance to liquid flow by a cakehaving a thickness of 1 m.

Substituting for Rc from Equation 8 into Equation 5, we obtain:

l^dV_ = = A/?A rfi p [ r Q x Q ( V / A ) + R f ] (9)

Filtrate volume, XQ can be expressed in terms of the ratio of the mass of solid particlessettled on the filter plate to the filtrate volume (xw) and instead of r0, a specific masscake resistance r w is used. That is, rw represents the resistance to flow created by auniformly distributed cake, in the amount of 1 kg/m2. Replacing units of volume bymass, the term r0 XQ in Equation 9 changes to rwxw.

Neglecting filter plate resistance (Rf = 0), and taking into account Equation 7, weobtain from Equation 3 the following expression:

At /u, = l N-sec/m2, hj. = 1 m and u = 1 m/sec, r0 = Ap. Thus, the specific cakeresistance equals the pressure difference required by the liquid phase (with a viscosityof 1 N-sec/m2) to be filtered at a linear velocity of 1 m/sec through a cake 1 m thick.This hypothetical pressure difference, however, is beyond a practical range. Forhighly compressible cakes, r0 can exceed 1012m2. Assuming V = 0 (at the start offiltration) where there is no cake over the filter plate. Equation 9 becomes:

*, ~~ &

At ju = 1 N-sec/m2 and u = 1 m/sec, Rf = Ap. This means that the filter plateresistance is equal to the pressure difference necessary for the liquid phase (withviscosity of 1 N-sec/m2) to pass through the filter plate at a rate of 1 m/sec. For manyfilter plates Rf is typically 10'° m"! .

For a constant pressure drop and temperature filtration process all the parameters inEquation 9, except V and T, are constant. Integrating Equation 9 over the limits of 0to V, from 0 to T, we obtain:

v %v f( r0Jt0 —- + Rf) dV = I bpAdi (12a)A J J

or

Cake Filtration and Filter Media Filtration

2A(12b)

Dividing both sides by jur0Xo/2A gives:

V2 + 2RfA

V = (13)

Equation 13 is the relationship between filtration time and filtrate volume. Theexpression is applicable to either incompressible or compressible cakes, since atconstant Ap, r0 and XQ are constant. If we assume a definite filtering apparatus andset up a constant temperature and filtration pressure, then the values of Rf, r0, p. andAp will be constant.

The terms in parentheses in Equation 13 are known as the "filtration constants", andare often lumped together as parameters K and C; where:

RA(15)

Hence, a simplified expression may be written to describe the filtration process asfollows:

(16)

Filtration constants K and C can be experimentally determined, from which thevolume of filtrate obtained over a specified time interval (for a certain filter, at thesame pressure and temperature) can be computed. If process parameters are changed,new constants K and C can be estimated from Equations 14 and 15.

Equation 16 may be further simplified by denoting TO as a constant that depends onKandC:

Liquid Filtration

K(17)

Substituting TO into Equation 16, the equation of filtration under constant pressureconditions is:

(V+Cf = (18)

Equation 18 defines a parabolic relationship between filtrate volume and time. Theexpression is valid for any type of cake (i.e., compressible and incompressible).

From a plot of V + C versus (T+IO), the filtration process may be represented by aparabola with its apex at the origin as illustrated in Figure 6. Moving the axes todistances C and TO provides the characteristic filtration curve for the system in termsof volume versus time. Because the parabola's apex is not located at the origin of thisnew system, it is clear why the filtration rate at the beginning of the process will havea finite value, which corresponds to actual practice.

Figure 6. Typical filtration curve.

Constants C and TO in Equation 18 have physical interpretations. They are basicallyequivalent to a fictitious layer of cake having equal resistance. The formation of thisfictitious cake follows the same parabolic relationship, where TO denotes the timerequired for the formation of this fictitious mass, and C is the volume of filtraterequired. Differentiating Equation 16 gives:


dV K2 ( V + C )

(19)

And rearranging in the form of a reciprocal relationship:

d^ _ 2V + 2CdV ~ K + K~

(20)

This form of the equation provides a linear relation as shown by the plot in Figure 7.The expression is that of a straight line having slope 2/K, with intercept C. Theexperimental determination of dt/dV is made simple by the functional form of thisexpression. Filtrate volumes V, and V2 should be measured for time intervals T, andT2, Then, according to Equation 16:

V -VV2 M V -VV v K

2C(V 2 -V, )

K

2CK

(21)

In examining the right side of this expression, we note that the quotient is equal to theinverse value of the rate at the moment of obtaining the filtrate volume, which is equalto the mean arithmetic value of volumes V, and V2:

V - VV2 M dV

Filtration constants C and K can be determined on the basis of several measurementsof filtrate volumes for different time intervals.

As follows from Equations 14 and 15, values of C and K depend on r0 (specificvolumetric cake resistance), which in turn depends on the pressure drop across thecake. This Ap, especially during the initial stages of filtration, undergoes changes inthe cake. When the cake is very thin, the main portion of the total pressure drop isexerted on the filter medium. As the cake becomes thicker, the pressure drop throughthe cake increases rapidly but then levels off to a constant value. Isobaric filtrationshows insignificant deviation from Equation 16. For approximate calculations, it is


Figure 7. Plot of Equation 20.

possible to neglect the resistance of the filter plate, provided the cake is not too thin.Then the filter plate resistance Rf = 0 in Equation 15, C= 0 (Equation 15) and TO = 0(Equation 17). Therefore, the simplified equation of filtration takes the following form:

V2 = K-c (23)

For thick cakes, Equation 23 gives results close to that of Equation 16.

Constant-Rate Filtration

When sludge is fed to a filter by a positive-displacement pump, the rate of filtration isnearly constant (i.e., dV/dt = constant). During constant-rate filtration, the pressureincreases with an increase in cake thickness. Therefore, the principal variables arepressure and filtrate volume, or pressure and filtration time. Equation 9 is the principaldesign relation, which may be integrated for a constant-rate process. The derivative,dV/dT, may be replaced simply by V/T:

Ap pR,At

(24)

The ratios in parentheses express the constant volume rate per unit filter area. Hence,Equation 24 is the relationship between time T and pressure drop Ap. For incompressiblecakes, r0 is constant and independent of pressure. For compressible cakes, therelationship between time and pressure at constant-rate filtration is:


(25)

Filtration experiments are typically conducted in pilot scale equipment and generallytests are conducted either at constant pressure or constant rate to determine axo, aswell as s and R f, for a given sludge and filter medium. Such tests provide empiricalinformation that will enable the time required tor the pressure drop to reach thedesired level for a specified set of operating conditions to be determined. In the initialstages of filtration, the filter medium has no cake. Furthermore, Ap is not zero, buthas a value that is a function of the resistance of the medium for a given flowrate. Thisinitial condition can be stated as:

/ v \(26)

For an incompressible cake (where s = 0), Equation 25 takes the form:

J?( —AT I f\ At

As noted earlier, for thick cakes, the resistance of the filter medium may be neglected.Hence, for Rt= 0, Equation 25 simplifies to:

* i . I v yV = paxQ\ --- I T (28)\ A T

An increase in pressure influences not only coefficient r0, but the cake's porosity aswell. Since the cake on the filter plate is compressed, residual liquid is squeezed out.Thus, for constant feed, the flowrate through the medium will not be stable, but willfluctuate with time.

The weight of dry solids in a cake is:

W = xQV (29)

where XQ = weight of solids in the cake per unit filtrate volume.

The concentration of solids in the feed sludge is expressed by weight fraction c. It isalso possible to evaluate experimentally the weight ratio of wet cake to its dry contentm. Hence, a unit weight of sludge contains me of wet cake. We denote y as the


specific weight of feed sludge. This quantity contains c amount of solids; hence, theratio of the mass of solids in the cake to the filtrate volume is:

cv(30)

1 - me

Thus, from the sludge concentration c and the weight of wet cake per kg of dry cakesolids m, XQ can be computed. If the suspension is dilute, then c is small; hence,product me is small. This means that XQ will be approximately equal to c. Accordingto Equations 29 and 30, the weight ratio of wet to dry cake will vary. Equation 30shows also that because XQ depends on the product me, at relatively moderatesuspension concentrations this effect will not be great and can, therefore, be neglected.However, when filtering concentrated sludges the above will play some role; that is,at constant feed, the filtrate changes with time.

Variable-Rate and -Pressure Filtration

The dynamics of variable-rate and -pressure filtrations can be illustrated by pressureprofiles that exist across the filter medium. Figure 8 shows the graphical representationof those profiles.

According to this plot, the compressed force in the cake section is:

P = Pi-Ps, (31)

where p, = pressure exerted on the sludge over the entire cakethickness

ps = static pressure over the same section of cake

p corresponds to the local specific cake resistance (rw)x. At the sludge-cake interfacepa = Pi and p = 0; and for the interface between the cake and filter plate ps, = ps( andP = Pi -p'st- p'a corresponds to the resistance of filter plate pf, and is expressed by:

Ap = pRfW (32)

where W = rate of filtration (mVm2-sec).

Note that Apf is constant during the operation.

Pressure p is also the driving force of the process. Therefore, starting from thegoverning filtration equations, the general expression for an infinitesimal incrementof solid particle weight in a cake of unit of area is xwdq (q = filtrate volume obtainedfrom 1 m2 filtering area, m3/m2). The responding increment dp may be expressed as:


F ILTRATE

Figure 8, Distribution of static pressure pst in liquid andp along the cake thickness and filterplate: /, // -boundaries between the cake and sludge at i:" and T'; III, IV-boundaries betweencake layers or cake and filter plate at T" and T'; V- boundary line between the cake and filterplate or free surface of filter plate; 1,3-curves Pst=f(hJ andp=f(hoc) at rf; 2, 4 -curvesPst=f(h0() andp=f(hoc) at T".

(33)

xw is not sensitive to changes in p.

In practice, an average value for xw can be assumed. Note that W is constant for anycross section of the cake. Hence, Equation 33 may he integrated over the cakethickness between the limits of p = 0 and p=p, -p'sl, from q = 0 to q = q:

dp(34)

Parameters q and W are variables when filtration conditions change. Coefficient (rw)is a function of pressure:


MARIONETTEBOTTLE

M E C H A N I C A LLOAD

VENT

TOP LOAD

PISTON

FIXED CELL BODY

POROUS MEDIUM

C A K E

POROUS MEDIUM

BOTTOM FLOATINGPISTON

T R A N S M I T T E DL O A D

Figure 9. Compression-permeability cell.

The exact relationship can be derived from experiments in a device called acompression-permeability cell which is illustrated in Figure 9. Once this relationshipis defined, the integral of the right side of Equation 34 may be evaluated analytically(or if the relationship is in the form of a curve, the evaluation may be madegraphically). The interrelation between W and P, is established by the pumpcharacteristics, which define q = f(W) in Equation 34. Filtration time may then bedetermined from the following definition:

(36)

Hence,

dq

W(37)


Constant-Pressure and -Rate Filtration

This mode of operation is achieved when a pure liquid is filtered through a cake ofconstant thickness at a constant pressure difference. Cake washing by displacementwhen the washing liquid is located over the cake may be considered to be filtration ofwashing liquid through a constant cake thickness at constant pressure and flowrate.The rate of washing is related to the rate of filtration during the last stages and maybe expressed by Equation 9, where Ap is the pressure at the final moment and V is thefiltrate volume obtained during filtration, regardless of the filtration method used (i.e.,constant-pressure or constant-rate operation). In the final stages, filtration usually isperformed under constant pressure. Then, the rate of this process may be calculatedfrom Equation 19. From filtration constants C and K, at constant pressure for a givensystem, the filtration rate for the last period is determined. If the washing liquid passesthrough the filter in the same pore paths as the sludge and filtrate, then the differencebetween the washing rate and filtration rate for this last period will be mostly due toa difference in the viscosities of the wash liquor and filtrate. Therefore, Equation 19is applicable using the viscosity of the washing liquid, p^. Denoting the rate offiltration in the last period as (dV/di), the washing rate is:

dV\ ( dW] u— I ' / T O \

j , (Jo)d-c )w \ dl} ^

Filter-Medium Filtration Formulas

Solid particles undergoing separation from the mother suspension may be capturedboth on the surface of the filter medium and within the inner pore passages. Thisphenomenon is typical in the separation of low-concentration suspensions, where thesuspension consists of viscous liquids such as sugar liquors, textile solutions ortransformer oils, with fine particles dispersed throughout. The penetration of solidparticles into the filter medium increases the flow resistance until eventually thefiltration cycle can no longer proceed at practical throughputs and the medium mustbe replaced. In this section standard filtration formulas are provided along withdiscussions aimed at providing a working knowledge of the filter-medium filtrationprocess.

Constant-Pressure-Drop Filtration

Constant-pressure drop filtration can result in saturation or blockage of the filtermedium. The network of pores within the filter medium can become blocked becauseof one or a combination of the following situations:

1. Pores may become blocked by the lodging of single particles in the pore passage.2. Gradual blockage can occur due to the accumulation of many particles in pore passages.3. Blockage may occur during intermediate-type filtration.


Proper filter medium selection is based on understanding these mechanisms andanalyzing the impact each has on the filtration process.

In the case of single-particle blockage, we first consider a i m surface of filtermedium containing Np number of pores. The average pore radius and length are rp

and S. p, respectively. For laminar flow, the Hagen-Poiseuille equation may be appliedto calculate the volume of filtrate V passing through a pore in a unit of time:

V1 -

Consequently, the initial filtration rate per unit area of filtration is:

W = V'N

Consider 1m3 of suspension containing n number of suspended particles. If thesuspension concentration is low, we may assume the volume of suspension and filtrateto be the same. Hence, after recovering a volume q of filtrate, the number of blockedpores will be nq, and the number unblocked will be (Np - nq). Then the rate offiltration is:

(41)

or

W = W. ~kq (49)III " V"T~/

where

V'n (43)

k' is a constant having units of sec"1. It characterizes the decrease in intensity of thefiltration rate as a function of the filtrate volume. For constant V, this decreasedepends only on the particle number n per unit volume of suspension. The totalresistance R may be characterized by the reciprocal of the filtration rate. Thus, W inEquation 42 may be replaced by 1/R (sec/m). Taking the derivative of the modifiedversion of Equation 42 with respect to q, we obtain:


dR k'

Tq = 7F~^7 (44>

Comparison with Equation 42 reveals:

<!!*. = _£.^ W2

or

dR = k'R2

dq

Equation 46 states that when complete pore blockage occurs, the intensity of theincrease in the total resistance with increasing filtrate volume is proportional to thesquare of the flow resistance.

In the case of multiparticle blockage, as the suspension flows through the medium,the capillary walls of the pores are gradually covered by a uniform layer of particles.This particle layer continues to build up due to mechanical impaction, particleinterception and physical adsorption of particles. As the process continues, theavailable flow area of the pores decreases. Denoting x0 as the ratio of accumulatedcake on the inside pore walls to the volume of filtrate recovered, and applying theHagen-Poiseuille equation, the rate of filtration (per unit area of filter medium) at thestart of the process is:

where

When the average pore radius decreases to r, the rate of filtration becomes:

W = BNr* (49)

For a finite filtrate quantity, dq the amount of cake inside the pores is x0dq, and thecake thickness is dr. That is:


xodq = -N2nrtdr (50)

Note that the negative sign indicates that as q increases, the pore radius r decreases.Integrating this expression over the limits of 0 to q, for rp to r we obtain:

P P(51)

And from Equations 47 and 49, we may define the pore radii as follows:

W.

BN,

1/2

(52)P/

or simply:

(53)

Substituting these quantities into Equation 51 and simplifying terms, we obtain:

12

where

xo B1/2

(55)

It is convenient to define the following constant:

2Cvl/2 (56)

From which Equation 54 may be restated as:

W= W.}(l-l/2Kq)2(57)


Since W= 1/R, we may write:

W.m(\ - \l2Kq)2

The derivative of this expression with respect to q is:

dR = k

dq W.n(l- \l2Kq)3

On some rearranging of terms, we obtain:

(60)

or

where

K" - K(Wit)m (62)

Equation 61 states that the intensity of increase in total resistance with increasingfiltrate amount is proportional to resistance to the 3/2 power. In this case, the totalresistance increases less sensitively than in the case of total pore blockage.

As follows from Equations 56 and 62:

K" = 2C (63)

Substituting Equation 55 for C and using Equation 48 for B, the above expressionbecomes:

K" = 2(W,) 1/2 (64)

Note that for constant Win, parameter K" is proportional to the ratio of the settledvolume of cake in the pores to the filtrate volume obtained, and is inverselyproportional to total pore volume for a unit area of filter medium.


Replacing W by dq/dt in Equation 57, we obtain:

i - —Kq I dq (65)W. I 2

Integration of this equation over the limits from 0 to T for 0 to q we obtain:

2q~W.n(2-Kq)

and on simplification:

K _ T _ 1T = (67)

Equation 67 may be used to evaluate constants K (m"1) and Win.

Finally, for the case of intermediate filtration, the intensity of increase in totalresistance with increasing filtrate volume is less than that occurring in the case ofgradual pore blocking, but greater than that occurring with cake filtration. It may beassumed that the intensity of increase in total resistance is directly proportional to thisresistance:

* = K'"Rdq

Integration of this expression between the limits of 0 to q, from Rf to R gives:

Rf

Substituting 1/W for R and l/Win for Rf, the last expression becomes:

Win K,,,

~W ~ e

or

W = W. e ~K""(I


Substituting dq/di for W in Equation 71 and integrating over the limits of 0 to tbetween 0 and q we obtain:

1 eK'"q - 1W. K'"in

Hence,

If ' ' T ~ __-_.—- -~ / --7 o \

u7~~ W ' •'

Accounting for Equation 70, the final form of this expression becomes:

_L = JL + #'"Tw w.

Filtration Mechanisms

To compare the different mechanisms of filtration, the governing equation of filtrationmust be rearranged. The starting expression is:

dV _ A/?

Adi p[rox0(V/A) + Rr]

Replacing V by q, and denoting the actual filtration rate (dq/di) as W, the governingfiltration equation may be rewritten for a unit area of filtration as follows:

W= ^

At the initial moment when q = 0, the filtration rate is

'" ~ ,. v (77)

From Equations 76 and 77 we have:

W:

Wl+K'"Winq


where

The numerator of Equation 79 characterizes the cake resistance. The denominatorcontains information on the driving force of the operation. Constant K'" (sec/m )characterizes tile intensity at which the filtration rate decreases as a function ofincreasing filtrate volume.

Substituting 1/R for W in Equation 78 and taking the derivative with respect to q, weobtain:

dR(80)

The expression states that the intensity of increase in total resistance for cake filtrationis constant with increasing filtrate volume. Replacing W by dq/di in Equation 78 andintegrating over the limits of 0 to q between 0 and T we obtain:

K'" T 1

Note that this expression reduces to Equation 74 on substituting expressions for Win

(Equation 77) and K'" (Equation 79).

Examination of Equations 46, 61, 68 and 80 reveals that the intensity of increase intotal resistance with increasing filtrate volume decreases as the filtration processproceeds from total to gradual pore blocking, to intermediate type filtration and finallyto cake filtration. Total resistance consists of a portion contributed by the filtermedium plus any additional resistance. The source of the additional resistance isestablished by the type of filtration. For total pore blockage filtration, it is establishedby solids plugging the pores; during gradual pore blockage filtration, by solid particlesretained in pores; and during cake filtration, by particles retained on the surface of thefilter medium.

The governing equations (Equations 42, 67, 74 and 81) describing the filtrationmechanisms are expressed as linear relationships with parameters convenientlygrouped into constants that are functions of the specific operating conditions. Theexact form of the linear functional relationships depends on the filtration mechanism.Table 1 lists the coordinate systems that will provide linear plots of filtration datadepending on the controlling mechanism.

In evaluating the process mechanism (assuming that one dominates) filtration data maybe massaged graphically to ascertain the most appropriate linear fit and, hence, the


type of filtration mechanism controlling the process, according to Table 1. If, forexample, a linear regression of the filtration data shows that q = f(t/q) is the best linearcorrelation, then cake filtration is the controlling mechanism. The four basic equationsare by no means the only relationships that describe the filtration mechanisms.

Table 1. Coordinates for representing linear filtration relationships.

Type of Filtration

With Total Pore Blocking

With Gradual Pore Blocking

Intermediate

Cake

Equation

42

67

74

81

Coordinates

q vs W

T VS T/q

T VS ]/W

q vs t/q

All the mechanisms of filtration encountered in practice have the functional form:

dR VT3b- - KR (82)

where b typically varies between 0 and 2.

Constant Rate Filtration

Filtration with gradual pore blocking is most frequently encountered in industrialpractice. This process is typically studied under the operating mode of constant rate.

We shall assume a unit area of medium which has Np pores, whose average radius andlength are rp and 0p, respectively. The pore walls have a uniform layer of particles thatbuild up with time and decrease the pore passage flow area. Filtration must beperformed in this case with an increasing pressure difference to compensate for therise in flow resistance due to pore blockage. If the pores are blocked by acompressible cake, a gradual decrease in porosity occurs, accompanied by an increasein the specific resistance of the deposited particles and a decrease in the ratio of cake-to-filtrate volumes. The influence of particle compressibility on the controllingmechanism may be neglected. The reason for this is that the liquid phase primarilyflows through the available flow area in the pores, bypassing deposited solids. Thus,the ratio of cake volume to filtrate volume (x0) is not sensitive to the pressuredifference even for highly compressible cakes.

From the Hagen-Poiseuille relation (Equation 39) replacing Win in Equation 40 withconstant filtration rate W and substituting APin for constant pressure drop AP weobtain:

W = B'WinN/p (83)


where

The mass of particles deposited on the pore walls will be x0dq, and the thickness ofthis particle layer in each pore is dr. Hence

(85)xadq = -N 2nr( di

Integration over the limits of 0, q from rp to r yields

N uH ,(86)

Radii rp and r are defined by Equations 83 and 85, respectively, from which we obtainthe following expressions:

q = W wB'LpN

(87)

or

™p

B'

1/2

A/?

1/2

Since q = WT, Equation 88 may be stated in a reduced form as:

1/2 / , \ 1/2CT =

1(89)

where

JV

1/2

A plot of Equation 89 on the coordinate of T vs (l/Apin)'/2 - (l/Apj/2 results in a

straight line, passing through the origin, with a slope equal to C. Thus, if experimentaldata correlate using such coordinates, the process is gradual pore blocking. Note thatat T= 0, Ap = Apin, which is in agreement with typical process observations.


The filtration time corresponding to total pore blockage, when Ap-+°° may beestimated from:

1T ™

C1 1/2

(91)

To express the relationship between AP and T more directly. Equation 89 is restatedin the form:

(A-Cif

where

1/2

It is important to note that pore blocking occurs when suspensions have the followingcharacteristics:

1. relatively small particles;2. high viscosity; and3. low solids concentrations.

Both particle size and the liquid viscosity affect the rate of particle settling. The rateof settling due to gravitational force decreases with decreasing particle size andincreasing viscosity. The process mechanisms are sensitive to the relative rates offiltration and gravity sedimentation.

Examination of the manner in which particles accumulate onto a horizontal filtermedium assists in understanding the influences that the particle settling velocity andparticle concentration have on the controlling mechanisms. The separation processthrough a cross section of filter medium is illustrated in Figure 10. "Dead zones" existon the filter medium surface between adjacent pores. In these zones, particle settlingonto the medium surface prevails. After sufficient particle accumulation, solids beginto move under the influence of fluid jets in the direction of pore entrances. This leadsto favorable conditions for bridging. The conditions for bridge formation become morefavorable as the ratio of particle settling to filtration rate increases.

An increase in the suspension's particle concentration also enhances accumulation in"dead zones" with subsequent bridging. Hence, both high particle settling velocityincreases and higher solids concentrations create favorable conditions for cakefiltration. In contrast, low settling velocity and concentration results in favorableconditions for gradual pore blocking.


The transition from pore-blocked filtration to more favorable cake filtration cantherefore be achieved with a suspension of low settling particles by initially feedingit to the filter medium at a low rate for a time period sufficient to allow surfaceaccumulation. This is essentially the practice that is performed with filter aids.

Figure 10, Suspension flow downward onto a filter medium. An initial accumulation of solidsoccurs around the pott entrance followed by particle bridging.

Suggested Readings

1. Cheremisinoff, P.N., Wastewater Treatment Pocket Handbook, PudvanPublishing, Northbrook, IL, 1987

2. Cheremisinoff, P.N., Pocket Handbook for Solid-Liquid Separations,Publishing Co., Houston, TX, 1984

3. Noyes, R., Unit Operations in Environmental Engineering, NoyesPublishers, NJ, 1994

4. Kirkpatrick, J. , Mathematics for Water and Wastewater Treatment PlantOperators, Ann Arbor Science Pub., Ann Arbor, MI, 1976

Gulf


5. Environmental Law Institute, Clean Water Deskbook, Environmental LawReports, Washington, DC, 1988.

Nomenclature

A = area (m2)B, B' = empirical parametersC = filtration parameterc = concentration (kg/m)Fs = force (N)hc = cake height (m)K,K",K'"

= filtration constantsL = cake thickness (m)!p = pore length (m)n = number of suspended particlesNp = number of poresp = pressure (N/m2)q = filtrate volumer per unit area of filter (m3/m3) or filtrate

volume (m3)r = specific resistance (m"1)r0 = specific volumetric cake resistance (kg/m2)rp = pore radius (m)rw = specific mass cake resistance (kg/m2)R = resistance (m/sec)Rc,Rt = cake and filter resistances, respectively (m"1)u = average velocity (m/sec)V = filtrate volume (m3)W = mass of dry solids (kg), or rate of filtration (m3/m2-sec)XQ = ratio of cake to filtrate volume.

Greek Symbols

e = porosityyt, = viscosity (P)II = ratio of filtration rate to gravity settingT = time (sec)TO = time constant (sec)

Introduction

Filtration equipment is commercially available in a wide range. Proper selection mustbe based on detailed information of the slurry to be handled, cake properties,anticipated capacities and process operating conditions. One may then select thepreferred operational mode (batch, semibatch or continuous), and choose a particularsystem on the above considerations and economic constraints.

Continuous filters are comprised of essentially a large number of elemental surfaces,on which different operations are performed. These operations performed in series aresolids separation and cake formation, cake washing, cake dewatering and drying, cakeremoval, and filter media washing. The specific equipment used can be classified intotwo groups: (1) stationary components (which are the supporting devices such as thesuspension vessel); and (2) scraping mechanisms and movable devices (which can bethe filter medium, depending on the design).

Either continuous or batch filters can be employed in cake filtration. In filter-mediumfiltration, however, where particulates are retained within the framework of the filtermedium, batch systems are the most common. Batch filters may be operated m anyfiltration regime, whereas continuous filters are most often operated under constantpressure.

In an attempt to organize the almost overwhelming number of different types offiltration equipment, two classification schemes have evolved for continuousoperations. The first scheme is based on operating pressure differentials and isprovided in Table 1. The second scheme is based on the relative difference betweengravity force and filtrate motion. Three orientations are possible: forces acting in

88

4

INDUSTRIAL FILTRATIONEQUIPMENT

Industrial Filtration Equipment 8

opposite direction (countercurrent), forces acting in the same direction (cocurrent),and forces acting normal or perpendicular to each other (cross-mode operation).

Table 1. Pressure differential scheme for filtration operating classification.

Source

Hydrostatic pressure of the suspension layer to be separated

Action of compressors

Action of pumps

Pressure Differential (N/mz)

Usually no more than 5

5-9

Up to 50 and higher

Because the influence of gravity is so important to most filtration operations,the second classification basis is used in this chapter. The operating principles andimportant features of filtration equipment are described in this chapter with the intentof providing the reader a background in the versatility and selection options available.

Rotary Drum Filters

Rotary-drum filters fall into the category of the countercurrent mode type operation;and are either vacuum- or pressure-operated. They are most frequently operated asvacuum filters. Although operated under pressure, they are rarely subjected toexcessive pumping pressures. The principal advantage of these filters is the continuityof their operation. Total filtration cycles are limited to narrow time intervals. Thisnecessitates maintaining nearly constant slurry properties. Changing slurry propertiescan lead to wide variations in the required times for completing individual operationsof the filtration process.

For separating low-concentration, stratified suspensions, rotary drum filters arenormally specified for a submergence rate of 50 %. Such slurries require only mildmixing to prevent particle settling. These filters are less useful in handlingpolydispersions containing particles with wide size ranges. Fouling by small solids isa frequent problem in these latter cases.

Drum Vacuum Filters with External Filtering Surfaces

These filters are characterized by the rate at which the drum is immersed in thesuspension. These are perhaps the most widely employed countercurrent operatedfilters in industry, an example of which is shown in Figure 1. As shown, the designconsists of hollow drum-1, with a slotted face, the outer periphery of which containsshallow tray-shaped compartments-2. The filter cloth is supported by a grid or a heavyscreen, which lies over these compartments. The drum rotates on a shaft with one endconnected to the drive-3, and die otiier to a hollow trunnion adjoining to an automaticvalve. The drum surface is partially immersed in the suspension contained in vesscl-6.The cake that is formed on the outer surface of the drum is removed by scrapcr-7 asthe drum rotates.

Figure 1. Rotary-drum vacuum filter with external filtration surface: 1 - hollow drum; 2 -filtration compartments; 3 - drive; 4 - hollow trunnion; 5 - automatic valve; 6 - tank farsuspension; 7 - knife for cake scraping.

Figure 2 illustrates a longitudinal view of the system. Compartments-2 of the drum-1 areconnected through the pipe-3, passing through the hollow trunnion-4 of shaft-5, withthe automatic valve-6. A stirring device-7 is mounted under the drum to preventparticle settling.

A diagrammatic cross section of the filter is illustrated in Figure 3. As the drumrotates clockwise, each compartment is connected by the pipe-2 with differentchambers of immobile parts of the automatic valve-4 and passes in series through thefollowing operating zones: filtration, first dewatering, washing, second dewatering,cake removal and cloth regeneration.

In the filtration zone, the compartment contacts the suspension in the tank-11, and isconnected to the pipe-10 hooked up to a vacuum source. The filtrate is dischargedthrough the pipe and space in the collector and the cake forms on the compartment'ssurface. In the first dewatering zone the cake comes in contact with the atmosphere,and the compartment is connected to the space-10. Because of the vacuum, air isdrawn through the cake, and for maximum filtrate recovery, the compartment remainsconnected to a collection port on the automatic valve.

In the washing zone the cake is washed by the nozzles (or wash headers)-^. Thecompartment is connected through the port-6, which is also tied into a vacuum source.The wash liquor is removed in the other collector.

In the second dewatering zone, the cake is also in contact with the atmosphere, andthe compartment is connected with the port-6. Consequently, the washing liquid isdisplaced from the cake pores and delivered lo the collector. To avoid cake crackingduring washing and dewatering, an endless belt-7 is provided, which moves over a set

Industrial Filtration Equipment

Figure 2. Longitudinal section of a rotary-drum vacuum filter with an external fiitratisurface: 1 - drum; 2 - compartment; 3 - connecting pipe; 4 - hollow trunnion; 5 - shaft; <automatic valve; 7 - stirring device.

of guide rollers. In the discharge zone, the compartment is connected with the port-5,which is supplied by a compressed air source. This reversal of pressure or "blow"loosens the cake from the filter medium, whence it is removed by a scraper or doctorblacle-3.

In the regeneration zone, compressed air is blown through the cloth; the air enters thecompartment through the pipe from the port-13. The automatic valve serves to activatethe filtering, washing and cake discharge function of the filter sections. It providesseparate outlets for the filtrate and wash liquid, and a connection by which thecompressed air blowback can be applied.

Cocurrent Filters

Cocurrent or top-feed filters employ flat and cylindrical filtering media. In flatdesigns, the angle between the directions of gravity force and filtrate motion is 0°, but

n Liquid Filtration

SECONDDEWAT£fttNG

5 CA#£REMOVAL

4

. WASHFIRST LI QUO

FILTRATIFILTRATION

Figure 3, Diagrammatic cross section of rotary-drum vacuum filter: 1 - drum; 2 - connectingpipe; 3 -scraper; 4 -automatic valve; 5, 13 -chambers of automatic valve connected with asource of compressed air; 6, 10 -chambers of automatic valve connected with a source ofvacuum; 7 -endless belt; 8 - wash header; 9 - guiding roll; 11 - tank for suspension; 12 -stirring device.

may vary to larger angles. In this class of filter, the directions of gravity force actionand filtrate flow are the same. Filter designs in this class are quite different from diecounterflow type. They include sophisticated rotary-drum filters, continuous-beltfilters, Nutsch batch filters and filter presses with horizontal chambers. These filtersare most often used for separating stratified slurries. Separation by filters of the firstsubgroup is based on intensive slurry mixing by agitators. It is especially advisable touse these filters for the separation of polydispersed systems. In this case, the cakeformed is properly stratified with large particles adjacent to the filter medium.

Flat filtering surfaces form a cake of uniform thickness and homogeneous structureat any horizontal plane. This permits highly effective washing.

Internal Rotary-Drum Filters

An example of an internal rotary-drum filter is illustrated in Figure 4. The filtermedium is contained on the inner periphery. This design is ideal for rapid-settlingslurries that do not require a high degree of washing.

Tankless filters of this design consist of multiple-compartment drum vacuum filters.One end is closed and contains an automatic valve with pipe connections to individual

Industrial Filtration Equipment 9 3

compartments. The other end is open for the feed entrance. The drum is supported ona tire with rigid rollers to effect cake removal. The drum is driven by a motor andspeed-reducer connected to a riding roll shaft.

The feed slurry is discharged to the bottom of the inside of the drum from thedistributor and is maintained as a pool by a baffle ring located around the open endand the closed portion of the outer end. As the drum revolves, the compartmentssuccessively pass through the slurry pool, where a vacuum is applied as eachcompartment becomes submerged. Slurry discharge is accomplished at the top centerwhere the vacuum is cut off and gravity (usually assisted by blowback) allows thesolids to drop off onto a trough. From mere, a screw or belt conveyor removes thesolids from the drum. This filter is capable of handling heavy, quick-settling materials.

Nutsch Filters

Nutsch filters are one design type with a flat filtering plate. This configurationbasically consists of a large false-bottomed tank with a loose filter medium. Olderdesigns employ sand or other loose, inert materials as die filtering medium, and arestill employed in water clarification operations. In vacuum filtration, these false-bottom tanks are of the same general design as the vessels employed for gravityfiltration. They are, however, less widely used, being confined for the most part torather small units, particularly for acid work. Greater strength and more carefulconstruction are necessary to withstand the higher pressure differentials of vacuumover gravity. This naturally increases construction costs. However, when high filteringcapacity or rapid handling is required with the use of vacuum, the advantages maymore than offset higher costs.

Construction of the vacuum false-bottom tank is relatively simple; a single vessel isdivided into two chambers by a perforated section. The upper chamber operates underatmospheric pressure and retains the unfiltered slurry. The perforated false bottomsupports the filter medium. The lower chamber is designed for negative pressure, andto hold the filtrate.

Nutsch filters are capable of providing frequent and uniform washings. A type ofcontinuous filter that essentially consists of a series of Nutsch filters is the rotating-tray horizontal filter.

Horizontal Rotary Filters

An example of a horizontal rotary filter is illustrated in Figure 5. These machines arewell suited to filtering quick-draining crystalline solids. Due to its horizontal surface,solids are prevented from falling off or from being washed off by the wash water. Assuch, an unusuaJly heavy layer of solids can be tolerated. The basic design consists of"a circular horizontal table that rotates about a center axis. The table is comprised of anumber of hollow pie-shaped segments with perforated or woven metal tops. Each of thesections is covered with a suitable filter medium and is connected to a central valvemechanism that appropriately times the removal of filtrate and wash liquids and the

Figure 4. Sectiua view of an interiar medium rolory-drum v a ~ ~ u m f&er.


dewatering of the cake during each revolution. Each segment receives the slurry insuccession. Wash liquor is sprayed onto each section in two applications. Then the cakeis dewatered by passing dry air through it. Finally, the cake is scooped off of the surfaceby a discharge scroll.

Belt Filters

Belt filters consist of a series of Nutsch filters moving along a closed path. Nutschfilters are connected as a long chain so that the longitudinal edge of each unit lias theshape of a baffle plate overlapping the edge of the neighboring unit. Each unit isdisplaced by driving and tensioning drums.

Nutsch filters are equipped with supporting perforated partitions covered with thefiltering cloth. The washed cake is removed by turning each unit over. Sometimes ashaker mechanism is included to ensure more complete cake removal.

In contrast, a belt filter consists of an endless supporting perforated rubber beltcovered with the filtering cloth. The basic design is illustrated in Figure 6. Supportingand filtering partitions-1 are displaced by driving drum-2 and maintained in a stretchedcondition by tensioning the drum-3, which rotates due to friction against the rubberbelt. Belt edges (at the upper part of their path) slide over two parallel horizontal guideplanks. The elongated chamber-4 is located between the guide planks. The chamberin the upper part has grids with flanges adjoining the lower surface of the rubber belt.The region under the belt is connected by nozzles-5 to the filtrate collector-6, whichis attached to a vacuum source. The chamber and collector are divided into sectionsfrom which filtrate and washing liquid may be discharged. The sludge is fed by thetrough-7. The cake is removed from the drum-2 by gravity or blowing, or sometimesit is washed off by liquid from the distributor nozzle-8. The washing liquid is suppliedfrom the tank-9, which can move along the filtering partition, it can be washed duringthe belt's motion along the lower path.

The filtering partition, illustrated in Figure 7, consists of riffled rubber belt-1 withslots-2, grooves-3 and filter cloth-4, which is fixed in a set of grooves by cords-5.Slots-2, through which the filtrate passes, are located over the grids of the elongatedchamber. The edges of the rubber belt are bent upward by guides forming a gutter ondie upper path of the belt.

The velocity of the filtering partition depends on die physical properties of die sludgeand the filter length. The cake thickness may range from 1 to 25 mm. The advantagesof belt filters are tiieir simplicity in design compared to filters with automatic valves,and die abilities to provide countercurrent cake waslling and removal of thin layers ofcake. Their disadvantages include large area requirements, inefficient use of die totalavailable filter area, and poor washing at die belt edges.

3

Liquid Filtration

Figure 7. Filtering partition for a belt fitter: 1 - rubber belt; 2 - slots; 3 - grooves; 4 -filteringcloth; 5 - cord; 6 - edges of rubber belt.

Cross Mode Filters

Filters of this group have a vertical flat or cylindrical filtering partition. In this case,filtrate may move inside the channels of the filtering elements along the surface of thefiltering partition downward under gravity force action, or rise along this partitionupward under the action of a pressure differential. In the separation of heterogeneoussuspensions, nonuniform cake formation along the height can occur because largerparticles tend to settle out first. This often results in poor cake washing due todifferent specific resistances over the partition height. The cake may creep down alongthe partition due to gravity; mis is almost inevitable in the absence of a pressuregradient across the filtering partition. The vertical filtering partition makes these filtersespecially aseful as thickeners, since it is convenient to remove cake by reverse filtrateflow.

Filter Presses

The common filter press is the plate-and-frame design, consisting of a metal framemade up of two end supports rigidly held together by two horizontal steel bars.Varying numbers of flat plates containing cloth filter media are positioned on thesebars. The number of plates depends on the desired capacity and cake thickness. Theplates are clamped together so that their frames are flush against each other, forminga series of hollow chambers. The faces of the plates are grooved: either pyramided orribbed. The entire plate is covered with cloth, which forms the filtering surface. Thefilter cloth has holes that register with the connections on the plates and frames, so thatwhen the press is assembled these openings form a continuous channel over the entirelength of the press and register with the corresponding connections on the fixed head.The channel opens only into the interior of the frames and has no openings on theplates. At the bottom of the plates, holes are cored so that they connect the faces ofdie plates to the outlet cocks. As the filterable slurry is pumped through the feedchannel, it first fills all of the frames. As the feed pump continues to supply fluid andbuildup pressure, the filtrate passes through the cloth, runs down the face of the plateand passes out through the discharge cock. When the press is full, it is opened anddumped. Cake cannot be washed in these units and is therefore discharged containing


a certain amount of filtrate with whatever valuable or undesirable material it maycontain.

Each plate discharges a visible stream of filtrate into the collecting launder. Hence,if any cloth breaks or runs cloudy, that plate can be shut off without spoiling the entirebatch.

If the solids are to be recovered, the cake is usually washed. In this case, the filter hasa separate wash feed line, and the plates consist of washing and nonwashing typesarranged alternately, starting with the head plate as the first nonwashing plate. Thewash liquor moves down the channels along the side of each washing plate, and movesacross the filter cake to the opposite plate and drains toward the outlet. This isillustrated in Figure 8.

t—WASHING PLATE-—*

WASHWATERIN

HEAD

CLOTHNON-WASHING

PLATE FRAME

(MM

Figure 8, Illustrates wash water outlets on a filter press.

To simplify assembly, the nonwashing plates are maked with one button and thewashing plates with three buttons. The frames carry two buttons.

In open-delivery filters the cocks on the one-button plates remain open and those onthe three-button plates are closed. In closed-delivery filters a separate wash outletconduit is provided. Figure 9 illustrates the basic design of a frame, a nonwashingplate and a washing plate. These plates and frames in closed-delivery filters are shownin Figure 10.


In terms of initial investment and floor area requirements, plate-and-frame filters areinexpensive in comparison to other filters. They can be operated at full capacity (allframes in use) or at reduced capacity by blanking off some of the frames by dummyplates. They can deliver reasonably well washed and relatively dry cakes. However,the combination of labor charges for removing the cakes and fixed charges fordowntime may constitute a high percentage of the total cost per operating cycle.

PLATE,NONWASHING

FEED PORT

FRAME

WASH PORT

PLATE,WASHING

Figure 9. Plates and frame of an open-delivery through-washing filter.

FEED

PLATE, NONWASHING

Figure 10. Plates and frame of a closed-delivery through-washing filter.

Leaf Filters

Leaf filters are similar to plate-and-frame filters in that a cake is deposited on eachside of a leaf (refer to Figure 11), and the filtrate flows to the outlet in the channelsprovided by a coarse drainage screen in the leaf between the cakes. The leaves areimmersed in the sludge when filtering, and in the wash liquid when washing.Therefore, the leaf assembly may be enclosed in a shell, as in pressure filtration, orsimply immersed in sludge contained in an open tank, as in vacuum filtration.

In operating a pressure leaf filter, the sludge is fed under pressure from the bottomand equally distributed. The clear filtrate from each leaf is collected in a commonmanifold and carried away. In filters with an external filtrate manifold (refer to thesketch in Figure 12), die filtrate from each leaf is visible through a respectivesightglass. This is not possible when the leaves are mounted on a hollow shaft thatserves as an internal filtrate collecting manifold. The filter cakes are built on each side


of the leaves and filtration is continued until the required cake thickness is achieved.For washing, the excess sludge is usually drained, simultaneously admittingcompressed air (3-5 Ib pressure), which serves mainly to prevent the cake frompeeling off the leaves.

Guard

Drainagescreen

Filtercloth

Frame

Figure 11. Sectional view of a filter leaf showing construction and approximate location ofcake.

Disk Filters

Disk filters consist of a number of concentric disks counted on a horizontal rotaryshaft and operate on the same principle as rotary-drum vacuum filters. The basicdesign is illustrated in Figure 13. The disks are formed by using V-shaped hollowsectors assembled radially around the shaft. Each sector is covered with filter clothand has an audet nipple connected to a manifold extending along the length of me shaftand leading to a port on the filter valve. Each row of sectors is connected to a separatemanifold. The sludge level in the tank should provide complete submergence to thelowest sector of the disks.


FILTRATEORASH

SIGHT GLASS

FEED ORWASH

DISCHARGE

Figure 12, Sweetland pressure filter.

Figure 13. Rotary-disk vacuum filter: 1 - section; 2-filtering disks; 3 - automatic valve; 4 -manifold for vacuum and filtrate discharge; 5 - piping for compressed air; 6 - doctor's knivesfor cake removal.


Compared to drum vacuum filters, the greatest advantage of the disk filter is that, forthe same filtering area, it occupies considerably less floor space. However, becauseof vertical filtering surfaces, cake washing is not as efficient as when a drum filter isused. The disk filter is ideal when the cake is not washed and floor space is at apremium.

Cartridge Filters

Cartridge filters are normally operated in the countercurrent mode, but because oftheir extensive use throughout the chemical and process industries in applicationsranging from laboratory-scale to commercial operations for flows extending to anexcess of 5000 gal/min, a separate discussion is warranted.

Typical applications include:

remove undispersed solids;remove precipitated solids;protect catalyst beds;protect instruments;remove DE filter carryover;keep spray nozzles open;filter recirculating water;remove particles from coalings;filter cooling tower water;remove char particles;filter condensate;filter bottle and can wash water;filter poultry and meal wash water;

remove oversize particles fromslurries;clean electrolytic solutions;filter waste oil for reuse;remove plastic fines from water;filter scrubber water;filter boiler feed water;filter pump seal water;protect glue applicators;protect reverse osmosis systems;protect chiller and air-conditioners;andremove pulp from juices.

Industrial applications of cartridge filters are as follows. The chemical industry usescartridge filters to handle:

acetic acid,calcium carbonate,brine,ethylene glycol,herbicides,hydrochloric acid,

latexes,resins,polymers,sulfuric acid,cooling tower water, andpelletizer water.

The food industry uses cartridge filters to handle:

com syrup,dextrose,lard,jelly,juices,milk sugar,edible oils,

tea liquor,city and well water,extracts,chocolates,soybean concentrate, andpeanut butter.


The paper industry uses cartridge filters to handle:

pigtnented coatings,white water,freshwater,size,starch,TiO, slurry.

dyes,cooling water,pump seal water,decker shower water,wet end additives, andclay slurry.

mill water,

The petroleum industry uses cartridge filters to handle:

a mine,feedstocks,reduced crudes,naphtha,fuel oil,motor oil,

hydraulic oil,injection fluids,completion fluids,cooling tower water,pump seal water, andsynthetic lubricants.

Miscellaneous industrial uses of cartridge filters include:

adhesives,resins,solvents,paints,shampoo,dyes,

cooling water,Pharmaceuticals,beverages,toothpaste,liquors, andbeer.

Table 2 provides a summary typical filtration ranges encountered throughout industry.Early designs, still widely used, consist of a series of thin metal disks that are 3-10 in.in diameter, set in a vertical stack with very narrow uniform spaces between them.The disks are supported on a vertical hollow shaft, and fit into a closed cylindricalcasing. Liquid is fed to die casing under pressure, whence it flows inward between thedisks to openings in the central shaft and out through the top of the casing. Solids arecaptured between the disks and remain in the filter. Since most of the solids areremoved at the periphery of the disks, the unit is referred to as an edge filter. Theaccumulated solids are periodically removed from the cartridge.

As with any filter, careful media selection is critical. Media that are too coarse, forexample, will not provide the needed protection. However, specifying finer mediathan necessary can add substantially to both equipment and operating costs. Factorsto be considered in media selection include solids content, type of contaminant,particle size and shape, amount of contaminant to be removed, viscosity,corrosiveness, abrasiveness, adhesive qualities, liquid temperature, and requiredflowrate. Typical filter media are wire mesh (typically 10-700 mesh), fabric (30 mesh- 1 /mi), slotted screens (10 mesh - 25 jttm) and perforated stainless steel screens (10-30 mesh). Table 3 provides typical particle retention sizes for different media.


Table 2. Typical filtration ranges encountered in industry applications.

