Delft 3D Environmental

Introductory courseEnvironmental

Version: 4.00.31088

28 March 2014

Introductory courseEnvironmental,

Published and printed by:DeltaresRotterdamseweg 185p.o. box 1772600 MH DelftThe Netherlands

telephone: +31 88 335 82 73fax: +31 88 335 85 82e-mail: [email protected]: http://www.deltares.nl

For sales contact:telephone: +31 88 335 81 88fax: +31 88 335 81 11e-mail: [email protected]: http://www.deltaressystems.nl

For support contact:telephone: +31 88 335 81 00fax: +31 88 335 81 11e-mail: [email protected]: http://www.deltaressystems.nl

Copyright © 2014 DeltaresAll rights reserved. No part of this document may be reproduced in any form by print, photoprint, photo copy, microfilm or any other means, without written permission from the publisher:Deltares.

Contents

ContentsFunctional specifications

Overview of topics

General introduction

Coupling GUI

Numerical aspects Delft3D-WAQ

Processes Library Configuration Tool (PLCT)

Modelling Oxygen - Bentic Oxygen Demand

Modelling suspended substances

Modelling Nutrients

Modelling Algae

Statistics

Toxic Substances

Open PLCT

Concepts of Particle tracking

Deltares iii

Contents

Deltares 1

Contents

Bract, odd page

Deltares 2

Contents

Bract, even page

Deltares 3

Delft3D Environmental

Functional Specifications

Version: 2.20.28529

27 March 2014

Delft3D Environmental, Functional Specifications

Published and printed by:DeltaresRotterdamseweg 185p.o. box 1772600 MH DelftThe Netherlands

telephone: +31 88 335 82 73fax: +31 88 335 85 82e-mail: [email protected]: http://www.deltares.nl

For sales contact:telephone: +31 88 335 81 88fax: +31 88 335 81 11e-mail: [email protected]: http://www.deltaressystems.nl

For support contact:telephone: +31 88 335 81 00fax: +31 88 335 81 11e-mail: [email protected]: http://www.deltaressystems.nl

Copyright © 2014 DeltaresAll rights reserved. No part of this document may be reproduced in any form by print, photoprint, photo copy, microfilm or any other means, without written permission from the publisher:Deltares.

Contents

Contents

1 Introduction 11.1 Areas of application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Delft3D framework overview . . . . . . . . . . . . . . . . . . . . . . . . . . 11.3 Utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 Water quality module 32.1 General introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.2 Module description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.3 Application areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.4 Coupling with other modules . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3 Sediment transport module 93.1 Module description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3.1.1 Cohesive sediment . . . . . . . . . . . . . . . . . . . . . . . . . . . 93.1.2 Non-cohesive sediment . . . . . . . . . . . . . . . . . . . . . . . . 103.1.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3.2 Application areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103.3 Coupling with other modules . . . . . . . . . . . . . . . . . . . . . . . . . . 10

4 Ecological module 114.1 Module description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114.2 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124.3 Coupling with other modules . . . . . . . . . . . . . . . . . . . . . . . . . . 12

5 Particle tracking module 135.1 Module description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135.2 Application areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

6 Pre-processing and post-processing 176.1 Visualisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

6.1.1 GPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176.1.2 QUICKPLOT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

6.2 Grid generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196.3 Grid data manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206.4 Grid aggregation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206.5 Tidal analysis and comparison with observations . . . . . . . . . . . . . . . 216.6 Tidal analysis and prediction . . . . . . . . . . . . . . . . . . . . . . . . . . 216.7 Nesting of Delft3D-FLOW models . . . . . . . . . . . . . . . . . . . . . . . 226.8 Nesting of D-Water Quality models . . . . . . . . . . . . . . . . . . . . . . . 226.9 Interfaces with other programs . . . . . . . . . . . . . . . . . . . . . . . . . 22

6.9.1 Interface to GIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236.9.2 Interface to Matlab . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

7 Hardware configuration 25

Deltares iii


iv Deltares

List of Figures

List of Figures

1.1 System architecture of Delft3D . . . . . . . . . . . . . . . . . . . . . . . . . 1

2.1 General overview of substances included in D-Water Quality. Substances areorganised in functional groups indicated by a grey header, except for somesubstances that form a group of their own. Major processes between sub-stances are indicated by arrows; note that many links are omitted . . . . . . 3

5.1 Example of a D-Waq PART oil spill simulation . . . . . . . . . . . . . . . . . 14

6.1 Example QUICKPLOT figure: 3D view of bed level . . . . . . . . . . . . . . 186.2 Example QUICKPLOT figure: Depth-averaged velocity vectors and drying and

flooding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196.3 Example QUICKPLOT figure: Depth-averaged velocity vectors and tidal ellips 24

Deltares v


vi Deltares

1 Introduction

Deltares has developed a unique, fully integrated modelling framework for a multi-disciplinaryapproach and 3D computations for coastal, river, lake and estuarine areas. It can carry outnumerical modelling of flows, sediment transport, waves, water quality, morphological devel-opments and ecology. It has been designed for experts and non-experts alike. The Delft3Dframework is composed of several modules, grouped around a mutual interface, while beingcapable to interact with one another.

Delft3D can switch between the 2D vertically averaged and 3D mode simply by changing thenumber of layers. This feature enables to set up and investigate the model behaviour in 2Dmode before going into full 3D simulations.

1.1 Areas of application

Delft3D can be applied, but is not limited, to the following areas of applications:

� flows due to tide, wind, density gradients and wave induced currents;� propagation of directionally spreaded short waves over uneven bathymetries, including

wave-current interaction;� advection and dispersion of effluents;� online morphodynamic computations (local scour, short time and length scales);� sediment transport of cohesive and non-cohesive sediment;� water quality phenomena including ecological modelling, the prediction of heavy metal

concentrations, interaction with organic and inorganic suspended sediment, interactionbetween the water and bottom phase (such as sediment oxygen demand), algae blooms;

� particle tracking, including oil spill and dredging plume modelling;� initial and/or dynamic (time varying) 2D-morphological changes, including the effects of

waves on sediment stirring and bed-load transport.

1.2 Delft3D framework overview

Delft3D is composed of a number of modules (see Figure 1.1), each addressing a specificdomain of interest, such as flow, near-field and far-field water quality, wave generation andpropagation, morphology and sediment transport, together with pre-processing and post-processing modules. All modules are dynamically interfaced to exchange data and resultswhere process formulations require. In the following chapters these modules are described inmore detail.

Flow/Mor Wave Water Quality Ecology Particles/Oil

Overall Menu

Tools

Figure 1.1: System architecture of Delft3D

Deltares 1


All features are embedded in Graphical User Interface suitable for Linux or the MS Windows.An application (model) can be completely defined, inspected and analysed through a menu-driven, user-friendly, graphical interface.

The basic processes covered by each of the modules are:

Delft3D-FLOW and MOR 2D and 3D hydrodynamic, salinity, temperature, transport and onlinesediment transport and morphology

Delft3D-WAVE short wave propagation (using SWAN)D-Water Quality general water qualityDelft3D-SED cohesive and non-cohesive sediment transportDelft3D-ECO complex eutrophication and ecological modellingD-Waq PART particle tracking, oil spill modelling

1.3 Utilities

The following utility programs are available for pre-processing and post-processing:

RGFGRID: for generating orthogonal curvilinear grids, in Cartesian or spherical co-ordinatesQUICKIN: for preparing and manipulating grid oriented data, such as bathymetry, initial condi-

tions for water levels, salinity, constituents and other parametersDelft3D-TRIANA: for performing off-line tidal analysis of time-series generated by Delft3D-

FLOWDelft3D-TIDE: for performing tidal analysis on time series of measured water levels or veloci-

tiesGPP: for visualisation and animation of simulation resultsDelft3D-QUICKPLOT: for visualisation and animation of simulation resultsGISVIEW: ArcGIS extension to export GIS-coverages to Delft3D format and to read, visualise

and process results from Delft3D (ArcGIS is not included)Delft3D-MATLAB: user interface and Matlab functions to read Delft3D files and to visualise or

process results in Matlab environment (Matlab is not included)D-Waq DIDO: interactive grid aggregation editor for coupling FLOW with WAQ models

For more information please contact:

e-mail: [email protected]: www.deltares.nl

2 Deltares

2 Water quality module

2.1 General introduction

The transport of substances in surface and ground water is commonly represented by theso-called advection-diffusion equation. The water quality module, D-Water Quality, is basedon this equation and it offers different computational methods to solve it numerically on anarbitrary irregular shaped grid, on a grid of rectangles, triangles or curvilinear computationalelements. D-Water Quality can be applied just as easily on 0D, 1DV, 1DH, 2DV, 2DH and 3Dschematisations of a water body. D-Water Quality includes the complete natural cycles of C,N, P, Si and O2, as well as cohesive sediments, bacteria, salinity, temperature, heavy metalsand organic micro-pollutants (see Figure 2.1).

Atmosphere

Water

Sediment

Salinity

Chloride

Temperature

Conservative

Tracers

1 2 3 4 5

Coli Bacteria

E.Coli FColi TColi

pH

CO2 Alkalinity

Decayable

Tracers

1 2 3 4 5

Nutrients

NO3 NH4 PO4 Si

Adsorbed PO4

Phytoplankton

C N P Si

Organic Matter (particulate)

C N P Si

Organic Matter (dissolved)

DOC DON DOP

Dissolved Oxygen Oxygen Demand

BOD COD

Inorganic Matter

IM1 IM2 IM3

Heavy Metals

As Cd HgCr Cu

Ni Pb V Zn

Organic micro-

pollutants

Atr BaP Diu Flu

HCB HCH PCB-153

Organic Matter (particulate)

C N P Si

Inorganic Matter

IM1 IM2 IM3

Heavy Metals

Organic micro-

pollutants

Micro-phytobenthos

C N P Si

Grazers

C N P Si

Sediment Oxygen

Demand

Figure 2.1: General overview of substances included in D-Water Quality. Substances areorganised in functional groups indicated by a grey header, except for somesubstances that form a group of their own. Major processes between sub-stances are indicated by arrows; note that many links are omitted

To proceed one step in time (t + ∆t), D-Water Quality solves equation 2.1 (a simplifiedrepresentation of the advection-diffusion-reaction equation) for each computational cell andfor each state variable.

M t+∆ti = M t

i + ∆t×(

∆M

∆t

)

Tr

+ ∆t×(

∆M

∆t

)

P

+ ∆t×(

∆M

∆t

)

S

(2.1)

The mass balance has the following components:

1 the mass at the beginning of a time step: M ti

2 the mass at the end of a time step: M t+∆ti

3 changes by advective and dispersive transport:(

∆M∆t

)Tr

Deltares 3


4 changes by physical, (bio)chemical or biological processes:(

∆M∆t

)P

5 changes by sources (e.g. waste loads, river discharges):(

∆M∆t

)S

The basic principles of D-Water Quality are the same whether you have one state variableand only two computational cells, or you have several tens of state variables and thousandsof computational cells. The only difference is the number of times that D-Water Quality has tosolve equation 2.1. D-Water Quality is capable of describing any combination of constituentsand is not limited with respect to the number and complexity of the water quality processes.

Water quality processes are described by linear or non-linear functions of selected state vari-ables and model parameters. Many process formulations are available in the form of a library,which smoothly interfaces with the water quality module. The library contains over 50 wa-ter quality process routines covering 140 standard substances. A graphical user interfacewithin the WAQ module enables you to select any combination of substances and associatedwater quality processes. Also, for less experienced users pre-defined sets are available tojump-start the water quality modelling.

2.2 Module description

D-Water Quality simulates a physical system that consists of a surface or ground water body.Strictly speaking, it models a body of a medium that is able to transport passive constituents.In this respect, "passive" means that the influence of the concentration of the constituents onthe transport coefficients may be neglected.

The transporting medium is characterised by its spatial and time dependent content (mass) ofthe modelled constituents. Some are transportable; some are non-transportable. An exampleof the latter is the bottom sediment in a surface water model. The concentration of the trans-portable constituents is computed by dividing the mass by the water volume. The mass is thestate variable and the model is mass conserving by definition.

Waste disposals are specified either as mass units per time unit or as a combination of wasteflow and concentration. They represent either point sources (urban, industrial, rivers) or dif-fuse sources (run-off, atmospheric deposition). The case of recirculating flows, as with coolingwater studies, is also taken care of: the water that was let in, will have the same quality at theoutlet.

The hydrodynamic characteristics of the transporting medium are expressed in terms of thevolume and the flux of the transporting medium ("flow"). The combination of water volumesand flows must be consistent, i.e. an increase of the water volume must be balanced by a dif-ference between inflow and outflow. D-Water Quality repeats hydrodynamic characteristics toextend to simulation times beyond the available hydrodynamic simulation time. Also, D-WaterQuality can combine several hydrodynamic simulations into a single water quality simulation.For example, a representative neap tide and a representative spring tide hydrodynamic simu-lation can be combined to create a complete spring-neap cycle.

As part of Delft3D, the coupling module can derive a set of consistent hydrodynamic flowsautomatically from Delft3D-FLOW, but the methods involved can be applied equally well tothird-party hydrodynamic models outside Delft3D.

In many cases, the water quality processes in the model are determined by meteorological

4 Deltares

Water quality module

conditions, by other (modelled or non-modelled) constituents or by other (modelled or non-modelled) processes. Examples are wind, water temperature, acidity (pH), primary productionand the benthic release of nutrients. These entities are referred to as "forcing functions".Water quality process formulations are often of an empirical or semi-empirical nature andcontain "model parameters" which are subject to tuning or calibration. Because of this, D-Water Quality allows complete freedom in selecting the set of water quality processes and therelevant forcing functions and model parameters may vary between individual applications. Ittherefore provides flexible input facilities for constants, spatially varying parameters, functionsof time and functions of space and time.

The physical system is affected by two types of processes:

� transport processes: these processes involve the movement of substances;� water quality processes: these processes involve a transformation of one or more sub-

stances.

The transport of substances in surface and ground water is commonly represented by theso-called advection diffusion equation. Advection is determined by the velocity field and dis-persion by the dispersion coefficient. These basic transport processes operate on all trans-portable substances in the same way. D-Water Quality offers the possibility to model othertransport phenomena as well which may differ between individual substances. Examples arethe gravity induced settling of particles and the autonomous motion of fish. These additionaltransport processes must be expressed as an extra, substance dependant velocity or disper-sion coefficient.

Water quality processes are incorporated in the advection diffusion equation by adding anadditional source in the mass balance (equation 2.1). Examples of water quality processesare:

� exchange of substances with the atmosphere (oxygen, volatile organic substances, tem-perature);

� adsorption and desorption of toxicants and ortho-phosphorous;� deposition of particles and adsorbed substances to the bed;� re-suspension of particles and adsorbed substances from the bed;� mortality of bacteria;� biochemical reactions like the decay of BOD and nitrification;� growth of algae (primary production);� predation (e.g. zooplankton on phytoplankton).

