+ All Categories
Home > Documents > N NT F TH PRFRN DNTTN PND b - Open...

N NT F TH PRFRN DNTTN PND b - Open...

Date post: 04-Jun-2018
Category:
Upload: lykhanh
View: 231 times
Download: 0 times
Share this document with a friend
151
An assessment of the performance of federally regulated sedimentation ponds Item Type Thesis-Reproduction (electronic); text Authors Vandivere, William Benton. Publisher The University of Arizona. Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author. Download date 05/07/2018 10:26:35 Link to Item http://hdl.handle.net/10150/191711
Transcript
Page 1: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

An assessment of the performance offederally regulated sedimentation ponds

Item Type Thesis-Reproduction (electronic); text

Authors Vandivere, William Benton.

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this materialis made possible by the University Libraries, University of Arizona.Further transmission, reproduction or presentation (such aspublic display or performance) of protected items is prohibitedexcept with permission of the author.

Download date 05/07/2018 10:26:35

Link to Item http://hdl.handle.net/10150/191711

Page 2: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

AN ASSESSMENT OF THE PERFORMANCE

OF FEDERALLY REGULATED

SEDIMENTATION PONDS

by

William Benton Vandivere

A Thesis Submitted to the Faculty of the

SCHOOL OF RENEWABLE NATURAL RESOURCES

In Partial Fulfillment of the RequirementsFor the Degree of

MASTER OF SCIENCEWITH A MAJOR IN WATERSHED MANAGEMENT

In the Graduate College

THE UNIVERSITY OF ARIZONA

1980

Page 3: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

gement

Assistant Profess .r of Hydrologyand Wate fesoAices

LOUIS . HEKMANAssistant Prof ssor of Renewable

Natur Resources

STATEMENT BY AUTHOR

This thesis has been submitted in partial fulfillment of re-quirements for an advanced degree at The University of Arizona and isdeposited in the University Library to be made available to borrowersunder rules of the Library.

Brief quotations from this thesis are allowable without specialpermission, provided that accurate acknowledgment of source is made.Requests for permission for extended quotation from or reproduction ofthis manuscript in whole or in part may be granted by the head of themajor department or the Dean of the Graduate College when in his judg-ment the proposed use of the material is in the interests of scholar-ship. In all other instances, however, permission must be obtainedfrom the author.

SIGNED:

APPROVAL BY THESIS COMMITTEE

This thesis has been approved on the date shown below:

, - 27 .

MARTIN MAI,U FOGEL

il/A-Date

GOJIJI-1(

X/C6/ 3 (t- ODONALD ROSS DAVIS Date'

Date

Page 4: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

ACKNOWLEDGMENTS

The author is greatly indebted to his principal advisor, Dr.

Martin Fogel, for providing the opportunity to develop and expand his

intellectual facilities. His avowed skepticism was at once refreshing

and realistic and will continue to endear him to his students. Addi-

tional thanks must be directed toward Dr. Louis Hekman and Dr. Donald

Davis, whose patience and guidance were well appreciated. The periodic

assistance from Dr. John Thames also contributed to the author's pur-

suit of a practical understanding of the field of hydrology.

Best wishes and sincere affection are ext4nded to friends and

colleagues of the author during his brief tenure with the School of

Renewable Natural Resources. A genuine comaraderie existed, and hope-

fully will continue between us, expecially Steve Blake, Jeff Franklin,

and Todd Rasmussen. Their input was always informative and much valued.

A special heartfelt thanks goes to Ms. Paula-Ann Cech who faith-

fully supported the author through good times and bad and without whose

help both his entire graduate and a significant part of his life expe-

rience would have been left unfulfulled.

Appreciation is expressed for the timely work done by Phyllis

Miller in preparing the final manuscript.

Also, the author wishes to acknowledge the aid received from

Andy Ward of the Agricultural Engineering Department of the University

of Kentucky who supplied access to the DEPOSITS sedimentation model.

Page 5: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

iv

Portions of this study were carried out under grant

#14-34-0001-9056 from the Office of Water Resources Technology

entitled "The Role of Hydrologic Variability in Complying with Regu-

latory Enforcement Standards for the Rehabilitation of Surface-mined

Coal Lands." The author wishes to thank the taxpayers of this country

for this funding.

Page 6: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

TABLE OF CONTENTS

Page

LIST OF ILLUSTRATIONS vii

LIST OF TABLES viii

ABSTRACT ix

1. INTRODUCTION 1

2. REVIEW OF LITERATURE AND PERTINENTREGULATORY STATUTES 4

Sedimentation Ponds and Manipulated Environments • • 4Animal Feedlots 5Construction Sites 7Surface Mining Operations 9

Uncertainty in the Sedimentation Process 13Applicable Regulatory Statutes 16

3. SITE CHARACTERISTICS AND FORMULATION OF STUDY 21

Hypothetical Watershed 21Development of Hydrologic Inputs 22Model Components and Operation:

Program INFLUX 25Precipitation 25Infiltration 27Runoff 34Sediment 36

Sedimentation Pond Design 38DEPOSITS Sedimentation Model 43

4. RESULTS AND DISCUSSION 49

Pond Sensitivity Analysis 49National Soil Material with

Untreated Pond Inflow 57Hydrologic Uncertainty and Implications

for Pond Performance 66Model Adaptability and Regional Bias 70

Page 7: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

vi

TABLE OF CONTENTS, Continued.

Page

5. CONCLUSIONS AND RECOMMENDATIONS 72

APPENDIX A: SEDIMENTATION POND DESIGN 77

APPENDIX B: INFLUX VARIABLE DESCRIPTION 109

APPENDIX C: PROGRAM LISTING OF INFLUX 112

APPENDIX D: PROGRAM AND INPUT LISTING OF DEPOSITS 116

LIST OF REFERENCES 137

Page 8: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

LIST OF ILLUSTRATIONS

Figure Page

1. SCS Type I and II 24-hr. rainfall distribution 26

2. Data points and infiltration curve for 0.5 year oldspoil material, J-3 experimental area 29

3. Generalized flowchart for hydrologic linkages . . 31

4. SCS triangular hydrograph 37

5. Schematic diagram of sedimentation pond anddewatering device 44

6. Plug flow routing for DEPOSITS sedimentationmodel 47

7. Predicted peak effluent concentrations and contoursin mg/2 for Black Mesa minspoil with altered inputsand untreated pond inflow

58

vii

Page 9: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

LIST OF TABLES

Table Page

1. Point precipitation-frequency values for Black MesaMine, Arizona, in inches 24

2. Discretization of Type II rainfall distribution for24 hr. duration 28

3. Discretization of J-3 infiltration capacity curve 32

4. Precipitation characteristics and predicted peakeffluent sediment concentrations for routedBlack Mesa storms, 5-10 yr. return periods:minespoil with untreated pond inflow 51

5. Precipitation characteristics and predicted peak effluentsediment concentrations for routed Black Mesa storms,12-25 yr. return periods: minespoil with untreatedpond inflow 53

6. Particle size distributions for undisturbed experimentalwaterhseds: Black Mesa, AZ 59

7. Predicted effluent sediment concentrations for routedBlack Mesa storms: natural soil material withuntreated pond inflow 61

8. Predicted peak effluent sediment concentrations forrouted, selected Black Mesa storms: minespoil withchemical treatment of inflow 63

9. Predicted event sedimPnt yield for simulated conditionsat Black Mesa, AZ: 5-10 yr. return periods 64

10. Predicted event sediment yield for simulated conditionsat Black Mesa, AZ: 12-25 yr. return periods 65

A.1 Size fraction distributions for sediment production:J-3 experimental watershed, Black Mesa Mine 84

A.2 Rating relations for final pond design 88

A.3 Stage-discharge relations for final pond design 91

viii

Page 10: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

ABSTRACT

A study was undertaken to evaluate the performance of federally

regulated sedimentation ponds, used in conjunction with surface mining

operations in the semi-arid southwest. Emphasis was placed on the assess-

ment of pond performance under conditions of hydrologic uncertainty rep-

resented by precipitation inputs of varying frequencies and durations.

A hypothetical watershed with characteristics common to the study area

functioned as the medium for surface water flux to the detention facility.

Pond design was based on accepted hydrologic and engineering procedure

and concurred with published federal reclamation statutes. Computer pro-

grams were utilized to model both the temporal characteristics of south-

western convective rainfall and the generation of water and sediment in-

flaws resulting from the application of storms over the watershed. A

previously developed sedimentation routine was then used to determine

effluent sediment concentrations corresponding to the modeled events.

Three watershed-pond conditions were investigated to assess the efficacy

of the sedimentation pond in meeting effluent quality standards. Results

indicated that poor pond performance ensured unless chemical treatment

was maintained. Since variations in precipitation intensity influenced

predicted pond performance, it was recommended that hydrologic uncertain-

ty be considered in the drafting of regional reclamation statutes.

ix

Page 11: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

CHAPTER 1

INTRODUCTION

As natural phenomena the fluvial processes of erosion and

sedimendation proceed at a rate determined by prevailing hydrometeoro-

logical and geologic conditions. A quasi-equilibrium is established

over time which affects a balance between hillslope development and

stream regimen. Man's intervention, however, often results in short

term distortions in this delicately adjusting mechanism.

The national addiction to non-renewable fossil fuels has led to

a recent expansion in possibly the most environmentally disruptive of

human activities; the surface mining of coal. An immediate consequence

of this disruption is an acceleration in the rates of erosion and

sedimentation observed in the effected areas. Extensive soil loss from

surface mine sites inhibits generation of protective vegetative cover

and befouls area stream flow, threatening indigenous aquatic life.

In an effort to mitigate the environmental damage wrought by

mining operations, the 95th Congress enacted preventative measures

embodied in the Surface Mining Control and Reclamation Act of 1977.

The Act expressed the intent to assure reclamation of mined land while

not unduly burdening mine operators. Of course, the high priority given

continued access to the coal resource was sustained.

The federal statutes represent baseline criteria for acceptable

mining and reclamation procedures. States have been presented with the

1

Page 12: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

2

option of developing their own standards, as long as they are at least

as stringent as those outlined by the Act. In addition, states have

been encouraged to consider regional variabilities in the drafting of

their federal counterparts.

A permanent regulatory program has been published by the Office

of Surface Mining (OSM) for purposes of setting standards for mining and

reclamation practices which are consistent with the legislative intent

of Congress. Considerable attention was paid to the maintenance of the

hydrologic balance in and adjacent to areas subjected to mining disturb-

ances. Central to this concept was the desire to minimize changes in

water quality and quantity, drainage patterns,and groundwater systems.

Consideration of feasible alternatives convinced OSM that the use of

sedimentation ponds in conjunction with other sediment control measures

provided the best available technology for removing suspended solids

from mine site runoff.

In the semi-arid regions of the western U.S., extreme variability

in the occurrence and nature of precipitation introduces an element of

uncertainty into the design process for sedimentation ponds. Since

assessment of the effects of regional hydrologic and climatic uncertainty

on reclamation efforts has been left unaddressed by federal statutes, a

need has arisen for evaluation of the uncertainties and their implica-

tions for the design and expected performance of mandated sedimentation

ponds.

The present study attempts to aid in the appraisal of the

effects of hydrologic uncertainty on the performance of federally

regulated sedimentation ponds. Specifically, its purpose is threefold:

Page 13: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

3

(1) Compile a model capable of simulating the performance of

sedimentation ponds designed in accordance with federal

statutes and generally accepted engineering-hydrologic

practice.

(2) Assess the effects of precipitation uncertainty, in particular,

the influence of variable rainfall intensity and rainfall-

runoff volume on sediment yields from minespoil watersheds

and on the effluent quality of event based discharges from

sedimentation ponds functioning on these watersheds.

(3) Evaluate current federal performance criteria for sedimentation

ponds functioning in a semi-arid environment where precipita-

tion influx is limited and predominantly convective in char-

acter.

Page 14: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

CHAPTER 2

REVIEW OF LITERATURE ANDPERTINENT REGULATORY STATUTES

The following section is devoted to a review of research and

legislation currently influencing the design and evaluation of the

performance of sedimentation ponds. Initially, studies dealing with the

application of sedimentation ponds to the problem of alleviating pollut-

ant migration under disturbed conditions are summarized. A brief

examination of work aimed at delineating the effects of uncertainty in

the sedimentation process follows. Concluding the chapter is a condensed

review of recently promulgated federal statutes especially relevant to

the present investigation.

Sedimentation Ponds and Manipulated Environments

Literature pertaining to the design and performance of sedimenta-

tion ponds was rather sparse until the advent of federal water quality

legislation in the early nineteen seventies. In response to newly en-

acted state and federal guidelines for pond design, researchers concerned

themselves with the evaluation of ponds, existing or hypothetical,

which conformed to applicable regulatory statutes. Relevant studies

have focused on the utilization of detention basins for purposes of

mitigating environmental degredation in three disturbed settings:

4

Page 15: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

5

(1) animal feedlots; (2) construction sites; and (3) surface mining

operations. The requirement for detention facilities at feedlot loca-

tions stems from the desire to limit discharge of organic waste products

as well as nutrient and sediment-laden water into receiving streams.

Discharge from areas disturbed by construction and surface mining, while

also potentially deleterious to existing chemical balance in streams,

is primarily undesirable from the standpoint of the intensive stream

sediment loading it engenders.

Animal Feedlots

Due to the objectionable nature of runoff from animal feedlot

areas, detention facilities are often designed to preclude any discharge

of runoff generated by the 10-yr, 24-hr. or 25-yr., 24-hr. rainfall

event. Wensink and Miner (1975), using simulated rainfall and tempera-

ture inputs for Oregon feedlot sites, analyzed the performance of a

hypothetical detention basin. Dewatering of stored runoff was accom-

plished solely by pumpage in response to irrigation demand which was

dependent upon antecedent temperature and moisture conditions. Reservoir

storage volume was determined using two design methods. A "retention

return period" method, employing the SCS rainfall-runoff relations and

rainfall frequency data published by the National Oceanic and Atmo-

spheric Administration (NOAA) was compared with a "sufficient design"

technique. The latter method sized the facility on the basis of total

storage of all runoff resulting from storms falling within the bounds

of the design event. The authors concluded that for most cases the

Page 16: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

6

"return period" design technique produced either insufficient storage or

resulted in unreasonably expensive basins. The "sufficient design

technique, on the other hand, was found to minimize the required pond

volume for the appropriate pumping rate while more adequately satisfying

environmental protection standards.

A similar investigation (Koelliker, Manges, and Lipper, 1975)

carried out in Kansas examined the effect of regional precipitation

variability on detention basin response. Evaporation, as well as

irrigation pumpage, was incorporated into the analysis. Basin perfor-

mance as measured by frequency of overflow was considerably worse in

regions experiencing lengthly periods of persistent rainfall. Drier

regions generally subject to single rainfall events with longer inter-

arrival times created fewer basin overflows. Consideration of chronic

wet periods was, therefore, viewed as the critical factor in the sizing

of detention facilities. It should also be noted that most discharges

resulted from storms substantially less than the design event.

Hydrologic conditions typical of North Carolina dairy feedlots

were modeled by Overcash and Phillips (1978) to evaluate established

guidelines for the animal production industry. Linear and non-linear

rainfall-runoff models were applied to assess the efficiency of using

the mandated 25-yr., 24-hr. storm for retention basin design. At

representative SCS curve number values for specific feedlot locations,

the rainfall magnitude at which incremental rainfall produced a

correspondingly high (95%) runoff response was determined. This

quantity was found to correlate closely with that corresponding to the

Page 17: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

7

appropriate 25-yr., 24-hr. rainfall value extracted from records of the

U.S. Weather Service and HISRAS (Hydrologic Information and Retrieval

System). On this basis, the authors concluded that the 25-yr., 24-hr.

storm appeared justifiable from a basin design viewpoint.

Construction Sites

Construction and urbanization denude substantial areas of land

exposing it to the heightened erosive capabilities of rainfall. Rapid

deterioration of water bodies adjoining these disturbed areas has ne-

cessitated regulatory controls directed at easing the impact of

accelerated sediment production on water quality. Because of the

extensive nature of the problem, a multitude of field evaluation and

model studies have ensued. Noteworthy, is the development of a sediment

discharge model (Curtis, 1976) describing water and sediment transport

in urbanizing areas. Model results revealed a tendency for both peak

sediment discharge and total volume of sediment discharge to increase

with increasing rainfall intensities.

The use of sedimentation ponds for reducing sediment concentra-

tions in construction site runoff and the turbidity of receiving streams

has become widespread. Oscanyon (1975) introduced a set of design

criteria for the design of sediment basins on construction sites in

Maryland. It was assumed that even a well designed and adequately main-

tained structure would remove no smaller than .005 mm diameter sediment.

Alternative on-site measures were cited as offering a greater margin of

sediment control where higher percentages of clay are contained in pond

Page 18: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

8

influent. An examination of the relative merits of in-stream and off-

stream sedimentation ponds was undertaken by Reed (1975). Both types

of ponds performed equally well in removing sediment, with reductions

of 80 percent for most storms. The in-stream pond, however, sustained

higher mean turbidity levels for longer periods of time than its off-

stream bounterpart. Curtis and McCuen (1977) derived a mathematical

model of a detention basin coupled with a watershed hydrologic model

for assessment of basin hydraulic efficiency. Trap efficiencies for

the basin were found to increase with decreasing proportions of smaller,

lighter particles in the inflow. In addition, both decreased initial

storage and smaller orifice diameters for perforated risers increased

modeled trap efficiencies. Decreased basin depth along with a concom-

mitant increase in surface area also produced increased trap efficien-

cies. The effect of basin depth and area on performance was reaffirmed

in a study by Bondurant, Brockway, and Brow (1975). Specifically, the

authors recommended that in order to achieve a reduction in forward

velocity and depth of settling, the design would have to include

(1) adequate sediment storage volume, (2) decreasing flow depth towards

the outlet, and (3) a means for reducing entrance velocities to the

pond. Higher removal efficiencies were obtained at higher flow rates

due to the greater relative proportion of finer particles transported

at low flow rates.

Page 19: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

9

Surface Mining Operations

Surface mining is one of the single most devastating operations

practiced by man upon his environment. Studies involving the application

of sedimentation ponds to mining operations and subsequent reclamation

efforts have again tended towards either the development of acceptable

guidelines or the appraisal of previously enacted legislation referring

to pond design and performance. Curtis (1974) conducted a study to

determine sediment production from mined areas and to propose criteria

for calculating detention basin storage volume. The first six months

following the termination of mining was indicated as the most critical

period for sediment production. Major factors contributing to sediment

yield were deduced to be methods of mining and handling of overburden

and rapid establishment of vegetative cover. The effectiveness of on-

site sediment control measures coupled with an in-stream sedimentation

pond was analyzed by White and Plass (1974) for a mining operation in

West Virginia. It was noted that pond removal efficiency was greatest

for low intensity stormllow.

A review of sedimentation mechanics and earlier methods for

determination of sediment basin trap efficiency was presented by Haan

and Barfield (1978). Among the methods examined were that of the EPA

(1976) and the DEPOSITS sedimentation model (Ward, Haan, and

Barfield 1977a). The EPA (1976) methodology, a derivative of the prior

work of Camp (1945), was described as plausible for steady state flows,

but inadequate in its representation of semi-dry basin performance.

Because of its superior capabilities for handling typical field

Page 20: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

10

conditions, the DEPOSITS model (Ward et al., 1977a) was favored by the

authors for more accurately depicting actual basin functioning. Another

conclusion reached by Haan and Barfield (1978) was that given identical

outflow riser configurations, a basin containing a permanent pool ca-

pacity will produce higher quality effluent than one lacking permanent

storage. This was attributed in part to the lessened probability that

resuspension of deposited sediments would occur.

The construction of the DEPOSITS model is described in detail

by Ward, Haan, and Barfield (1977b). Model verification was completed

using data published in a report by the EPA by Hittman Associates, Inc.

(1976a). The authors of the DEPOSITS formulation expressed misgivings

over the methods used in the EPA study. In particular, data collection

techniques were deemed unacceptable and the method for determining actual

basin performance was questioned. The equation developed for determina-

tion of trap efficiency assumed instantaneous flow through the basin.

This assumption neglected the effect of varying detention times for the

different portions of throughflow and depended entirely on simultaneous

readings of influent and effluent sediment concentrations over a short

time period. Since the DEPOSITS model accounted for varying flow rates

and reservoir detention times, the authors felt that it embodied a more

realistic conception of the actual sedimentation process and was, there-

fore, of greater value as a tool for evaluation of pond performance.

The ground breaking investigation on the performance of sedi-

mentation ponds at eastern mining sites by Kathuria, Nawrocki, and Beck-

er for the EPA (1976a) contributed much to the recent discussion on

Page 21: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

11

performance standards prompted by OSM. In-field evaluation of nine

functioning ponds in W. Virginia, Kentucky and Pensylvania was conducted

in an attempt to determine trap efficiencies and to identify character-

istics influencing pond behavior. Sampling was carried out during both

baseline and rainfall operating conditions. Theoretical removal effi-

ciency computed by means of Ideal Settling Theory was compared to a

measure of actual removal efficiency expressed as:

106

R(% solids removed) = 1cl

X 100 (1)

610C2

where C1

is the concentration of suspended solids in the influent in

mg/1 and C 2 is the concentration of suspended solids in the effluent in

mg/l. Poor maintainance and lack of conformance to approved design

plans were cited as major factors inhibiting attainment of desired trap

efficiencies. It was recommended that either a ten hour minimum de-

tention time or a maximum overflow velocity of 2 X 10-5 m/sec. be

maintained in order to achieve higher suspended solids removal efficien-

cies. Maximization of pond surface area and continuous provision for a

minimum depth of 1.0 m. (3.3 ft.) were advised to limit resuspension of

settled sediment. The authors also acknowledged that in most cases

flocculating agents would be required for removal of fine grained sedi-

ments.