105

Industry and Liquid

Chemical IndustryAlumBrineEthyl AlcoholFerric ChlorideHerbicides/PesticidesHydrochloric AcidMineral OilNitric AcidPhosphoric AcidSodium HydroxideSodium HypochloriteSodium SulfateSulfurie AcidSynthetic Oils

Drugs and CosmeticsAcetic AcidAerosolBath OilCitric AcidGlycerineLipstickShampooSoapSuntan LotionTallowToothpaste

Food and BeverageApple JuiceBeerBrineChocolateCorn SyrupFructose SyrupFruit Juices with PulpJellyLardLemon EffluentLiquorsVegetable OilWash Water

Petroleum IndustryAtmospheric Reduced Crude Completion FluidsCompletion FluidsDBADistil lated OilDecant OilDiesel FuelGas OilGasolineHydrocarbon Wax

Typical Filtration Range

60 mesh-60 pm100-400 mesh5- 10pm30-250 mesh100-700 mesh100 mesh to 5-10 pm400 mesh40 mesh to 5-10 pm100 mesh to 5- 10pm1-3 to 5- 10pm1-3 to 5- 10pm5- 10 pin250 mesh to 1-3 pm25-30 pm

40- 150 mesh60-200 mesh400-700 mesh60 mesh to 1-3 pm5- 10pm60- 150 mesh100-250 mesh10-250 mesh15-20 pm700 mesh to 25-30 pm100 mesh

5- 10 pm250-400 mesh400 mesh to 15-20 pm10-400 mesh80 mesh to 5-10 pm5- 10 to 25-30 pm10- 100 mesh700 mesh500 mesh to 5-10 pm60- 150 mesh700 mesh to 15-20 pm150 mesh to 5-10 pm20-250 mesh

25-75 pm200 mesh to 1-3 pm250 mesh to 5- 10 pm200 mesh60 mesh100 mesh25-75 pm1-3 pm25-30 pm


Table 2. (Continued)

Petroleum industry (continued)IsobutaneMEANaphthaProduced Water for Injection ResidualOilSea waterSteam InjectionVacuum Gas Oil

Pulp and PaperCalcium CarbonateClarified White WaterDyeFreshwaterGroundwood Decker RecycleHot Melt AdhesivesLatexMiliwaterPaper CoatingRiver WaterStarch SizeTitanium Dioxide

All IndustriesAdhesivesBoiler-feed WaterCaustic SodaChiller WaterCity WaterClay Slip (ceramic and china)Coal-Based SynluelCondensateCoolant WaterCooling Tower WaterDeionized WaterEthylene GlycolFloor PolishGlycerineInksLiquid DetergentMachine OilPelletizer WaterPhenolic Resin Binder PhotographicChemicalsPump Seal WaterQuench WaterResinsScrubber WaterWaxWeil water

250 mesh200 mesh to 5-10 fim25-30 fim1-3 to 15-20 /xm25-50 pm5- 10 jiin5-10 /xm25-75 /zm

30- 100 mesh30- 100 mesh60-400 mesh30-200 mesh20-60 mesh40-100 mesh40-100 mesh60-100 mesh30-250 mesh20-400 mesh20-100 mesh100-200 mesh

30- 150 mesh5-10 jum250 mesh200 mesh500 mesh to 1-3 fim20-700 mesh60 mesh200 mesh to 5- 10 /im500 mesh150-250 mesh100-250 mesh100 mesh to 1-3 /im250 mesh5-10/iin40-150 mesh40 mesh150 mesh250 mesh60 mesh25-30 ;xm200meshto5-10ftm250 mesh30-150 mesh40- 100 mesh20-200 mesh60 mesh to i-3 urn

Single filters may be piped directly into systems requiring batch or intermittentservice. Using quick-coupling connectors, the media can be removed from thehousing, inspected or cleaned. Also, filtering elements are interchangeable. Hence,while one is being cleaned, another can be placed into service.


Multiple filters are also common, consisting of two or more single filter units valvedin parallel to common headers. The distinguishing feature of these filters is the abilityto sequentially backwash each unit in place while the others remain on stream. Hence,these systems are essentially continuous filters. These units can be fully automated toeliminate manual backwashing. Backwashing can be controlled by changes indifferential pressure between the inlet and outlet headers. One possible arrangementconsists of a controller and solenoid valves that supply air signals to pneumatic valveactuators on each individual filter unit. As solids collect on the filter elements, flowresistance increases. This increases the pressure differential across the elements and,thus, between the inlet and outlet headers on the system. When the pressure dropreaches a preset level, an adjustable differential pressure switch relays informationthrough a programmer to a set of solenoid valves, which in turn sends an air signal tothe pneumatic valve actuator. This rotates the necessary valve(s) to backwash the firstfilter element. When the first element is cleaned and back on stream, each successivefilter element is backwashed in sequence until they are all cleaned. The programmeris then automatically reset until the rising differential pressure again initiates thebackwashing cycle.

Filter cartridges or tubes are made from a variety of materials. Common designs arenatural or synthetic fiber wound over a perforated plastic or metal core. A precisionwinding pattern covers the entire depth of the filter tube with hundreds of funnel-shaped tunnels, which become gradually finer from the outer surface to the center ofthe tube and trap progressively finer particles as the fluid travels to the center. Thisprovides greater solids retention capacity than is associated with surface filter mediaof the same dimensions. Typical cartridge materials are cotton, Dynel, polypropylene,acetate, porous stone and porous carbon filter lubes. Supporting perforated cores forcotton, Dynel or polypropylene are stainless steel, polypropylene or steel. Supportingcores for acetate tubes are tin-plated copper with voile liner. Porous stone and porouscarbon filter tubes do not require supporting cores. Stainless steel cores arerecommended for mildly acid and all alkaline solutions, pH 4-14. Polypropylene coresare used where all metal contact must be eliminated or where stainless steel isattacked, such as high chloride and sulfuric acid solutions. It is recommended for allacid and alkaline solutions, pH 0-14. Two types of polypropylene cores are available:mesh polypropylene and rigid perforated polypropylene. Mesh polypropylene issatisfactory for temperatures below 140°F. The more expensive rigid polypropylenecores are used for temperature applications over 140°F, and for double- and triple-tiered filter chambers because their greater strength is needed here. Perforated steelcores are used for dilute alkaline solutions, solvents, lacquers, oils, emulsions, etc.

Table 4 can serve as a rough guide to filter cartridge selection. The following generalguidelines are useful:

• Cotton filter tubes are recommended for moderately acid and alkaline solutions in the pH range 3-11.

• Polypropylene, Dynel and porous carbon filter tubes are recommended for concentrated acid andalkaline solutions and for all fluoborate solutions over the entire pH range (0-14).


• Polypropylene filter tubes are also recommended for electropolishing solutions, as well as certainother highly corrosive solutions.

» Porous stone Filter tubes are recommended for concentrated acid solutions.

» Acetate filter tubes are recommended for water,

Table 3. Typical filter retentions based on averages reported by different equipment suppliers.

Wire Mesh

Perforated

Slotted

Fabric

MeshEquivalent

102030406080100150200250400700

102030

10152030406080100120150200325

6080100150250500

Nominal Particle Retention

in.0.0650.0350.0230.0150.0090.0070.00550.00460.00330.00240.00180.0012

0.0630.0450.024

0.0630.0450.0350.0240.0150.0090.0070.0060.0050.0040.0030.0020.001

0.0090.0070.09550.00460.00240.00160.0010-0.00120.0006-0.00080.0002-0.00040.00004-0.00012

/*m165089058538023018014011584604530

15751125600

16001140890610380230180150125100755025

230180140115604025-3015-205-101-3

Open Area

564641362732303733363625

151812

5043363020182513119753

NANANANANANANANANANA

NA = percentage of open area not applicable to fabric media.


Strainer and Filter Bag Baskets

Strainer filter baskets and filter bag baskets are used as prefiltering devices. Thisprestraining or prefiltering stops the larger contaminated particles and thus extends thelife of the entire system.

Single-stage strainers and bag filters differ only in the basket design. Strainer basketshave solid flat bottoms, and baskets for filter bags have perforated bottoms to acceptstandard size filter bags.

Dual-stage straining/filtering action is achieved by insertion of a second, inner basket.It is supported on the top flange of the outer basket. Both baskets can be strainers(with or without wire mesh linings) or both can be baskets for filter bags. They mayalso be a combination: one a strainer basket, the other a filter bag basket. Dual-stageaction increases strainer or filter life and reduces servicing needs.

Figure 14 shows details of the basket seal, which prevents unfiltered liquid frombypassing the strainer or filter bag basket. The seal is maintained during operation bya hinged basket bail handle being held down under the closed cover, which holds thebasket down against a positive stop in the housing.

There are a variety of strainer and filter bag basket arrangements. Figure 15 showsdifferent single- and double-stage basket units.

Fabric bag filter baskets are capable of providing removal ratings from 20 mesh tonominal 1 /xm, for both Newtonian and viscous liquids. Wire mesh or fabric basketscan be cleaned and reused in many applications, or are disposable when cleaning is notfeasible.

Side-entry models feature permanent flanged connections, for line pressures to 150psi. These filters are fabricated to American Society of Mechanical Engineers (ASME)codes for applications that must comply with piping standards established in manyprocessing plants.

Top-entry models feature the inlet connection as an integral part of the lid. The inletcan be equipped with different types of quick disconnects for fast basket removal.

Strainers should be selected so that the pressure drop incurred does not exceed aspecified limit with a clean strainer basket (typically 2 psi). Pressure drop versus flowcapacity curves for basket strainers are given in Figure 16. This plot provides grosspressure drop for different capacities of water flow at suitable strainer pipe sizes. Thevalue obtained must be corrected on the basis of the actual fluid viscosity and straineropening size to be used. These corrections are given in Table 5 and the procedure isas follows:

» Under the pressure drqj value from the bottom scale, with the specified flowrate. Read up to whereits vertical line intersects the diagonal representing a strainer pipe size that gives a reasonablepressure drop, which is found by following the horizontal line to the pressure drop scale at the left.


• To correct this figure to match the actual fluid viscosity (and strainer media choice), use Table 5.Read down the appropriate viscosity column, and across from the appropriate strainer mediadescription, to find the correction factor.

• Multiply die pressure drop figure found in step 2 by the factor found in step 2 to obtain die adjustedpressure drop.

• If you are using a mesh-lined (not pleated) strainer basket, the pressure drop can be lowered byusing a 3Q-in.-deep basket instead of the 15-in.-deep one: divide the pressure drop figure from step3 by 1.5.

Diaphragm Filters

Diaphragm filters are specially designed filter presses. That have the ability to reducesludge dewatering costs by a squeezing cycle using a diaphragm. Instead of theconventional plate-and-frame unit in which constant pumping pressure is used to forcethe filtrate through the cloth, diaphragm filters combine an initial pumping followedby a squeezing cycle that can reduce the process cycle time by as much as 80%.

The operating cycles for this design are illustrated in Figure 17. During the filtrationcycle, sludge is fed at approximately 100 psi into each chamber through an inlet pipein the bottom portion of the filter plate. The number of chambers can range from afew dozen to more than 100. The sludge feed pump continues to feed sludge into thechamber until a predetermined filtering time has been achieved. Filtrate passes throughdie cloth on both sides of the chamber. The filtration cycle is completed independentlyin each chamber. Short filtration cycles produce cake thicknesses of 0.5-0.75 in,(12.7-19.1mm).

COVERGASKET

CONTAMINAT60FLUID

HOUSING

BASKET •^___^_ j,:I MillillillllMMlMBllllilillMKi' I .

ouTLrr

Figure 14. Details of basket seal.


Table 4. Guide to filter cartridge selection.

AcidFluoboratesNonfluoborates

AlkalineCyanide

PyrophosphateEletroless

Acids

Alkalies

Misc. Chemicals

Organic Liquids

Petroleum Products

Electroplating Solutions

Cu, Fe, Pb, SnCu, Sn, Zn: <6 oz/gal H2SO4

Cu, Sn, Zn: >6 oz/gal H2SO4

CrAu, In, Rli, PdFeCl2 (190°F)Ni (Woods)Ni (Watts type & bright)Ni (high-chloride)Ni (sulfamate)Electrotype Cu and NiSn (stannate)Brass, Cd, Cu, Zna

Au, In, Pt, AgCu, Fe, Sn, etc.Ni plating: < 140°FNi plating: > 140°FCu: <140°FCu: > 140°F

Chemicals

Acetic: diluteAcetic: concentratedBoricChromic, hydrochloric, nitric,phosphoric, sulfuricHydrofluoric, fluoboricNaOH or KOHNH4OH: dilute.NH4OH: concentratedBiological solutionsElectropolishing solutionsPharmaceutical solutionsPhotographic solutionsRadioactive solutionsUltrasonic cleaning solutionsNickel acetate (190°F)Food productsCC14

DichloroethyleneHydraulic fluidsLacquersPer- and trichloroethyleneSolventsFuel oil, diesel, kerosene,gasoline, lube oil

Filter Tube (Material/Core)

Polypropylene (PP) or Dynel/PPPP or cotton/PPPP or Dynel/PPPP or Dynel/PPPP or Dynel/PPPP/rigid PP (RPP) or porous stonePP or Dynel/PPPP or cotton/PPPP or cotton/PPPP or cotton/PP 'PP or cotton/PPCotton/stainless steel (SS)Cotton/SS, PP or Dynel/PPCotton/SS, PP/PPCotton/SS or PPCotton/SS or PPPP/RPP, cotton/SSPP/PPPP/RPP

Uter Tube (Material/Core)

Cotton/SS, PP/PPPP or Dynel/PPCotton/SS, PP/PP

PP or Dynel/PP, porous stoneb

PP or Dynel/PPPP/PPCotton/SS, PP/PPPP or Dynel/PPCotton/SS, PP/PP, porous stone11 Porousstone, PP/PPCotton/SS, PP/PP, porous stone"Cotton/SS, PP/PPCotton/SS, porous stoneh

Cotton special B compound/SSCotton/SSCotton/SS, PP/PPCotton/steel or SSCotton/steel or SSCotton/steel or SSCotton/steel or SSCotton/steel or SSCotton/steel or SS

Cotton/steel or SS

a When operated as high-speed baths at high temperatures (> 140°F) or with high alkali content, use PPor Dynel/PP.'' Porous stone is recommended for all acids except hydrofluoric and fluoboric.


Figure 15. Dgferent one- and two-stage basket configurations.


Strainer pine size: 2" 3" 4"

5 5 8 88 SS8 38

Flow, flpm

Figure 16. Pressure drop vsflow capacity for basket strainers.

Table 5, Viscosity correction factors.

Air unlined baskets, with orwithout pleated inserts40- mesh lined60- mesh lined80- mesh lined100-mesh lined

Viscosity (cP)10

0.650,730.770.931.00

50

0.850.951.001.201.30

100

1.001.201.301.501.60

200

1.101.401.601.902.20

400

1.201.501.702.102.40

600

1.401.802.102.402.70

800

1.501.902.202.603.00

1000

1.602.002.302.803.30

2000

1.802.302.803.504.40

Once the filtration cycle is complete, the sludge pump is stopped and a diaphragm inthe chamber is expanded by water pressurized up to 213 psi. This compresses thesludge on both sides of the chamber into a thin, uniform cake with a solids content ofmore than 35%. The uniform water content of the thin cake (no wet cores) results ineasier shredding and conveying and makes it much more adaptable to self-sustainedthermal destruction or landfill. Optimum filtering and squeezing time cycles vary,depending on the type of sludge, and can be determined accurately by bench tests.

Liquid Filtration

CAKE DISCHARGE CLOTH WASHING

Figure 17, Operating cycles of a diaphragm press.

Squeezing water is recycled. A hydraulic ram keeps (he chambers in position during bothcycles.

On completion of the filtration and squeeze cycles, the chambers are automatically openedand the cakes are discharged, usually onto a belt conveyor. No precoating is required. Twochambers are normally opened at a time in sequence. This reduces the impact loading onthe belt conveyor. Any sludge or filtrate remaming in the feed and filtrate lines isautomatically purged by high-pressure (100-psi) air before the next cycle begins. Thispurging prevents wet sludge from discharging and keeps sludge lines from plugging.


Cake discharge from filter presses is fast. After a number of cycles (depending on thesludge type), the filter cloth will require cleaning. This can be accomplished manuallyor can be performed automatically at preset frequencies with an automatic clothwasher using a jet of 1000-psi wash water.

Where even faster cake discharge is desired or where sludge cakes may tend to besticky, automatic cloth vibrators can be provided. These units help speed mechanicaldischarge and help remove cakes where poor sludge conditioning causes excessivesticking. This reduces the need for continuous monitoring by operations personnel. Clothvibrators also simplify cloth selection, since cloths can be selected to assure clearerfiltrate or better filtering qualities rather than sacrificing these advantages for a cloth thatallows for filter discharge characteristics. Cake discharge is illustrated in Figure 18.Typical capacities and dimensions for diaphragm presses are given in Table 6.

High Pressure, Thin Cake Filters

Thin-cake staged filters have been used effectively at high flowrates per unit area formany years in both Eastern and Western Europe. Use of ultrathin cakes is a usefultechnique for increasing flowrates.

The basic elements of the filter are shown in Figure 19. Filtration surfaces arerecessed plates equipped with rotating turbines that maintain permanent precoat-typethin cakes throughout the filter. Cake thickness is prevented from growing beyond thein situ precoat formed during the first few minutes of the operation by blades on arotating shaft passing through the axis of the filter.

Table 6. Capacities and dimensions for diaphragm presses".

Number ofChambers'1

304052667892104118130

Filter Area0

m2

115154200254300354400454500

ft2

1,2371,6562,1522,7323,2283,8084,3044,8845,380

Cake 'Volume at 0.7Thickness

liters

1,0021,3371,7362,2112,6113,0793,4793,9474,353

ft

3647617892109123139154

Weight"

kg53,20059,40067,80079,60088,70099,000110,400122,700130,200

tons

5865758898109121135143

Length"

mm

7,3557,9758,72010,03510,78011,65012,89513,76014,505

ft-in.

24-126-228-732-1135-438-242-345-147-7

" Based on average values reported by different machine suppliers.h Presses available from 30 to 130 chambers in 2-chamber increments.c Nominal plate size is 1500 X 1500 mm (4 tt-11 in. square).d Weight of the press only (without sludge)."Overall length. All presses have an overall width of 3000 mm and an overall height of 4200 mmin. X 13 ft-9in.).

(9 ft-10

IU Liquid Filtration

VIBRATION UNIT.

CAKE

WITH DtA PHRAGM

PLATE

Figure 18. Cake discharge from a diaphragm filter press.

PUD

TYPICAL STAGEDELAYED CAKE

FILTER-THICKENER

Figure 19. Principal components of a staged, thin-cake filter thickener.

Industrial Filtration Equipment 1 1 "

Slurry flows into the first stage and then flows around the turbines and through theclearances between the shaft and active filter surfaces. As liquid is removed, diethickening slurry moves from stage to stage. The unit acts as a filter-thickener,producing a continuous extruate that may contain a higher solids content than isnormally encountered in conventional filters.

The turbine plates sweep close to the filter cloths, leaving a thin, permanent cake oneach stationary plate (see Figure 19). Even in the last stage of the filter, where theslurry is highly non-Newtonian, a thin, easily identifiable, hard cake is maintained.

At low turbine velocities, the blades serve as scrapers that limit the cake thickness tothe'dimensions of the clearance. At higher velocities, the cake thickness is reduced andcan be as thin as 1.0 mm with 3-mm clearance. For filters that depend entirely on fluidaction, shear forces at the cake surface depend on fluid properties and velocitydistributions.

The combination of high pressure (300 psi) and thin cakes produces high rates.Washing is accomplished either cocurrently or countercurrently. Separate filters inseries can be employed in a manner similar to conventional thickeners. Washing mayalso be performed within a single unit, whereby, an initial portion of the filter mustbe used to remove liquid. In this case, the final stages are used for concentration.Clean wash liquid may be injected after the initial filtering at one or severalintermediate stations. Injection wash tends to increase the overall filtrate rate butdecreases the cake output rate.

Thickeners

Many existing filtration applications can be greatly enhanced if their presentequipment, such as plate-and-frames and rotary vacuums, is used in conjunction witha thickening operation. Table 7 illustrates this point. Case 1 shows that if a feed slurryof 2% is concentrated in a filter to 50% (by volume), a total of 98% fractionalremoval of water is needed. If, however, a thickener is employed to concentrate from2 to 10%, the fractional removal, of water is 82%, thus leaving only 16% of the filter.This means that the present filter could be used about three times more effectively itsupplemented with a thickener.

In Case 2, a 1% slurry is concentrated to 30% solids. A single filter would require98% fractional removal of the water. By use of a thickener concentrating first from1 to 7%, we fractionally remove 87% of the water. This leaves only 11% fractionalremoval of the present filter to go from 7% to the required 30%.

Many filter-thickeners are simple settling tanks or decanters. Thickeners of this typeare generally large and bulky and have relatively slow rates. Centrifuges have agreater driving force but, in general, are expensive to operate and can deliver cloudyoverflow if fine particles are present.


High-velocity cross-flow thickeners are available, but operating experience oftenshows them to be highly dependent on the rheology of the slurry. Sometimes a slightincrease in outlet concentration can result in filter blocking.

Table 7. Examples of thickener operation improvements to filtration.

Case

i

2

Solids(vol%)

21050

1730

Void Ratio, e*

499i

9913.292.33

Fractional Removal ofWater (F%)b

082160

8711

a e= vol liquid/vol solids.11 F% =( er^^ted,

Dynamic Thickeners

Dynamic thickeners have become a popular machine option. Special dynamic elementshoused inside a thickening chamber keep the slurry continuously moving; hence, aconcentrate of a paste-like consistency is possible without the danger of filter blocking.High flowrates per unit area, resulting from very thin cake formation, allow such unitsto be designed of relatively small size.

These systems are designed for little or no cake formation at lower levels ofconcentration. It can be shown that filtration rates increase with a reduction in cakethickness. However, some materials (especially gels such as aluminum hydroxides)are so compressible that 90% of the available pressure drop is absorbed by a "skinlayer" formed on top of the filter media, while the remainder of the cake remainssoupy and unconsolidated. Consequently, reducing the cake thickness (in examplessuch as unwashed "gels") in equipment that uses techniques of "thin cakes" would notresult in any significant improvements. It is, therefore, advantageous to minimize theformation of a "skin layer".

Dynamic thickeners operate in a recycle mode of operation. The feed enters thethickener and the filtrate leaves the filtering plates while the steady-state-rumiing,concentrated paste comes out of the modulating cake valve and reenters the feed tank.When the feed tank solids reach a predetermined concentration, e, the thickeningoperation is complete. To make the thickening operation continuous, one would installtwo feed tanks, so that after the first tank is completed the product can then be fed tocontinuous filtration equipment for further liquid removal as the second tank of feedsolution is processed through the thickener. These two feed tanks can be set up withhigh- and low-level audible signals and automatic switching three-way plug valves sothat continuous operations are possible with any continuous filter. The operatingscheme is illustrated in Figure 20.


FEED

FEED TANK 1

\

WASH WATER(IF REQUIRED)

FEED TANK 2

RECYCLE

Figure 20. Typical operating scheme for a dynamic thickener.


Solids Washing

Washing of chemical solids in filtration is employed to enhance the purity of theproduct. High washing efficiency is the ultimate goal, along with minimum use ofenergy and wash liquid, clear filtrates, maximum flowrates, and a homogeneouswashed product. Washing is usually accomplished in conventional cake-formingsystems by forcing wash liquid through existing filter cakes. The initial efficiency ishigh as the mother liquor is being displaced; however, after breakthrough, the processis controlled by the diffusion rate of the solute, which explains why washing efficiencydrops so rapidly with time.

Cake sagging in plate-and-frame filters reduces the effectiveness of wash as most ofthe liquid flows through the area having the least solids buildup of cake. In rotarydrum filters, an even more detrimental effect is cake cracking. Massive amounts ofwash liquid short-circuit directly through the filter cloth, thus, crippling the entirewashing process. In contrast, with dynamic thickeners, the slurry or paste is washedinstead of a cake. Solubles in the feed are dispersed into die wash liquid by strongagitation,

The time required to reduce the solubles in a slurry to a desired level is a function ofthe feed solids concentration. The optimum washing concentration can be determinedfrom the slurry's filtration characteristics.

For example, as shown in Figure 21, washing should begin at the feed concentrationif the data produce a concave curve. On the other hand, a convex curve implies thatwashing at the higher solids concentration is best. Finally, the washing curve maycontain an inflection point—an indication that the slurry should be thickened to apredetermined concentration before washing begins. Note that each curve shows itsoptimum slope line.

As the first stage is run, information is automatically recorded (with a data logger)concerning temperature, torque and filtrate weights. Thus, in one run filtrate rates canbe obtained as a function of solids concentration.

Centrifugal Filtration

Filtering centrifuges are distinguished from standard centrifugation by a filteringmedium incorporated into the design. Slurry is fed to a rotating basket or bowl havinga slotted or perforated wall covered with a filtering medium such as canvas or metal-reinforced cloth. The angular acceleration produces a pressure that transports theliquor through the filtering medium, leaving the solids deposited on the filter mediumsurface as a cake. When the feed stream is stopped and the cake spun for a short time,residual liquid retained by the solids drains off. This results in final solids that areconsiderably drier than those obtained from a filter press or vacuum filter.


»*>

Curve with InflectionPoint

Figure 21. Optimum wash curves for a dynamic thickener: Wash at feed concentration (top);wash at thickened concentration (middle); wash at semithickened concentration A (bottom).i/ris a function of rate; yis an inverse function of solids concentration.


Principal types of filtering centrifuges are suspended batch machines, automatic short-cycle batch machines and continuous conveyor centrifuges. In suspended centrifuges,the filter medium is usually canvas or a similar fabric, or woven metal cloth.Automatic machines employ fine metal screens. The filter medium in conveyorcentrifuges is usually the slotted wall of the bowl itself.

Figure 22 shows a widely used design. The system combines the features of acentrifuge and a screen. Feed enters the unit at the top and is immediately brought upto speed and distributed outward to the screen surface by a set of vanes. Water orother liquid is forced by the sudden centrifugal action through the screen openings intoan effluent housing. As solids accumulate, they are gently moved down the screen bydie slightly faster rotating helix. With the increase in screen diameter, highercentrifugal gravities are encountered and solids are dispersed over a graduallyincreasing area, thus forming a thin, compact cake from which the remaining liquidis extracted. The relatively dry solids are blown out the bottom of the rotor by a setof vanes into a conical collection hopper.

HOPPER

LIQUID FILTRATE

MOTOft

SOLIDS DlSCHAftQE

Figure 22. Cutaway view of one type of filter centrifuge.

The theory of constant-pressure filtration may approximately be applied to filtrationin a centrifuge. The following are assumed:


1, Effects of gravity and changes in the liquid kinetic energy arc negligible.2. The pressure drop developed from centrifagal action is equivalent to the drag of die

liquid flowing through the cake.3. Particle voids in the cake are completely tilled with liquid.4, The resistance of the filter medium is constant.5, Liquid now is laminar.6. The cake is incompressible.

Dewatering is not only an important step in a filtration process — it is also one of meprimary operations in processing materials. The necessary first step in the efficientdrying or processing of many products is the extraction of excess moisture byscreening and pressing. Sludge dewatering can be accomplished in several ways.However, in general, pressing tends to be a more energy-efficient operation thanevaporation or other heat transfer methods.

Multistage screw presses can be used for dewatering chemical cellulose and for theremoval of "black liquor" from kraft pulp, employing a recycling system with liquorflow countercurrent to the flow of stock, thus, producing a much higher percentageof solids in the liquor fed to evaporators. Presses can also be employed in thecontinuous rendering industry, as well as in reconstitution processes, as, for example,flax shive slurries, where four presses are used in conjunction with four slurryblending tanks, operating as a four-stage countercurrent washing or leaching step forupgrading an otherwise waste material. On certain products, continuous four-stagepresses can accomplish multistage counterflow washing in a single unit.

Screw presses may be used in the diffusion process for sugarcane, wherein the liquidsfor the diffusion of sugar solutions and fresh makeup water are extracted from the cutcane chips by the single or multistage unit.

Some other products that can be handled by continuous screw presses are:

reclaimed and synthetic rubber,wood pulp,waste paper pulp,drugs,miscellaneous chemicals,brewer's spent grains and hops,distiller's spent grain,packing house cracklings,paunch manure,soybean and cereal by-products,beet pulp,tomato pulp,

citrus pulp and peels,sweet and white potato pulp,tobacco slurries,cooked fish and fish cannery offal,copra,peat moss,corn germ,nitrocellulose,castor seed or beans,coffee grounds,alpha-cellulose.

Presses are available with many types of casings, designed to suit the characteristicsof the material to be pressed, such as:

1. heavy 1-in.-thick carbon or stainless steel slatted casing;


2. 3/16-in,-thick naval alloy brass drilled screens, tapered for self-cleaning;3. stainless steel perforated screens;4. stainless steel narrow bar type super-drainage casing; or5. tine mesh Dutch twilled filler cloth in stainless steel.

The narrow bar type drainage casing has approximately three times the drainage areaof the steel slatted casing, and twice the drainage area of the perforated or drilledscreen casings.

Low oil and moisture contents can be obtained with a continuous press, althoughoutput and final moisture content vary with the material being pressed, the speed atwhich the press is rotated, the uniformity of the feed to the press and themanufacturing process. A variable speed mechanical feeder with screw feed isavailable for forced feed when gravity feed is inadequate.

Presses are of extremely rugged construction. Various parts of the press (the screws,for example) may be chrome-plated, of stainless steel or Monel, or furnished in othermaterials where corrosion and abrasion are severe. Provisions can be made for steam,water, press liquor or other liquids to be injected for cleaning or improved processingresults. Figure 23 illustrates a typical screw press design.

The material enters the press through the intake hopper from a surge tank or conveyorand drops on the feed flights (with wide pitch) of the screw. The flights of the screwbecome progressively closer together and the cones of the various stages increase indiameter as they approach die discharge end. Each successive stage presses thematerial harder; the high pressure extracts the liquid, which passes through theperforated screens or other types of casings and leaves the press in ail directionsaround the casing.

Ultrafiltration

Three kinds of submicrometer semipenneable membranes can be delineated. The typewith the largest pores is used for microfiltration (MF). MF typically lies in the rangeof 0.02-10 /xm. MF separation generally involves removing particles from fluids basedon size; osmotic pressure is negligible. Ultrafiltration (UP) generally involvesseparation of large molecules from smaller molecules, and overlaps somewhat withdie porosity range of membranes used for reverse osmosis (RO). RO usually involvespurification or concentration of small molecules or ionic constituents in a solvent.Thus, we have microfilters, ultrafilters and membranes used for RO.

The overlap of the definitions for RO and UF membranes arises from the followingconsiderations. The "pores" in the skin of a membrane intended for removal of salt byRO are generally larger (e.g., 10-40 A) than the hydrated ions (e.g., Na+ C l , Ca + ,SO4

2 ) they are intended to repulse. However, these pores are filled with water thatis strongly influenced by the polymeric walls of the pores. Such water becomes"ordered water", which, because of its ordering, has too low a dielectric constant to

Industrial Filtration Equipm

ent


dissolve salt ions, in contrast to the bulk water. Thus, salt rejection in a useful ROmembrane (e.g., more than 85% salt rejection) is based, on the lack of solubility ofthe hydrated ions in the ordered water within the pores, not on the size of the pores,it is not hard to imagine that the same membrane, or (at least) an inferior ROmembrane (e.g., 5-20% salt rejection), would pass small molecules and reject largermolecules based primarily on size (UF) rather than on solubility (RO), hence theoverlap of RO and UF ranges shown in Figure 24.

Figure 24. Chart showing microporous filtration ranges.

RO, UF and MF membranes are generally a tew mils in thickness; however, thediscriminatory layer may be either a tight skin supported by an open substructure (i.e.,a very thin effective thickness and, thus, low frictional resistance to flow) or it maybe the entire thickness of the membrane or gel involved in the pass/rejectionmechanism. In the latter case, the friction factors are much higher, i.e., the entirethickness equals the effective thickness.

Osmotic pressure across a semipermeable membrane arises from differences inconcentration, which in turn arise from relative ratios of the numbers of impermeableindividual ions or molecules on the two sides of the membrane. These osmoticpressures are dominant when salts are to be removed by RO. Osmotic pressures varyfrom 3.5 psi for good tapwater to 350 psi with average sea water, as the number ofions per unit volume is very high (35,000 ppm). At the other extreme (MF), there areessentially no dissolved species that cannot permeate through the membrane; it followsthat the osmotic pressures are minimal. UF membranes lie in between, usually withvery few impermeable species of very high molecular weight; and therefore, muchlower osmotic pressures exist across the membrane. Exceptions can exist involving UFand may be circumvented, e.g., the pervaporation process.


It follows that operations, such as RO, involving high osmotic pressures require higherpressures (i.e., consume more energy) than do low-osmotic-pressure operations.Therefore, very thin effective thicknesses are desirable for practical industrial orcommercial RO installations, to cut down on the frictional resistance to flow clue toeffective thickness. Concentrated brine is continually swept out of the RO elementsand away from the membrane to avoid plugging and concentration polarization. At theother extreme, MF membranes involved in dead-end flow require low drivingpressures; therefore, thicker membranes with higher dirt-holding capacities aregenerally found most useful. Skinned or pseudoskinned varieties of MF membranesplug rapidly and account for some commercial failures of selected microfilters.

UF membranes lie in between RO and MF membranes and are of two kinds; both areuseful. Of industrial importance, are the thin-skinned membranes, which allowenhanced fiowrates (low friction factors) at given pressure differentials. Such UFmembranes have larger pores in the thin skin than most RO membranes, andmolecules of different molecular weights may be separated. Shape, size and molecularweight are important. As the osmotic effect is less important with UF membranes thanin the case of RO membranes, lower pressures (generally less than 100 psig) aresufficient to promote permeation, and molecules that differ by a factor of ten in theirmolecular weights may usually be separated. Fractionation of cheese whey intosolutions of protein and lactose is one familiar example.

Of medical and biotechnical importance are the thicker homogeneous gel membranes,such as Cuprophane™, which are used in the artificial kidney and/or concentrationdialysis. With the Cuprophane membranes, diffusional migration, driven byconcentration differences across the membrane, effects the transport of the variousspecies across the membrane and little, if any, pressure differential is applied.

In kidney dialysis, toxic "middle molecules" diffuse across the Cuprophane membraneand out of the blood, while the larger desirable species are retained. Almost as muchsalt diffuses out of the blood as diffuses into the blood from the dialysate during (hisprocedure. A small pressure is imposed that depletes the patient of a few pounds ofaccumulated water over a period of hours. Such processes are considered to beprimarily concentration-driven.

These thick gel membranes are biotechnically very important; however, the pressure-driven thin-skinned UF membranes, while perhaps somewhat less selective, produceproduct streams so much more rapidly that they are the materials of choice forindustrial processes.

The above discussions primarily considered the physical parameters of the variousmembranes and their porous properties. Particularly in the case of UF, seriousconsideration must be given to the species that penetrate or are rejected by the UFmembrane.

Figure 24 shows that different sources attempt to relate molecular weight and poresize. Note that 10 A is presumably the cutoff point for either 300- or 500-molecularweight molecules, depending on the reference. Both sources could be correct, and the


reasons that such uncertainty exists become even more important when largermolecular weights are involved.

As examples only, consider the behavior and properties of proteins that one mightwish to separate. Proteins are said to have primary, secondary, tertiary and,sometimes, quaternary structures. An oversimplified description of proteinconfigurations in solution is useful. The various individual amino acids (about equalin number to the letters in the alphabet) may be strung together head-to-tail in analmost infinite number of sequences, just as randomly hitting the keys of a typewriterwill give nonsensical words hundreds of letters long. Each of these random chains ofamino acids (words) would correspond to the primary structure of a different protein.A particular sequence of amino acids depicted in two dimensions is considered theprimary structure of a specific protein. There are, moreover, highly selective sitesalong these chains that are attracted to other specific sites along these same chains,forming loops held together by hydrogen bonds. A rendition of what sites areconnected to what other specific sites and hence whether the resulting protein moleculewould be forced to assume either helical or pleated sheet configurations reveals thesecondary structure. As a result of these same hydrogen-bond interactions, the helicalchains or pleated sheets become twisted, coiled chains, rods or globular shapes. Thismorphology constitutes the three-dimensional or tertiary structure.

On occasion, two to four independent chains (based on primary structure) becomeintertwined via hydrogen bonds and van der Waals forces and these also assumevarious three-dimensional morphologies. These multiple-strand agglomerates are saidto have quaternary structure.

To complicate matters still further, these protein molecules may assume differentmorphologies in different environments or solutions. Table 8 shows the intrinsicviscosities of various proteins where the intrinsic viscosity is defined as volume permass of a given protein; it may be seen that the molecular weights of proteins bearlittle relationship to die intrinsic viscosity.

Note that ovalbumin (44,000 molecular weight) is a compact globular particle thatoccupies 3.7 cnrVg; if the sulfur-sulfur bonds are decoupled it further opens toencumber 54 cmVg of protein.

Probable dimensions of variously sized particles are listed in Table 9. Furtherdiscussion exceeds the scope of this chapter. These examples illustrate that one shouldnot jump to any filtrative conclusions based on molecular weight.

Therefore, while it is safe to say that a given UF membrane could separate the muchsmaller lactose from the much larger protein in whey, it is dangerous to assume fliatselected proteins could be separated from each other without experimental evidence.All microporous filtration (MF, UF and RO) deals with purification, fractionation,concentration or partition. An example of purification is pressure-driven UF removalof particles and liigh-molecular-weight species from water subsequently to be used inhollow-fiber RO desalination. An example of pressure-driven fractionation isseparation of protein and lactose from cheese whey for use as food additives (in the

Industrial Filtration Equipment 12?

case of protein) and subsequent fermentation into alcohol (in the case of the lactose).Were the lactose merely defined as waste and dumped into a sewer, the process wouldbe defined as pressure-driven UF protein concentration.

Another concentration-driven UF process is the concentration of protein solutions indie laboratory where an aqueous solution of protein is placed in a dialysis bag or tubeand left for a period of hours in concentrated salt solution. Kidney dialysis alsoexemplifies concentration-driven partition filtration.

Table 8. Intrinsic viscosities for macromolecules.

Compact Globular ParticlesPolystyrene Latex ParticlesRibonucleaseLysozymeMyoglobinP-LactoglobulinOvalbuminSerum AlbuminHemoglobinLiver Alcohol DehydrogenaseHemerythrinAldolaseRibosomes (yeast)Bushy Stunt Virus

Randomly Coiled ChainsPolystyrene in Toluene

Reduced RibonucleaseOxidized RibonucleaseOxidized Ribonuclease in UreaOvalbumin in UreaSerum Albumin in UreaReduced Serum Albumin in UreaMyosin in Guanidine HydrochlorideRNAHeat- denatured DNA

Rodlike ParticlesFibrinogenCollagenMyosinDNATMV

Molecular Weight

109

13,70014,40017,00035,00044,00065,00067,00083,000107,000142,0003.5 X 106

8.9 X 106

45,00070,00013,70014,10014,10044,00066,00066,000200,0001.5 X 10*5 x 106

330,000345,000620,0005 X 106

4 X 107

IntrinsicViscosity(cm3/g)

2.43.33.03.13.44.03.73.64.03.63.85.04.0

283714.411.613.934225393100150

271150230500029

Having distinguished between MF, UF and RO, and identified the two prevalent kindsof UF membranes, we will now discuss modes of operation: cross-flow (also tangeutial-tlow and/or split-stream) filtration versus dead-end filtration. The numbers of particlesper unit volume generally diminish in (he order: RO (ions) > UF (molecules) > MF(bacteria, etc.); when the retained particles are comparatively small in number, as isusually the case in MF filtration, dead-end filtration is suitable (Table 10). At the other


extreme, as is the case in RO, concentration buildup always demands cross-flow,tangential-flow or split-stream treatment. The concentrate is continuously swept away,providing a relatively unchanged surface concentration. Pressure-driven UP also usessplit-stream filtration to avoid membrane plugging or concentration polarization (alsoknown as gel polarization). More recently, cross-flow filtration coupled withbackwashing has also been implemented in MF filtration when the particulate load isparticularly heavy or when long lifetimes of the MF membranes are desired.

Table 9. Dimensions of various particles.

ParticleYeasts, Fungi

BacteriaViruses

Proteins (lOMO6 mol wt)Enzymes

Antibiotics, PulypeptidesSugarsWater

Dimensions (/im)MO

0.3-100.03-0.3

0.002-0.10.002-0.005

0.0006-0.00120.0008-0.001

0.0002

Table 10. Dead-end versus cross-flow filtration.

Dead-end

Crosstlow,Tangential Flowor Split Stream

RO-

+

UF.

+

MF+

-f(energizing)

All three kinds of membranes (RO, UF and MF) may be manufactured in either flatsheet, tube or hollow tubular form. Generally, the hollow fiber (RO and smaller) orhollow tubular (UF and larger) configurations are less effective per unit area than arethe flat sheet configurations, but this is offset by the greater effective area that can bepacked into a volume of hollow tubules or fibers. The flat sheet configurations areusually plate-and-frame, spiral-wound or pleated cartridges.

The third configuration, large tube intermediate, possessing the performancecharacteristics of the flat sheet but lacking the surface-to-volume advantage of hollowfibers. Tube configurations can, however, cope with the most contaminated streams,primarily because they can be cleaned mechanically. At the two extremes, tiny hollowtubules and most flat sheet configurations can be cleaned by reverse flow, but certainclogging contaminants are difficult to remove.

The previous discussion brings us to one of the most important features of UF: gelpolarization, which is important when the separation of macromolecules is involvedin either flat sheet, tubular or hollow-fiber UF membrane configurations. As permeatecontaining the smaller molecules passes through the membrane, a layer of solutioncontaining the larger rejected molecules accumulates adjacent to the membrane surface


and may reduce the flow by plugging or fouling the membrane and/or forming agelatinous filtration medium in series with the original membrane, increasing frictionalresistance and sometimes reducing its effective pore size and not allowing the passageof smaller molecules that were intended lo pass through the unencumbered membrane.

In some cases the problem is so severe that UF is precluded. However, threeapproaches have been used successfully in restoring the utility of the fouledmembranes and/or keeping them from becoming fouled. They are (in order ofdecreasing difficulty of application) periodic purging with cleaning solutions (e.g.,chemicals or enzymes), introduction of turbulence (see below) by one of a number ofbaffling arrangements and periodic backflushing. Backflushing is most readily appliedto hollow tubular devices and is responsible in no small part for their growingacceptance. Turbulence promoters, generally inapplicable in hollow tubule devices,are most commonly employed in flat sheet configurations where, for example,Vexar™, a coarse webbing, is placed next to the membrane surface to induce asweeping action or eddy currents, which promote rapid mixing of the incipientboundary layer back into the bulk fluid. There is, of course, a maximum concentrationof potential gel-forming material that can be tolerated, at which point further UFbecomes ineffective. Such induced sweeping is employed in plate-and-frame, spiral-wound and pleated-membrane devices.

Periodically, the cumulative effects of gel polarization, dirt accumulation of biologicalgrowth, render it necessary to renovate or clean the UF assemblies. These cleaningor antifouling techniques are of three kinds: chemical, reverse-flow or mechanical.Combinations of these can be used. All are practicable, depending on element, moduleor cartridge configuration (refer to Table 11).

Chemical cleaning techniques are applicable to all configurations, although care mustbe taken to make certain that the membrane and other materials of construction arecompatible with the chemical agents used. Reversing the flow is usually practicable,but with certain flat-sheet, spiral-wound, fluted and tubular configurations, inadequatemembrane support during reverse-flow operation may cause problems.

Table 11, Cleaning techniques for UF,

Flat SheetSpiral Wound FlutedTubularHollow Fiber

Chemical4-++

+

Reverse-Flow

±±±±

Mechanical±-

+

Because plate-and-frame and tubular configurations are used with the mostcontaminated fluids, mechanical cleaning techniques are used. In the case of plate-and-franie systems, the equipment may be disassembled and scrubbed, while in the tubeconfigurations oversized soft foam plugs are driven through the tubes by pressure.


The three methods of cleaning fouled UP membranes have been discussed above andwhile induced mild turbulence (considered below) may be seen as a preventivemeasure, all of the procedures result from the necessity to counteract the effects of ge!polarization.

At least five additional techniques are being investigated that fall into the preventivecategory:

1. The tube pinch effect no doubt takes place during rapid laminar flow in hollow tubules wherehydrodynamic forces tend to cause particles to migrate toward the centers of the tubules and, hence,away from the walls.

2. Enzymes, which decompose protein deposits, have been incorporated into the UF membranes,either by postimmobilization or by inclusion during the membrane's manufacture. Such membranesmay be considered as self-cleaning to some extent.

3. Immobilized positive or negative charges have been attached to UF membranes. By repelling like-charged species, the tendency to foul is diminished (see section on electrodeposition of paints).

4. Electric fields have been imposed such that potentially fouling macromolecules or particles areattracted away (electrophoretically) from the UF membrane surface.

5. Emulsified surfactants are injected into the feed. The surfactants are selected depending on thespecific surfactant's enhanced ability to attract specific foulants to the water-surfactant interfacerather than to the membrane-water interface.

The hollow tubule configurations with lumens frequently on the order of 0.5-2,0 mmin diameter present a different set of constraints but also present opportunities.Consider a cartridge (Figure 25) composed of a large number of hollow tubules pottedat each end and encased in such a manner that the process stream can enter a plenum(A) at either end of the bundle of hollow tubules, proceed through the length of thetubules, losing fluid through the walls (UF), into the encasement and exit into either,a drain or reticule tank (B). Provision is made for removing the permeate (Q.

Figure 26 (left) illustrates a similar situation, where 90% of the material issues aspermeate (C). When the permeate flow decreases below a certain point due to fouling,the device may be renovated (Figure 26, right) by closing transiently the permeatevalve, reducing the average transmembrane pressure to zero, and concoinitantlyincreasing the fluid through die tubules fourfold. In this fashion die fast flush mayremove the accumulated debris.

A close look at fast flushing (Figure 26, right) reveals that backflushing is also takingplace. The hollow tubular bundle has a substantial friction factor due to die smalldiameters; hence, there is a pressure drop between A and B. Assuming forconvenience a 20-psi pressure drop down the tubules (from A to B) under fast flowconditions, what would be the pressure in the encasement? Assuming symmetry, thepressure would be around the average at A and B. Thus, at the tubules near A therewould be a 10-psi pressure drop between A and the encasement (encouragingultrafiltration permeation), while the pressure would be reversed at the tubule endingsnear B, encouraging ultrafiltered fluid to backflush the tubules near B. Reversing the"fast flow" direction through the tubules would backflush in turn each end of thedevice (refer to Figure 27).


B PROCESS FLUID OUT

IIPROCESS FLUID IN

Figure 25. Hollow tubule uaraftttration.

10%WASTE OUT

SINGLEPASS

OPERATION

BACKFLUSHZOfte

ULTRA*FILTBATIOII

ZONfi

A 100* FEED IM A 100% FEED IN

Figure 26. Hollow tubule utirafiUration.


FEED OUT

UPWARDFLOW

FEED IN

FILTRATION

UlTflA-FtLTftATlON

BACKFtlHWf

INFEED OUT "V

Figure 27. Fast flushing.

As backflushing near the middle would be nil from time to time, ultrafiltered fluid orother cleaning solutions could be injected through D and reclaimed or dumped throughA and/or B (refer to Figure 28).

Although UF was first thought to be primarily applicable to the treatment ofwaste waters, such as treated sewage, to remove particulate and macromolecularmatter, it is now known to be useful industrially in producing high-grade waters,recycling electrocoat paint particles, separations involving whole and skim milk,vegetable protein isolates (especially soybean), fermentation products, fruit juices,biochemicals such as pyrogens, phages in general, and human chorionicgonadrotropin.

Reverse Osmosis

Reverse osmosis (RO) for water and wastewater treatment and for reuse at electricity-generating power plants is a standard application. Uses of this unit operation include:recirculating condenser water, ash sluice water, boiler blowdown, boiler makeup andwet sulfur dioxide scrubber waste.

Use of RO for desalination of seawater for boiler makeup is a typical application. Theavailability of this system has opened up the use of heretofore unavailable watersupplies, and it has been used by the industry as a pretreatment to ion exchangedemineralization. RO acts as an economical roughing demineralizer, bringing downthe overall cost and improving the life of resins and operation of the ion exchangeequipment.