Special attention is paid to the treatment of the interaction with the bottom:

� all suspended sediment is modelled as cohesive sediment that can be transported withthe water flow just like a dissolved substance;

� all particulate inorganic matter can be represented by three size fractions or components;� all particulate organic matter is represented by separate components, namely detritus

carbon, other organic carbon, diatoms, non-diatom algae (Green), adsorbed phosphorusand organic carbon from loads;

� the bottom sediment is modelled via two separate layers. Each layer is considered homo-geneous (well mixed). The different layers can have different compositions. The densityof a layer is variable depending on the sediment layer composition, which is also variable.The porosity within a given layer is constant (user-defined).

Deltares 5


� a third (deeper) layer exists (but is not explicitly modelled) which can supply sediment forupward sediment transport ’digging’;

� sedimentation and resuspension are modelled using the Krone-Partheniades approach(see the description of the sediment transport module Delft3D-SED).

2.3 Application areas

D-Water Quality can be applied to the following areas:

� Environmental Impact Assessment (EIA), objective evaluation of alternatives� Water balance studies, including the identification of the origin of water, residence times

and flushing capabilities� Swimming water quality, bacterial decay processes;� Sewage and/or storm water outfall studies;� Nutrient cycling and eutrophication;� Sedimentation and resuspension of particulates, dredging plumes;� Sediment-water interaction (including diffusive and benthic mixing);� Bio-availability of heavy metals (and organic micro-pollutants;� Recirculation of cooling water from power and desalination plants including the release of

bio-fouling chemicals such as chlorine

The processes always require input in the form of rate constants and/or simulation resultsfrom other substances. The input could come from:

� one of the other modelled substances;� a user-specified spatially distributed time function;� a user-specified time function for the whole area;� a user-specified spatially distributed constant;� a user-specified constant for the whole area;� a process flux originating from one of the water quality processes from the library;� output from one of the other processes in the library;� a default value from the database containing default values.

The pre-processor will report the origin of the input for each process. If information for aprocess is missing, so that the process cannot be evaluated, it will detail what information isactually required in addition.

Results can be presented as spatial patterns or time-series. Statistical processing can becarried out during the simulation for minimum, average and maximum values, percentiles,percentage of exceedance, geometric mean, etc. These values can be calculated for pre-selected periods, so that it is possible to calculate for example winter and summer averages.

As a special type of output, mass balances can be generated. These can be used to analysethe fate of substances (e.g. nutrients or heavy metals) or to quantify the contribution of wasteloads or certain processes to the ambient concentration of a substance.

6 Deltares

Water quality module

2.4 Coupling with other modules

Hydrodynamic conditions can be prescribed by Delft3D-FLOW, SOBEK (Deltares 1D softwareand SIMONA (2D/3D hydrodynamic model). Transfer of results goes via a so-called communi-cation file. This file usually spans a representative period such as a tidal cycle, a spring-neapcycle or even a full year. D-Water Quality is able to calculate beyond the period by repeatingthe hydrodynamic cycle. Also, D-Water Quality is able to make use of multiple hydrodynamicfiles by either combining them simultaneously or successively. For example, a representativelow river discharge hydrodynamic file and a high river discharge hydrodynamic file may becombined to simulate all river discharges in between.

The grid aggregation tool D-Waq DIDO can be used to fit the grid resolution to the specificneeds of the application. D-Waq DIDO allows regular and irregular aggregation (in the hori-zontal) and will reduce the number of computational elements and therefore computation time.D-Waq DIDO is frequently used to reduce the grid resolution in areas far away from the areaof interest, while keeping the area of interest at the highest possible resolution.

QUICKIN can be used to generate space-varying initial conditions or space-varying processparameters (e.g. wind speed).

Deltares 7


8 Deltares

3 Sediment transport module

The sediment transport module, Delft3D-SED, can be applied to model the transport of cohe-sive and non-cohesive sediments, i.e. to study the spreading of dredged materials, to studysedimentation/erosion patterns, to carry out water quality and ecology studies where sedimentis the dominant factor.

It is in fact a sub-module of the water quality module, that is all processes contained in thesediment transport module are also present in the water quality module. For a detailed de-scription of the general aspects we refer to the description of the water quality module.


3.1.1 Cohesive sediment

This section describes the implementation of the physical processes in some detail. For co-hesive sediment transport sedimentation, erosion, burial and digging are taken into account.

For sedimentation the following assumptions apply:

� sedimentation takes place when the bottom shear stress drops below a critical value;� there is no correlation between the sediment components (i.e. each of the particulate

fractions can settle independently);� sedimentation always results in an increase of sediment in the uppermost sediment layer;� the total shear stress is the linear sum of the shear stresses caused by water velocity and

wind effects. Effects of shipping and fisheries can also be included.

The effects of ’hindered settling’ (i.e. decrease in sedimentation velocity at very high sus-pended solids concentration) can be included.

For resuspension the assumptions are:

� the bottom sediments are homogenous within a layer. Therefore, the composition of theresuspending sediment is the same as that of the bottom sediment;

� the resuspension flux is limited based on the available amount of sediment in a sedimentlayer for the variable layer option. The resuspension is unlimited if the fixed layer option isused;

� as long as mass is available in the upper sediment layer, resuspension takes place fromthat layer only;

� resuspension flux is zero if the water depth becomes too small.

Burial is the process in which sediment is transferred downward to an underlying layer. Thesediment layer is assumed to be homogeneous, therefore the composition of the sedimentbeing buried is the same as that of the (overlying) sediment layer.

Digging is the process in which sediment is transferred upward from an underlying layer. Thesediment layers are homogeneous, therefore the composition of the sediment being trans-ported upwards is the same as that of the (underlying) sediment layer. A third and deeperlayer allows for an unlimited ’digging’ flux to the second layer. The quality of this third layermust be defined by you and is not modelled by Delft3D-SED.

Deltares 9


3.1.2 Non-cohesive sediment

For non-cohesive sediment (sand) the transport rate is calculated according to the transportformulae of Engelund-Hansen and Ackers-White. These (semi-)empirical relations describethe total transport (bed load and suspended load) in the situation of local equilibrium.

The implementation recognises two options: unlimited supply of sand via the boundaries andthe presence or absence of bedrock.

3.1.3 Limitations

To apply the sediment transport module the following limitations must be observed:

� in the sedimentation process, there is no correlation between the cohesive and non-cohesive components, i.e. between sand and silt; each is treated independently;

� the effect of short waves must be taken into account through the hydrodynamic moduleor through a localised wave effect estimation (that is, the waves are considered to be inequilibrium with local circumstances);

� Delft3D-SED should only be used for short- or medium-term (days, weeks, months) mod-elling of erosion and sedimentation process as the changes on bottom topography andits effects on the flow are neglected. For long-term processes (years), whereby the flowchanges induced by changing bottom topography is significant, the separate morpholog-ical and sediment module (Delft3D-MOR) should be used. This module has advancedonline coupling capabilities with the hydrodynamic flow and wave modules.


Delft3D-SED can be applied to the following application areas:

� effects of dredging on the environment;� sedimentation and resuspension of sediment in general;� sand transport.


See Section 2.4.

10 Deltares

4 Ecological module

The far-field water quality module D-Water Quality models algae using an approach basedon Monod kinetics and is routinely included in the process library. The Delft3D-ECO modulecontains the more sophisticated algae model BLOOM II (???) that is based on an optimisationtechnique.


Delft3D-ECO distributes the available resources (nutrients and light) in an optimal way amongthe different types of algae. A large number of groups and/or species of algae and even dif-ferent phenotypes within one species can be considered. In the same way, algae living inthe water column (phytoplankton) and algae and water plants living on the sediment (benthicspecies) can be included with their specific ecophysiological characteristics. With BLOOMII, apart from the calculation of biomass concentrations, the dynamics of algae communi-ties including competition for light and nutrients, adaptation to environmental conditions andspecies composition can be simulated. The eutrophication processes in Delft3D-ECO can becombined with the general water quality processes in D-Water Quality.

Delft3D-ECO can be used to calculate eutrophication phenomena, including:

� the competition between several groups of algae species;� adaptation of algae to changes in the environment, in terms of stoichiometry and growth

characteristics. (This can be of particular importance if the simulation of possible devel-opment of nuisance algae is an aim of the modelling.)

� steep gradients in algae biomass due to temporal or spatial variations;� phytoplankton blooms;� chlorophyll concentrations;� species composition;� limiting factors for algae growth;� oxygen kinetics, including daily cycles;� nutrient concentrations.

Algae blooms usually consist of various species of phytoplankton belonging to different taxo-nomic or functional groups such as diatoms, microflagellates and dinoflagellates. They havedifferent requirements for resources (nutrients; light) and they have different ecological prop-erties. Some species are considered to be objectionable for various reasons. Among theseare phaeocystis, which causes foam on the beaches and various species of dinoflagellates,which among others may cause diurethic shell fish poisoning. To deal with these phenomenait is necessary to distinguish different types of phytoplankton in the algae model.

Delft3D-ECO is based upon the principle of competition between different species, or groupsof species. The basic variables of this module are called types. A type represents the physi-ological state of a species under strong conditions of limitation. Usually a distinction is madebetween three different types: an N-type for nitrogen limitation, a P-type for phosphorus limi-tation and an E-type for light energy limitation. Usually for each (group of) species the threedifferent types are modelled.

The solution algorithm of the model considers all potentially limiting factors and first selectsthe one, which is most likely to become limiting. It then selects the best adapted type for

Deltares 11


the prevailing conditions. The suitability of a type (its fitness) is determined by the ratio ofits requirement and its growth rate. This means that a type can become dominant either be-cause it needs a comparatively small amount of a limiting resource (it is efficient) or becauseit grows rapidly (it is opportunistic). Then the algorithm considers the next potentially limitingfactor and again selects the best adapted phytoplankton type. This procedure is repeateduntil it is impossible to select a new pair of a type and limiting factor without violating (i.e.over-exhausting) some limiting factor. Thus the model seeks the optimum solution consistingof n types and n limiting factors. The optimal distribution of biomass over the types cannotalways be reached within one time step due to growth and mortality limitations. Special timedependency constraints are imposed upon all types to take their potential growth and mortal-ity rates into account. As they represent different stages of the same species, the transitionof one type to another is a rapid process with a characteristic time step in the order of a day.Transitions between different species is a much slower process as it depends on mortalityand net growth rates. It is interesting that the principle just described, by which each phyto-plankton type maximises its own benefit, effectively means that the total net production of thephytoplankton community is maximised.

4.2 Applications

The BLOOM model has been extensively used to model the Southern North Sea and hasbeen validated for 25 years of data in the Dutch coastal zone. Furthermore the model resultshave been validated for a wide range of both freshwater and marine systems. The following(groups of) algae or macrophyte species have been modelled using the BLOOM module forsalt waters:

� Diatoms� Flagellates� Dinoflagellates� Phaeocystis� Ulva (on the bottom)� Ulva (floating)

For Ulva two life forms are distinguished: Ulva that is rooted in the sediment and Ulva thatfloats on the water after it has been cut loose by strong winds or currents. The process ofcutting loose has been incorporated in the model. Up to now 6 types of algae or macrofyteshave been modelled and calibrated. As in BLOOM the properties of the algae are adjusted tothe light climate and nutrient availability, you do not need to adjust the parameters for this bycalibration. The default parameter values obtained by calibration of one model can thereforebe applied in a wide range of other model applications. For this reason, if one needs tomodel an area that resembles a water system that has been modelled with BLOOM II before,choosing the model that has proven successful under those conditions can be particularlyhelpful.


See Section 2.4.

12 Deltares

5 Particle tracking module

The particle tracking module, D-Waq PART, is a 3-dimensional mid-field water quality model. Itestimates a dynamic concentration distribution by following the tracks of thousands of particlesin time. The model is fit for a detailed description of concentration contours of instantaneousor continuous releases of salt, oil, or other conservative or simple decaying substances. Thissection gives a brief introduction to the computer module and its applications.


D-Waq PART simulates transport processes and simple chemical reactions of substances.The module allows the simulation of detailed shapes of patches of wasted material.

D-Waq PART can operate in a 2.5DH or 3D mode, in which D-Waq PART is coupled toa 2-dimensional (one layer) or 3-dimensional (multi-layer) Delft3D-FLOW model. In the 2-dimensional mode, the flow is extended with an analytical vertical velocity profile for bottomshear and wind. The module calculates concentrations with a resolution that is higher thanthat of the underlying hydrodynamic grid.

In D-Waq PART, two modules are available:

� Tracer module: simulation of conservative or first order decaying substances;� Oil spill module: simulation of oil spills with floating and dispersed oil fractions (special

license required).

The physical components in the system are:

� the water system: a lake, estuary, harbour or river, possibly with open boundaries to otherwater systems. Tidal variations are included;

� discharges due to human activities that may be instantaneous and/or continuous;� chemical substances like rhodamine dyes, salt, oil or suspended solids;� wind fields;� settling and erosion of suspended matter;� concentration dependent settling velocity.

In terms of physical processes or phenomena D-Waq PART can represent:

� the dynamics of patches close to an outfall location;� simple first-order decay processes like the decay of several fractions of oil;� vertical dispersion for well-mixed systems;� horizontal dispersion due to turbulence. According to turbulence theory this dispersion

increases in time.� the effects of time-varying wind fields on the patches;� the effects of bottom-friction on the patches;� the existence of a plume at the outfall (rather than a point-source) by starting the simulation

from a circular plume with an estimated or field-measured radius.� settling of particles, where a concentration dependent settling, subject to a minimum and

maximum settling velocity, can be specified;� transport and fate of spilled oil. Processes that are included in the oil module are: advec-

tion of floating oil by wind and currents, dispersion (entrainment in water) of oil induced by

Deltares 13


wind waves (depending on wind speed and oil characteristics), evaporation of floating oil,emulsification, decay, sticking of oil to the coastline or seabed. Changes of oil properties(density, viscosity, water content) due to these processes are included in the oil module.Additional advection due to wind drag is implemented in 3D (for 2D the logarithmic profileof the current speed results in a similar wind driven transport), whilst in 3D a deflectionangle is included to represent the Coriolis effect of wind induced advection due to waves.

.

Figure 5.1: Example of a D-Waq PART oil spill simulation

� the model may be started from a known initial distribution of oil, e.g. a remote sensingimage of an oil spill.

D-Waq PART can in theory simulate an unlimited number of particles and substances. Theonly restriction is the available memory of the hardware. For an application with approximately400,000 particles and 8 substances, about 64 Mbytes internal (hard core) computer memoryis required. Under these conditions, a computer simulation requires for most applications lessthan one hour, and takes most often less than 200 Mbytes of disk space. The requirementscan increase significantly when the numbers of particles and substances increase. Post-processing is done with the general post-processing program GPP or QUICKPLOT. Graphicalmaps can also be generated with advanced methods like point spread functions. Visualisationis off-line. The coupling between the hydrodynamic module, Delft3D-FLOW, and D-Waq PARTis streamlined, but is off-line.