Page 22: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

12

Inefficiencies inherent in settling finer sediment particles

were again recognized in an EPA study (1976). Referring to Ideal Set-

tling Theory, the study suggested that required settling area for a

sediment detention structure be computed as: A=00 /V s , where Qo is the

pond overflow rate and Vs

is the critical settling velocity of the

smallest particle to be retained. Factors causing deviations from Ideal

Settling Theory were outlined. Additionally, a number of design innova-

tions accruing from years of experimentation with existing ponds were

presented.

Ward, Haan and Barfield (1978) furthered understanding of the

basin design process with their work on the hydrology and hydraulics of

sediment basins. Various hypothetical basin geometries together with

different riser configurations were analyzed. Predictive equations for

estimation of peak effluent sediment concentration and basin trap effi-

ciency were derived through regression analysis on data generated by the

DEPOSITS sedimentation model (Ward et al., 1977b). Three baseline con-

ditions for simulation of pond performance were examined: (1) a dry

basin prior to the storm event; (2) a permanent pool below the riser

crest prior to the storm event; and (3) a permanent pool followed by a

base flow event occurring after the storm event. Notably, the authors

were of the opinion that effluent standards could not be met with per-

forated risers (principle spillway), and thus, they were not evaluated.

A number of illuminating conclusions were advanced regarding basin

design-performance interaction. Where the percentage of finer than 20

micron (II) particles exceeded 30 percent, it was the authors contention

Page 23: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

13

that trap efficiencies would not exceed 80 percent in basins providing

a detention time of less than 12 hours for the 10 yr.-24 hr. design

event. Moreover, it was felt that if sediment in the inflow contained

greater than 20 percent of particles finer than 20g, it was unlikely

that water quality standards would be achieved unless flocculating

agents were utilized or storage in excess of 24 hours was possible. An

investigation pursued by McCarthy (1977) was mentioned as having indica-

ted that flocculants could provide an economical solution to achievement

of water quality standards. In his work on sediment control on three

watersheds near Centralia, Washington, the author estimated chemical

treatment costs of $10/ac.-ft. of runoff.

Direct reference to the current federal statutes concerning

sedimentation pond design has been made by Krishnamurthi and Blazer

(1978). The authors contended that trap efficiency was more dependent

on functional design characteristics than on a particular magnitude of

flow. It was recommended that instead of requiring use of the design-

storm concept for basin design, modeling of the physical characteristics

of the storm events should dictate design logic. Redirected emphasis on

sediment concentrations in streamflow as opposed to point source

pollutant concentrations in pond effluent was also suggested.

Uncertainty in the Sedimentation Process

Transport of sediment from contributing watersheds to receptor

watercourses is closely correlated with overland water discharge. Thus,

the uncertainties involved in the sedimentation process are related, in

Page 24: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

14

part, to those germane to hydrologic systems. There are, however, other

sources of uncertainty introduced when a physical conceptualization of

the process is entailed. No attempt has been made in this review to

encompass all of the existing volumes devoted to empirical and fully

deterministic treatment of reservoir sedimentation. Only those formu-

lations which recognize the uncertainties inherent in the sedimentation

process are surveyed.

Apart from readily discernable sources of random hydrologic

behavior, uncertainty affecting sedimentation also resides in factors

such as type of land use, vegetative cover, soil structure and erodi-

bility, gulley headcutting, and other natural and man-induced processes

which contain elements of randomness (Woolhiser and Renard, 1978). The

uncertainty involved in appraising the erodibility of soil subjected to

the dynamic conditions of surface mining has been acknowledged by field

researchers of the SCS (EPA 1977). Another factor in the determination

of erosion from watersheds is rainfall energy (Wischmeier and Smith,

1965). The effect of regional precipitation uncertainty on the rainfall

erosion index (El) of the USLE has been examined by Renard and Simanton

(1975). Spatial and temporal variability in El values resulting from

air-mass thunderstorms were shown to be considerable even for water-

sheds located in close proximity to one another.

Woolhiser and Blinco (1975) discussed a study by Krumbein

(1968) in which the author classified three stages of statistical

development in sedimentology; descriptive statistics, analytical

Page 25: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

15

statistics, and application of stochastic process models. Descriptive

statistics emphasizes the characteristics of the sample while analytical

statistics concerns itself primarily with extracting sample information

for the purpose of inferring population characteristics. The stochastic

model derives from consideration of the random properties of the

phenomena.

The analytic category is exemplified by the work of Shirley and

Lane (1978), Flaxman (1972), and Weber, Fogel, and Duckstein (1976).

Shirley and Lane (1978) derived a mathematical erosion simulation model

and made a least squares fit to observed data for a small watershed near

Tombstone, Arizona. In his study of sediment yield characteristics

for the western U.S., Flaxman (1972) utilized multiple regression

analysis on logged reservoir and stock pond sedimentation data to derive

an expression for watershed sediment production. Independent variables

were assumed to be: the ratio of average annual precipitation to

average annual temperature, watershed slope, the percent of soil par-

ticles coarser than 1 millimeter (mm) in the surface 2 inches of soil,

and a descriptor of aggregation potential in that same 2 inch soil

surface layer.

The utility of multiple regression models for the prediction of

sediment yield has been scrutinized by Weber et al. (1976). Data

obtained from Flaxman's (1972) study was used to assess the applicability

of four linear and logarithmic transformation models. It was concluded

on the basis of regression analysis and economic loss function analysis

that the linear model was preferable to the other log transformation

models evaluated.

Page 26: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

16

Referring to the work of Parzen (1962), Woolhisen and Blinco

(1975) state that: "A stochastic process is the dynamic part of prob-

ability theory and we observe a stochastic process whenever we examine

a process developing in time in a manner controlled by probabilistic

laws." There exists a number of partial and wholly stochastic models

which can be included in Krumbein's (1968) third category. Model studies

linking stochastic rainfall-runoff relations with deterministic sediment

yield relations have been constructed by Auernhamer et al. (1977),

Renard and Lane (1975), and Fogel, Duckstein and Musey (1976). A method-

ology for determining reservoir sediment yield based on limited rainfall

data and a derivative of William's (1975) sediment yield model was

elaborated by Smith, Davis, and Fogel (1977). Effective rainfall, event

duration, and number of events per season were viewed as random variables,

thus dictating the random nature of the computed sediment yield.

Mathematical derivation of stochastic process models has appeared in the

work of Woolhiser and Todovoric (1971), Woolhiser and Blinco (1975), and

Woolhiser and Renard (1978). Mathematical representations were advanced

by Woolhiser and Blinco (1975) for the stochastic processes of pre-

cipitation influx, evapotranspiration, porous media flow, and surface

streamflow. A distribution function for sediment yield resultant from

the modeling of watershed stochastic processes was also presented.

Applicable Regulatory Statutes

Since the purpose of this analysis is the assessment of sedi-

mentation ponds designed in accordance with federal statues for surface

Page 27: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

17

mining and reclamation operations, a brief overview of the applicable

design, performance, and effluent standards set forth in those statues

is offered. All referenced quotations have been excerpted from the

Federal Register (1979). Although more recent editions may have appeared

during the interim period, it is assumed that no major alterations in

the text have ensued.

Perhaps the most straightforward of all the design criteria is

that pertaining to required sediment storage volume for the pond (Federal

Register, p. 15400):

Sedimentation ponds shall provide a minimum sediment storagevolume equal to

(1) The accumulated sediment volume from the drainage area tothe pond for a minimum of 3 years. Sediment storage volumeshall be determined using the Universal Soil Loss Equation,gully erosion rates, and the sediment delivery ratio con-verted to sediment volume, using either the sedimentdensity or other empirical methods derived from regionalsediment pond studies if approved by the regulatoryauthority, or

(2) 0.1 acre-foot for each acre of disturbed area within thethe upstream drainage area or a greater amount if requiredby the regulatory authority based upon sediment yield tothe pond. The regulatory authority may approve a sedimentstorage volume of not less than 0.035 acre-foot for eachacre of disturbed area within the upstream drainage area,if the person who conducts the surface mining activitiesdemonstrates that sediment removed by other sediment con-trol measures is equal to the reduction in sedimentstorage volume.

More controversial is the detention time provision (Federal Register,

p. 15400):

Sedimentation ponds shall provide the required theoreticaldetention time for the water inflow or runoff entering the

Page 28: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

18

pond from a 10-year, 24-hour precipitation event (design event).Theoretical detention time is defined as the average time thatthe design flow is detained in the pond and is further definedas the time difference between the centroid of the inflow hydro-graph and the centroid of the outflow hydrograph for the designevent. Runoff diverted under Sections 816.43 and 816.44, awayfrom the disturbed drainage areas and not passed through thesedimentation pond need not be considered in sedimentation ponddesign. In determining the runoff volume, the characteristics ofthe mine site, reclamation procedures, and on site sedimentcontrol practices shall be considered. Sedimentation ponds shallprovide a theoretical detention time of not less than twenty-fourhours, or any higher amount required by the regulatory authority,except as provided under sub-paragraphs (1), (2), or (3) ofthis paragraph.

(3) The regulatory authority may approve a theoreticaldetention time of less than 24 hours to any level ofdetention time, when the person who conducts the surfacemining activities demonstrates to the regulatory authoritythat the chemical treatment process to be used - (i) Willachieve and maintain the effluent limitations; and (ii)Is harmless to fish, wildlife, and related environmentalvalues.

Dewatering requirements governing the design and performance of

the pond spillway systems or other modes of discharging stored storm

runoff are outlined as follows (Federal Register, p. 15400):

The water storage resulting from inflow shall be removed by anonclogging dewatering device or a conduit spillway approvedby the regulatory authority, and shall have a discharge rateto achieve and maintain the required theoretical detentiontime. The dewatering device shall not be located at a lowerelevation than the maximum elevation of the sedimentationstorage volume. (e) Each person who conducts surface miningactivities shall design, construct, and maintain sedimentationponds to prevent short-circuiting to the extent possible.(g) There shall be no outflow through the emergency spillwayduring the passage of the runoff resulting from the 10-year,24-hour precipitation event or lesser events through thesedimentation pond. (h) Sediment shall be removed fromsedimentation ponds when the volume of sediment accumulatesto 60 percent of the design sediment storage volume.

Page 29: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

19

(i) An appropriate combination of principal and emergencyspillways shall be provided to safely discharge the runofffrom a 25-year, 24-hour precipitation event, or larger eventspecified by the regulatory authority. The elevation of thecrest of the emergency spillway shall be a minimum of 1.0 footabove the crest of the principal spillway.

Finally, the requirement which ultimately assures compliance with water

quality guidelines (Federal Register, p. 15400):

(0 The design, construction, and maintenance of a sedimenta-tion pond or other sediment control measures in accordance withthis Section shall not relieve the person from compliance withapplicable effluent limitations as contained in 30 CFR 816.42.

The purpose of dictating the use of sedimentation ponds is,

of course, to clarify polluted water delivered from those areas disturbed

by mining. Following are passages associated with water quality standards

and effluent limitations (Federal Register, p. 15398):

(a) (1) All surface drainage from the disturbed area, includingdisturbed areas that have been graded, seeded, or planted, shallbe passed through a sedimentation pond- or a series of sedimenta-tion ponds before leaving the permit area.

(2) Sedimentation ponds and other treatment facilities shallbe maintained until the disturbed area has been restoredand the vegetation requirements of Sections 8.6.111-816.117 are met and the quality of the untreated drain-age from the disturbed area meets the applicable Stateand Federal water quality standards requirements for thereceiving stream.

(7) Discharges of water from areas disturbed by surface miningactivities shall be made in compliance with all Federaland State laws and regulations and, at a minimum, thefollowing numerical effluent limitations:

Page 30: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

Effluent limitations, in milligrams per liter (mg/1) exceptfor pH

Effluentcharacteristics

Average ofdaily values

Maximum for 30allowable consecutive

dischargedays

Iron total 7.0

3.5

Manganese total 4.0

2.0

Total suspended solids. 70.0

35.0pH Within range of 6.0 to 9.0

To be determined according to collection and analytical proceduresadopted by the Environmental Protection Agency's regulations forwastewater analysis (40 CFR 136).

Based on representative sampling, The manganese limitationsshall not apply to untreated discharges which are alkaline asdefined by the Environmental Protection Agency (40 CFR 434).

In Colorado, Montana, North Dakota, South Dakota, Utah andWyoming, total suspended solids limitations will be determinedon a case-by-case basis, but they must not be greater than 45 mg/1(maximum allowable) and 30 mg/1 (average of daily value for 30consecutive discharge days) based on representative sampling.

(b) A discharge from the disturbed areas is not subject to theeffluent limitations of this Section, if -

(1) The discharge is demonstrated by the discharger to haveresulted from a precipitation event equal to or largerthan a 10-year, 24-hour precipitation event; and

(2) The discharge is from facilities designed, constructed, andmaintained in accordance with the requirements of this Part.

(c) Adequate facilities shall be installed, operated, andmaintained to treat any water discharged from the disturbed areaso that it complies with all Federal and State laws and regula-tions and the limitations of this Section.

20

Page 31: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

CHAPTER 3

SITE CHARACTERISTICS ANDFORMULATION OF STUDY

This chapter begins with a description of the hypothetical

watershed which functions as the medium for this study. Regional and

site-specific characteristics which aid in defining the resultant

hydrologic regime are detailed. Next, the procedure followed in

evaluating the magnitude of precipitation associated with particular

frequency-duration storms is outlined. The construction and methodology

of the INFLUX program which generates the required inputs for utiliza-

tion by the DEPOSITS sedimentation routine is then examined. Procedural

aspects of sedimentation pond design, including composition of rating

curves for the reservoir, are described. A somewhat abbreviated analysis

of the DEPOSITS routine concludes the chapter. The reader is advised

to refer to the appendices for clarification of the computational logic

of the aforementioned routines.

Hypothetical Watershed

The characteristics of the hypothetical watershed developed

herein for hydrologic analysis are indicative of reclaimed surface-

mined watersheds on the Black Mesa in northeastern Arizona. Black Mesa

coal seams, being mined presently by Peabody Coal Co., range from 5 to

28 feet in thickness (Fogel, Heckman, and Vandivere, 1979). Following

21

Page 32: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

22

the extraction of the mineral, the spoil material is recontoured so as

to approximate, as closely as possible, the original topography.

The study drainage area was assumed to encompass 50 acres of

graded spoil material. Average watershed slopes of 6.7 percent and a

slope length of 250 feet have been chosen. Contour-grid determination

(Williams and Brendt, 1977) of existing basin slopes on experimental

watersheds at Black Mesa was used to calculate the former value while

the latter fell within the range suggested to the author by Hamon (1979).

Spoil material was assumed similar to that found on the J-3 spoils

experimental watershed located on the mesa. Lack of structure,

relatively high clay and low organic content, and low infiltration

capacity distinguish this material from surrounding natural soils.

Natural precipitation which varies from 9 to 13 inches in the Black

Mesa region has been assumed to be the only available source of moisture

for production of water and sediment discharge from the watershed.

Approximately half of the annual moisture influx is derived from summer

convective storm activity. The remaining portion is delivered by frontal

storm systems in the form of rain or snow. Runoff-generating events

are few and often far between, thus delineating the largely ephemeral

nature of streamflow in the area.

Development of Hydrologic Inputs

In order to adequately assess the overall performance of the

sedimentation pond, it was felt that a wide spectrum of hydrologic

events should be incorporated into the investigation. It was deemed

Page 33: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

2 3

appropriate that the precipitation-frequency maps published by NOAA

(1973) be used for determination of point rainfall volumes at the study

site, since it is common practice among designers possessing limited

data to utilize thie material.

An extensive network of both recording and non-recording rain-

gages provided the Weather Bureau (NOAA) researchers who compiled the

maps with point rainfall volumes for specific sites in the western U.S.

Isopluvials were then constructed on the basis of extrapolated data

drawn from multiple linear regression equations relating topographic

and climatologic factors to variations in precipitation frequency values.

Frequency analysis was carried out using the annual series method and

empirically derived factors for conversion to partial duration series.

A total of 38 precipitation events were defined through use of

the maps and accompanying equations and diagrams. Precipitation in

the form of snowfall was neglected under the assumption that runoff and

sediment production resulting from snowmelt at the site is minimal.

Point precipitation values were left unaltered by depth-area analysis

due to the small area of the watershed. The suggested one or two

percent reduction due to areal distribution of rainfall seemed question-

able in light of the uncertainties involved in the Weather Bureau

procedure. A summary of these estimated precipitation volumes for the

chosen durations and return periods appears in Table 1.

Page 34: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

Table 1. Point precipitation-frequency values for Black Mesa Mine,Arizona, in inches*.

Duration Return period (T), yr.**hr. 5 8 10 12 15 25

0.17 0.46 0.52 0.55 0.56 0.59

0.25 0.59 0.66 0.70 0.71 0.75

0.50 0.81 0.92 0.97 0.99 1.04 1.13

1.0 1.03 1.16 1.23 1.25 1.32 1.43

2.0 1.15 - 1.37 - 1.48 1.60

6.0 1.39 - 1.65 - 1.80 1.93

12.0 1.63 - 1.88 - 2.00 2.19

24.0 1.85 - 2.10 - 2.24 2.45

*All values were obtained through subjective interpolation ofisopluvial contours.

**T = 11PE , where PE is the exceedence probability.

24

Page 35: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

25

Model Components and Operation: Program INFLUX

The INFLUX program has been composed for this thesis with the

expressed purpose of integrating the limited data base available for

the mine site into the evaluationof storm-watershed response. It was

the intention of the author that the physical characteristics of storm

events be given fuller consideration, at least where the mathematical

modeling was concerned.

Precipitation

As in the case of the methodology picked for determination of

precipitation frequency values, an attempt was made to enlist a scheme

for temporal distribution of rainfall which was founded on essentially

valid statistical analysis. The Type II rainfall distribution developed

by the SCS (Kent, 1973) was chosen due to its large data base and

qualified recognition by the hydrologic community.

Rainfall depth-duration relationships outlined in Weather Bureau

technical papers (1953, 1954, 1956) were applied by the SCS to the

analysis of cumulative rainfall and duration for recorded storms through-

out the U.S. The resultant Type II curve represented the curve of best

fit for the bulk of the continental United States, including all of

Arizona. As can be discerned from Figure 1, the greatest 30-minute depth

occurs near the middle of the 24-hour period. Because the selection of

the period of maximum intensity was intentionally related to hydrologic

design considerations, meterological relevance may not always be

retained (Kent, 1973).

Page 36: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

26

t7Zd / xd 1 1V.101. 01. Tit/AN IV a3ivintAmov ou_va

Page 37: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

27

For purposes of modeling, the distribution has been broken down

into 22 intervals of varying lengths. Any storm regardless of duration

can be apportioned through time with the aid of this discretized Type II

distribution. A sample distribution for a storm of 24-hour duration

has been entered in Table. 2. Use of this distribution, in conjunction

with the infiltration component which is discussed in the next section

enables the investigator to determine both the rainfall excess and the

duration of that effective rainfall. This is necessary if the effect

of individualized storm rainfall intensity on sediment yield and sub-

sequent pond performance is to be realized.

An index of maximum storm rainfall intensity has been calculated

as the ratio of the rainfall volume for the fifteenth increment to its

corresponding duration. Its description as an index derives from the

lack of actual influence it exercises over runoff and sediment production

as will become evident later in this chapter.

Infiltration

A component infiltration model which relates infiltration rate

to the availability of soil moisture storage has been applied to the

INFLUX routine. Drawing on the prior work of Blumer (in prep.), an

infiltration curve for the J-3 experimental watershed (see Figure 2)-

was utilized for determining that portion of storm precipitation which

translated into runoff. The J-3 watershed is monitored for precipitation

runoff, infiltration and sediment yield by the School of Renewable

Resources, University of Arizona. The Blumer data, in lieu of sufficient

Page 38: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

28(1)Table 2. Discretization of Type II rainfall distribution for 24 hr.

duration.

Time(hrs)

Time(pdf) (2)

Precip.(pdf) (2)

Time(cdf) (3)

Px/P24(Ratio of accumulatedrainfall to total)

0 0.000 0.000 0.000 0.0002.0 .083 .022 .083 .0224.0 .084 .026 .167 .0486.0 .083 .032 .250 .0807.0 .042 .020 .292 .1008.0 .041 .020 .333 .1208.5 .021 .013 .354 .1339.0 .021 .014 .375 .1479.5 .021 .016 .396 .1639.75 .010 .009 .406 .172

10.0 .011 .009 .417 .18110.5 .020 .023 .437 .20411.0 .021 .031 .458 .23511.5 .021 .048 .479 .28311.75 .010 .104 .489 .38712.0 .011 .276 .500 .66312.5 .021 .072 .521 .73513.0 .021 .037 .542 .77213.5 .020 .027 .562 .79914.0 .020 .027 .583 .82016.0 .084 .060 .667 .88020.0 .166 .072 .833 .95224.0 .167 .048 1.000 1.000

(1) from Kent (1973)

(2) probability density function

(3) cumulative distribution function

Page 39: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

00 0.2 0.4 0.6 0.8 1.0

TIME, hours

Fig. 2. Data points and infiltration curve for 0.5 year oldspoil material, J-3 experimental area.

(Blumer, in prep.)

29

Page 40: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

30

documentation, was assumed to represent an average antecedent soil mois-

ture condition for the watershed. Repeated attempts of fitting this

data to established infiltration models (Huggins and Monke, 1966; Holtan

1961) proved futile owing to the uncharacteristically rapid decay dis-

played by the J-3 infiltration curve. Consequently, a tabular

representation of infiltration rate vs. available soil moisture storage

was opted for use in the computer routine. A listing of the discretized

time-storage relations for infiltration can be found in Table 3. The

interactive procedure involving the rainfall and infiltration components

is now described. The reader is referred to the flowchart of Figure 3

for clarification during the ensuing discussion.