BACKFLUSH IN

WASTE

Figure 28. Back/lushing.

As noted earlier, osmosis is the spontaneous passage of a liquid from a dilute to amore concentrated solution across an ideal semipermeable membrane that allowspassage of the solvent (water) but not the dissolved solids (solutes) as shown in Figure29. If an external force is executed on the more concentrated solution, the equilibriumis disturbed and the flow of solvent is reversed. This phenomenon, RO, is depicted inFigure 30.

A basic RO treatment system consists of the components illustrated in Figure 31.Feedwater to the RO system is pumped first through a micrometer filter. This is areplaceable-cartridge element filter. The purpose of this filter is to remove anyturbidity and particulate matter from the feed water before it enters the RO system.

The filtered raw water then flows to a high-pressure pump, which feeds the rawwater at a typical pressure of 400 psi through the RO membrane system. Valvesand pressure gauges between the micrometer filter, die high-pressure pump andmembrane modules control the flow of water through the system and monitor itsoperation.

The RO system consists of two stages. The raw water is pumped through the firststage, which contains twice the number of membrane modules as the second stage.The first stage purifies 50% of the water fed to the system and rejects the remaining


50%, which contains all of the contaminants. This reject water from the first stage Kthen passed through the second stage, which purifies 50% of the water fed to it andrejects the remaining 50% to waste. This second stage reject now contains all of thecontaminants removed by both stages. Thus, the total flow through the system is 75%purified product water and 25% reject water.

IOSMOTIC

PRESSURE

SEMIPERMEABLEMEMBRANE

CONCENTRATED

• SOLUTION .DILUTE

SOLUTION

Figure 29. Osmosis: normal flow from low to high concentration.

SEMIPERMEABLEMEMBRANE

CONCENTRATED'- SOLUTION

DILUTESOLUTION

Figure 30. Reverse osmosis: flow reversed by application of pressure to high-concentrationsolution.

The RO system removes 90-95% of the dissolved solids in the raw water, togetherwith suspended matter (including colloidal and organic materials). The exact percentof product purity, product recovery and reject water depends on the amount ofdissolved solids in the feedwater and the temperature at which the system operates.

RO membrane performance in the utility industry is a function of two major factors:the membrane material and the configuration of the membrane module. Of the four

High PressureMicron Filter Switch -»j — i

O /"^ / •— *— ̂ ; ' -. Pump O

FEED ^T V ? ( p~~£* — ̂ —V~V £ Pump Pres

Control Va

Legend

9 Pressure Gauoe P"3 Ntsnua! Valven

CJ Pressure Control C*3 Solenoid Valve

2 Sampling Port XxZ Flow Meter(Water and/Of Pressare! | j

1ST STAGE

p— jj4 RQ Module ' €

X__-^vJ RO Module Ur-J-̂0~1T J,

rft. Inter Stage/-vfr Shut-Off Valve

sure __j RO Module Hr-iive O" 2ND STAGE ¥*

9 1 flil £^3 I 1

x VReject *- RejectControl Valve Flow f

O

LS=-o- REJEC

i/Ieter

4 Total Product' Flow Meter

u2R,— H h^PRODUCT

• Second StageProduct FlowMeter

r

Ift

. Typical reverse osmosis process. B.


RO membrane module types, most utility applications use either spiral-wound orhollow-fiber elements. Hollow-fiber elements are particularly prone to fouling and,once fouled, are hard to clean. Thus, applications that employ these fibers require agreat deal of pretreatment to remove all suspended and colloidal material in the feedstream. Spiral-wound modules, due to their relative resistance to fouling, have abroader range of applications. A major advantage of the hollow-fiber modules,however, is the fact that they can pack 5000 ft2 of surface area in a 1 f t 3 volume,while a spiral wound module can only contain 300 ft2/ft3.

The hollow fine fiber configuration consists of a bundle of porous hollow fine fibers.These fibers are externally coated with the actual membrane and form the supportstructure for it. Both ends of each fiber are set in a single epoxy tube sheet, whichincludes an O-ring seal to match the inside diameter of the pressure vessel. Influentwater enters one end of the pressure vessel and is evenly distributed along the lengthof the vessel by a concentric distributor tube. As the water migrates out radially, someof it permeates the fibers and exits the pressure vessel via the tube sheet on theopposite end. The direction of permeate flow is from outside to inside the fibers. Theconcentrated solution, or reject, completes its radial flow path and leaves the vesselat the same end at which it entered. Figure 32 is a representation of this configuration.For clarity, the vessel and inlet distributor have been omitted. The actual outsidediameters of individual fibers range 3-10 mils, depending on manufacturer. Figure 33depicts a complete module.

The spiral-wound configuration consists of a jelly roll-like arrangement of feedtransport material, permeate transport material and membrane material. At the heartof the wall is a perforated permeate collector tube. Several rolls are usually placed endto end in a long pressure vessel. Influent water enters one end of the pressure vesseland travels longitudinally down the length of the vessel in the feed transport layer.Direct entry into the permeate transport layer is precluded by sealing this layer at eachend of the roll. As the water travels in a longitudinal direction, some of it passes inradiaily through the membrane into the permeate transport layer. Once in the transportlayer, the purified water flows spirally into the center collection tube and exits thevessel at each end. The concentrated feed continues along the feed transport materialand exits the vessel on the opposite end from which it entered. A cross section of thespiral configuration is depicted in Figure 34 and a typical module assembly is shownin Figure 35.

The two types of membrane materials used are cellulose acetate and aromaticpolyamide membranes. Cellulose acetate membrane performance is particularlysusceptible to annealing temperature, with lower flux and higher rejection rates athigher temperatures. Such membranes are prone to hydrolysis at extreme pH, aresubject to compaction at operating pressures, and are sensitive to free chlorine above1.0 ppm. These membranes generally have a useful life of 2-3 years. Aromaticpolyamide membranes are prone to compaction. These fibers are more resistant tohydrolysis than are cellulose acetate membranes, but they are more sensitive to freechlorine.


Rotated Pprous BlockEnd Plato

Spacer

Fiber Bundle

Porous Distributor

End Plato 0' Ring

Plf»*

Figure 32. Hollow-fiber module.

FEEDTRANSPORTSPACER

PERMEATE TRANSPORT LAYER

PERMEATE TUBE

GLUE LINE

MEMBRANE

Figure 33. Spiral-wound membrane.


PurifiedWaterOutlet

Concentre!Outlet

PurifiedWaterOutlet

Concentrate Outlet

Figure 35. Spiral-wound module.

Intereonnector ft O-Rlng*

Sealed,End Cap

OnProduct Tube

Closure

The range of different filtration equipment is broad and it is difficult to generalizeselection criteria. Machinery selection depends largely on the application, theproperties of the slurry, the degree of separation or intended efficiency, throughputcapacities and solids loadings, and the economics of the process. Economicconsiderations should include the capital investment in the filtration unit andsupporting equipment, operating costs (in particular, energy costs), maintenancerequirements (including estimated life expectancy of parts and costs for replacement),and labor (operator attention time and cost of training qualified operators). Chapter6 provides an analysis for cost estimating filtration systems. Although the analysis ispresented for a specific filtration technology, the reader can readily generalize theanalysis for application to other filtration technologies.

APPLICATION OF FILTRATION TOWASTEWATER TREATMENT

Introduction

This chapter provides an overview of applications of filtration operations towaste water treatment applications. The most widely used filtration application inwastewater treatment is granular media filtration, although other methods are alsoused. Filtration may be applied as the primary treatment method, or more commonlyas both a pre-treatment step and as a final or finishing stage, depending on the cleanupobjectives and criteria. When employed as a finishing operation, the filtration processis referred to as polishing. Other operations that are often used with filtration includecarbon adsorption, sedimentation, disinfection, biological methods, and others. Thereader should consult the list of references at the end of this chapter for discussionson other unit operations used in wastewater treatment.

Granular Media Filtration

Granular media filtration is most often used for treating aqueous waste streams; thefilter media consists of a bed of granular particles (typically sand or sand withanthracite or coal). The bed is contained within a basin and is supported by anunderdrain system which allows the filtered liquid to be drawn off while retaining thefilter media in place. As water containing suspended solids passes through the bed offilter medium, the particles become trapped on top of, and within, the bed. Thefiltration rate is reduced at a constant pressure unless an increase in the amount ofpressure is applied to force the water through the filter. In order to prevent plugging,the filter is backflushed at high velocity to dislodge the particles. The backwash watercontains high concentrations of solids and is sent to further treatment steps.

142

5

Application of Filtration to Wastewater Treatment 143

Filter application is typically applied to handling streams containing less than 100 -200 mg/liter suspended solids, depending on the required effluent level. Increased-suspended solids loading reduces frequent backwashing. The suspended solidsconcentration of the filtered liquid depends on particle size distribution, but typically,granular media filters are capable of producing a filtered liquid with a suspendedsolids concentration as low as 1 - 10 mg/1. Large flow variations will affect theeffluent's quality.

Granular media filters are usually preceeded by sedimentation in order to reduce thesuspended solids load on the filter. Granular media filtration can also be installedahead of biological or activated carbon treatment units to reduce the suspended solidsload and in the case of activated carbon to minimize plugging of the carbon columns.

Granular media filtration is only marginally effective in treating colloidal size particlesin suspensions. Usually these particles can be made larger by flocculation although thiswill reduce run lengths. In cases where it is not possible to flocculate such particles(as in the case of many oil/water emulsions), other techniques such as ultrafiltrationmay be nessesary.

Filtration is an effective means of removing low levels of solids from wastes providedthe solids content does not vary greatly and the filter is backwashed at appropriateintervals. The operation can be easily integrated with other treatment steps, andfurther, is well suited to mobile treatment systems as well as on-site or fixedinstallations,

A typical physical/chemical treatment system incorporates three "dual" medial (sandanthracite) filters connected in parallel in its treatment train. The major maintenanceconsideration with granular medial filtration is the handling'of the backwash. Thebackwash will generally contain a high concentration of contaminants and requiresubsequent treatment.

In this application, the operations of precipitation and flocculation play importantroles. Precipitation is a physiochemical process whereby some, or all, of a substancein solution is transformed into a solid phase. It is based on alteration of the chemicalequilibrium relationships affecting the solubility of inorganic species. Removal ofmetals as hydroxides and sulfides is the most common precipitation application inwastewater treatment. Lime or sodium sulfide is added to the wastewater in a rapidmixing tank along with flocculating agents. The wastewater flows to a flocculationchamber in which adequate mixing and retention time is provided for agglomerationof precipitate particles. Agglomerated particles are then separated from the liquidphase by settling in a sedimentation chamber, and/or by other physical processes suchas filtration. Precipitation is often applied to the removal of most metals fromwastewater including zinc, cadmium, chromium, copper, fluoride, lead, manganese,and mercury. Also, certain anionic species can be removed by precipitation, such asphosphate, sulfate, and fluoride. Note that in some cases, organic compounds mayform organometallic complexes with metals, which could inhibit precipitation.Cyanide and other ions in the wastewater may also complex with metals, makingtreatment by precipitation less efficient.


The process of flocculation is applicable to aqueous waste streams where particlesmust be agglomerated into larger more settleable particles prior to sedimentation orother types of treatment. Highly viscous waste streams will inhibit the settling ofsolids. In addition to being used to treat waste streams, precipitation can also be usedas an in situ process to treat aqueous wastes in surface impoundments. In an in situapplication, lime and flocculants are added directly to the lagoon, and mixing,flocculation, and sedimentation are allowed to occur within the lagoon.

Precipitation and flocculation can be integrated into more complex treatment systems.The performance and reliability of these processes depends greatly on the variabilityof the composition of the waste being treated. Chemical addition must be determinedusing laboratory tests and must be adjusted with compositional changes of the wastebeing treated or poor performance will result.

Precipitation is nonselective in that compounds other than those targeted may beremoved. Both precipitation and flocculation are nondestructive and generate a largevolume of sludge which must be disposed of. Coagulation, flocculation, sedimentation,and filtration, are typically followed by chlorination in municipal wastewater treatmentprocesses.

Coagulation involves the addition of chemicals to alter the physical state of dissolvedand suspended solids. This facilitates their removal by sedimentation and filtration.The most common primary coagulants are alum ferric sulfate and ferric chloride.Additional chemicals that may be added to enhance coagulation include activate silica,a complex silicate made from sodium silicate, and charged organic molecules calledpolyelectrolytes, which include large-molecular-weight polyacryl-amides, dimethyl-diallylammonium chloride, polyamines, and starch. These chemicals ensure theaggregation of the suspended solids during the next treatment step-flocculation.Sometimes polyelectrolytes (usually polyacrylamides) are also added after flocculatioeand sedimentation as an aid to the filtration step.

Coagulation may also remove dissolved organic and inorganic compounds. Thehydrolyzing metal salts may react with the organic matter to form a precipitate, orthey may form aluminum hydroxide or ferric hydroxide floe particles on which theorganic molecules adsorb. The organic substances are then removed by sedimentationand filtration, or filtration alone if direct filtration or inline filtration is used.Adsorption and precipitation also removes inorganic substances.

Note that flocculation is a purely physical process in which the treated water is gentlystirred to increase interparticle collisions and, thus, promote the formation of largeparticles. After adequate flocculation, most of the aggregates will settle out during the1-2 hours of sedimentation.

The process of sedimentation involves the separation from water, by gravitationalsettling of suspended particles that are heavier than water. The resulting effluent isthen subject to rapid filtration to separate out solids that are still suspended in thewater. Rapid filters typically consist of 24 - 36 inches of 0.5- to 1-mm-diameter sandand/or anthracite. Particles are removed as water is filtered through the media at rates


of 1 - 6 gallons/minute/square foot. Rapid filtration is effective in removing mostparticles that remain after sedimentation. The substances that are removed bycoagulation, sedimentation, and filtration accumulate in sludges which must beproperly disposed of.

Coagulation, flocculation, sedimentation, and filtration will remove many contaminants.Perhaps most important is the reduction of turbidity. This treatment yields water of goodclarity and enhances disinfection efficiency. If particles are not removed, they harborbacteria and make final disinfection more difficult.

FILTER TANK

GRADEDGRAVEL

PERFORATED LATERALS

FILTER FLOOR

CAST-IRON MANIFOLD

Figure 1. Cutaway view of a rapid sand filter.

The hydraulic performances required of the sand with slow filters are inferior to thosefor rapid filters. In the case of slow filters, one can use fine sand, since the averagefiltration velocity that is usually necessary lies in the range 2 - 5 in/day.

In slow filtration, much of the effect is obtained by the formation of a filtration layer,including the substances that are extracted from the water. At the early stages of theoperation, these substances contain microorganisms able to effect, beyond thefiltration, biochemical degradation of the organic matter. This effect also depends onthe total surface of the grains forming the filter material. The probability of contactbetween the undesirable constituents of the water and the surface of the filter mediumincreases in proportion to the size of the total surface of the grains.


The actual diameter of the sands used during slow filtration typically lies between 0.15and 0.35 mm. It is not necessary to use a ganged sand. The minimum thickness of thelayer necessary for slow filtration is 0.3 - 0.4m, and the most efficient filtrationthickness typically is at 2 - 3 cm.

The actual requirements for the sand in slow filtration are chemical in nature. Purityand the absence of undesirable matters are more important than grain-sizedistribution in the filtration process. On the other hand, the performance of rapidfilters requires sands with quite a higher precise grain size. In the case of rapidfiltration, the need for hydraulic performances is greater than in slow nitration. Thismeans that the grain-size distribution of the medium is of prime concern in the lattercase.

Sand often contains undesirable impurities, and additionally it can have broad particlesize distributions. Sand that is used in filtration must be free of clay, dust, and otherimpurities. The ratio of lime, lime-stone, and magnesium oxide will have to be lowerthan 5 weight percent. The standard guide value of the quality of fresh sand is to bebelow 2% soluble matter at 20 °C within 24 hours in hydrochloric acid of a 20 weightpercent concentration.

In waste water treatment plants, the purity of the sand media used must be examinedregularly. In addition, both the head loss of the filter beds and an analysis of the washwater during the operation of washing the filters must be checked regularly. Specialattention must also be granted to the formation of agglomerates. The presence ofagglomerates is indicative of insufficient washing and the possible formation ofundesirable microbiological development zones within the filter bed.

The primary mechanisms that control the operation of sand filtration are:

StrainingSettlingCentrifugal actionDiffusionMass attraction, or the effect of van der Waals forcesElectrostatic attraction

Straining action consists of intercepting particles that are larger than the freeinterstices left between the filtering sand grains. Assuming spherical grains, anevaluation of the interstitial size is made on the basis of the grains' diameter (specificdiameter), taking into account the degree of nonhomogeneity of the grains. Porosityconstitutes a important criterion in a description based on straining. Porosity isdetermined by the formula VL/VC, in which Vc is the total or apparent volumelimitated by the filter wall and VL is the free volume between the particles. Theporosity of a filter layer changes as a function of the operation time of the filters. Thegrains become thicker because of the adherence of material removed from the water,whether by straining or by some other fixative mechanism of particles on the filteringsand. Simultaneously the interstices between the grains dimmish in size. This effectassists the filtration process, in particular for slow sand filters, where a deposit is


formed as a skin or layer of slime that has settled on the bed making up the activefilter. Biochemical transformations occur in this layer as well, which are necessary tomake slow filters efficient as filters with biological activity.

Filtration occurs correctly only after buildup of the sand mass. This formation includesa "swelling" of the grains and, thus, of the total mass volume, with a correspondingreduction in porosity. The increases and swellings are a result of the formation ofdeposits clinging to the empty zones between grains.

The porosity of a filter mass is an important factor. This property is best defined byexperiment. A general rule of thumb is that for masses with the effective size greaterthan 0.4-0.5 mm and a specific maximum diameter below 1.2 mm the porosity isgenerally between 40 and 55 % of the total volume of the filter mass. Layers withspherical grains are less porous than those with angular material.

The second important mechanism in filtration is that of settling. From Stoke's law oflaminar particle settling, the settling velocity of a particle is given by :

18 v p w

where :p = volumetic mass density of the waterp+Ap = volumetic mass density of the particles in suspensionD = diameter of the particlesg = 9.81m/s2

v = kinematic viscosity (e.g., 104 m/s at 20°C)

In sedimentation zones the flow conditions are laminar. A place is available for thesettling of sludges contained in the water to be filtered.

Although the total inner surface that is available for the formation of deposits in afilter sand bed is important, only a part of this is available in the laminar flow zonesthat promote the formation of deposits. Usually material with a volumetic massslightly higher than that of water is eliminated by sedimentation during filtration. Suchmatter could be, for example, organic granules or particles of low density. In contrast,colloidal material of inorganic origin-sludge or clay, for instance—with a diameter of1 - 1 0 (Jim is only partially eliminated by this process, in which case the settlingvelocities in regard to the free surface become insufficient for sedimentation.

The trajectory followed by water in a filter mass it is not linear. Water is forced tofollow the outlines of the grams that delineate the interstices. These changes indirection are also imposed on particles in suspension being transported by the water.This effect leads to the evacuation of particles in the dead flow zones. Centrifugalaction is obtained by inertial force during flow, so the particles with the highestvolumetic mass are rejected preferentially.


Diffusion filtration is another contributor to the process of sand filtration. Diffusionin this case is that of Brownian motion obtained by thermal agitation forces. Thiscompliments the mechanism in sand filtration. Diffusion increases the contactprobability between the particles themselves as well as between the latter and the filtermass. This effect occurs both in water in motion and in stagnant water, and is quiteimportant in the mechanisms of agglomeration of particles (e.g., flocculation).

The next mechanism to consider is the mass attraction between particles which is dueto van der Waals forces. These are universal forces contributing to the transport andfixation mechanism of matter. The greater the inner surface of the filters, the higheris the probability of attractive action. Van der Waals forces imply short moleculardistances, and generally play a minor role in the filtration process. Moreover, theydecrease very quickly when the distance between supports and particles increases.Nevertheless, the indirect effects, which are able to provoke an agglomeration ofparticles and, thus, a kind of flocculation, are not to be neglected and may becomepredominant in the case of flocculation-filtration, or more generally in the case offiltration by flocculation.

Electrostatic and electrocinetic effects are also factors contributing to the filtrationprocess. Filter sand has a negative electrostatic charge. Microsand in suspensionpresents an electrophoretic mobility. The value of the electrophoretic mobility, or ofthe corresponding zeta potential, depends on the pH of the surrounding medium.Usually a coagulation aid is used to condition the surface of microsand. In filtrationwithout using coagulant aids, other mechanisms may condition the mass more or lesssuccessfully. For instance, the formation of deposits of organic matter can modify theelectrical properties of the filtering sand surfaces. These modifications promote thefixation of particles by electrokinetic and electrostatic processes, especiallycoagulation. Also, the addition of a neutral or indifferent electrolyte tends to reducethe surface potential of the filtering sand by compression of the double electric layer,This is based on the principles of electrostatic coagulation. The sand, as the carrier ofa negative charge spread over the surface of the filter according to the model of thedouble layer, will be able to fix the electropositive particles more exhaustively. Thishas a favorable effect on the efficiency of filtration of precipitated carbonates or offloes of iron or aluminum hydroxide-oxide. Optimal adherence is obtained at theisoelectric point of the filtrated material. In contrast, organic colloidal particle carriersof a negative charge such as bacteria are repulsed by the electrostatic mechanism ina filter with a fresh filter mass. In this case, the negative charges of the sand itselfappear unchanged. With a filter that is conditioned in advance, there are sufficientpositively charged sites to make it possible to obtain an electrochemical fixation of thenegative colloids.

Bed Regeneration

In addition to washing the bed, a degradated mass containing agglomerates orfermentation zones (referred to as mud balls) can be regenerated by specific treatmenttechniques. Among the regeneration techniques that are usually used are sodium


chloride, regeneration through application of chlorine, and treatment with potassiumpermanganate, hydrogen peroxide, or caustic soda.

Cleaning methods based on the use of caustic soda are aimed at eliminating thinclay, hydrocarbons, and gelatinous aggregates that form in filtration basins. Afterthe filter has been carefully washed with air and water or only with water, accordingto its specific operating scheme, a quantity of caustic soda is spread over a waterlayer approximately 30 cm thick above the filter bed. The solution is then diffusedin the mass by slow infiltration. After about 6 -12 hours, the filter is washed verycarefully.

Sodium chloride is used specifically for rapid filters. The cleaning solution is spreadin solution form in a thin layer of water above the freshly washed sand bed. After 2or 3 hours of stagnation, slow infiltration in the mass is achieved by opening an outletvalve for the filtered water. The brine is then allowed to work for about a 24 hourperiod. The filter is placed back into service after a thorough washing. Sodiumchloride works on proteinic agglomerates, which are bacterial in origin.

The use of potassium permanganate (KMnO4) is applied to filters clogged with algae.A concentrated solution containing potassium permanganate is spread at an effectiveconcentration over the surface of the filters to obtain, a characteristic pink-purplecolor on the top of the mass and allowed to infiltrate the bed for a 24 hour period.After this operation, the filter is carefully washed once again.

Hydrogen peroxide is typically used in the range of 10 - 100 ppm. The cleaningmethod is similar to that used for permanganate. The addition of phosphates orpolyphosphates makes it easier to remove ferruginous deposits. This method can beused in situ for surging the isolation sands of the wells. Adjunction of a reductor asbisulfite can be useful to create anaerobic conditions for the elimination of nematodesand their eggs when a filter has been infected.

Hydrochloric acid solution is applied to the recurrent cleaning of rapid filters for sand,iron, and manganese removal. This operation has the advantage of causing the formationof chlorine in situ which acts as a disinfectant.

Instantaneous cleaning of a filtering sand bed can be accomplished by the use ofchlorine. A water layer is typically used as a dispersion medium. Further infiltrationof the solution is obtained by percolation into the bed. The action goes on for severalhours, after which the filter is washed. Chlorine is used from concentrated solutionsof sodium hypochlorite. An alternative method involves the application of dioxide.This method has the advantage of arresting the formation of agglomerates of biologicalorigin by permanent treatment of the filter wash water with chlorine.

Fiocculation Filtration

The sand filtration process is normally comprised of a clarification chain includingother unit operations which precede filtration in the treatment sequence and can not


be conceived of completely independent of the filtration stage. The conventionaltreatment scheme consists of coagulation-flocculation-settling followed by filtration.When the preceding process, in this case flocculation and/or settling, becomesinsufficient, subsequent rapid filtration can be used to ensure a high quality of theeffluent treated. However, this action is achieved at the expense of the evolution offilter head loss. Problems in washing and cleanliness of the mass may arise. Filtrationis often viewed as serving as a coagulant flocculator. This is referred to asflocculation-filtration.

The presence of thin, highly electronegative colloids (e.g., activated carbons) introducedin the form of powder in the settling phase may be a problem for the quality of thesettled effluent. The carbon particles, which are smaller than 50 /^m, penetrate deeplyinto the sand filter beds. They may rapidly provoke leakage of rapid filters. The sameholds for small colloids other than activated carbon.

Activated silica, which may have a favorable or an unfavorable effect on filtration, iscomposed of ionized micella formed by polysilicic acid-sodium polysilicate. Thisbecome negatively charged colloidal micella. The behavior of activated silicas dependson the conditions of neutralization and the grade of the silicate used in the preparationof the material. Activated silica is a coagulant aid that contributes to coalescence ofthe particles. Hence, it brings about an improvement in the quality of settled orfiltrated water, depending on the point at which it is introduced.

Preconditioning of the sand surface of filters by adding polyelectrolytes is an alter-native use of sand filters as a coagulator-flocculator. In the treatment of drinking waterthe method depends on the limitations of these products in foodstuffs.

The addition of polyphosphates to a water being subjected to coagulation usually has anegative effect; specifically the breaking of the agglomeration velocity of the particlesduring flocculation will occur in sand filtration. The addition of polyphosphatessimultaneously with phosphates can be of value in controlling corrosion. This sometimesmakes it possible to avoid serious calcium carbonate precipitation at the surface of filtergrains when handling alkaline water. The application concerns very rapidly incrastingwater while maintaining high hardness in solution. The addition of polyphosphatesinvolves deeper penetration of matter into the filter mass. Hence, the breaking offlocculation obtained by the action of polyphosphates enables the thinner matters topenetrate the filters more deeply. These products favor the "in-depth effects" of thefilter beds. Their use necessitates carefully checking that they are harmless from ahygienic point of view.

The depth penetration of material in coagulation-filtration is almost opposite to theconcept of using the filter as a screen. Precipitation initiated by germs plays asignificant role. Empirical relations are normally relied on in the design of filters asa function of the penetration in depth of coagulated material. The concentration ofthose residual matters in filtered water (Cf) depends on several factors: the linearinfiltration rate (vt), the effective size of the filter medium (ES), the porosity of thefilter medium (e), the final loss of head of the filter bed (Ah), the depth of penetrationof the coagulated matter (/), the concentration of the particles in suspension in the

Application of Filtration to Wastewater Treatment 15!

water to be filtered (C0), and the water height (H). The following generalized relationis often found among the filtration engineer's notes.

= / ( vfx(ES) x Q x / / (2)

It should be noted lhat the total loss of head of a filter bed is in inverse ratio to the depthof penetration of the matter in suspension.

In a normal wastewater treatment plant, the water is brought onto a series of rapid sandfilters and the impurities are removed by coagulation-flocculation-filtration.Backwashing is typically performed in the counterflow mode, using air and water. Onetype of common filter is illustrated in Figure 2, consisting of closed horizontalpressurized filters.

HITRATIQN PHASE PHASE lir *i!ve

f l i tcren

/ ' \I2L./ I \ X

nishhnl

Figure 2. Cross section of a typical filtration unit.

Slow Sand Filtration

Slow sand filtration involves removing material in suspension and/or dissolved in waterby percolation at slow speed. In principle, a slow filter comprises a certain volume ofarea! surface, with or without construction of artificial containment, in which filtrationsand is placed at a sufficient depth to allow free flow of water through the bed. Whenthe available head loss reaches a limit of approximately 1 m, the filter must be pulledout of service, drained, and cleaned. The thickness of the usual sand layer isapproximately of 1 - 1.50 m, but the formation of biochemically active deposits andclogging of the filter beds takes place in the few topmost centimeters of the bed.


The filter mass is pored onto gravels of increasing permeability with each layer havinga thickness of approximately 10 - 25 cm. The lower-permeability layer can reach atotal thickness of 50 - 60 cm. So-called gravels 18 - 36 cm in size are used and theirdimensions are gradually diminished to sizes of 10 - 12 cm or less for the uppersupport layer.

The sand filter must be cleaned by removal of a few centimeters of the clogged layer. Thislayer is washed in a separate installation. The removal of the sand can be done manuallyor by mechanical means. The removed sand may not be replaced entirely by fresh sand.Placing preconditioned and washed sand is recommended as this takes into account thebiochemical aspects involved in slow filtration. An alternative to manual or mechanicalremoval involves cleaning using a hydraulic system as illustrated in Figure 3.

WATER UNDER PRESSURE

-*«-—*>»-—*

11

ASPIRATION

Figure 3. Hydraulic cleaning device for slow sand filters.

Sometimes slow filtration is used without previous coagulation. This is generallypracticed with water that does not contain much suspended matter. If the water isloaded (periodically or permanently) with clay particles in suspension, pretreatmentby coagulation-flocculation is necessary. Previous adequate oxidation of the water, inthis case preozonization producing biodegradable and metabolizable organicderivatives issuing from dissolved substances, can be favorable because of thebiochemical activity in slow filters.

There are several disadvantages to the use of slow filters. They may require asignificant surface area and volume, and may therefore involve high investment costs.


They are also not flexibile — mainly during the winter, when the open surface of thewater can freeze. During the summer, if the filters are placed in the open air, algaemay develop, leading to rapid clogging during a generally critical period of use. Algaeoften cause taste and odor problems in the filter effluent. Additional construction coststo cover slow filters are often necessary.

Rapid Sand Filtration

Rapid filtration is performed either in open gravitational flow filters or in closedpressure filters. Rapid pressure filters have the advantage of being able to be insertedin the pumping system, thus allowing use of a higher effective loading. Note thatpressure filters are not subject to development of negative pressure in a lower layerof the filter. These filters generally support higher speeds, as the available pressureallows a more rapid flow through the porous medium made up by the filter sand.Pressure filtration is generally less efficient than the rapid open type with free-flowfiltration. Pressure filters have the following disadvantages. The injection of reagentsis complicated, and it is more complicated to check the efficiency of backwashing.Work on the filter mass is difficult considering the assembly and disassembly required.Also, the risk of breakthrough by suction increases. Another disadvantages is thatpressure filterts need a longer filtration cycle, due to a high loss of head available toovercome clogging of the filter bed.

Another option is to use open filters, which are generally constructed in concrete.They are normally rectangular in configuration. The filter mass is posed on a filterbottom, provided with its own drainage system, including bores that are needed forthe flow of filtered water as well as for countercurrent washing with water or air.There are several types of washing bottoms. One type consists of porous plates whichdirectly support the filter sand, generally without a layer of support gravel. Even ifthe system has the advantage of being of simple construction, it nevertheless suffersfrom incrustation. This is the case for softened water or water containing manganese.Porous filters bottoms are also subject to errosion or disintegration upon the filtrationof aggressive water.

The filter bottom is often comprised of pipes provided with perforations that areturned toward the underpart of the filter bottom and embedded in gravel. The lowerlayers are made up of gravel of approximate diameter 35 - 40 mm, decreasing up to3 mm. The filter sand layer, located above this gravel layer, serves as a support andequalization zone. Several systems of filter bottoms comprise perforated self-supporting bottoms or false bottoms laid on a supporting basement layer. The formerconstitutes a series of glazed tiles, which includes bores above which are a series ofgravels in successive layers.

All these systems are surpassed to some extent by filter bottoms in concrete providedwith strainers. The choice of strainers should in part be based on the dimensions of theslits that make it possible to stop the filter sand, which is selected as a function of thefiltration goal. Obstruction or clogging occurs only rarely and strainers are sometimesused.


Strainers may be of the type with an end that continues under the filter bottom. Thesedo promote the formation of an air space for backwashing with air. If this air spaceis not formed, it can be replaced by a system of pipes that provide for an equaldistribution of the washing fluids.

Pressure filters are worth noting. These are usually set up in the form of steelcylinders positioned vertically. Another variation consists of using horizontal filtrationgroups. This has the drawback that the surface loading is variable in the differentlayers of the filter bed; moreover, it increases with greater penetration in the filter bed(the infiltration velocity is lowest at the level of the horizontal diameter of thecylinder). The filter bottom usually consists of a number of screens or mesh sieves thatdecrease in size from top to bottom or, as an alternative, perforated plates supportinggravel similar to that used in the filter bottoms of an open filter system.

Filter mass washing can influence the quality of water being filtered. Changes may beconsequent to fermentation, agglomeration, or formation of preferential channels liableto occur if backwashing is inadequate.

Backwashing requires locating a source that will supply the necessary flow andpressure of wash water. This water can be provided either by a reservoir at a higherlocation or by a pumping station that pumps treated water. Sometimes an automatedsystem is employed with washing by priming of a partial siphon pumping out thetreated water stored in the filter itself. An example is shown in Figure 4. The washwater must have sufficient pressure to assure the necessary flow.

Washing of the filter sands is accomplished followed by washing with water and inmost cases including a short intermediate phase of simultaneous washing with air andwater. Due to greater homogenization of the filter layer and more efficient washing,the formation of fermentation areas and agglomerates in the filter mass of treatmentplants for surface water (mud balls) is diminished. The formation of a superficial cruston the filter sand is avoided by washing with air.

After washing with air, water flow is gradually superimposed on the air flow. Thisoperational phase ends at the same time that the wash air is terminated, to avoid thefilter mass being blown away.

The wash water contains materials that eventually require treatment in a sludgetreatment plant. Their concentration varies as a function of the washing cycle.Accounting for the superficial load in filtration, velocity of the wash water, and lengthof the filtration cycle, it may be assumed that the water used for washing will notattain 5 % of the total production.

For new installations the first washing cycles result in the removal of fine sand as wellas all the other materials usually undesirable in the filter mass, such as particles ofbitumen on the inner surface of the water inlet or other residuals from the crushing orstraining devices of the filter media. Consequently, it is normal that at the beginningof operation of a filter sand installation, dark colored deposits appear at the surfaceof the filter mass. In the long term they have no consequence and disappear after a few


Figure 4. Automatic backwashing filter with a partial siphon system: 1-filtered water(reserve); 2-partial siphoning; 3-initiation; 4-restitution.

filtration and wash cycles. If, after several weeks of filtration, these phenomena havenot disappeared, it will be necessary to examine the filter sand. The elimination of finesand must stop after 1 or 2 months of activity. If this sand continues to be carriedaway after the first several dozen washings it is necessary to reexamine the hydrauliccriteria of the washing conditions, the granulometry of the filter mass, and the filter'sresistance to shear and abrasion.

Chemical Mixing, Flocculation and Solids Contact Processes

Chemical mixing and flocculation or solids contact are important mechanical steps inthe overall coagulation process. Application of the processes to waste water generallyfollows standard practices and employs basic equipment. Chemical mixing thoroughlydisperses coagulants or their hydrolysis products so the maximum possible portion ofinfluent colloidal and fine supracolloidal solids are absorbed and destabilized.Flocculation or solids contact processes increase the natural rate of contacts betweenparticles. This makes it possible, within reasonable detention periods, for destabilizedcolloidal and fine supracolloidal solids to aggregate into particles large enough foreffective separation by gravity processes or media filtration.

These processes depend on fluid shear for coagulant dispersal and for promotingparticle contacts. Shear is most commonly introduced by mechanical mixingequipment. In certain solids contact processes shear results from fluid passage upward


through a blanket of previously settled particles. Some designs have utilized shearresulting from energy losses in pumps or at ports and baffles.

Chemical Mixing

Chemical mixing facilities should be designed to provide a thorough and completedispersal of chemical throughout the wastewater being treated to insure uniformexposure to pollutants which are to be removed. The intensity and duration of mixingof coagulants with wastewater must be controlled to avoid overmixing or undermixing.

Overniixing excessively disperses newly-formed floe and may rupture existingwastewater solids. Excessive floe dispersal retards effective flocculation and maysignificantly increase the flocculation period needed to obtain good settling properties.The rupture of incoming wastewater solids may result in less efficient removals ofpollutants associated with those solids. Undermixing inadequately disperses coagulantsresulting in uneven dosing. This in turn may reduce the efficiency of solids removalwhile requiring unnecessarily high coagulant dosages.

In water treatment practice several types of chemical mixing units are typically used.These include high-speed mixers, in-line blenders and pumps, and baffled mixingcompartments or static in-line mixers (baffled piping sections). An example of a high-speed mixer is shown in Figure 5. Designs usually call for a 10-30 second detentiontimes and approximately 300 fps/ft velocity gradient. Variable-speed mixers arerecommended to allow varying requirements for optimum mixing.

In mineral addition to biological wastewater treatment systems, coagulants may beadded directly to mixed biological reactors such as aeration tanks or rotating biologicalcontactors.

Based on typical power inputs per unit tank volume, mechanical and diffused aerationequipment and rotating fixed-film biological contactors produce average shearintensities generally in the range suitable for chemical mixing. Localized maximumshear intensities vary widely depending on the speed of rotating equipment or onbubble size for diffused aeration.

Flocculation

The proper measure of flocculation effectiveness is the performance of subsequentsolids separation units in terms of both effluent quality and operating requirements,such as filter backwash frequency. Effluent quality depends greatly on the reductionof residual primary size particles during flocculation, while operating requirementsrelate more to the floe volume applied to separation units.


DRIVE MECHANISM

MOTOR

SUPPORT BEAMS

FEED

Figure 5. Example of an impeller mixer.

Flocculation units should have multiple compartments and should be equipped withadjustable speed mechanical stirring devices to permit meeting changed conditions. Inspite of simplicity and low maintenance, non-mechanical, baffled basins are undesirablebecause of inflexibility, high head losses, and large space requirements. Mechanicalflocculators may consist of rotary, horizontal-shaft reel units as shown in Figure 6.Rotary vertical shaft turbine units as shown in Figure 7 and other rotary or reciprocatingequipment are other examples. Tapered flocculation may be obtained by varying reel orpaddle size on horizontal common shaft units or by varying speed on units with separateshafts and drives,

In applications other than coagulation with alum or iron salts, flocculation parametersmay be quite different. Lime precipitates are granular and benefit little from prolongedflocculation.

Polymers which already have a long chain structure may provide a good floe at lowmixing rates. Often the turbulence and detention in the clarifier inlet distribution isadequate.


INFLUENT

CONTROL VALVE

W.L.

, PADDLES

jr irr^jLEFFLUENT

..&.„. „..«,JJ. JL.

Figure 6. Mechanical flocculation basin horizontal shaft-reel type.

MOTORIZED SPEED REDUCER

WAT Cft PftESSURE tUBRICATCO

Figure 7. Mechanical flocculator vertical shaft-paddle type.


Solids Contacting

Solids contact processes combine chemical mixing, flocculation and clarification in asingle unit designed so that a large volume of previously formed floe is retained in thesystem. The floe volume may be as much as 100 times that in a "flow-through" system.This greatly increases the rate of agglomeration from particle contacts and may alsospeed up chemical destabilization reactions.

Solids contact units are of two general types: slurry-recirculation and sludge-blanket.In the former, the high floe volume concentration is maintained by recirculation fromthe clarification to the flocculation zone, as illustrated in Figure 8. In the latter, thefloe solids are maintained in a fluidized blanket through which the wastewater undertreatment flows upward after leaving the mechanically stirred-flocculatingcompartment, as illustrated in Figure 9. Some slurry-recirculation units can also beoperated with a sludge blanket.

Solids contact units have the following advantages:

1. Reduced size and lower cost result because flocculation proceeds rapidly athigh floe volume concentration.

2. Single-compartment flocculation is practical because high reaction rates and theslurry effects overcome short circuiting.

3. Units are available as compact single packages, eliminating separate units.4. Even distribution of inlet flow and the vertical flow pattern in the clarifier

improve clarifier performances

Equipment typically consists of concentric circular compartments for mixing,flocculation and settling. Velocity gradients in the mixing and flocculationcompartments are developed by turbine pumping within the unit and by velocitydissipation at baffles. For ideal flexibility it is desirable to independently control theintensity of mixing and sludge scraper drive speed in the different compartments.

Operation of slurry-recirculation solids contact units is typically controlled bymaintaining steady levels of solids in the reaction zone. Design features of solidscontact clarifiers should include:

1. Rapid and complete mixing of chemicals, feedwater and slurry solids must beprovided. This should be comparable to conventional flash mixing capability andshould provide for variable control, usually by adjustment of recirculator speed.

2. Mechanical means for controlled circulation of the solids slurry must be providedwith at least a 3:1 range of speeds. The maximum peripheral speed of mixerblades should not exceed 6 ft/sec.

3. Means should be provided for measuring and varying the slurry concentration inthe contacting zone up to 50 % by volume.

4. Sludge discharge systems should allow for easy automation and variation of vol-umes discharged. Mechanical scraper tip speed should be less than 1 fpm withspeed variation of 3:1.

5. Sludge-blanket levels must be kept a minimum of 5 feet below the water surface.

RAPID MIXING AND RECIRCUUTtON SLOW MIXING AND FLOC FORMATION

TREATED WATEREFFLUENT

CHEMICAL INTRODUCTION

\

SLUDGE RECIRCULATION

CLARIFIEDWATER

SEDIMENTATION

r•§'a.

SLUDGE REMOVAL

Figure 8. Solids contact clarifier without sludge blanket filtration.

Application of Filtration to W

astewater

Treatm

ent161


6. Effluent launders should be spaced so as to minimize the horizontal movementof clarified water.

Further considerations include skimmers and weir overflow rates. Skimmers shouldbe provided on all units since even secondary effluents contain some floatable solidsand grease. Overflow rates and sludge scraper design should conform to therequirements of other clarification units. The reader may refer to Chapter 9 forexamples of typical flow sheets and auxiliary filtration equipment schematics,including process flow sheets for chemical feeding operations described above.

Suggested Readings

1. Anon., Water Sewage Works, 6, 266 (1968).2. Maeckelburg, D., G.W.F., 119,23 (1978).3. O'Mella, Ch. R., and O.K. Crapps, J. AWWA, 56,1326 (1964).4. Drapeau, A.J., and R.A. Laurence, Eau Quebec, 10, 314 (1977).5. Burman, N.P., H2O, 11, 348, (1978).6. Cleasby, J.L., J. Arboleda, D.E. Burns, P.W. Prendiville, and E.S.

Savage, J. AWWA, 69,115 (1977).7. Cheremisinoff, P.N., Pollution Engineering Flow Sheets: Waste water

Treatment, Pudvan Publishing Co., Northbrook, IL, 1988.8. Cheremisinoff, N.P., Biotechnology for Waste and Wastewater Treatment, Noyes

Publication, Park Ridge, NJ, 1996.9. Cheremisinoff, N.P. and P.N. Cheremisinoff, Carbon Adsorption for Pollution

Control, Prentice Hall Publishers, Inc., Englewood, NJ ,1993.10. Cheremisinoff, N.P. and P.N. Cheremisinoff, Liquid Filtration for Process and

Pollution Control, SciTech Publishers, Inc., Morganville, NJ, 1981.11. Cheremisinoff, N.P. and P.N. Cheremisinoff, Chemical and Non-Chemical

Disinfection, Ann Arbor Science Publishers, Ann Arbor, MI, 1981.12. Cheremisinoff, P.N. and R.B. Trattner, Fundamentals of Disinfection for

Pollution Control, SciTech Publishers, Inc., Morganville, NJ, 1990.

ADVANCED MEMBRANETECHNOLOGY FOR WASTEWATERTREATMENTIntroduction

This chapter discusses a new membrane filtration system technology based on usinga formed-in-place hyperfiltration membrane. The technology has been used to treat acreosote and pentachlorophenol (PCP) contaminated groundwater. The membranetechnology described can be used as an integral part of a remediation system tosignificantly reduce the volume and toxicity of contaminated wastewater. Thetechnology is particularly suited for the treatment of contaminated groundwater as partof a pump and treat system. The technology reduces risks to human health and theenvironment by transferring the contaminants to a smaller volume facilitatingdestruction or detoxification by other technologies. The technology is particularlyapplicable to the treatment of dilute waste steams, where the concentration of thecontaminants into a reduced volume would result in significant cost savings as well asminimize off-site treatment. The reduced-volume concentrated residual could befurther treated on-site, or transported off-site for treatment and disposal.

The system is simple to operate, reliable and requires a minimum of operator attentionor maintenance once the membrane has been formed. The stability of the systemmakes it particularly suitable for long-term use as is necessary for extended pump andtreat remedial programs.

The information provided in this chapter was largely obtained from a reported studyby the United States Environmental Protection Agency (USEPA) from their Office ofResearch and Development in Washington, DC. The reader may contact the RiskReduction Engineering Laboratory of the Office of Research and Development inCincinnati, Ohio for detailed information on this filtration technology. A specificreference that the reader can refer for detailed information is EPA/540/AR-92/014 -

163


August 1993 - Membrane Treatment of Wood Preserving Site Groundwater by SBPTechnologies, Inc.: Application Analysis Report.

Overview of Technology Case Study

Based on the results of a demonstration project at the American Creosote Works sitein Pensacola, Florida and information concerning other studies provided by thevendor, SBP Technologies, Inc., for different wastes at other sites, severalconclusions can be drawn. The conclusions are organized based on the evaluationfactors of volume reduction and contaminant reduction. These factors are critical inapplying the technology to other sites and wastes. The SBP filtration unit (asconfigured) effectively removed high molecular weight compounds from the feedstream, but smaller molecular weight compounds were not removed. The technologyuses a formed-in-place membrane system which is quite effective (92%) at removingpolynuclear aromatic hydrocarbons (PAHs) found in creosote from the feed water andproducing a permeate with little of these materials.

However, the membrane was found not to be very efficient at removing phenolics.Rejections were in the range of 18% for phenolics. Overall, based on a comparisonof total concentrations of a pre-designated list of creosote-derived PAH and phenolicsemivolatile contaminants in the permeate versus the feed water, the system did notmeet the claimed rejection efficiency of 90%.

On the basis of the PAH rejections of over 90%, the permeate would be expected tobe acceptable for discharge to POTWs (Publically Owned Treatment Works) with littleor no polishing. Other pollutants found in contaminated waters at wood treatmentfacilities (e.g., polychlorinated dioxins and furans) also are concentrated in the rejectstream. Other constituents commonly encountered at such sites including colloidal oilsand suspended solids are also extensively removed by the membrane process. Removalefficiencies for oil and grease were 93%. Suspended solids were removed to non-detectable levels. These materials did not appear to have an adverse effect on thefiltration process.

The system was found to effectively concentrate organic contaminants into aconcentrate of much smaller volume. The volume of wood preserving wastecontaminated wastewater was reduced by over 80%. This means that only 20% of thevolume of the feed water would require further treatment to immobilize or destroy theorganic contaminants.

The filtration unit operated consistently and reliably over a brief testing period. Theunit was easy to operate and maintain. The filtration unit operated in a batch mode forsix hours each day, for six days, and processed approximately 1000 gallons of feed perday. Over the six day test period, permeate flux was relatively constant. Based on atotal membrane area of 300 ft2 for the system, the permeate flow rate for the fourmodule filtration unit averaged 2.6 gpm. Excessive fouling of the membrane,necessitating frequent cleaning or regeneration, was not encountered. However, the

Advanced Membrane Technology for Wastewater Treatment 165

membrane system did exhibit a gradual and controllable fouling which requiredperiodic cleaning.