D-Waq PART can be applied to cases in which the effects of the discharge from a limitednumber of discharge points (sources) are studied and that focus on the behaviour of theeffluent plume within the mid-field range (order of 1-10 kilometers). Simulation periods aregenerally in the order of days-weeks.

Note that in all applications the results of a dynamic two- or three-dimensional flow calculation(including an accurate description of tidal variations), such as from Delft3D-FLOW, have beencoupled with the PART module.

Examples of applications are:

� dilution of conservative tracers (e.g. from effluent discharges);� dispersion of first-order decaying substances (such as BOD, coliforms);� dispersion of suspended solids resulting from dredging or dumping operations, including

14 Deltares

Particle tracking module

the effects of settling and resuspension;� effects of continuous or instantaneous spills;� oil spill simulations

Deltares 15


16 Deltares

6 Pre-processing and post-processing

In this chapter several pre-processing and post-processing programs available in Delft3D aredescribed in some details. These programs concern visualisation, grid generation, manipula-tion of grid related data and data analysis and manipulation.

6.1 Visualisation

6.1.1 GPP

The general post-processor (GPP) module of Delft3D allows uniform access to all kinds ofdata files to select and visualise simulation results and measurement data. More specificallythe program allows to:

� select the map and/or time histories you want to visualise;� select the lay-out and composition of the plot figure to be produced;� select the type of output medium, i.e. screen for inspection, plotter or printer for hard copy

output.

The type of presentation depends on the character of the data set:

� vector plots for flow velocities, bottom shear stress and other vector quantities, with auto-matic or user-defined scaling of s-axis, y-axis and vector scale;

� time history plots, from a single run, from various runs in the same plot or simulationresults in combination with measurement data. Depending on the data files, these can betypical hydrodynamic quantities, such as water levels, velocity magnitude and direction,but also water quality parameters like salinity, temperature and E.coli concentration. Thescaling can be determined automatically or set by you;

� contour and isoline plots of scalar quantities like the depth, water levels or algae growthrates. Again you can choose automatic scaling or set the contour classes manually;

� vertical profiles for quantities defined on a three-dimensional grid;� geometric plots of the grid itself, tidal flats, land boundaries;� mass balances and limiting factors for displaying the details of water quality models.

Data sets can be plotted in any (sensible) combination, as long as there is a common co-ordinate system. Layouts may contain more than one viewport, allowing several independentplots on one page. It is noted that the overview above is by no means complete but it gives ageneral idea about the possibilities.

The program has been designed to be general enough to handle different kinds of underlyinggeometries and data files of widely varying formats.

The program is capable of producing high quality colour plots. It is also able to produce a plotfile in various standard formats. At the same time a print-out of the results in ASCII format canbe made, enabling the data to be imported in other post-processing programs.

For the use of ArcView and Matlab to visualise and further process Delft3D results, see Sec-tion 6.9.

Deltares 17


6.1.2 QUICKPLOT

The post-processing program Delft3D-QUICKPLOT allows you to easily plot and animate datafrom most output files and some input files of Delft3D and several other software packages ofWL |Delft Hydraulics (such as SOBEK and PHAROS). Furthermore, it supports some simpleASCII formats such that you can combine model output and measurement data in one plot,and it is possible to load bitmap data as a backdrop for your 1D or 2D plots.

Typical plots created using Delft3D-QUICKPLOT are 2DH or 2DV plots and time-series plots,although it also has basic support for 3D plots. Scalar results may be presented using con-tour lines, contour patches, grid cell based patches, interpolated continuous shades, colouredmarker or value fields. Vector results may be presented as vectors, coloured vectors or nor-malised vectors or as scalar quantities by selecting a single component (e.g. x-component,y-component, magnitude, direction) of the vector.

Data sets can be linked to animate single or multiple data sets in a figure. Animation framescan be stored in various bitmap formats. Data sets can be exported to various in-house or 3rdparty formats.

Delft3D-QUICKPLOT is a standalone program based on technology of The MathWorks Inc.It can be seamlessly integrated with the MATLAB environment via the Delft3D-MATLAB inter-face.

.

Figure 6.1: Example QUICKPLOT figure: 3D view of bed level

18 Deltares

Pre-processing and post-processing

.

Figure 6.2: Example QUICKPLOT figure: Depth-averaged velocity vectors and dryingand flooding

6.2 Grid generation

RGFGRID is a program to generate orthogonal, curvilinear grids of variable grid size, that areto be used in combination with each of the modules of the Delft3D suite. The grid-generatorincludes a graphical interface and an orthogonalisation module, providing easy control of thegrid generation process.

RGFGRID supports the following features:

� graphical user interface;� generation of grids in Cartesian or Spherical co-ordinate systems� display of grid features as orthogonality, smoothness, aspect ration etc.;� several user-functions have been implemented to provide easy control over the grid shape;� keyboard and mouse driven events are supported;� iterative way of working, each cycle providing more definition in the grid shape.� generation of multi-domain interfaces.

Deltares 19


6.3 Grid data manipulation

To create, visualise and modify grid based data, such as bathymetries, and other grid relateddata the program QUICKIN is provided. QUICKIN is used in combination with the modules ofDelft3D.

QUICKIN supports the following features:

� graphical user interface;� several interpolation options (averaging, triangulation, diffusion);� suitable for different ratios of grid-density versus sample-density;� various display possibilities: isolines, dots, perspective, etc.;� implementation of various user-functions to provide easy control over the final bathymetry;� sample data from different sources can be interpolated in sequence, thus, starting with

the best quality data available, an optimal bathymetry can be created.� Definition of dredge and dump sites with their characterristics.

6.4 Grid aggregation

The program DIDO enables you to span coarser, irregularly shaped, grid segments for waterquality modelling, starting from the fine grid of e.g. the grid used by the hydrodynamic model.For ecological modelling with large numbers of state variables, a coarser schematisation, fol-lowing ecological and transport separation lines rather than grid lines, is often preferable. Thefine grid of the hydrodynamic model serves as input, integer multiples of the input grid areused for the description of the coarse grid. The procedure is fully mass-conserving. Aggrega-tion is only supported in a plane surface.

DIDO provides the following features:

� zoom in locally;� separate a working area from the remainder of the schematisation;� aggregate regularly (e.g. every 2 segments in the one and 3 in the other direction);� aggregate irregularly (by rubber band lines comparable to the bulls hide);� fine tune by point and click on single elements;� select a subset of the hydrodynamic area for water quality modelling;� display information of a selected segment;� save intermediate results on the fly;� resume unfinished work from saved files;� save the final result for water quality simulation.

The final result of DIDO will be used as input to the coupling program between the hydro-dynamic module Delft3D-FLOW and the water quality module D-Water Quality enabling thelatter to run on a coarser grid using the fine grid hydrodynamic database. Water quality simu-lations are converted back to the fine grid in post-processing software. This gives spatial plotswith the fine resolution (although aggregated areas will still show equal concentration values).

20 Deltares


6.5 Tidal analysis and comparison with observations

Analysis and interpretation of a hydrodynamic simulation in terms of tidal amplitudes andphases can be performed by the program Delft3D-TRIANA. Delft3D-TRIANA performs of-fline tidal analyses on time-series of either water levels and/or velocities. Moreover, Delft3D-TRIANA compares the results from these analyses with observation data supplied by you.Amplitude ratios and phase differences as well as objective statistics are determined.

6.6 Tidal analysis and prediction

The program Delft3D-TIDE is used for the analysis of tidal recordings and the preparation oftidal predictions.

The main module TIDE/ANALYSIS performs tidal analysis on time-series of water levels orcurrents. A variety of features is included, such as:

� the coupling of closely positioned astronomical components;� the simultaneous analysis of successive records of different instruments;� the discrimination of sub-series to account for gaps in measurement recordings;� the appreciation of linear trends and an accuracy analysis.

In a tidal analysis of a time-series of one year with a 10 minutes interval, 100 or more tidalconstituents can be prescribed simultaneously. The constituents are selected from the internaldatabase that contains 234 constituents that may be important at locations world-wide.

The module TIDE/FOURIER performs Fourier-analyses on any type of time-series. This fea-ture can be used to investigate the series of residual levels or velocities which has beenidentified during the tidal analysis on remaining tidal components.

Using a set of tidal constants, such as computed in the analysis module, the TIDE/PREDICTmodule predicts water levels or tidal currents as a function of time.

The module TIDE/HILOW may provide the production of tide tables with the dates, times andheights of the High and Low Waters. Using a word-processor or desktop publishing softwarepackage, the basic tide tables can be processed further and combined with other relevantinformation like tidal stream data.

Whereas in the regular analysis part of the package you pre-define the constituents that will beconsidered, the program also features an option (TIDE/ASCON) to compute the astronomicarguments and node amplitude factors for all 234 internally defined constituents.

The package is accompanied with a comprehensive User Manual, exemplifying the use of theprogram and its scientific backgrounds. A number of examples is added in the form of inputand data files.

Deltares 21


6.7 Nesting of Delft3D-FLOW models

At the open boundaries of a Delft3D-FLOW model, boundary conditions are required for thevertical and/or horizontal tide and the substances if applicable. In case these open boundariesare located within a (coarser) overall Delft3D-FLOW model, then the overall model can beused to generate the boundary conditions for the detailed model. In this case we say thedetailed model is nested within the overall model.

The procedure to generate nested boundary conditions consists of 3 steps:

1 Using the program Delft3D-NESTHD 1 a list of monitoring stations in the overall model,needed for the interpolation, will be generated. In addition to this, the program generatesthe nest administration, i.e. the link between the boundary support points in the detailed(or nested) model and the monitoring stations in the overall model.

2 Run the overall model with the list of monitoring stations generated by Delft3D-NESTHD1.

3 The actual boundary conditions for the nested model are generated by Delft3D-NESTHD2 using the history file of the overall model and the nest administration.

6.8 Nesting of D-Water Quality models

The transfer of data from an encompassing or ’overall’ numerical model to an embeddedor ’nested’ numerical model is called nesting. In general the overall model has a coarseresolution of grid cells, whereas the nested model has a higher resolution. At the boundarylocations of the nested model the results from the overall model are required as boundaryconditions for the nested model. The boundary conditions can be water levels, currents,fluxes or discharges in case of hydrodynamic models, and water quality parameters in caseof water quality models.

The procedure of nesting through concentrations between D-Water Quality (or D-Waq PART)models is performed by the system D-Waq NESTWQ. In this procedure two steps can bedistinguished which are handled by separate subsystems:

1 D-Waq NESTWQ 1, for the determination of nest segments and nest weights in the overallmodel. The concentrations at these segments are used by the next subsystem.

2 D-Waq NESTWQ 2, for the generation of boundary conditions for the boundary segmentsin the nested model from the results at the nest segments in the overall model.

6.9 Interfaces with other programs

It should be emphasised that even though these extensions can be quite useful as a sup-plement to the Delft3D tools, real benefits are gained mostly if you are familiar with both theDelft3D environment and the external environment.

22 Deltares


6.9.1 Interface to GIS

While pre-processing and post-processing can be done quite adequately using the specifictools offered by Delft3D, recently a link have been established with ArcView. The link is in-tended as a supplement to the existing tools rather than a replacement. It adds the ability toview and manipulate model results and model input in a different environment.

The link to ArcView implies:

� exporting a GIS line coverage as land boundary outline and depth data as contained inArcInfo/ArcView map layers to a format suitable for RGFGRID and QUICKIN;

� importing the model grid and the corresponding depth field as generated by RGFGRIDand QUICKIN, so that they can be presented in a geographical context;

� importing the grid-based model results (scalar and vector quantities) with a user interfacequite similar to that of GPP in the ArcView environment for presentation or further analysis

All data files are read directly by this ArcView extension and stored as shape files. There isno need to convert or process the model result files.

6.9.2 Interface to Matlab

In a similar way as with GIS it is possible to import the results produced with Delft3D directlyinto Matlab. This gives the opportunity to visualise or use the results for further analysis usingthe facilities offered by Matlab.

The Delft3D-MATLAB interface allows you to seamlessly integrate the simplicity of simulationdata access by Delft3D-QUICKPLOT with the flexibility of the MATLAB environment developedby The MathWorks Inc. The combination of these two tools allows you to use the full power ofMATLAB for analysing, processing and visualising the simulation results.

Deltares 23


.

Figure 6.3: Example QUICKPLOT figure: Depth-averaged velocity vectors and tidal ellips

24 Deltares

7 Hardware configuration

Delft3D and its accompanying programs is supported on the following platforms:

� Windows 32-bit platforms� Linux Redhat 3.4

Configuration item Minimal Preferred

Processor IA32, 1 GHz IA32, 3 GHz or more

Internal memory 1024 MB 2 GB or more

Swap space 2.0 × internal memory 4.0 × internal memory

Hard disk 10 GB 80 GB or more

Monitor 17 inch colour 19 inch colour

Display 800 × 600 pixels256 colours

1280 × 1024 pixels16 million colours

Remark:� Delft3D will run on a Windows 64-bit platform but it will not benefit the 64-bit architecture,

because Delft3D is not yet compiled for a 64-bit platform.

Deltares 25


26 Deltares

Contents

Bract, odd page

Deltares 36

Contents

Bract, even page

Deltares 37

WAQ_Topics_02

Introductory CourseDelft3D-WAQ

Overview of topics

WAQ_Topics_02 2

Overview of topics (1)

functional specifications Delft3D-WAQwater quality modelling, introduction and conceptssteps in water quality modellingtransport modelling and numerical aspectsProcess Library Configuration Tooloxygen - BOD, processes and formulation

WAQ_Topics_02 3

Overview of topics (2)

Suspended sediment, sedimentation and erosionNutrients, cycles of nitrogen, phosphorus and siliconGeneral introduction on algae growthStatistical outputToxic substances (heavy metals and organic micropollutants)Principles of OpenPLCTexercises

WAQ_Introduction_03


Water quality modelling,introduction and concepts

WAQ_Introduction_03 2

TOOLS and add-ons

FLOW WAVE WAQ PART ECO CHEM SED MOR

Delft3D Overall Menu

Delft3D system overview

WAQ ECO SED


Set up of training

General idea of Delft3D-WAQUser Interface of Delft3D-WAQTransport modellingWater Quality Processes


Issues in Water Quality Modelling (1)

Quantitative aspects water management• drinking water, flooding (dikes, dams, etc.)