At the outset of a storm run, an estimate of the initial avail-

able moisture volume for the topsoil unit is made. This unit is assumed

to be underlain by a geologic stratum of greatly reduced permeability

which effectively impedes the downward progress of infiltrated water.

Upon saturation of the topsoil unit, additional precipitation influx

is converted entirely to runoff. Available soil moisture was computed

as the difference in stored water between saturation and the wilting

point for plant life. Pressure potentials for these levels were assumed

to be zero and -12 bars, respectively. Since the interarrival time for

convective rainfall activity is generally short, it was felt that this

lower bound for available soil moisture better reflected probable site

conditions. incorporating volumetric water content values for Black Mesa

spoil material developed by Fischer (1976), an available soil moisture

storage volume of 1.8 inches of water was computed for an accompanying

topsoil layer thickness of 6 inches.

Page 41: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

Write infil. rateinfil. vol.

Accumulate stored volume

YesAccumulate

rainfall excessand duration

Set moisture deficit =

Increment discretizedrainfall

Compute incrementalrainfall volume and duration

Yes•

Determine infil.rate for computeddeficit

Calculate infiltratedvolume

Updated moisture deficit

Accumulated stored volume

Set infiltration ratesteady-state

Calculate infiltrated volume

Set infiltrated volume =remaining storage

NoNo

Yes

Computemax. stormintensity

Write total rainfallexcess and duration

Construct triangularhydrograph

( START

Read in parameters

/and initialize

31

Fig. 3. Generalized flowchart for hydrologic linkages.

Page 42: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

32Table 3. Discretization of J-3 infiltration capacity curve.

Time(hr.)

Ave. Infiltration Ratefor Interval (in/hr)*

Infiltrated Volume(in.)

0-.10 1.93 .21.11-.20 1.21 .12.21-.30 .76 .08.31-.40 .50 .05.41-.50 .36 .04.51-.60 .29 .03.61-.70 ** .25 .02.71-.80 .22 .02**.81-.90 .22 .02.91-1.00 .22 .021.00 .22 .02

* Using trapezoidal approximation for area determination** Threshold for steady-state infiltration rate

Page 43: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

33

Following initialization of the stored soil moisture variable,

TFILL, the routine enters into a loop which calculates successive incre-

mental values of rainfall volume, RAINV. After each iteration of RAINV,

the infiltration rate corresponding to the present value of the soil

moisture deficit (AVAIL-TFILL), is applied over the proper time incre-

ment, RAINPD. An interpolation factor, TERP, has also been included to

better approximate the continuous decay process of infiltration.

A check has been introduced at this juncture to determine whether

or not sufficient storage has accumulated for the steady state infiltra-

tion value, SSINF, to be activated. As can be observed in Table 3, the

steady-state infiltration rate corresponds to a cumulative infiltrated

volume of approximately .55 inches of water. The entered value for

SSINF is thus referred to if the computed soil moisture deficit falls

below (1.8 - .55), or 1.25 inches.

Next, the volume of infiltrated water for the interval, FILL, is

added to the preceeding value of TFILL. This volume is equal to the

product of RAINPD and either ACTFIL or SSINF, both of which represent

infiltration rates for non-steady and steady state behavior, respectively.

The updated value of TFILL is then used in evaluating the following

iteration of rainfall. This procedure is repeated until the SSINF is

attained, from whenceforth the infiltrated water is introduced at that

rate and TFILL is set equal to AVAIL, thereby assuring maintenance of

the steady-state condition.

Page 44: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

34

Runoff

Once the abstraction is satisfied, if there remains any effec-

tive rainfall for the increment, it accumulates as rainfall excess,

TRAINX. In addition, the duration for any increment over which excess

rainfall is generated is added to the value representing the duration

of storm rainfall excess, expressed by the variable, DUREX. As TRAINX

is actually the depth of water per unit area of the watershed which is

available for transport overland, its extrapolation over the entire

basin establishes the volume of total runoff delivered to the detention

facility.

Use of the DEPOSITS sedimentation routine, described at the end

of this chapter, requires as an input the distribution of incoming flows

to the pond. The triangular hydrograph method (Kent, 1973) used by the

SCS for hydrologic design of conservation and drainage structures was

chosen for this purpose because of its simplicity and applicability to

ungaged watersheds. Limitations on its general use include a maximum

drainage area of 2000 acres and average slopes of less than 30 percent.

Calculation of the peak discharge from which constrUction of the inflow

hydrograph.can•be accomplished is expressed by the equation:

KAQ q= t

where q = peak flow rate in cubic feet per second.

K = watershed parameter, a function of hydrograph geometry

A = watershed area in square miles

Q = volume of rainfall excess in inches over duration D

t = time from initiation of runoff to attainment of peakflow

Page 45: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

35

The time to peak is closely related to the time of concentration

for a watershed and is calculated by the following expression:

Dt = 2 +L

where D = duration of rainfall excess in hours, and

L = basin lag time in hours.

The drainage basin lag time, L, is computed by the equation:

.8 (S + 1) 0.7

L= 1900 Y0 • 5

where L = basin lag time in hours

1 = length of mainstream to farthest divide in feet,

_ 1000 10CN ,

CN = A retardance factor approximated by the curve numberrepresenting the watershed hydrologic soil cover com-plex, and

Y = average slope of watershed in percent

An empirical relationship derived from small watershed data

describes the hydraulic length:

1 = 209a0.6

where 1 = hydraulic length in feet, and

a = drainage area in acres

When peak flow has been computed for an event, the program simu-

lates construction of the triangular inflow hydrograph at a chosen time

interval of .05 hours. The choice of this particular interval satisfied

the minimum routing requirement of 5-10 time steps for the rising limb

of the hydrograph for all of the storms examined. Two simple linear

Page 46: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

36

expressions are necessary for hydrograph constructions:

ORD(J) = PEAK/TPEAK x CUMIN(J), for the rising limb

and

ORD(J) = PEAK - (PEAK/1.67 x TPEAK) x (CUMIN(J) - TPEAK) ),for the receding limb

where ORD(J) = the inflow hydrograph ordinate for the jth stepin cfs,

PEAK = peak flow rate in cubic feet per second,

TPEAK = time to peak in hours, and

CUMIN(J) = cumulative time at the jth step in hours.

A graphical representation of the triangular hydrograph is pre-

sented in Figure 4.

Sediment

Since the primary objective of the study was to assess the

effect of uncertainties in the physical characteristics of precipitation

on sediment yield and pond performance, a desirable sediment yield model

had to incorporate a certain degree of specificty with regard to these

factors. To this end, the modified Universal Soil Loss Equation (USLE)

assembled by Williams (1975) for determination of event based sediment

yield was chosen. A hydrologically more specific formulation of the

USLE, the modified equation is given as:

.Y = 95x(Qxq)

56 xKxLSxCxP

where Y = the event sediment yield in tons,

Q = runoff volume in acre-feet

q = peak flow rate in cfs

K = soil erodibility factor,

Page 47: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

_ 484 A(Q) Ag ap

L

37

INCREMENT OF EXCESSRAINFLOW OR INFLOW

OUTFLOW HYDROGRAPH

D

I-<

/ID TIME

Tp

tbi

Fig. 4. SOS triangular hydrograph.

Page 48: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

38

LS = length-slope factor

C = crop management factor, and

P = erosion control practice factor

The last four variables are equivalent to those utilized in the USLE and

their values must be estimated according to the methods outlined by the

SCS. Peak flow rate and runoff volume are obtained from earlier pro-

gram components.

The modified USLE resulted from experiments conducted on small

watersheds in the Texas Blacklands which are also part of the semi-arid

zone of the Western U.S. Uniform prediction accuracy was maximized in

development of the equation. Predictive ability was found to be greater

for larger storms than for smaller ones. This, however, coincides with

the importance ascribed to larger storms in the production of sediment.

Because the formulation was based on sediment yields and not gross

erosion, use of the modified equation negates the requirement for appli-

cation of a delivery ratio for the purpose of determining the transport

efficiency of overland flow.

Sedimentation Pond Design

Upon completion of computer evaluation of the 10-yr., 24-hr.

design storm event, sufficient information exists to enter the pond de-

sign process. A trapezoidal basin configuration has been chosen which,

within the restrictions of the study, offered a simplified, yet not

overly inaccurate, representation of actual pond geometries.

Page 49: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

39

A detailed description of the design procedure is presented in

the DEPOSITS design manual (Ward, Haan, and Tapp 1979), therefore, only

highlights of the procedure are addressed here. The reader is referred

to Appendix A which contains the final design aspects for the pond used

in subsequent analysis.

There are six basic steps involved in the design procedure used

herein:

(1) Determination of design storm characteristics including water and

sediment volume and the accompanying inflow hydrograph. In addi-

tion, accumulated 3-yr. sediment storage volume must be computed

using the USLE and a sediment delivery ratio. Due to difficulties

in determining the sediment volume derived from gully erosion,

that factor has been neglected.

(2) Site selection based on surrounding topography, location with

respect to disturbed area and active alluvial systems, and hydrau-

lic design considerations. (Basin length, as measured from inlet

to outlet, should equal or exceed twice the average basin width).

(3) Preliminary design of dam embankment with provision for an emer-

gency spillway and freeboard, and delineation of the prinipal

spillway configuration.

(4) Compilation of operating curves for the reservoir, specifically

those relating reservoir stage to surface area and spillway

discharge.

Page 50: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

40

(5) Routing of the design storm through the reservoir using any estab-

lished procedure- The DEPOSITS sedimentation routine was applied

in this case.

(6) Repetition of steps 2-5 until sufficient detention time has been

achieved.

Keeping with the practice of using established methods for actual pond

design, inflow volume was computed with the aid of the SCS rainfall-

runoff equation described by Kent (1973). The SCS equation is given as:

where Q = accumulated direct runoff in inches

P. = accumulated rainfall in inches

Ia = initial abstraction including surface storage,interception, and infiltration.

S = potential maximum retention

The initial abstraction has been empirically approximated as a fraction

of the potential maximum retention by the relation:

Ia = .2S

Therefore, the resultant expression for determination of runoff volume

is:

(P - .2S)2

Q = (P + .8S)

A great deal of freedom has been allowed by the regulatory

agency in working out the particulars of pond design. Prudent

Page 51: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

41

engineering practice is relied upon to minimize inefficiencies in hydrau-

lic behavior, consequently providing a maximum of design flexibility.

Some assumptions adopted in the process of modeling the hydraulic

behavior of the dewatering system used for pond simulation should be

stated. A perforated pipe spillway was employed because of the ease it

exhibited in achieving adequate detention times for routed basin inflows.

Characteristics of the structure include a corrugated metal riser and

two sets of circular perforations comprised of three perforations per

set, each perforation measuring two inches in diameter. Three hydrau-

lic conditions are assumed to adequately describe flow through the

principle spillway.

Initially, as the reservoir stage exceeds that corresponding to

the successive dewatering sets, the discharge can be described by the

orifice flow equation:

= Ca (2gH)05

where Q = discharge in cubic feet per second,

C = orifice discharge coefficient, assumed = .60,

a = cross-sectional area of orifice in square feet,

g = acceleration due to gravity in feet per secondsquared, and

H = head on the orifice, measured from the orifice centerto reservoir water level, in feet.

When water level adjacent to the riser transcends the spillway

crest elevation, weir flow ensues. Morris and Wiggert (1972) suggest

use of the following equation for description of this condition:

Page 52: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

where

a(2gH)Q = (1 +K+K+ K L )1/2

e b c

H = head on conduit outlet measured from the reservoirlevel to six-tenths the conduit diameter above theinvert.

421/2

Ke = entrance loss coefficient,

Kb = correction factor for energy losses in bends, and

Kc = friction factor

The pipe flow values were extraced from Table 5.3 of the DEPOSITS user's

manual (Ward et al.,1979) which presents discharge values as a function

of conduit diameter, head, and conduit length for (Ke + Kb ) = 1.0.

As is true of larger impoundment structures, the hydraulically

active portion of the pond excludes that volume which must be allocated

for deposited sediment, referred to as "dead storage." Some of the de-

sign details affecting hydraulic performance have been set out in the

regulatory statutes, highlighted in Chapter 2, however, others such as

positioning of the spillway and maintainance of minimum depths or

permanent pool volumes have been neglected. The accumulations of

further data on actual and simulated pond performance may expectedly

lead to amendments to the current criteria covering pond design.

To enable relative differentiation between storm characteristics,

and their effect on pond performance, maintainance of a baseline condi-

tion for the pond was essential. Established federal statutes relating

to pond design and effluent limitations convinced the author that a

"critical" condition is defined by the following elements: (1) a dead

storage equal to 60 percent of the calculated 3-year sediment

Page 53: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

43

accumulation corresponding to the specified cleanout level for the pond;

and (2) a permanent pool consisting of the volumetric difference between

the cleanout level and that of the first orifice set. A graphic repre-

sentation of general pond features is presented in Figure 5.

DEPOSITS Sedimentation Model

The DEPOSITS sedimentation routine (Ward et al.,1977b) is a

mathematical model used to describe the transport and deposition of sed-

iment delivered to and routed through a sedimentation pond. Impetus

for model development came mainly from the desire to provide a means

with which the user could more adequately assess the prospective per-

formance of sedimentation ponds, and as a gage to weigh the necessity of

additional or alternative methods of sediment control. The model has

been verified using data from 11 functioning ponds and was found to ex-

plain greater than 90 percent of the variation in basin trap efficiency.

Flow within the pond is described in terms of the idealized plug

flow concept. This assumes that no mixing takes place between plugs and

that each successive plug entering the pond assumes the position main-

tained by its immediate predecessor. Although it simulates basin

performance, the model does not pretend to describe the complexities of

actual hydraulic behavior in the basin.

The sedimentation process defined in the model can be elucidated

by tracing the movement of a sediment-laden plug of inflow through the

basin. Upon entering the pond, the plug is assumed to possess a speci-

fied sediment distribution with depth. Furthermore, each plug is

subdivided into four layers at 0.125, 0.375, 0.625, and 0.875 of the

Page 54: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

LaLu

o>0

Lu

<txo

0co I—

Ew w0 (r)I. I.

O (,)>>

44

Page 55: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

45

average flow depth. As the plug progresses, fall velocities required

for particles to reach these distances are computed as the depth tra-

versed divided by the detention time for the plug. Particle diameters

corresponding to the computed fall velocities are then determined by

applying Stokes Law of Ideal Settling:

V x D - 51.5 x (SG -1)

where D = particle diameter in millimeters,

V = corrected fall velocity in feet per hour,

= water viscosity in centimeters squared per second

SG = particle specific gravity, and

51.5 is equal to 0.8 times the gravitational acceleration (32.3 ft/sec)

times a conversion factor to achieve equation dimensionality. The

correction factor of 0.8 compensates for the effect of non-spherical

particles on settling theory.

The proportion of sediment remaining in any layer of the plug is

subject to the law of continuity:

dsI - 0 = —dt

where I = mass of incoming sediment

0 = mass of outgoing sediment

dl= differential of mass of stored sediment with respectdt to time

Deposition occurs and is assumed irreversible as soon as a particle

reaches the reservoir bed.

.5

Page 56: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

46

When the plug is discharged, its sediment concentration is thus

dependent not only on the withdrawal characteristics employed, but also

on the detention time and the redistributed sediment profile within the

plug. A graphic conceptualization of the plug flow routing process is

presented in Figure 6.

Required inputs to the DEPOSITS model are readily available and

are minimally dependent on in-field evaluation:

1. Inflow hydrograph.

2. Viscosity of the flow.

3. Stage-area curve for the basin.

4. Stage-discharge curve for the basin.

5. Stage-discharge distribution curve.

6. Degree of dead storage or short circuiting.

7. Sediment inflow graph or load.

8. Particle size distribution and specific gravity of the sus-pened sediment.

Stage, as referred to in the model, is the depth of water above

the lowest level on the basin bed. A uniform outflow rate with depth

is assumed by the model if a stage-discharge distribution curve for the

basin is not specified.

If sediment distribution data is lacking, sediment concentration

is made proportional to the water inflow rate. For effluent concentra-

tions to be determined, the total incoming sediment load must be inputed.

In addition, the model has the capability to simulate changes

in basin geometry resulting from deposition during the event. This makes

Page 57: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

o

SAO 31V2:1 MO1 A

47

Page 58: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

48

it especially desirable from the standpoint of prospective time series

analysis of hydrologic events and corresponding pond performance.

A caveat should be inserted here concerning the computation

of detention time in the DEPOSITS routine. As is required by the reg-

ulatory statutes, detention time is calculated as the time between the

centers of mass of the inflow and outflow hydrographs and is represented

by the program variable, CENTME. In situations where a permanent pool

exists, which remains as storage after dewatering, the computed cen-

troidal detention time will only approximate the average theoretical

detention time for the storm event. This is due to the fact that the

detention time for the volume of stored flow discharged during the

routing process has been maintained for an extended period of time. The

DEPOSITS routine has included another measure of detention time which

gives special consideration to the permanent pool volume, identified in

the program as the variable STRMTM. Because of the complexity and sheer

bulk of the DEPOSITS sedimentation model responsibility for a more de-

tailed explanation of program operation has been delegated to the

DEPOSITS Design Manual (Ward et al., 1979).

Page 59: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

CHAPTER 4

RESULTS AND DISCUSSION

The following pages document the results of model runs for the

Black Mesa hypothetical watershed and companion sedimentation pond. A

discussion addressing the implications of hydrologic uncertainties on

pond performance is then offered. The chapter ends with some thoughts

on model adaptability and regional bias.

Pond Sensitivity Analysis

Prior to initiation of the present study, the author, along with

his colleagues, expected that only one watershed condition would need

be examined in order to assess the effect of hydrologic variability on

pond performance. This, as the study unfolded, turned out not to be

the case. To, in effect, unmask the hydrologic uncertainties as they

affected the efficacy of basin performance, two additional sets of runs

had to be made. Thus, the following modeling schemes were investigated:

(1) minespoil material with untreated pond inflow,

(2) natural soil material with untreated pond inflow, and

(3) minespoil material with chemically treated pond inflow.

The reader is directed to Appendix B for a sample listing of required

DEPOSITS input parameters and their associated values.

49

Page 60: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

50

Minespoil Material with Untreated Pond Inflow

A total of 38 precipitation events encompassing varying

frequency-duration relationships were evaluated for this condition

using the INFLUX and DEPOSITS models. Peak effluent sediment concentra-

tions resulting from routing the specified event flows through the pond

are listed along with precipitation characteristics in Tables 4 and 5,

while sediment yields for all conditions simulated are compiled in

Tables 9 and 10, pages 64 and 65.

Since the Williams equation (1975) for event-based sediment

yield was used in generating incoming sediment loads, estimates had to

be made for the equation's variables corresponding to the proper site-

inflow conditions. These values were based on projected 3-yr. averages

as follows: K = 0.35, LS = 1.26, C = 0.63, P = 0.35. The estimated

crop management factor (C) value of 0.63 represents an assumption that

during the first year no vegetative growth occurred, while in the second

and third years, a meager 10 percent cover was established. Justifi-

cation for the remaining variable values listed can be found in Appendix

A covering pond design. On first inspection, these values may appear

low, however, it must be remembered that irrigation has been neglected

and natural rainfall is limiting. The distribution used in apportion-

ing incoming sediment is equivalent to that utilized in the pond design

process and appears in Appendix A.

Data presented in Tables 4 and 5 suggest that pond performance

is linked less to rainfall volume than to its intensity, represented by

Page 61: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

51

Co7:1

cU4-1

0

Cl

4-1Cl

co-Xc•-4

-Xso co Nj al

-;--.CO

-X -X(041

4-• cd0 di

CU 4-1CJ CL) CIO0 C11

COnr)

Ce)CO

Ce)Lin

C11ce)

Ln0")

cnce) cn

c0r-,

.1-

CO01CO 1/40 Co Q

0Q

c•In.o cf-) cf)

00 0 C.)CU 0Q 4-10 4 Cl0 .6-+

•,1

(1)cr)

0

-00)0)Co

4-J E

(1)• •

,"-:••o..z

4-10.)

14-1 C/G EG-1

c-‘)

0 r-1no

O '%

CO

CclCT1CA

cr)CO

CA

CACO

(Ni

1-10coCe)CN1

010un.1-

NI

1-nCr,0

coCr)

Ce)

r•-•C•1

r-1

CACCr")

r-1r.1--1

N.h-

CO

1/1

10Le-1v-1coCe)

(-.1431a,c*, )

cnC

CN)

L.0CN1Ill

C•1

Co

1-41 0Cla)

1.1 •1-1

cd a

-G10)4-J

u

)-; Co X •

Iai Ln

a) $.4a) -G

4-3 ClCl I— Cl

1-1 -ri

o•Nco

N.

c-s)

01

s.0

0

CO

Ln

... COr-1 ce)

CNun

.0

NI

1/4.0

N.t-4

CT1C1.1

Co1-1

4-1CO

Le)CA

0N

CO1/4.0

CO

1/40CO

0Ce)

Gr, cG

CoCi Co

E

ci) 04-3 Cl

0•,-4 •

4-) Clc.) cd

4-J Cl(71

Cl

cd CL)Z

C.) CiCl

4-1

a. 4.1-1 4-JQ0)0)

0

C71

0

1-4Co

0

Ce) L11

,-I

01c'n

Cr).0

•CO

.r-1

CALn

0

,,43.0

0

C110•L

0

s.0

1-1

unL.r)

0

0N-

0

h,crn

0

cnN.J

r-1

cn

r-1

uN

0 P:1

4-) "0 0Cl 1-14-) 4-) 14-1

04 0 .1-4 L-1W. W ;-• -G •

C.)0) 7-1

Tl0

-0 4 4W. W

•W

w4

.W4

4In

W.0

•W

4 4 Cl$-1

0f:1-1 14-1

0a. .LJ

0 I

N.1-1

te)CA

0ul

4

t-i

4•

CsJ

4

1/4.0C•I1-1

.CN1

N.r-I

NI• tin

,-C

r-1

N,--I

lnC.1

01-el

CA• 1 •

C) Clr.z.) Cl 0

I I I I I I I I I 1 1 1 1 1 1 1 1

CL)

r-I4-1

cr• triO0)

1-1>1

1-1>1

1-1>,

1-1 1-1 3-4>1

1-1>1 >1 >1 >1

1-1>1 >,

••1

CO 4-)cil 4-1

Ln Ln in Ln un ln CO CO CO CO Ctri

0ri

0ri

0ri

04

0tri

Page 62: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

52

$.4 :: a)0 0

'10 a) ,--44../ Ca

0 Ca ›4-1 • ri

0 ..1 1-1.0 0 P. 04-1 1-4 0 4-1a. 4.J W C.1CU z a. cd

44•ri cd

.