The operating cost for the membrane process as used at American Creosote Works isin the range of $220 - $1,740/1,000 gallons, depending on system size. Major costcontributors are labor and residuals disposal. Labor costs decrease significantly as thescale of the process increases. Auxiliary equipment that could be needed to supportthis process is comparable to that which would be needed for other above-groundtreatment systems such as oil/water separators and clarifiers for pretreatment, andfilters, carbon adsorbers, etc, for effluent polishing as required.

With membranes similar to those manufactured for the American Creosote Works site,the system could be well suited for the concentration of polynuclear aromatichydrocarbons from wastewaters (groundwater, process wastes, lagoon leakage, etc.)found at coke plants, wood preserving sites, and some chemical plants. Based on theexpected mechanisms of membrane filtration, the technology also may be useful forwastewaters containing other large molecules such as polychlorinated biphenyls(PCBs) and polychlorinated dioxins and furans, particularly where these are associatedwith oil or particulate matter. It probably is also highly effective for oils, colloidalsolids, and greases.

According, to the developer, the formed-in-place membrane can be easily modifiedto conform to waste characteristics and the degree of contaminant removal desired.Therefore, the membrane can be tailored to the unique characteristics of the wastesteam.

Extensive data were collected on primary pollutants (phenols, and PAHs) and onsecondary pollutants (oil, suspended and dissolved solids, COD, dioxins, and VOC's).The results of this project demonstrated the ability of the formed-in-place membrane,operating in a cross-flow mode, to minimize fouling, and to remove polynucleararomatic hydrocarbons from the contaminated feed water. As operated, rejection ofthe PAHs appears to increase with the number of aromatic rings. However, similarcorrelations appear to exist with molecular weight as well as with the partitioncoefficient reflecting hydrophobicity. The permeate, accounting for approximately80% of the feedwater, contained only about 12% of the predominant PAHs,naphthalene and phenanthrene.

The removal of phenol and methyl phenols was not comparably high under theconditions of the demonstration, with an average rejection of 18%. The concentrationsof phenolics in the permeate could present a regulatory problem in the United States,depending on the concentrations in the feedwater and the final disposition of thepermeate. However, the vendor states that different membranes and tubeconfigurations could resolve this.

Secondary constituents, such as oil, suspended solids, and dissolved solids, did notappear to interfere with the operation of the process at the concentrations present inthe waste water studied during the demonstration. Decreases in chemical oxygendemand (COD), total organic carbon (TOC) and oil and grease (O&G) indicated that


the system removes other organic species as well as PAHs, but not necessarily withthe same efficiency.

The SBP membrane process would be most applicable to wastewaters containing largemolecular weight organic compounds (PAHs, dioxins/furans, polychlorinatedbiphenyls, and certain pesticides/herbicides). The system can remove smallermolecular weight compounds (phenols, benzene, toluene, ethylbenzene, xylenes) iflarger molecular weight compounds are not abundantly present. Removal of smallermolecular weight compounds can be accomplished by modifying the structure of theformed-in-place membrane. For these applications, the pores of the membrane arereduced, resulting in higher retentions of smaller components as well as a reduction inthe flux (throughput) of the system. To compensate for the reduced flux, eitheradditional membrane modules can be added or more time will be required to accomplishthe remediation. In either case, the overall cost may be higher.

The system may be most suitable to treating relatively dilute, but toxic, waste streamsin which the percent reduction of contaminants will allow discharge of the permeatewithout further treatment. This feature makes the unit highly suitable for polishingeffluents as part of a multi-technology treatment train. In this system, the primarytreatment technology can be utilized to remove the bulk of the contamination, with thefiltration unit being used as a final polishing step.

A major attribute of the system is its ability to minimize fouling. The systemeffectively controlled excessive fouling, in spite of the problematical nature of thewood preserving waste feed, through a combination of cross-flow operation andmembrane cleaning. The membrane cleaning process effectively regenerated themembrane to its original clean permeate flux conditions. This enabled the membraneto be reused, without the necessity to reformulate.

The ability to repeatedly regenerate the flux after the cleaning procedure is a goodindication that the forrned-in-place membrane is stable and can be used over anextended length of time. In the unlikely event of an irreversible fouling, the membranecan be cost-effectively and easily reformed on-site with a minimum of downtime.

The technology uses a proprietary formed-in-place membrane technique. Themembrane is formed on porous sintered stainless steel tubes by depositing microscopiclayers of inorganic and polymeric chemicals. The properties of the formed-in-placemembrane can be varied by controlling the type of membrane chemicals used, theirthickness, and the number of layers. This important feature allows for customizationof the membrane system to a wide variety of waste characteristics and clean-upcriteria. The formed-in-place membrane can be quickly and economically reformulatedin the field to accommodate changes in waste characteristics or treatmentrequirements.

The formed-in-place membrane is compatible with a wide variety of contaminantsoften encountered in hazardous wastewater streams. The formed-in-place membraneis stable under most chemical environments and will not degrade even at highcontaminant concentrations.


The extent of contaminant reduction required (overall and for individual pollutants)can also be an important factor in system design and operation. This will impactmembrane selection, and operational requirements such as the number of cyclesnecessary to achieve the targeted volume reduction. Generally, as the desired level ofvolume-reduction increases, the overall quality of the permeate decreases, so a balancemust be maintained between throughput and permeate quality. This will also affect thethroughput capability (as permeate) for a particularly sized system.

Other factors that could affect the removal of PAHs or other contaminants may includethe presence of other organics, oil and grease, suspended solids, and dissolved solidsin the feed water. While the levels of such contamination encountered in thedemonstration project had no apparent adverse effect, it is unclear how much rejection(of PAHs) was due to molecular size or weight and how much was due to solubilityin oil that was rejected and coalesced by the membrane. Additional or alternativemechanisms also may be operative.

Case Study Specifics

The EPA's Office of Solid Waste and Emergency Response (OSWER) and the Officeof Research and Development (ORD) established the Superfund InnovativeTechnology Evaluation (SITE) Program in 1986 to promote the development and useof innovative technologies to clean up Superfund sites across the country. The SITEProgram is helping to provide the treatment technologies necessary to meet newfederal and state cleanup standards in the United States that are aimed at permanent,rather than temporary, remedies. The SITE Program is composed of two majorelements: the Demonstration Program and an Emerging Technologies Program. Inaddition, the Program includes research on analytical methods that can expeditecleanups at Superfund sites.

The USEPA demonstration programs are designed to provide engineering and costdata on selected technologies. EPA and the developers participating in the programshare the cost of demonstrating their innovative systems at chosen sites, usuallySuperfund sites. Developers are responsible for the operation of their equipment (andrelated costs). EPA is responsible for sampling, analyzing, and evaluating all testresults and comparing these results to claims originally defined by the developer. Theresult is an assessment of the technology's performance, reliability, and cost. Inaddition to providing the developer with carefully documented information useful inmarketing, the information, in conjunction with other data, also will be used to selectthe most appropriate technologies for the cleanup of other Superfund sites.

Developers of innovative technologies apply to the Demonstration Program byresponding to EPA's annual solicitation. To qualify for the program, a new technologymust have a pilot or full scale unit and offer some measurable advantage over existingtechnologies. Mobile technologies are of particular interest to EPA.

Once EPA has accepted a proposal, EPA and the developer work with the EPARegional offices and state agencies to identify a site containing wastes suitable fortesting the capabilities of the technology. EPA's contractor designs a detailed sampling


and analysis plan that will thoroughly evaluate the technology and ensure that theresulting data are reliable. The duration of a demonstration varies from a few days toseveral months, depending on the type of process and the quantity of waste needed toassess the technology. While meaningful results can be obtained in a demonstrationlasting one week with some technologies, others, may require months. On completionof a demonstration, EPA prepares reports. Ultimately, the Demonstration Programleads to an analysis of the technology's overall applicability to Superfund problems.

The second principal element of the SITE Program is the Emerging TechnologiesProgram, which fosters the investigation and development of treatment technologieswhich are still at the laboratory scale. Successful validation of these technologies couldlead to the development of systems ready for field demonstration. A third componentof the SITE Program, the Measurement and Monitoring Technologies Program,provides assistance in the development and demonstration of innovative techniques andmethods for better characterization of Superfund sites.

In this study it was demonstrated that SBP's membrane technology can be used as anintegral part of a remediation system to significantly reduce the volume and toxicityof contaminated wastewater. The technology is particularly suited for the treatmentof contaminated groundwater as part of a pump and treat system. The technologyreduces risks to human health and the environment by transferring the contaminantsto a smaller volume facilitating destruction or detoxification by other technologies.

The vendor uses a proprietary formed-in-place membrane technology. The membraneis formed on porous sintered stainless steel tubes by depositing microscopic layers ofinorganic and polymeric chemicals. The properties of the formed-in-place membranecan be varied by controlling the type of membrane chemicals used, their thickness, andthe number of layers. This important feature allows for customization of themembrane system to a wide variety of waste characteristics and clean-up criteria. Theformed-in- place membrane can be quickly and economically reformulated in the fieldto accommodate changes in waste characteristics or treatment requirements.

Contaminated feedwater is recirculated through the filtration unit until the desiredlevel of volume reduction is attained. The filtration unit generates two process wastestreams. A relatively clean stream, called the "permeate", passes through themembrane while a smaller portion of the feedwater, retaining those species that do notpass through the membrane, is retained in a stream called the "concentrate". Thepermeate stream should be clean enough for disposal as a non-hazardous waste withlittle or no additional treatment. The concentrate would require further treatment toimmobilize or destroy the contaminants.

Technology Application

This technology lends itself as a means of concentrating organic contaminants inaqueous waste streams. The prime benefit of concentrating contaminants is tominimize costly treatment of the entire wastestream. In addition, by concentrating the


organic contaminants into a smaller volume, alternative treatment technologies maybe feasible based on technical and/or economic criteria.

The ability of the filtration unit to concentrate organic contamination from aqueouswaste streams was demonstrated on a groundwater contaminated with wood preservingwastes (phenolics, PAHs, and PCP). The results from the demonstration, inconjunction with information supplied by the vendor, were used to assess theapplicability of the technology for a variety of waste types and site conditions.

The process uses a formed-in-place hyperfiltration membrane on a stainless steelsupport to separate and concentrate higher molecular weight contaminants.Contaminated groundwater (feed) is pumped through the modules under pressure. Aportion of the feed passes through the formed-in-place membrane forming a permeate.The membrane retains certain contaminants resulting in a permeate that is cleanrelative to the feed. The bulk of the contamination remains in the "concentrate"fraction. The concentrate is recycled through the unit until the desired concentrationor level of volume reduction is attained, or the level of contaminants in the recyclingconcentrate inhibits the filtration process (fouling). The system relies on cross-flowfiltration to minimize fouling of the membrane and, thus, maximize throughput.

The properties of the two process streams (permeate and concentrate) are of particularimportance since these characteristics define waste disposal options. The permeatestream should exhibit significant reductions in contamination so as to allow economicaldischarge to local wastewater treatment facilities without extensive pretreatmentrequirements. The concentrate stream should be volumetrically small, relative to theoriginal feed, in order to minimize the volume of waste requiring further treatmentprior to disposal. Furthermore, the filtration process should enable the use ofadditional disposal options for the concentrate (as compared to the raw feed).

The following subsections summarize observations and conclusions drawn from thereported study. Included in the discussion are factors such as the application ofmembrane processes for wastewater reduction, benefits of the system, other applicablewaste waters, site characteristics and constraints, and unique handling requirements.

Mechanisms of Membrane Separations

Membranes are semi-permeable barriers that are used to isolate and separateconstituents from a fluid stream. The separation process can be accomplished througha number of physical and chemical properties of the membrane as well as the materialbeing separated. Separation can occur through processes such as size, ionic charge,solubility, and combinations of several processes. Membranes can remove materialsranging from large visible particles to molecular and ionic chemical species.Membrane materials are diverse and can consist of synthetic polymers, natural fabrics,porous metals, porous ceramics, or liquids. The surface of the membrane can bechemically or biologically altered to perform separations on specific chemical


compounds. The interaction of the components of the fluid stream with the membraneis the mechanism controlling the outcome of the separation process.

There are two basic modes of membrane separation. In dead-end filtration specificspecies are trapped within the matrix of the membrane material. The membrane"filters-out" these species producing a relatively clean effluent. In dead-end filtrationthe components that are trapped are usually not recovered and remain within themembrane matrix. In addition, the membrane eventually becomes pluggednecessitating the replacement of the membrane. Dead-end filtration is principallyutilized to purify a fluid in applications where the removed species is relatively dilute.

In cross-flow filtration the fluid steam is directed parallel to the surface of themembrane. This action inhibits the accumulation of components within the matrix ofthe membrane. The cross-flow action of the fluid keeps the surface of the membraneclean allowing for the passage of species smaller than the pores of the membrane.Cross-flow filtration produces two effluent streams. The permeate is the steam thatpasses through the membrane and is relatively depleted in species larger than the poresize of the membrane. The concentrate is the cross-flow stream that contains the largerspecies that are unable to pass through the membrane and accumulate. The concentratecan be recycled allowing for progressive concentration of species over time. Due tothe ability of the cross-flow system to concentrate components from the feed stream,it is commonly used as a method to separate and recover these components.Furthermore, the cross-flow action minimizes plugging of the membrane (fouling) byconstantly sweeping the membrane's surface. This cleaning action extends the life ofthe membrane and minimizes degradation of flow through the membrane.

Membrane systems have many applications for the pretreatment and treatment ofhazardous wastes. Membrane separation is a volume reduction technology. Thistechnology can separate and concentrate specific contaminants from a waste stream,resulting in a significant reduction in the volume of waste requiring treatment. Theconcentrated contaminants can then be destroyed or rendered non-toxic. The utility ofa membrane based technology is based on its ability to reduce the volume of waste byremoving contaminants from the feed stream and producing an effluent stream thatwould require little or no further treatment. The greater the volume reduction, themore effective the technology is in reducing ultimate disposal costs. However, thereis a balance between the magnitude of the volume reduction, the quality of the effluentstream, and the size and operation of the unit. A higher volume reduction wouldrequire additional recycling, reducing the overall flow through the system. In addition,higher levels of contaminant removal will usually result in lower fluxes through themembrane requiring either more membrane area or longer processing time. Thebalance between throughput and effluent quality is dictated by clean-up standards andtreatment costs. This balance will impact such factors as the size and type of theequipment, mode of operation, time required for remediation, treatment requirementsfor the permeate, and ultimate disposal mechanism for the concentrated contaminants.

171

Membrane processes have many applications in the treatment of contaminated wastestreams. The most common applications involve the removal and concentration oforganic and inorganic contaminants from liquid waste streams. The waste steams canoriginate from industrial processes, contaminated groundwater, contaminated surfacewater bodies, or as by-products of other treatment processes.

Membrane and filtration processes have historically been utilized for the treatment andpurification of drinking water. For this application, filtration is used to remove a widevariety of constituents, ranging from visible particulates (sand filters, refer to Chapter5) to ionic species (reverse osmosis, refer to Chapter 4). From these conventionalapplications, new uses of membrane separations have recently been applied to thetreatment of hazardous waste streams.

Membranes can be used to separate and concentrate organic contaminants from wastestreams. In these applications, the organic contaminants are removed based on theirsize (molecular weight) or polarity. Size separations rely on membranes with specificpore size distributions. The smaller the pores, the greater will be the removal of smallmolecular weight compounds. However, as the membrane's pore size decreases, theflux (flow per unit membrane area) also decreases impacting the overall economicsand efficiency of the process. The polarity of an organic constituent is a measure ofit's ability to ionize in solution. Examples of polar molecules are water, alcohols, andcompounds with hydroxyl (e.g. phenols) and carboxyl groups (e.g. organic acids).Aliphatic hydrocarbons and polynuclear aromatic hydrocarbons are examples on non-polar organic molecules. The chemical characteristics of the membrane can be usedto separate non-polar constituents in a waste stream from polar constituents. Forexample, a membrane whose surface is hydrophilic will allow passage of polarcomponents while retaining the non-polar components. These membranes can be usedto separate dissolved and emulsified oils from aqueous waste streams.

Inorganic contaminants, such as salts and heavy metals, can be removed andconcentrated from waste streams by membrane processes. Suspended inorganics canbe easily removed through the use of microfiltration membranes. These membraneshave pore sizes ranging from as low as 0.01 up to several microns. Dissolvedinorganics can be removed either through the use of hyperfiltration (reverse osmosis)membranes, or by precipitation followed by microfiltration. Conventional reverseosmosis membranes may require extensive prefiltration to avoid fouling, and thereforecan only be used on relatively clean feed solutions. Chemical precipitation, followedby microfiltration, allows for the use of microfilters which exhibit higher fluxes andare not as sensitive to fouling. Membrane processes can be helpful in solving manyremediation problems at hazardous waste sites.

Contaminated Groundwater

Containment and/or remediation of contaminated aquifers typically utilizes pump andtreat technologies to control contaminant plume migration and ultimately restore the


quality of the groundwater. The recovered groundwater usually requires treatmentprior to discharge. Treatment alternatives for the recovered groundwater aredependent on the nature and extent of the contamination. Membrane systems can beeffectively used to significantly reduce the quantity of groundwater requiring costlytreatment.

The contaminants of concern can be isolated and concentrated into a reduced volumewhich can be more easily handled. Another potential benefit of the concentrationprocess is that additional destructive treatment alternatives may become feasible. Forexample, the concentration of hydrocarbons from a contaminated groundwater canproduce a reduced volume waste with a high BTU value allowing for fuel blending asa disposal alternative. This not only reduces the quantity of groundwater that must betreated, but also produces a more easily treatable final waste product. As anotherexample, heavy metals can be concentrated from an aqueous stream by membraneprocesses and immobilized by solidification/stabilization technologies.

Membrane processes can be potentially used to recover organic and inorganicconstituents for recycle/reuse. In these applications, the separation scheme must bedeveloped to produce a high quality concentrate.

Membrane processes can be applied to the removal of many organic contaminantsfrom waste streams. Organic contaminants that can be removed include petroleumderived hydrocarbons (benzene, toluene, ethylbenzene, xylenes), polynuclear aromatichydrocarbons, PCBs, dioxins/furans, pesticides, and chlorinated hydrocarbons.Generally, membrane process are more easily applied to removing larger molecularweight, non-polar organic components because larger pored membranes can be utilizedand surface chemistry interactions can augment size separations.

Removal of hazardous inorganic species from contaminated groundwater requires adetailed knowledge of the water chemistry in order to optimize the separation. In manycases, addition of precipitating chemicals must be added in order to induce particulateformation. Furthermore, groundwater containing high concentrations of innocuousinorganic constituents such as iron and divalent cations (e.g., potassium and calcium)may compete with and interfere with the removal of toxic heavy metals. Conventionalreverse osmosis membranes are fragile and must be protected from the corrosivenature of many highly contaminated aquifers.

Integration with Other Technologies

Membrane processes are particularly amenable to integration with other remedialtechnologies enabling applications to additional waste matrices. Ease of integration isfacilitated by the modular and scalable properties of membrane systems. Thesesystems can be readily integrated with other remedial process equipment to enhancethe effectiveness and economy of these systems.

Membrane processes can be used as a final polishing tool for remedial technologiesinvolving discharge of process water. In this capacity, the membrane system is utilizedto remove contaminants from a relatively dilute waste stream. The benefit of using this


polishing step is to avoid costly overdesign of the primary remedial technology. Forexample, a membrane system can be implemented as a final polishing step on abioreactor. The bioreactor can be designed to cost-effectively treat the bulk of theorganic contamination, while the polishing membrane can be designed to treat theaqueous phase prior to discharge.

Membrane processes can be used as a pre-treatment step for other remedialtechnologies. The purpose of the pretreatment would be to concentrate thecontaminants to a level that is amenable for specific remedial technologies. Forexample, organic contaminants in dilute aqueous streams (e.g., groundwater, leachate)can be concentrated to a level that could support an efficient biomass forbioremediation technologies.

Membranes can be integrated with remedial technologies as a component in theprocess. For example, membranes can be used to recycle and recover extraction fluidsused to concentrate organic and inorganic contaminants in soil extraction technologies.

Features of the Hyperfiltration System

The hyperfiltration system has several unique features which provides advantages overconventional membrane processes in wastewater treatment applications. Thetechnology uses a proprietary formed-in-place membrane technique. The membraneis formed on porous sintered stainless steel tubes by depositing microscopic layers ofinorganic and polymeric chemicals. The properties of the formed-in-place membranecan be varied by controlling the type of membrane chemicals used, their thickness, andthe number of layers. This important feature allows for customization of themembrane system to a wide variety of waste characteristics and clean-up criteria. Theformed-in-place membrane can be quickly and economically reformulated in the fieldto accommodate changes in waste characteristics or treatment requirements.

Conventional membranes rely on rigid polymeric, ceramic, or porous stainless steelmembranes. These membranes are available in discrete pore sizes and cannot becustomized to the characteristics of the feed. Furthermore, once installed on-site it isdifficult and costly to modify their separation properties in response to variable feedcharacteristics.

The formed-in-place membrane is compatible with a wide variety of contaminantsoften encountered in hazardous wastewater steams. Many conventional reverseosmosis membranes are made from materials such as cellulose acetate and exhibit poorcompatibility with reactive substances often encountered in hazardous wastes. Theseconventional membranes will degrade and become inoperative when challenged withmany organic compounds. The compatibility problem becomes more critical as thelevel of concentration increases. The formed-in-place membrane is stable under mostchemical environments and will not degrade even at high contaminant concentrations.


A major limitation of many membrane systems is their propensity to irreversibly foul.Fouling is the uncontrolled build up of materials on the surface of the membrane.Fouling leads to a loss of flux and eventually results in cessation of flow. If amembrane fouls, it must be cleaned in order to restore flux. If cleaning isunsuccessful, then the membrane is replaced.

The technology discussed in this chapter utilizes a cross-flow filtration mechanism tocontinuously clean the surface of the membrane, hence minimizing fouling. In thismode, the feed stream is directed parallel to the membrane's surface resulting in acleaning action which minimizes the buildup of materials on the membrane's surface.

Since all membranes eventually foul, a cleaning cycle is necessary to restore flux andoperability. Many membrane systems have limited abilities to be regenerated due torestrictions in the choice of cleaning chemicals. The formed-in-place membrane iscompatible with a wide range of chemical cleaning methods, enabling in-placeregeneration of flux. In situations where the membrane becomes irreversibly fouled,the formed-in-place membrane can be stripped and reformulated on-site.

The membrane technology can be used as an integral part of a remediation system tosignificantly reduce the volume and toxicity of contaminated wastewater. Thetechnology is particularly suited for the treatment of contaminated groundwater as partof a pump and treat system. The technology reduces risks to human health and theenvironment by transferring the contaminants to a smaller volume facilitatingdestruction or detoxification by other technologies.

The system is simple to operate, reliable and requires a minimum of operator attentionor maintenance once the membrane has been formed. The stability of the systemmakes it particularly suitable for long-term use as is necessary for extended pump andtreat remedial programs.

The demonstration at the American Creosote Works was designed to evaluate the twomost critical process parameters for membrane systems; volume reduction andcontaminant reduction. A summary of the demonstration results for these criticalprocesses parameters are presented below. A discussion of the demonstration resultsand process performance, as they relate to applicability to other wastes and sites alsofollows.

The claim that the system can be operated to recover 80% of the feedwater volume aspermeate was achieved in the demonstartion program. Average water recovery(volume reduction) for the first five runs was 83 %. The volume reduction for theextended ran was 96%, and represents the maximum volume reduction capability ofthe unit for the waste steam tested.

The process did not achieve the developer's claim of 90% overall removal of thesemivolatiles present in the feedwater (on the average, a 74% reduction wasachieved). However, the process does effectively remove polynuclear aromatichydrocarbons from the feedwater and place them in the concentrate. Overall, removal


of polynuclear aromatic hydrocarbons averaged 92%. Removals of individual PAHsrange from 78% to well over 94% for individual two, three, and four ring PAHs,

Other high molecular weight pollutants, such as oils and dioxins, are also rejectedfrom the permeate with high efficiency (93% for oils and >99% for dioxins).However, removal of low molecular weight phenols is much less effective, with valuesbetween 15 and 21%.

Depending on how a system is used, i.e., level of volume reduction and quality ofpermeate, operating plus capital cost could be as low as $200/1,000 gallons. Capitalcost for an averaged size system is approximately $300,000.

The demonstration was designed to evaluate the innovative features of process as avolume reduction technology. The demonstration took place at the American CreosoteWorks in Pensacola, Florida and utilized groundwater contaminated with creosote andpentachlorophenol. Creosote was chosen as a testing material for two reasons.

1. Creosote is a complex mixture of over 250 individual compounds, dominatedby polynuclear aromatic hydrocarbons and phenolics, and exhibits a wide rangeof chemical and physical properties. The wide molecular weight distributionof the organic contaminants is an excellent challenge material for a membraneprocess, allowing for analysis of removal efficiencies over a wide range of feedcharacteristics.

2. Wood preserving waste contaminated aquifers represent a significant andwidespread environmental problem. Results from this demonstration could bedirectly applicable to other wood preserving waste sites.

A pumping well recovered the creosote and contaminated groundwater from the site.The groundwater, which contained aqueous and dense free product fractions, wasallowed to settle and the aqueous phase retained for the study. The aqueous phase wasdiluted with carbon-treated potable water in order to adjust the concentration of thesemivolatiles in the feed to fully test the concentrating capabilities of the filtration unit.

The utility of a membrane system is its ability to remove contaminants from awaste water stream and concentrate them into a reduced volume. The contaminantreduction is the percent decrease in specific contaminants from the feed to thepermeate (discharge). The higher the percent contaminant reduction, the moreeffective is the membrane at removing contaminants from the waste steam.

It is important to note that the applicability of the technology cannot be made solelyon the percent contaminant reduction. Since contamination is reduced as a percentageof the concentration in the feed, the quality of the permeate is dependent on feedconcentrations. In order to assess applicability, the predicted quality of the permeatecan be estimated by calculating contaminant reductions from the feed. The estimatedpermeate quality can then be compared to site specific discharge standards.


For the demonstration, the total concentrations of semivoiatile contaminants for eachrun are summarized in Table 1 for the feedwater and permeate. The system wasevaluated by comparing the total concentrations of these compounds in the feedwateragainst the permeate. Over the six day period, an average overall rejection of 74%was achieved. Thus, starting with a feedwater containing on the average 90 mg/L oftotal designated semivoiatile components, the composited permeate, accounting for80% of the original feedwater volume, contained on the average 23 mg/L. This didnot meet the vendor's claim for 90% removal, largely because of the notedinefficiency with phenolics. This is not totally unexpected since the membrane, asformulated, was not expected to remove species with molecular weights less than 200.

Table L Feed and permeate semivolatiles — total concentration and contaminant reduction.

Run 1

Run 2

Run 3

Run 4

Run5

Run 6

Total Semivoiatile Concentrations (mg/L)

Feed

104

91

92

104

85

60

Permeate

18

24

26

22

23

24

ContaminantReduction (%)

83

74

72

79

73

60

A summary of the average concentrations for individual semivoiatile compounds in thefeed and permeate, along with the associated rejections, for the six day demonstrationare presented in Table 2. The results of the demonstration indicated that the pilot unitwas capable of removing over 94% of some PAHs but only 15 - 21 % of the phenolics.The permeate generated during the process was discharged directly to the local POTW(publically owned treatment works).

These results indicate, as expected, that the membrane is more effective in removinglarger molecular weight components (PAHs) than the smaller molecular weightmolecules (phenolics). With a complex feed such as creosote, it is difficult to achievehigh reductions of all components and at the same time deliver adequate throughput.In this application, the membrane was formulated to maximize reduction of the moretoxic polynuclear aromatic hydrocarbons. Passage of the phenolic compounds into thepermeate did not pose a significant disposal problem since the local POTW couldaccept the phenols in their treatment system. At other sites, careful attention shouldbe made to local discharge requirements and available treatment facilities.


Table 2. Individual semivolatile concentration and rejections (average of six daily runs).

Analyte

Phenol

2-Methyl phenol

4-Methyl phenol

2,4-Dimethyl phenol

Benzole Acid

Pantachlorophenol

Naphthalene

2-Methyl Naphthalene

Acenaphthylefle

Acenaphthene

Dibenzofuran

Fluorene

Phenanthrene

Anthracene

Fluoranlhene

Pyrene

Benzo(a)anthracene

Chrysene

Benzo(b)fl uoranthene

Benzo(k)fluoranthene

Benzo(a)pyrene

Feed

4.90

2.31

6.92

1.82

(1.42)

(2.42)

12.87

4.52

(0.14)

6.84

4.88

5.92

17.08

1.98

7.01

4.70

1.24

1.13

(0.46)

(0.43)

(0.31)

Permeate

3.88

1.93

5.75

1.54

2.16

1.88

2.87

0.46

(0.02)

0.57

0.41

0.37

0.59

0.07

0.10

0.05

*0.03

"0.03

*0.03

*0.03

'0.03

Rejection

20.8

16.5

16.9

15.4

-

»

77.7

89.8

*

91.7

91.6

93.8

96.6

96.5

98.6

98.9

>97.6

>97.4

»

»

Values in parentheses represent analytes with estimated values that are above instrument limits but belowquantitation limits.

Analytes not detected are presented by an *, and the values represent one-half the quantitation limit.;t Individual rejections not calculated due to estimated values.

This type of membrane process would be most applicable to wastewaters containinglarge molecular weight organic compounds (PAHs, dioxins/furans, polychlorinatedbiphenyls, and certain pesticides/herbicides). The system can remove smallermolecular weight compounds (phenols, benzene, toluene, ethylbenzene, xylenes) iflarger molecular weight compounds are not abundantly present. Removal of smallermolecular weight compounds can be accomplished by modifying the structure of the


formed-in-place membrane. For these applications the pores of the membrane arereduced, resulting in higher retentions of smaller components as well as a reductionin the flux (throughput) of the system. To compensate for the reduced flux, eitheradditional membrane modules can be added or more time will be required toaccomplish the remediation. In either case, the overall cost may be higher.

The system may be most suitable to treating relatively dilute, but toxic, waste streamsin which the percent reduction of contaminants will allow discharge of the permeatewithout further treatment. This feature makes the unit highly suitable for polishingeffluents as part of a multi-technology treatment train. In this system, the primarytreatment technology can be utilized to remove the bulk of the contamination, with thefiltration unit being used as a final polishing step.

If the concentration of contaminants in the permeate does not meet clean-uprequirements, then the permeate can be recycled back through the membrane toachieve the targeted effluent quality. Recycling of the permeate has the disadvantageof requiring additional membrane modules, or additional time, both of which increasetreatment costs.

A number of mechanisms could explain the contaminant reduction results, includingrejection by the membrane on the basis of molecular weight or molecular size,rejection and coalescence of dispersed oil in which specific components are soluble,or even rejection simply by adsorption of the PAHs on inert suspended solids.

Examination of the results for the conventional parameters tested in the feed andpermeate (Table 3) provides some insight into the separation mechanism. Highconcentrations of oil and grease found in the feedwater suggests that considerable oilremained in a dispersed or colloidal form. This oil would be removed by a membranewith ultrafiltration or hyperfiltration characteristics. Since the PAHs are more solublein oil than in water, concurrent removal of the PAHs entrained within the oil may haveoccurred. The phenols with relatively high solubility in water are, also as expected,removed more poorly. This also is reflected in the poor rejections calculated for TOCand COD. Other contaminants, not quantified by the semivolatile analysis, also maycontribute to the high TOC and COD in the permeate.

Tables 3. Conventional parameters (values are averages of six runs).

Analyte

IDS

TSS

OIL/GREASE

TOC

COD

Feed

237

34

191

121

379

Permeate

190

<4

14

92

35

Rejection %

20

>88

94

24

2 7


Vo/ume Reduction

The utility of a membrane separation system for treating hazardous waste streams isalso dependent on the magnitude of volume reduction. The volume reduction is ameasure of the percent of the feed water that can be generated as cleaner permeate.The higher the volume reduction, the greater the potential utility of the membranesystem. Volume reduction cannot be solely used as an indicator of membraneperformance. The quality of the permeate must also be considered when evaluating theapplicability of the technology. A high volume reduction with low permeate qualityis not acceptable since the permeate will not be dischargeable and will require furthertreatment. When designing a membrane separation system, volume reduction andpermeate quality must be balanced in order to develop a cost-effective treatmentmeeting site-specific clean-up criteria.

For the demonstration at the American Creosote Works an 80% volume reduction wasachieved each day. This level of volume reduction was set as a target prior to thedemonstration and was easily attained. The level of volume reduction was achievedby continuously recirculating the concentrate through the system. On the last day ofoperation the process was allowed to run until the unit could no longer function,representing the maximum volume reduction for that feed. The maximum volumereduction was 96%.

The relationship between volume reduction and permeate quality is exemplified byresults from the demonstration. During the demonstration, were grabbed samples ofthe permeate steam, were collected at the beginning, middle, and end of each run. Thepurpose of these samples is to document changes in permeate quality during the courseof the batch filtration. The analysis of the data reveals an increase in total semivolatilecontent of the permeate from the beginning to the end of each run. Six day averagepermeate concentrations of total semivolatiles were 19.24 mg/L at the beginning of therun, 24.17 rng/L in the middle, and 29.95 mg/L at the end of the run. In addition, onday six, when the unit was allowed to ran to a maximal volume reduction of 96%, thefinal permeate semivolatile concentration was 47.25 mg/L.

These changes in permeate quality during the filtration are due to increasingsemivolatile contents of the recirculating concentrate. As the batch filtration proceeds,the surface of the membrane is challenged with progressively higher concentration ofcontaminants. Since the membrane can only reject a certain proportion of the feedstream, the concentration of contamination in the permeate will increase.

When applying a membrane solution to a wastewater problem it is crucial to evaluatethe balance between permeate quality and volume reduction. Maximizing volumereduction is important since it impacts economics by minimizing the volume ofwastewater requiring treatment. However, the quality of the discharged water iscritical and must be maintained during the filtration process. Treatability testing isnecessary to determine the optimal balance between permeate quality and volumereduction.


Fouling Control

Fouling is the loss of flux due to the buildup of components on the surface of themembrane. All membranes exhibit some degree of fouling and eventually requirecleaning to restore flux. Many membranes foul readily and are not amenable tocleaning for flux restoration. If flux cannot be restored, then the membrane must bereplaced resulting in considerable expense and downtime.

A major attribute of the technology is its ability to minimize fouling. The processeffectively controlled excessive fouling, in spite of the problematical nature of thewood preserving waste feed, through a combination of cross-flow operation andmembrane cleaning. Flux and pressure data collected during the demonstrationindicated gradual and slight fouling of the membrane. This slight fouling was reversedafter each two-run cycle by a membrane cleaning procedure. Analysis of thewashwaters from the cleaning process indicated that approximately 8% of the mass ofsemivolatiles remained in the system and were removed during the washing process.The membrane cleaning process effectively regenerated the membrane to its originalclean permeate flux conditions. This enabled the membrane to be reused, without thenecessity to reformulate.

The ability to repeatedly regenerate the flux after the cleaning procedure is a goodindication that the formed-in-place membrane is stable and can be used over anextended length of time. In the unlikely event of an irreversible fouling, the membranecan be cost-effectively and easily reformed on-site with a minimum of downtime.

Operational Reliability and Implementability

Operational reliability and implementability are important in deciding the applicabilityof the technology to other waste streams and sites. The system proved to be quitestable and required a minimum of attention over the demonstration period. Systemperformance was relatively constant during the six day test. With feed concentrationsof total semivolatiles ranging from 60.4 - 103.8 mg/L, the percent rejection averaged74%, with a narrow standard deviation of 7.5. Other than adjustment of the pressureto maintain flux and the cleaning of the unit, which consumed about 2 hours everyother day, there was little need for an operator. In a commercial installation somemeans of on-line monitoring (e.g., changes in pressure, contaminant concentration,etc.) could alert the operator to out-of-specification operation or out-of-compliancepermeate. It is estimated that the unit could be run by two operators (health and safetyrequirements). Additional units could easily be operated by the existing personnel.

Other than the cleaning operation every other day, there was no downtime during thedemonstration. With the exception of the pump there are no moving parts to breakdown or require service.

The process equipment and supplies for the system are commercially available. Thisincludes the filtration modules, membrane forming chemicals, pumps, tanks, processcontrols, gauges, and flowmeters.


The membrane formation procedure requires a high level of expertise and may requiretrial and error methods to achieve the desired separation characteristics. However,these are not obstacles to implementation since the process is inexpensive and rapid.

The process is easily scalable and can be modified by adding or deleting modules inresponse to processing requirements. The addition of modules does not affect the modeof operation, except for additional support equipment (pumps, tanks, plumbing).

Based on the observations from the demonstration, it is feasible that the membranesystem can be effectively and reliably operated over an extended time period as wouldbe necessary for pump and treat remediations.

Applicable Wastes

Although the hyperfiltration unit was limited to a single wastewater study for thegroundwater available at the American Creosote Works site, the results of the studyalong with other results provided by the vendor suggest that the technology would haveapplicability to other contaminated groundwaters and process waters. The developerbelieves the system can teat wastes with 100 - 500 mg/L of COD where the molecularweight of the contaminants to be concentrated are over about 200. However, thecharacteristics of the membrane can be modified to treat smaller molecular weightcompounds. More dilute feedwater will necessitate additional cycles to achieve thedesired concentrations in permeate and concentrate streams. However, the more dilutefeedwaters would also allow for higher fluxes. Other than having an impact on costand throughput, this should not adversely affect operation.

Waste streams exceeding the target concentration range (100 - 500 mg/L COD) wouldrequire reduced cycling to achieve the required level of concentration. The effect ofelevated feedwater concentrations on the rejection of individual components may alsoneed to be determined by laboratory testing. Data from this study indicates a reductionin permeate quality as the concentration of the feed increases.

Groundwater rich in PAHs would probably be suitable while feedwater where smallermolecular weight compounds are a major pollutant would probably not be appropriatefor this technology. However, membranes could be formulated to separate smallmolecular weight species (BTEX) such as those found in hydrocarbon contaminatedwaste waters.

Cross-flow filtration using the formed-in-place membrane may also be applicable toother waste streams containing different high molecular weight organic contaminants.This might include polychlorinated biphenyls (PCBs) as might be encountered froma spill from a PCB transformer leak, particularly since the same preferential solubilityin oil noted earlier may prevail. On the same basis, the system may be useful forseparating other emulsified or dispersed organics which do not lend themselves tosimple physical phase separation. The system is also well suited to significantly reducethe concentration of dioxins and furans in wastewater. Reduction of dioxins/furansencountered in this demonstration was greater than 99.9%.


The developer believes the membrane can be customized to achieve different rejectioncharacteristics that could be applied to a wide range of contaminants.

Site Characteristics of the Pilot Demonstration

The pilot-plant unit used in the demonstration program required a level base largeenough to accommodate the unit, and storage tanks for the feed, concentrate, andwash water. A covered concrete pad is recommended to protect the equipment fromthe elements as well as contain the accidental release of contaminated materials. Cleanwater and power are the only utilities needed. If necessary, the relatively small amountof clean water needed for washing of the membrane can be trucked in and power forthe compressor can be provided by an on-site generator. While it was not studied, itmay be practical to use permeate for washing. Where the unit is being used to treatgroundwater, power also would have to be provided for the well pumps.

Acquisition of groundwater for the unit may require the development of an extractionwell network, consisting of the appropriate pumps, regulators, and plumbing. Permitrequirements and the mode in which the filtration unit is operated may make itnecessary to have additional space for storage tanks for equalization of the permeateuntil analyses can confirm acceptability for the POTW or surface water bodydischarge.

Materials Handling Requirements

Materials handling requirements for the unit involve 1) the acquisition of feed materialfor the unit, 2) pretreatment, and 3) residuals (permeate and concentrate) management.

If the filtration unit is part of a system used to treat groundwater, the first need is awell drilling rig to provide the well or wells from which the feedwater is to beobtained. Once the wells are drilled and developed, each must be equipped with apump to draw up the necessary feed water. Local well drilling requirements wouldhave to be taken into consideration.

At some sites pretreatment may be necessary to remove free oil and even suspendedsolids. Since the developer has indicated that the filtration unit is most effective whenoperating with a feed water having a COD range of 100 - 500 mg/L and is mosteffective in rejecting materials with molecular weights greater than about 200, pre-testing will be necessary to assure that these requirements are consistently met. If thevendor's system is provided with relatively clean ground or process water, nopretreatment may be necessary.

The applicability of this membrane technology at a site is dependent on the quality ofthe permeate, site-specific discharge criteria, and the availability and accessibility oflocal public or industrial waste water treatment facilities. It is important to conduct atreatability study to assess the quality of the permeate and to determine options fordisposal. If the permeate quality is not amenable for discharge to surface waters orlocal treatment facilities, then the technology is not applicable to the site.


Prior to the initiation of the demonstration at the site, a limited filterability test wasconducted on contaminated groundwater to determine if the permeate would beaccepted by the local POTW. The permeate was subjected to biological testing(Ceriodaphnia) and chemical analysis to determine its suitability for discharge. Thepermeate passed the local POTW's criteria and was directly discharged to a localsewer hook-up.

Several additional options are available for permeate disposal and are dependent onwaste and site conditions, as well as local discharge regulations and treatment options.The permeate quality may meet local standards for direct discharge to local surfacewater bodies. This would occur only if the level of contaminants in the permeate wasextremely low and meeting the strict requirements for surface discharge.

The permeate could be treated on-site with additional treatment equipment to reducecontaminant levels for either surface water body discharge or sewer discharge.Treatment, such as with activated carbon, may be necessary to reduce contaminationto acceptable limits. The use of additional treatment equipment will increaseremediation costs and may necessitate additional disposal requirements.

The permeate may be recycled through the filtration unit, or processed through asmaller unit, to further reduce contaminants for surface water body or sewerdischarge. The secondary filtration unit may have different membrane characteristicsas the primary unit to remove species that were not retained or require greaterreductions. This option would also add to the overall cost of the remediation sinceadditional equipment and time would be required.

If it is not feasible to reduce contaminant concentrations to levels adequate for on-sitedischarge, and if no local sewer hook-up is accessible, then it may be necessary totransport the permeate by tanker truck to an acceptable treatment facility. This optionwould only be economically feasible if the membrane process drastically reduced thevolume of a waste stream that is very costly or difficult to teat (e.g., dioxincontaminated wastewater).

Concentrate Disposal Options

The membrane process minimizes the quantity of waste requiring extensive treatmentby concentrating the contaminants into a reduced volume while producing a cleanerpermeate for discharge. Since the contaminants are not destroyed by the process it isnecessary to consider disposal options for the reduced volume concentrate stream. Ifthe treatment options for the concentrate steam do not reduce overall treatment costsor provide a reduction in risk to human health and the environment, then themembrane system is not a feasible remedial technology. Optimally, a disposal optionthat can permanently destroy or immobilize the contaminants in the concentrate streamon-site is preferable to off-site transportation and disposal.

A portion of the concentrate from the demonstration was utilized to develop abioremediation technology that could be coupled to the filtration unit to produce atreatment system for on-site destruction of a major portion of the waste. The system


uses a two-stage bioreactor containing several naturally occurring strains of soilbacteria capable of mediating PAH contamination. The membrane system is used toreduce the quantity of wastewater input into the bioreactors and to optimizecontaminant concentrations to support the biomass. The use of the concentrate as afeed to the bioreactors extends the utility of this volume limiting technology byreducing the volume of wastewater that must be processed, therefore reducingequipment costs and site space requirements.

The concentration process enhances the calorific value of most organic wastes. Thisenables the utilization of thermal technologies as a means of destroying the organiccontaminants. The feasibility of using and choosing a thermal technology is based onthe nature of the organic contaminants. Concentrates from petroleum basedcontamination could be readily used for fuel blending, while concentrates from othersources (such as wood preserving wastes) would require careful testing to determineselection of the appropriate thermal technology. Thermal destruction could beaccomplished on-site (mobile units) or transported off-site.

Concentrates containing highly toxic constituents, such as PCBs and dioxins/furans,which are not amenable to biodegradation or thermal treatments, can be chemicallyneutralized by processes such as dechlorination. The neutralized waste could then bedisposed of in a conventional manner.

Process Economics

The primary purpose of this economic analysis is to estimate costs (excluding profit)for commercial-scale remediation using the filtration unit. With realistic costs and aknowledge of the basis for their determination, it should be possible to estimate theeconomics for operating similar-sized systems at other sites utilizing scale-up costformulas. Among such scale-up cost formulas for chemical process plant equipmentis the "six-tenths rule". The six-tenths rule is an exponential method for estimatingcapital costs from existing equipment costs. If the cost of a piece of equipment of sizeor capacity q, is Cj, then the cost of a similar piece of equipment of size or capacityq2 can be calculated from:

The value for n in this discussion is taken as 0.6.

It is assumed that the performance of commercial-scale equipment will be the sameas that demonstrated.

Cost figures provided here are "order-of-magnitude" estimates, and are representativeof charges typically assessed to the client by the vendor, exclusive of profit. The totalannual cost to operate a 12-module filtration unit ranges between $514,180 and$1,209,700, depending on whether effluent treatment and costs are considered, theflow rate through the unit, the cleanup requirements, and the cost of effluent treatment

185

and disposal (if required). Effluent treatment and disposal costs, if considered, couldaccount for up to 60% of the total cost. Labor can account for up to 40% of totalannual costs. Processing costs are more dependent on labor costs than equipmentcosts. The cost per 1,000 gallons can be broken down by flow rate as follows (for withand without effluent treatment and disposal costs):

With Effluent Treatment Costs

24gpm 12gpm 7.2 gpm$228-522/1, OOOgal $456-1,44o/l,000gal $760-1,7397 l.OOOgal

Without Effluent Treatment Costs

24gpfn 12gdm 7.2gpm$222/1,000 gal $444/1,000 gal $739/1,000 gal

As expected, the cost category having the largest impact and variability on total costwas effluent treatment and disposal.

The demonstration used a four-module filtration unit. For a full-scale remediation,twelve of the same modules instead of four would be used with a portable generatorfor power, a mix tank, and a single pump and motor.

No assumptions as to the site size or volume of waste to be treated were made. It wasassumed that the same unit would be operated at different flow rates for a one yearperiod to obtain the desired results. For example, at the maximum assumed flow rateof 24 gpm, 2.6 million gallons of waste would be treated in 230 days of operation.The annual cost was then divided by the volume of waste that would be treated at aparticular flow rate to obtain $/1,000 gal.

No assumptions regarding percent rejection or outlet contaminant concentrations weremade. Based on results from the demonstration, a volume reduction of 80% betweenwaste and concentrate was assumed. Costs per 1000 gal, treated were calculated for24, 12 and 7.2 gpm flow rates; the last corresponding to what was demonstrated in thepilot program. Flow rates, the amount of recycle, and the initial concentration ofcontaminants may impact costs significantly.

One equipment operator/supervisor and one technician will operate the unit and be on-site eight hours per day, although the system will be operated only seven hours perday, five days per week. The extra hour each day will be used for cleaning andmaintaining the unit. A site supervisor will visit the site for approximately two to threedays each month for oversight purposes. The two-person crew could operate up tothree 12-module systems. If more modules are required, additional manpower wouldbe needed.

The filtration unit was assumed to be utilized for 230 days out of a possible 365 daysa year. Scheduled maintenance was assumed to be performed during normal operatinghours.


For the purpose of this analysis, capital equipment costs were amortized over a 7-yearperiod with no salvage value. Interest rates, time-value of money, etc, were not takeninto account.

The following is a list of additional assumptions used in this study.

• Access to the site is available.« Utilities, such as electricity, water, telephone, is easily accessible.• The permeate stream will not require further treatment.• A hook-up to the appropriate outlet (sanitary sewer, storm sewer, surface water

body) is available on or near the site.• There are no waste water pre-treatment requirements.

Basis for Economic Analysis

In order to compare the cost-effectiveness of technologies in the pilot demonstrationprogram, costs were broken down into 12 categories shown in Table 4 using theassumptions already described. The assumptions used for each cost factor aredescribed in more detail below.

Table 4. Estimated costs for the filtration unit studied.