Salinity (affects ecosystem)Bacterial pollutionOrganic material (BOD, COD)Nutrient enrichment (eutrophication)Algae bloomsOxygen depletion


Issues in Water Quality modelling (2)

Aesthetic criteria (taste, colour, smell)Sediment plumesToxic substances

• HM and OMP (toxic, persistent, bio-accumulation)Thermal pollutionRadioactive pollution


Issues in Water Quality modelling (3)

Concentration of substances (compared to water quality objectives)Distribution of pollutants (how does a discharge of pollutants

influence water quality)Effect of changes (Environmental Impact Assessment)

WAQ_Introduction_03

Concept of Delft3D-WAQ

How do we model water quality?


Conceptual background

Administrate the mass balance of a substance in a segmentComponents of the mass balance

• changes by transport• changes by processes (physical / chemical)• changes by sources / discharges


Conceptual background

To proceed one step in time, solve for each segment:

• Tr: change due to transport• P: change due to processes• S: change due to sources


Transport

Derived from hydrodynamic model (e.g. Delft3D-FLOW or SOBEK)both advective and dispersive transport


Processes

Physical: reaeration, settling, resuspensionChemical: denitrification, decay of organic matterBiological: algae growthProcesses can

•remove a substance from or add a substance to the system(denitrification)

•convert a substance (nitrification NH4+ ® NO3

-, settling IM1 ®IM1S1)


Sources / discharges

Waste loadsRiver loadsBoundary loadsDiffusive loads (atmospheric deposition)


Mass and concentrations in a segment

Process 1

Concentration ofsubstance 1

Extra Process

Process 2 Concentration ofsubstance 2

Advective transport (FLOW)Advective transport (FLOW)

Dispersive transport (FLOW)Dispersive transport (FLOW)

Output

Input Input

Externalsource

Segment

Output

Input

Contents

Deltares 47

Contents

Bract, odd page

Deltares 48

Contents

Bract, even page

Deltares 49

WAQ_CoupUI_04


Steps in Water Quality Modelling

WAQ_CoupUI_04 2

First of all ...

Understand the problem(and make sure your understanding corresponds with that of theclient)

Understand the system you have to deal withStart with data analysis (if possible)Define the question you want to answer with the water quality model

WAQ_CoupUI_04 3

Steps in Delft3D water quality modelling

Steps in Delft3D water quality modelling

WAQ_CoupUI_04 4

1) Pick up the hydrodynamic result

2) Choose substances & processes

3) Set-up the WAQ simulation

4) Run the simulation

5) Check the results

6) Visualise the results

WAQ_CoupUI_04 5

Delft3D-WAQ Framework

WAQ_CoupUI_04 6

Delft3D-WAQ Framework with direct coupling

Hydrodynamics(Delft3D-FLOW)

Aggregation(DIDO)

Water Quality input

(WAQ-GUI)

Simulation(DELWAQ)

Visualise results(QUICKPLOT,

GPP)

Inspect results(editor)

Processes Tool(PLCT)

*.dwq

*.hyd

*.sub

*.0*.inp

map/history-files *.lsp, *.lst, *.mon

tri-diag.*

*.scn

WAQ_CoupUI_04 7


WAQ_CoupUI_04 8


WAQ_CoupUI_04

Step 1

WAQ_CoupUI_04

Hydrodynamic TransportHow do we include the Delft3D-FLOW results in the WAQsimulation?

WAQ_CoupUI_04 11


WAQ_CoupUI_04 12

Coupling between hydrodynamics and WAQ

Result from hydrodynamic model is used® communication file

The communication file normally covers a representative period suchas a tide, a spring-neap tide or an average river flow.

The communication file can be aggregated in both space and time.

WAQ_CoupUI_04 13

Coupling with the hydrodynamics (1)

Possibilities:

different times scales• longer time steps• tidally average modelling on residual flow fields• selected period

different spatial scales horizontally through DIDOdifferent spatial scales vertically

WAQ_CoupUI_04 14

Types of spatial aggregation

surface layer 1 1

2

3 2

4

5 3

6

7 4

8

9 5

bottom layer 10

Vertical aggregation Horizontal aggregation

1 x 1 2 x 2

WAQ_CoupUI_04 15


WAQ needs finite grid volumes, exchange areas and flows,dispersion lengths to solve advection diffusion equation

coupling derives information from hydrodynamic database (com*file) and converts water level and velocities to WAQ requiredunits

Coupling allows for:• hydr: 2DH water qual: 2DH• hydr: 3D water qual: 2DH• hydr: 3D water qual: 3D

WAQ_CoupUI_04 16

Mass balance

The selected period of the FLOW model can be rewound. So youcould for example simulate a whole year of water quality repeatingone representative tide or spring-neap cycle.

The water levels / volumes should be similar at the beginning and theend of this selected period.

You should always check the mass balance of your model, bychecking these water levels and modelling the substance‘Continuity’.

WAQ_CoupUI_04 17


Time scales:write hydrodynamic files for water quality with time step of n * dt

of the hydraulic run.• saves space• some info is lost, decide yourself• this new time step is in the UI• the selected period of the FLOW model can be rewound-> beware

of mass balance errors.run water quality with this new time step divided by m

• flows constant for the sub-steps• dt based on stability criteria

WAQ_CoupUI_04 18


3 time steps:Dthydro

Dtcoupling = n . Dthydro n ³ 1Dtwaq = Dtcoupling /m m ³ 1

errors defined:• mass balance error (e):

• volume error:

D DV Q ti i ji j

- ×- >- >å

D V Vi i/

WAQ_CoupUI_04 19

What is DIDO?

Interactive grid editor that allows you to spatially aggregate thehorizontal hydrodynamic grid

Hydrodynamic models have different requirements than waterquality (spatial gradients are “lazy”)

Reasons for aggregation:• decrease computation time• decrease computational storage (renumbering of grid cells)• Zoom in on a particular region

WAQ_CoupUI_04 20

DIDO Options

Regular aggregation:• Two by two cells together• Three by four etc.

Irregular aggregation:• Retain as much detail as possible• Aggregate the small grids only

Whole grid or part of it onlyCreates (*.dwq) aggregation file to be incorporated in Delft3D-WAQ

Contents

Bract, odd page

Deltares 60

Contents

Bract, even page

Deltares 61

WAQ_Num_Asp_04


Transport modelling andNumerical aspects

WAQ_Num_Asp_04 2

Transport modelling

Several aspects are of importance for adequate transportmodelling:

horizontal gridvertical layeringtime stepdispersionnumerical scheme

WAQ_Num_Asp_04 3

Grid size and time step (1)

Considerations on horizontal grid size:

computation time ® more segments means a longer computationtime

measure of detail ® clustering of segments means loss of detailresidence time ® larger segments have longer residence times and

the larger the time step can be (important for some numericalschemes)

accuracy ® loss of accuracy at too large time steps

WAQ_Num_Asp_04 4

Stability criterion

For explicit schemes:

“The volume of water replaced within any grid cell within one timestep should always be smaller than the volume of the grid cell”

(Courant-Friedrichs-Lewy condition)

Note: Also for dispersive transport and processes

WAQ_Num_Asp_04 5


If the water package from segment 1 has passed segment 2, theconcentration in 2 can not be calculated from concentrations in 1and 3.

t = t

t = t + dt

u m/s

1 2 3 4

1 3 42

waterpackage

waterpackage

WAQ_Num_Asp_04 6


If the time step is too large, negative concentrations can result:

dt

conc.

time

WAQ_Num_Asp_04 7

Vertical layering

In case of stratification the pycnocline should ideally be matched bythe interface between two layers.

Within one layer concentrations are averaged, so this meansnumerical vertical mixing.

The numerical schemes in Delft3D-WAQ differ in their degree ofnumerical dispersion.

WAQ_Num_Asp_04 8

Vertical dispersion

Vertical diffusion can be used from the hydrodynamic model.Minimum diffusion values can be imposed.Additional diffusion can be imposed to avoid oscillations.

WAQ_Num_Asp_04 9

Horizontal dispersion

Horizontal dispersion generally about 1 m2/s in 3D;In 2D, horizontal dispersion should reflect spreading by velocity

differences over depth.

WAQ_Num_Asp_04 10

Numerical schemes

Current number of numerical schemes in Delft3D-WAQ: 14Important features:

• accuracy• stability• positivity• efficiency

The suitability of a scheme for a problem depends on the problem.

WAQ_Num_Asp_04 11

Transport (1)

Advective and dispersive transportbasic equation:

(So the mass change is the sum of the mass transports over allsegment borders)

WAQ_Num_Asp_04 12

Transport (2)

Numerical implementationÞ discretisation in time and space

• Numerical schemes differ in q and ~C

WAQ_Num_Asp_04 13

Time discretisation

q ® time discretisation:• q = 0 explicit scheme• q = 1.0 fully implicit scheme• q = 0.5 semi-implicit scheme

Criteria• q ³ 0.5 Þ unconditionally stable (robust)• q < 0.5 Þ can be unstable if stability criterion is not satisfied• q = 0.0 or q = 1.0 required for positivity• q = 0.5 Þ chance of oscillations and negative concentrations

WAQ_Num_Asp_04 14

Space aggregation

~

~C C for Q

C C for Qi j i i j

i j j i j

® ®

® ®

= >

= <

0

0

~C• ® space aggregation• Upwind

• Central

• Higher order Upwind

~CC C

i ji j

® =+

2

WAQ_Num_Asp_04 15

Numerical schemes in Delft3D-WAQ (1)

Schemes 1 to 5:

Explicit (stability criterion!)In practice only applied in 2D simulations as the vertical direction

usually results in a very small time step (notably the influence ofdispersion)

scheme 1 is most robustscheme 5 is most accurate (2nd order scheme)schemes 2 to 4 should only be used in special cases (refer to

manual)

WAQ_Num_Asp_04 16


Scheme 10:

As scheme 1 but implicitUses a direct inversion of the matrix to find the solution at t+DtCan be applied in 3D simulations, but direct inversion may be very

time consuming

WAQ_Num_Asp_04 17


Schemes 11 to 14:

Separate solution of horizontal and vertical transportHorizontal transport explicit either according to scheme 1 or scheme 5

(stability criterion!)Vertical transport implicitCentral or Upwind discretisation in the verticalUpwind schemes may cause numerical mixing which is undesirable

when a significant vertical gradient existsScheme 12 (in horizontal as scheme 5 (thus most accurate), in

vertical central) is preferred

WAQ_Num_Asp_04 18


Schemes 15 and 16:

Implicit (both in the vertical and in the horizontal)Iterative solvers of the matrix (very time efficient)Upwind (scheme 15) and Central (scheme 16) in the verticalNumerical mixing of steep gradients occurs

WAQ_Num_Asp_04 19


Schemes 19 and 20:

Schemes are also used in Delft3D-FLOWADI (Alternating Direction Implicit) usedAggregation of the hydrodynamic database is not allowedVery accurate, but iterative solver may require small time step to

comply with accuracy criterion

Scheme 21 and 22:

Dynamically switch between explicit and implicit avoiding instabilitySophisticated variation on flux-corrected transportScheme 21 uses the flux limiter of Salezac, while scheme 22 uses

the flux limiter of Boris and BookAccurate and relatively fast

WAQ_Num_Asp_04 20

Considerations for choice of numerical scheme (1)

Accuracy:• decreases with time step (due to numerical dispersion)• upwind schemes have numerical dispersion• central schemes may produce negative concentrations• implicit schemes smooth gradients

Stability:• explicit schemes require smaller time steps

WAQ_Num_Asp_04 21

Considerations for choice of numerical scheme (2)

Positivity:• some schemes may produce negative concentrations

Efficiency:• smaller time step ® longer computation time• aggregation not always possible• convergence of solver

WAQ_Num_Asp_04 22

Rule of thumb choice of numerical scheme

1. Scheme 19/20 most accurate, but computation time may belong

2. Scheme 12 (3D) or scheme 5 (2D)accurate, but must fulfil stability criterion

3 . Scheme 21/22 fast but accuracy dependent on time step, nostability criterion

4a. Scheme 1 less accurate than scheme 5, but morerobust (larger time step)

4b. Scheme 16 Very fast scheme, but numerical dispersionmay be large

4c. Scheme 10 only 2D, less accurate than scheme 5, but nostability criterion

WAQ_Num_Asp_04 23

Salinity Pearl Riverbottom layer

Scheme 16:very efficient: ± 300 days in 19 hours

computationless accurate: salinity wedge not

reproduced

Scheme 20:accurate: steep gradient and salinity

wedgevery inefficient: 25 days in 5 days

computation time

WAQ_Num_Asp_04 24

Numerical approach to Processes and Sources

Both are expressed as fluxes (g/m3/d).The calculation of process fluxes is based onconcentration at time t (not t+Dt), so the numericalschemes do not affect the process fluxes, just thetransport fluxes.The change in concentration by processes and sources iscalculated by:

WAQ_Num_Asp_04 25

Transport modelling as a start (1)

For understanding of the system it is useful to look at:distribution of tracer substancesresidence timesdepthsalinity

Contents

Deltares 75

Contents

Bract, odd page

Deltares 76

Contents

Bract, even page

Deltares 77

WAQ_PLCT_03


Selecting Water Quality Processeswith theProcesses Library Configuration Tool (PLCT)

WAQ_PLCT_03 2


WAQ_PLCT_03 3

Water quality processes

physical and (bio)chemical processes:sedimentation/resuspensionmineralisationnitrification/denitrificationreaeration (air-water exchange)etc.

WAQ_PLCT_03

WAQ_PLCT_03

WAQ_PLCT_03 6

Processes Library Configuration Tool

Graphical interface allows you to select substances and associatedprocesses

Multi-level interactive editor• substance groups

> substances– associated processes

• process parameters• extra processes

by selecting substances they are included in the model

WAQ_PLCT_03 7

Steps in PLCT

4 steps to operate the Processes LibraryConfiguration Tool (PLCT):

Step 1 ® Selection of substancesStep 2 ® Selection of processes

Step 3 ® Selection of process parametersStep 4 ® Specification of extra processes

Every process is described in the Delft3D-WAQ manual.

A good knowledge of the processes is essentialfor setting up a good model.

WAQ_PLCT_03 8

Substance groups

WAQ_PLCT_03 9

Select substances

Select processes

WAQ_PLCT_03 10

Specify process

WAQ_PLCT_03 11

WAQ_PLCT_03 12

Notes on PLCT (1)

Substances are subject to fluxes of material between substancesFluxes are computed by process modules which need specific inputCoefficients may have defaultsPLCT can determine if a process can be switched onThe user may specify constants, functions, parameters or segment

functions. The default values are assigned automatically.

WAQ_PLCT_03 13

Notes on PLCT (2)

Besides concentrations, also all fluxes and many derived variables areavailable for post-processing

Per substance, all fluxes form a closed mass balance, available foranalysis

In the WAQ-manual the formulations for every process are described.Also the meaning of all the input parameters is explained here.

WAQ_PLCT_03 14

Available Processes

Transport of conservative and decaying substancesBacteriological pollution: E.ColiSedimentation and erosionOxygen depletion and biological decayEutrofication & algae growthTemperatureHeavy metals and organic micropollutants

In the water column and two sediment layers.