0 P04-Jri CD4-J -r-4c0 -0

-000

r.C.)=

.7)gQ3 4-1 g 0

-C1 .1-1 ,--I 0)-I 0.-4 -,--i 00 .1-1 Cd 4-1 0

"Cl C.) 4-i 0 1-1g (1) g 40 • 4-10 1-_4 -,-I (f) 04.4 P. 0 1--4 g .0Cl) 1-1 .4-JO •,-1

ID) al -H $4>1 4 F-1 _4 4-2 4-1

4-1 4.1 1-1 '0 ri COri "0 1-1 Ca1-1 4-4Q) CD 0 Mt 4-3

CO 01:. NO Cl0 7, ”-I c.) Q) Q)0' 0E-4 CO N4-4

•,-1 ca r1 4-_44-1 4-1 Cf) Q) ci) (t ID)gQ)

cci C..))4 En

_Q)u -'.— a) 0

0 -H -40r-1 0) I.) 4-J4 c.) 1-14-1 -1: W c.) -,-1 4-144 4.+ 4- ) al cd .L.J cd

ID) 0.'-1 3-1 4-1CO LI-1 CO CCI ••-I

-‹ 4-J O -cl 0. 1:.a 0 a) -0 -7- 4w fa.) 4-1 Cl 'OC)

0 0 ct › Q) Q)QI Q1 •,-1 E1--i 14 /-14.) 1--44-J `,.4 a) 11) 0.QJ 1:2.4-4 4J C)) 4../I)) Q)0 al ,o —.4 4-4z G4 0. W 0 <Oic ,—I c.i ri

Page 63: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

cnG cUO +.1.r-I cd4-) a)cd

4-)

GG 0

c.) 4G 4.4

cJO .r4

4.)G •*-1C) 0

1-1

ai"Licl) G

4-1GQ)..• di

'4-40L4-1

(1)C)

cia) G• 1-1

^0 4-1cU

4-1

CD >1

014CV

"0 Ig CN0 ri

Cl) ••U C/1

E4-1 1.4Cf) 0

.1-1 4-11-i CC)

4-1 cl:(.) c1)Q)QJ• Z

(J c.)

O 1-1O pO .

4-1 "0 0c1 CL)

4-1 4-4**-1 g gfa. 0

1-1L) "0a) 14 G$-1 0 0P4 44 PLI

53

4.: 4c ic 4( .x 4C 4'. 4.: •X 4:N. s.0 cr) N. 01 CV CV Cr1 CI1 n--1 ,.0 0 Cn 0 al k.0 CO ...1-4 .

Cr) CV %.0 Cr) 0 C1I V) ....1- V) CO 0 ri CO 0 ri 1-1 01CO 01 01 ‘.0 01 01 0 ,./D ....1* Cr) CV -..1- ri VO ..1- Ce) Ce) CV

i-1 r--1

cn .0 .0 on Ln co Cr. 01 CV CO L4-1 • 0 ulIf) 0-1 N. ce) rl ON ce) 0 ce) CO N. ri N. 1.11cr.% CO %.0

r-I ce)- tr) CO Q c\

I Ln a, 0 Ln 0• 0 on.0 if) 01 ol .0 Ln cc) on c.1 c.1 CV Cr) CV CN cN

tr) CO s..0 CO N. 01 CV N. Ce) CO CO N. N. CO u.0(N s.0 Cr) 0 (N 1-1 11) 111 1-1 Ce) N. CO 0 0 lx) uN

ON N. if) (N Cr) CO N. (N • crt 0 CO (NCO

N. ce) CO N. u") ce) 1-1 tr) Cr) CV

s-0 ri Cr) 14.1 Oin CV CO 0 0 -4 Cr) C,") 0 Ce) CS) V)Ln N. C1 CV if1 N. 0 cn CO Q CV CN

• • • • •0 0 0 0 0 •-1 cN1 CN1 n-1 c.1 cN1

P P 1-4

P . 4.44 4 • 4 4 4;-4 ...1 4 P

P -'.4 -4 P .4 4 1-1 )-1 )-1 4 -GN. Ln 0 4 N. Ln 0 -0 -G 4 0 4 4 4,--1 c \I 1.r) n--I cNi in c.1 --.1^ Ln C1I ...?

v-I ri CN %.0 ri C,4 • ri CV s.0 ri CV

I I I I 1 I 1 1 I I I I I I 1 I I I

1.4 4-1 1-1 1-1 1-1 1-1 4-4 4-4 4-4 ).4?, P., P. ›, >, ?, >1 ›, ›, P., P., P.. P., P.

c•I (N 1-41 Ir) Ir Lrl in in Ir) u-1 Le) 1.r) Ifn-1 1-1 ri ri ri CV CV

Page 64: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

54

the aforementioned intensity index. This index, as previously defined,

represents the ratio of the rainfall depth to duration for the critical

intensity index. This index, as previously defined, represents the

ratio of the rainfall depth to duration for the critical fifteenth

time increment of the discretized SCS Type II rainfall distribution.

The standard rainfall intensity expressed as the rainfall rate

in inches/hr. over the corresponding time period was felt to be too cum-

bersome for comparative purposes. Consequently, the cited intensity

index was chosen in its stead. Readers are cautioned not to confuse the

two concepts, lest unwarranted protests ensue.

Tracing causal connections, it becomes evident that the link

between intensity and pond performance is due to the ability of short

duration-high intensity storms to generate relatively large amounts of

sediment accompanied by high peak discharge rates and short inflow hy-

drograph base times. Examination of successively routed plugs through

the reservoir indicates an accelerated introduction of the plug corre-

sponding to the observed peak inflow rate and peak influent sediment

concentration. Because the shortened time base of the inflow hydro-

graph, this rapidly introduced plug affects a greatly reduced detention

time which is responsible for the higher peak effluent sediment concen-

trations simulated by the model. Low intensity-longer duration storms,

on the other hand, display the attenuated inflow hydrograph characteris-

tics which contribute to the attainment of increased plug detention

periods, especially for the plug associated with peak inflow rate and

sediment concentration.

Page 65: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

55

In both cases, the clay fraction was responsible for the bulk of

that sediment extant in the pond effluent. There was, however, a

greater percentage of silt-sized particles in the effluent of simulated

low intensity-longer duration events. Although an apparent inconsistency,

this appears to accrue from the difference in discharge characteristics

associated with the two storm types. The greater volume of stormflow

produced by the low intensity-extended duration events results in an

increased time period over which higher stage levels and subsequent

discharge rates are registered. Since the increased range of depths

associated with these events now encompasses both dewatering orifice

sets, relatively greater plug detention times are offset by the increased

settling depths which must be traversed by entrained particles to avoid

incorporation into the pond effluent.

Another aspect of data listed in Tables 4 and 5 which merits

attention is the gulf which separates the peak effluent concentrations

computed for all the events modeled and the current OSM-EPA effluent

water quality standard of 70 mg/l. This is attributable to two inter-

related factors. Coincidental with the required 24-hr. theoretical

detention time is the notion that sedimentation ponds can only be

depended upon to remove sediment corresponding to minimum particle size

of 20 g (0.02 mm). Corresponding to the middle of the silt sized

particle range, this lower limit precludes the maintainance of high

trap efficiencies for watershed soils with a high proportion of par-

ticles residing in the clay and lower silt fractions. Thus, ponds

Page 66: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

56

designed according to the federal guidelines cannot be expected to

even remotely approach Federal effluent limitations where these con-

ditions are present.

In an effort to clarify the projected effect storm character-

istics have on the actual attainment of pond effluent standards, it was

necessary to negate the overriding influence of the soil particle size

distribution on simulated pond performance. A sensitivity analysis was

conducted on pond peak effluent concentrations to determine the size

fraction distribution and crop management factor value which when

inputed would enable differentiation between storms that successfully

achieved water quality standards and those which failed. The results of

this analysis are shown in Figure 7. A clear distinction can now be

made between the two storm regimes studied.

Inspection of Tables 4 and 5, listing storm characteristics and

peak effluent concentrations from which Figure 7 was derived, provides

further evidence of the critical nature of high intensity-short duration

storms with regard to pond performance. For the particle size distribu-

tion and "C" factor value noted in Figure 7, intensity index (PINTMX)

values in excess of 40 in/hr. resulted, for all but one case, in a

failure to meet the 70 mg/1 standard. Conversely, all modeled events

which were calculated to have intensity index values less than 40 in/hr.

satisfied the requirement. The crop management factor value used

represented cover conditions which in the opinion of the study

advisers could exist only under irrigated conditions. Even for these

conditions, the resulting particle size distribution would in reality be

unattainable.

Page 67: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

57

The most obvious departure the data exhibit from the stated OSM

performance criteria relates to the water quality exemption for storms

possessing proven generated inflow volumes in excess of that produced

by the applicable 10 yr.-24 hr. event. The demarcation of event con-

centrations represented by the vertical line in Fig. 7 suggests a marked

preference for the exempted stormf lows in attaining effluent sediment

concentration standards. Moreover, a substantial number of storms which

exhibited high effluent concentrations spawned only nominal inflow

volumes.

National Soil Material with UntreatedPond Inflow

Because of the overwhelming failure of all routed stormflows in

achieving federal water quality statutes, modeling of a control situa-

tion for comparative purposes was viewed appropriate. All factor values

included in the sediment yield prediction equation, with the exception

of the "C" factor, were left unchanged for condition 2 model runs. The

"C" value chosen, 0.09, correspond to a cover percentage of 40, describing

a vegetal mix of grassy surface and limited canopy. In addition, the

particle size distribution calculated for the incoming sediment was

derived from data for natural experimental watersheds located at Black

Mesa. A tabulation of this data appears in Table 6.

Generally, similar observations can be made in viewing natural

condition data as for those concerning untreated minespoil conditions.

Although peak effluent sediment concentrations are somewhat lower than

those resulting from minespoil simulation, they still exceed the

Page 68: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

60.0

50.0

>-F—ET) 40.0

LLI1—Z

30.0

90.0

58

80.0

70.0

PARTICLE SIZE DISTRIBUTION:

<.002 mm = 0 %.002 mm 5- X 5 .063 mm = 5 %.063mm <X 5-.125 mm =87%

>.125 mm = 8 %

Crop management factor = .09

20.0

I, 111.3

102.2

100• 94.4

0. q

90

80

70

ca.;60

50

4044: 1 • 40.4

10.0

0.0 3170 4930.0 1.0 2.0 3.0 4.0 5.0

INFLOW VOLUME, AC.-FT.

Fig. 7. Predicted peak effluent concentrations and contoursin mg/2. for Black Mesa minespoil with altered inputsand untreated pond inflow-

Page 69: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

59

Table 6. Particle size distributions for undisturbed experimentalwatersheds: Black Mesa, AZ.

* * * *Watershed Event Date %Clay%Silt %V.f.Sand%Sand

7-11-77 14.6 62.0 12.1 11.3

8-12-77 22.7 32.0 32.6 12.7

8-15-77 40.6 36.4 18.5 4.5

J-3 Natural 8-15-77 16.7 49.7 27.5 6.1

8-15-77 36.9 29.0 28.4 5.7

8-12-77 37.6 30.8 22.2 9.4

8-7-77 32.4 39.1

8-12-77 13.4 6.7

8-15-77 16.2 33.9

J-7 Natural8-15-77 27.7 43.0

8-17-77 42.2 38.0

8-17-77 37.3 20.8

7-22-77** 47.1 31.9

8-5-77** 51.6 17.3

8-16-77** 71.5 5.5

Average fraction %: 28.2 35.1

19.1 9.4

33.1 46.8

29.8 20.1

14.2 15.1

9.4 10.4

23.4 18.5

13.0 8.0

7.6 23.5

17.6 5.4

22.5 14.2

*clay: <.002 mmsilt: .002 mm x < .063 mmv.f. sand: .063 mm < x < .125 mmsand: >.125 mm

**assumed outliers

Page 70: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

60

effluent standard by two orders of magnitude. The precipitation

characteristics were unaltered for those runs, therefore, it is not

surprising that the previously cited tendencies relating storm charac-

teristics to pond performance are again evident. A listing of effluent

concentrations can be found in Table 7.

The question which begs attention here is: given the magnitude

of sediment concentrations in runoff generated under natural conditions,

why should sediment concentrations in runoff from disturbed areas be

required to register at levels corresponding to two orders of magnitude

less? Rationale for such a restriction seems to derive more from

difficulties inherent in determining ambient stream sediment concentra-

tions than from any conceptual analysis of the hydraulic regime. Even

a cursory review of current sediment transport theory reveals the long-

accepted principle that a stream-hillslope system will strive towards

maintenance of a dynamic equilibrium between the supply of sediment and

the stream's capacity to transport it. Thus, it appears conceivable

that the high quality water discharged from sedimentation ponds into

streams or appurtenant drainageways carrying natural runoff could

actually lead to downstream scouring which in turn could result in re-

suspension of previously deposited channel material. Acceptance of

Federal rationale aside, the inadvertant distortion of the existing

hydrologic regime in this manner seems a distinct possibility.

Minespoil Material with Chemically Treated Pond Inflow

A selected group of representative storms was simulated for

minespoil conditions and the introduction of a cationic polyelectrolyte

Page 71: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

61

Cic.n 0

461

CO r-1 1-r1 0 0 1-r) (T 4-1 -0- CO 01

0 VD 4.4-4 COO CVc43 V) 04 n.0 CO 1/40 c,44 4-1 4-1 tr)01 CO C'n I-I CN1 Cr) (.31

k k 1-4 kl • k,a •k

,.= ,.= ..cI.; kl kl

s-i4

4-,4

.0$.4 }-; W

S-I40

In•

-Q

4-4-1

r--,--i

incv

0LC i

.=

44-4

,.=

c.1

•-a

,..0CVr-4

-0-

c*44

0cr-4

,.--

4-1

.=

CV

,.=

,.0CV4-1

I I I I I I I I i I I i I i I

kl k 4-4 k k 4-1 kl k kl k 4-1 k kl 4..1 k>1 >1 >1 >1 >1 ›N ›, >1

CA cv Ln in ii-) in if-) Lc) Li-L In Lin Ln Lc) L.r) Li-)ri ri ri 4-1 4-1 4-1 4-1 4---1 ri ri CV CV CV CV c‘l

4.4C1.4

EuCO 0

4-4 4-1 L0 CO cn co Ln re) Cr) •

C)

co tr) co cD co L.r) sz) n.00 Clco n0 O CA C•4 ‘--1 CO 01 In CO Ln cv L.0

)-1 Cr‘

• o

O O Cr crN 0

0

41-1 ce) c-V CV 4-1 -0- cn (-A In cr4 cv4-14-4 Cit-4-4 UC44 0

0

CiP-1

0o

-4-4 •4-4 4-4 4-4 $-4 4-4 4 ..4 4_: ..4 4-4al -C .0 ..0 • kl k .0 .0 .0 • 0 .0 .0

4.1 $.40 0 r-- Ln 0

w.0

4-1.0

4-1.0

.0 -04--- Ln O ,t1 N- in 0 4.-IC. .ti

$11.L.(1) 77 ri CA In C1I , ..t rk C1I in ,--I cv Li-,

› I • • • 4-1 CV L.0 ,--1 cv ,-i ,--I cv n.0Cr.3

U 111 1111111111111111

..1 a)00 4-4 4-1 k k kl 4-1 kl kl kl 4.1 k 1- 1 k 4-1 k k ki k.kcia a)

4_iv-L

›N

Ln

>1

Ln

>-n

Ln

›N

Ln LE)

›,

in Ln Ln 00

P',

CO

>1

00

4--1

CO

>1

0,-i

›,

0n-i

›N

0r-I

0,-1

0,--1

›,

0,-1

046-4CO

cl

aJ

cd

-oCi

0

• •

o$.4

4-4 4-1o 14-1

-HCOO -00

0

ca)4 -04_i CIO 4-1Ci raO C)

)-1O 4-i

Ci

r..)

• 4-.1E

(1.1 4-41CO 0

444-141-4 4.4

(1)CI) .1-44

O 0r-i E4-44-4 4-4

0.10

Ci4-J r-1

"Cl4-i

P-1 O

(71 4- 0 0 it cv s.0 ce) u") O cr, r-- CA r-- c0 r--

4-) cv -.I- (' ) cv Lc)

Cr) CV CA rk rk Cr) CA rkrki(1)

Page 72: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

62

into the pond influent. The addition of a flocculant of this type

affects a change in the effective particle size distribution which

defines the settling characteristics of incoming sediment.

Actual determination of the effects of a coagulant or flocculant

on particle size distributions should be based on lab experimentation.

Explanation of this procedure is presented in Ward et al. (1979). For

study purposes, and estimate of the flocculant effect was made as a

result of conversations the author had with Rick Ellwagner (1980), a

sales representative for the Tucson office of American Cyanamid. These

discussions produced the following estimates:

clay fraction -> reduced to 0.1%

silt fraction -> effective size increased to 0.1 mm.

The value listed for the unaffected clay fraction percentage is a con-

servative one, and should produce equally conservative peak effluent

concentrations. The amount of flocculant required to produce this

change was quoted to the author at 0.1 lb./ton of solids at a cost

of $.60/1b.

Results for scheme 3 runs appear in Table 8. Pond trap

efficiency for all routed events was extremely high, on the order of

99.9%. Only one of the events resulted in abrogation of the effluent

standard. For the above mentioned rate of application and per pound

cost of flocculant, the total cost for treatment of an event which

generates on incoming sediment load of 30 tons is approximately

Page 73: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

63

Table 8. Predicted peak effluent sediment concentrations for routed,selected Black Mesa storms l : minespoil with chemicaltreatment of inflow. 2

Storm Event, Peak Effluent Sedimentfrequency-duration Concentration, mg/),

5 yr. - 1 hr.

8 yr. - .50 hr.

10 yr. - 2 hr.

10 yr. - 24 hr.

15 yr. - .17 hr.

25 yr. - .50 hr.

25 yr. - 24 hr.

53.5

63.7

44.3

46.8

112.2

65.8

40.1

'Event sediment yields appear in Tables 9 and 10.

2Estimated effective particle size distribution: < .002 mm = .1%

Crop management factor, C = .63 > .1 mm = 99.9%

Precipitation characteristics for all stormsappear in Tables 4 and 5.

Page 74: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

64

Table 9. Predicted event sediment yield for simulated conditions atBlack Mesa, AZ.: 5-10 yr. return periods. 1

Storm Eventfrequency-duration

Event Sediment Yield, tonscondition 12

5 yr. - .17 hr. 88.5 12.6

5 yr. - .25 hr. 112.7 16.1

5 yr. - .50 hr. 133.5 19.1

5 yr. - 1 hr. 157.7 22.5

5 yr. - 2 hr. 132.0 18.0

5 yr. - 6 hr. 125.5 17.9

5 yr. - 12 hr. 121.7 17.4

5 yr. - 24 hr. 113.5 16.2

8 yr. - .17 hr. 101.8 14.5

8 yr. - .25 hr. 126.9 18.1

8 yr. - .50 hr. 161.4 23.1

8 yr. - 1 hr. 186.3 26.6

10 yr. - .17 hr. 110.9 15.8

10 yr. - .25 hr. 138.9 19.8

10 yr. - .50 hr. 174.8 25.0

10 yr. - 1 hr. 197.7 28.2

10 yr. - 2 hr. 170.7 24.4

10 yr. - 6 hr. 153.4 21.9

10 yr. - 12 hr. 136.8 19.5

10 yr. - 24 hr. 137.2 19.6

lYields for condition 3 are equivalent to those cited for condition 1,unaltered.

Yields for condition 2 are equivalent to those cited for condition 1,altered

2Minespoil material without treatment of inflow for original and alteredinputs, respectively.

Page 75: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

65

Table 10. Predicted event sediment yield for simulated contitions atBlack Mesa, AZ: 12-25 yr. return periods.*

Storm Eventfrequency-duration

Event Sediment Yield, tonscondition 1

12 yr. - .17 hr 114.0 16.3

12 yr. - .25 hr. 141.9 20.3

12 yr. - .50 hr. 180.2 25.7

12 yr. - 1 hr. 202.2 28.9

15 yr. - .17 hr. 122.3 17.5

15 yr. - .25 hr. 153.9 22.0

15 yr. - .50 hr. 193.8 27.7

15 yr. - 1 hr. 218.0 31.1

15 yr. - 2 hr. 190.4 27.2

15 yr. - 6 hr. 166.1 23.7

15 yr. - 12 hr. 139.0 19.9

15 yr. - 24 hr. 150.6 21.5

25 yr. - .50 hr. 218.5 31.2

25 yr. - 1 hr. 237.8 34.0

25 yr. - 2 hr. 190.3 27.2

25 yr. - 6 hr. 158.2 22.6

25 yr. - 12 hr. 159.6 22.8

25 yr. - 24 hr. 144.4 20.6

*Footnotes are equivalent to those cited in Table 9.

Page 76: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

6 6$2.00/event. Initial capital investment to cover purchase of the

application mechanism would be the only significant additional cost

required. The resultant reduction or elimination of pollution fines

would accelerate the expected amortization of this expenditure.