COST COMPONENT TOTAL

1. Site Preparation Costs *2. Permitting & Regulatory Costs *3. Equipment Costs (amortized over 7 years) 4.Startup *5. Labor6. Consumables and Supplies

Health & Safety GearMaintenance Supplies

7. UtilitiesTelephoneElectricitySewer/Water

8. Effluent Treatment & Disposal (Concentrate)9. Residuals/Waste Shipping, Handling

and Transport Costs10. Analytical Costs11. Facility Modification, Repair & Replacement12. Demobilization Costs *TOTAL (without concentrate Disposal)TOTAL (with concentrate disposal)

$85,000$15,000$42,850$5,000$199,080

$3,000$500

$6,600$2,000$2,000$13,915-$695,520$46,000

$60,000$37,150$10,000$514,180$528,09541,209,700

* one-time costs


Site Preparation Costs - The amount of preliminary preparation will depend on the siteand is assumed to be performed by the responsible party (or site owner). Sitepreparation responsibilities include site design and layout, surveys and site logistics,legal searches of access rights and roads, preparations for support facilities,decontamination facilities, utility connections, and auxiliary buildings. Thesepreparation activities are assumed to be completed in 500 staff hours. At a labor rateof $50/hr. this would equal $25,000.

Other significant costs associated with site preparation include construction of a padand cover, well drilling as well as buying and installing a groundwater pump, holdingtanks, and associated plumbing. The cost to construct a concrete pad and cover tosupport the unit and protect the unit from the elements is estimated to be $20,000.

Based on the demonstration, the cost to drill a well was assumed to be $5,000. Toachieve the appropriate maximum groundwater extraction rate of 24 gpm, threerecovery wells are required, resulting in a cost of approximately $15,000. A 5200gallon, holding tank cost $5,000. Using the "six-tenths rule" to scale-up, the cost ofa 10,000 gallon tank for a full-scale remediation was assumed to cost $7,400. Threetanks will be required, resulting in a cost of $22,200. A V2 horse-power pump cost$1,035 for the demonstration. A pump for each well would cost a total of $3,105.These additional costs amount to about $40,000.

Therefore, the total site preparation costs for a full-scale remediation would be about$85,000 as shown in Table 4.

Permitting and Regulatory Costs - Permitting and regulatory costs include actualpermit costs, system health/safety monitoring, and analytical protocols. Permitting andregulatory costs can vary greatly because they are very site- and waste-specific. Forthis cost estimate, permitting and regulatory costs are assumed to be 5% of theequipment costs. This assumption is based on operation at a Superfund site. At RCRA(Resource Conservation and Recovery Act) corrective action sites permitting andregulatory costs may be higher and an additional 5 % of the equipment cost should beadded.

Equipment Costs - Capital equipment costs are for a twelve-module filtration unitequipped with a portable generator for power, a mix tank, and a single pump andmotor all mounted on a trailer with associated instrumentation, alarms and controls.Variation in equipment costs from site-to-site should not be significant. However,based on the cleanup requirements and the material being treated, the flow ratethrough the system may vary dramatically resulting in a wide range of costs per unittreated.

Based on a capital cost estimate of $300,000 for 12 modules, each module would cost$25,000. Equipment costs were amortized over 7 years, with no salvage value at theend of that time period, giving an annual cost of $42,850 as shown in Table 4, withoutany interest factor.


Startup ~ Filtration units are mobile and designed to move from site-to-site.Transportation costs are only charged to the client for one direction of travel and areusually included with mobilization rather than demobilization. Transportation costs arevariable and dependent on site location as well as on applicable oversize/overweightload permits, which vary from state-to-state. The total cost will depend on how manyand which state lines are crossed.

The system is designed to be ready to operate as mounted on the trailer so mobilizationcosts should be primarily the cost of travel and the time to connect the plumbing andadjust the membranes, if necessary. The startup labor cost is included in the total laborcost component and includes relocation and/or hiring expenses.

The cost of health monitoring programs has been broken down into two components— OSHA (Occupational Safety and Health Act) training, estimated at $l,000/person,and medical surveillance, estimated at $500/person for a total cost of $l,500/person.For two people, on-site, this would be $3,000. Depending on the site, however, localauthorities may impose specific guidelines for monitoring programs. The stringencyand frequency of monitoring required may have significant impact on the project cost,A conservative estimate of $5,000 was assumed as shown in Table 4.

Labor - Labor costs may be broken down into two major categories: salaries and livingexpenses. It is estimated that the equipment will require two on-site personnel foroperation and maintenance. Due to the extended time requirements for majorgroundwater restoration projects, plans to hire local operators or relocate personnelto the site may be necessary. These actions would minimize costs associated withliving expenses. A cost of $5,000 is estimated for hiring and/or relocation.

Site supervision will require periodic visits from the mam or regional office to overseethe progress of the remediation. Per diem is assumed to be $125 per day per person,but may vary widely by location. This rate is a liberal estimate assuming that cleanupsmay occur in some of the more expensive areas of the country. Travel to and from thesite (periodic supervision) is estimated to be $800/visit. One rental car is assumed tobe obtained at a rate of $55/day.

Supervisory and administrative staff will consist of an off-site program manager at$75/hour. The filtration system will operate 7 hours per day, 5 days per week. Oneequipment operator/ supervisor at $50/hr. and one technician at $35/hr. will be on-site8 hr./day. The labor requirements and rates are detailed in Table 5.

Consumables and Supplies - There are two items to consider under this cost category.The first is health and safety gear which include hard hats, safety glasses, respiratorsand cartridges, protective clothing, gloves, safety boots, and a photoionization detectormonitor, all estimated at $l,500/person. For two people this totals $3,000.

The second item is maintenance supplies (spare parts, oils, greases and otherlubricants, etc.) estimated at 1% of the annual amortized capital costs or


approximately $500, The cost of membrane forming chemicals are inconsequential(less than $200),

Utilities - Telephone charges are estimated at $500/month plus an additional 10% forfax service or $550/month. This will total $6,600 annually.

Electric usage is estimated to cost about $ 10/day or $2,000 annually. Combined sewerand water usage costs is assumed to be about $0.05/1000 L ($0.20 per 1000 gal).Based on the demonstration results, approximately 150 gallons of water were used toflush a 4-module system. Hence a 12-module system was assumed to use three timesas much water or about 500 gallons/day. This would cost about $ 10/day or $2,000 ayear as well. This does not consider discharge of permeate, which may incuradditional cost.

Table 5. Labor requirements and rates to operate the filtration unit,

Living and Travelling Expenses: 3 days/month for 12 months:Per Diem $125/day/person x 1 person x

3 days/week x 12 weeks = $4,500Rental Car $55/day x 7 days/week x

52 weeks = $1,980Travel $800/trip x 12 months = $9,600

Salaries:Program Manager - $75/hr(*} x

8 hr/day x 36days = $21,600Operator/Supervisor - SSO/hr^ x

8 hr/day x 230 days = $92,000Technician - $35/hr(*} x

8 hr/day x 230 days = $64,000Relocation/Hiring = $5,000

Total Labor = $199,080Includes salary, benefits, and administration/overhead costs but excludes profit.

Effluent Treatment and Disposal - Two process streams are produced by the filtrationunit. The permeate is considered to be essentially free of contaminants and is assumedto meet standards appropriate for discharge to a POTW. The concentrate is thereduced-volume portion of the waste stream containing the enriched contaminants.This stream would require further treatment such as biological degradation,incineration, fuel-blending, or some other process appropriate to the type andconcentration of contaminants.

The filtration system is a volume reduction technology, and as such minimizes thevolume of wastewater that would require treatment. The technology was demonstratedas a method to reduce the volume of wood preserving waste contaminatedgroundwater. Therefore, treatment of the concentrate is not part of the demonstrated


technology and it is not necessarily appropriate to consider costing for this parameter.However, the cost for treating these effluents can be a substantial factor in designinga remediation program. Based on these issues, overall costing will be calculated bothwith and without effluent treatment and disposal costs.

Two concentrate disposal options are considered in this exercise. The first isbioremediation which provides on-site destruction of PAH contaminants. A projectedcost estimate of 10-40 cents per gallon of groundwater contaminated with 100-2000ppm of PAHs is appropriate for a full-scale bioremediation system.

The second disposal option for the concentrate is more conventional. Based on thecharacteristics of the concentrate, fuel blending is considered a viable disposal option,resulting in a cost of $1.50/gallon.

It is important to note that effluent treatment costs can be very high and are dependenton specific waste and site conditions. Cost estimates for this exercise are based onwaste and site characteristics of the demonstration.

Based on the demonstration, the concentrate accounts for 20% by volume of thecontaminated groundwater influent stream to the filtration unit. The volume ofconcentrate generated each day and the range of costs for the three different flow ratesare shown below for the bioremediation system and conventional disposal:

Gallons of Waste Treated/Day

Gallons of ConcentrateGenerated/Day (assumes 20%)

Annual Treatment CostsBioremediation

Annual Treatment CostsConventional

24gpm

10,080

2,016

$46,370-$185,470

$695,520

12gprn

5,040

1,008

$23,180-$92,736

$347,760

7.2gpm

3,024

605

$13,915-$55,660

$208,725

Effluent treatment and disposal costs can range from $14,000-$700,000 depending onthe flow rate through the filtration unit, the mode of treatment, and the cost oftreatment in the bioremediation system.

Residuals/Waste Shipping, Handling and Transport Costs - Waste disposal costsincluding storage, transportation and treatment costs are assumed to be the obligationof the responsible party (or site owner). It is assumed that residual or solid wastesgenerated from this process would consist only of contaminated health and safety gear,used materials, etc. Landfilling is the anticipated disposal method for this material andcosts were once again derived from the demonstration test. Approximately four drumsof solid waste were generated each day of operation. However, due to intensivesampling activities during the demonstration, excessive solid waste was generated.


Under actual remediation conditions, substantially less waste would be generated. Itis estimated that approximately one drum of solid waste would be generated each dayof operation. At a disposal cost of $200/drum, the total yearly cost of disposal isestimated to be $46,000.

Analytical Costs - Standard operating procedures do not require planned sampling andanalytical activities. Periodic spot checks may be executed to verify that equipment isperforming properly and that cleanup criteria are being met, but costs incurred fromthese actions are not assessed to the client. The client may elect, or may be requiredby local authorities, to initiate a sampling and analytical program at their ownexpense.

For this cost analysis, one sample per day for 100 days at $600/sample was assumedto be required by local authorities for monitoring and permitting purposes. This wouldtotal approximately $60,000.

Facility Modification: Repair and Replacement Costs - Since site preparation costswere assumed to be borne by the responsible party (or site owner), any modification,repair, or replacement to the site was also assumed to be done by the responsible party(or site owner). The annual cost of repairs and maintenance was estimated to be$37,150.

Demobilization Costs - Site demobilization will include shutdown of the operation,final decontamination and removal of equipment, site cleanup and restoration,permanent storage costs, and site security. Site demobilization costs will varydepending on whether the treatment operation occurs at a Superfund site or at aRCRA-corrective action site. Demobilization at the latter type of site will requiredetailed closure and post-closure plans and permits. Demobilization at a Superfund sitedoes not require as extensive post-closure care; for example, 30-year monitoring is notrequired. This analysis assumed site demobilization costs are limited to the removalof all equipment and facilities from the site. It is estimated that demobilization wouldtake about two weeks and consist primarily of labor charges. Labor costs includesalary and living expenses. Demobilization is estimated to be $10,000.

Grading or recompaction requirements of the soil will vary depending on the futureuse of the site and are assumed to be the obligation of the responsible party (or siteowner).

Overall Economic Analysis

Table 4 shows the total annual cleanup cost to range between $514,180 and$1,209,700. This is based on the assumption that the remediation will take one year.Most applications for this technology will require several years, as in pump-and-treatremedial projects. Since many of the cost factors are one-time, the overall $/galloncost will go down as the length of the project increases. This is illustrated in thehypothetical site example in the subsequent sub-section. The total cost is also highlydependent on whether concentrate treatment and disposal is considered as part of the


nitration's technology and responsibility. Concentrate disposal costs can vary widely,and are dependent on technical and regulatory issues related to the wastecharacteristics. Therefore, if concentrate disposal costs are considered, this categorycould account for up to 60% of die total costs. Without concentrate disposal, labor isthe dominant cost, accounting for approximately 40% of the cost. Equipment costsrepresent a relatively minor component. Furthermore, the system can be easily scaled-up by adding 12-module units. Up to three 12-module units can be operated withoutadding additional labor. This would significantly reduce overall treatment costs. Thesmallest cost categories appear to be those associated with startup, and consumablesand supplies. All other cost categories appear to contribute to the total cost aboutequally (i.e., 5-10%).

The costs per 1,000 gal is dependent on the flow rate, the duration of the project,whether concentrate disposal is being considered, and the cost of effluent treatmentand disposal. These ranges are shown below and are based on a one year project.

With ConcentrateDisposal

Without ConcentrateDisposal

Cost per thousand gallons of feed

24gpm

$2284522

$222

12gpm

$456-$1,044

$444

7.2gpm

$760-$1,739

$739

In all of the above analyses, it should be remembered that costs for 10 out of the 12cost components were considered. One of the cost components not included here waspermitting and regulatory expenses. Additional effluent treatment and disposal for thepermeate was assumed to be not required. If these factors are taken into account, costscould significantly increase.

Remediation of a Hypothetical Site

The economic analysis presented in the preceding section is based on costs for a oneyear remedial project. The dominant application of the membrane system is expectedto be for groundwater restoration projects. Since groundwater restoration projects canlast for ten to twenty years, a hypothetical economic analysis is presented to illustratethe application of the twelve factors in developing a multi-year project.

The hypothetical site contains groundwater contaminated with wood preserving wastesin composition and concentrations similar to the feedwater tested in the demonstration.The remedial plan calls for containment of the groundwater plume, with eventualaquifer restoration. A hypothetical model predicts that approximately two milliongallons of groundwater is contaminated, and that twenty million gallons must betreated to restore the aquifer. The groundwater will be extracted from the shallowaquifer (ten to thirty feet below surface) through three wells.


The remedial design will utilize three 12-module filtration units operating at 7.2gpm/unit for a combined throughput of 21.6 gprn. Treatability testing identified thatan 80% volume reduction could be achieved, with the permeate meeting dischargestandards to the local POTW. The concentrate from the process will be treated on-siteby a bioremediation technology at a cost of 40 cents/gallon. Based on these conditions,and the economic assumptions previously stated, the remedial time-frame will be tenyears. Approximately 2,100,000 gallons of groundwater will be treated each year bythe filtration unit, and 420,000 gallons of concentrate by the bioremediation system.The total volume of groundwater to be treated for the ten year project is 21 milliongallons.

Table 6 is a summary of the costs for each of the twelve criteria as they relate to theconditions set forth hi the hypothetical analysis. Based on the requirements of thehypothetical site, the overall treatment costs for the remediation is $300/1000 gallons.It is important to note the overall $/gallon treatment cost is highly dependent on thelength of the remediation project. The longer the project, the lower the $/gallontreatment cost.

Table 6. Hypothetical site cost analysis for a ten year project.

1. Site Preparation $85,0002. Permitting and Regulatory $15,0003. Equipment $900,0004. Startup $5,0005. Labor $1,990,8006. Consumables and Supplies $35,0007. Utilities $106,0008. Effluent Treatment and Disposal $1,669,2489. Residuals $460,00010. Analytical $600,000I t . Facility Modification Repair

and Replacement $371,50012. Demobilization $10.000

Total $6,247,548

Detailed Process and Technology Description

Membranes are being used increasingly for the removal of dissolved and colloidalcontaminants in wastewater streams. Reverse osmosis (hyperfiltration) is well knownfor its ability to concentrate ionic species while ultrafiltration has found broad utilityfor the removal of dispersed colloidal oil, non-settlable suspended solids, and largerorganic chemical molecules. One of the major problems these processes have facedis the fouling or blinding of the membranes after limited use. Various approaches havebeen developed in an effort to minimize this deterrent. Cross-flow filtration, wherethe contaminants are constantly flushed or washed from the membrane surface by thefeedwater stream, is one of these approaches. The unit goes farther. Rather than a thin


polymeric membrane requiring careful handling to avoid perforation, a membrane iscreated on a stainless steel microfilter support by the introduction of a mixture ofcarefully selected (and proprietary) chemicals. This approach imparts specialproperties and allows a degree of customization that may be difficult to achieve withconventional membranes. The resulting "formed-in-place" membrane can be designedto provide properties similar to conventional hyperfiltration or ultrafiltration, asneeded for a specific application.

The filtration unit consists of porous sintered stainless steel tubes arranged in amodular, shell-and-tube configuration. Multi-layered inorganic and polymeric"formed-in-place" membranes are coated at microscopic thickness on the insidediameter of the stainless steel tubing by the recirculation of an aqueous slurry ofmembrane formation chemicals. This "formed-in-place" membrane functionally actsas a hyperfilter, rejecting species with molecular weights as low as 200. In addition,surface chemistry interactions between the membrane matrix and the components inthe feed play a role in the separation process. A relatively clean stream, called the"permeate", passes through the membrane while a smaller portion of the feedwater,retaining those species that do not pass through the membrane, is retained in a streamcalled the "concentrate" or "reject".

For efficient operation of a membrane filtration system, it is necessary to prevent thebuildup of dissolved and particulate species on the surface of the membrane and in themembrane pores. The buildup of contaminants, termed "fouling", can lead to a steadydecline in the permeate flux (flow per unit area of membrane surface), eventuallycausing cessation of flow. To prevent, or retard excessive fouling, the filtration unitis operated in a cross-flow mode (as illustrated Figure 1). In cross-flow mode the feedstream is directed parallel to the surface of the membrane. Material larger than thesurface porosity is temporarily retarded on the membrane surface and then swept cleanby the cross-flow action — if the fluid velocity is sufficient. Meanwhile, the portionof the stream containing the smaller species passes through the membrane. The goalof cross-flow filtration is not to trap components within the pore structure of themembrane.

The test unit used operates with four modules aligned in parallel. The filtration unitis approximately 13 feet long, 5 feet wide, and 7 feet high and contains an estimatedtotal membrane area of up to 300 square feet. Automatic level controls provide forunattended operation with continuous feed to the tank. Concentrate recycle flow alsocan be controlled automatically. Figure 2 provides a schematic of the filtration unit.

At the American Creosote Works site, groundwater was pumped from a well to anabove ground storage tank where a quiescent period of several hours allowed oil andsuspended solids to coalesce and separate. The feedwater stream to the filtration unitwas drawn from the mid-section of the storage tank to minimize introduction of thesematerials. The pump that drew material from the tank also provided the compressionfor the system to operate, approximately 750 psig.

The permeate leaving the filtration unit was sampled as required and then dischargedin accordance with permit requirements. The concentrate was collected in a smaller


tank until the desired volume was accumulated. It is then recycled as feed until thedesired final concentration and volume are achieved. This mode of operation wasselected for the demonstration in anticipation of a companion study of biodegradationprocess for the concentrate. Alternate operating modes can be used to achieve othergoals, depending on disposal plans and options for the permeate and the concentrate.

Permeateo

Influent Concentrate

• <%«» e*e?e*

Figure 1. Illustrates cross-flow filtration mode of operation.

Groundwater

, Membrane%• Filtration -O

Feed Tanki i

9-̂ MembraneFiltration Unit

I*Permeate 1Disposal V

Concentrate

Wfc^ Qtsvran*

Concentrate Recycle 1

Tank

r

Sampling Points

Figure 2. Schematic of the filtration system.

The hyperfiltration system, which consists of porous stainless steel tubes internallycoated with specially formulated chemical membranes, has been demonstrated tosuccessfully treat water contaminated with a number of hazardous or toxic materials.In this system, contaminated ground and surface waters are pumped through the


filtration system tubes and contaminants are collected inside the tube membrane while"clean" water permeates the membrane and tubes. The system has been shown to behighly versatile and able to effectively remove a variety of materials fromcontaminated waste streams, including petroleum hydrocarbons, benzene, toluene,ethylbenzene, xylene, dioxins, chlorinated furans, and heavy metals. This extent ofversatility is provided largely by the application of several types of chemicalmembranes with distinct permeability and ion exchange capabilities.

In the field study discussed in this chapter, one type of chemical membrane was usedto coat the porous stainless steel tubing, with an aim to provide the optimal separationefficiencies for the higher molecular weight contaminants. These are generallyregarded to be most "hazardous" molecules and are generally more resistant todegradation in the environment. The results confirmed that this membrane system wasvery effective in removing more than 95 % of the high molecular weight polynucleararomatic hydrocarbon (PAH) contaminants, the most carcinogenic components, duringthe monitored demonstration run. However, on the average, approximately 20% ofthe lower molecular weight phenolic contaminants were removed by this membranetype. The vendor feels that this percentage can be improved by adjusting membranesand flows in the field.

In the study, the relatively heavy concentrations of contaminants in the feed materialfor the hyperfiltration unit precluded effective use of other "tighter" types ofmembranes as the first pass barrier due to the potential for fouling. Optimally, acombination system employing the initial membrane used here to remove highmolecular weight contaminants, followed by a "tighter" membrane to remove lowerweight phenolics from the permeate of the first membrane would have been morelikely to provide the full spectrum of contaminant removal desired.

Petroleum Hydrocarbon Hyperfiltration

Other waste streams, highly contaminated with lower molecular weight molecules, canalso be hyperfiltered and concentrated by using alternate chemical membranetypes. One example is the hyperfiltration of waste water from certain oil refiningoperations containing significant amounts of benzene, toluene, ethylbenzene, andxylenes (BTEX), as well as other petroleum hydrocarbons.

A summary of the hyperfiltration results of this treatment is as follows:

Contaminant

BenzeneTolueneEthylbenzeneXylenesTPH

Original Feed Sample

2200ppb7640ppb2590ppb12200ppb 483ppm

Permeate Sample

34ppb82ppb15ppblOOppb3.5ppm

%Removal

98.5%98.9%99.4%99.3%99.3%

Note the TPH stands for total petroleum hydrocarbons.


In a separate study, condensate from a methane-recovery operation at a municipallandfill was treated for the removal and concentration of a large spectrum ofcontaminants, including naphthalene, heavy metals, and BTEX. The results of this are:

Contaminant

TolueneEthylbenzeneXylenesNaphthalene

Original Feed Sample

177ppm268ppm561ppm90.4ppm

Permeate Sample1

<5ppm< 5ppm< lOppm< lOppm

%Removal

>97,2%>98.1%>98.2%>89.0%

All permeate readings were below the detectable limits for the analysis method used in this study.

These results demonstrate the capacity and versatility of the hyperfiltration system intreating a variety of waste steams and achieving effective volume reduction in removalof contamination from groundwater, municipal landfill leachates, or contaminatedpetroleum waste streams. The value of this technology can be further enhanced bycoupling it with a biodegradation process. This can be achieved by using thehyperfiltration concentrate as a bioreactor feed stream, as well as by using thehyperfiltration system to polish bioreactor effluent to yield two streams: one, a cleanstream suitable for discharge; and the other, a polished concentrate to feed to thebioreactor. This creates a closed loop for targeted contaminants, and provides for anefficient continuous flow remediation design.

In addition to the application of hyperfiltration technology to the remediation ofcreosote-contaminated groundwater, effective biodegradation of creosote andpentachlorophenol has also been achieved using specially selected, non-engineeredmicroorganisms in a bioreactor system. The combination of these two systems,hyperfiltration and bioremediation, provides a novel and reliable means to firstconcentrate the waste feed to the bioreactor to an optimal level for efficientbioremediation activity, as well as to provide for a final polishing step usinghyperfiltration of bioreactor effluent.

For this study, a bi-phasic bioreactor design was utilized, operating in a semi-continuous flow process, having a hydraulic retention time of four days. Groundwater,with contaminant concentrations as high as 7000 ppm creosote, was treated on-site.This demonstration achieved a removal efficiency of greater than 99% for totalpolynuclear aromatic hydrocarbons (PAHs). This includes a removal rate of 98% forthe most recalcitrant, and most hazardous fraction of the PAHs, and 88% for PCP.The field test proved that biotechnology application for hazardous waste remediationcan be effective at an actual waste site.

PAHs are a widespread contaminant of soil and groundwater typical of creosote woodtreating facilities, manufactured gas plants, refineries and related industries.Bioremediation has been attempted for PAH constituents in several studies and fieldapplications, but until now, biodegradation using indigenous bacterial strains has beenable to achieve only 50% - 75% removal of PAHs. The untreated portion is generallycomprised of the recalcitrant high molecular weight (HMW) PAHs which are those


compounds with 4, 5 or 6 fused rings and are the PAH components which are knownor suspected carcinogens (see Table 7). Similarly, PCP, a common wood preservative,has proven difficult to biodegrade under field conditions.

Table 7. Classes and characteristics of PAHs.

Low Molecular Weight PAHs — relatively non-hazardous, relatively easy todegrade.

naphthalene2-methylnaphthalene1-methyinaphthalene biphenyl2, 6-dimethylnaphthalene2, 3-dimethylnaphthalene

Medium Molecular Weight PAHs — some potentially hazardous to humanhealth, more complex but still biodegradable by many bioremediation systems,

acenaphthyleneanthracenefluorenephenanthrene2-methylanthraceneanthraquinone

High Molecular Weight PAHs — several carcinogens, very slow rates ofbiodegradation without specialized microbes.

fluoranthenepyrenebenzo(b)fluorenechrysenebenzo(a)pyrenebenzo(a)anthracenebenzo(b)fluoranthenebenzo(k)fluorantheneindeno( 1,2,3-c, d)pyrene

USEPA scientists have isolated and identified several strains of naturally occurring soilbacteria capable of mediating PAH and Pep degradation at rates in excess of thoseachieved by undifferentiated communities of indigenous microbes. The strains areidentified as:

CRE 1-13: low and medium weight PAH degraders comprised of anassemblage of 13 Pseudomonads.

Advanced Membrane Technology for Wastewater Treatment 15*9

EPA 505: a strain of Pseudomonas palleimobilis capable of high ratedegradation of HMW PAH constituents.

SR3: a strain of Pseudomonas sp. which degrades PCP.

All organisms have been shown to mineralize their target contaminants. Additionally,EPA 505 has been patented for use in the degradation of high molecular weight PAH,The bioremediation process was tested on groundwater at a Superfund site wherewaste liquids from the manufacturing process were placed in unlined surfaceimpoundments on-site. These impoundments often overflowed into drainage ditcheswhich discharged into local waterways. In addition, wastes have migrated into theshallow aquifer, contaminating both soil and groundwater.

The Superfund site has large volumes of shallow groundwater contaminated bycreosote and Pep. In order to prove the capability of the organisms to degrade thesecontaminants under field conditions, a highly contaminated groundwater was chosenas the test matrix.

Groundwater was pumped from the aquifer via an existing monitoring well. Theextracted groundwater was stored in an equalization tank prior to the test. From thistank, the contaminated feed was pumped to the two-stage bioreactor treatment system(that is illustrated in Figure 3). Each bioreactor had a hydraulic capacity of 200gallons and was designed to provide mixing and up-lift type aeration.

The bioreactors were operated sequentially, i.e., the contaminated water wastransferred to Bioreactor 1 (BR1) at a pre-set flow rate of four days. After four days,when BR1 was full, the water was allowed to overflow into Bioreactor 2 (BR2).Laboratory grown concentrates of CRE 1-13 (specialized degraders for the low andmedium weight PAHs) were added to BR1, along with nutrients and sparged air, andthe tank was mixed. Similarly, BR2 was inoculated with EPA 505 (HMW PAH-degrading strain) and SR3 (Pep degrader) during its eight days of operation. Treatedflow from BR2 was held in a tank for testing; after testing, the water was dischargedto the city sanitary sewer.

During operation of the bioreactors, samples were collected for the analysis of keyoperating parameters, such as dissolved oxygen, nutrient levels, total organic carbonand suspended solids. Microbial analysis was performed to assess the cellconcentration of the specialized bacteria being added. All cultures were prepared inadvance and added to the bioreactors. Additional samples were collected to measurethe contaminant concentration across the bioreactor, as well as in the various portionsof the treatment system, in order to calculate a mass balance.


ContaminatedGroundwater

ORE 1-13

EPA 505

SR3

Discharge to CitySanitary Sewer

BioreactorNo. 1

lBioreactor

No. 2

Nutrients &Air

Figure 3. Simplified process flow for SBP Technologies Inc. 's bioreactor system.

The overall PAH and PCP degradation performance of SBP's treatment system isshown in Table 8.

Table 8. Summary of bioremediation results of PAN and PCP removal.

Contaminant

Low Molecular Weight PAHs

Med Molecular Weight PAHs

High Molecular Weight PAHs

Pentachlorophenol

Influent(mg/L)

31

539

368

256

Effluent(mg/L)

8.1

1.6

5,2

31

%RemovaI

<99

>98

98

88

These results represent a significant advancement in PAH bioremediation. Not onlyhas the total PAH been reduced to <1% of its original concentrations, the HMWPAH, a class of compounds which are typically not removed in biological remediationsystems, have been reduced by 98-plus %. This removal has the additional significant


characteristic of having been mediated by an organism specially and specificallyisolated, cultured and introduced into the bioremediation system to remove this classof compounds, and this has been accomplished under typical field remediationconditions.

To further validate the treatment technology, the contaminated groundwater wassubjected to tests for toxicity and teratogenicity before and after treatment (sampledbefore carbon treatment). As determined by Microtox™, embryonic Menidia beryllina,Mysydopsis bahia, and Ceriodaphnia dubia bioassays, these indicators of potentialthreat to human health and the environment were significantly reduced. This datashows that the treatment system was effective in removing the hazardous propertiesof the waste material, while simultaneously demonstrating that no metabolic by-products or toxic intermediates were created by the microbial biotransformation of thewaste constituents.

A mass balance was performed by comparing the total influent loading of PAHs to theresidual PAHs left in the treatment system at shutdown. The HMW PAHs were foundto be 80% removed (20% remained in the residual). In a scaled-up system with highervolume of wastewater treated, the removal efficiently at steady state would approachthe percent removal from influent to effluent (98%).

The approach to bioremediation of hazardous waste sites is a "multi-phase" strategy,meaning that the technology is adaptable to the treatment of soil, sludges, leachates,groundwater and surface water. By the multi-phase approach, the waste matrix ismodified and managed to support the greatest level of specialized organisms survivaland highest biodegradation rates. Examples are given below.

Bioremediation is applicable to a wide range of organic contamination. SBP hasfocused on research, development and field implementation of biotechnology-basedapproaches to some of the more difficult hazardous waste constituents, in particularwood preserving wastes and solvent contamination. SBP has bioremediation solutionsfor:

PAHs (e.g., creosote, coal tar)Chlorinated Aliphatic (e.g., TCE)PentachlorophenolPetroleum Hydrocarbons

Treatment of contaminated liquids such as water, leachate, filtrate, groundwater,storm water, surface water and industrial process waste water can be accomplishedby several of SBP's technologies. Certain waste streams can be concentrated by thehyperfiltration units; the permeate is clean water and the "concentrate" contains thereduced volume of the pollutants. These concentrated contaminants can often bebioremediated, thus minimizing the waste stream. Other liquids may be treateddirectly, either by biological processes, or in the case of volatiles, air stripping withbiotreatment of the pollutants in a gas phase bioreactor, or "biofilter".


There are several approaches for the handling of contaminated soil. Solid phasesystems are used when biostimulation of indigenous bacteria is appropriate, Slurryphase reactors, either in-vessel or in situ are used for soils or sludges requiring bettercontrol of microbial systems. A multi-phase approach combining soil washing andhyperfiltration to reduce volume and bioreactor treatment of the wash water isappropriate for heavily contaminated wastes.

Summary of Case Study Analytical Results

The goal of the demonstration project was to study the effectiveness of the formed-in-place hyperfiltration membrane unit, operating in a cross-flow mode, for theconcentration of PAHs (and PCP) from contaminated groundwater found at a woodtreating site. The original plans called for a companion study of a proprietarybiodegradation process for the concentrate; however, a decision was made during thisstudy not to carry out that portion of the study.

Based on available information, including a Remedial Investigation / Feasibility Study(RI/FS), the American Creosote Works site in Pensacola, Florida was selected for theinvestigation. The site was included on the National Priorities List on the basis of theRI/FS study which suggested that the groundwater under the site was heavilycontaminated with both PAHs and PCP.

The American Creosote Works site had been used for wood treatment from 1902 to1981. Creosote was the preservative used until the 1970's when a shift was made topentachlorophenol in a light oil. Over the years, wood impregnation was carried outin open toughs, resulting in spills and drippage which seeped into the ground.

A well selected for the study yielded sufficient groundwater volume and flow but theoutput was contaminated with free product. Field COD tests were used to determinethat five-fold dilution of the groundwater with city water was necessary to meet the100-500 mg/L COD input guideline for feedwater to the pilot filtration unit.Approximately 700 gallons of the groundwater was pumped from the well on alternatedays. The groundwater was allowed to settle for several hours, and the aqueous phasepumped to a separate tank. The contaminated aqueous phase was diluted with 2800gallons of carbon-treated city water to bring the contaminated level of the feed in thedesired range of 100-500 mg/L COD. Approximately 1000 gallons of this dilutedstream was used as the feedwater to the filtration unit per day. The averagecharacteristics of this feed stream are shown in Table 9, and are based on compositesamples taken on each day over the six days of the demonstration.

To achieve the desired volumes and concentrations, the feedwater was introducedduring the first two hours (approximately) of each day's run. For the remainder ofeach day's run, the reject (concentrate) was recycled, becoming the feed, while thepermeate was continuously removed and discharged after samples were compositedfor analysis. Each day's run was terminated when the flowmeters indicated that the


desired 800 gallons of permeate had been discharged, leaving approximately 200gallons of concentrate.

Table 9. Average characteristics of feed stream to the filtration unit.

AverageTDS(ppm)TSS(ppm)TOC(ppm)COD(ppm)Oil/Grease (ppm)

Semivolatiles (ppb)Phenol 4,9002-Methylphenol 2,3084-Methylphenol 6,9172,4-Dimethylphenol 1,817Pentachlorophenol (2,425)Benzole Acid (1,421)Naphthalene 12,8672-Methylnapthalene 4,525Acenaphthylene (138)Acenaphthene 6,842Dibenzofuran 4,875Fluorene 5,925Phenanthrene 17,083Anthracene 1,983Fluoranthene 7,008Pyrene 4,700Benzo(a)Anthracene 1,235Chrysene 1,127Benzo(b)Fluoranthene (460)Benzo(k)Fluoranthene (428)Benzo(a)Pyrene (312)

( ) estimated value

Samples were analyzed for oil and grease, dissolved and suspended solids, volatileorganics, TOG, COD, and dioxins/furans in addition to the list of designatedsemi volatile organics used in evaluating the developer's claims.

As noted earlier, the system was operated by introducing approximately 1000 gallonsof the diluted feedwater (about 1 volume of groundwater to 4 volumes of city water)over a two hour period at a rate of about 8 gpm. At that time, recycle of theconcentrate was initiated and continued until approximately 800 gallons of permeatewas generated and approximately 200 gallons of concentrate was collected.


Based on operational parameters (flux, pressure), the developer determined thatwashing of the unit was necessary every other day. At the end of the second day, thefourth day, and the sixth day approximately 200 gallons of city water, to whichcleaning chemicals had been added, was flushed through the system and collectedThis material was also analyzed.

Grab samples of the feed stream were collected every 15 minutes during the initialtwo-hour single pass phase, while permeate grab samples were collected every 45minutes over the whole run. Laboratory samples were prepared by compositing thegrab samples on a flow proportional basis. After the initial single pass, the concentratewas recycled for approximately four hours in each test and one composite sample wastaken at the end of each run. All samples were transferred to bottles, inhibited orpreserved as called for by the individual test methods, labeled, sealed, and shipped inice-filled coolers to off-site laboratories by overnight express.

Instantaneous flow and accumulated volumes of feed and permeate were measuredautomatically using calibrated flowmeters. Volume of the final concentrate wasdetermined by the difference between the feed and permeate volumes. Thetemperature and pH in the three steams were measured in the field and recorded toassure that there was no gross change in characteristics that could affect filtrationeffectiveness.

At the end of the sixth day of operation, after all the required samples had been taken,the process was allowed to continue for several hours, further concentrating thecontaminants. Additional samples of the permeate were taken about every 25 gallonsto observe whether there was a fall-off in contaminant rejection rate.

The filtration unit operated in a batch mode for six hours each day, for six days, andprocessed approximately 1000 gallons of feed per day. Over the six day test period,permeate flux was a relatively constant 0.0085 gallons/min/ft2 (coefficient of variation< 10%). Based on a total membrane area of 300 ft2 for the system, the permeate flowrate for the four-module filtration unit averaged 2.6 gpm.

The unit was cleaned every two days with approximately 200 gallons of city water andmembrane cleaning materials. Over the two day period, the surface of the membranegradually fouled, requiring an increase in feed pressure in order to maintain a constantpermeate flux. The cleaning at the end of each two-day period was sufficient to restorethe original flux.

Semivolatile Contaminant Analysis

Contaminant Reduction

A selected list of semivolatile organic compounds was used to assess the effectivenessof the filtration unit. The analytes consisted of all quantifiable semivolatile compoundsdetected in the feed.


Rejection efficiency was defined as the change in total concentration for thesecontaminants between the feed and the permeate:

% Rejection =c - cJ-L_,_1£——— xlOO (2)

Cf = total concentration of contaminants in feedCp = total concentration of contaminants in permeate

On the basis of composite samples, rejection for the total designated contaminantsaveraged 74% over the six days, as shown in Table 10.

Analysis of the data reveals that the system is very effective for the rejection ofpolynuclear aromatic hydrocarbons and much less effective for phenols. Separated inthis fashion, rejection for the PAHs averaged 92% over the six days while rejectionfor the phenols averaged 18%.

Table 10. Overall semivolatile rejection for the filtration unit.

Day

123456

Feed (ppm)

10491921048560

Permeate (ppm)

182426222324

Concentrate (ppm)

206585248242199538

Rejection %

837472797360

A more accurate means of quantifying and graphically representing rejection is tocalculate the reductions on a logarithmic basis. For example, attempting to graphdifferences between 90% and 99% rejections does not fully depict the fact that a ten-fold reduction has occurred. In order to accurately represent this process, it isnecessary to calculate the order of magnitude of rejection (ORD) as the log value ofthe ratio of the contaminants concentration in the feed to the concentration in thepermeate. The equation is written as:

ORD = Log (CyCp) (3)

Therefore, an ORD of 2.0 means that a 100 fold decrease in contamination hasoccurred in the permeate relative to the feed. An ORD of 1.0 means a 10-folddecrease has occurred.


For the semivolan'le constituents, it appears that rejection follows the molecular weightof the chemical. Figure 4 presents the average ORD over the six days versus themolecular weight of the individual compounds.

.2._u3

"O0)cco>T33+*"EO5CO

0)"So

90 110 130 150 170Molecular Weight

190 210

Figure 4. Molecular weight versus order of magnitude reduction (ORD).

Mass Balance

Combining the masses of pollutants in the permeate and the concentrate on a daily basisdid not provide a good material balance, relative to the feed. Only when the material inthe washwater was also included was all the material accounted for. Based on thevolume of washwater used in each washing operation, approximately 200 gallons, about8% of the masses of the designated pollutants are retained on the membrane or in theliquid in the system. Table 11 provides a summary of the contribution of the differentstreams for each two-day period between wash operations. The relative distribution ofpollutants in the washwater was very similar to mat in the concentrate, with PAHs farexceeding the phenols. However, the concentration of pollutants in the washwater wassimilar to the feedwater.

Extended Operation

After the system yielded the desired 200 gallons of concentrate on the sixth (last day)of the test and all necessary samples had been taken, operation of the system wascontinued with further recycle of the concentrate as long as possible to observe thebehavior of the unit (i.e., the rate of folding and the quality of the permeate). Theextended operation resulted in only thirty seven gallons of concentrate remaining,


representing a 97% volume reduction. However, during this time the rejectionefficiency decreased, as expected, as concentration continued (Table 12).

Table 11. Mass contribution of total semivolatiles in all streams (2 day totals).

Days

1&23&45&6

Feed(Influent)gms

111778581

Permgms

140159151

Conegms

556320527

Washgms

468432

Totalgms

742563710

Recovery%

9572122

6 day recovery = 94%

Table 12. Behavior on extended filtration.

Additional Gallons ofPermeate Produced

255075100150

Semivolatile Rejection %(calculated from initial feed)

47.548.560.032.321.7

Conventional Parameters

In addition to the designated semivolatile pollutants, the removal of several conventionalparameters from the feedwater was also evaluated. These included total dissolved solids(TDS), total suspended solids (TSS), chemical oxygen demand (COD), total organiccarbon (TOC), and oil and grease (O&G). Comparison of feed and permeate confirmthat the membrane is capable of removing larger molecules such as oil and suspendedsolids, but is not nearly as efficient in the removal of dissolved material (TDS) orotherwise-unidentified organic species (COD) and TOC. Such results, summarized inTable 13, suggest that the membrane is operating predominantly as a hyperfiltrationmembrane. In addition, considerable concentrations of organics remain in the permeateeven though the PAHs are efficiently removed. For example, the ratio ofsemivolatiles/COD decreases significantly, from about 0.2 in the feedwater to about 0.07in the permeate, demonstrating that semivolatiles are not the major contributor to theCOD, and that significant unidentified, smaller molecular weight compounds can passthrough the membrane.


Pofychforinated dibenzo-p-dioxins/dibenzofurans

Polychlorinated dioxins (PCDDs) and furans (PCDFs) are often encountered in themanufacture of pentachlorophenol and can remain with the commercial product. Toinvestigate this matter, selected samples of the three streams taken on the first, second,and sixth day of the demonstration were scanned for the various dioxins and furansusing high resolution GC coupled with low resolution MS. A number of the morehighly chlorinated dioxin and furan species were found in the feed water and in thereject at significantly higher concentrations, but analyses of the permeate indicatedefficient removal of these pollutants. The 2,3,7,8-TCDD isomer was not found abovethe detection limit in any sample of feed, permeate, or concentrate. The 2,3,7,8-TCDF isomer was found in only one sample of feed and one sample of concentrate,in both cases the value was <0.5 ng/L (ppt). The major dioxins and furans were theocta species. These results are consistent with rejection by molecular weight or oilsolubility, but not by number of rings.

Table 13. Dioxins/furans in process streams.

Day 1

2378 TCCD12378 PeCDD123478 HxCDD123678 HxCDD123789 HxCDD1234678HpCDDOCDD 148702378 TCDF12378 PeCDF23478 PeCDF123478 HxCDF123678 HxCDF234678 HxCDF123789 HxCDF1234678 HpCDF1234789 HpCDFOCDF 1030

Feed (ng/L)

.--53.074.6198038.9---11.23.28.5-34628.33.3

Perm. (ng/L)

----1.642355---------2595

Cone. (ng/L)

.--277.713.3383599.7.21--528.730.4-10667699.7

Removal %

99.9

An1-' indicates the isomer was absent or below detection limits in stream.

Advanced Membrane Technology for Wastewater Treatment

Table 13 (Continued),

209

Day 3

2378 TCDD -12378 HxCDD123478 HxCDD123678 HxCDD123789 HxCDDS 234678 HpCDDOCDD 132502378 TCDF12378 PeCDF23478 PeCDF123478 HxCDF123678 HxCDF234678 HxCDF123789 HxCDF1234678 HpCDF1234789HpCDFOCDF 22.4

Feed(ng/L)

.-2.447.76.714807.9.37--8.1---300

Perm.(ng/L)

_----

3413----.--

1223

Cone.(ng/L)

--1169.9363099.95-.-28.44.217.0-68940.1

Removal%

* An'-' indicates the isomer was absent or below detection limits in stream.

Day 6

2378 TCDD12378 PeCDD123478 HxCDD123678 HxCDD123789 HxCDD1234678 HpCDDOCDD 92002378 TCDF12378 PeCDF23478 PeCDF123478 HxCDF123678 HxCDF234678 HxCDF123789 HxCDF1234678 HpCDF1234789 HpCDFOCDF 473

Feed(ng/L)

_-1.930.64.413708.8.39--5.11.55.1-19814.1.50

Perm.(ng/L)

_----.3690400---------5235

Cone.(ng/L)

_.8.133427.6895099.901.5-1.7111.734.242.6-2055126.299.89

Removal%

99.97

An'- ' indicates the isomer was absent or below detection limits in stream.


Volatile Organics

Analyses for volatile organics (VOCs) were carried out during the runs on days 1,3,and 6 because the developer was somewhat concerned that the high pressures used inthe system might force volatiles through the membrane into the permeate. All VOCconcentrations were low in the feedwater; the same species were also found in thepermeate and the concentrate but at lower concentrations. The principal VOC wasacetone, but the expected BTEX species were also present. Refer to the data in Table14.

Table 14. Votatiles in process streams.

Methylene ChlorideAcetoneCarbon Disulfide2-ButanoneBenzeneTolueneEthylbenzeneStyreneXylenes

Feed(Mg/U

**2321620121843

Permeate0*g/L)

**<531716(7)726

Concentrate0*g/L)

**<532<5<5<5<5<5

probable laboratory contamination) estimated value

Closure

The technology reviewed in this chapter represents a relatively new process with theintended application of wastewater treatment. The process is best applied, asdemonstrated, with a pump and treat system and can be integrated with othertechnologies such as bioremediation. The economics of the process are favorable forthe intended application, however the reader should apply separate cost studies to eachintended application to assess the applicability of this technology. Further informationcan be gained from the United States Environmental Protection Agency.

SLUDGE DEWATERING OPERATIONS

Introduction

This chapter provides a brief overview of alternative dewatering technologies usedin wastewater treatment and sludge concentrating applications. Many of the techniquesdescribed can be used in conjuction or integrated with filtration operations. Theobjective of dewatering is to increase the solid content (decrease the water content) ofthe paniculate matter separated out from filtration for one or more of the followingreasons:

1. Dewatered solids are more easily handled.2. Dewatering is normally required prior to incineration to reduce fuel

requirements.3. Dewatering reduces the cost of subsequent treatment processes, particularly

thermal processes.4. Dewatering is required prior to land disposal of solids.5. Dewatering reduces the costs of transporting wastes recovered from filtration

to their ultimate disposal by reducing their volume and weight.

In most cases, the percent solids content of a dewatered solids is set by therequirements for subsequent treatment and disposal. Each treatment technology hasan optimum range of percent solid, above or below which the technology will notoperate efficiently and economically, if at all. For example, combustion requires thatthe solid content be greater than 24 %, preferably in the 28-30 % range for moreeconomical operation. Particulates will vary in percent solid depending on thetechnology and application.

Dewatering technologies can be subdivided into two general processes: air dryingprocesses and mechanical processes. This chapter provides an overview of thesetechnologies and concentrates on typical sludge dewatering operations used forwastewater treatment.

211

7


Overview of Dewatering Technologies

"Air drying" refers to those dewatering techniques by which the moisture is removedby natural evaporation and gravity or by induced drainage. Air drying is less complex,easier to operate, and requires less operational energy than mechanical dewatering.Air drying also can produce a dryer sludge than mechanical dewatering; up to 40 %solids under normal operation and over 60 % solids with additional drying time or withthe use of underdrainage systems. Air drying processes do, however, require a largerland area and are more labor intensive than mechanical processes. Contaminantreleases via seepage, drainage, and volatilization during the dewatering process mustalso be considered.

The most widely applicable and economical air drying process available for sludgesis an appropriately managed confined disposal facility (CDF). CDFs are engineeredstructures designed to retain solids during dredging and provide storage time forgravity drainage, consolidation, and evaporation. The rate of dewatering may beaccelerated by using underdrains, pumps, or wick drains.

Mechanical dewatering involves processes in which water is forced out of the sludgethrough mechanically induced pressures. Mechanical dewatering processes include thefollowing: filtration, including belt filter presses, chamber filtration, and vacuumrotary filtration; centrifuges, including solid bowl and basket; and gravity thickening.Table 1 provides a summary of the primary dewatering techniques. A brief descriptionof each technique for dewatering applications follows.