WAQ_PLCT_03 15

Finally ...

Substance files are available for standard applications (sedimentation,oxygen)

When setting up a model, work incrementally• start with salinity or a tracer• add substances along the way• a proven sequence is: salinity ® temperature ® suspended sediment® total nutrients ® oxygen ® algae

It should not be your goal to make the model as complicated aspossible. Simple is beautiful !

Inspect the diagnostics output (the *.lsp file)Always understand the result of your simulation

Contents

Bract, odd page

Deltares 86

Contents

Bract, even page

Deltares 87

WAQ_Oxygen_03


Oxygen - BODProcesses and formulations

WAQ_Oxygen_03 2

Introduction

Problems associated with oxygen:Oxygen depletion in the water column causes fish mortalityAnoxic conditions may occur in stratified systemsBad smell (H2S formation)Accumulation of heavy metals under anoxic conditions

WAQ_Oxygen_03 3

Processes

… that increase the oxygenconcentration in the water columnprimary production(photosynthesis)reaeration over the atmosphere-water interface (in case ofundersaturation)uptake of nitrate by algaedenitrification

… that decrease the oxygenconcentration in the water columnrespirationreaeration (in case ofoversaturation)decay of organic material in thewater columnoxidation of reduced substances(such as S2-)sediment oxygen demandnitrification

WAQ_Oxygen_03 4

Reactions (1)

Decay of organic carbonCH2O + O2® CO2 + H2O

Primary productionCO2 + H2O ® CH2O + O2

molar ratio C : O2 = 1 : 1weight ratio C : O2 = 12 : 32 = 1 : 2.67

Oxidation of ammonia (nitrification)NH4

+ + 2 O2 ® NO3- + H2O + 2 H+

molar ratio N : O2 = 1 : 2weight ratio N : O2 = 14 : 64 = 1 : 4.57

WAQ_Oxygen_03 5

Reactions (2)

Denitrification

NO3- + H+ ® 0.5 N2 + 0.5 H2O + 1.25 O2

molar ratio N : O2 = 1 : 1.25weight ratio N : O2 = 14 : 40 = 1 : 2.86

WAQ_Oxygen_03 6

‘Negative’ oxygen concentrations

Reduced substances (e.g. sulphides) are not modelled in Delft3D-WAQ

Instead the oxygen concentration can drop below zeroThis reflects anoxic conditions and should be interpreted as an

oxygen deficit

WAQ_Oxygen_03 7

Spectrum of organic material

Fresh organic material (e.g. mortality algae, faeces)Untreated sewage dischargeTreated sewage discharge:

• screening• primary treatment• secondary treatment• chemical treatment

Dissolved organic carbonParticulate particles (peels, wood, etc.)

WAQ_Oxygen_03 8

Biological Oxygen Demand (1)

Lumped parameters are often used to estimatethe quantity of organic matter present:

BOD: Biological Oxygen DemandCOD: Chemical Oxygen DemandTOC: Total Organic Carbon

COD includes nitrogenous oxygen demand (organic nitrogen,ammonia) as well as carbonaceous oxygen demand: COD = CBOD +NBOD

(other reduced substances are not included in Delft3D-WAQ)

WAQ_Oxygen_03 9

Biological Oxygen Demand (2)

A value often used, is BOD5:the amount of oxygen that is consumed within 5 days after sealingof a water sample

Nitrogenous oxygen demand generally starts after 5 days and isthus not included in BOD5 measurements.

WAQ_Oxygen_03 10

Types of organic material in Delft3D-WAQ

Living organic matter (Algae)Detritus Organic Carbon (DetC)Other Organic Carbon (OOC)several BOD fractionsAll fractions have corresponding substances in the sediment

Usually either the combination Algae/DetC/OOC is modelled or BODis modelled. When modelling both at the same time be aware ofdouble counting of the organic fraction

WAQ_Oxygen_03 11

Formulation for decay (1)

dMin = Z + Rc ´ C ´ f(T)

[gC/m3.d] [gC/m3.d] [1/d] ´ [gC/m3] ´ [-]flux zeroth order first order temp.

dependency

f(T) = Tc (Temp- 20)

WAQ_Oxygen_03 12

Formulation of decay (2)

range of decay rates (Rc)0.75 d-1 for fresh material0.25 d-1 for average material0.10 d-1 for older material

Temperature dependency (Tc)range of values 1.02 - 1.09

WAQ_Oxygen_03 13

Formulation for reaeration

dREAR = RcRear ´ (SaturOXY - OXY) ´ f(T)

[gO2/m3.d] [1/d] ´ [gO2/m3] ´ [-]flux rate oxygen temperature

constant deficit dependency

RcRear = f (flow velocity, wind speed, water depth)® choice of 11 formulations

SaturOxy = saturation concentration= f (Salinity, Temperature)® choice of 2 formulations

WAQ_Oxygen_03 14

Important parameters in oxygen - BOD modelling

Decay / mineralisation rates organic matter (including BOD) both inthe water column and the sediment

Switch for reaeration (which formulation for reaeration do you use)

Others:zeroth-order Sediment Oxygen Demand (fSOD)primary production (algae)

Contents

Deltares 95

Contents

Bract, odd page

Deltares 96

Contents

Bract, even page

Deltares 97

WAQ_Suspended_04


Suspended SedimentSedimentation and erosion

WAQ_Suspended_04 2

What is sediment?

Sediment is a mixture of• clay particles• silt particles• sand particles• organic material• water• (gas)

Composition and properties vary in time and spaceclassified as cohesive and non-cohesive sediments

WAQ_Suspended_04 3

Introduction

Cohesive sediment: silt - clay (< 63 mm)

Very fine fraction that can be suspended easily and is transported byadvection and turbulent motion.

Non-cohesive sediment: sand - gravel (63 mm - 20 mm)

Coarse fraction that is mainly transported by bed-load.

The focus of Delft3D-WAQ is mainly on cohesive sediment

WAQ_Suspended_04 4

What is mud?

mud is cohesive due to physico-chemical properties of the sedimentand water:

• a diffusive double layer is formed• mud particles form flocs (time dependent)• flocs dramatically increase the settling velocity• cohesive sediment has a much longer residence time in the

water than sand: only suspension transport• appearances:

> dilute suspensions (< 1 g/l)> high concentrated suspension (1-10 g/l)> flocs (5-100 g/l)> fluid mud (150 - 400 g/l) (density currents, not in WAQ!!)

WAQ_Suspended_04 5

Issues

Issues associated with suspended sediment:•high turbidity disturbing underwater life•silting up of fairways•transport of pollutants adsorbed to suspended sediment•release of adsorbed materials after resuspension of sediment•clogging of intakes

Modelling issues•large spin-up time of models•initial availability at bed needed to get water column values

>errorneous sources keep on acting>start with a starved bed, and let model calculate bed composition

WAQ_Suspended_04 6

Relevant processes

erosion of the bed by currents and wavesdeposition of suspended mattervertical transport: turbulence and stratificationhorizontal transportsettling and consolidation (self-weight)

vertical concentration profile with low densities at the surface andhigh at the bottom

bottom layer may get its own gravity-driven flowsconstant settling velocity (v) and vertical diffusity (D):

c c ebottom

vD

z= ×

-

WAQ_Suspended_04 7

Processes

SedimentationResuspension

Critical parameterÞ Bottom shear stress (Tau)

® Sedimentation occurs if Tau is lower than a critical value;® Resuspension occurs when Tau exceeds a critical value

WAQ_Suspended_04 8

Shear stress

Force exerted on the sediment through:• flow of water over the sediment - water interface• wave action (important in shallow areas)• other turbulence (such as ships or fishes)

Tau can be derived from the hydrodynamic calculation; orTau can be calculated by Delft3D-WAQ

Especially spatial and temporal distributions of the shear stress matter

WAQ_Suspended_04 9

Bottom shear stress

tauSed tauRes

Bottom shear stress (tau)

Sedimentation Resuspension

FLOWl

2

2tau = g VelocityChezy

r . .

WAQ_Suspended_04 10

Model implementation

contains 3 inorganic sediment fractions (IM1, IM2 and IM3)2 (homogeneous) sediment layers S1 and S2shear stresses from Delft3D-FLOWsedimentation/erosion processes at water/bottom interface includedburial/digging between sediment layers includedmodel computes:

• sedimentation/erosion fluxes• concentration in the water column• layer thickness of bed

WAQ_Suspended_04

WAQ_Suspended_04 12

Formulation for sedimentation

( ) ( ) ( )

2 2 3[ / / ] [ ] [ / / ] [ / ] [ / ]

SED

SED

g m d g m d m d g m

fSedIM1 P (ZSedIM1 VSed IM1)

P Max 0 , 1cs

VSed VSed0 f Sal f flocculation f T

tt

-

= ´ + ´

æ öç ÷= -ç ÷è ø

= × × ×

WAQ_Suspended_04 13

Sedimentation (2)

Important parameters•Settling velocity (VSedi in m/d)

normal range of values: 0.1 - 20 m/d•critical shear stress for sedimentation (TaucSi in Pa)

normal range of values: 0.05 - 1.0 Pa

Note:•Zero-order sedimentation rate should not be used•Settling velocity can be a function of salinity, total suspendedsediment concentration and temperature

WAQ_Suspended_04 14

Effects on settling velocity (models)

2D and 3D models need different values, in 3D the ‘fall velocity’ isneeded

Settling velocity depends on the number of fractions usedSettling velocity depends on the calibration. What kind of data is

used, and what are the criteria (concentrations, fluxes, …?)Does a model simulate tide and waves, or are these effects

parameterised or lumped?

Formulation for resuspension

WAQ_Suspended_04 15

2 2 2[ / / ] [ ] [ / / ] [1 / ] [ ] / [ ]

Re ( )

0, 1

RES

RES

g m d g m d d g m

VResDM DMS1f sDM P ZResDMSurf

P MaxcR

tt

-

´= ´ +

æ öæ öç ÷ç ÷= -

ç ÷ç ÷è øè ø

WAQ_Suspended_04 16

Resuspension (2)

Important parameters•critical shear stress for resuspension (in Pa)

normal range of values: 0.05 - 1.0 Pa•zero order resuspension rate (in g/m2/d)

NOTE that for resuspension the total resuspension flux is calculated.The resuspension of individual substances is equal to the weightfraction in the sediment times the total resuspension flux.

The first order resuspension rate (VresDM) should be used with care:it mimicks a partial coverage of a sand bed with thin mud foils(patches, spots), so the erosion is proptional to the area covered.

Erosion/sedimentation under tidal conditions

WAQ_Suspended_04 17

WAQ_Suspended_04 18

Burial/Digging

Mimicks effects of consolidation: different time scales

Downward sediment transport from one sediment layer to anunderlying layer (burial)

Upward sediment transport to a sediment layer from an underlyinglayer (digging)

Two options:• Constant burial and/or digging rate• Burial and/or digging flux determined by the thickness of the

bottom layer.

Contents

Deltares 107

Contents

Bract, odd page

Deltares 108

Contents

Bract, even page

Deltares 109

WAQ_Nutrients_04


NutrientsCycles of nitrogen, phosphorusand silicon

WAQ_Nutrients_04 2

Introduction

Nutrients are the building blocks for life

Only macro-nutrients are considered:• nitrogen ® ammonia (NH4

+) and nitrate (NO3-)

• phosphorus ® phosphate (PO43-)

• silicon ® silicate (SiO2)

WAQ_Nutrients_04 3

Nitrogen cycle

Pools of nitrogen• Phytoplankton• Organic nitrogen• Ammonia• Nitrate• Sediment

Cascade :Algal-N ® Detritus-N ® NH4 ® NO3 ® N2

WAQ_Nutrients_04

WAQ_Nutrients_04 5

Formulations (1)

Mineralisation : DetN ® NH4

dMin = Z + Rc ´ C ´ f(T)

[gN/m3.d] [gN/m3.d] [1/d] ´ [gN/m3] ´ [-]flux zeroth order first order temp.

dependency

f(T) = Tc (Temp- 20)

WAQ_Nutrients_04 6

Formulations (2)

Nitrification : NH4 ® NO3

NH4+ + 2 O2 ® NO3

- + H2O + 2 H+

dNitrif = Z + RcNit ´ NH4 ´ f(T) ´ f(O2)

zeroth-order flux (Z) + first-order term

WAQ_Nutrients_04 7

O2-function nitrification

WAQ_Nutrients_04 8

Formulations (3)

Denitrification : NO3 ® N2

NO3- + H+ ® 0.5 N2 + 0.5 H2O + 1.25 O2

water column :dDenitr = Z + RcDenWat ´ NO3 ´ f(T) ´ f(O2)

sediment :dDenitr = (Z + RcDenSed ´ NO3 ´ f(T) ) / Depth

WAQ_Nutrients_04 9

O2-function denitrification

WAQ_Nutrients_04 10

Formulations (4)

Denitrification in the sediment is modelled as a sink for nitrate only.There is no link with oxygen or organic carbon.

Note:Sophisticated sediment modules for the calculation of sediment-waterexchange of nutrients and oxygen will be available shortly.

WAQ_Nutrients_04 11

Phosphorus cycle (1)

Pools of phosphorus• Phytoplankton• Organic phosphorus• dissolved Phosphate• adsorbed Phosphate• Sediment

Cascade :Algal-P ® Detritus-P ® PO4 « adsorbed-P

WAQ_Nutrients_04 12

Phosphorus cycle (2)

Alg-P DetP

PO4 AAP

DetPS1 AAPS1

Water

Sediment

1

2

34

55

5

6

78

1. Autolysis2. Uptake3. Mortality4. Mineralisation water column5. Sedimentation6. Adsorption7. Mineralisation sediment8. Desorption

WAQ_Nutrients_04 13

Adsorption (1)

Two options :1. Partition coefficient

Partitioning instantaneousPartitioning independent of available adsorption places

K POAAPD = 4

WAQ_Nutrients_04 14

Adsorption (2)

2. Langmuir

Adsorption capacity (MaxPO4) depends on the characteristics of theinorganic material (especially iron content)

dAds Rc AAP AAP

AAP TIM K PO MaxPOK PO MaxPO

equilibrium actual

equilibriumD

D

= ´ -

= ´´ ´´ +

( )

4 4

4 4

WAQ_Nutrients_04 15

Silicon cycle

Pools of siliconPhytoplanktonOrganic Sidissolved SiSediment

CascadeAlg-Si ® Organic Si ® Si

WAQ_Nutrients_04

WAQ_Nutrients_04 17

Important parameters in nutrient modelling

Decay / mineralisation rates organic matter both in the water columnand the sediment

(De)nitrification rateAdsorption coefficient PO4 - AAP

Contents

Deltares 119

Contents

Bract, odd page

Deltares 120

Contents

Bract, even page

Deltares 121

WAQ_Algae_03

Modelling of Algae in Delft3D-WAQ

General introduction onalgae growth

WAQ_Algae_03 2

General introduction on algae growth (1)

General aspects of algae growth:requirements for growthdiatoms versus flagellatesseasonal dynamicsgradients (horizontal, vertical)

WAQ_Algae_03 3


Algae need for their growth:macronutrients (nitrogen, phosphorus, silicate)micronutrients (vitamins, iron, etc.)light

Simple conclusion (too simple):the more nutrients, the more algaethe more light, the more algae

WAQ_Algae_03 4


If there is no phosphorus in the water, additional nitrogen will notincrease algae concentrations.