It can therefore be inferred that under the conditions applied

in this investigation, the addition of chemical flocculants is

absolutely necessary to approach a no-risk situation for fully contained

(no emergency spillway overflow) pond inflows. Of course, federal

regulatory policy has rarely, if ever, been directed towards a total

elimination of risk. The next few paragraphs address the consequences

of hydrologic uncertainty for pond performance.

Hydrologic Uncertainty and Implications for Pond Performance

In assessing the uncertainty in modeling any natural phenomena,

three sources must be considered: (1) uncertainty in model choice,

(2) sample uncertainty, and (3) imperfect understanding of the physical

process itself. Of the three, only the first two are partially con-

trollable by the modeler.

The models used for rainfall distribution and sediment yield

are both indicative of the uncertainty involved in model selection.

It is entirely possible that the SCS rainfall distribution scheme

utilized in the determination of rainfall excess fails to adequately

describe the temporal characteristics of southwestern convective storms.

Other distributions are available which differ in the timing and extent

Page 77: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

67

of the high intensity burst exhibited by observed storms. This could

have a significant effect on production of runoff and sediment yield,

because of the effects of varied intensities on infiltration rates and

satisfaction of prevailing soil moisture deficits. At best, these pro-

jected scenarios would only slightly reduce the measured pond effluent

concentrations, for their effects would be damped by the persistence

of the large clay fraction.

Another model application providing a source of uncertainty

for the study is that of the modified USLE (Williams, 1975). George

Foster, a hydraulic engineer with the U. S. Department of Agriculture,

who specializes in the use of soil loss equations, has stated that

Williams' (1975) formulation may not perform well in situations where

raindrop impact accounts for a significant portion of soil erosivity,

since this factor is neglected in Williams' equation (oral communication

1980). However, sediment yields computed via Williams' (1975) equation

are well within the range of 3-4 tons/acre observed by members of the

University of Arizona research team at Black Mesa (Fogel, 1980).

Model uncertainty also enters into consideration in the case of

the DEPOSITS sedimenation model (Ward et al.,1977b). As acknowledged

by its developers, the plug flow concept used for the routing process

is an imperfect device for modeling actual basin hydraulic and sedi-

mentation processes. The authors feel that a partial mixing model

would provide a more accurate description of pond performance but its

comparable complexity would be substantial. Inclusion of a partial

Page 78: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

68

mixing model could be expected to result in no more than a moderate

reduction in simulated peak effluent concentrations because of the mix-

ing which would take place between incoming sediment-laden plugs and

the better clarified water stored in the permanent pool.

Differences in the relationships describing Black Mesa sediment

yields for the present study and previous work by Fogel et al. (1979)

exemplify the uncertainty involved in the use of sampled data. Limited

data utilized in the prior study suggested a linear relationship between

event sediment yield and the product of runoff volume and peak flow.

This study, however, employs the non-linear relation between the afore-

mentioned variables cited by Williams (1975). Both the uncertainty

inherent in the limited amount of data available to Fogel et al. (1979)

and that residing in the infiltration data used herein make determination

of the actual sediment yield relationship difficult. Retention of the

linear relation for sediment yield computation in this investigation

would have led to inordinately high sediment yields and correspondingly

high effluent sediment concentrations. Due to the lack of adequate data,

the non-linear sediment yield relationship was assumed to satisfactorily

represent existing conditions at Black Mesa.

Uncertainty due to an imperfect understanding of the physical

phenomena is extreme for the examined processes linking rainfall with

erosion sediment transport and reservoir sedimentation. It must be

remembered that the precipitation events modeled in the study are

relevant in the statistical sense. That is, their occurrence is not

an established certainty. An idea of this variability is provided

Page 79: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

69

by computation of the probability associated with at least one 5 yr.

return period storm occurring in the next 10 years: Assuming indepen-

dence of annual rainfall and a probability of exceedance, p, equal to

the reciprocal of the return period:

PE = 1/T = 0.20

For a geometric distribution:

p (at least one 5 yr. storm occurs in 10 yrs.) =

1-p (no occurrences in 10 yrs.) = 1 - (1-p) 10

= 1 - (0.80) 10

= 0.89

Entire summers void of any runoff producing rainfall whatsoever

are not unheard of for the region. Conversely, storm sequences could

occur during which overlapping of storm routing periods might produce

emergency spillway discharges. Moreover, all individual inflow volumes

within the sequence could conceivably fall below that of the computed

design storm, resulting in simultaneous satisfaction and abridgement of

federal statutes. The stochastic nature of precipitation characteristics

such as intensity and rainfall depth should also be recognized since

their association has been shown by this investigation to greatly in-

fluence the predicted performance of sedimentation ponds.

Other variable factors which contribute to the uncertain com-

prehension of physical phenomena include soil erodibility, complex

hillslope flow systems, gully formation, and the degree to which natural

Page 80: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

70

in-pond aggregate formation affects settling rates. Vandivere, Davis,

and Fogel (1979) investigated the uncertainty involved in applying the

USLE to semi-arid surface mining sites. Results of hydrologic simula-

tion on Black Mesa spoils showed an extremely large variance in the pre-

dicted 3-yr. sediment yields. The USLE overpredicted the 3-yr. sediment

accumulation. This means that for the present study the basis for allo-

cation of dead storage for the pond entails additional uncertainty. Also,

variation in the particle size distribution with flow rate makes obtain-

ing a representative size distribution problematic (dilmoth, Hill, and

Ettinger, 1979).

Model Adaptability and Regional Bias

This model has been designed with general application in mind.

Use of accepted mathematical descriptions for the component parts should

assure it a high degree of adaptability to strip mine situations through-

out the semi-arid U. S. Its reliance on field data is minimal, there-

fore, a limited data base should not preclude its utilization. As pond

sampling data increases, verification of the model can be undertaken

along with regional optimization of parameter values. Lack of verifica-

tion notwithstanding, the model appears to perform satisfactorily in

generating sediment yields and simulating basin trap efficiencies con-

sistent with the imputed particle size distributions.

Inherent regional bias is evident in all phases of the model due

primarily to the simplicity afforded the present study in neglecting

winter precipitation and spring snawmelt. For application to areas where

frontal storm systems produce a significant portion of annual

Page 81: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

71

precipitation, runoff, and sediment yield, a winter precipitation model

coupled with a snawmelt accounting procedure can be added to the SCS

Type II rainfall model used herein for convective storm activity. Alter-

native sediment yield models can also be substituted for the one used

for this study, although regionalization of the parameters in the Williams

(1975) equation is fairly straightforward, given adequate erosion and

sediment yield data.

Regardless of the methods used for determining water and sediment

inflows to the sedimentation pond, the DEPOSITS routine (Ward et al.,

1979) can be applied with equivalent accuracy to any region, as long as

the modeled pond is designed properly. Provisions are available for

simulating chemical treatment, density currents and short-circuiting and

numerous withdrawal conditions. An additional attribute of the routine

lies in its ability to model reservoir deposition over time. This en-

ables its conjunctive use for purposes of hydrologic time series analysis,

especially where event sequences related to frontal storm systems are

simulated.

The application of the INFLUX program in its present form is

limited to small watersheds of 2000 acres or less where the bulk of

runoff and sediment yield is derived from summer convective storms.

Also, an infiltration curve based on representative field conditions is

required.

Computer time and consequent costs for the model are low. The

combined cost of INFLUX and DEPOSITS models for the 38 selected precipi-

tation events was under $1.50 per run, excluding print costs. This

should make the package more desirable to prospective users.

Page 82: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

CHAPTER 5

CONCLUSIONS AND RECOMMENDATIONS

A model has been compiled herein which enables simulation of

sedimentation pond response to inflows generated by selected precipita-

tion events on surface mined and natural watersheds. Procedures util-

ized in design of the pond reflect both the consideration of published

federal criteria and recognition of current engineering practice. In

the development of hydrologic inputs to the model, attention was directed

toward an adequate description of the physical characteristics of convec-

tive precipitation. Application of the model is limited to small water-

sheds where spring snowmelt is negligible in the production of sediment

and convective rainfall accounts for the bulk of annual runoff. The

model was shown to successfully reproduce observed sediment yields and

simulate pond effluent sediment concentrations which coincided with

computed trap efficiencies.

Simulation was undertaken for three watershed-pond conditions:

(1) minespoil material with untreated inflow; (2) natural soil material

with untreated inflow; and (3) minespoil material with treated inflow.

Results indicated that in every case pond discharge for conditions 1 and

2 grossly exceeded federal standards for concentrations of suspended

solids. Only when condition 3 was evaluated did a majority of routed

pond inflows succeed in meeting the established standards. This suggests

that in most semi-arid mining environs, current Federal guidelines for

72

Page 83: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

73

the design of sedimentation ponds are totally incompatible with the

attainment of established effluent quality criteria. Therefore, if the

logic of a national effluent water quality standard is upheld, mine

operators will have to rely heavily on additional measures in order to

both reduce erosion and create depositional opportunities for transient

flaw.

For the semi-arid regions of the west, precipitation uncertainty

makes it highly improbable that the healthy vegetative cover required

to forestall excessive erosion and sedimentation can be achieved without

an effective irrigation schedule. However, as the present study has

shown, maintenance of a good stand of vegetation by no means assures

compliance with published Federal effluent criteria. A crop management

factor as low as 0.09 which corresponds to irrigation-derived vegetal

cover proved insufficient in attaining Federal standards under prevailing

soil and precipitation conditions. The overriding factor affecting pond

performance is the character of the incoming sediment. In situations

where the clay fraction of the particle size distribution is as clearly

dominant as that observed in runoff from graded minespoils at Black Mesa,

attainment of the desired effluent sediment concentrations will definitely

require the addition of flocculants or coagulating agents to pond inflows.

Application of chemical additives results in a significant increase in

pond trap efficiencies due to the effective redistribution of particle

sizes. Acceptance of this prerogative would assure compliance with

federal regulations at minimal cost. Furthermore, the initial capital

Page 84: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

74

expenditure required would likely be offset by a concomitant elimination

of federally or state imposed pollution fines.

In an effort to clarify the effect of precipitation uncertainty

on projected pond performance, an alteration of documented sediment and

management characteristics was applied to condition 1. Analysis of sub-

sequent runs revealed a tendency for high-intensity-short duration storms

accompanied by lesser inflow volumes to exceed effluent limitations. Con-

versely, the low intensity-long duration storms associated with greater

inflow volumes which are exempted from federal regulations consistently

displayed pond effluent concentrations within the allowable limits.

Thus, the use of the 10 yr-24 hr. rainfall event as a design criteria

for sedimentation ponds which do not operate according to a total con-

tainment policy appears to be unsubstantiated. The obvious dependence

of predicted pond performance on the characteristics of individual pre-

cipitation events along with the uncertainty associated with the occur-

rence of these events offers ample justification for the consideration

of hydrologic uncertainty in the sedimentation pond design process.

Only then will designers and regulators be able to assess their position

in the light of hydrologic reality.

It must be recognized that even conditions existing on reclaimed

watersheds are dynamic over time. Changing cover densities and the ini-

tially accelerated transport of fines could affect a substantial reduc-

tion in the influent clay fraction for a large portion of the projected

pond lifetime. This decrease, however, would have relatively little

influence on observed peak effluent sediment concentrations.

Page 85: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

75

The poor quality of pond effluent simulated for undisturbed

natural watershed conditions calls into question the applicability of

current federal statutes pertaining to pond effluent limitations to the

semi-arid regions of the western U. S. Ambient sediment concentrations

measured in eastern streams are indicative of the comparatively lush

vegetative cover and overlying canopy which effectively reduce both rain-

fall energy and the overland transport of sediment. Typically, semi-

arid zones exhibit sparse vegetal cover and a restricted protective

canopy which effectively reduce both rainfall energy and the overland

transport of sediment. Typically, semi-arid zones exhibit sparse vegetal

cover and a restricted protective canopy, conditions which favor greater

erosive activity, transport capacities, and consequently, higher ambient

stream sediment concentrations. In addition, releasing effluent at con-

centrations considerably below that of ambient levels may lead to changes

in downstream channel morphology as adjustments are made to accommodate

the new flow regime. It seems sensible, therefore, to encourage the con-

sideration of regional diversity in the establishment of effluent criteria

for state regulatory proposals.

Further research is required to assess the probabilities associ-

ated with the occurrence of rainfall events incorporating specific

intensity-volume relationships. Risk analysis could then be attempted

so that the effect of hydrologic uncertainties on the design and expected

performance of detention facilities can be concretized. Since the highest

intensity storm events have been shown to produce the highest peak efflu-

ent sediment concentrations, the critical precipitation scheme for

Page 86: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

76

semi-arid regions would most likely include a close sequence of these

events. Consecutive days of runoff-producing rainfall would lead to

an overlapping of routing periods. Greatly reduced detention times for

the routed inflows could then result in pipe discharge of heavily

sediment-laden flow through the principal spillway and eventual emergency

spillway overflow.

Another subject which requires investigation is the principle of

total containment regarding all inflows less than that corresponding to

the computed design event. Although this design would produce increased

detention times for runoff generated by the high intensity precipitation

events, the critical scenario outlined above still applies because storage

capacity is limiting and the potential for emergency spillway discharge

remains. Moreover, despite reassurances to the contrary (Nadolski, 1980),

the author suspects that resuspension of deposited sediments by the accom-

panying horizontal dewatering device is likely, therefore increasing the

risk of violating water quality standards.

The model described in this study maintains a slight regional

bias, but can be easily adapted to other areas and conditions. It re-

tains enough flexibility to enable its use in the analysis of hydrologic

time series and is also relatively inexpensive to run. Because of its

lack of reliance on recorded data, the model is applicable to ungaged

watersheds, thereby increasing its range of utilization. Validation of

the model in its present form is necessary so that expansion of the data

base for semi-arid lands can be realized. This is important if the

western states are to draft mining and reclamation regulations which are

relevant to native conditions.

Page 87: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

APPENDIX A

SEDIMENTATION POND DESIGN

A comprehensive discussion of the procedural elements contributing

to the finalized sedimentation pond design is presented below. The design

process is subdivided into three components. Determination of water

storage volume is followed by a description of the procedure used to com-

pute the required sediment storage volume. The basin dimensioning is

outlined and the operating characteristics of the reservoir, in the form of

stage-area-storage volume and stage-discharge relationships, are formulated.

A listing of DEPOSITS output resulting from routing the design storm

through the basin not only confirms the legitimacy of the design, but also

provides the reader with a feeling for the power of the model in describing

basin performance. References are cited in parenthesis adjacent to the

source material used for design purposes.

Basin Capacity and Dimensioning

The procedure outlined here for calculation of basin storage

volume is presented in Ward et al. (1979). Based on an approximation of

inflow-outfow relationships by triangular hydrographs, the required design

storage volume is given as:

V = 0.0413 tb. (q . - q )s 1 pi po

where

77

Page 88: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

78

Vs = water storage volume in acre-feet,

bi = base time for the inflow hydrograph in hours, and-

q .i,qpo

=p, peak inflow and outflow rates, respectively, incubic feet per second.

As pointed out in Chapter 3, the design method used for com-

putation of the inflow hydrograph was the SCS triangular hydrograph method

described by Kent (1973). The geometry of the hydrograph dictates the

following relationship for its time base:

ADT = — +Lp 2

where

AD = duration of rainfall excess in hours, and

L = basin lag time in hours.

A 30 minute storm duration was assured representative for design pruposes

and the lag time was calculated to be .17 hours, therefore:

0.5+ 0.17 = .42 hours

tbi

= 2.67 x .42 = 1.12 hours.

Runoff was determined with the use of the SCS rainfall runoff

relation (Kent, 1973):

T=

and

2

(P - .2S) 2

= (1) + .8S)

where the maximum potential retention, S, is related to a curve number

index which assesses the runoff producing potential of the watershed.

Page 89: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

79

Based on soil type, vegetative cover, and antecedent moisture condition,

the relationship is expressed as:

1000S = - 1CN 0

where CN represents the watershed curve number.

For average antecedent conditions and characteristics common to

graded spoil material at Black Mesa, a curve number of 89 was estimated.

Thus, a value of 1.24 was calculated for the variable S. The resultant

runoff volume was computed as:

(2.10 - .2(1.24))2

Q 2.10 + .8(1.24)

= 1.12 inches

Next, using the SCS equation for peak runoff determination,

KOA

where

. = peak inflow rate in cubic feet per second,

K = a watershed parameter, assured equal to 484for chosen hydrograph geometry,

Q = rainfall excess in inches, and

A = drainage area in square miles.

Substituting for equation unknowns:

Page 90: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

80484(1.12)(.078)

qpi .42

= 100.83 cfs

The remaining factor in the storage volume equation is the project-

ed peak outflow rate from the reservoir given by Ward et al. (1979):

q = t . q ./tP0 bi pl bo

where

q = peak outflow rate,po

tbo = time base of outflow hydrograph in hours, and

tbo = (3 x td) + tbi

where

td = required detention time in hours.

As stated in the federal statutes (Federal Register, P. 15400), a theo-

retical detention time of 24 hours is the minimum acceptable period for

pond design, therefore:

tbo

= (3 x 24) + 1.12

= 73.12 hours

Additionally: q = (1.12 x 100.83)/73.12po

= 1.54 cfs

Inflow runoff volume for a triangular hydrograph is:

V = 1/2 q . t .Pi bi

where V = inflow volume in cfs-hr.

Page 91: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

Thus, - 100.83 X 1.12 V 2

81

= 56.5 cfs-hr.

Converting to acre-feet:

V = 56.5 cfs-hr. X ac-ft. 12 cfs-hr.

= 4.7 ac.-ft.

In lieu of the aforementioned storage equation, empirical relationships for

inflow and storage volumes on page 57 of the design manual are used:

S/V = .83 for q= .015

This results in a required storage capacity, neglecting sediment storage,

of:S = .83(4.7)

= 3.9 ac.-ft.

Sediment Storage Volume

The level of the lowest dewatering device is required to be no less

than that corresponding to 100% of the 3-year accumulated sediment storage

volume. Since the regulations state that the Universal Soil Loss Equation

(USLE) and an appropriate delivery ratio must be utilized for this calcula-

tion (Federal Register, p. 15400) its application to the hypoethetical

situation is now developed.

The USLE is an empirically derived formula based on thousands of

plot years of data including natural and simulated rainfall conditions for

Page 92: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

82

a variety of cultivation and management practices. Recent accumulation

of data has provided for an extension of its usage to surface mine con-

ditions (EPA 1977). The USLE is expressed as follows:

where

A = RxKxLSxCxP

A = annual soil loss in tons per acre,

R = rainfall erosion index,

K = soil-erodibility factor,

LS = length and steepness of slope factor,

C = cropping management factor and

P = erosion control practice factor

The equation is dimensionally correct and the reader is referred to the

current USDA-ARS users guide (Wischmeier and Smith, 1978) for a detailed

discussion of factor development and application.

Selection of parameter values was predicated upon 2 assumptions:

1. Only post-mining conditions were to be examined and no unnatural

disturbances or hydrologic inputs were affected.

2. Essentially no vegetative cover is established during the first

year and only 10% cover is generated over the next two years.

Estimated factor values and justification for the choices made are:

R =

K =

LS =

30 evaluated from isoerodent map for Arizona(SCS 1976)

.35 estimated value for Black Mesa grade spoilmaterial (Fogel et al., 1979)

1.26 derived from estimated average slopes of 6.7%and a slope length of 250 ft.

Page 93: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

83

C = 1.0, .45 estimated values for zero and 10% coverrespectively (EPA, 1977)

P = .35 estimate based on practice of contourgouging or pitting currently employedat Black Mesa (EPA, 1977)

Thus, yearly annual soil loss per acre for the 3 years is:

A = (30 x .35 x 1.26 x 1.0 x .35) .9 = 4.17 tons/acre/year1st year

A = (30 x .35 x 1.26 x .45 x .35) .9 = 1.88 tons/acre/year2nd, 3rd years

where .9 represents the assumed delivery ratio defined as the ratio of

sediment delivered to the watershed outlet to the gross watershed erosion.

Over the entire extent of the 50-acre watershed the expected 3-yr. sedi-

ment yield would be:

50 ac.(4.17 tons/ac./yr. x 1 yr. + 1.88 tons/ac./yr./x 2 yr.)

= 396.5 tons

For storage to be allocated in the pond, the sediment yielded must be

converted into a volumetric quantity reflecting the properties of the

incoming sediment (Federal Register, p. 15400). An accounting of re-

corded incoming sediment size fraction distributions for the J-3 experi-

mental watershed is offered in Table A.1. Particle diameters in

millimeters for the size fraction descriptors appearing in the table

are:

sand:

silt:

clay:

coarse-medium:

very fine:

> .125

.063 < x < .125

.002 < x < .063

< .002

Page 94: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

84

Table A.1. Size fraction distributions for sediment production:J-3 experimental watershed, Black Mesa Mine.

Storm date Size fraction, % of total

Sand Silt Clay

Coarse-med. Very fine

7-11-77 14.5 2.3 39.1 44.17-11-77 1.9 2.0 47.4 48.77-11-77 2.9 2.0 45.4 49.77-11-77 4.3 2.2 48.0 45.57-19-77 2.4 3.0 48.9 45.77-19-77 5.5 2.9 44.8 46.87-19-77 5.7 1.9 46.4 46.07-19-77 8.0 2.7 43.8 45.57-22-77 2.9 1.8 36.6 58.77-22-77 2.4 1.5 40.1 56.07-22-77 4.3 2.9 36.7 56.47-22-77 2.7 .9 28.8 67.68-05-77 5.4 2.4 41.1 51.18-05-77 5.7 2.2 41.2 50.98-05-77 11.4 6.3 24.7 57.68-05-77 3.9 1.7 40.7 53.78-12-77* 1.6 2.3 10.3 85.88-12-77* .8 .9 4.7 93.68-12-77* 2.8 1.6 2.5 93.18-15-77 .5 .9 51.7 46.98-15-77 2.2 1.9 49.8 46.18-15-77 .5 .8 36.9 61.88-15-77 .7 1.2 44.0 54.18-22-77 .4 .8 52.3 46.58-22-77 .4 .8 51.1 47.78-22-77 2.6 2.1 45.5 49.88-22-77 2.9 1.8 51.3 44.0

Average 3.9 2.0 43.2 50.9

*Unrepresentative, assumed outliers.