Belt filter presses accomplish dewatering by carrying the sludge between twotensioned porous belts and squeezing out the water as the sediment/belt "sandwich"passes over and under various size rollers. All belt filter presses incorporate thefolio wing features: a polymer conditioning zone, a gravity zone, a low pressure zone,and a high pressure zone. Polymer conditioning produces a superflocculationphenomenon that allows water to drain more efficiently. The gravity drainage zoneand the low pressure zone prepare the sediment for the high pressure zone, wheremost of the water removal actually takes place.

Belt filter presses are common dewatering choices in Europe and the United States.They can dewater sediment rapidly and do not take as large an area as the air dryingprocesses. Belt filter presses are, thus, often used in confined locations (such as cities)and where large volumes of sediment must be dewatered. They do not dewater ascompletely as air drying processes, however, and may be limited by the percent soliddemands of the treatment to be used. Furthermore, if the sludge is very gritty, thebelts may wear out rapidly.

Table 1. Summary of Dewatering Techniques.

Technique

ConfinedDisposalFacility

Belt FilterPress

ChamberFiltration

Applications

Dewatering sediment of any grain size to a solids content of upto 60 % up to 99 % solids removal.

Generally used for large scale dredging operations where landspace is available.

Used to dewater fine grained sediments. Capable of obtainingrelatively dry filter cake obtaining up to 45 - 70* % solids; ableto achieve solids capture of 85 - 95 % .

Generally best suited of filtration methods for mobile treatmentsystems.

Used to dewater fine grained sediments.

Capable of obtaining a relatively dry filter cake with a solidscontent up to 50 - 80*%; able to achieve a high solids capturerate of up to 98%.

Limitations

Requires large land areas.

Requires long set-up time.

Labor costs associated with removal or dewateringsediments are high.

Systems using gravity drainage are prone toclogging.

System using vacuums require considerablemaintenance and supervision.

Systems based on electroosmosis are costly.

Performance is very sensitive to incoming feedcharacteristics and chemical conditions.

Belts can deteriorate quickly in presence orabrasive material.

Costly and energy intensive.

Replacement of filter media is time consuming.

SecondaryImpacts

Potential forgroundwatercontamination.

Potential forlocalized odorand air pollutionproblems.

Generates asubstantialamount of wastewater that mustbe treated.

Generates a washwater that mustsubsequently betreated.

RelativeCost

Low to High

Medium

High I

o•g

Table 1 Continued.

VacuumRotaryFiltration

Solid BowlCentrifuge

BasketCentrifuge

GravityThickening

Used to dewater fine grained sediments capable of obtaining afilter cake of up to 35 - 40% solids and a solids capture rate of88-95%.

Thickening of dewatering sediments; able to obtain a dewateredsludge with 15-35% solid; solids capture typically ranges from90-98%.

Suitable for areas with space limitations.

Thickening of dewatering sediments; able to obtain a dewateredsludge with 10-25% solids. Solid capture ranses from 80 -98 %.

Suitable for areas with space limitations.

Good for hard-to-dewater sludges.

Thickening of sediments slurries to produce a concentrate thatcan then be dewatered using filtration or dewatering lagoons.Able to produce a thickened product with a solids concentrationof 15-20%.

Least effective of the filtration methods ofdewatering.

Energy intensive.

Not as effective in dewatering as filtration orlagoons.

Process may result in a build up of fines in effluentfrom centrifuge.

Scroll is subject to abrasion.

Not as effective in dewatering as solid bowlcentrifuge, filtration, or dewatering lagoons.

Process may result in a build up of fines in effluentfrom centrifuge.

Units cannot be operated continuously withoutcomplex controls.

Least effective method for dewatering sedimentslurries.

Requires use of a substantial amount of land.

Generates a washwater that mustbe treated.

No significantsecondaryimpacts.

No significantsecondaryimpacts.

Potential forlocalized odorand air pollutionproblems.

High

Medium toHigh

Medium toHigh

Low toMedium

iCL

* Percent solids achievable may represent values for optimal conditions and do not necessarily represent normally expected values. Dredged sediments are often fine-grainedand difficult to dewater to the maximum indicated values.

Sludge Dewatering Operations 215

Chamber filter presses also use positive pressures, but apply the pressure to thesediments inside rigid, individual filtration chambers operated in parallel.

Vacuum rotary filtration uses negative pressure to pull the water to the interior of adrum while the sludges adhere to the exterior.

Centrifuge^ use the centrifugal forces created by a rapidly rotating cylindrical drumor bowl to separate solids and liquids based on variations in density. There are twotypes of dewatering centrifuges: the solid bowl centrifuge and the basket centrifuge.

Gravity thickening is accomplished in a continuous flow tank. Sediments settle to thebottom and are removed by gravity or pumping. Water overflows the tank and leavesthrough an effluent pump. Gravity thickeners are used primarily in tandem with otherpre-treatment technologies: they reduce the hydraulic toad to subsequent pre-treatmentoptions.

Particle Classification

Particle classification separates the slurry according to grain size or removes oversizematerial that is incompatible with subsequent processes. Classification by grain sizeis important in the management of sludges contaminated with toxic materials since thecontaminants tend to adsorb primarily onto fine grain clay and organic matter. Thesmall grain solids of a specific size or less can be treated while the relatively non-contaminated, coarser soils and solids can be disposed of with minimal or noadditional treatment. Separation technology depends on the following: volume ofcontaminated sludges; composition of the sludge, such as gradation, percent clays, andpercent total solids; characterization of the contaminants; and site location andsurroundings, including available land area. Particle classification options include.screening processes that depend on size alone, processes that depend on particle sizeand density or density alone, and processes that depend on conductive or magneticproperties of the particles.

Particle classification technologies include: impoundment basins, hydraulic classifiers,hydrocyclones, grizzlies, and screens. Following is a description of thesetechnologies. Applications, limitations, and relative costs for the first three listedparticle classification technologies are identified in Table 2.

Impoundment basins allow suspended particles to settle by gravity or sedimentation.A slurry of dredged material is introduced at one end of me basin; settling of solids— depending on the particles' diameters and specific gravities — occurs as the slurryflows slowly across the basin. The flow resulting at the opposite end has a greatlyreduced solids content. Multiple impoundment basins in a series can separatesediments across a range of sizes.

Hydraulic classifiers are rectangular tanks that function similarly to impoundmentbasins. They have a series of hoppers along the length of a tank which collectssediments of various sizes. Motor-driven vanes sense the level of solids and activatedischarge valves as the solids accumulate in each hopper. Hydraulic classifiers may


be used in tandem with spiral classifiers to separate fine grained materials such asclay and silt. Portable systems that incorporate hydraulic and spiral classifiers areavailable.

Table 2. Summary of sediment/water separation techniques.

Technique

ImpoundmentBasin

HydraulicClassifier

Hydrocyclones

Applications

Used to removeparticles down to agrain size of 20 - 30microns withoutflocculants, and downto 10 microns withilocculants.

Provide temporarystorage of dredgedmaterial.

Allow classification ofsediments by grainsize.

Used to removeparticles from slurriesin size range of 74 -149 microns (finesand to coarse sand) .

Used to separate andclassify solids in sizerange of 2000microns or moredown to 10 micronsor less.

Limitations

Requires large landareas.

Requires long set-up time.

Hydraulicthroughput islimited to about250 - 300 tphregardless of size.

Not capable ofproducing a sharpsize distinction.

Requires use oflarge land area forlarge scaledredging or wheresolidsconcentrations arehigh.

Not suitable forslurries with solidsconcentrationsgreater than 10 -20 %.

SecondaryImpacts

Potential forground watercontamination.

No significantimpacts.

No significantimpacts.

RelativeCost

High

Medium

Medium

Hydrocyclones use centrifugal force to separate sediments. A hydrocyclone consistsof a cone-shaped vessel into which a slurry is fed tangentially, thereby creating avortex. Heavier particles settle and exit at the bottom while water and sediments exitthrough an overflow pipe. Hydrocyclones may be useful where a sharp separation byparticle size is needed.


Grizzlies are vibrating units reliable in the removal of oversized material, such asbricks and rocks. Grizzlies are very ragged and are useful in reducing the amount ofabrasive material in order to minimize wear on subsequent, more delicate,technologies.

Screens may be vibrating or stationary and operate by selectively allowing particlesto pass through them. As the slurry passes over the screen, fine-grained particles andwater sift through the screen and larger particles slide over the screen. Screens comein a variety of types with a variety of applications to contaminated sediments.

Handling and Rehandling

The amount of handling and rehandling required by various pretreatment options willalso influence pretreatment decisions. Especially with severely contaminated sludges, allequipment mat comes in contact with the sediments will require subsequentdecontamination. For example, air drying heavily contaminated sludges requires that thesediments be put in the drying structure and later rehandled when the dewatered sludgesare removed in a more highly concentrated form. Rehandling also mechanically disruptsthe sediments and increases die probability of introducing contaminants into theenvironment. Conversely, a series of pretreatment steps requiring rehandling may be themost efficient way of separating the contaminated sediments and preparing them fortreatment. For example, the Dutch are using particle classification (hydrocyclones) anddewatering (belts, filter presses, chamber filter presses, and others) to improve thequality of solids prior to treatment or disposal. Although methods of treatment are notaddressed in this chapter, the water generated during dewatering generally containscontaminants and suspended solids, and may require further treatment.

Use of Drying Beds

Drying beds are among the first method of sludge dewatering. They consist simply ofshallow ponds with sand bottoms and tile drains as illustrated in Figure 1. Sludge ispumped to these beds at a depth of 6-12 in. and the time required for the sludge tode water to a liftable consistency ranges from several weeks to several months. Innorthern climates, sand beds can be covered with a greenhouse roof to facilitateevaporation.

The removal of water from sludge in drainage beds is in two stages. Initially, thewater is drained through the sludge into the sand and out tile drains. This processmight last a few days, until the sand is clogged with the fine particles and/or all thefree water has drained away. Further dewatering occurs by evaporation.

During the initial drainage a considerable portion of the water is first drained bysettling of the solids and compaction. Secondary drainage occurs by the formation ofchannels that facilitate the movement of water.


c

A

Tf

3

3

6

***»..*•, J*

i

i

.,.„-»,„»,

1

1

..»......," r *i

'i

L»•

i

i

_..«.....- in ,' vitrified pipe l*kf -^ \with plcttic joints

]

J

!

\\

\I:

Spl»»h boxii

V JJ...

J.

ZFSft&li"*fet

1

1

|

\i|1

(

\

j-ictr*ct*

i

tif*s

1

B.iA ft

TQ

• 9

;

J |« 1la1iII.5 *

i

»....»...,

1

1

^_jTTTTr "IT». tmnôu

shear oaie1

••*

:;

».i"

•_ __UL

...

1

.**»«...».»

1

1

1

1

.p

£

c

™-:-:̂ {|

1

J

^""^yti""^^

!

j

II EII: iII

1!H|i t15

IsCJj|w.,.̂ . J

A

1

PLAN

6->n. fine sand3-trt coarse sand33 *n medium grivnl3 to 6-in COWM 9'§w«l

2-lA,

Pip* column forQlMt-OVW

S-in.with open jam**

S€CTION A-A

Figure 1. Typical sand drying bed.

As with other sludge handling methods the design of sludge drying beds is based onexperience or, if the sludge is available, by scale-up from laboratory or pilot tests.Most often it is necessary to rely on experience because sludges are not available insufficient quantities for testing.

For open beds rainfall must be taken into account. A rule of thumb is that about 57%of the rain is absorbed by the sludge and must later be evaporated, the remainder


draining away. On the average, therefore, 57% of the rainfall must be added to theamount of water requiring evaporation.

Use of Vacuum Filtration

Vacuum filters have been used for decades in wastewater treatment for the dewateringof sludges. It is generally necessary to first digest sludge to increase filterability,although this is not always the case. Digestion does, however, reduce the odorproblem, thus making the operation less objectionable than it might be otherwise.

The filter operates as illustrated in Figure 2. The chemicals used to condition thesludge and, thus, make it easier to filter are combined with the feed sludge anddumped into a trough, which is situated underneath a large rotating drum. The drumis covered with a permeable fabric or other materials. A vacuum is drawn inside thedrum, and water is sucked through the fabric into vacuum lines inside the drum andpumped out as the filtrate. The solids that cannot get through the fabric are caught onthe surface of the drum and removed as the filter cake. Belt filters (where the fabricis lifted off the drum during cake discharge, as in Figure 2) tend to require greaterchemical doses than the older drum filters, where the sludge is scraped off with adoctor blade. The higher doses are required due to problems with cake release.

The objectives of vacuum filtration are to obtain a

a. high solids concentration in the filler cake,b. clean filtrate, and anc. acceptable filter yield.

The principle operational variable is the sludge to be filtered, since chemicalconditioning can be considered a means of changing the sludge. The type of sludgefiltered is within the control of the operator, and it is also within his province todictate what happens to the sludge before it gels to the vacuum filter.

The sludge solids concentration has a significant influence on the filter yield which isdefined as the pounds of sludge obtained from the vacuum filter per square root offiltered area per hour. An increase in the teed sludge solids concentration usually resultsin a substantial increase in filter yield. The filter yield in terms of pounds per square footper hour is often numerically equal to the percent of solids in the feed sludge. There is,a practical limit of about 8-10% because at greater solids concentrations, the sludgebecomes difficult to pump. An increase in sludge solids concentration also tends todecrease the amount of chemical needed, hence making it doubly important for thetreatment plant operator to try to introduce as concentrated a sludge as possible to thevacuum filter.

The amount of time that the sludge spends out of the treatment process decreases thefilterabilily of a sludge. For example, if aerobically digested sludge is being filtered,removing the sludge from the aeration tank for any length of time tends to decreasesludge filterability. Similarly, digested sludge tends not to filter as well if the sludge

220L

iquid Filtration

•» «3

5»•» «*

s

=

o

-5

a

I1I

Sludge D ewatering Operations 2 21

is first taken out of the digester and allowed to cool. The action of chemicalconditioners also deteriorates with time.

Dram submergence can be adjusted or the depth at winch the drum is set inside thetrough that contains the sludge can be controlled. If the drum is dropped lower into thesludge, a greater percentage of the cycle time is devoted to the pickup of solids fromthe trough and, thus, thicker but wetter filter cakes result. Decreasing the drumsubmergence tends to decrease the amount of time that the wet sludge is in contactwith the filter, and, thus, produce thinner but drier cakes.

By increasing the speed of the drum, the length of time the sludge is in contact withthe filter is decreased as is the time for drainage once die sludge has left the pool, thusobtaining a wetter cake. However, this tends to increase the sludge yield.Correspondingly, a slower drum speed tends to result in a drier cake and a lower filteryield.

The sludge in the trough must be agitated to prevent solids from settling and the extentof this agitation can be controlled. If the agitation is too violent, the floes willdisintegrate and poor filtration will result. If the agitation is insufficient, settling willoccur resulting in operational problems.

One of the most important considerations is the selection of the proper filter medium.The available media can be categorized as open or tight. Open media have large poreswhile tight media have small openings. Although a tight filter will remove a higherpercentage of the fines, a medium can be so tight as to make filtration impossible, andthus, blind, or stop up completely, preventing further filtration. The types of mediain use were described in an earlier chapter.

A method of improving filter efficiency is to use compounds that can be mixed withthe sludge to improve its filterability. Filter aids were also described earlier. Thedisadvantage of filter aids is that the quantity of sludge will increase. The advisabilityof using filter aids is strongly dependent on the local cost, as well as the increase infiltration efficiency.

Precoat filters are widely used in the chemical process industry. In this operation,instead of mixing the filter aid with the sludge, the filter aid is placed onto the filterfabric before the sludge is introduced. This prevents rapid premature clogging of thesmaller pores and allows for a greater filter yield. Diatomaceous earth is an effectivebut expensive filter aid for pre-coat filters.

In dewateriug digested sludge, vacuum filters are typically able to form cakes ofbetween 20% - 40%. The filtrate quality can vary anywhere from 100-20,000 mg/1of solids, corresponding to solids recoveries within ranges between 50 - 99%. Thefiltrate is almost always returned to the head of the plant for processing because itcontains significant BOD (biological oxygen demand) and, thus, cannot be dischargedwith the plant effluent.


Waste activated sludge, unless mixed with primary sludges, is not normally dewateredon vacuum filters. An exception to this is pure oxygen sludges that can be filteredsuccessfully with the use of ferric chloride conditioning in a filter leaf.

Vacuum filtration is an old established technology. Although for the most part it is asuccessful technique, it does have several disadvantages. Some of these are highoperating costs, and from the cost of chemical conditioners. Many treatment plantshave had problems with blinding to the point where vacuum filtration was no longerpossible. Another problem associated with vacuum filtration is that the sludge lias astrong odor. The sludge in vacuum filtration is always open to the atmosphere andoften warm which are two conditions producing die most severe odor problems.

Use of Pressure Filtration

Pressure filtration differs from vacuum filtration in that the liquid is forced throughthe filter medium by a positive pressure instead of a vacuum. Among the most widelyused in the chemical process industry (and widely used in Europe for wastewatertreatment) is the filter press. As shown in Figure 3, the filter press operates bypumping the sludge between plates that are covered with a filter cloth. The liquidseeps through the filter cloth leaving the solids behind between the plates. When thespaces between the plates are filled, the treatment plant operator separates the platesand removes the solids.

UOVASLC MUD

CLEAR FILTRATEOUTLET

MATERIAL ENTERS JflP™ÛNOCft PftCSCUIIC lli***—

Figure 3. Schematic of a filter press machine.

Filter pressing is a cyclic operation. Different designs enable automatic cake removal.One method is to blow the cake off by sending compressed air through the filtratetubes. Another method is to make the filter cloth a moving belt. After each operation,the belt is shifted to allow the cake to drop off on either side.

Sludge Dewateriiig Operations 223

Use of Centrifugation

The solid bowl centrifuge (or the decanter) is the type of machine used in wastewatertreatment. This machine has the attribute of being able to dewater or at least separateout any solid from any liquid, as long as the solids are heavier. It is possible to use acentrifuge for many purposes in a wastewater treatment plant, for example, thickeningactivated sludge and when this is not needed, for the dewatering of digested sludge.

The conventional solid bowl centrifuge is shown schematically illustrated in Figure 4.It consists of an outer bullet-shaped bowl that rotates at high speed. The sludge ispumped through a central pipe into this rotating bowl and because of centrifugal force,this sludge hugs the inside walls of the bowl. The heavier solids will sink to the bottom(that is, to the inner bowl wall) and the lighter liquid will remain pooled on top. Thebowl acts as a highly effective settling tank, and nothing more. It is necessary toremove the sludge to make the operation successful and in the solid bowl centrifuge,(his process is accomplished by the scroll, or screw conveyor, that is placed inside themachine and rotates only slightly slower than the bowl. This screw action tends toconvey the solids up onto the inclined beach and out the open end. The centrate, orclear liquid, flows out the holes on the other end of the bowl.

The objectives of centrifugation are similar to those of other dewatering devices. It isnecessary to obtain a dry cake, a clear centrate, and a reasonable throughput or, in thelanguage of filtration, a centrifuge yield.

The variables involved in centrifugation are listed in Table 3. These may be classifiedas machine variables or operational variables, as before.

The bowl diameter is a machine variable controlled by the design engineer when theunit is purchased. Increasing the bowl diameter (and maintaining the same centrifugalforce, that is, slowing down the machine) will result in a longer retention time withinthe machine. This results in a higher solids recovery, much as it would in a settlingtank, but this solids recovery is at the expense of cake dryness. As the solids recoveryis increased, the smaller particles will also escape as cake, increasing the moisturecontent of the cake and decreasing the solids concentration.

Increasing the bowl length will also increase the residence time which will generallyresult in a high solids recovery, but the changes in cake dryness are not alwayspredictable. It is also possible under some circumstances, to also increase the cakesolids by increasing the bowl length.

Bowl speed is one variable that can be designed into the machine simply by selectingthe proper gear ratio. Increasing the bowl speed increases the centrifugal force and thesolids recovery, with possibly a concurrent increase in the cake solids as well. Thiswill occur only if solids recovery already approximates 100%, that is, when almostaU the solids are driven out the cake end. An increase in centrifugal force at that pointwill then increase the cake solids concentration, but this increase is often marginal,depending on the sludge.


Table 3. Variables affecting centrifuge performance.

Machine Variables Operational Variables

Bowl diameterBowl lengthBowl rotational speedBeach angleBeach lengthPool depthScroll rotational speedScroll pitchFeed point of sludgeFeed point of chemicalsCondition of scroll blades

Residence timeSludge characteristics (includingsludge conditioning)

MAIN SCREW

EFFLUENTCAKE

Figure 4. Two solid bowl centrifuge configurations.


The pool depth is a variable that can be varied while die machine is in operation. Somemachines require the removal of plugs at the end of the bowl to increase or decreasethe pool depth. An increase in pool depth will result in higher retention time, hencebetter solids recovery and a wetter cake.

Increasing the conveyor speed will force the solids out of the machine more quickly,thus leaving behind some of the wetter solids and increasing the solids concentrationin the cake while decreasing solids recovery. Similarly, the conveyor pitch willinfluence solids recovery and cake dryness. If the scroll pitch is increased, the solidswill be moved out faster but only the larger, heavier solids will he pushed out leavingthe wetter solids in the centrate. Increasing the number of leads in the conveyor willlikewise increase cake dryness at the expense of solids recovery.

It is also possible to change the feed point of the sludge within the machine. If the feedpoint is changed toward the beach, a wetter cake will often result while solidsrecovery will increase; this is because the sludge will have a longer travel time to dieend of the machine where the centrate exists.

A drawback of centrifugation is that the conveyor will be subjected to quite severewear because of its grinding against the sludge, especially if die sludge contains sanclor other gritty materials. A worn conveyor will result in a greater cake dryness sinceonly the heavier particles are moved out, but this is, of course, at the expense of solidsrecovery,

The bowl angle, or the angle at which the conical section comes off the cylinder, alsohas a strong influence on both cake dryness and solids recovery if the sludge to becentriraged comprises soft, fluffy materials. An increase in the angle will preventsome of the softer, fluffier solids from making their way up the beach, and these solidsare hence discharged in the centrate, thus, decreasing solids recovery and increasingthe cake solids concentration.

The sludge on the inclined beach lias, in addition to the force directed radiallyoutward, a force component called slippage force that pushes die sludge back into thecylindrical section of the bowl. If the solids still can flow, they will move under theconveyor blades, back to the cylinder, and not be expelled as cake. The critical pointof solids slip is reached when the solids emerge from the pool. Because of thisproblem, machines that are designed to handle soft sludges have small beach angles,or the machines operate with the pool level raised above the outlet. This method,called super pool, utilizes hydraulic force to help the solids out.

The centrifuge operates because two basic processes occur within its bowl. One isclarification, or the settling of solids from the mother liquid. The second is diesuccessful movement of these solids out of the bowl. If either process is not performedsuccessfully, the centrifuge will not work. It is necessary then to measure in thelaboratory test how a sludge will clarify as well as to estimate how well a sludge willbe moved out of the bowl.


Centrifugation has many advantages over other dewatering methods. In addition tolower capital cost, a reasonable relative operating cost can be achieved, especially iflabor costs are taken into account. A centrifuge is an enclosed unit and is, thus, odorfree and can be used for dewatering such materials as heat treated sludges, which areobnoxious. Centrifuges can be accommodated in small buildings, or for that matter,outside. Chemicals are not necessary for operation but are generally recommended forcontinuous good operation and to avoid the buildup of fine solids in a treatment plait.A centrifuge is a flexible unit and is easy to start and clean up by running clean waterthrough the machine.

The disadvantages of centrifuges include the following. When chemicals are neededfor operation they can be costly. Trash can become embedded in the sludge and cloggthe machine very badly, causing many hours of downtime. Maintenance, especiallywith the wear and tear on the conveyor, can also lead to substantial maintenance costs.

Alternative Mechanical Dewatering Techniques

Other centrifugal type devices can be used for sludge diickening and/or dewatering.The cyclone consists of a cone into which the sludge is pumped tangentially asillustrated in Figure 5. The heavy sludge solids tend to move to the inside wall, down,and drop out die bottom while the clean liquid moves up and out the center. Cyclonesare quite useful to recover heavy materials and have found a use in degritting of sludgeprior to centrifugation.

The disc centrifuge machine has also been applied to thicken activated sludge. Thedisc machine, shown in Figure 6A has the advantage of being compact and efficientin its solids recovery of slow solids feeds due to high hydraulic capacity. The solidsare settled outwardly, and the lighter solids separated by settling against the undersideof the discs, gaining in density, and also sliding outward. The clarified liquid flowsout through the disc stack. The solids are removed through nozzles in the bowl. Thesenozzles can be so constructed to open intermittently, expel the solids, and close beforethe liquid can escape, or a portion of die solids can be recycled through the machine,thus, maintaining a high solids inventory. The latter arrangement results in bettersolids control as well as permitting the use of larger nozzles, thus, reducing the chanceof clogging.

The disadvantage of the disc machine lies in the clogging problem. Trash in the sludgeand unexpected power failures can cause clogged internals, and cleaning the machineis a tedious process. The stack must be disassembled one disc at a time.

Another centrifuge of some interest in wastewater treatment is the basket centrifugemachine shown in Figure 6B. The basket is simply a vertical rotating bowl without aconveyor. Periodically, as the solids collect on the sides of the bowl, a scoop comesby and scrapes out die solids as the cake.

Sludge Dewatering Operations

SAND REMOVAL TANK l',M

SANDY WATER

Figure 5. Cyclones used for degritting operations.

B

Disk Centrifuge Basket Centrifuge

Figure 6. Disc and basket centrifuge basket systems.

Suggested Readings

1. Francingues, N.R., M.R. Palermo, C.R. Lee, andR.K. Peddicord. ManagementStrategy for Disposal of Dredged Material: Contaminant Testing and Controls.Miscellaneous Paper D-85-1, USAGE Waterways Experiment Station.Vicksburg, MS, 1985.

2. Averett, D.E., B.D. Perry, and EJ. Torrey. Review of Removal, Containmentand Treatment Technologies for Remediation of Contaminated Sediments in the


Great Lakes. Miscellaneous Paper EL90-25, U.S. Army Corps of Engineers,Vicksburg, MS, 1990.

3. Fairweatlier, V. The Dredging Dilemma. Civil Engineering, August, pp. 40-43, (1990).

4. Palermo, M.R. et al. Evaluation of Dredged Material Disposal Alternatives forU.S. Navy Homeport at Everett, Washington. EL-89-1, U.S. Army Corps ofEngineers, Vicksburg, MS, January, 1989.

5. Cullinane, M.J., Jr., D.E. Averett, R.A. Sharer, J.W. Male, C.L. Truitt, andM.R. Bradbury. Alternatives for Control/Treatment of Contaminated DredgedMaterial. In: Contaminated Marine Sediments — Assessment and Remediation,National Academy Press, Washington, D.C., pp. 221-238, (1989).

6. Truitt, C.L. Engineering Considerations for Capping Subaqueous DredgedMaterial Deposits - Background and Preliminary Planning: EnvironmentalEffects of Dredging, Technical Notes. EEDP-09-1, U.S. Army EngineerWaterways Experiment Station, Vicksburg, MS, 1987.

7. Truitt, C.L. Engineering Considerations for Capping Subaqueous DredgedMaterial Deposits - Design Concepts and Placement Teclniiques:Environmental Effects of Dredging, Technical Notes. EEDP-09-1, U.S. ArmyEngineer Waterways Experiment Station, Vicksburg, MS, 1987.

8. Palermo, M.R. Design Requirements for Capping, Environmental Effects ofDredging, Technical Notes. In preparation, U.S. Army Engineer WaterwaysExperiment Station, Vicksburg, MS, 1991.

9. Culiinane, M.J., D.E. Averett, R.A. Schafer, J.W. Male, C.L. Truitt, andM. R. Bradbury. Guidelines for Selecting Control and Treatment Options forContaminated Dredged Material Requiring Restrictions. Final Report, U.S.Army Engineer Waterways Experiment Station, Vicksburg, MS, September,1986.

10. Sukol, R.B. and G.D. McNelly. Workshop on Innovative Technologies forTreatment of Contaminated Sediments, June 13-14, 1990. EPA/600/2-900/054,U.S. Environmental Protection Agency, Cincinnati, OH, 1990.

11. Miller, J.A. Confined Disposal Facilities on the Great Lakes. U.S. ArmyCorps of Engineers, Chicago, IL, February, 1990.

8

INDUSTRIAL WASTEWATER SOURCESIntroduction

General discussions on the applications of filtration to wastewater flows were givenin previous chapters. These discussions have provided the reader with anunderstanding of the objectives of wastewater treatment and how filtration can be usedas one of the unit operations for solids removal. Filtration must, however, be used inconjunction with other treatment technologies or operations, and does not necessarilyrepresent the primary treatment method. It is beyond the scope of this volume todiscuss the other techniques used in the treatment of wastewater flows. For generalinformation and background on other unit operations and processes used with filtrationin the treatment of waste waters, the reader may refer to the suggested references atthe end of this chapter.

The decision as to what type of filtration equipment to use, and indeed the decision onwhether filtration is even applicable in the treatment of a particular waste stream,depends on the characteristics of the stream and the requirements for treatment which,in fact, are often legal standards and specific to the wastes. This in part requiresknowledge of the pollutants generated during a manufacturing operation. Legalstandards for wastewater discharges exist in many countries. For the purposes ofdiscussions in this chapter and for the newcomer to this subject, these will be referredto as priority pollutants. They include biochemical oxygen demand (BOD), suspendedsolids (SS), total dissolved solids (TSD), pH, fecal coliform bacteria, chemical oxygendemand (COD), total organic carbon (TOC), nitrogen and nitrogen compounds. Theremay also be concern for such pollutants in wastewater streams as fats, oils and greasesof animal or vegetable origin because these could interfere with the operations ofcertain treatment works.

This chapter provides an overview of the characteristics of typical waste streams frommajor industry sectors. A review of the discussions in this chapter will provide thereader with the understanding of the sources of these wastes from each industry sector,

229


their general characteristics and types of pollutants, and when filtration is applicable.In most of the cases presented, filtration will play a major role in the pretreatment ofwaste streams and also, to a lesser degree, as a finishing stage for final solids removal.Industry terminology typically refers to this as polishing.

Paper and Allied Products Industry Wastes

Industry Description

This industry sector includes the manufacture of pulp from wood, rag, and othercellulose fibers and the manufacture of paper, paperboard, and building products. Themanufacture of converted paper and paperboard products from purchased paper is notincluded since it involves a relatively simple process, whose wastewater flows andloadings are not generally significant to the design of treatment systems. Therefore,the plants making converted paper and paperboard products are excluded fromdiscussions.

The manufacture of paper and allied products involves the preparation of wood andother raw materials, separation and recovery of cellulose fibers, and blending of thefibers with proper additives to produce "furnish", which is formed into paper. Theadditives include: sizing materials such as alum and resins, sodium aluminate, and waxemulsions; synthetics, such as acrylics, isocyanates, and fluocarbons; and fillers suchas clays, calcium carbonate and sulfate, talc, barium sulfate, aluminum compounds,and titanium oxide. When fillers are used, retention aids (starches or synthetic resins)are added to increase the retention of the filler.

The principal operations involved in the manufacture of pulp and paper are:

Wood PreparationPulping (mechanical, chemical, semi-chemical, and deinking)Pulp Washing and ScreeningStock PreparationPapermaking

The pretreatment sub-groups for this industry areas follows:

• Integrated pulp and paper mills using mechanical pulping processes (bleachedand unbleached)Integrated pulp and paper mills using chemical pulping processes (unbleached)Integrated pulp and paper mills using chemical pulping processes (bleached)Integrated pulp and paper mills using deinked pulpPaper and paperboard millsBuilding products mills

The following industrial practices can significantly influence pretreatment:

Industrial Wastewater Sources 231

Pulping Process — The pulping process determines the pulp yield and quality, and theprobable organics loss in the wastewater from a pulp mill. (Mechanical pulping resultsin minimum dissolution of wood components, while chemical pulping solubilizes thenon-cellulose components of wood tannins, lignins, wood sugars, and hemicellulose).

Recovery and Reuse of Spent Cooking Liquor — Most chemical pulping processesinvolve recovery and reuse of spent cooking liquor and, therefore, do not generatesignificant quantities of wastewater. The dissolved organics present in the liquorusually are burned in the chemical recovery furnace. The only pulping process wherethe spent cooking liquor is not suitable for recovery is the calcium-base acid sulfiteprocess. The spent cooking liquor from this process, with the dissolved organics, isgenerally discharged with other process wastes.

Bleaching -— When the desired quality of the final product requires bleaching of pulprecovered from wood, it is usually done by the addition of oxidizing chemicals, suchas chlorine, chlorine compounds, peroxides, and hydrosulfites. The oxidizingchemicals react with the non-cellulose portion of the pulp, rendering it soluble in wateror in alkaline solutions. As a result, the bleaching step adds to the wastewater volumeand pollutant loading.

Plant Pollution Control Methods — Control of spills and leaks, and recovery and reuseof chemicals constitute the major pollution control practices within the paper and alliedproducts industry. The extent of pretreatment required is largely dependent on theextent and effectiveness of the in-plant control processes adopted,

Wastewater Characteristics

The characteristics of the process wastewaters from each pretreatment sub-groups arelisted in Table 1. Integrated pulp and paper mills generally operate continuouslythroughout the year except for the shutdowns for preventive maintenance andequipment repair and replacement. Modem practice is to employ continuous pulpingprocesses, however, some mills around the world are still using batch pulpingprocesses which result in intermittent discharges of wastewater.

The overall wastewater characteristics from wood pulping processes may varyseasonally because of the changes in characteristics of wood and variations in thetemperature of the water. The volume and characteristics of the process wastewaterdepend upon the degree of water reuse, chemical recovery systems, and the type andquality of paper involved.

The wastewaters generated from the paper and allied products industry contain BOD,COD, suspended solids, dissolved solids, color, acidity or alkalinity, and heat.Chemical pulping processes may produce wastewaters with heavy metals (Cr, Ni, Hg,Pb, Zn). If pulp bleaching is part of the operation, the wastewaters may containadditional heavy metals (Hg) and dissolved solids (chlorides). Mercury may be presentin the caustic used in pulping and bleaching operations. Zinc is used in the bleachingof ground wood pulp. Chromium, nickel, and iron may be introduced from thecorrosion of process equipment.


When spent cooking liquor recovery is not practiced, the wastewaters may be acidic(pH in the range of 2 - 3) and have high concentrations of dissolved organics andinorganics. The solubilized organics and the type of cooking liquor will determine thecharacteristics of the wastewater.

Pretreatment Operations

The pretreatment unit operations which may be necessary for various types of jointtreatment processes are summarized in Table 2. The neutralization requirementsdepend primarily on the pulping process and on the related recovery and reusepractices. When spent cooking liquor is discharged to municipal sewers, special caremust be taken to insure control of pH, organic shock loads, and color (if present). Ifbleaching is part of the operation, pH adjustment may be necessary before mixing withother wastes. Spills of spent liquors and pulp washing water may introduce shockloads of pH and organic material if their discharge into the municipal sewer systemis not carefully controlled. In the absence of effective in-plant control procedures,adequate equalization and extensive pH and conductivity control may be required toprotect the operation of a joint treatment facility such as POTW which handle bothmunicipal and industrial wastes from paper mills.

The heavy metal concentrations in the paper and allied products industry wastewatersare very low and generally do not require pretreatment. Nevertheless, their levelsshould be determined to insure that effluent limitations are not exceeded where heavymetals are a significant factor.

Dairy Products Industry Wastes


This industry sector includes bulk handling, packaging, and processing (pasteurizing,homogenizing, and vitaminizing) of milk, and the manufacture of dairy productsincluding butter, cheese, ice cream, condensed evaporated milk, and dry milk andwhey. The manufacture of dairy products involves receiving and storing raw milk,separation of excess cream, pasteurization and homogenization, fluid milk packaging,and making butter, ice cream, and cheese. In the separation step, excess cream maybe skimmed off in order to standardize the butter fat content, or the raw milk may beseparated by centrifuge into cream and skim milk. Separated cream is then used inbutter or ice cream making, while the skim milk may be used in the production ofcottage cheese and non-fat dry milk solids. Natural cheese (i.e., not cottage cheese)is made with whole rnilk.

Some of these processes generate by-products which may be recovered and utilizedin other food manufacturing operations. Buttermilk, skim milk, and whey areproduced from butter and cheese making. The regional market potential for these by-products often determines the amount of in-plant recovery.

Table 1. Wastewater Characteristics Paper and Allied Products.

Characteristics

Industry Operation

FlowBODTSSIDSCOD

GritCyanideChlorine DemandpHColor

TurbidityExplosivesDissolved GasesDetergentsFoaming

Heavy MetalsColloidal SolidsVolatile OrganicsPesticides

PhosphorusNitrogenTemperaturePhenolSulfides

Oil & GreaseColiform (Total)

Mechanical Pulping(Bleached & Unbleached)

Year-round

ContinuousExtremely HighHigh t

AverageHigh

PresentAbsentHighNeutralLow

HighAbsentAbsentAbsentPresent

AbsentPresentAbsent4Absent

DeficientDeficientHighAbsentAbsent

PresentAverage

Chemical Pulping(Unbleached)

Year-round

ContinuousAverage-Ext, HighLow-HighHighHigh

PresentAbsentHighAcid-alkalineHigh

HighAbsentPresentAbsentPresent

PresentPresentPresenÂbsent


PresentAverage

Chemical Pulping(Bleached)

Year-round

ContinuousAverage-Ext. HighLow-HighHighHigh

PresentAbsentHighAcidicHigh

HighAbsentPresentAbsentPresent

PresentPesentPesent4Absent


PresentAverage

DeinkingPulp

Year-roimd

ContinuousHighHighHighHigh

PresentAbsentHighAlkalineLow


PresentPresentAbsentAbsent


PresentAverage

Paper andPaperboard

Year-round

ContinuousAverage-HighAverage-HighLow-averageHigh



AbsentPresentAbsentAbsent


PresentAverage

BuildingProducts

Year-round

ContinuousExtremely HighExtremely HighLowHigh


Very HighAbsentAbsentAbsentPresent

AbsentPresentAbsentAbsent


PresentAverage

High if bleaching of pulp is practiced.~ Acidic if bleaching of pulp is practiced.

Present if bleaching of pulp is practiced.Present only from log washing operations.

Table 2. Pretreatment Unit Operations for the Paper and AlUed Products Industry.

Suspended BiologicalSystem

Fixed BiologicalSystem

Independent Physical-ChemicalSystem

Mechanical Pulping(Unbleached)

Mechanical Pulping(Bleached)

Chemical Pulping(Unbleached)

Chemical Pulping(Bleached)

Deinking Pulp 1

Paper & Paper Board

Building Products

Coarse Solids Separation+ Grit Removal

Coarse Solids Separation4- Grit Removal +

Neutralization

Coarse Solids Separation+ Grit Removal +

Neutralization

Coarse Solids Separation+Grit Removal +

Neutralization

Coarse Solids Separation+ Neutralization

Coarse Solids Separation

Coarse Solids Separation

Where pulp and paper waste waters constitute more than about 10 % ofthe total wastewater flow, fixed biological systems or independentphysical chemical systems normally are not used. Where the pulp andpaper wastewater constitute less than this proportion, the pretreatmentrequirements will be the same as for suspended biological systems.

&SX

Equalization may be required in addition to those shown when batch pulping processes are used.


The dairy industry is a year-round operation. Raw milk receiving stations handle milkfrom the local farms for subsequent transfer to tank trucks. These stations are operatedon an intermittent daily basis. The rest of the industry will operate either continuouslyor on an intermittent basis depending on the economics of the individual facility. Ingeneral, larger processing plants tend to be integrated plants producing more than oneproduct.

The pretreatment sub-groups for this industry are as follows:

• Cottage and Natural Cheese Products• Milk Handling and Products.

Product recovery is the major method for reducing waste water loadings. The dairyindustry, however, must maintain sanitary conditions and this tends to limit the amountof wastewater recycle which is practicable.

The following industrial practices can have a major impact on the wastewatercharacteristics:

Whey Handling — Whey can be condensed and dried, and used as a food and feedsupplement. However, there are inherent difficulties in drying the acid whey derivedfrom cottage cheese because of its lactic acid content.

Operating Procedures — Spillage, overflow, and leakage caused by improperlymaintained equipment and poor operating procedures can result in major pollutionalloads due to the concentrated nature of the dairy products, e.g., whole milk has aBOD over 100,000 mg/L (0.8 pounds of BOD for every gallon of milk lost).

Cleaning — High efficiency in sanitizing operations is important to minimize the usageof sanitizers and detergents. In addition, rinses can be collected and used as make-upwater for sanitizers. Milk processing lines should be sloped to central collection pointsso that the milk product may be collected before the lines are sanitized.


The characteristics of the process wastewaters from each pretreatment sub-group areshown in Table 3. Daily wastewater flows are characterized as intermittent becausesome major unit processes, e.g., cheese and butter making, are batch, and becausemilk processing equipment must be shut down daily for sanitizing to maintain rigidhealth standards. Relatively clean water may be a substantial portion of the totalwastewater from a dairy plant. These waters are from condensers, refrigerationcompressors, milk coolers, and air conditioning systems.

The major types of wastewaters from the dairy industry are:

1. Wash and rinse water from cans, tank trucks, equipment, and floors. Ingeneral, the pH of the wastewater will be affected by the cleaning compound


(either acid or alkali) and the proportion of the wash water in the total plantwaste water flow.

2. By-products (such as buttermilk, skim milk, or whey) are sewered rather thanrecovered. Buttermilk and skim milk have BOD values as high as 70,000mg/L, while the BOD of whey is 50,000 mg/L.

3. Entrainment from evaporators and the sewering of spoiled or damagedproducts.

Table 3. Wastewater Characteristics Dairy Products.

Characteristics

Industrial OperationFLOWBODTSSIDS

CODGritCyanideChlorine DemandpH

ColorTurbidityExplosivesDissolved GasesDetergents

FoamingHeavy MetalsColloidal SolidsVolatile OrganicsPesticides

PhosphorusNitrogenTemperaturePhenolSulftdes

Oil & GreaseColi form

Milk HandlingMilk Products

Year-round (Batch)IntermittentAverage-HighLow-AverageAverage-High

Average-HighPresentAbsentHighAcid to Alkaline

HighHighAbsentAbsentPresent

PresentAbsentHighAbsentAbsent

PresentAdequateNormal-HighAbsentAbsent

PresentPresent

Natural and CottageCheese Product

Year-round (Batch)IntermittentExtremely HighAverage-Ext. HighHigh

Extremely HighPresentAbsentHighAcid to Alkaline

HighHighAbsentAbsentPresent

PresentAbsentHighAbsentAbsent

PresentDeficient ?

Normal-High"AbsentAbsent

PresentPresent

There are possible bio-static effects in the joint treatment plant attributable to large amounts of sanitizersand detergents in the dairy products wastewater.

' Temperature equal to or higher than domestic wastewater, may affect design but not harmful to jointtreatment processes.

The wastewaters generated in the dairy industry can be characterized generally ascontaining high concentrations of BOD, COD, and TDS. Settleable solids are not animportant consideration in most dairy wastewaters, since all the organic material is ina colloidal or dissolved state. However, filtration practices are still important becausesand or other gritty material may be present from tank truck washings. Cheesewastewaters, on the other hand, have higher concentrations of settleable solids due tothe presence of curd solids.


Design considerations in the joint treatment of dairy and domestic wastewaters are thehigh chlorine demand and the presence of surface-active agents. In addition, cheeseproduction wastewaters are usually nitrogen deficient. Septicity should be aconsideration in the design of any equalization or clarification facilities.

Low-rate trickling filters are generally not effective in treating dairy wastes, becausethese wastes produce large quantities of biological solids which clog the filters. Thisproblem can be overcome by high hydraulic loading and high recirculation rates.

Pretreatment

The pretreatment unit operations which may be necessary for various types oftreatment facilities are shown in Table 4. In general, dairy wastes are amenable tobiological as well as to chemical treatment if equalization and neutralization areprovided as pretreatment for the dairy wastes. Where whey is present, it may benecessary to control the rate of discharge and to provide nutrient supplements at thetreatment plant.

Table 4. Pretreatment Unit Operations for the Dairy Products Industry.

PretreatmentS ub-Group

Milk HandlingMilk Products

Natural andCottage Cheese Products


Equalization +Neutralization



Equalization -fNeutralization


Independent Physical-Chemical System



Textile Industry Wastes


The textile industry involves the manufacture of fabrics from wool, cotton, andsynthetic fibers; the synthesis or spinning of synthetic fibers is not included in thisgroup, but rather is included under synthetic organic chemicals. Of the three majortextiles, wool represents the smallest market and synthetic textiles the largest.

The major unit processes of the woolen textile industry include scouring, dyeing,fulling, carbonizing, bleaching, and weaving. Raw wool is scoured to remove greaseand dirt. The process employs a detergent and mild alkali at temperatures of 130°F.This operation is responsible for 55-75 % of the total BOD load from wool finishing.

The dyeing process uses dyes, dyeing assistants, (e.g. acetic acid, ammonium sulfate),and dye carriers containing heavy metals. The dye carriers will be present only if thewool is being combined with a synthetic fabric, which requires a dye carrier tofacilitate dye penetration. Various chemicals (e.g. sulfuric acid, hydrogen peroxide,and olive oil) may be added before and during the fulling operation. These chemicals


then enter the wastewater during a subsequent washing step. Carbonizing impregnatesthe wool with sulfuric acid to remove any traces of vegetable matter. Bleaching maythen be accomplished, with either sulfur dioxide, hydrogen peroxide, or opticalbrighteners.

The major unit processes employed in the cotton and synthetic textile industry includesizing, weaving, desizing, scouring, dyeing, and finishing. Chemicals used in thesizing process include starch, polyvinyl alcohol, carboxymethyl cellulose, andpolyacrylic acid. After weaving, the fabric is desized using an acid enzyme reaction.Desizing removes the chemicals added during sizing by hydrolyzing them to a solubleform.

During scouring cotton wax and other non-cellulosic components of cotton areremoved by using hot alkaline solutions. Synthetic materials require only lightscouring because of the absence of chemical impurities.

Both cotton and synthetic fabrics are treated with special finishes, using formaldehydeand urea, and with fire retardants, such as triaziridyl phosphine oxide.

The pretreatment sub-groups for this industry are:

• Wool• Cotton and Synthetic Fabrics

The following industrial practices can significantly affect the wastewatercharacteristics:

Segregation of Waste Streams — The segregation of waste streams permits recoveryof heavy metals, caustic recovery and reuse, and control of toxic spills (such asdieldrin used for moth-proofing). Many of the older textile mills have a commoncollection system with chemical reuse, but the modern mills have a segregatedcollection system to permit chemical recovery and reuse.

Alkaline Wool Scouring — Alkaline wool scouring may be used in place of neutralscouring. In alkaline scouring, soda ash is added to the wash water and subsequentlycombines with a portion of the wool grease to form natural soap. This procedurereduces the amount of detergent required and reduces the BOD and the concentrationof residual surface-active agents.

Chemical Sizing — The substitution of polyvinyl alcohol or carboxy methylcellulosefor starch in the sizing of cotton reduces the overall COD in the wastewater.

Pressure Dyeing Decks — The use of pressure dyeing decks in the place ofatmospheric units permits reduction in the amount of dye carriers required, thereby,reducing the BOD and heavy metal concentrat ions.



The characteristics of the process wastewaters from each pretreatment sub-group areshown in Table 5. The textile industry generally operates continuously throughout theyear except for scheduled shutdowns. However, many of the individual unit operationswithin the industry are batch oriented. Daily waste water flows are continuous withpeak flows occurring if certain batch-type operations are used. The textile industry isusually in a state of flux, with the manufacturing process continually being modifiedto reflect changes in consumer trends.