Phosphorus is then the limiting factor

Algal strategies to cope with limiting factors:more efficient use of the limiting factorstorage of nutrients or energy

WAQ_Algae_03 5


Shifts in limiting factors lead to:adaptations of algal physiologyshifts in species composition towards species that are better

adapted (for example shift to flagellates when silica isexhausted)

WAQ_Algae_03 6


Seasonal dynamics:

00 w inter summer winter

nutrients

light

WAQ_Algae_03 7


Dutch coastal waters

Chlorofylconcentrations 1975-1998, Dutch coastal waters

0

10

20

30

40

50

60

70

dec-74 jan-75 feb-75 mrt-75 apr-75 mei-75 jun-75 jul-75 aug-75 sep-75 okt-75 nov-75 dec-75

month

chlo

rofy

lcon

cent

ratio

n(u

g/l)

WAQ_Algae_03 8


Horizontal gradients in coastal waters:

01 off-shore shore

more light, less nutrients less light, more nutrients

nutrients

light

WAQ_Algae_03 9

General introduction on algae modelling (8)

Vertical gradients in an estuary:

WAQ_Algae_03

Algal physiology and WAQ-approach

WAQ_Algae_03 11


Processes associated with algae growth:uptake and release of nutrients and oxygengrowth: increase of biomassmortality: decrease in biomassrespiration, energy for metabolism

Biomass increase = growth - mortality - respiration

(unity for algae is not numbers but biomass: gram carbon)

WAQ_Algae_03 12


Uptake and release of nutrients is proportional to the biomass createdor released, according to the (fixed) stochiometry.

Release of nutrients during mortality:autolysis fraction: inorganic nutrientsdetritus fraction: organic decaying nutrientsother organic compounds: slow decay

WAQ_Algae_03 13


Growth (primary production):

Growth is equal to maximum growth (PPMax) corrected fortemperature and limitations due to nutrients, light and daylength.

PProd = LimDl LimNut LimRad TFGro PPMaxi i i i i i´ ´ ´ ´

WAQ_Algae_03 14


Temperature function:

i iTemp-20TFGro = TCGro

WAQ_Algae_03 15


Nutrientlimitation:

i i i i

ii

ii

ii

i

i i

L i m N U T = ( L im N , L im P , L im S i )

L i m N = D I ND I N + K m D I N

L i m P = ( P O 4 )( P O 4 ) + K m P

L i m S i = ( S i )( S i ) + K m S i

D I N = ( N H 4 ) + ( N O 3 )P r fN H 4

i f K m S i = - 1 t h e n L im S i = 1 .0

M in

WAQ_Algae_03 16


Respiration has 2 components:maintenance respirationgrowth respiration

RcResp = PP rod GResp + MResp TFMrt (1 - GResp )i i i i i i´ ´ ´

WAQ_Algae_03 17


Mortality:Constant mortality rate (1/day), only corrected for temperature.

With temperature function:

RcMrt Mort TFMrti i= ´

i iTemp-20TFMrt = TCDeC

WAQ_Algae_03

Light and Extinction

WAQ_Algae_03 19

Light and extinction

Extinction with depthExtinction by different components

® inorganic material® organic material® algae® background (water and dissolved substances)

Daylight limitation(Spectral distribution)

Contents

Bract, odd page

Deltares 132

Contents

Bract, even page

Deltares 133

WAQ_Statistics


Statistical output

WAQ_Statistics 2

Statistical output

Defined on periods:Standard statisticsGeometric meanPercentiles and quantiles

Defined every output time:Depth Average, max. and min. over the depthPeriodic average

WAQ_Statistics 3

Statistics defined on periods

Defined on periods:You can define as many periods as you like. For example simulation

period, annual period, wet season, dry season, every month.You specify the start and stop time of the period and a suffix used to

identify the periodOutput will be written to two extra output filesFor the observation points to a ASCII monitoring file with the name

<runid>-stat.monFor the complete grid to a binary map file with the name <runid>-

stat.map

WAQ_Statistics 4

Statistics per output time

Defined every output step :Output will be written to the standard output filesYou specify a suffix which is used to identify the statistical operationOutput can take a lot of disk space(NEFIS output file has a maximum of 100 output parameters)

WAQ_Statistics 5

General features

You can get statistics on every substance and output parameterThe name of the statistic variable will be a concatenation of the

parameter name, the period suffix and the suffix used to identifythe statistical operation

The names are reported in the <runid>.lst fileYou can not have statistics on statistics

WAQ_Statistics 6

Standard statistical quantities

Minimum value over all periodsMaximum value over all periodsMean value over all periodsStandard deviation over all periods (stdev)

WAQ_Statistics 7

Geometrical mean

Log averaged value over all periodsOften used in objectives for bacteriaTends to suppress the influence of extreme valuesA zero value in concentration leads to an undefined geometric, so we

give a threshold value below which the value is not used or set tothis threshold value.

WAQ_Statistics 8

Percentiles

Exceedance timeYou specify the critical concentration of exceedanceYou can get the reverse (for example in case of oxygen)

WAQ_Statistics 9

Quantiles

Exceedance concentration for a specific percentage of time within theperiod

You specify the quantile percentageThe concentration frequency is stored in a number of classes which

must be specifiedOperation uses a lot of in-core memory when you increase the

number of classes

WAQ_Statistics 10

Depth averages

You get the depth averaged concentration and the minimum and themaximum over the water column

You get the same output for every layer

WAQ_Statistics 11

Periodic averages

You get the averaged concentration of the last finished periodOften used for daily average or tidal averageYou can specify a time offset to synchronise the periodsThis is not a moving average

Contents

Bract, odd page

Deltares 140

Contents

Bract, even page

Deltares 141

WAQ_Toxic_substances


Toxic substances

WAQ_Toxic_substances 2

Toxic substances

Detrimental effects:

• Acute toxicity (LD50, no effect level)• Chronic toxicity (LC50, no effect level)• Decreasing growth-rate, mobility, etc.• Decreasing fertility, increase of dead borne• Carcinogenity


Main groups of substances

Heavy Metals (HM)

Organic Micro Pollutants (OMP)

Radioactive Substances


Heavy metals

no decay or decompositiononly evaporate as part of organic complexesIon-binding to inorganic suspended sedimentsedimentation with suspended sedimenterosion with suspended sedimentremobilisation (ionic strength, pH, anaerobic conditions)


Groups of heavy metals

Sulphide forming heavy metals Cd, Cu, Zn, Ni, Hg, Pb (group 1)

Hydroxide forming metal Cr (group 2)

Anion forming “metals” As and Va (group 3)


Partitioning

fdis = fraction dissolvedfads = fraction adsorbedj = porositySS = suspended sedimentKd = partitioning coefficient

310(1 )

disd

ads dis

f K SS

f f

j

j=

´+

= -


Heavy metals

Partitioning coefficient Kd

for every particulate fractionalternative formulations see the manual


Organic micropllutants

VolatilizationPhotolysisHydrolysisBiodegradationDecompostionBinding to organic material


Organic micropollutants

Volatilization uses Henry’s law for gas exchange

Degradation lumped in one first order decay

Equilibrium partitioning implemented

Non equilibrium partitioning implemented

Contents

Deltares 147

Contents

Bract, odd page

Deltares 148

Contents

Bract, even page

Deltares 149

WAQ_OpenPLCT


Open PLCT – define your own processes

WAQ_OpenPLCT 2

The open processes library

Two parts:Library of computational routines:

• Define your own water quality processesUser-interface to set up the administration:

• Provide skeleton source code

Once done, the library contains all of the standard processes and your ownprocesses.

WAQ_OpenPLCT 3

Principles

All process routines work the same:Rely on the framework to provide the concentrations and values for

process parameters (that way, a process parameter can be aconstant, a function of time or something varying over the modelarea)

Determine the change, called a flux, in the concentration of thedependent substances. (Note: the unit is g/m3/day)

Determine any extra outputStore the results in the relevant arraysDefine the “stoichiometric” constants

WAQ_OpenPLCT 4

Example

Decay of organic matter (BOD) and its influence on dissolved oxygen:

The flux is “k*BOD”The stoichiometric constant for BOD is -1, as the flux is to be subtractedfrom the concentration.The stoichiometric constant for DO is -1 as well. Same reason

d BOD k BODdt

d DO k BODdt

= -

= -

WAQ_OpenPLCT 5

Some details

Any number of substances as inputAny number of process parametersAny number of fluxes as outputAny number of output parameters

But:The more a water quality process routine is doing, the more difficultit is to use it.Often small routines are better

WAQ_OpenPLCT 6

Demonstration

Implement the BOD-DO process …

Contents

Deltares 153

Contents

Bract, odd page

Deltares 154

Contents

Bract, even page

Deltares 155

PART_Concepts_07

Introductory Course Delft3D-PART

CONCEPTS of particle tracking

PART_Concepts_07 2

• Delft3D-PART applications.• Difference between Delft3D-WAQ and Delft3D-PART• Pros and cons Delft3D-PART.• Tutorial (a 1st introduction).• Concepts• Exercises• Summary• Example of an oil application.

Contents

PART_Concepts_07 3

Delft3D-PART

Delft3D-PART applications …

PART_Concepts_07 4

The Delft3D-PART model concept is commonly applied forplume modelling.

Examples of Delft3D-PART studies:Sediment plumes due to dredging activities.Plume dilution studies with spatial scales of several kms (from

outfalls)Oil spill modelling.

Delft3D-PART

PART_Concepts_07 5

Sediment plume

PART_Concepts_07 6

Production water

Release of production water from an oil platform:

PART_Concepts_07 7

Oil spill

Oil tanker: Prestige

PART_Concepts_07 8

Difference between Delft3D-WAQand Delft3D-PART

PART_Concepts_07 9

Delft3D-WAQ• Deterministic• Basic unit is mass / concentration per segment• Far-field• Time scales (spring-neap to annual)• (Eulerian approach)

Delft3D-PART• Deterministic and stochastic• Basic units are mass and location of every particle• Mid-field: 15 m – few kms• Time scales (hours to days)• (Lagrangian approach)

WAQ-PART Differences

PART_Concepts_07 10

Advantages:• Measure of detail not limited by hydrodynamic grid size• Rapid calculation time.

Disadvantages:• Not suitable for long term simulations.

> only a few days and a few kms, otherwise too many particles(and simulation time) are needed.

• Only simple processes.

Pros and cons Delft3D-PART

PART_Concepts_07 11

Concepts

Concepts

PART_Concepts_07 12

Particle tracking

A particle is nothing more than a coordinate with a mass.Particles are transported and dispersed in the water.The path of each individual particle is tracked:

every time step the exact position of each particle is known ® P (t, x, y, z)

P (t,x,y,z)

P (t+Dt,x,y,z)

P (t+2Dt,x,y,z)

P (t+3Dt,x,y,z)

(Dx,Dy,Dz)

(Dx,Dy,Dz)

(Dx,Dy,Dz)

PART_Concepts_07 13

Delft3D-PART can model different types of substances:• tracer/sediments,• oil-spill

Each type of substance contains different processes.

Each particle has properties.These properties can influence both transport (Dx, Dy, Dz) and

mass.

PART_Concepts_07 14

General equation:

Mass (t+Dt, x+Dx,y+Dy,z+Dz) =Mass (t,x,y,z) +advective transport (Dx,Dy,Dz) +dispersive transport (Dx,Dy,Dz) +processes

PART_Concepts_07 15

Substances in the water are modelled by particles that aretransported and dispersed in the water:

Advective transport from Delft3D-FLOW.Dispersion as ‘random’ (stochastic) turbulent motion.Processes affect the mass or ‘type’ of the particles

• sediments and oil change types;• tracers do not change types.

PART_Concepts_07 16

Advective transport 1(2)

Advection (velocity and direction) according to the underlyingDelft3D-FLOW model.

Horizontal aggregation of the hydrodynamic database is notallowed.

Velocities are linearly interpolated within the segment.

Velocity indirection x

Dx

Mass (t+Dt, x+Dx,y+Dy,z+Dz) =Mass (t,x,y,z) +

advective transport (Dx,Dy,Dz) +dispersive transport (Dx,Dy,Dz) +processes

PART_Concepts_07 17

Advective transport 2(2)

In 3D flow models, horizontal and vertical transport arecalculated in the same way (linear interpolation of velocitiesat the segment boundaries).(3D models can be coupled to 2D)

In 2D flow models, Delft3D-PART creates a vertical velocityprofile, taking into account the effect of wind and bottomfriction.So: with 2D flow Delft3D-PART can run in semi-3D mode.• Logarithmic profile for bottom shear.• Parabolic profile for wind stress.


PART_Concepts_07 18

Horizontal dispersion

Random step in a square with a width and height of 2x maxstepwith a uniform probability density function.

,6 x ymaxstep D t= D

Dx,y : Dispersion coefficient

Dt : time stepmaxstep

maxstep


PART_Concepts_07 19

Vertical dispersion

Same idea only different value for maxstep in vertical.

Maxstep is now calculated from the vertical dispersioncoefficient DZ (2 options: user-defined or depth averagedalgebraic)


maxstep

max

step

PART_Concepts_07 20

Demonstration: A particles track plotted on the hydrodynamicgrid combination of advection and dispersion.

PART_Concepts_07 21

Processes

Tracer/sediment:• decay (1st order)• sedimentation• (resuspension)

Oil (advanced course):Different processes and properties.


PART_Concepts_07 22

Decay

• First order decay:

• Mass per particle decreases in time.

1with k: decay rateday

dm k mdt

= - ×

æ öç ÷è ø

PART_Concepts_07 23

Sedimentation and resuspension

• All particles get an equal settling velocity in the vertical.• When the shear stress is below a critical value, particles

hitting the bottom settle into the sediment layer.• When the bottom shear stress exceeds a critical value, all

settled particles go back into suspension (!).

PART_Concepts_07 24

Output: from particles to concentration

• The distribution of the particles is translated intoconcentrations according to:

• Using an output grid that can be:• the hydrodynamic grid• a user-defined zoom grid

i

Pnr

i

i Volume

PMassionConcentrat

i

å=

_

1_

PART_Concepts_07 25

Exercise 1

PART_Concepts_07 26

Tutorial: a first introductionto the user interface

PART_Concepts_07 27

Step-by-step instruction to complete a simulation with twoconservative tracers:One substance is released at the surface as an instantaneous

release.The other substance is released continuously at the bottom

layer of the model.