Page 95: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

P age Missingin Original

Volume

Page 96: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

86

An obvious inconsistency appears in the computed average size fraction

percentages when matched against those used below in figuring the sedi-

ment density value. This is the result of assumptions made later in the

study regarding the possibly anamolous character of the indicated mea-

sured sediment distributions for the indicated storm events. This, how-

ever, only lends to the conservative nature of the pond design.

The density of stored sediment is assumed to accrue from conditions

outlined in the DEPOSITS manual (Ward et al., 1979) for "reservoir opera-

tion type III" representing a normally dry reservoir:

W = WP +WP +WPcc mm s s

where W = density of sediment in pound per cubic foot,

W ,W Wc m s unit weights for clay, silt, and sand fractions,respectively, for a dry reservoir in pounds percubic foot, and

P P P = proportion of each of the same three abovec m smentioned constituents, expressed as a decimalfraction.

Therefore:

W = 40(.553) + 72(.39) + 97(.06)

= 55.7 lb/ft 3

Finally, the required sediment storage volume is:

ft 3- ac.-ft.396.5 tons x 2000 lb/ton x 55.7 lb. x43,560 ft j = .33 ac. ft.

When the required water storage capacity is combined with the sediment

storage volume, the total design storage capacity results:

Page 97: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

87

Total storage volume = 2.9 + .33

= 4.23 ac.-ft.

Stage-area-discharge Relations

Assuming the reservoir bed is constant at stage 0.00 feet, the

surface area and storage capacity are listed as a function of elevation

above the bed in Table A.2. The required sediment storage volume evi-

dently corresponds to a stage of .97 feet. This stage height was desig-

nated as the location of the first de-watering orifice set. Results of

earlier trial designs suggested a riser crest elevation of approximately

9.0 feet. A second orifice set was also added at a stage level of 6.0

feet.

With the riser configuration set, the only information lacking

in the determination of the stage-discharge relationship is that of the

head relation for pipe-full flow as described in Chapter 3. This entails

completion of a preliminary basin dimensioning and embankment design so

that the positioning of the culvert and its outlet elevation can be ob-

tained.

Given the computed basin capacity, dimensioning of the basin can

be undertaken. Assuming a trapezoidal configuration with bottom width,

b o , length, 1, and design depth, yd , accompanied by embankment slopes of

1 vertical to 2 horizontal (Federal Register, p. 15400), the following

dimensions have been chosen:

yd = 15 ft.

bo

= 50 ft.

L = 285 ft.

Page 98: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

88

Table A.2. Rating relations for final pond design.

Stage(ft.)

Surface Area(ac.)

Storage Capacity(ac.-ft.)

0.00 .33 0.00

0.97 .35 0.33 Trapezoidal Basin:

1.00 .35 0.34

2.00 .38 0.71 Surface Area =

3.00 .40 1.10(bo + 4)L

Y

4.00 .43 1.52

5.00 .46 1.96

6.00

7.00

.48

.51

2.43

2.93

Storage Capacity =b L + 2 2 L0 y y

8.00 .54 3.45

9.00 .56 4.00

9.50 .58 4.29

10.00 .59 4.58

11.00 .61 5.18

12.00 .64 5.81

13.00 .67 6.46

14.00 .69 7.14

15.00 .72 7.85

Page 99: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

89

Check for length to width ratio at design depth:

b15 • 50 + 2(15 x 2) = 110 ft.

L/b15 • 285/110

• 2.6 which satisfies the suggested ratio of greaterthan 2.0 (Ward et al., 1979).

Allowing 5 ft for embankment freeboard and emergency spillway section:

H = total height of embankment = 20 ft.

Minimum top width must be >((H + 35)/5) (Federal Register, p.15400)

(H + 35)/5 = 55/5 = 11 ft.

Total width of the embankment = 2(2 x 2) + 11 ,.- 90 ft.

Assume a culvert length = 90 ft. and culvert diameter = 8 in.

Optimum critical slope to produce

pipe full flow (S )c op.n2= 111 (Portland Cement Associa-

D1/3 tion, 1964)

where n = Manning's roughness factor, assumed = .024

D = culvert diameter in feet

(S c ) op for an 8 inch diameter corrugated metal pipe = .079 ft/ft.

Fall height of conduit over projected length = 90 x .079 = 7.1 ft.

Assuming a distance of .5 ft. between the center of the lowest

dewatering orifice and the center of the culvert at its connection with

the riser,

Page 100: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

90

head for pipe flaw condition = 7.1 + .5 + 8.03 + h

= (15.6 + h) ft.

where h is the head on the spillway crest.

Stage-discharge relations based on the hydraulic equations given

in Chapter 3 for the finalized pond design are presented in Table A.3.

This information, along with the stage-area data appearing in Table A.2,

provides the primary inputs required by the DEPOSITS model for routing

of sediment laden water through the reservoir. An output generated by

the DEPOSITS routine for the 10 yr-24 hr. design event is listed in the

following pages. Variables appearing in the output are defined in the

DEPOSITS "glossary of terms" which is located in Appendix C.

Page 101: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

91

Table A.3. Stage-discharge relations for final pond design.

Stage(ft.)

Q orifice(cf s)

Q weir Q pipe

(cfs.) (cfs.)total(cfs.)

head1 a1 head2 a2 head a H

-pipe

0.00

0.97

-

-

-

- -

-

-

-

-

-

-

-

-

-

-

0.00

0.00

1.00 0.03 .05 - - - - - - 0.05

2.00 1.03 .32 - - - - - - 0.32

3.00 2.03 .44 - - - - - - 0.44

4.00 3.03 .54 - - - - - - 0.54

5.00 4.03 .63 - - - - - - 0.63

6.00 5.03 .70 - - - - 0.70

7.00 6.03 .77 1.0 .31 - - - - 1.08

8.00 7.03 .83 2.0 .44 - - - - 1.27

9.00 8.03 .89 3.0 .54 - - - - 1.43

9.50* 8.53 .09 3.5 .06 0.5 2.44* 16.1 2.51* 2.66

10.00 9.03 .09 4.0 .06 1.0 6.91 16.6 2.54 2.69

11.00 10.03 .10 5.0 .07 2.0 19.54 17.6 2.62 2.79

12.00 11.03 .10 6.0 .08 3.0 35.90 18.6 2.70 2.88

13.00 12.03 .11 7.0 .08 4.0 55.28 19.6 2.77 2.96

14.00 13.03 .11 8.0 .09 5.0 77.26 20.6 2.84 3.04

15.00 14.03 .12 9.0 .09 6.0 101.56 21.6 2.90 3.11

*Transition to pipe flow occurs at stage = 9.5 ft; orifice dischargeat stages > 9.5 ft. are reduced by 90% due to submergence of orifices.

Page 102: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

a.

oQ

LIJ

0

o 00

0 •

1-• 0a 0

Z1.•

CJX

•.-4

Li 0 tn.2 0QC

0

o1-•Li! •4,1

oX 01•11 •

511

t... 0..-1 0

0 •-1 •LI. .--I

0

in 0o.cc

0•= 0-r

a t-

UIz 41 •

U. ••1

n••VI

4. 0UJ •

Ut00

4.3

92

..doLC 0 •ILI tet NI

03. ect

Page 103: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

DC 000 00000000000 000eCD C) C) C) C) C) C) C) C) t) Ci C) C) CD C) CDCD C) C) C) <D C) C) C) CD C) C) C) C) CD 0 CD 0.J • • • • • . • • • • • • • • • • • •U. 41 Ir. 41 41 III Ul 41 41 g4 0 al al al 41 Ul 41 41 41O. 04 N em OW VW 04 04 C4 P4 VW ow 04 1,4 04 04 04 NJ 0,1

In

• 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0••n 00000 00 3 300000000 3

.4 • • • • • • • • • • • . • • • • • •U. gl g, al 41 41 41 41 4:1 41 41 41 *1 41 41 41 u4 41 41r- *I *J VO 04 es., Al 04 04 CQ Al 04 04 NI VW 04 04 04 "4

CD

93

• C) CD 0 C) C) C) Ci C) C) C) C5 C) C) C) CD CD CDCD .n 3 C) 0 C) en o o o o o c) o o D 0 4).J • , • • • • • • • • • • • • • • • •LL ••• 41 41 11 IN m 41 41 41 41 41 al al 0 41 JA.('• NJ 1M NI 04 04 N "4 Al ra 04 N N c* *I 04 04 04 CS,

-•3 0 0 0 0 C) C) 0 0 0 LD C5 0 C) C) C) C)C) m C) C) 0 7 7 7 0 C) 0 0 0 3 0 D

Lu.J • • • • • • • • • • • • • • • •

••n• t11121‘i, na, gl Mg, L/1 UN

▪ 04 04 *4 NJ mi C4 04 V 04 04 04 04 04 (MN 04 NJ 0400 00 •

O

0 0

•••ON..000 000000 0 .)00 3001.41 sQ r- Cr 7 D CD C) C) 41 0 C> C) C) 15 CiN4 U. 41I U. • • • • • y • • • • . • . • . •

.4 r- C) -4 NJ 0.1 UN .0 a) Cr Cr •4 CQ sn 40 41V) 41

Page 104: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

- 0 0 al ry 4'V/ 0 .3 0 In 4u. • • • • •...) 0 0

r-cs.LiJ

0 m1— 0 Cr O 0 0.

11/ U. • • •••• 0 6.4 •••I N

1.1.1

0000 000000000000000 0 ol m m m en m m en m m 0 co cO a) cO co0 CrOal a-s0.r.al0.nulONN fs ty Nr-0 eu .7 0 0.0 en try r.• 0• • • •0 • y

.-• vs N esi 4 O. en e-WI• Q.•

.1'•

O.• 1. CI` N• In..1 1. e4 N In M 1 47

.7 en 0 co r. In 0 Crin .0 r• 0 N 1 .0 0• • • • •

- .1

•NI

•f4

In In e4 0 Inœ r.• 0 al In N 0 0•

In• •

4%

• •

Is•

02 co 0.

Ps ir al m m... . .7 .0 0• • • . •In 1/11 NO Is t-

0 c0 .0 4 .4Is 0 0' 0 .-I

• • • •N hl IV In In

co .-0 NI NI ena)

• • • • •

Cr ni mm-4 I.

94

cn00000 00 0 00000 30000.x...100000003 3 300000000uicco0003000 000 .3030000

cdm ,g1 al CO 0 M c0 ,r .0 co Q. Is cy. esj.4 41E In 1 7 7 4 u".. Itt LI1 43 .0 0 .0 Pn

• • • • . • • • • • • • • •

.1J...000000000000 00 0 300-Di— 3 30300 3003,3000 3030a .A.0 r.- 3003000000000300i—....0 ,7.0000 -1.3030 al0.3 3000cn • • . • • • • • • • • . •

0 .1 N en 4 Les 0 r.. cO Cr Cr 0 -4 NI PI 7 al. . 4 . .1 .

Page 105: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

95

I- /-• I- 1- I-U. U. U. U. U.

•I I I I I114 U., Iff LU U4 . -I ...1Cc CC CC CC CC 46 an .... n .... S. in vl v., zÇ) ro c., Li C., I- U. U. CD CD CD CD Cc cc cC 0. a .4 •IL .4 .4 U. V V 2 2 2 2 2 2 2 I-

al 0 .0 a.n u.% AN 0 Cr . 0 M I.-m ..,1 en ..1 tet 0. a) .0 0 07• . . . . • . In UN •

4.• 4. 4' 7. o riis m en .NI r•0 r rfl m N. ..1- NI. to o

o• LU coen .

• • • • • • •

In O•

Sr r-ru n.j

aa

Page 106: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

96

W m

J00000.4 m0ONWIMMW, WW0••1W 0 4.0.1.M0..MTNNMW,g000MSU. = d1M00.1,041-1JMNIMOPOMW m .0ftNINNI ......... 000

IN NI NJ ftl MINN NI N1Nft NM hi NO NI

LUwoo

M1,41014100 0000000000000000,43AMN

W 00ftlaNC1WANT

meINM

00000 n n ^ CP NW 41 n MOM N =0.7.001• n 0 0 0 0 ,110A M 01, 41 N O. .0N =M WONP.0 ...r w• • • • • • • • • • • • • • • • • • • • •A. 0 m00000 M 7'- N101.4.4.‘0.0.0MP•=M0.0.0.* W* 441

= m.100.4.1,11,1,101.7.MOMMWM40M M0.0M/M•00M,AM0.00 ...... ft

Wegre“ooelPeeeeoweee•ee.

M W m0M.S.,MMMM,819,40WMODOCOaa0 0

4,11

WW 0 m=fttnO.MILIMOMW,COOM•OPMMN0M0.10 .4 M0,00010.0.CON..041AMM000.0.C M ,,os4w0e.o•eow0e00....o..N mN14112M0.0.0.0'0.0•0.0.0.0.0.0.7.0.WCO

0

C W

o Mn

VIZ n

* t"00000 aCP<D0 r- men mil,:* ce.-1m,ANCP-4,44...000,N.4..ocoON.r.omocm* m......................• Z n ..... eqcmcmcsiNmmmmm.4.4.• w

O

Ui

W mUlMNWO.WWW1,.rn ,01.•0.MMMOM1,0nNft▪ N ,l1M.P.0 ,0 ,0 ,04 ,04041Mft00.VDM.,SSZ 61. • • • • • • • • • • • • • • • • • • • • • •

O MMNft NI NI ftVftftftNNINIMMMM,MMVI m

O

m0 v103000030000.130 0 000000 3

wcoN (5 0 0 00003003003 -30 -30W w••••••••••••••••••••••Z ttI ts00000000000 .)• Or- r -i

W L10000300003300033)003 30WNJ7.0ONS ,OCOON -7 .0M0N.743MONJ

m Coeo.0..soolko..o.elkoves▪ m MM-OMMNINNANNIMMM.IMST-4.

Page 107: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

OOOOOOOOOOOOOOOOO 000000 OOOOOO 00000

02.4.4.4•N41M4ONMIMMC30.4.e ^ 4,434. 0.1010.6401,.. 04,.ONN 0.40NM4)(113. 0' MWC.9MMJNI M 9 MMIFJTO3Tn 41.1mmpim,,0> 434CONN.MlfsNTOMCP.ONIC041040..0 MONOC4 04411.1r40.1,40m -4M4 4.0JPJ443 LOtOMMNNM.ImMM-40000-PTPCPCP4030,P1MMM7 0 000 00 0 0 00 0 00 0 0000 0 0 00.04 0. a O. 0, 001mm-.. mNI CU nj ,4 P4MJC11 MJNO4 MIN ININNN04 OOOOOO MINMJ M3 MI MJ

OOOOO....... ........ ......0

0000003000 0 00000000 0000 0 000000000

CP 9.4 P. 04 r- 04 r- .4 41 CP M NO m 43 cg -4 m 41 r- cr c) N m ,r ,4') .0 T4 r- r.. r- mm C. 04 43 CP CO 4D CD .1 -00 mon (a on 43 0 .1 .00' V4 40' Og 4, C( .4 CO or un n cs,• • • • • • • • • • • • • • • • • • • • . • • • • • • • • • • •300 .4 or o.1 ou mm .t un un un 44944D D Pft Pft r- o co cc. cl. a C. 33 0.4.4

curuniN ............. 00000003C/0' 4c0..................................404comm omm en cOnnnww49 4 pouslcca.o./...

09 T4 CO en po un .4 p. on c. un .4 N M C. on .4 IS M 43I •4 02 4. 0.0 04 4 -#0 .0 mcp 43 n T. p. '04 43 41 U1 4. m 04 VU C13 .4 .4 C) 00 cr C. CO 32 r.. r- r- .0 41• • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

CD C3 CP DP CP n CO r• t- r-49 940 q) Un n 41 .7 m m 04 .4 04 .4 .4 -40000'CD <D C. C. C. C.43 N O' M 74 4, P. CP .0 RI ul r- c. .4 .1 r- a. J4 n 41 NO. -4 JI 41 P. CP .4 04 43 CO• • • • ••• • • • ..• • ..• • • •...• • • . ••• • • •J.4.4..1' 414,4141141,443,043.40n 0-- PJP.P•10.3 034133.0.0.0.0.03000

-4.4.1.4.4J4 .4 .4

.4 7 00' 9 4 PJ 43 0 41 .7-7 m m '1 .4 .4or 3 0' c. 4 0 NO 41 V1 NP 74 T4 N .- C) 3 C.44 4 menmenrIm 74 erwmmenrnm 74 441-4 ,4 , ./ NCO MI M4 MI M1 V N N N rs)N• • • • • • • • • • • • • • • • • • • • • • • • • • • • •

-4

Ci C) CI Ci 0 C) C) 3 3 0 3 0 ,3 C) C) 0 0 C3 CD 3000003030000003 03000 C3 3 0 0 :3 -3 C) C) 0 3 3 '3 a c) a 0 7 7 CD 7 C) C) 3 ,7O 7 7 7• • • • • • • • • • • • • • • • • • • • • • • • • • • • •C) _D Ci 0 Cf .2 00 000C3 Ci C) 03 C) O .3 0 30 0 3 00 CD Z) 0 7 CD 0 3 0

3 000000 000 0 00033303000003003003300.09 p ou .s. n 4 3 o.4 .1.,0ccON.7.0 9 ON.7 ..04.Do.0.7.0nON747 -00 ,4

• • • • • • • • • • • • • •.11. 41 41 4, 41 J9 0.09 43 rn N 1.n P. ...z333.13 4 DC. 7 7. CP aa.ac) 0 Z3 M •4

97

Page 108: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

4fl 0'04)ki, Cd030.1.0.0.-404,40...ea,a140 .r.c.Oulc041) .) eM 4Do) NACPmmmoW0.0P40. .0m0WWJC>0. 03P-0 ,04,slCMP- 11,e..4CD434pmNOCPW.4)Ame4.-.10mr-43rn..00.DP.mmiNcumpammj ...... 000000000'0. 0' 0' 0'M0)03m.,14144 0,4141414, 414114141404040401,14141414141444.1 -441N0101NNNN41 44 414.1414.14.141N414'J4.14141014.10N044.104N410,14.141NN,14141NN

000000 0000c3 000000 0 000 00 000 000 00 0 0

4) C).1.1 P4 OM.) 0 .. Oq W4 P. CP C3 v. memernea c) cp P. .4. ...I 43 1.4 44 P. 44 P4 0441 CP ea 471 CP 04 14 OD 0. 111D .. oc •1 ,r rn 0 01 OP 04 ,r P. C2 111 14 /0 .1 (0 .0 /0 .4• • • • • • • • • • • • • ••• • • • ••• • • • • • • • ••••

v. v4 .4 04 04 401 e er,-7 ,r WI oN 41 4) 4) 43 P. p. P. P. '0D o) CP Cp cp CP CD C) C) CD -404 04 P4 44 04 04 V4 41 04 04 04 04 04 0 04 04 04 04 04 04 04 0, 04 04 04 04 04 1,1 tv 01 (01

41 41" e, .4 CP 0 42 A C ('JO CP p- 04 44 CP OD 43 e 04 c3 o) P. A-7 NI -4 0' a:I00001.n Pn p. p. P. P. P- 4) .0 .0 40 .0 .0 key al UN 14 14 UN 111 ,r ,r ,r 4,41(41R1...••••••••••••••••••••••••••••••

.....

r4 ,r CD 1.4 e C) .0 04 CP 111 04 D41 04 UN -4 CC 47 44 04 40 .4 03 ,r v. CD UN CP 43

41 11 40 ,r e e 40 04 4.1 v4 4. -4 000CO CP cp o' D Da) c0 4) 4) 0 41 41 ,r• • • • • • • • • e • . • • • • • • • • • • • • • • • • • • • 0 •

P4 P4 04 0.4 44 44 P4 04 04 P4 04 44 P4 P4 44 04 4) 4) 42 •D 4) -0 .0.0.0.0.0 4) .0 0.0 .0.0 .0 .04) 4) 4)

13/0/0020.4).4)43.1. 4r141. m/44,4104414.14.1 N.4.1-4.41

0414. 10W04.140 ,0W 0 4140/0 00JNS -0 /000444,0/00414.,0023414.

4. • ... ow...... 044J N01.14141404, 41 01 -0 .7 ,7 14 1141 ^1443'0 0 4 0N.-1

co V:. P4 43 a, fl '? m ,4 44 -4 3 OP CP P4 10 rn -4 C) c0 0 a, .740 ('4 c, o

... .4 •4 C) C) C5 .) C) C5 C5 CD CP cp CP cp cp cp op o) c0••••••••••••••••••••••••••••••••• -4 .

00000000000000007 00030000000000

03,01000333,000 0 0 0)000,010 0 00........ ..... ....................