Table 5. Wastewater Characteristics of the Textile Industry.

Characteristics

Industrial Operation

FlowBODTSSIDSCOD



Heavy MetalsColloidal SolidsVolatile OrganicsPesticidesPhosphorus

NitrogenTemperaturePhenolSulfidesOil & GreaseColiform (Fecal)

Wool

Year-round (batch)

Intermittent -ContinuousHighHighHighHigh

PresentAbsentHighBasicHigh

HighAbsentAbsentPresentPresent

PresentPresentAbsentAbsentPresent

DeficientNormal-High 2

Absent 3

AbsentHigh5

Present

Cotton and Synthetics

Year-round (batch)

Intermittent -ContinuousAverage-HighLow-AverageHighAverage-High

AbsentAbsentHighBasicHigh

HighAbsentAbsentPresentPresent

PresentPresentAbsentAbsentPresent

DeficientNormal-High 2

Absent 3

AbsentAbsent-PresentAbsent

Wastewater flow characterized by an intermittent pattern over the day.

Temperature equal to or higher than domestic wastewater. Hay affect design but not harmful to jointtreatment processes.Phenol may be present in dye carriers.Oil present in wastewaters from synthetic textiles only.Wool processing wastewaters contain high concentration of animal grease.

Wastewaters generated in the textile industry are high in BOD, COD, TDS, and color.For synthetics, dyeing results in the largest BOD contribution, attributed to the use ofdye carriers such as methyl naphthalene, biphenyl, and orthophenyl phenol, all of


which have a high BOD. For cotton finishing, desizing contributes about 45 % of thetotal BOD, scouring about 30 %, and dyeing about 17%. The most significantdifference between wool process wastewaters and those from the rest of the industryis the release of wastewaters with high concentrations of suspended solids and greasefrom the wool-scouring operation.

Textile wastewaters can vary from slightly acid to highly alkaline depending on theindividual processes carried out within the plant. They generally are alkaline whencaustic scouring or mercerizing is involved. Heavy metals such as copper, chromiumand zinc result from the use of certain dye carriers in the dyeing operation of syntheticfabrics and of blended fabrics, e.g., cotton and rayon. The pretreatment unitoperations which may be necessary for various types of joint treatment facilities arelisted in Table 6.

Table 6. Pretreatment Unit Operations for the Textile Industry.

PretreatmentSub-Group

Wool

Cotton &Synthetics


Coarse Solids Separation+ Grease Removal +Chemical Precipitation(color, heavy metals) +Equalization +Neutralization

Coarse Solids Separation+ Chemical Precipitation(color, heavy metals) +Equalization -t-Neutralization


Coarse Solids Separation +Grease Removal 4-Chemical Precipitation(color, heavy metals) +Equalization +Neutral izat ion

Coarse Solids Separation +Chemical Precipitation(color, heavy metals) 4-Equalization +Neutralization


Coarse Solids Separation +Grease Removal +Chemical Precipitation(color, heavy metals) +Equalization +Neutral izat ion

Coarse Solids Separation 4-Chemical Precipitation(color, heavy metals) +Equalization +Neutralization

Pharmaceutical Industry Wastes

industry Description

This industry produces medicinal chemicals and pharmaceutical products, includingsome fine chemicals which are marketed outside the pharmaceutical industry asintermediates. In general, the pharmaceutical industry may be divided into two broadproduction categories: chemical synthesis products and antibiotics (penicillin andsteroids are examples). The manufacturing operations for synthesis products may beeither dry or wet. Dry production involves dry mixing, tableting or capsuling, andpackaging. Process equipment is generally vacuum cleaned to remove dry solids andthen washed down. The production of wet synthesis products and antibiotics is verysimilar to fine chemicals production, and uses the following major unit processes:reaction, extraction and concentration, separation, solvent recovery, and drying.

Wet synthesis reactions generally are batch types followed by extraction of theproduct. Extraction of the pharmaceutical product is often accomplished through


solvents. The product may then be washed, concentrated and filtered to the desiredpurity, dried, capsulized, and packaged.

Some antibiotics are produced in batch fermentation tanks in the presence of aparticular fungus or bacterium. The culture frequently is filtered from the medium andmarketed in cake or liquid form as an animal feed supplement. The antibiotic isextracted from the culture medium through the use of solvents, activated carbon, etc.The antibiotic is then washed to remove residual impurities, concentrated, filtered, andpackaged.


• Synthesis• Fermentation

The following industrial practices can significantly influence the wastewatercharacteristics:

1. Solvent recovery is practiced in both the synthesis and the fermentationproducts segment of the industry. Certain products may require a high-purity solvent in order to achieve the required extraction efficiencyrequired. This increases the incentive for making the recovery processhighly efficient.

2. Some solvent streams which cannot be recovered economically, areincinerated. Incineration is also used to dispose of "still bottoms" fromsolvent recovery units.


The characteristics of the process wastewaters from each pretreatment sub-group areshown in Table 7. The pharmaceutical plants operate continuously throughout the yearand are characterized by batch and semi-batch operations with significant variationsin pollutional characteristics over any typical operating period. The major sources ofwastewaters are product washings, concentration and drying procedures, andequipment washdown. Wastewaters generated from the pharmaceutical industry canbe characterized as containing high concentrations of BOD, COD, TSS, and volatileorganics. Wastewaters from some wet chemical syntheses may contain heavy metals(Fe, Cu, Ni, V, Ag) or cyanide, and generally have anti-bacterial constituents, whichmay exert a toxic effect on biological waste treatment processes. Considerationssignificant to the design of treatment works are the highly variable BOD loadings, highchlorine demand, presence of surface-active agents, and the possibility of nutrientdeficiency.


Pretreatment

The pretreatment unit operations which may be necessary for various types oftreatment facilities are shown in Table 8. The specific type and degree of pretreatmentfor heavy metals and cyanide will be governed by the industrial effluent guidelines forthe pharmaceutical industry. Cyanide removal or control is especially important.

Pharmaceutical industries generate wastewaters on an intermittent basis and equalizationmay be needed as pretreatment. When solvents are used for extraction, solvent removalcan be accomplished by using gravity separation and skimming. Neutralization may berequired to neutralize acidic or alkaline wastewaters generated from the production ofspecific pharmaceutical products.

Table 7. Wastewater Characteristics of the Pharmaceutical Industry.

Characteristics

Industrial OperationFlowBODTSSTDS




PhosophorusNitrogenTemperaturePhenolSu If ides

Oil & GreaseColifonn (Total)

Synthesis

Year-round (Batch)IntermittentAverage-HighHighAverage-High

Average-HighAbsentPresentAverage-HighAcid-Basic

Average-HighAveragePresentAbsentPresent

PresentPresentHighHighAbsent

DeficientDeficientNormal-HighAbsentAbsent

Absent-PresentAbsent

Fermentation

Year-round (Batch)IntermittentExtremely HighExtremely HighHigh

Extremely HighAbsentAbsentHighAcid-Basic

Average-HighHighPresentAbsentPresent

PresentAbsentHighHighAbsent

Deficient-HighDeficient-HighNormal-HighAbsentAbsent

Absent-PresentAbsent-Present

Temperature equal to or higher than domestic wastewater. May affect design but not harmful to jointtreatment processes.


Table 8. Pretreatment Unit Operations for the Pharmaceutical Industry.

PretreatmentSub-Group

Synthesis

Fermentation


Chemical Precipitation(Heavy Metais) + SolventSeparation \- Neutralization+ Cyanide Oxidation

Solvent Separation +Equalization +Neutralization

Fixed Biological System

Chemical Precipitation(Heavy Metals) + SolventSeparation + Equalization+ Neutralization +Cyanide Oxidation


IndependentPhysical-Chemical

System

ChemicalPrecipitation(Heavy Metals) +SolventSeparation + Equali-zation +Neutralization+ Cyanide Oxidation


Leather Tanning and Finishing Industry Wastes


This industry sector includes tanning, curing, and finishing hides and skins intoleather. A major portion of the output from the tanneries is from cattle-hideprocessing, with the other sources being pigskin, calfskin, goatskin, and sheepskinprocessing. The tanning process involves conversion of animal hide and skins intoleather. The grain layer and the cerium portion of the skins constitute the leather-making material and consist mainly of the protein collagen. During the tanningprocess, the collagen fibers are reacted with tannin, chromium, alum, or other tanningagents to form the leather. Four basic operations are involved in tanneries:

1. Beam House2. Tan House3. Retan, color and fat liquor4. Finishing

The beam house operation involves; storage and trimming of hides; washing andsoaking to remove dirt, salt, blood, manure and non-fibrous proteins; green fleshingfor the removal of adipose fatty tissues and meat; unhairing to remove epidermis andhair; bating to remove non-collagenous proteins; and pickling in some operations tostabilize and preserve the unhaired stock for subsequent operations. The beam houseoperation is typical of hide and skin processing with cattle-hide processing being themost important in many parts of the world.

The tan house operation consists of preparing the stock for tanning. Pickling is doneto make the skin acid enough to prevent precipitation of chromium during tanning.Two types of tanning are common; vegetable tanning and chrome tanning. Vegetabletanning is carried out in a solution containing plant extracts (such as vegetable tannin)


to produce heavy leathers such as sole leathers and saddle leathers. Light leathers,such as shoe upper leathers, are usually chrome-tanned by immersion in a bathcontaining proprietary mixtures of basic chromium sulfate.

Hides which have not been fully tanned in the chrome-tanning process may beretanned with either chrome, vegetable, or synthetic tanning agents. The fat liquorprocess involves the addition of many types of oils and greases to the tanned hides toprevent cracking and to make the leather soft, pliable, strong, and resistant to tearing.Coloring or dyeing of tanned leather may be done either before or after fat liquoringand uses either natural or synthetic dyestuffs. Finishing operations such as drying,coating, staking, and plating follow the foregoing wet processes. The pretreatmentsub-groups for this industry are chrome tanning and vegetable tanning.

In-plant pollution control techniques and chemical recovery practices in tanneries varydepending on the tanning process and the economics of chemical recovery systems.In vegetable tanning, it is common practice to recycle the tanning solution. (In chrometanning, tanneries usually practice recycling of the tanning solution.) Recovery ofgrease is normally practiced in pigskin and sheepskin tanneries.

The wastewater characteristics from the unhairing process will depend on whether theindustry is practicing a "save hair" or "pulp hair" operation. A low amount of sulfideremoves the hair with minimal damage, while a high amount pulps and partially dis-solves the hair. The "save hair" operation involves mechanical pulling and recoveryof hair. Dissolution of hair through chemical reactions is referred to as "pulping" or"burning",


The characteristics of the process wastewaters from each pretreatment sub-group areshown in Table 9. Most sub-processes within the tanneries are batch operated, and,therefore, the wastewater flow and characteristics fluctuate during the industryoperation. In addition, weekend shutdowns in some tanneries will result in wastewaterflow only during weekdays. The seasonal variations in wastewater flow are limited tothe variations in hide characteristics. In general, the waste characteristics and volumevary widely throughout the day and throughout the week in tanneries. The concen-trated waste fractions (lime liquors and spent tan solutions) are derived from batch-type processes. These fractions are therefore discharged intermittently.

Liquid process wastes are generated in tanneries from soaking and washing, fleshing,unhairing, bating, pickling, tanning, coloring, and fat liquoring. Auxiliary wastewatersfrom tanneries result primarily from boiler blowdown and from cooling, and representonly a minor fraction of the total waste load from tanneries. Therefore, only processwaste streams are considered for establishing wastewater characteristics andrecommending pretreatment.


Table 9, Wastewater Characeristics Leather Tanning and Finishing Industry.

Characteristics

Industry OperationFlowBODTSSIDS





Oils & GreaseConform (Total)

Chrome Tanning


Extremely HighPresentAbsentHighAcid- Alkaline

PresentPresentAbsentPresentPresent

AbsentPresentPresentPresentAbsent

DeficientAdequateNormal1

AbsentPresent

High-Low

Veqetable Tanning


Extremely HighPresentAbsentHighAcid-Alkaline

PresentPresentAbsentPresentPresent

AbsentAbsentPresentPresentAbsent

DeficientAdequateNormal1

AbsentPresent

High-Low

'-Temperature equal to domestic wastewater.•;-Oil and grease (animal origin) are significant only in pigskin and sheepskin processing wastewaters,

The characteristics of wastewaters from tanneries vary according to the type of hideprocessed and the tanning process (vegetable or chrome tanning) used. The tanningprocess is more of an art than science, and as a result the wastewater characteristicscan vary widely for the same type of hide and tanning process. The processwastewaters from tanneries contain as major constituents: BOD, COD, chromium, oiland grease, sulfide, suspended and dissolved solids, alkalinity, hardness, color, andsodium chloride. Significant pollutants that may be present in tannery wastes include:hair, hide, scraps, bits of flesh, blood, manure, dirt, salts, suspended lime, solubleproteins, sulfites, sulfides, amines, chromium salts, vegetable tannin, soda ash, sugarand starches, oils, fats and grease, surface active agents, mineral acids, and dyeswhich contribute to the BOD and COD.

In general, washing, fleshing, and unhairing operations produce more than half, 56% of the total volume and approximately 70 % of the poUutional load from tanneries.The tanning process generates from 5 - 20 % of wastewater volume and loading. The


dry finishing operations produce only minor quantities of wastewater from clean-upoperations.

The major wastewater sources from tanneries are the beam house and the tan houseoperations. The beam house wastewater is highly alkaline due to the large quantitiesof lime used in the process. The wastewater generated from the tan house is generallyacidic due to the discharge of spent tanning solution. Normally, these wastewaters aredischarged to a common collection sewer before treatment or discharged to amunicipal system. The unregulated batch dumping from beam house and tan houseresult in a total waste stream with varying pH values. Mixing alkaline waste streams(from the beam house) with the acidic chrome tanning wastes would result in partialprecipitation of chromium, which can be removed by clarificaftion. There is also thehazard of the evolution of hydrogen sulfide gas.

The other major pollutant in tannery waste is the effluent from the lime-sulfideunhairing operation. The concentration of sulfides in tannery wastes may vary between30 and 100 mg/L in the total effluent. The consequences of release of H2S gas in thesewer lines and the effect of reducing characteristics of the sulfides on biologicaltreatment processes should be taken into consideration in the design of treatmentworks. Hydrogen sulfide is readily released from solution as a corrosive and extremelytoxic gas with an obnoxious odor. By controlling the pH of the solution above 10.0,the H2S release can be minimized. If the pH of the wastewater is expected to be lower,the sulfide concentration should be reduced below 1.0 mg/L to prevent H2S odorproblems,

Pretreatment

The pretreatment unit operations which may be necessary for various types oftreatment processes are shown in Table 10. Screening to remove debris, equalizationto provide uniformity of effluent, and neutralization with precautions for possiblegeneration of hydrogen sulfide gas, to prevent excessively high pH values aregenerally necessary prior to discharge to a municipal collection system. Chemicalprecipitation may be needed to reduce the amount of chromium in the effluent.

The considerations in Table 10 are based on the assumption of fat and grease recoveryas a by-product. Where this is not practiced, grease removal facilities may also beneeded.

Petroleum Refining Industry Wastes


This industry is engaged in producing gasoline, kerosine, fuel oils, residual fuel oils,lubricants, and other products through distillation of crude oil, cracking, or otherprocesses. Petroleum refining is a combination of several interdependent processes,many of which are highly complex. There are more than two dozen separate processes


fundamental to the operation of a refinery producing the full spectrum of productsfrom crude oil. The major operations within a refinery include: crude oil storage;desalting; fractionation by pressure, atmospheric and vacuum distillation; thermal andcatalytic cracking; reforming; polymerization; and alkylation. Other operationsgenerally involving separation and finishing of the products to specifications, includeacid treatment of lubricating oil stocks, sweetening of gasoline, extraction, andstripping. Storage of crude oil to provide adequate working supplies and to equalizeprocess flow involves the separation of water and suspended solids from crude oil.

Table 10. Presentment Unit Operations for the Leather Tanning and Finishing Industry.

PretreatmentGroup

Chrome Tanning

VegetableTanning


Coarse Solids Separation+Gmt Removal +Equalization+ Chemical Precipitation(Heavy Metals) + SolidsSeparation +Neutralization

Coarse Solids Separation+Grit Removal +Equalization•f Neutralization


Coarse Solids Separation+Grit Removal +Equalization+ Chemical Precipitation(Heavy Metals) + SolidsSeparation 4-Neutralization

Coarse Solids Separaration+Grit Removal +Equalization4- Neutralization

IndependentPhysical-Chemical

System

Coarse SolidsSeparation +Grit Removal +Equalization

Coarse SolidsSeparation +Grit Removal +Equalization

1 Pretreatment requirements assume fat and grease recovery as saleable by-product.

The first major operation in a refinery is the crude oil desalting process for removinginorganic salts and suspended solids from the crude oil prior to fractionation. Water isused in the desalting process as a sequestering agent. The crude oil after desalting isgenerally passsed through atomspheric and/or vacuum distillation to separate lightoverhead products, side-stream distillates, and residual crude oil. Steam is used in thisprocess, and the steam condensate from the overhead accumulation is discharged aswaste water. The heavy fractions removed during the crude oil fractionation anddistillation process can be cracked using either thermal, catalytic, or hydrocrackingprocesses to yield light oil fractions such as domestic heating oil. The low-octanefractions obtained from the foregoing processes can be converted to yield high-octanegasoline blending stock by reforming, polymerization, and/or alkylation processes. Thereforming converts naphthas to finished high-octane gasoline. Reforming is a relativelyclean process producing a low volume of dilute wastewater. The polymerization processproduces waste waters containing sulfide, mercaptans, high pH materials, and nitrogencompounds. Phosphoric acid or sulfuric acid is used in the polymerization process andgenerates solid wastes. The alkylation process uses a sulfuric acid or hydrofluoric acidcatalyst to convert isoparaffins and olefins into high-octane motor fuel. Solvent refiningis used in a refinery to extract lubricating oil fractions and aromatics from feedstockscontaining various types of hydrocarbons.


The petroleum refining industry uses very large quantities of water for process andcooling purposes. Approximately 90 % of the water used in refineries is for coolingpurposes. Lesser water uses include: steam generation (boiler-feed.), directprocessing, fire protection, and sanitary uses. Steam is used in stripping anddistillation processes, where it comes in contact with petroleum products, therebycontributing to the total waste water flow from refineries.

Oily process wastes and oil-free wastes are collected separately in some refineries sothat the oily wastes can be treated for oil removal before mixing with other wastestreams. The spent caustics and spent acids are generally collected and sold ordisposed by other means. Few refineries neutralize these wastes for discharge, to thewaste water collection system.

The sour water (condensates from various fractionation units) containing sulfides andammonia is generally steam or air-stripped before being discharged to the sewer lines.Depending on the pH of the sour water, the stripping can reduce the sulfide andammonia concentrations in the final effluent.

In general, the in-plant control methods employed by the industry (sour-waterstripping, spent-caustic neutralization and oxidation, slop oil recovery, etc.) willdetermine the final effluent characteristics and the level of pretreatment required fordischarge to the municipal collection system. The following specific in-plant practicesare frequently employed:

1. Sour-condensate stripping is used to remove sulfides (as hydrogen sulfide,ammonium sulfide, and polysulfides) before the waste water enters the sewer. Thesour water is usually treated by stripping with air, stream, or flue gas. Hydrogensulfide released from the wastewater can be recovered as sulfuric acid or can beburned in a furnace. Hydrogen sulfides at concentrations in the range of 10 to 15mg/L can cause upsets in biological treatment plants, and removal of sulfides fromthe sour water by stripping would prevent such upsets.

2. Spent caustic neutralization is applied to both phenolic and sulfidic waste streams,but oxidation of spent caustics is limited to sulfide waste streams, since phenolsinhibit the oxidation of sulfides in spent caustics.

3. Spent acids (generally sulfuric) can be recovered for reuse or sold to acidmanufacturers, thereby avoiding their discharge to sewer systems. Spent catalystssuch as aluminum chloride and phosphoric acid can either be regenerated for reuseor disposed of as landfill.


The characteristics of the process wastewaters from the industry are shown in Table11. Petroleum refineries use gravity oil separators to recover free oil from processeffluents. The effluents from gravity oil separators are therefore used to define thewastewater characteristics hi the table. The petroleum refineries operate on acontinuous basis throughout the year except for separate process shutdowns for


preventive maintenance and equipment repairs. Many refineries have a segregatedwastewater collection system, with separate subsystems for clean and polluted wastestreams. The 'clean' waters may include pollution-free cooling waters, boilerblowdown, and cooling tower blowdown.

Table 11. Wastewater Characteristics of the Petroleum Refining Industry.

CharacteristicsIndustry Operation

FlowBOOTSSTDSCOD




PhosphorusMitrogenTemperaturePhenolSulfides


Year-Round

ContinuousAverage*LowHighHigh

PresentPresentHighAcid-AlkalineLow

LowPresentPresentLowAbsent

PresentLowPresentAbsent

Deficient'AdequateHigh2

HighHigh

HighLow

After gravity oil separation (API Separators).1 The refinery wastewaters have high COD/SOD ratios indicating the presence of biologically resistantorganic chemicals.: If cooling tower blowdown is also discharged with the process wastewater, phosphates may be presentdepending on water treatment.

The characteristics of wastewater drawn from storage tanks will depend on the qualityof the crude oil stored and may contain dissolved inorganics, oil, and suspendedsolids. The steam condensate from the overhead accumulator can be characterized ashaving oil, sulfides, mercaptans, and phenol. If barometric condensers are used invacuum distillation, the condenser water will have very stable oil in emulsion.However, if the barometric condensers are replaced by surface condensers, thecondenser water will be essentially free of oil.


The major wastewater from the alkylation process is the spent caustic from theneutralization of the hydrocarbon stream from the reactor. Even though the spent acidsare recovered as salable by-product, leaks and spills of acid catalysts could reach thesewer lines. The major pollutants from the solvent refining processes include solventssuch as phenols, glycols, and amines. Other processes for the manufacture of waxesand asphalt, and for finishing and blending of gasoline, produce relatively lowvolumes of dilute wastewater.

In general, the most significant process waste waters from petroleum refining are:crude oil desalting waste, storage tank draw-off, steam condensates, spent caustics,spent acids, product losses, and leaks and spills of solvents used in extractionprocesses. The process wastewaters, which come in direct contact with petroleumhydrocarbons, contain free and emulsified oil, sulfides, phenols, ammonia, BOD,COD, heavy metals, and alkalinity as major waste constituents.

Refinery wastewaters generally are susceptible to conventional biological treatmentmethods after adequate pretreatment. In addition, phosphorus supplementation may berequired in some cases to provide a nutrient-balanced system for biological treatment.This supplemental nutrient requirement will depend on the phosphorus content of thecooling tower blowdown and its inclusion in the process wastewater.

Pretreatment

The pretreatment unit operations for various types of wastewater treatment facilitiesare shown in Table 12. These pretreatment operations assume the following industrialpractices:

1. Sour condensate stripping to reduce sulfides and/or ammonia2. Spent-caustic neutralizations3. Spent-acid neutralization and recovery or disposal by other means4. Separate collection and disposal of acid sludges, clay, and spent catalysts5. Gravity separation of free oil from process effluents

Table 12. Pretreatment Unit Operations for the Petroleum Refining Industry.


Equalization + Coagulation -Solids Separation1 +Neutralization


Equalization + Coagulation -Solids Separation +Neutralization


Equalization -fCoagulation - |Solids Separation2 + jNeutralization 1

!' Pretreatment Unit Operations apply to API Separator Effluent.2 Combined with oil removal to insure oil concentration below 50 mg/1.

Depending upon the in-plant control methods used within a refinery it may benecessary to add sulfide removal and neutralization to the pretreatment operationslisted in Table 12. The oil concentration in the wastewater should be reduced to 50mg/L in order to insure trouble-free operation in secondary biological treatmentfacilities. In addition, the presence of oil in a sewer would constitute a fire and


explosive hazard. For this reason, sewer ordinances generally have prohibited thedischarge of refinery wastewaters to municipal facilities.

The heavy metals present in refinery effluents (As, Cd, Cr, Co, Cu, Fe, Pb, Ni, andZn) are generally in such low concentrations that they do not pose serious problemsfor conventional treatment methods. If heavy metals reduction should be required byeffluent guidelines, provisions should be made for their removal before dischargingto municipal systems. The biological sludge developed from refinery wastewaters canbe thickened and dewatered by conventional methods such as vacuum filters andcentrifugation.

The pretreatment unit operations which may be necessary for various types oftreatment facilities are shown in Table 12. The most prevalent in-plant treatmentmethods are sour-water stripping, neutralization and oxidation of spent caustics,ballast water treatment, and slop-oil recovery. These measures substantially reduce thewaste loadings and to a significant degree are required to protect subsequent treatment.In addition to these in-plant control methods, refineries use gravity oil separators torecover free oil from process effluents.

Food and Meat Packing Industry Wastes


This industry includes slaughterhouses, packinghouses, processing plants (beef,poultry, hog, and sheep), and rendering plants. Live animals are usually held inholding pens for less than one day prior to slaughter. In the killing area, the animalsare slaughtered and the carcasses drained of blood. The processes of skinning,defeathering or dehairing, and eviscerating follow the slaughtering of the animals.Depending on the desired product, carcasses may be cut into smaller pieces, e.g., hogsare cut into parts such as hams, sides, loins, and shoulders. These parts may be furtherprocessed (e.g., smoked or pickled), or they may be shipped directly to wholesalerswithout further processing. Because of the similarity of the wastewaters from meatproducts industry, there are no separate pretreatment subgroups for this industry. Thefollowing industrial practices can significantly influence the wastewatercharacteristics:

Sanitation Requirements - Unlike most other industries, the food industry is requiredto maintain strict sanitary conditions, which limits the amount of process wastewatersthat can be recycled. The industry practices in-plant recovery (e.g. blood or grease)to reduce wastewater strength. In most plants, blood is collected and subsequentlyprocessed. The recovery of blood represents an in-plant practice which is extremelydesirable since whole blood represents a BOD concentration of over 150,000 mg/L.In addition, the dry handling of such wastes as manure and bedding materials fromholding pens, paunch manure from eviscerating, and meat cuttings and trimmings willsignificantly reduce the quantity of waste materials discharged to the sewers.


Implementation of these practices will affect the concentration and quantity of wasteconstituents, but not the quantity of waste water,

Rendering — Rendering is a major unit process where in-plant modifications cansignificantly influence the pretreatment considerations. Wet and dry rendering are twosubprocesses presently used within the industry. In the wet process, the meat by-products in a batch tank are cooked by direct injection of steam. Dry rendering usesonly heat, and little wastewater is produced. In wet rendering, the solids in the waterphase are screened out, and the remaining tank water maybe evaporated or sewered.The tank water is a major source of organic pollution, when sewered, and has a highBOD value. Evaporation of wet-rendering tank water and the installation ofentrainment separators on barometric condensers may reduce the need for pretreatmentof wet-rendering process wastewaters.

Wastewater Segregation — Wastewater originating within a meat products plant willgenerally be made up of wastewater from the operations, sanitary wastes, andwastewaters from auxiliary sources (e.g., cooling water from ammonia condensers inthe refrigeration systems). Many large plants providing their own complete or partialtreatment have found it economical to segregate wastewaters into blood, clean water,manure-free water, and manure waters.


The characteristics of wastewaters from the meat products industry are shown inTable 13. The meat products industry is a year-round operation with daily operationon an intermittent basis. Plants usually shut down daily for an extensive clean-upperiod following the processing period. This practice results in the generation ofintermittent wastewater flows.

Each plant may have a number of operations, depending on the products and degreeof processing. However, the wastewaters generated from any meat products plant canbe characterized as containing high concentrations of BOD, COD, TSS, TDS, andgrease. Washdown of holding pens, mainly to remove manure and urine, will add tothe BOD and suspended solids concentrations.

The processes of skinning, defeathering or dehairing, and eviscerating are sources ofBOD, grease, and suspended solids. The disposal of paunch manure and the washingof carcasses during eviscerating are particular operations which generate substantialpollutants. The processing of hogs and fowl produce a floating solids problem causedby hair and feathers.

Dressing and processing operations are minor wastewater sources compared to theslaughterhouse operations. Wastewaters from these operations contain grease andsolids originating from equipment washdown and product losses. Druing theseoperations, various trimmings, fat, and fleshings are produced.

Considerations in the treatment of meat product wastewaters with domestic wastes arethe high chlorine demand, the presence of surface-active agents, focal conform,


intermittent flows, and the high septicity potential due to the high organic content ofmeat product wastewaters. In general, meat product waste waters are amenable toeither biological or chemical treatment.

Table 13. Wastewater Characteristics Meat Products Industry,

Characteristics



ColorTurbidityExplosivesDissolved GasesDetergentsFoamingHeavy MetaisColloidal SolidsVolatile OrganicsPesticides


Oil & GreaseColiform (Fecal)

Meat Products

Year-roundIntermittent1

High-Ext. HighHighHigh

High-Ext. HighAbsentAbsentHighNeutral

HighHighAbsentAbsentPresentAbsentAbsentHighAbsentAbsent

PresentPresentNormal-High2

Present3

Absent

PresentPresent

1 Wastewater flow is intermittent over the day or week.' Temperature equal to higher than domestic wastewater; may affect design but not harmful to joint treatmentprocesses,3 Phenols may be present in sanitizers used for clean-up.

Pretreatment

The pretreatment unit operations which may be necessary for various types oftreatment processes are shown in Table 14. In addition to screening and free-floatinggrease removal, various in-plant control practices, such as blood recovery, andseparate handling of paunch manure as solid waste would greatly reduce the wasteconstituents in process wastewaters. When meat product wastewater is combined withdomestic wastewater, pretreatment should include equalization to reduce organic andhydraulic fluctuations. The equalization basin should be aerated to prevent septicconditions.


Table 14. Pretreatment Unit Operations for the Meat Products Industry.

Suspended Bioloqical System

Coarse Solids Separation +Grease Removal2

Fixed Biolodical System

Coarse Solids Separation +Grease Removal2


Coarse Solids Separation+ Grease Removal

1. Assumes in-plant recovery and separate handling of blood, manure, and paunch manure.2. Only free-floating oil and grease.

Beverages Industry Wastes


This industry is engaged primarily in the manufacture of malt, malt beverages (ale,beer, and malt liquors), wines (table wine, dessert wine, and brandy), distilled spirits,bottled and canned soft drinks, and flavoring extracts and syrups. These products canbe classified under two major groups according to their basic manufacturing processesas:

1. Fermentation Products (beer, wine, distilled spirits, malt)2. Extraction Products (soft drinks, flavors, and extracts)

The fermentation products are made from grains or fruits, while the extractionproducts are made from flavor substitutes of oils such as cocoa, vanilla, and orangeoil. The fermentation products derived from grains are manufactured by cooking thegrains, fermenting the cooking liquor with a yeast culture, and separating thefermented alcohol by clarification and filtration.

The manufacturing processes used for the production of flavoring extracts and syrupsare normally proprietary in nature. The basic processes for the recovery of naturalflavoring can be listed as follows:

1. Steam distillation and petroleum ether extraction (essential oils).2. Expression (hydraulic pressing) and petroleum ether extraction (fruit syrup).3. Expression and evaporation (jams).4. Alcoholic extraction of vanilla and other tissue.

The soft drink baffling and canning plants use flavor extracts and purchased syrups.The bottling and canning process involves bottle washing and sterilizing, mixing offlavor extracts and syrup, carbonation, and filling.


• Malt, malt beverage, and distilled spirits (except industrial alcohol).• Wine and brandy.• Bottled and canned soft drinks, and flavors and syrups.


During the brewing and fermentation process, malt and hops are added to convertstarch to sugar and to incorporate a bitter taste to the product. Water is used in theprocess for cooking, cooling, container washing and other miscellaneous uses. Bothsolid wastes and liquid wastes are generated in the process. Spent grains, excess yeast,and spent hops are the solid wastes, and are generally hauled away or dried forlivestock and poultry feed. A variation of this type of disposal is where certaindistilleries manufacturing distilled wine or spirits, the stillage is discharged with otherliquid wastes. Liquid process wastes result from fermentation, aging, filtration andevaporation, and washing and clean-up operations. Liquid wastes are also dischargedfrom auxiliary operations such as cooling, boiler blowdown, and water softening.

The fermentation process results in the generation of "lees", which is a mixture ofwine, yeast cells, and other sediment. The lees is considered a liquid or semi-liquidwaste, which is either discharged directly to the sewer or recovered in the case oflarge wineries. In order to improve the quality of the wine, the fermented liquid isoften processed by a sequence of racking, filtration, and fining operations. The wastesfrom racking (clarification) and filtration often produce a sludge containing significantquantities of wine. When these wastes are sewered, they add significantly to the BODand solids concentration of the wastewater. Brandy is produced by distillation of wineand condensation of the overhead in order to obtain a beverage with high alcoholcontent. The stillage from such distillation is a significant liquid waste.

The manufacture of flavoring extracts and syrups also generates both liquid and solidwastes. Solid wastes are the residues after extraction of flavors and syrups. Waste-waters from normal extraction operations include:

Fruit Expression:

1. Water used for washing fruits.2, Hydraulic press clean-up.

Evaporation:

1. Evaporator condensate.2. Kettle wash water.

Steam Distillation:

1. Boiler blowdown.2. Bottoms from packed column.

The major wastewater sources in the bottling industry are the bottle washing andclean-up operations. Auxiliary wastewaters such as cooling, air conditioning, andboiler blowdown are also generated. The predominant method of disposing of liquidwastes from beverage plants is by discharging to municipal sewerage systems. Atypical plant collects all its wastewaters in a common sewer and discharges them tomunicipal sewers.



The characteristics of process wastewaters from each pretreatment sub-group areshown in Table 15. The beverage industries operate throughout the year. However,the volume of waste production and the loading will vary with the season dependingupon product demand.

Table 15. Wastewater Characteristics of the Beverages Industry.

Characteristics





PhosphorusNitrogenTemperaturePhenolSul fides

Oil & GreaseColiform (Fecal)Coliform (Total)

Mait Beverages andDistilled Spirits

Year-roundIntermittent-Continuous2

HighLow - HighHigh

HighPresentAbsentNo DataAcid-Neutral

PresentPresentAbsentPresent4

Present

PresentAbsentPresentPresentAbsent

DeficientDeficientNormal-High5

AbsentAbsent

AbsentAbsentPresent

Wine and Brandy

SeasonalIntermittentHigh-Ext. HighLow - Ext. HighHigh

High-Ext. HighPresentAbsentNo DataAcid-Alkaline3

PresentPresentAbsentAbsentPresent4



AbsentAbsentPresent

Soft Drinks Bottling1

Year-roundIntermittentAverage - HighLow - HighLow - High

Average - HighPresentAbsentNo DataAlkaline3

PresentPresentAbsentPresentPresent4


DeficientDeficientNormalAbsentAbsent

AbsentAbsentPresent

Pollutants characteristics represent only Bottling Industry; no data available for flavors and syrups.1 Malt beverages generate wastes on a continuous basis; distilled spirits waste flow will be cyclic.3 Alkaline pH due to caustic detergents used for bottle washing.4 Surface active agents are discharged primarily from bottle washing.5 Temperature equal to or higher than domestic wastewater, may affect design but not harmful to jointtreatment processes.

Major considerations in the treatment of beverage wastewaters are die presence oflarge paniculate matter in suspension and the fluctuations in hydraulic and organicloads. The following is a brief description of the wastewater from each pretreatmentgroup:


Malt Beverages and Distilled Spirits — The wastewaters generated from the malt andmalt beverages industries have as major constituents BOD, SS, pH, and temperature.The waste solids from the malt house, and the excess yeast, spent grains, and spenthops from the malt beverages industry are disposed of either by hauling away or byon-site drying to make cattle feeds. If the spent wet grain is dried in the brewery, thespent grain liquid must be disposed of; generally, it is discharged to municipal sewerswithout pretreatment. The distilleries produce wastewaters from cooking andfermentation of grains, the stillage or slops from distilling operations, and fromwashing and baffling operations. The stiliage from the distilleries contain yeast,proteins, and vitamins. Depending upon the size of the plant, complete or partialrecovery of stillage is practiced. The major constituents in distillery wastewatersinclude BOD, suspended solids, acidity, and heat.

Wine and Brandy — The wine and brandy industries produce wastewaters fromcrasher-stemmer, pressing, fermentation, clarification and filtration, distillation, andbottling operations. Brandy is manufactured by distillation of wine and, therefore,results in the generation of stillage or "still slop". The stillage is a significant liquidwaste in the manufacture of brandy. The wastewaters are high in organics and havethe potential of introducing shock loads in treatment works.

Soft Drinks, Flavors, and Syrups — The wastewaters generated from the manufactureof flavor extracts and syrups are generally discharged to municipal sewerage systemswithout treatment. The bottling and canning of soft drinks generate wastewatersprimarily from bottle-washing operations. These wastes contain BOD, suspendedsolids, and alkalinity. Bottling operating practice involves recirculation of final rinsewater for pre-rinsing, thereby reducing the volume of waste water discharged to thesewers.

Typical pretreatment operations include screenings, grit removal, and equalization.The pretreatment unit operations are listed in Table 16. The addition of auxiliarywastes (cooling, boiler blowdown, and water softening) will lower the strength of totaleffluent from the industry. In general, the wastewaters from the beverage industriesare amenable to treatment by conventional processes, such as activated sludge andtrickling filters. The pretreatment unit operations listed in Table 16 are based on theassumption that the following in-plant pollution control methods are practiced:

1. Hauling or drying of spent grains, hops, and stillage.2. Separate solids-handling and disposal of crusher'-stemmer and pressing wastes.

Spent grains, hops, stillage, crusher-stammer, and pressing wastes can be characterizedas solid wastes rather than liquid wastes. Therefore, it is desirable to collect themseparately for disposal.


Table 16, Pretreatment Unit Operations for the Bevaerages Industry.

Pretreatment Sub-Group

Malt Beverages

Wine and Brandy

Soft Drinks Bottling

SuspendedBiological System

Coarse SolidsSeparation+ Grit Removal +Equalization +Neutralization


Grit Removal +Neutralization

FixedBiological System




Independant Physical-Chemical System

Coarse Solids Separation+ Grit Removal +Equalization *+Neutralization

Coarse Solids Separation+ Grit Removal +Equalization +Neutralization


Plastic and Synthetic Materials Industry Wastes


Discussion on this industry sector covers the manufacture of plastic and syntheticmaterials, but not the manufacture of monomers, formed plastic products (other thanfibers), and paint formulations. The manufacture of resins used in paints is alsoincluded.

Plastics and resins are chain-like structures known chemically as polymers. Polymersare synthesized by one or more of the following processes: bulk, solution, emulsion,and suspension. After polymerization, the products undergo separation, recovery, andfinishing before being marketed. There are numerous plastics and syntheticsmanufactured in this industry and only a few are covered in this discussion.

The industry can be divided into the following pretreatment sub-groups:

• Rayon Fibers• Nylon Fibers• High- and Low-Density Polyethylene Resins• Urethane Resins, Polyolefin Fibers, Polyvinil Acetate Resins, Poly vinyl Alcohol

Resins,Polyester Fibers• Cellulosic Resins, Cellophane, Polypropilene Resins, Cellulose Acetate Fibers,

Polyvinil Chloride Resins, Polystyrene, ABS, SAN Resins, Phenolic Resins,Nylon Resins, Polyacetal Resins, Acrylic Fibers.

The following industrial practices can significantly influence the wastewatercharacteristics:


Suspension Polymerization — In suspension polymerization, a monomer is dispersedin a suspending medium consisting of a mixture of water and suspending agents suchas poly vinyl alcohol, gelatin, etc. The suspension is heated, and polymerizationoccurs. The polymer is then separated, washed, and dried. The concentrate from theseparation process may contain suspending agents, surface-active agents, catalysts,(e.g., benzoyl, lauroyl) and small amounts of unreacted monomers.

Emulsion Polymerization — Emulsion polymerization consists of solubilization anddispersion of the monomer in a solvent (e.g., water, cyclohexane, or tetrahydrofuran)with the appropriate emulsifiers (e.g., soaps or surfactants). Before polymerizationoccurs, initiators such as persulfates, hydrogen peroxides, etc, are added. Whenpolymerization is complete, the product is a milk-like latex of permanently dispersedpolymer, from which the polymer particles are recovered, generally by spray dryingor by coagulation and centrifugation.

Solution Polymerization — Solution-polymerization relies on a solvent to dissolve themonomer, catalyst and co-catalysts reaction, the polymer is precipitated using anantisolvent (e.g., n-hexane, methanol). The polymer is then filtered and dried.

Bulk or Mass Polymerization — Bulk or mass polymerization is different from theforegoing processes in that no carrier liquid is used. Therefore, there is generally littleor no process waste water associated with this process.


The characteristics of the process wastewaters from the manufacture of plastic andsynthetic materials are shown in Table 17. The plastic and synthetic materials industryis typically a continuous year-round operation. Because it is technically andeconomically advantageous, many firms manufacture several different, but relatedchemical products at one location. For example, a typical complex makes ethylene,polyethylene, sulfuric acid, ethyl chloride, ammonia, nitric-acid and phosphoric acid.

In the first group the wastewater has relatively high BOD, COD, and TSS; heavymetals (Zn, Cu) and synthetic fiber losses. The second group's wastewater has lowBOD and COD, may be either acidic or alkaline. The third group is characterized aseither having no process water or as having process wastewater containing virtuallyno pollutants. The fourth group's wastewater has variable BOD and COD, may beeither acidic or alkaline, and contains synthetic fibers. Discharge of faulty batchesfrom synthetic fiber plants may introduce shock loads into the wastewater treatmentfacility.

Conditions significant in the design of treatment facilities include high chlorinedemand, the presence of surface-active agents, high solids concentrations, and nutrientdeficiency. The process diversity and complexity, as well as the proprietary nature ofmany of the process chemicals, require that the pretreatment be established on a case-by-case basis after thorough investigation.


Table 17. Wastewater Characteristics for Plastics and Synthetic Materials Industries,

Characteristics

Industry OperationFlowBODTSSIDSCOD





Oil & GreaseColiform (Fecal)

Sub-Group 1

Year-roundCont.-VariableHighHighHighHigh

AbsentAbsentHighAcid-BasicLow-Average

HighAbsentAbsentAbsentAbsent

PresentHighAbsentAbsent


AbsentAbsent

Sub-Group 2

Year-roundContinousLowLowLowLow

AbsentAbsentLowAcid-BasicLow-Average

LowAbsentAbsentAbsentAbsent

AbsentLowPresentAbsent


PresentAbsent

Sub-Group 3

Year-roundContinousLowLowLowLow

AbsentAbsentLowNeutralLow

LowAbsentAbsentAbsentAbsent

AbsentLowAbsentAbsent


AbsentAbsent

Sub-Group 3

Year-roundContinuousAverage-HighLow-HighLow-HighAverage-High

AbsentAbsentAverage-HighAcid-BasicLow-Average

Low-HighAbsentAbsentPresentAbsent

AbsentAveragePresentAbsent

DeficientDeficientNormal-HighPresentAbsent

AbsentAbsent

Table 18. Pretreatment Unit Operations for the Plastic and Synthetic Materials Industry.

Pre-treatment

Sub-Group

1

2

3

4


Coarse Solids Separation +Neutralization -1- ChemicalPrecipitation (heavymetals)

Oil Separation +Neutralization

Pretreatment Not Required

Coarse Solids Separation+Neutralization


Coarse Solids Separation+ Neutralization +ChemicalPrecipitation(heavy metals)


Pretreatment NotRequired

Coarse Solids Separation+Neutralization


Coarse Solids Separation +Neutralization 4- ChemicalPrecipitation(heavy metals)


Pretreatment Not Required

Coarse Solids Separation +Neutralization

Oil separation required to reduce mineral oil (petroleum sources) concentration below 50 mg/L.


Blast Furnaces, Steel Works, and Rolling and Finishing Wastes


This industry sector includes: pig iron manufacture; manufacture of ferro-alloys fromiron ore and from iron and steel scrap; converting pig iron, scrap iron, and scrap steelinto steel; hot rolling; and cold finishing. Blast furnaces and by-product (or beehive)coke ovens are also included under this category, although these are almost non-existent in the United States today. The complex and interdependent operationsinvolved in a steel industry around the world can be listed as follows:

1. Coke Works2. Iron Works3. Steel Works4. Hot Forming5. Cold Finishing

Significant quantities of water are used, both for processing and for cooling purposes.The steel industry generates enormous volumes of waste water. In older plants aroundthe world, coke is used in large quantities for the production of pig iron. Old stylelarge iron and steel manufacturing operations include the production of coke fromcoal. There are two methods generally used for the production of coke: 1) the beehiveprocess; and 2) the by-product or chemical recovery process. The beehive processuses air in the coking oven to oxidize the volatile organics released from coal and torecover the heat for further distillation. The by-product or chemical recovery processis operated in the absence of oxygen, and the heat required for distillation is providedfrom external fuel sources.

In the byproduct process, coal is heated in the absence of air to a temperature at whichthe volatile matter is driven off. At the end of coking cycle, the hot residual coke isconveyed to a quenching station, where it is cooled with a spray of water. The off-gases from the coke oven are cooled in a cooling train, where tar and ammonia liquorseparate out. The tar contains a large proportion of phenols removed from the furnace.

Iron is manufactured from iron ore (iron oxide) in blast furnaces, with carbonmonoxide (from coke) as a reducing agent, again in older plant operations as in Russiaor the Ukraine. The major impurity (silica) in the iron ore is removed from the blastfurnace as molten slag, through the use of limestone.

Steel is manufactured from pig iron by adjusting the carbon content of the alloy toapproximately 1 %. The three principal steel-making units are the electric arc furnace,the open-hearth furnace, and the basic oxygen furnace. All three methods use thesame raw materials and produce similar wastes. Pure oxygen or air is used to refinethe hot iron into steel by oxidizing and removing silicon, phosphorus, manganese, andcarbon from the iron.


The steel ingot obtained from the furnace is reheated to provide uniform temperaturefor further processing or hot forming. The ingot steel is generally processed in ablooming mill or slab mill to form plates, sheets, strip, skelp, and bars.

The cold finishing operations are used for the conversion of hot-rolled products to givedesired surface, shape, or mechanical properties. These operations include pickling,cold rolling, tinplating, coating, shaping, and drawing to make various finishedproducts. Integrated iron and steel mills may operate many different subprocessesgenerating wastewaters of varying characteristics.

The only two types of waste stream susceptible to joint treatment are the coke ovenwastewaters and the pickling liquor. Most process wastes occur in such largequantities and contain only suspended impurities that it is more logical to treat themon-site and discharge directly to surface waters. Coke oven wastewaters and thepickling liquor are often found sent to municipal treatment works in many plants of theworld.

The strong acid pickle liquor, containing iron salts of mineral acids, is usuallycollected separately for other means of disposal or for use in waste treatment plants.The acid salts of iron are useful either as a flocculant aid, as a precipitating agent forphosphorus removal, or as a neutralizing agent in waste treatment plants. Any suchuse should be investigated before discharging the spent pickle liquor as a waste stream.Recovery of strong acid pickle liquor for reuse is also practiced for hydrochloric acidsystems, although this is rare.


The characteristics of die process wastewaters from various operations within the steelindustry are shown in Table 19. The wastewaters generated in the steel industry varywidely between operations and they are generally segregated for treatment. Someolder steel mills, however, still have common collection systems for discharging thetotal plant flow. The steel industry operates throughout the year and generateswastewaters over a 24-hour day. The volume and characteristics of waste water aresubject to hourly variations from batch dumping of acid baths and still bottoms.

The major constituents present in the wastewater are phenol, cyanides, ammonia, oil,suspended solids, heavy metals (Cr, Mi, Zn, Sn), dissolved solids (chlorides,sulfates), acidity, and heat. The process wastewaters are generally treated on-sitebefore disposal. Joint treatment of these wastes with municipal wastes is limited tosmall installations. A significant portion of the wastewater generated containssuspended solids and dissolved solids that are inorganic in nature.