Demonstration

PART_Concepts_07 28

For a Delft3D-PART simulation the following info needs to bespecified:

hydrodynamicssubstancestime framenumerical parameters: number of particles, vertical

dispersion…releases: location, number of particles, mass, typeprocess parameters: which processes, horizontal dispersionoutput: time interval

Input PART model

PART_Concepts_07 29

PART_Concepts_07 30

PART_Concepts_07 31

There are no boundary conditions: particles which aretransported outside model domain cannot return.

Delft3D-PART assumes that there are initially no particlespresent in the computational domain.

Generally work through screens from top to bottom.

PART_Concepts_07 32

PART_Concepts_07 33

‘hyd’ file from coupling required.

PART_Concepts_07 34

2 substances: each substance can be assigned different properties.

PART_Concepts_07 35

Time step must be less than orequal to time step in hydrodyn.

Use to adjust the presentation ofthe output results.

PART_Concepts_07 36

Important for required resolution

For well-mixed systems. Stratification is not accounted for.

PART_Concepts_07 37

Actual total number of particles may be greater or less than 100%.

One substance is released at the surface as an instantaneous release.

PART_Concepts_07 38

PART_Concepts_07 39

Specify mass of each of the substances to be released.Here, only one substance is released as an instantaneous release.

PART_Concepts_07 40

The other substance is released continuously at the bottom layer of the model.

PART_Concepts_07 41

Specify time-series of discharge for each release for the entiresimulation period. The first time breakpoint must be equal to thesimulation start… etc

PART_Concepts_07 42

PART_Concepts_07 43

A time-dependent decay rate can be specified.Linear interpolation between these breakpoints.

For conservative tracers the decay rate is zero.

PART_Concepts_07 44

Sedimentation and erosion according to Parheniades and Krone.

PART_Concepts_07 45

PART_Concepts_07 46

Roughness length to account for bottom friction.

Wind direction given as direction fromwhich wind is blowing.

PART_Concepts_07 47

On subgrid scale though indicated in segment

PART_Concepts_07 48

Output types:

• History file: for selected observation points at specified time intervals.

• Map files: for the entire computational grid.

• Zoom grid: a user-defined subgrid area at specified time intervals.

PART_Concepts_07 49

Zoom grid cell: 100m x 100m

grid definition: zoom grid

PART_Concepts_07 50

3D tracks: creates an extra output filethat contains information about the 3Dlocation of every particle in thesimulation at every time step.

Time breakpoints for which outputis written to zoom grid

PART_Concepts_07 51

Input files:GUI sets up the input files for Delft3D-PART• *.mdp

Output files:• his: time series for observation points.• map: spatial plots for whole hydrodynamic grid.• plo: outputs for the zoom grid.• trk: outputs for the 3D particle tracks.

Summary

PART_Concepts_07 52

Tracer concentration (instantaneous release) in the surface layer.

05 Aug 1990 12:45 05 Aug 1990 13:00

Upper: map file on flow gridLower: plo file on zoom grid.


Upper: map file on flow grid.Lower: plo file on zoom grid.

PART_Concepts_07 53

05 Aug 1990 14:00 05 Aug 1990 16:00



PART_Concepts_07 54

05 Aug 1990 18:00 05 Aug 1990 20:00



PART_Concepts_07 55

Notice the higher concentrationson zoomgrid:

• Mass per particle is constant.

• c=m/V.

• V zoomgrid << V hydrodynamic

grid.

05 Aug 1990 20:30



PART_Concepts_07 56

Exercise 2

PART_Concepts_07 57

Exercise 3

PART_Concepts_07 58

Demonstration: Effect of dispersion

PART_Concepts_07 59

A particle track:• Blue: without dispersion: Dx,y = 0• Green: with dispersion (Dx,y = 100)• Orange: with dispersion (Dx,y = 100 different realisation)

PART_Concepts_07 60

Continuous release

PART_Concepts_07 61

Instantaneous release

PART_Concepts_07 62

Exercise 4

PART_Concepts_07 63

Demonstration : Effect of zoom grid/output

PART_Concepts_07 64

Output of cont-release01 on 1 dec 2005 00:00:• left: concentrations on hydrodynamic grid (from *.map).• right: concentrations on zoom grid (from *.plo).

PART_Concepts_07 65

Exercise 5

PART_Concepts_07 66

Demonstration: Effect of number of particles

PART_Concepts_07 67

Release rate: 1 kg/s x 86400 s = 86400 kg; zoom grid volume:10mx1000mx1000m = 107 m3.

• Left: 1000 particles. 1 Particle = 86.4 kgà 8.64 x10-6 kg/m3 =blue (2 particles: 1.7x10-5 = green).

• Right: 100.000 particles. 1 Particle = 0.864 kgà 8.46x10-8

kg/m3.

PART_Concepts_07 68

Exercise 6

PART_Concepts_07 69

Demonstration: Effect of number ofparticles, part 2

PART_Concepts_07 70

Time series: continuous release:• B: D=0; 1000 particles• G: D=100; 1000 particles• R: D=100; 100.000 particles

PART_Concepts_07 71

Exercise 7

PART_Concepts_07 72

Demonstration: Maximum step as a function ofdiffusion and timestep

PART_Concepts_07 73

particle tracks: inst-release01:

PART_Concepts_07 74

1 Nov 2005 00:00 1 Nov 2005 02:00

1 Nov 2005 11:00 2 Nov 2005 00:00

PART_Concepts_07 75

Inst-release: maxstep• Blue: track of 1000 particles at t = 1, 2 and 3• Green: track of 1000 particles at t=2 (1 timestep after

release)

PART_Concepts_07 76

Exercise 8

PART_Concepts_07 77

• Recapitulation of exercises and demos.For Delft3D-PART modelling very important:• zoom grid• number of particles• duration of simulation

• Delft3D-PART mainly used for oil spill modelling. Similar totracer only with extra processes.

PART_Concepts_07 78

The fate of an oil spill: processes

PART_Concepts_07 79

Demonstration of an oil applicationHong Kong

PART_Concepts_07 80

PART_Concepts_07 81

• Advanced Delft3D-PART course.• Delft3D-PART under continuous development:

• new processes: resurfacing of oil• stochastic winds• reverse modelling

Contents

Deltares 197

Contents

Bract, odd page

Deltares 198

Contents

Bract, even page

Deltares 199

Contents

Exercises introduction course Environmental,far-field water quality

0 Calculation of a simple mass balance

There is a lake with dimension 200 m×1000 m and an average depth of 2 m. A river with adischarge of 10 m3/s and a cross-section of 20 m2 flows into the lake. Through evaporation0.5 m3/s is lost from the lake. A factory re-circulates water from the lake for cooling purposesat a rate of 1 m3/s. Five kilometres upstream of the river mouth, another factory has a licenceto discharge an organic pollutant up to a concentration of 1 mg/l at a maximum discharge of15 l/s.

1 What is the average residence time of water in the lake?2 What is the average concentration of the organic pollutant in the lake?3 If the decay rate of the organic pollutant is 0.5 d−1 (assuming a first order decay rate:C = C0 exp(−kt)), what will be the average concentration in the river as it enters thelake?

4 What will be the average concentration in the lake then?

Tips:

� Assume that the lake is well-mixed.

Deltares 200

Contents


1 Coupling of the hydrodynamic database

For horizontal aggregation

� Select GRID in the Delft3D main window� Select DIDO in the Grid and bathymetry window

For coupling the hydrodynamic datbase

� Select Water Quality in Delft3D main window� Select General in the Far-field water quality window� Select Coupling in the Water quality (WAQ) window� Select ‘Define input’� Select ‘Hydrodynamics’ in coupling module� Select communication file <com-∗.dat>

1.1

Carry out a 2×2 regular aggregation of the grid for the Bornrif area. It consists of an ”outer”sea part, a narrow channel between two islands and a more or less enclosed area, near themainland (there are two so-called ”wantijs” to the west and east, areas where two tidal wavescancel each other).

Tips:

� You need to set the working directory in the Delft3D menu to that containing the MDF file(this has to do with a limitation of DIDO)

� Then you can load the MDF file that is found in <far field/hydro/bornrif-fine>� By selecting the MDF file all relevant information is automatically loaded: the open bound-

aries, the dry points and the depth.� Create a regular aggregation - use the right mouse button to make the aggregation defini-

tive.� Zoom in on the landboundaries to see what individual cells are left.� Select the ”Irregular” mode from the menu. Select grid cells that should be taken together

via the left mouse button. When finished, click with the right mouse button to make theaggregation permanent.

Deltares 201

Contents

1.2

Start with the same grid, but now create an irregular aggregation, using a 1×1 aggregation forthe inner region and a 2×2 aggregation in the outer region. Set the option to view the depthto ”on”. This will give you more insight in the characteristics of the region.

Wad

den

sea

North sea

The Netherlands

Create an irregular aggregation that complies with the following specifications:1×1 for Depth < 10 m,2×2 for 10 m < Depth < 20 m,4×4 for Depth > 20 m.

1.3

Load the monitoring file<far field/hydro/bornrif-fine/trih-rif-waq.dat> into Delft3D-QUICKPLOTto inspect the timeseries for the various monitoring points.

Questions:

� Determine the rough dimensions of the inner region via QUICKIN. What is the total volumeof this region?

� Given the tidal range, give an estimate of the residence time of water in the region. Whatis the consequence for a water quality computation?

� Load the communications file <far field/hydro/bornrif-fine/trih-rif-waq.dat> into the cou-pling user-interface. What can you tell about the stored period in relation to the estimatedresidence time? What about the stored period in relation to the tidal period?

Now: Run the coupling program.

Deltares 202

Contents

We will use the results for this region for several other exercises.

� Check the coupling report file <couplnef.out>. What does it tell you about the maximumtimestep for a water quality computation?

Deltares 203

Contents


2 Working with the Delft3D-WAQ and Delft3D-SED User Interface

2.1

Using the coupled hydrodynamic files generated in exercise 1.4 for the Bornrif area, check themass balance and calculate the residence time for the segments. You can use the substancefile:

<course models/substance/continuity 3.sub>.

Tips:

� We check the hydrodynamic water (or mass) balance by assigning a concentration of 1mg/l to all water. This means that the initial condition is 1 mg/l, but also the water enteringand leaving the model area (boundary conditions, discharges) has a concentration of 1mg/l. For this we use a substance called Continuity, which is a non-reacting substances.So, as nothing happens to the substance, the concentration should remain 1 mg/l, whenwe run the simulation. Usually, this will indeed be the case, but if not, something has gonewrong which has to be solved first, before continuing.

� Work accurately from top to button through the buttons in the User Interface, as it is easyto miss an item.

� When you have made a mistake or when you want to change something in a previouslysaved scenario file <∗.scn>, you can simply open the file and make your change. It isrecommended that you save the new scenario file under another name.

� When you have to do a lot of simulations or calibration runs, you should keep a logbookto keep track of the changes you have made and the simulations you have run.

� Also, it is advisable to set up a directory structure in which to save all your work. To storeall files in just one directory leads quickly to a disordered file management, including therisk to inadvertently overwrite some of your files.

2.2

Using the coupled hydrodynamic files generated in exercise 1.4, determine whether materialdumped just outside the estuary will enter the estuary or if it will be transported along thecoast. You can use the substance file <tracer 3.sub>.

Question:

� How would you address this issue?

Tips:

� You may want to define several locations for dumping the substance.� You can use numerical scheme number 16.

Deltares 204

Contents

Wad

den

sea

North sea

The Netherlands

Deltares 205

Contents


3 Numerical aspects of water quality modelling

3.1

We will use the 2D hydrodynamic database of the Bornrif area. Compare the results of a 1×1aggregated hydrodynamic database with the results of a 4×4 aggregated database. You willneed to prepare both (see exercise 1). Use numerical scheme 16.

Tips:

� Choose some monitoring points in the whole modelled area.� Once you have set up the 1×1 simulation, you can use this scenario file as a basis for the

4×4 simulation. To do so, open the scenario file, load the 4×4 hydrodynamic databaseand save the scenario under a new name. The user-interface takes care of all waste loads,boundary conditions, monitoring points, etc.

3.2

Release a tracer at location A in the modelled area at 100 g/s, using the 1×1 aggregated grid.Compare the results of a simulation with numerical scheme 16 (implicit) with a simulation withnumerical scheme 5 (explicit).

Wad

den

sea

North sea

The Netherlands

A

Deltares 206

Contents

3.3

Repeat the exercise with the 4×4 aggregated grid. Compare the results with the simulationsfrom the previous exercise.

Tips:

� Choose some monitoring points spread over the whole area but also in the vicinity of therelease location.

� Set up water quality scenario (<∗.scn> file)� Load the substance file in WAQ-GUI: <age 3.sub>� Add concentration cTR1 and dTR1 of 100 mg/l at the discharge at location A.� Uniform dispersion of 5 m2/s� Define 2 monitoring stations, one in the direct vicinity of the discharge and one more off

shore� Set monitoring timers

3.4

Carry out the following computations based on exercise 3.2, they differ in the numerical op-tions:

� Allow dispersion with zero flow and allow dispersion over open boundaries� Do not allow dispersion with zero flow, but do allow dispersion over open boundaries

Questions:

� Are there any differences? (Look at the mass balance reported in the monitoring file)� Use QUICKPLOT or GPP to make contour and time series plots and compare the results.

Deltares 207

Contents


4 Modelling water quality E.Coli

4.1

The sewage discharge at the Palmar outfall in Lake Maracaibo releases waste water with aconcentration of Ecoli bacteria of 1012 MPN/100ml. What is the E.Coli concentration at beachlocation (200250, 1142450) at 15 Oct. 1992 12:30 in Lake Maracaibo? How does the mortalitydue to radiation contribute to the total mortality of Ecoli?

Tips:

� The communication files can be found in <course models/hydro/maracaibo>� Create a substance file containing EColi as a state variable and mortality and extinction

uv-light as processes. Take IM1 and Temp editable. Select mortality rates as output.� Assume that all others discharge are zero.� Take a uniform IM1 concentration of 25 mg/l, a UV radiation of 25 W/m2 and a temperature

of 298 K. Set the uniform dispersion to 10 m2/s, the integration option to 5 and a time stepof 15 min.

4.2

Extend exercise 4.1 with the modelling of suspended matter.

Tips:

� Open the previously created EColi substance file and add inorganic suspended matter(IM1) in water and in sediment (IM1S1) as substances.

� Activate the processes sedimentation and resuspension and activate the calculation ofshear stress (Tau).

� Set settling velocity to editable� Adapt the previously created scenario file by selecting the newly made substance file.� Set IM1 at boundary at 2 mg/l and at the rivers at 50 mg/l. The settling velocity is 0.05

m/h.