0000300000..10000-1000 000 0 000000 3

)000 ,, ,D300000000 000 3 0000000,04

4.43=0P4-r-Om0,14. -ODO ,J1-0mCD P4.4. Z -00P14.0,00(.44..0=

.4 4 -4 MA MA 04 MA MA WI 40 OM 41 40'? r ,r 41 41 41 al 41 43 0 0 .0 .n N N N 44

-4.444 ..... .v4

98

Page 109: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

OnNMED.WCP.4 41 74NN m CMNONCPMCP4SMMS4ON nuM4c14untrotaDOEmNmuna. mi s P-0140.N40.N4•CuN.rnm00.WP-N—OuNSM N m m 0 0. 0c01, 44onsmmpu m m0

..... ..... P-44444444444444NNMJN.NANNNNMN(MNINJN 1NINNNINNNN4NJMNNNNINCNIN ryININNMSOJNNNftCNANCNirsI NNNNNNNNeNINN1NN(N, rym

<D 0 CD C) CD C) CD CI 0 CD C) CA 0 CD 0 CD C) C2 CD CD 0 CA CD C) CD 0 C) C) C) C) CD 0 CD

rn ,4 UN W Og 4 CP cm Mo n 4) r. m m m m m r. PP 40 4) 4, 41 .7 on p4 F. 0 CO GE, .0m 43 omun r- N .7 43 c0 0N .7 ,0 o ON qr GIDON P-• • • . • . • • • • • • • • • • • • • • • • • • • • • • • • • • •'.4_5 P4 P4 04 1,J 04 M M mm .7 .7 .7 .7 ,r un un 41 41 41 43 4 4) 42 40 P- P- r- PPMMMMIviMMMOmM mmMoIMMmMolMMMmMMmmMmmm m M

nM4,400CPMp4OPPPZSM-10 CI MMIDCPO4041SP4M0MMMIMONNNIMINNAN ....... 0000000 0' 0.0.0. 0 0. 0' 0'.. oeogegoo.“.4rO o do evesoOoogrOel000.

M O c0 UN si OP P- 4r •4 CP 4) 41.cp .0 S P40' P- 4r P40. r- .7 P40 p- 4N 04 CD a:, 41 M,r 04 p, P4..4 p4 .'40 0000'CD Cr. CP CP aDa202 cooDp. P. P. .0 4D .0• • • • • ••• • • • n • • • • • • • • , • n ••• • ••• •4) 40 .0 4D 40 43 .0 a 42 42 40 40 '0 G 42 .0 0 4N MN WN .n UN UN 41 UN UN .41 UN UN dl UN 41 .N

-4000000.0.0.0'CD CD CD CD CP 1. CP CP CP OP .0 a3 aD CO a3 PP PP P. P. p. P. .4, 42 4, ,0 4,'0 00 0 r4 4r 43 P. cp MO 41 PP Ce. ,4 M .CN P. Ce. M un P- 0. ..4 on un r- cr r4 en 4, P.. CD.•••••••••••••••••••••••••••••••••P- P.23 ,37, ciDaDcc 7 7 0. 0. 0. 00C) CC 00.'l..... r4 OJ rJ 04 PWRIMMMMe4 -4 ,4 4 -• ...... 04 r4 04 04 e4 N4 CW r4 rJ OW r4 04 04 P4 04 n4 P4 P4 ,11

49 un .rm04 0 0. N 33 SI ,rm04 0 0304) CD 7 7 7 0.aDa)0a) g0 a) P. p-DomoWooP- P- N. NP- P- NNIP- NP-43004243400.0.0003O• • • • • • • • • • • • • • • . • • • . • • • • •

CD 0 CD CD CD CD 0 3D cD C7 4) CD C) C) CD CD C) 3 0030000003030a 3 3 3 3 C) 3 0 3 0 C) CD 3 C) CD C) C) 3 3 3 C) 3 0 0 C) 3 3 CD• • • . • • • • • • • • • • • • • • • • • . . • • . • • • • •7 CD CD 3 CD CD C) 3 0 0 0 CD 0 0 CD -D ca a C) 3 3 0 0 C) 00 ,3 cD CD 0 C) CD

000300 0 0 0000000300030000303000003OoDON,r Dm .3 N .7.0 DON.). ,) a, 0 P4 s P2a3 3N-7 0 .r) ON-7............................•••••

,DOOO=CPM3. 000 ,30 ..... NNNN.vm,mnmrnr .,-4-4.- NNNNNNANNNNNNNNNNNNNNNNv

99

Page 110: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

7-4-m,D4204Z,C00.0n00 M0 1.10. 0010070.00 43 0.ZODCPMr...-.0.-1,00^0041.aW10.4104Nm0m0...0.0.000.-0010MMNPJ,.

MOOTTWODP-P-0041NIM MN,CluNoNuNtiNOUNOIAtriA4Nulul .....NPJOJNINNININNNNNNNO4NNNNINNNPUNNPJP.INCYMINNNNNNOINNO4NNO4N,NINP1NPUNINNINNNNNNNNtNINNI,N N

........... 00000000000000000000 .0000000 c

.

a0 0 00c000000000c00000 000

4n 4.010.400.0n 0.0 NM 0.0.- UNMNOwnm..4cpp.4 0. 0. 4.,Mm4,0...0.oft.j.42020mmuNP- Tm 0.1 4. .000' 4 M4,N00Nmull...0.•••••••••••••••••••••••••••••••••MO3M WMCP 00. 0.0. 000 0 00 000000 NNNNounIMMmCIMm me0 MMMMMM01 -r -r ....... ......

03P.04.M04a0.00.UN...0M MOCT=0 ,014. 01•400.M0W(4.M.-40CP c0- 07 OD '0'0'0 ..... P. Pn P. 4D 43 00 o .0 Ln UN 111 %I IA IA IA IA .1"• • • • • • • • • • • • • • • . • • • • • • • • • • • • • • • . •4) 43 4D .0 43 42 43 43 43 4D 43 43 43 4a 41 -0

c4 CO O 01 p4 cp .41. eu 10. P. UN n11 CD CC up •4 CP P. 04 O'0 n On CP 42 04 3. P.CIA LIN UN IA 4. 4.mmm0101Cu eu 04 04 ..... 0 CD CD CD CP CP CP Topa,•••••••••••••••••••••••••••••••••41 MN n UN ul crt UN ul A UN 4N 11 UN LIN .r1 UN al Ul UN ul IA UN A LIN 41 4N UN .r 4. 4.4.

4N 41 UN Ul ....... 01 1.0/0/111(110111104 04 04 04 P.1 04 ....... cDe4 0/ IA P. CP M ri 01 e4 In MI No' MI A P. CP 04 01 UN 01 CP WN CP c4 en UN....•••••••••••••••• n •••• n ••••••••4. 4. 40 4. 43 UN y, 4N 1N 43 43 0 .0 P. 0. A. P. P. co cf co co co ci. cr 0.• nD' CI. CD CD CDOJ r4 04 04 04 04 04 04 r4 PJ o4 04 0.1 04 OW P4 04 04 C4 04 0.4 Oj 04 04 04 04 C4 CU 0j rn

0.n P. P. P. '0,00,0,0,042 43 42 43 Po A UN UN uN UN 4.4. 41- m nI on On on on 04 04 0.1 .40 0 .0 43 410 43 43 0 0 000000.0000000 00• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

0000000 a 000 0 3 J 0 0000000000 3 0000 3 0013-3a300.003030 003a030003000300.3009•••• ••••••••••••••• ••••••••••••••O33O., 000 0Q003 O3O003O a (D CD 'a CD (D (3 D 0

3300000393 ° 1300 3 0 030000000000000MOaDa0.1.4.0 p 3 N 4.000 N 4. 43M3 04.043.90N4. 0000J-T0M0••• •••••• ••••••• •••••••••••••••• •

4. 41 uN uN JI ul 40 43 4200 P. 0- 0- 33WMDCP T C. CP C) a0D0-4eu y4 04 pu e4 c.4 04 04 04 04 e4 PI 04 04 PJ 04 04 04 0.1 o4 mi 04 ,j 04 c4 In In In M In 'I

100

Page 111: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

4,1.-.PaDŒWPCP.O.M•74000J4..(7. 0. 41mN,-4NM•ap40.4101•4^Jp•om01,4% No..emoaDisomp41.-.P..moN414%WrnmelmNNNN.-,...000 0171, 4341MN0pW414sWW 40.71 4.4. ............... MmmmmmoINN,ANNNNInitlAIOXIWIA MU, n4C41.M.17truin

000 n 00.000 0 . 0 00 .................0 0 0 0 00 00 0000 0 00 00 0 0 00000000000000

CO,nr-CPO NWP ,, WP ,"\IMOINGI.. ..0`0,4NOW(L.Pao0NMW,ONWP0,....14' 41.0N WPO .... M•rWQr- COP 0 ....,10, 4',Ors.................................

..... ..... 44.4.114..rw..r.r1M1WW

...... mmwommmcnmmtNNNNNNINN ......... 0.1poogeooe0000000

.oeeee.ro.000.oeoe

W, 40 C) 4) ,nt n- CD .73 n• r4 CO nJ C) 0 .7 74 0. v.. te,o3comNis l...-1,43,043 .43 .43w) ,414% ,A4. 4. 4. 4. Wmelmmme4N,JN•4.-$.-4• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

00000. a. C.0. 0. 0J3 mMa.0.000LAAn tn) 414.WWM1CP.-4m4- •MNS.omaNW.D.30N4".0.20N4. .0120N4..000

00 .---4. .4 4'.r 4' 444 fl fl4040 mmolmm 0 melmmmmtm 10 40 M 40 0 404004040 en 40 7440 rnm 40

N ..... 000000. 0. 0=flœwn- n-1,-0,04.),,4) .0 .C4 0 43.4:100.41.4741A0 1 35nOulto 45 ns5444••••••••••••••••••••••••••.......

0000033000300 30000003 03 00003030 Q 0O 730 303 333033000003 )330333700333................ ..... ............

00000ao..loo.poa 3030 J 3 70000 33

00000000300-1 000 00000000300 .)03J00'44 10M AJS .000 N4'003 74 S000NSZW3NSO03NS0.. ..........................e.o..

▪ .-4 ,4-4NNNAJ NIMMTMMWS .74' S LA A A , 444.0.00Mr,,,,, mMMMTMMM.NMMNMcNTTMMMMMMMMM.NM

101

Page 112: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

................ 00000000000000000

0,0„, w 4Dmais .00,10 ma, m..0. r... .0 .7 mA , CrgailtnWmcv....0Crcor-0J ,-. 0,1D 1... 4)10 m 01 . 0 .1Dr•Dulm 0,1 •40(2. 0dtc“70101,-100.9D.CIt0.:Nrveso 00000000 0000000aa.PC).0.0.Craa.Crm-awa.Mtr% 414141 LOW% tO 41.0.4) tO inv n 41 41 ..... -3 -r

00000 .400.00 0000000000 OV100..010.0.*

0000000o300o0o0po00c000c000000000

CD CD .4 .4 r.a, .7. 0 pl .4

.4 rv{PI 44

ra r,43

rw ra .4 .4sp C. CD .4 04

.4 .4 CD CDal ,r 41 4D

.7 .747

4, ce P. '0.4 41 VP an 04 .4 O.7cp CP CD .4 .4 en ,r 41 4D Is co ,1)

• • • • • • • • • • • • • • • • • • • • • • • • • • . • • • . • •1.- N.cccomsmmommPaCraCIMO.0 . 0000000000,r 4. ,r ,r ,r ,r ........ ...... ,r MN Ul 40 11 W1414141

000 00000 o.o.—pacpero.o.com=wpcommor-NI, Nr..1..r.r......... . . . . . . . . . . . . .

.o 4D 4) 0 43 '0.4 D0 al ua 41 al al VA al 44 .004w4 44 al a, ul al 0 41 UN .0 .0 .1 0.0

0.),-,0.1, L0m,...a.rs0m•10.0...re400 ,04. NOma ,r0J0D4D-1-401D.4.000000.a' CP0.0. CO CO CO Doe P.P.P.P. 4:1 .0 ,0 42 4:1 11 AO in 4100.00008001.0000.04000. Gre000.000 0 .00

4. 4. 4. 4. 4. 4. 4. MMMO/MM ,, MMMMMOIMMMO/M ,IMMOIMMM ,/

Nis, 001,00r..racc001 ,r43400n4m ,0ncr.m ,0

• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •r.-r-r-r-DpmMWCra.C1.7.0.00 -a00 ..... ,,J0inJe4mcnmmmmmmmmmmmmmm,r ,rm ,rrn ,rrnm ,r.r4.

ul Ul 0 ,r ,r Ar 4. Men an an en .4 .4 .4 ('4 -4 -4 -4 .4 .4 CD "3 CD CD CD I. D. Cr .7 .7Len LA LA• • •

.1141 41 O• • • •

LA Lt, LC* 41 n1• • • • • • •

L0 041• • •

nu, n.e% Le, 41 LA rs 14, 4141 '74'• • • • • • • • • • • • • • • •

00000000000000005500p530000000535033 50005'3 303050305305303 355353. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

3 CD CD CD a 0 0 CD 0 CD CD CD 0 35 5,300.5 3 500000500.5C. CD CD C. -1 CD CD -3

00000030 ,33050000 -700505 0 0030 000

ce. ,,, 4.4300a44. 424DONt-CIDC3 ,4444. .003 ,14.0 = 3,1• • •

102

Page 113: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

0' MN0400 .7 0W ,0 ,7 mmmn4ICIO‘r 0' ,7 C.Qmmcwaw..r.4342I7% iCtmenmMenmmmmmm ,sr-cccr 0'3 mn4m.s.ennamotroonni..0Cror-rnmni.-.00.03P-4743n4rmr4ammaanNr-N ........... 4.4-rn.rr ..... ..........

O 000* ........... olOosee100•0 .......

000000000000e000000000000000000 0

cia r- .0 44 Je 04 .4 0 co r- 0.7 P4 v4 r. 4) 4. P4 CD CP. P. 40 el r4 en 4. 04 C co sp m

0'0 mn4MW 4i L42P- 00'0 m..-, Nm.7 ws4a.GNOCrOC.-, Nm.r.r..N.0•••••••••••••••••••••••••••••••••

.... r4 .. micsiminJoANINNINNNNIMmmmmmmmmwaux,r,rt,r.m,n47,47,1AiruNtnuN4ligainuNtA4an4Nuarmlour“rtuawsw , 144 ,cl

- CD Cr r. 4D 41 4r en 04 r4 CD CP. CD 1.n 410 A r P4 CD 7. oa) 4D trt ,7 en 04 CD 0'N- P. 43 4) 4D 0 43 40 40 43 44 40 41 ci 40 WN ALIN 41 '7 '7 4r ,r 4r .4 .renco•••••••••••••••••••••••••••••••••

kcim isn 146 uci Leila al 7% I14 .Cn tn "41 le, n ^4.* al 41 .1 In in in 41 41

WJOZ.0.",M -40. f.. X1MNOM .12.7 N 100.1:14.M-4a.rn

J 4r.7 .rmmmenmn.INn4Nmsc 000000' 0

• • • • 8 0. • *e.g. • .0* • ••• 00.0••••• .0o

0.1TRITRIMIMM1•1111 Infnn menenm rnrnm momenr4 N N

/I 4D CD OD OD ...... e• 414 40 .0 .0 471 4D 44 41 111 Ul 0.0 .7.7.7.7.7.7.74r 4r 4r ni

r- Cr -4 en i P- Cr 44 m 47% r- Cr -4 M ql r- e4 RI 44 r- Cp. -4 01 UN Na' el 444 Na'• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •41M14./.2.4.41104441 40 4,43 4, 4 4-1,CDOODD W MC.C./.0.C.0

r.r.r ....... -7 4r -7 4rm.s.4.4r4r-r 4? '7 4r4r,rw,

Croamoor-r-47143.0 ,0-043uNnWs 4r 4r .M molMmmr4

4r,r4rMs-rm ,r ,rWM 4r ,r 4..74.M4r,rrn,7444.4444..444. .4.44.44 4..44.4.444

c) 3 000000000000 000.70000002.30,0-300

00 .1 03 ...)3 -7 -7,0000 ,0 3 -000-3,00000

4 4 4.444...4.. ...4444,.. 4 4 •

.DC),3000°-)0000J 0 .300 J 0000.30.3.)0op

0 00000,300 3 000 000000000000000

TZ 00 NJ43400N4.090N/4303 04.4.440304 / 4333N140

00000. 00.000000000000 00111&00000.

.7 4. 41., 10 .4:04aZ O O DCOM0' 0.0000,3

103

Page 114: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

41.0 MN•0414.01 MMADODCOPIN ,,,ACTSCraNODO1CPQN1n14DP- MCPOrANa 4.414Dr- MCraNOICrOaM4. M4. CPICOdDNI

w.

a44444J44 4. 4 dratflanlmagatarnmalaNIQICI•r .7 4 4.

.... o. 000000000 . ..ove 0000000000 •••000000000000000000000000000000000

.4404m....a4,mor-4-4cromomor..4.4r-4 ....... DaNr... 031,00a1101M ,741414.NOOM, CrOardrds14. .e4144414D4D4D1,•••••••••••••••••••••••••••••••••

a la r a 4 4 ,r .7 .1 .7 4 ..... 41 41 41 41 41 41 A 41 A A A 41 UN 41 41Q A 44 .4 41 41 41 A 41 4:1 41 41 41 41 41 W. .1 41 A A 41 A 41 A A 41 14. 41 41 75 4, 41 41

O a M 04 C) cr 0 47 41 dr RI rd a CD Cr Q r- 41 dr a eu A CD Cr n rn a a -I. aMI en tri en 0, rA AIMAI rd r4 n4 Al .......... 0 Do 0 0 o o• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

J‘ 4, 41 41 di 41 n 44 41 41 41 41 di 44 41 41 41 41 A di 41 41 dl A A 41 a Ui A 41 Cl

41 .7 rd CD 011 41 a a CD OD a a a n4 le a JO /1 en o. CD cID 43 UN ta r4 0 nin no a .7ac022,o0T-T-2-2.-1....2.-a4340aa ..nulininaintn.r.r.s. -74. 4- manma• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

r4 c4 eu ou ru n4 c4 CU eJ eJ 04 cq r4 r4 ru C4 rg Cg ,4 rd r4 Al V e4 rd rd rd rd

11 me en a on am ro ru ro r-454 ru rd ........ CD CD CD CD 0 CD CT Ce CT CO 7.MLANCe-AMUlr•CraMaNCram41 M1PAMUZINCraM 411,•ODONS •C• • • • • • • • • • • • • • • • • • • 0 • • • • • • • • • • • • •

CD o CD . '..4 ... ru r4 n4 n4 ,4marram,r 4 Jr 41 41 UN 41 41 4D a a .0▪ in un un 41 a 41 a 4% 41 un 41 UN A 41 41 ul dl 41 41 44 41 41 41 41 41 41 41 41 41 44 LA

cv,Jrua..000001.7.Tcr, Cra0aactaaaal.r,.-T ,T.2.4. 4. 1 ..2.W. SMaMnfmmolmenm en enamm olmmmn• • • ••••••••••• ••••• ••••••••••

C) 0 CD C) 0 0 0 C) CD 0 CD C) C) C) 3 CD C) C) 0 C) 0 C) CD CD CD 0 CD C3 DO 7 7 DOD7 CD C) 7 7 7 7 7 C) 7 0 CD CD C) 3D 3003 7 7 7 CD 7• • • •• • • • • • • • •• • • • • • • • • • • • • • •• • • • •..ac.0..30.„10o03o.au,Dao.a.D00000 0 0J 000

330000CD CD CD CD CD 7 CD CD 0 CD CD 30003CD CD 0 70CDCD 7 CD 0 CD CD CD 330 3

0 '4 a o3 0 rd dr a) a) 0 rd dr dO 40 3 rd Jr '0 0 C) r4 4 a co D 04 Jr 43 co cl ni Jr• • • • • • • • • • • • • • • • • • • • • • • •• • • • • • • • •

n4 n4 -I a y-4 c. rd r4 r4 r1 çn tS m en ,r 4. t '7'74 41 41 41 d% dl 4.0 a a Cr'- r-41 ..r1 41 al 41 A A 41 41 al 41 41 A A 41 A UN 41 ul 41 A 41 n 4, ot A 41 al A 4-1 41 41 41

104

Page 115: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

• 000000000000000000000 ........

77.111 - 77';JTV'"mœ""'".............. '",1?=7.= ;"!=

o O.o.oese ........ 0.oeooOfi0000oe.0*

0000000000000 0 00 0 0 000 0 00 0 00000000

03 .4 Is a Ar r. CD 04 AD OP 01 4D CP mi uN a) .4 4' Pn 0 CO V CO Pft CV. n4 W. 0p- r- 03 aL 03 IT CI. CP CD 0 C7 CD orq ej cu eu mm m .r Al AN AN W3 40 40 .0 r-• • . • • • • • • • • . • • • • • • • • • • • • • • • • • • •tr, a.Ala. ,rnir,.M 434DX)4D447.C ,043474DX)4D ,C.4D4Da4444, 43,44m4,11 41447.44w,c4 ,44A 404044444004.4400A4,4440 M

oa .4 CD CP CP 03 P- 43 04 ar Pa —0 CP a, P. 4D 44 .r n ms -4 CD CP r. 4D 4D a. .r m ra00-1CPCPCPCPaCPCPCPCPCPCP30co40m40aDmWai40r..np-M1p-r.p.pAr••••••••••••••••••••••••••••••••••ANIAN4p14. .r.r.p ........

.4 CP. aD 47 u. I ea .4 04. 03 .00 m pa .4 cp. 00 0 40 .? pa ....I CP. .2 uN .r Pa .4 C) aa r. 4D,1 Oa N r, rq Oa 0,4 ea ....... CD 3 00000CD CD CD 0 cr cr cr cp GP CP CP aD co to• •.• • • • •.• • • • • • • ••••• • • • • • • • ••• • •

NI Cara na 454 P0 Psi Para pa NJ Nina 443 444 Pa (Ni 4-4 m 4 "4.-44-4.4

MOD ..... 474,4000na4 ,44Aa.marsmmmnMera0:40044. .c4mON.:4:1WONIS4DaDOMON4.4Dm3m.r4DaDON...................................... maDmWm<P00030 ..... p,DrarqN ,J.Im

44,44n.r.a4444.414DAD474242474D474D4D.47

4 X) ANO) ,4 01 n Ar .r .r .4mmmmmr4 ea r4 ru -4 w4 -4 3 0 lk CP aP 40 aDmmmmmenmmM.ImolmM.Immnmnmmmmmmmmc ,INNINN• • • ••• ••• • • • • • • •.. • • • • • • • • • • .