The only two waste streams generally susceptible to joint treatment are the coke ovenwastewaters (ammonia liquor, still bottoms, and light oil recovery wastewaters) andthe pickling liquor. The other process wastewaters are high in solids (sub-micron ironoxide dust) and heavy metals. Pretreatment to reduce these waste constituentsgenerally results in an effluent which can be discharged directly to surface waters.


The coke oven process wastewaters are amenable to biological treatment when theyconstitute only a minor fraction (approximately 25 %) of the total waste water flow tothe treatment facility. The cyanide concentration (with its resulting aquatic toxicitycharacteristics) is the primary consideration in the treatment of coke oven processwastes.

Table 19, Wastewater Characteristics for Steelmaking Operations.

Characteristics

Industrial OperationFl owBODTSSTDS




CharacteristicsPhosphorusNitrogenTemperaturePhenolSu! fides


Coke Works

Year-roundIntermittentLow-AverageLowLow

Low-AveragePresentPresentHighNeutral

AbsentPresentAbsentPresentAbsent

AbsentAbsentPresentPresentAbsent

Coke WorksDeficientAdequateHighPresentPresent

PresentAbsent

Iron Works

Year-roundContinuousLow-Aver.Low-HighLow

Low-Aver.PresentPresentLowNeutral


AbsentPresentAbsentAbsentAbsent

Iron WorksDeficientAdequateHighPresentPresent

AbsentAbsent

Steel Works

Year-roundContinuousLowAver. -HighLow

LowAbsentAbsentLowNeutral



Steel WorksDeficientDeficientHighAbsentAbsent

AbsentAbsent

Hot Forming

Year-roundContinuousLowLow-HighLow

LowAbsentAbsentLowNeutral



Hot FormingDeficientDeficientHighAbsentAbsent

PresentAbsent

Cold Finishing

Year-roundIntermittentLow-AverageLow-HighHigh

Low-AverageAbsentPresentLowAcidic

AbsentPresentAbsentAbsentPresent

AbsentPresentPresentPresentAbsent

Cold FinishingDeficientDeficientHighPresentAbsent

PresentAbsent

The following are characteristics of the process wastewaters:

The still bottoms, containing phenol, constitute the major wastewater source from thecoke oven process. Since the beehive oven process utilizes the heat value in the off-gases, only the quench water is. discharged as wastewater. The gases (CO2, CO, N2,and HCN) leaving the furnace are hot and contain dust particles. The gases alsocontain water vapor and traces of hydrogen sulflde. In order to clean the exit gas fromthe blast furnace operations, the gas is generally passed through dust collectors,scrubbers, and coolers. The water used in the scrubbers and coolers is the primarywastewater source in iron manufacture. The waste products from this process are slagand the oxides of iron released as submicron dust particles. Precipitators or venturiscrubbers are used to clean the exit gas, and the characteristics of the wastewater


discharged from the process will depend primarily on the gas cleaning system. Ifscrubbers are used, the wastewater generated will be acidic in nature due to thepresence of sulfur oxides in the exit gas.

Water is used under high pressure to remove scale and for cooling purposes. Theprimary wastewaters are the scale-bearing waters and cooling waters containingprimarily scale and oil.

Steel pickling to remove oxides and scales is accomplished through solutions of H2SO4,HC1, or hydrofluoric acid. The pickled steel is then rinsed with water and coated withoil before proceeding to the next step in the process.

The cold rolling process involves passing unheated metal through rolls for reducingsize or thickness, and improving the surface finish. Plating of steel products is doneelectrolytically, and is accomplished in either an alkaline or an acid electrolytesolution. If acid electrolyte is used, the process system will consist of alkalinewashing, rinsing, pickling, plating, quenching, chemical treating, rinsing, drying, andoiling. The most commonly used metallic coatings are tin, zinc, nickel, chromium,cadmium, copper, aluminum, silver, gold, and lead. Wastewaters generated from coldfinishing operations include: rolling solutions, cooling water, plating wastes, picklingrinse waters, and concentrated waste-acid baths. Rolling solutions and cooling watergenerally contain oil and suspended solids as pollutants. Plating wastes, pickling rinse,and concentrated acid baths may contain various heavy metals (Cr, Cd, Ni, Zn, Sn)as well as cyanides, acids, and alkali.

The pretreatment unit operations which are often employed are listed in Table 20.Note that the wastewaters from the coke oven and cold finishing operations containcyanide and heavy metals and therefore require special consideration in waster watertreatment works.

Table 20. Pretreatment Unit Operations for Steelmaking Processes.


Coke Production

Cold Finishing


Equalization +Solids Separation

Equalization + OilSeparation orSkimming+ Chemical Precipita-tion (heavy metals) +Neutralization



Equalization + OilSeparation orSkimming+ ChemicalPrecipitation (heavymetals) -fNeutralization

IndependentPhysical-

Chemical System


Equalization -f OilSeparation orSkimming+ Neutralization


Organic Chemicals Industry Wastes


This industry sector includes the manufacture of a wide variety of products rangingfrom industrial gases and fertilizers to dyes, pigments, and petroleum compounds. TheOrganic Chemicals industry consists of such a complex combination of processes andproducts that a "typical or average" plant exists only in a statistical sense. The productmix and output of an industry depends primarily on the total economic activity and thedemand for products. The organic chemicals industry is very dynamic in developmentof new products and processes. Examples of only a few of the high volume productsmanufactured are:

l.Ethylene 15. Ethanol2. Benzene 16. Isopropanol3. Propylene 17. Acetic Acid4. Ethylene Dichloride 18. Cumene5. Toluene 19. Cyclohexane6. Methanol 20. Phenol7. Ethylbenzene 21. Acetaldehyde8. Styrene 22. Acetic Anyhdride9. Formaldehyde 23. Terephthalic Acid10. Vinyl Chloride 24. Dimethyl Terephthalate11. Ethylene Oxide 25. Acetone12. Xylene (Mixed) 26. Adipic Acid13. Butadiene 27. Acrylonitrile14. Ethylene Glycol

Depending upon the sequence of production from petroleum sources, chemicals arereferred to as either feedstocks or intermediate petrochemicals. Of the 27 examplechemicals listed there are 22 intermediate chemicals and five feedstocks (i.e.,ethylene, propylene, benzene, toluene, andxylene).

A review of wastewater characteristics indicates that certain products can be groupedtogether on the basis of pollutants present in the wastewater. Accordingly, the 27product chemicals covered under this category are divided into three subgroups asfollows:

Sub-Group 11. Benzene 9. Terephthalic Acid (TPA)2. Toluene 10. Dimethyl Terephthalate (DMT)3. Xylene 11. Ethylene4. Cyclohexane 12. Ethylene Dichloride5. Adipic Acid 13. Vinyl Chloride (Monomer)6. Ethylbenzene 14. Ethanol7. Styrene 15. Acetaldehyde8. Phenol 16. Acetic Acid


17. Acetic Anyhdride 21. Cumene18. Propylene 22. Ethylene Oxide19. Isopropanol 23. Ethylene dyed20. Acetone

Sub-Group 21. Butadiene2. Methanol3. Formaldehyde

Sub-Group 3Acrylonitrile

The chemical reactions involved in the production of the foregoing chemicals includepetroleum reforming, thermal or catalytic cracking, oxidation, alkylation, dehydrogenatton,hydration, and chlorination. Most processes use proprietary catalysts to increase productyield and to reduce severe operating conditions and pollution. Water is used extensivelyboth in the process and for cooling purposes.


The characteristics of process wastewaters from the manufacture of products undereach pretreatment group are shown in Table 21.

The characteristics of wastewaters vary from plant to plant, according to the productsand processes used. The organic chemicals plants generally operate 24 hours a daythroughout the year. Depending upon the product mix and the manufacturing process,hourly variations in wastewater volume and loading may occur as a result of certainbatch operations (filter washing, crystallization, solvent extraction, etc.). Thewastewater collection systems are generally segregated, to permit separate collectionof process wastewaters and relatively clean cooling waters. The process wastewatersare usually discharged to a common sewerage system for treatment and disposal.

The process wastewaters from the manufacture of chemicals under subgroup 1generally contain free or emulsified oil, while under subgroup 2 generally do notcontain oil. Acrylonitrile manufacture (subgroup 3) produces a wastewater containingcyanides and substantial quantities of acids. These wastewaters, in general, containunreacted raw materials and losses in products, by-products, co-products, and anyauxiliary chemicals used in the process. Detailed analyses for every specific chemicalpresent in the wastewater is difficult and are not generally used to describe thecharacteristics of wastes. In general the wastewaters contain: BOD, COD, oil,suspended solids, acidity, alkalinity, heavy metals, and heat.

The wastewaters discharged from the manufacture of products under subgroup 1 maycontain oil and grease and a series of heavy metals (Fe, Cd, Cu, Co, V, Pd). Thetypes and amounts of heavy metals in the wastewater depend primarily on themanufacturing process and the amount and type of catalysts lost from the process.


Most catalysts are expensive and, therefore, recovered for reuse. Only unrecoverablecatalysts (heavy metals), generally in small concentrations, appear in the waste water.The wastewaters generated from the manufacture of products under subgroup 2contain: BOD, acidity or alkalinity (pH in the range of 4 to 11), and heavy metals (Cr,Cu, Zn, Hg). These wastewaters are amenable to biological treatment afterequalization and neutralization. The production of butadiene may produce awaste water containing free or emulsified oil; an oil separation device may be requiredas pretreatment when the oil content in the waste water exceeds 50 rng/L. Onlyunrecoverable heavy metals (catalysts), generally in small concentrations, appear inthe waste water.

Table 21. Wastewater Characteristics of the Organic Chemicals Industry.

Characteristics

Industrial OperationFlowBODTSSIDS

CODGritCyanideChlorine DemandPH


FoamingHeavy MetalsColloidal SolidsVolatile Organ icsPesticides



Sub-Group 1

Year-roundContinuous-Varia bleAverage-Ext. HighLow-HighHigh

Average-Ext. HighAbsentAbsentHighAcidic-Alkaline

Low-AverageLowAbsentPresentPresent

PresentPresentAbsentPresentAbsent

DeficientDeficientNormal-High3

Low-HighPresent

Low-HighLow

Sub-Group 2

Year-roundContinuous-VariableAverage-HighLowLow-High

Average-HighAbsentAbsentHighAcidic-Alkaline

Low-AverageLowAbsentPresentPresent


DeficientDeficient2

High'PresentPresent

Low-HighLow

Sub-Group 3

Year-roundContinuous-VariableLow1

HighHigh

HighAbsentPresentHighAcidic

LowLowAbsentPresentPresent


DeficientAdequateNo DataAbsentAbsent

AbsentLow

1 Low BOD is probably due to the toxicity characteristics of this waste.! Adequate when butadiene is manufactured.' Temperature equal to or higher than domestic wastewater; may affect design but not harmful to jointtreatment.

The manufacture of acrylonitrile produces a highly toxic wastewater which is difficultto treat biologically. The toxicity characteristics are attributed to the presence ofhydrogen cyanide in excessive quantities. In addition, the wastewater is generallyacidic and contains high concentrations of organic carbon. These wastewaters are


generally segregated from other process wastes and disposed of by other means (e.g.,incineration).

Pretreatment

Table 22 shows the pretreatment unit operations which are often used. The heavymetals present in organic chemical wastes are in many cases so low in concentrationthat heavy metals removal is not required from the standpoint of treatabilitycharacteristics. However, the effluent limitations for heavy metals and toxic pollutantsmay require additional pretreatment (chemical precipitation) for removal of thesematerials,

The pretreatment unit operations generally consist of equalization, neutralization, andoil separation. In addition, phenol recovery (to reduce the phenol concentration) andspill protection for spent acids and spent caustics may be required in some cases.

Table 22. Pretreatment Unit Operations for the Organic Chemicals Industry.


1

2


Oil Separation +Equalization +Neutralization +Spill Protection 4Chemical Precipita-tion'

Oil Separation" 4Equalization 4Neutralization


Oil Separation +Equalization 4Neutralization 4-Spill Protection +Chemical Precipita-tion1

Oil Separation2 4Equalization 4Neutralization


Equalization +Neutralization 4ChemicalPrecipitation'

Equalization 4Neutralization

1. Need for chemical precipitation depends on extent of catalyst recovery.2. Oil separation required for butadiene manufacture only.

Meta! Finishing Industry Wastes

industry Description

This industry includes various types of plating, anodizing, coloring, forming, andfinishing operations. The metal-finishing industry operations are related closely tothose of many other industries, including transportation (automobile parts and ac-cessories), electrical, and jewelry.

The metal-finishing operation involves cleaning, conversion coating, organic coating,plating, anodizing, coloring, and case hardening. Acid pickling is the most commontype of cleaning of metal being prepared for plating. Sulfuric acid is the mostcommonly used pickling agent, but phosphoric, hydrochloric, hydrofluoric, and otheracids are used as well. Alkalies, dichromates, and numerous proprietary compoundsare also used in various combinations for descaling, degreasing, stripping,

Industrial Wastewater Sources

brightening, or otherwise preparing different metals (zinc, steel, brass, copper,for plating or anodizing.

269

etc.)

The plating solutions for nickel, chromium copper, cadmium, zinc, tin, and silver maybe basically cyanide, acid, or alkaline. Anodizing is done either in sulfuric acid or inchroniate solutions. Colorizing is accomplished with dyes, nickel acetate, andchromates. Cyanides are used in case hardening.


The characteristics of the process wastewaters from the industry are shown in Table23. The metal-finishing industry usually generates a continuous stream of rinse waterscontaining dilute concentrations of heavy metals and cyanide and intermittent batchdumpings of spend acid and cleaning solutions. The nature of metal-finishingoperations and the consequent fluctuating (cyclic) characteristics of the wastewatershould be taken into consideration in the design of treatment facilities.

Table 23. Wastewater Characteristics of the Metal Finishing Industry .

CharacteristicsIndustrial Operation

FlowBODTSSIDSCOD

GritCyanideChlorine DemandPHColor





Year-round (BATCH)

Continuous-VariableLowAverage-HighHighLow

PresentHighHighAcidicPresent


HighAbsentPresentAbsent

PresentPresentNeutralLowAbsent

PresentAbsent


Water is used extensively in metal-finishing processes to clean, strip, pickle, and rinsethe metal products before and after plating operations. The rinse waters constitute themajor volume of wastewaters, while spent solutions discharged intermittently addmajor pollutants to the total effluent. The wastewaters contain, in general, spent acids,alkalis, oil and grease, detergents, cyanides, and various heavy metals (Cr, Ni, Cu,Ag, Fe, Zn, and Sn). The metal-finishing plants differ from one another with respectto their processes, metals, and chemicals, and the characteristics of waste water mayvary widely from one to another. However, their wastewaters all contain primarilyinorganic pollutants, particularly heavy metals. In addition, the wastewaters frequentlyare highly toxic due to the presence of cyanides and heavy metals.

In general, the types of wastewaters from metal-finishing industries are:

1. Acid wastes2. Alkaline wastes3. Heavy metals wastes4. Cyanide-bearing wastes5. Miscellaneous wastes (dyes, soluble and floating oils, etc,)

Any of these wastewaters may occur as either dilute rinse waters or concentratedbaths. Except for the cyanide-bearing wastes, the wastewaters are generally connectedto a common sewerage system for treatment and disposal. The cyanide wastes usuallyare collected in a segregated sewer system in order to prevent the release of toxichydrogen cyanide gas under acidic conditions. However, the cyanide wastes can bemixed with other waste streams provided that any acid streams are neutralized priorto mixing with the cyanide waste stream.

The major constituants in the wastewaters generated from metal-finishing operationsare cyanides, metal ions, (Cr6+, Ni, Fe, Cu, Ag, and Sn), oil and grease, organicsolvents, acids, and alkalis. The wastewaters characteristically are so toxic andcorrosive to sewers and equipment that they require pretreatment before discharge tomunicipal sewers. A wide variety of processes are used in metal finishing operations,resulting in widely varying wastewater characteristics. Typically, these wastewatershave poor treatability characteristics without adequate pretreatment.

Pretreatment

The pretreatment unit operations for various types of treatment facilities are shown inTable 24, The pretreatment processes generally involve separate treatment of cyanidewastes and other acid wastes containing metal ions. The cyanide wastes can be treatedwith ferrous sulfate and lime to convert highly toxic cyanides to less toxic cyanates orcyanide complexes, or can be oxidized to CO2 and N2 with chlorine under alkalineconditions. The acid waste streams are treated first to reduce hexavalent chromium totrivalent chromium, using ferrous sulfate, scrap iron, or sulfur dioxide, and thenprecipitating the metal ions (Cr3+) as metal hydroxides.


Table 24. Pretreatment Unit Operations for the Metal Finishing Industry.


Equalization + Neutralization +CyanideRemoval + ChromiumReduction + ChemicalPrecipitation (Heavy Metals) +Solids Separation


Equalization 4- Neutralization+ Cyanide Removal+ Chromium Reduction +Chemical Precipitation(Heavy Metals) +Solids Separation


Equalization +Cyanide Removal+ Chemical Precipitation+ Neutralization

1. Chemical precipitation may not be needed, depending on the processes used in the independent physicalchemical joint treatment plant.

In addition to the effluent limitations and the processes shown in Table 24 the degreeof reduction in heavy metals waste loadings should consider the sludge handling anddisposal methods used for the metal finishing wastewaters. Some processes (e.g.,anaerobic digestion) concentrate these metals, and this can lead to process failureunless adequate pretreatment is provided. Dewatering operations discussed in Chapter7 should be referred to.

Closure

Filtration is an important unit operation hi the treatment of many industrial wastestreams. In the industry sources described in this chapter, filtration is often relied uponfor the removal of coarse and fine particles, and as both a pretreatment andposttreatment step. One must recognize however, that most filtration operationsinvolve physical separation, or in some instances, are combined as a part of biologicaltreatment or with chemical pretreatment methods as described in earlier chapters.Because many of the waste streams described are highly toxic in nature, chemicaltechniques including extraction, precipitation and others are the primary methods oftreatment. The references cited below provide more in depth coverage of this subjectalong with additional examples on the use of filtration with chemical treatmentmethods.

Suggested Readings

1. Profile of the Fabricated Metal Products Industry, US EPA Document 310-R-95-007, September 1995

2. Profile of the Non-Fuel, Non-Metal Mining Industry, US EPA Document EPA310-R-95-011, September, 1995.

3. Profile of the Stone, Clay, Glass, and Concrete Industry, US EPA Document 310-R-95-017, September 1995.

4. Profile of the Metal Mining Industry, USEPA Document 310-R-95-008,September 1995.

5. Profile of the Iron and Steel Industry, US EPA Document 310-R-95-005,September 1995.

FILTRATION EQUIPMENT ANDPROCESS FLOW SHEETS

introduction

This chapter provides a compendium of filtration machinery and auxiliary equipment,along with typical process flow sheets in diagrammatic form. The principle designfeatures and configurations for different filter machines are summarized along withexamples of flocculators, chemical feedstock systems for conditioning operations, andcentrifuges. The material presented in this chapter is designed to aquaint the readerwith generic design configurations, operating principles, mode or scheme of operation,and typical process flow sheets. This collection of schematics and process flow sheetswill assist the newcomer in establishing preliminary design concepts and process flowsystem layouts. The author has been careful not to discuss specific manufacturer'sequipment because it is not the intent of this book to provide specific endorsements orrecommendations for suppliers or vendors. Because of the overwhelming numbers,types, and variations of filtration equipment, not all commercially available systemsor process schemes are included, however, the reader will find many examples of themost commonly used systems throughout industry. To effectively use this chapter, thereader should first review the list of schematics below, and then turn to the page andfigure number on which the drawing appears.

Index to Equipment and Flow Sheet Diagrams

Figure 1. Design details of a Rapid Sand Filter 274Figure 2. Details of a typical Vertical Leaf Vacuum Filter 275Figure 3. Details of a typical Vertical Leaf Pressure Filter with

Vertical Tank Orientation 276Figure 4. Common Filter Operating Configurations 277Figure 5. Cross section details of an Upflow Filter System 278Figure 6. Details of a typical Pressure Filter 279

272

Filtration Equipment and Process Flow Sheets 273

Figure 7. Flow Control Systems : Influent flow splitting and variabledeclining rate filtration , 280

Figures. Design details of Underdraws for Sand Filters 281Figure 9. Cutaway view of a Rotary Drum Vacuum Filter 282Figure 10. Illustration of the Cake Processing Phases of a Rotary Vacuum

Filter .' 283Figure 11. Example of a Laboratory-Scale Vacuum Filter Apparatus . . . . 284Figure 12. Process flow sheet for a typical Vacuum Filter 285Figure 13. Process flow sheet for a typical Rotary Vacuum Filter System . 286Figure 14. Illustration of the Operating Zones of a Vacuum Filter 287Figure 15. Diagram showing the cross-section details of a Coil Filter . . . 288Figure 16. Diagram illustrating various Classifications of Centrifuges . . . 289Figure 17. Diagram of a Disc Type Centrifuge 290Figure 18. Illustration of a Continuous Countercurrent Solid Bowl

Conveyor and Discharge Centrifuge . 291Figure 19. Illustration showing cross-section of a Countercurrent Flow

Solid Bowl Centrifuge 292Figure 20. Schematic diagram of a Basket Centrifuge 293Figure 21. Typical process flow sheet for a Filter Press System . . . . . . . 294Figure 22. Side view of a Filter Press 295Figure 23. Cutaway view of a Filter Press 296Figure 24. Cross section of a Belt Filter 297Figure 25. Example of a Moving Screen Concentrator System 298Figure 26. Example of a typical Microstrainer Unit 299Figure 27. Schematic of a Moving Bed Filter 300Figure 28. Illustrates an approach to upgrading a Low-Rate Trickling

Filter to a High-Rate Trickling Filter 301Figure 29. Wastewater Treatment System after upgrading a two-stage

trickling filtration system 302Figure 30. Illustrates upgrading a single-stage trickling filter to a

Two-Stage Filtration System 303Figure 31. Illustrates upgrading a high-rate trickling filter using

a super-rate trickling filter as a Roughing Unit 304Figure 32. Common flow diagrams for single and two-stage

High-Rate Trickling Filters 305Figure 33. Chart listing types of Chemical Feeders 306Figure 34. Example of a typical Dry Feeder System 307Figure 35. Example of a typical Lime Feeder System 308Figure 36. Example of a Flocculant Diagram , 309Figure 37. Example of a Manual Dry Feeder System 310Figure 38. Example of an Automatic Dry Polymer Feed System 311Figure 39. Example of a Caustic Feed System 312Figure 40. Examples of alternative Liquid Feeder Systems for overhead

and ground storage 313Figure 41. Examples of a Mechanical Flocculation Basin and Flocculator 314Figure 42. Example of a Rotary Drum Conditioner 315

Operatingfloor

Pipe galleryfloor

Perforalaterals

Cast-iron,manifold

Rate of flow and tossof head Filter bed wash-

water troughs

Influent to filters

Concrete fittertank

Pressure lines tohydraulic vafoes fromoperating tables

Effluent toclear well

Drain

Figure 1. Design details of a rapid sand filter.


Figure 3. Details of a typical vertical leaf pressure filter with vertical tank orientation.


- OVERFLOW TROUGH

EFFLUENT

30-402-

UNDEHDRAIN—CHAMBER

EFFLUENT

(a), CONVENTIONAL FILTER

INFLUENT

0»), UPFLOW FILTER

SINGLE MEDIA FILTERS

i INFLUENT

(c), U-aOW FILTER

30-40" 21-48"

GAR NET SAND

(d), DUAL MEDIA FILTERS {*),MlXID-WiD»A FILTERSCTR1M.E MEDIA)

Figure 4. Common filter operating configurations.


COVER OPTIONAL(FOR CLOSED SYSTEM)

"GRID"

DEEP SAND LAYER

GRAVEL LAYERS

i

INLET RAW WATER

WASH WATER

FILTRATE OUTLET

SAND "ARCHES

SPECIAL VENT

AIR FORSANDFLUSH CLEANING

Figure 5. Cross section details of an upflow filter system.


50PSIS PRESSUREVESSEL-^.

I COUPLINGAIR RELEASE

MEDIA

o a a a a a

12 « IS" MANHOLEON VERTICAL£ OF TANK FILTER SUPPORTS

AT 1/4 POINTS

E L E V A T I O N

10 FLAN6E INFLUENTBACKWASH WASTE

o o Q a o o o

2"FLANQESURFACE WASH

10 FLANQf EFFLUENTAND BACKWASH

2 FILTER DRAIN

B'-0"0,0,

DISTRIBUTOR- 12"* l«" MANHOLE

SURFACEWASH

MIXED MEDIA

SUPPORT 0RAVEL

CONCRETE

LATERALS

S E C T I O N

Figure 6. Details of a typical pressure filter.


Flow SpliiiiAg T»nk

OBs

•I f»ltr CM

Wttti Trough

"I finer c««

wunTraygn

}

o[oClaud

eillutn!irflttw

M«BM*I 1noKtina "•

_*.

»^p

ellOpen

O

1 f i«« CtaNo. J

1 r,iw CM

V

41"Claud '

I IManujl IIKilalin̂ *"Vllv*

O

INFLUENT FLOW SPLITTING

COMMON IMn.UINTHE«OEIIPIK OH CHAMHCL

oKinee IH.ATC onSHORT VCNTIMI ranNATtWOI CATIONONL*.

VARIABLE DECLINING RATE FILTRATION

Figure 7. Flow control systems : influent flow splitting and variabledeclining rate filtration.


A. HEADER LATERALS(COUHfTESf OF THE *WWA)

•/•' DM COMTttOL WWICI1Ox.tFC* M. rr.

B. LEOPOLD BLOCK SYSTEM

(Covrttty r. B- Leopold Co-)

Figure 8. Design details of underdrains for sand filters.


AIR AND FILTRATE

CLOTH CAULKINGSTRIPS

DRUM

FILTRATE PIPING

CAKE SCRAPER

AIR BLOW-BACK LINESLURRY FEED

Figure 9. Cutaway view of a rotary drum vacuum fitter.

VACUUM

FilTER MEDIA

CAKE

COMB

CONVEYOR BELT

CHEMICAL

CONCRETE

S7« SLUDGE

Figure 10. Illustration of the cake processing phases of a rotary vacuum filter.


ALUMINUM TUBE

GASKET

PERFERATEDSUPPORT

PLATE

LEVELING SCREW

VACUUM TIGHT VALVE

TO VACUUM1̂

r~L_D1

•*• i ii***•

sV

1 1

— —.

,»-—

u•«•

"»

s .̂\ ^

X.

*

pHHrt

^^

CALIBRATEDPLEXIGLASSTUBE WITH

DRAIN

Figure 11. Example of a laboratory-scale vacuum filter apparatus.

COAGULANT POLYMER FLO* CONTROL

SLUDGE91 9,

c »

s-•*

o

9 I

V^

^J

SLUDQE CONDITIONING TANK*

riLTSITE REfURHTO PUN?

RECEIVER

FILTMTE

COUVEIfOB

FILTRATEPUMP

AIR TOATMOSPHERE

HATER

SILENCER

WATER TO PLANT

VACUUMPUMP

Figure 12. Process flow sheet for a typical vacuum filter.

A** TO ATMOSTHHf

SlUNCft

SlUOGE

\VACUUM PUMP

Figure 13. Process flow sheet for a typical rotary vacuum filter system.


Figure 14. Illustration of the operating zones of a vacuum fitter.

COIL SPRINGFILTER MEDIA

WASH WATERSPRAY PIPING DRUM

VACUUM ANDFILTRATE OUTLETS

CAKE DISCHARGE

AGITATOR DRIVE

Figure 15. Diagram showing the cross-section details of a coil filter.

FUtration Equipment and Process Flow Sheets 289

GEAR BOX r D R I V E SKEIV?

FEED

SOLID BOWL CENTRIFUGE

CAKE DISCHARGE «-

FIID

• C L A f l l F J E D EFFLUEMT

— FEED

i » EFFLUENT DISCHARGE

LUOOI DI9CMAHOI

DISC TYPE CENTRIFUGE

Figure 16. Diagram illustrating various classifications of centrifuges.


Figure 17. Diagram of a disc type centrifuge.

COVER

DIFFERENTIAL SPEEDGEAR BOX MAIN DRIVE SHEAVE

ROTATINGCONVEYOR

"~7- - FEED PIPESU~^ (SLUDGE AND

CHEMICAL)

CiNTRATCDISCHARGE

SLUDGE CAKEDISCHARGE

Figure 18. Illustration of a continuous countercurrent solid bowlconveyor and discharge centrifuge.

JSE.

292L

iquid Filtration

•


FEED

POLYMER!

SKIMMINGS

CAKE CAKE

Figure 20. Schematic diagram of a basket centrifuge.

10 •c

CL

11

1 Sludge In2 Mechanical Screen3 Sludge Storage Tank

4 Chemical Storage Tank5 Chemical Measurement

and Dilution Tank

6 Chemical Pumps7 Conditioning Tank8 Sludge Pumps

9 Filter Presses10 Cakes Out11 Filtrate Drain

Figure 21. Typical process flow sheet for a filter press system.

FIXED END TRAVELLING ENDELECTRIC

CLOSING GEAR

OPERATING HANDLE,

i

Figure 22. Side view of a filter press.


F I L T E R ClOTHSFIXED END

SLUDGE IN

FILTRATE DRAIN HOLES

Figure 23. Cutaway view of a filter press.


.DRUM

DISCHARGE ROIL

DISCHARGE ZONE

WASH ROU

WASH TROUGH

Figure 24. Cross section of a belt filter.

298L

iquid Filtration


D R I V E UNIT WASH WATERIETS

EFFLUENT WEI*

iHFLUEHT CHAMBER

EFFLUENT CHAMBER

Figure 26. Example of a typical microstrainer unit.


I N F L U E N T

I I A J H WH imtD I S C H I R C t

Figure 27. Schematic of a moving bed filter.

PRIMARY EFFLUENT185,OOO

FUtration Equipment and Process Flow Sheets 301

TRICKLINGFILTER SECONDARY

CLARIFIES

SLUDGE

'FINAL EFFLUENT

TREATMENT SYSTEM BEFORE UPGRADINGLOW-RATE TRICKLING FILTER

EXIST INGTRICKLINGFILTER

PRIMARY EFFLUENT -*-y370.010 GPD

NEW RECIRCULATION PUMPING STATION

EXISTING SECONDARYCLARIF IER

•FINALEFFLUENT

^CIRCULATION 185.000

ADDITIONALREQUIREDC A P A C I T Y

SLUDGE

TREATMENT SYSTEM AFTER UPGRADING

Figure 28. Illustrates an approach to upgrading a low-rate tricklingfilter to a high-rate trickling filter.


INTERMEDIATE-RATETRICKLING FILTER

PRIMARYEFFLUENT6.0 MGD SECONDARY

CLARIFIES

FINALEFFLUENT

TREATMENT SYSTEM BEFORE UPGRADING

SINGLE-STAGE INTERMEDIATE-RATE TRICKLING FILTER

REC1RCBLATION 7.5 HGDHE* RECIRCULATION PUMPING STATION

SLUDGE

P R I M A R YEFFLUENT6.0 MSB

1 ST. STAGE-NEW HE1HIGH-RATE FILTER INTERMEDIATE

CLARIFIES

2ND STAGE-EX 1ST INC EXISTINGINTERMEDIATE-RATE SECONDARYFILTER CLARIFIER

Figure 29. Wastewater treatment system after upgrading a two-stagetrickling filtration system.


RECIRCUUTION 6.0 KGO

PR1«ARTEFFIUEST2.0 MCO

TRICKLING FILTER

SLUDGE

F I H J t lEFFLUENT

SECQHUWCHRIFIER

TREATMENT SYSTEM BEFORE UPGRADINGHIGH-RATE TRICKLING FILTER

PR mmEFFUHHT—A2.0 MOD

E X I S T I N GT R I C K L I M OFILTER

r-NE»\ COMPLETELY MIXED/\ UHADOH TANK /

A

HEN 100* SLUDGEREClfCLE FHILITIES

•EXISTINGSECONDARY

11u.- _

11t

-«.T

F I N A LE F F L U E M T

Figure 30. Illustrates upgrading a single-stage trickling filter to atwo-stage filtration system.


P R I K H 8 YEFFLUENT,2.0 MGD

H E C I R C U U T I f l H 6 ,8 HCD

T R I C K L I N GFILTER

S E C O N D A R YC L A R I F I E S

R E C I R C U U T 1 0 NP U M P I H S S T A T I O N

F I H A LEFFLUENT

TREATMENT SYSTEM BEFORE UPGRADINGHIGH-RATE TRICKLING FILTER

PRIMiYEFFLUENT2.0 UGD

NEW RE CIRCULATIONPUMPING STATION

N E WSYNTHETIC MEDIAROUGHfm FILTER

1.S MGO

EXISTINGTRICKLINGFILTER

EXISTINGS ECO HO ARTCLARIFIED

EXISTINCR E C I R C U L A I I O HPUMP I KGS T A T I O N

F I N A LE F F L U E N T

TREATMENT SYSTEM AFTER UPGRADINGROUGHING FILTER PRECEEDING EXISTING HIGH-RATE FILTER

Figure 31. Illustrates upgrading a high-rate trickling filter usinga super-rate trickling filter as a roughing unit.


S I N G L E - S T A G ER

i s

c —*i K>

T i a - S T A G E

L E 6 E K O$ SLOOCE RETWNi KCitetutiB FUJIO FHIMWY CUItrtElO TRICKL1N8 FILTER

IKTERIIEOUTE CUKIFIEUFINil CIMIFIE*

"m: -ars

Figure 32. Common flow diagrams for single and two-stagehigh-rate trickling filters.


TYPES OF CHEMICAL FEEDERS

Type of Feeder

Dry feeder:Volumetric

Oscillating pbte

Oscillating throat (universal)

Rotating disc

Rotation cylinder (star) . . . .

Screw

Ribbon.

B«H .

Gravimetric:Continuous-belt and scale

Lou in weigh;

Solution feeder:Nonposilivf displacement:

Decanter (lowering pipe) - - -OrificeRotameter (calibrated valve)

Lou in weight (tank withcontrol valve).

Positive displacement:Rotating dipper

Proportioning pump:Diaphragm

PistonG« feeders

Solution feed

Direct feed.

1 Use special heads and valves for (lurries.

Use

Any material, granules orpowder.

Any material, any particlesize.

Most materials including NaF,granules or powder.

powder.

powder or granular.Dry, free flowing material,

powder, granular, or lumps.Dry, free flowing material up

to I'/i-inch size, powder orgranular.

Dry, free (lowing, granularmaterial, or floodablematerial.

Most materials, powder,granular or lumps.

Most solutions or light slurries

Most solutions or slurries . . . .

for i% slurries.'Most solutions, light slurries. .

Chlorine

Sulfur dioxideCarbon dioxide

Carbon dioxide

General

loader forarching.

Use hopperagitator tomaintainconstantdensity.

LimitationsCapacitycu ft/hr

0.01 to 35

0.02 to 100

0.01 to 1 0

8 to 2 000or

7.2 to 300005 to 18

0.002 to 0.16....

O.I to 3,000

0.02 to 2

0.02 to 80

0.01 to 100. 1 6 to 50005 to 0.16

or0.01 to 200.002 to 0.20

0. 1 to 30

0.004 to 0.1 5 .

0.01 to 170

8000 Ib/day max2000 Ib/day max7600 Ib/day max

6000 lb/d«y max300 Ib/day max1 20 Ib/day max

1 0.000 Ib/day max

Range

40 to 1

40 to (

20 to 1

10 to ior

100 to I20 to 1

SO to i

10 to ior

100 to 1

100 to 1

100 to 1

100 to 110 to !10 to 1

30 to 1

100 to i

100 to i

20 to 1

20 to 120 to 120 to i20 to 110 to 17 to 1

20lo 1

Figure 33. Chart listing types of chemical feeders.


COLLECTOR

PIPE (PNEUMATIC)

DAY HOPPERFOR DRY CHEMICALFROM BAGS OR DRUMS

BIN GATE

FLEXIBLECONNECTION

ALTERNATE SUPPLIES DEPENDINGON STORAGE

D R A I NSOLENOID VALVE

CONTROLVALVE

DUST COLLECTOR

6 FILL

SCREENWITH BREAKER

SCALE OR SAMPLE CHUTEFEEDER

__ GRAVITY TOAPPLICATION

PUMPTO APPLICATION

Figure 34. Example of a typical dry feeder system.


DUST COLLECTOR

FILL PIPE (PNEUMATIC)

NOTE: VAPOR REMOVERHOT SHOWN FOR CLARITY

UN GATE

FLEX I8LECONNECTION

FLOW RECORDERWITH PACIWTRANSMITTER,OR SAMPLE CHUTE

ROTAMETERS

SUKINQ NfcTEK

GRAVITY FEED

RECIRCUUTION

BACKPRESSURE

VALVE

Figure 35. Example of a typical lime feeder system.

TO CENTRIFUGE

1V 7•1-I f

ROTOMETERFLOC MIXING

EDUCTOR FUNNEL

FLOCCULANTFEED PUMP

Figure 36. Example of a flocculant diagram.

FtOCCULANTA MIXING TANK

FRESH WATEROR PLANT EFFLUENT


WATER SUPPLY

-DRYFEEDER

DISPERSER

MIXER

DISSOLVIHQ-AQIHQ

TANK

HOLD IMG TANK

SOLUTION FEEDER

POINT OFAPPLICATION

Figure 37. Example of a manual dry feeder system.

Filtration E

quipment and Process Flow

Sheets 3! 1

-I

ft,


TRUCK FILL LIKE

VENT, OVERFLOWAND DRAIN

DILUTIONWATER

SODIUM HYDROXIDESTORAGE TANK

POINT OFAPPLICATION

Figure 39. Example of a caustic feed system.


FLOAT

OVERHEADSTORAGETANK

CONTROLVALVE

5~-_f0 c'ROTAMETER

NR0TODIP-TYP£ FEEDER

GRAVITY FEED GRAVITY FEED

"METERI KG PUMP

G R A V I T Y FEED

PRESSURE FEED

X

VITY

,-ROTODIP-TYPE

!$£?>— rt̂ gjhy

FEED

TRANSFER PUMP

FE

/

f

-J

EDER /nHf—

i

/RECIRCUL

/GRt

-> STO^ T4

, CONTROL V A L V E

ROTAMETERrmo*

)UND

RAGENX

(

i -•<

U. 3*-

_i *- !f uui i 5 i I PR«It *• o **= £ FEQC 0 X r 5K

x 'uj °̂ T14J n UJ 1 ^^

o »— ! o_^ uJ 7 -*D * f m

H t f K ,1 C3 '

SUREED

,

Figure 40. Examples of alternative liquid feeder systems for overheadand ground storage.


MECHANICAL FLOCCULATION BASINHORIZONTAL SHAFT-REEL TYPE

MOTORIZED SPCtO NKDUCtft

NANOftAIL

auioeWATER PRESSURE LUBRICATED

MECHANICAL FLOCCULATOR

VERTICAL SHAFT-PADDLE TYPE

Figure 41, Examples of a mechanical flocculation basin and flocculator.


CHEMICAL FEEDCONNECTION

CONDITIONINGTANK SUPPORT

CONDITIONING TANKFEED CHUTE

FILTER VAT{REAR SIDE)

Figure 42. Example of a rotary drum conditioner.

INDEX

backflushing, 132-134backwashing, 155bed regeneration, 148-149belt filters, 96-98, 297belt filter presses, 212beverage industry wastes, 254-258biological activity, 147

cake filtration, 59-70cake filtration dynamics, 60-70cake formation, 11, 17, 62cake permeability, 50cake properties, 13, 46cake resistance, 65cake specific resistance, 45cake structure, 61cake volume, 65cartridge filters, 103-108cellulose, 51centrifugal filtration, 120, 122-123centrifugation, 223-226centrifuges, 120, 215, 289-293ceramic filter media, 41chamber filter presses, 215charcoal, 42, 51chemical feeders, 306-313chemical mixing, 155, 156, 159chemical sizing, 238clarifiers, 160-161cloth media, 45coagulation, 16, 144coagulation filtration, 150coal media, 41,42cocurrent filters, 91-98coil filter, 288

coke media, 42colloidal clays, 50colloids, 15compressible cakes, 63compression-permeability cell, 74concentrate disposal, 183-184cocurrent filters, 91connectivity, 5constant pressure drop filtration, 75-81constant pressure filtration, 66constant rate filtration, 57, 70-72,83-86contaminated groundwater, 171-172corrosion resistance, 16cotton cloth chemical resistance, 22-23cotton cloth filters, 22-23cross flow filtration, 170, 181, 195cross mode filters, 98-103crushed stone media, 43cyclones, 227

dairy industry wastes, 232, 235-237Darcy's law, 3-4, 64depth type filter media, 19dewatered cakes, 62dewatered solids, 211de watering operations, 12, 13dewatering technologies, 212-217,226diaphragm filters, 110-115diatomaceous earth, 41, 42diatomite, 50diffusion, 146

316

Index 3 1 7

disc filters, 101, 102drainage beds, 217-219drum vacuum filters, 89-90drying beds, 217drying operations, 14dynamic thickeners, 118-119, 121

ebonite media, 41electrokinetic forces, 11electrostatic attraction, 146equipment selection, 13, 16-18

fixed rigid media, 34flexible filter media, 20-34flexible metallic media cloths, 25flocculation, 144, 155, 156, 157-158flocculation filtration, 149flocculation units, 157flow control systems, 280fly ash, 51foam plastic media, 41food industry wastes, 251-254fouling, 166fouling control, 180

fabric filter media, 20fabric surface properties, 21fiber cloth filter media, 23filter aid applications, 48-50filter aid efficiency, 53-55filter aid precoating, 56filter aid requirements, 48filter aid selection, 51-57filter aids, 15, 19, 20, 47-57filter cake resistance, 65filter centrifuge, 122filter media, 11, 19filter media filtration, 59filter media selection criteria, 43-47filter media washing, 88filter medium, 88filter medium selection, 76filter plate, 66filter presses, 92, 98-100, 294-296filtrate, 8filtrate motion, 88, 91filtrate quality, 77filtration classification, 10filtration conditions, 12filtration constants, 67filtration cycles, 113filtration equipment, 13, 88filtration formulas, 75filtration mechanisms, 81-83filtration rate, 10filtration tests, 49filtration time, 74finishing industry wastes, 243-246,247

glass cloths, 21-22granular media filtration, 142-148gravity forces, 10, 60gravity thickening, 215grizzlies, 217groundwater remediation, 175

heat-resistant filter media, 26horizontal rotary filters, 93, 95-96hydraulic classifiers, 215, 216hydraulic conductivity, 4hydraulic resistance, 44, 50, 64hydroclones, 216hydrogen peroxide, 149hyperfiltration, 163, 173-190

impeller mixing, 157impoundment basins, 215, 216incompressible fluids, 9incompressible systems, 68internal rotary drum filters, 92-93

kinematic viscosity, 148Kozeny constant, 6

leaf filters, 100-101leather tanning industry wastes, 243-246, 247loose rigid media, 41


manmade packed media, 8materials handling requirements,182-183mechanical flocculation basin, 158membrane filtration, 163membrane processes, 172membrane separations, 169-170mesh size, 34metal finishing industry wastes, 268-271metallic cloths, 25-26metallic filter media, 34, 40methyl phenols, 165microporous filtration range, 126microstrainer units, 299model pore sizes, 2

nitrated cotton cloths, 22nitrogen compounds, 229nonmetallic cloths, 25-26nonwoven filter media, 26, 34Nutsch filters, 93, 96nylon cloth, 24

packing arrangement, 2paper production wastes, 230-232,233-234paper pulp, 23particle bridging, 19-20particle classification, 215-217particle settling, 85particle size distribution, 11perlite, 50permeability, 4, 8, 52petroleum hydrocarbonhyperfiltration, 196-202petroleum refinery industry wastes,246-247, 248-251pH control, 16pharmaceutical industry wastes, 240-243pilot plant filter assemblies, 17plastics industry wastes, 258-260pollution control, 1pollution prevention, 1polychlorinated biphenyls, 165

poly vinyl chloride, 41pore blocked filtration, 86pore blocking, 84, 85pore clogging, 12, 60pore size distribution, 45pore structure, 2porosity, 61, 62porous media, 1, 2, 3-9, 34porosity of perlite, 50potassium permanganate, 149powdered metal, 34precoat applications, 15-16, 52precoat filters, 221precoating, 49preconditioning, 150pressure drop across cakes, 64pressure filter, 279pressure filtration, 222 processeconomics, 184-193pulping process wastes, 231pumping wells, 175purification, 14

rapid sand filters, 145, 274rapid sand filtration, 153-155reservoir models, 3retentivity, 44-45reverse osmosis, 134, 135-141Reynolds number, 3rigid filter media, 34, 40-43rotary disc vacuum filter, 102rotary drum conditioner, 315rotary drum efficiency, 55rotary drum filters, 49, 89-91, 94rotary drum vacuum filter, 282-283roughing units, 304rubber media filters, 24

sand and gravel media, 42sand drying bed, 218sand filters, 281sawdust, 51screw presses, 123-124, 125sedimentation, 144-145semivolatile contaminants, 204-206slow sand filtration, 151-153

Index 319

sludge blanket filtration, 161sludge dewatering operations, 211-228solid bowl centrifuges, 224solids contact processes, 155, 159solids recovery, 14solids washing, 120specific capacity, 17specific resistance, 55specific volume, 65spiral wound membranes, 139-140steel industry wastes, 261-264strainers, 109-110straining operations, 146surface-type filter media, 19suspended solids, 229suspension properties, 13Sweetland pressure filter, 102synthetic fabrics, 238synthetic fiber cloths, 24-25

temperature control, 16tensile strength, 45textile industry wastes, 237-240theory of pore packing, 2thickeners, 116-119thin-cake filters, 115, 116-117thin-cake thickener, 116tortuosity, 6total dissolved solids, 229total organic carbon, 229total suspended solids, 229tubular ultrafiltration, 133

UF membranes, 124, 126 -134ultrafiltration, 124, 126-134unit operations, 1upflow filter system, 278

vacuum filter operation, 220vacuum filters, 46, 285-287vacuum filtration, 219vacuum rotary filtration, 215variable pressure filtration, 72-74variable rate filtration, 72-74vertical leaf vacuum filter, 275volcanic glass media, 50volume reduction, 179

washing techniques, 12waste shipping, 190-191wastewater sources, 229-271waste water treatment technology,142, 163wool cloth filter media, 23


Other Books from Butterworth-Heinemann

Adsorption Calculations and Modelling

by Chi Tien

1995 288pp he 0-7506-9121-2

Chemical Process Equipment Design

by Stan Walas

1988 755pp he 0-7506-9385-1

Colloid and Surface Engineering

Applications in the Process Industries

by R.A. Williams

1994 240pp pb 0-7506-1940-6

Engineering Processes for Bioseparations

by Lawrence Weatherby

1994 25Qpp he 0-7506-1936-8

Purification of Laboratory Chemicals, Fourth Edition

by W.L.F. Armarego and D.D. Perrin

1997 512pp pb 0-7506-3761-7

Detailed information on these and all other BH Engineering titles may befound in the BH Engineering catalog (Item #725), To request a copy, call1-800-366-2665. You can also visit our web site at: http://www.bh.com

These books are available from all good bookstores or in case of difficultycall: 1-800-366-2665 in the U.S. or +44-1865-310366 in Europe.

E-Mail Mailing ListAn e-mail mailing list giving information on latest releases, specialpromotions, offers and other news relating to BH Engineering titles isavailable. To subscribe, send an e-mail message [email protected] Include in message body (not in subject line):subscribe bh-engineering

Date post:	04-Dec-2014
Category:	Documents
Upload:	shesharam-chouhan
View:	111 times
Download:	2 times

127425110 Liquid Filtration

Documents