Deltares 208

Contents


5 Dredging and dumping of material in a channel

Problem description

The government of Hong Kong Special Administrative Region intends to perform maintenancedredging a major shipping channel. During dredging and dumping activities, material will bespilled in the water column and suspended sediment concentrations will increase. Further-more, the organic material in the sediment will be released in the water column as well. Thedecay (mineralisation) of the organic matter will result in a decrease of the oxygen concentra-tions.

Apart from shipping, tourism and fishing are the main economic activities in the area. The wa-ter quality at the nearby beaches should not be adversely affected by the dredging activities.The water quality objectives for swimming water have to be met. Furthermore, fish stocks maynot be adversely affected by the dredging.

Question to the consultant

Using the hydrodynamic model of the area already available, you should determine the impactof the dredging and dumping activities. A number of scenarios are defined. Please determineif the scenarios fulfill the water quality objectives given below:

� Suspended matter → the suspended matter concentration at the beaches should notincrease more than 30% above the ambient background concentration.

� Oxygen→ the oxygen concentration in the surface layer should not be below 4 mg/l. theoxygen concentration in the bottom layer should not be below 2 mg/l.

Characteristics of the area are:

1 Water temperature 25 ◦C2 Salinity 30 g/kg3 Wind speed 5 m/s

Sediments samples were collected and analysed. Also, literature was reviewed to supplyaverage values for the area:

1 background concentration of suspended matter 5 mg/l %2 fraction organic matter 2.5 weight %3 Decay rate of organic matter 0.15 1/d4 Settling velocity of organic matter 1 m/d

The bottom material in the channel is largely fine-grained. The size distribution of the sedimentsamples is determined as follows:

� Sand fraction 70 %� Silt fraction 25 % (settling velocity 0.05 mm/s, τ for sedimentation 0.26 Pa)� Clay fraction 5% (settling velocity 0.01 mm/s, τ for sedimentation 0.1 Pa)

Deltares 209

Contents

The sand fraction settles immediately and is not subject to erosion, unless by storm events.The other two fractions have the settling velocities and critical shear stresses as specified.The critical shear stress for resuspension is 0.4 Pa. The resuspension rate is 30 g/m2 perday.

The table below specifies the dredging and dumping discharges to be considered. Both dredg-ing and dumping are assumed to occur continuously and simultaneously.

Dredging Dumping 1 Dumping 21000 ton/d 500 ton/d 500 ton/d

During dredging a spill of 50 ton/d is assumed at the dredging location. The dredging anddumping sites are indicated in the figure below. Beaches and other sensitive receivers areshown in the attached figure as well.

Hints:

a Define the problem and make a plan how to use the model to get an answerb Couple the hydrodynamic file <far field/hydro/d3d>c Which substances and processes do you include in you substance file?d Set up substance filee Set up input file, including monitoring points at sensitive receiversf Run modelg Check output:

� continuity� sediment concentrations� sedimentation rates� oxygen concentrations

A Dredging siteB Dump site 1C Dumpsite 2

S1 - S8Sensitivity receivers

S1

S2S5

S6

S7

S4

S8

S3

AB

C

Deltares 210

Contents


6 Environmental Impact Assessment Western Scheldt

Situation

The Western Scheldt is situated in the south of the Netherlands and forms the seaward entryto Antwerp in Belgium. The estuary is therefore frequented by large ships and also in theDutch part there is plenty of industrial activity.

The region happens to be important for various species of birds and the common seal. As aconsequence of the rich environment, tourism is another major economic activity.

Problem description

In the past years, both tourism and the economical activity in the area has increased enor-mously. Development of heavy industrial activities and expansion of the larger cities haveresulted in an increased impact on the local environmental and adjacent marine system. Theseaport at Flushings has recently been expanded with a large terminal and several tran-shipment docks. In order to keep the Antwerp harbour attainable for heavy container ships,significant dredging of the estuary is foreseen. According to estimates of the local dredgingcompany more than 15 million m3 of sediment has to be dredged, annually. Analysis of thesediment reveals that the sediment is heavily contaminated with cadmium and is thereforecharacterised as chemical disposal which may not be stored on land. Since high concen-tration of cadmium are lethal to the blue seal, the local Environmental Authority is seriouslyconcerned about the welfare of the common seal.

Due to the expected number of 300.000 inhabitants of Flushings in the near future, the do-mestic waste water discharge is expected to be a problem concerning the bacterial pollution inthe area. Especially, the Ministry of Tourist affairs is concerned about the water quality alongthe bathing beaches.

Environmental impact assessment

Water quality specialists are asked by the Environmental and the Tourist Ministry to make anenvironmental impact assessment study. The following requirements are set by the Ministries:

� Cadmium concentration in the entire area may never exceed the target value of 0.8 mg/kg-DW;

� The E.Coli concentration may never exceed 1×104 MPN/100ml along the bathing beaches;� operational costs must be kept to a minimal.

The corporation has to define/analyse and report several scenarios concerning the wastewater treatment efficiency and the dumping locations related on the specified requirements.

Deltares 211

Contents

Assumptions

Cost estimates:

� operational dredging, shipping and replacement costs: 20000 $ /mile;� waste water treatment costs:

� efficiency 50%: 10000 $/yr� efficiency 90%: 50000 $/yr� efficiency 98%: 250000 $/yr

� quality of dredged material: 1.8 µg Cd/gDW� density of sediment: 2.5 ton/m3

� porosity on volume basis: 0.75� fraction of sand: 0.45� settling velocity: 0.2 m/d� partitioning coefficient Cd: 50 l /g� solar radiation: 200 W/m2

� domestic discharge 200 l /day · inhabitant� concentration E.Coli in influent of treatment plant 1×1012 MPN/100ml� location of wastewater outfall (x, y) = (36312, 387231)� location of bathing beaches around (39404, 381524) and (28941, 384734)

Hints

� estimate amount (load) of dry weight fine sediment based upon porosity, sand percentageand sediment density;

� estimate load of the Cd dumping� create substance file containing IM1, Cd and Ecoli� define monitoring locations along the entire estuary and near the bathing beaches� use the Western Scheldt 2DH tidal hydrodynamics (see first exercise)� create matrix of scenarios with risk-index (i.e. exceedance factor PEC/target)

treatment efficiency(%)

Risk-indexcommon seal

Risk-indexbathing beach

Cost estimate($)

dumping location 1dumping location 2. . . . . .

5095

Deltares 212

Contents

Deltares 213

Contents

Bract, odd page

Deltares 214

Contents

Bract, even page

Deltares 215

Contents

Exercises introduction course Environmental,mid-field water quality

Exercise 1: Introduction

A factory discharges 2 m3/s into the sea. The concentration of a certain substance in theoutfall is 10 mg/l.

� The number of particles to be used in the simulation depends on a number of things.Name three of them.

� The zoom grid has a resolution of 100 m × 100 m and the average water depth is 10 m.In a 15 day 2D simulation, how many particles are needed for a resolution of 0.1 mg/l. Isthis a reasonable resolution? What can you say about the concentration levels in relationto this resolution?

� What if the simulation is 3D with 5 equidistant layers?

Tip:

� Assume that the sea is well-mixed.

Deltares 216

Contents

Answers� Minimum concentration you want to resolve = f(mass per particle, volume control volume

(e.g. zoom grid) and the runtime.Remark: aim at at least 5–10 particles per grid cell to get a nice representation. If 1particle/cell then the randomness starts become important.

� Maximum resolution you can resolve is if only 1 particle is present in 1 zoom grid cell.Required solution:

V = 100m · 100m · 10m = 100000m3

c =M

V→ 0.1

mg

l= 0.1

g

m3=

M

100000m3

M = 10000g mass of 1 particle.

Amount discharged:

2m3

s· 10

g

m3= 20

g

s

20g

s· 15 · 86400s = 25 920 000g mass of all particles

Nr of required particles:

25 920 000g

10 000g= 2 592

� In 3D; with 5 equidistant layers Nr of required particles:

5 · 2 592 = 12 960

Required solution:

V = 100m · 100m · 2m = 20 000m3

c =M

V→ 0.1

mg

l= 0.1

g

m3=

M

20 000m3

M = 2 000g mass of 1 particle.

Amount discharged:

2m3

s· 10

g

m3= 20

g

s

20g

s· 15 · 86 400s = 25 920 000g mass of all particles

Nr of required particles:

25 920 000g

2 000g= 12 960

So, increased resolution in vertical, requires a larger number of particles to resolve thesame resolution of 0.1 mg/l.

� Reduction in grid resolution→ reduction in accuracy of calculation.� Reduction in grid resolution with constant number of particles results in a reduction of

resolution.

Deltares 217

Contents


Exercise 2:

2.1

Set-up a non-decayable tracer run with a continuous release of 1000 particles at the watersurface and no horizontal and vertical dispersion.

Background info:

� The hydrodynamic database can be found at <course models/mid field/hydro/fk01>Hydrodynamic FLOW simulation details:

� 2D basin of dimensions: l · w · d = 67.5 km · 40.5 km · 10 m� The water is flowing from west to east with an average speed of ≈ 0.5 m/s.� Coupled period: 30 11 2005 00 00 00 - 01 12 2005 00 00 00

(time step: 12 hrs)

� PART simulation period is 4 days with a timestep of 1 hour.� Release details:

� Location:x = 86477.3 m; y = 480609 m (no initial spreading around the release location).

� Duration:All particles are released within the first 24 hours with a nett release rate of 1 kg/s.

� Particles are only advected by flow:

� No decay.� No sedimentation/erosion processes (no settling of particles).

� No wind and a roughness length of 0.02 m; density of water 1024 kg/m3

� Output:

� Time series every hour.� Spatial plots: every 12 hours.� 3D tracks of the particles.� Zoom grid: cell size of 1000 × 1000 m covering the majority of the plume and output

every hour.

2.2� � Visualize the particles tracks on the hydrodynamic grid by using QUICKPLOT.

Tip:

� Get the hydrodynamic grid from the <∗.map>.� Use the <trk-∗.dat> (of type Delft3D output file) for the particle tracks.� Clip all zeros.� Use markers

Deltares 218

Contents

Answers

2.1

Screenshots including explanation. Only treat the “problem sheets”. Don’t go through allscreens (overkill).

Indicate: <∗.inp> and the <∗.mdp>.

Explain: <∗.out>

2.21 Aggregated grid:

� cont-release00.map (Files of type: Delwaq binary file);� <∗.lga> from the coupled hydrodynamics: <course models/mid field/hydro/fk01>

2 Particles tracks:Open the <trk-∗.dat> (Files of type: Delft3D output file)

� Select tracer and all time steps.� Presentation type is marker circle; colour is blue and fill is green� Clipping Values is 0

ADD TO PLOT3 Results in <course models/mid filed/scen/cont-release00/Particle track>

Notice that:

� All particles follow the same track: horizontal dispersion is set to 0.� Particles which are transported outside the model domain stick at the boundary

Deltares 219

Contents


Exercise 3: (Horizontal) dispersion

Set the horizontal dispersion to 100 (a = 100) and visualize the particles tracks on the hydro-dynamic grid by using QUICKPLOT.

Tip:

� Create a directory under <course models/mid field/part/scen/cont-release01>.� Copy and rename <∗.inp> and <ast.mdp> from the previous exercise at<course models/mid field/part/scen/cont-release01>.

� Load the <∗.mdp> in the user interface.

Answers

Same as 2.2 but now for release: cont-release01.

Deltares 220

Contents


Exercise 4: Zoom grid/output1 Visualize the results of the simulation as concentrations on the hydrodynamic grid by using

QUICKPLOT.Tip:

� Clip 0 values

2 Visualize the results of the simulation as concentrations on the zoom grid by using QUICK-PLOT.Tip:

� Plot the hydrodynamic grid first.

3 Explain the difference.

Answers1 Plot tracer from <∗.map>. Clip 0 values.2 Plot aggregated grid from <∗.map>

Plot tracer from <∗.plo>. Clip 0 values.3 . . . .

Deltares 221

Contents


Exercise 5: Number of particles

Increase the number of particles from 1000 to 100 000 and visualize the results of the simu-lation as concentrations on the zoom grid by using QUICKPLOT.

Tip:

� Create a directory under <course models/mid field/part/scen/cont-release02>.� Copy and rename <∗.inp> and <∗.mdp> from the previous exercise at<course models/mid field/part/scen/cont-release02>

Answers

Plot aggregated grid from <∗.map>

Plot tracer from <∗.plo>. Clip 0 values.

ADD TO PLOT

Deltares 222

Contents


Exercise 6: Number of particles

6.1

Use QUICKPLOT to plot time series of tracer concentrations for observation point Loc2 for:

1 the simulation without dispersion (cont-release00)2 the simulation with dispersion (cont-release01)3 the simulation with dispersion and increased number of particles (cont-release02).

6.2

Why does the tracer concentration for the first simulation remain zero?

Tip:

� Plot all 3 time series in one plot.

Answers: 6.2

Location of the observation point (Loc2): (104420, 479712)

Location of release location: (86477, 480609)

The locations seem to be similar because they are indicated in the same hydrodynamic gridcell, but they are not.

Deltares 223

Contents


Exercise 7: Instantaneous release (optional)

7.1

Release the same amount of particles and the same amount of mass as in cont-release00only now in one go (instantaneous). Use for the remainder the same details as in exercise 2.Release at the start of the simulation. Visualize the particles tracks on the hydrodynamic gridby using QUICKPLOT.

Tip:

� Create a directory under <course models/mid field/part/scen/inst-release00>.� Copy and rename <∗.inp> and <∗.mdp> from the previous exercise at<course models/mid field/part/scen/cont-release00>.

7.2

Set the horizontal dispersion to 100 (a = 100) and visualize the particles tracks on the hydro-dynamic grid by using QUICKPLOT.

Tip:

� Create a directory under <course models/mid field/part/scen/inst-release01>.

Answers

7.1 & 7.2 Same as 2.2 but now for instantaneous release: inst-release00 & 01. See alsoanimation inst-release00 vs 01

Deltares 224

Contents


Exercise 8: Plume dispersion (optional)

An old sewage treatment plant is located at the tip of an island (see figure below). Its dischargepipe stretches a few kilometres into the sea. Still, fears exist that the plume may affect thenearby nature reserve.

Use the particle tracking model to demonstrate whether or not the discharge plume is a po-tential threat to the nature reserve.

What if a decay rate of 0.2 d−1 can be assumed for the pollution; what will be the decrease inconcentration at the nature reserve?

Answers

The hydrodynamics results have not been converted by the coupling program yet. You willhave to do that yourself first.

Ameland

Schiermonnikoog

Waddensea

North sea

Sewage plantwith submerged outfall

nature reserve

Tip:

� The hydrodynamic database can be found at <course models/mid field/flow/fti sewage>

Deltares 225

Date post:	17-Jul-2016
Category:	Documents
Upload:	tcp-vn
View:	56 times
Download:	11 times

Delft 3D Environmental

Documents