00 7 <D 0 CD CD 3 30 300030030000033000330700330003303307770033003000703• • • • • • • • • • • . • • . • • • • • • • • • • • . • • • • •3000000077000.7J 0 070000303 0 7 7000.3

0 -Do 303 330030-3003 3 -33.33000000000

• • • • • • • . • • • • • • • • • • • . • • • • • • . • • • . •...n.°0340ax.CPCPPTcp30703-4-4,-4-404 ,4cavraMmmmm4.11 ,o4a.nuluNAN4-i J%Zi4743447 ,04D ,O ,M470.0474D4D ,0 ,0.O

105

Page 116: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

O 0000 00000.*90.4. 00000000000 oog000.

mo .umCP 4,414,4 0. 4NCP4AJO,,OmCr45m0NrON.r.a,4"7* -4 .4 mmlmmc,Jem.....0000C). 0.mmmpNNP-4.43.041,n)r.4mmmmmeomelmmmmmmMMNNN ,JoJesiedim,imNNNNNNNNN4.4.4-Nr.r. 00000 Nr.p.4-4-r-Nr-NNNNIN

. • . . • • . • ..................... • •c) 0 CD C) 0 CD CD 0 (D CD CD C) 0 CD C) CD a CD 0 C) CD 0 C) C3 0 0 (D 0 0 0

m4 41 r- 0 ru .4 r- a- ,4 a5 CD PJ m3 P- OP M 41 r- 4.4 1,1 .4 43 (0 Ca 04 rr P- Crr- r- r. OD CID .2 D D CP a- o' Ct 0 0 0 0 CD em .4 ry N. 04 V4 P.g m m m m• • . • • • • • • • • • • • • . • • • • • • • • • • • • • • • •

.0 43 43 -0 43 4) 4) .0 4) .0 4) P. r- Pn rn rn Tn Tn Pn r- r- 4. t- r- p. r. N"r w aC%aIflÏ um .n an uN un uN un n0 41 41 un 41 41 41 41 U1 41 4 41 nn.nn c,

.4 CD CP qD r. 4 uN mn mJ .4 CD C0 CO 42 4) um m ri .4 0 CP CP o T.. 4) 1.12 .4 m MP. P. 9.0 4 0 4 42 42 43 9 In UN 41 UN 41 41 41 41 41 41 41 .4 4 .4 4 .4 .7 .7 4' 4-

• • • • • • • • • • • • • • • • • • • • • • • • • • • g • • • • •wr S .1r 4. 4. .4 .4 .4 ,t r .4 S .4 4. r .4 4. ,r .4 .4 .4 4. 4. 4. .4 ,4

aN m 4 ,r <3 (0 rp uN M 04 ,r CP N. O 41 40 M 554 .4 a a. m t- 4 dl r m nuN- p_ r. p. p. p. p. 4 4 .0 9 4 4 9.0 4 4 In In tc.% 41 441 41 an 41 un In

• • • • • ••• ••• • • • • ••••• ••• • • • • ••• 4

'J 000CD 7 <3 CP CP CP CP cp cp 03 a, a2 az) a5 co 42 110 P. r- r- r- r-443m0m.44053,4.44a00, ,mm3r-cr-omnr-cp ,sm.ANCr•-iMw, N•••••••••••••••••••••••••••••••••

mmm .4 4. .4 ,4 UN n 0) aN 41 U1 4D 4D 4) 4D 4) ............ '2.0 43 43 43 .0 JD 43 .0 (3 da A) 4) 43 42 4.0 43 43 42 4D .0 4D 4) 43 43 .0 4) 40 A, 4)

r-r- 494 41 41 In ,r .4 .4 r mm m N. 4.) N. .4 .4 .4 0 00 0 Cr CP CPN. 0) N. 04 N. 04 ma ma ea ma MJ Oa ra 04 155 04 ma r( ro N. fg N. N. mJ N. 04 r4 ru 54 .4 .4 4••• • • • • ••• • • • • • • • • ••• • • • • •••• •

00 0 0 C3 C) 0 0 C3 0 0 3 C) C) 0 C5 C5 0 0 0 0 0000 30C) 3 0 C3 0 C)7 7 7 3 C3 C3 3 3 3 3 3 C3 0 3 3 CI -3 3 3 3 C3 'D 7 7 3 C3 D 0 3 3 3 3•••••••••••••••••••••••••••• .....0003 00300000.20a00004 0 0 0 0 0 0 0300000

c0700000030703303)0 .3 -30000070 0 0 0N4. 4).23P44. 43MONW4M n'0044.-0220 ,44. .000 04 4. 4)DPN4.43• • • • • • • • • • • • • • • • 'i 4.4.41 41 In 41 41 .0430 09P. N. N. r-mxamma5 c• cr 0. T 0 3 0434343434309009 ,000004143-0434343940r-nm. ,

106

Page 117: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

CONN ......... (11MODMWWOrs.M05541MNGPMWP.P.554LI N C7.4)410 P. J ... 04.1 NACP ,41,PP. O. NO1P.OMPMMOD.4.4.01010101 NNNP4 ,14400°. 531, h4OW415.4flNI1j440 0 0PWP.P511,04,14111NNNNININNINCO ..... ....... 00.5000

..... Ps P. Ps P.P.. P. P. P. P.. P.. N P. P.. P. P. 11.. ssh rs.

000 0 0 0c0oc00o0o0oco0oc00000000000

CD P4 OM 21 Pn JD C) p- 40 r- Cr C) N. -s- JD N.0,75. 44 m .p r. ar CD N n, .r.r A A A A 41 41 43 -0 40 45 .0 45 45 40 r- r- pn r- 7. Ps P. CG GO CO CO

• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

.1 5,51.415, 5'111A1111111415‘5151415'4I4l^515410A5.51514104- 5145A

CI500. 57521..4,51SMCOP.1 44 00.0.4. MN 4....00. 51.7 .7 101TRIRIc4) MMRrM4IMOINNINNINN 7.5 NINNINNN .....

4151/..PWS/MTS.4.414141MMCOM 41 41MMMOIMJINNN....ow ee.e.o.....•• n ••••••••• •••••• e

r4 44 4 4

P. P. P. JD JD 4D JD .0 .0 .0 .2 JD JD J1 41 44 41 51 41 51 1.15 41 41 ,r ,T .r -I`M n rn r4 RI UN CP ol 54 r- e4 OM P. CP s4 41 41 P. CP s4 m 1.11 CP s4

•••••••••••••••••••••••••••••••••CP CD 0 CD CD .4 -4.4 44 C4 (11 04 ro n, m m m 5. -7 op UN 51 41 51 U1 0 .0

p. p. p. p. P. P. P. P. Ps P. r- P. P. P. P. Ps P. P. rs. P. r. r- P. P. p. N. r. r- P. Ps r-

.0 CO CO CD CO P. P. P. P. JD JD 40 .0 1 UN UN .P1 al 41 .1. 41 41 41 41 01 m 04

.4 .4 .4 .4 ..... 4 '44 4 . •••••••••••••••••••••••••••••••••

0000000.D000000.0.0000000 0 D0 0 0 00 ° 004)

,7o0a -a0073 ,7,0Q0 -00777 -D'7 0a 000,10 ca.......... .....................

0 0 C) .1 0 C) C) C) C) 0 0 0 0 -5 0 CD 0 ,4 C) a ,4 0 0 C) 30.0

000O3 -D000°,000'a003000-00300900000,40 ,11am3r4.1. ..030C)N.rom0 ,o.r.0.00 ,40 ,00l.................................

.......

107

Page 118: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

• • • • • • • • • • • • •a) cr e) .4. JO au er en r- cr P. P.P- C2 rn C3 en JD o On o en ,-•uN uN .7 en on V 4. n••1 CD CP CO 47 in0 CD CD C: CD 0 CD 0 CD CP Cf 00

• • • • • • • . • • • • •CD CD CD (D CD CD C/ CD CD CD 0 CD CD

• JD P. CO CP CD 04 M UN 47 Is".ap CO op CG cco CP CP CP CP CP CP CP CP• • • • • • • • • • • •

Pn P. r. r- ....... p. 1,uNuN aN,r4 un in '00 ,0%

4 .. 041 -4 -4 C7 CP CP CO P. JD... CD C7 a CD CD

• • • • • • • • • • • • •• Or .7 .7.7 .7 .7

P. P. JD 40 UN UN en en n4 04 e4V po y4 04 no 04 P4 04 po 04 n, ni eq• • . • . • • • . •

_r _r Inlelo Otelolemm mm1 41 0. 4-11 r•-•

• • . • • • • • • • • • •z 4) JD 1.n P. P. r- op ma) co CD,n p. o. r. o. r. rn r- r. r. r-

VVVV NI ..... -4

• • • • • • • • • 0 • • •

7 CD CD CD D 3 0 CD CD CD 0 D CDO 7 7 3 CD CD D <3 <3 CD D CI 7• • • . • • • . • • • •a CD CD CD CD 0 3 CD CD CD CD CD CD

CD CD 7 CD 0 <3 D CD CD 0 3 CD C7.74 ro .4) a C3 V 4' 43 CD

• • • • • • • • • • • • •

pn Se .0 .000 Cr Cp CPp. r- p- p. p. p. p. p. p. P. p. P. r-

108

Page 119: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

APPENDIX B

INFLUX VARIABLE DESCRIPTION

Program INFLUX computes event-based rainfall, infiltration,

runoff, sediment yield, and the inflow hydrograph required for utiliza-

tion of the DEPOSITS sedimentation routine. Included within are com-

ponent models which are described in Chapter 3. The program requires

the following inputs: event return period, precipitation volume and

duration, and infiltration rates and corresponding soil moisture storage

values. The variables appearing in INFLUX are defined below. Where

appropriate, values for the hypothetical watershed and units of the

variables are noted. Arrays are indicated by an asterisk.

Value andVariable Definition Units

in./hr.

inches

hours

50.0 acres

hours

hours

*ACTFIL

Actual value for infiltrationrate for a time increment

AVAIL

Initial soil moisture storagecapacity.

*CUMIN

Cumulative time for inflowhydro graph

DAREA Draingage area for hypotheticalwatershed.

*DURA Duration of precipitation event

DUREX Duration of rainfall excess

FACTC USLE crop management factor

FACTK USLE soil erodibility factor

109

Page 120: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

Value andVariable Definition Units

FACTLS USLE length-slope factor

FACTP USLE erosion control practicefactor

*FILL

Volumetric infiltration for a inchestime increment

*ORD Discharge corresponding to .05 hr.increments

*PCTDC Cumulative density function fortotal rainfall duration

*PCTDP Probability mass function forrainfall duration

PCTR Proportion of total rainfall fortime increment

PEAK

Peak flow rate for the event cfs

110

PINTMX

*PRECIP

*RAINCD

*RAINEX

*RAINPD

*RAINV

*RETURN

RUN VOL

Intensity index representingthe highest storm rainfallintensity

Total rainfall volume for the inchesevent

Cumulative density function for

hoursduration of rainfall

Volume of rainfall excess for an inchesincrement

Probability mass function for

hoursduration of incremental rainfall

Volume of rainfall occurring over

inchesa time increment

Return period for a precipitation yearsevent

Runoff volume for a precipitation ac.-ft.event

Page 121: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

111

Variable Definition Value and

Units

SEDYD Sediment yield for a precipitation tonsevent

SSINF

Steady-state infiltration rate in./hr.

*STOFIL

Infiltration rate corresponding

in. /hr.to the prevailing soil moisturedeficit for an increment

*STORE

TBASE

TERP

TFILL

TLAG

Available moisture storage capacity inchesfor an increment

Time base of the inflow hourshydrograph

An interpolation factorfor infiltration ratecomputation

Accumulated soil moisture inchesstorage volume

Watershed lag time hours

TPEAK Time to reach peak flow hours

rate for a precipitationevent

TRAINX Volume of rainfall excess inches

for a precipitation event

Page 122: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

APPENDIX C

PROGRAM LISTING OF INFLUX

112

Page 123: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

PROGRAM INFLUX (INPUT,OUTPUT,TAPE 5.INPUT,TAPE 6 -OUTPUT)C THIS COMPONENT ANALYZES PRECIPITATION EVENTS OF CERTAIN FREQUENCY ANDC DURATION AND DETERMINES CHARACTERISTICS OF MATER AND SEDIMENT INFLUXC TO THE SEDIMENTATION POND

DIMENSION PCTR(22),PCTDP(22),PCTDC(22),ACTFIL(22),FILL(22)DIMENSION RAINP0(22),RAINCD(22),RAINV(22),RAINEX(22),STORE(8)DIMENSION STOFIL(8),RETURN(38),DURA(38),PRECIP(38)DIMENSION CUMIN(100),ORD(100)READ(5,1)(PCTR(I),PCTDP(I),PCTDC(I),I.1,22)

1 FORMAT(3F10.3)READ(5,2) STOFIL

2 FORMAT(8F5.2)READ(5,3) STORE

3 FORMAT(8F5.2)DO 150 M..1,38READ(5,4) RETURN(M),DURA(M),PRECIP(M)

4 FORMAT(I3,2F7.2)WRITE(0,9) RETURN(M),OURA(M)

9 FORMAT("1",20X,"**** HYDROLOGIC SUMMARY:",I4," YR.",F6.2," HR. EVE).NT ****")DUREX.0.0JRAINX.0.0SSINF..22

C SSINF REPRESENTS STEADY STATE INFILTRATION RATE IN INCHES/HR.TFILL.0.0

C AVAIL REPRESENTS INITIAL MOISTURE STORAGE CAPACITY IN INCHESAVAIL-1.8

C STORE REPRESENTS AVAILABLE MOISTURE STORAGE CAPACITY IN INCHESFILL IS THE VOLUMETRIC INFILTRATION IN INCHES FOR THE TIME INCREMENT

C APPORTION RAINFALL VOLUME AND DURATION ACCORDING TO SCS TYPE IIC STORM DISTRIBUTION

WRITE(6,10)lù FORMAT(////, 5X," RAINPD(N)",10X," RAINCD(4)",10X," RAINV(N)",10X,"I INFIL. RATE"plUX," INFIL. VOLUME")WRITE(6,11)

11 FORMAT(6X,9("-"),11X19("-."),11X,8("--"),11X , 11(" - ") , 11X ,13 ("- "))DO 80 N.1,22RAINPD(N)=PCTDP(4)*DURA(M)RAINCD(N)=PCTDC(N)*DURA(M)RAINV(N).PRECIP(M)*PCTR(N)IF(N •NE. 15) GO TO 13

MAX. PRE:IF • INTENSITY, PINTMX, OCCURS AT 15TH DURATION INCREMENTPINTMX.RAINV(N)/RAINPO(N)

13 WRITE(6,15) RAINPD(N),RAINCD(N),RAINV(N)15 FORMAT(8X,F5.2,14X,F5.2,16X , F5.2)

C FIND 'NFU-. RATE CORRESPONDING TO ACTUAL AVAILABLE MOISTURE STORAGEIF((AVAIL-TFILL) .GT. 1.25) GO TO 18

C ACTFIL IS THE ACTUAL VALUE FOR INFILTRATION RATE OVER THE INTERVAL IN

C INCHES/HR.AZTFIL(N).SSINFFILL(N).SSINF*RAINPD(N)GO TO 30

18 DO 2 0 K=1,8IF((AVAIL-TFiLL) •GE. STORE(K)) GO TO 23

20 CONTINUE23 IF(K .EQ. I) GO TO 25

TERP.((AVAIL-.TFILL)-STORE(K))/(5TOREST 0 RE ( K ))ACTFIL(N)=STJFIL(K)+TERP*(STOFIL(K - 1) -STUFIL ( K ))

GO TO 2625 ACTFIL(N)=STOFIL(K)26 FILL(N).ACTFIL(N)*RAINPD(N)

IF(CAVAIL-•(TFILL+FILL(N))) .LE. 0.00) GO TO 27TFILL=TFILL+FILL(N)

113

Page 124: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

GO TO 4527 FILL(N)wAVAILTFILL30 TFILLwTFILL+FILL(N)

AVAIL.TFILL45 WRITE(6,50) ACTFIL(N),FILL(N)50 FORMAT("+",63X,F9.2,15XpF8.2)

C CALCULATE VOLUME OF RAINFALL EXCESS FOR INCREMENTRAINEX(N)=RAINV(N)FILL(N)IF(RAINEX(N) .GT. 0.0) GO TO 70GO TO 80

70 DUREX.DUREX+RAINPD(N)TRAINX.TRAINX+RAINEX(N)

80 CONTINUEWRITE(6,82) PRECIP(M),PINTMX

82 FORMAT("0"," PRECIP. VOLUME • " 1 F6.2," INCHES",/," MAXIMUM PRECIP.1INTENSITY •",F6.2," IN./HR.")WRITE(8,85) DUREX,TRAINX

85 FORMAT("0"," DURATION OF RAINFALL EXCESS w",F5.2," HRS.",/p" TOTAL2 RAINFALL EXCESS P",F5.2," INCHES")

C CALCULATE PEAK FLOW RATE, TIME TO PEAK, AND TIME OF HYDROGRAPH BASEDAREA-50.0

C TLAG REPRESENTS THE LAG TIME IN HRS. FROM CENTROID OF RAINFALL EXCESSC DURATION TO HYDROGRAPH PEAK

TLAGw.17C TPEAK REPRESENTS TIME IN HRS. FROM THE INITIATION OF RUNOFF TO THEC HYDROGRAPH PEAKC TBASE REPRESENTS TIME IN HRS. CORRESPONDING TO THE HYDROGRAPH BASE

TPEAKm(DUREX/2+TLAG)PEAKw(484.0*DAREA/64000*TRAINX)/TPEAKTBASE.2.67*TPEAK

C CALCULATE EVENT BASED SEDIMENT YIELD VIA REGIONALIZED WILLIAMS EQN.C RUNVOL REPRESENTS RUNOFF VOLUME IN AC. FT.

RUNVOLw(TRAINX*DAREA)/12.0FACTLS.1.26FACTCw.09FACTKw.35FACTPw.35

C SEDYD 15 THE EVENT SEDIMENT YIELD IN TONS5EDYUw95.0*(RUNVOL*PEAK)*w.564.FACTC1FACTK*FACTLS*FACTPWRITE(6,90) PEAK

90 FORMAT("0"," PEAK FLOW FOR EVENT • "..F7.2," CFS")WRITE(6p93) TPEAK

93 FORMAT(" "p" TIME TO PEAK DISCHARGE w",F5.2," HRS.")WRITE(6,95) RUNVOL

95 FORMAT(" "." VOLUME OF INFLOW TO POND • ",F5.2," AC. FT.")ORITE(6,96) TBASE

96 FORMAT(" "," TIME BASE OF INFLOW HYDROGRAPH • ",F5.2p" HRS.")WRITE(0/98) SEDYD

;a FoRmArin " 1 " SEDIMENT YIELD FROM WATERSHED FOR EVENT • ",F9.1," TON3S")

C CUMIN REPRESENTS THE CUMULATIVE TIME FOR THE INFLOW HYDROGRAPHC CALCULATE ORDINATES OF INFLOW HYDROGRAPHC WITH 4 TIME INCREMENT OF .05 HRS.

WRITE(6,110)£10 FORMAT("1"p2OX," INFLOW HYDROGRAPH COORDIUTES")

WRITE(6,111)111 FORMAT(21Xp29("".."))

WRITE (6,112)112 FORMAT("0",22X," TIME(HRS.)",1X," DISCHARSE(CFS)")

WRITE(8,113)113 FORMAT(24X,4("-"),8X,9("-"))

INPERwINT(TBASE/.05+1.0)DO 140 .1.1,INPER

114

Page 125: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

CUMIN(J).J/20.0IF(CUNIN(J) • GT. TBASE) GO TO 150IF(CUMIN(J) .GT. TPEAK) GO TO 115

C Q0(J) REPRESENTS THE DISCHARGE CORRESPONDING TO J INCREMENTS OF .05HRORD(J)*PEAK/TPEAK*CUMIN(J)GO TO 120

115 ORD(J).PEAK....(PEAK/(1.67*TPEAK)*(CUMIN(J)-TPEAK))120 WRITE(15 , 125)CUMIN(J).ORD(J)125 F3RMAT("0",22X,F5.2.11X,F5.1)140 CONTINUE150 CONTINUE

STOPEND

115

Page 126: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

APPENDIX D

PROGRAM AND INPUT LISTING OF DEPOSITS

Input Listing of Deposits

The following is a listing and description of the inputs required

by the DEPOSITS sedimentation model (Ward et al., 1977b). Typical vari-

able values are presented, and their implications for model functioning

appear parenthetically.

Variable Definition Value

NSTORM Number of inflow events required 1.0

CONSED Control variable determining the 1.0calculation of the inflow sedimentconcentrations. (concentrationsare approximated by the model)

DEPOST Control on use of the deposition 1.0option. (no deposition option)

MASS Total mass of incoming sediment 16.1in tons

FLOW Indicator of desired outflow con- 1.0ditions. (uniform withdrawal)

TRP Control variable providing for 1.0testing of several outlet structures(model straightforwardly determinestrap efficiency)

FILTER

DENSTY

SG

Enables alteration of initial stage- 1.0discharge curve due to deposition.(no disposition effect considered)

Density of deposited sediment .89

Specific gravity of sediment par- 2.65ticles in g/cm3

116

Page 127: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

117

VariableDefinitionValue

VISCOS Viscosity of the flow in cm2/sec. .0152

DELTAT Time increment for the inflow hydro- .20graph and inflow sediment-graph inhours.

DELPLG

Time increment for outflow plug rou- .20ting in hours

SET

Dictates the relationship between the

2.0sediment load and the inflow rate.(Sets inflow sediment concentrationproportional to the inflow rate)

SHORT

Basin short-circuiting option

1.0(plug flow through basin isassumed)

FIX

FLOWAV

FRACTN

SLAG

DEAD

Correction factor for simulationof turbulence or flocculation(turbulence and flocculation areneglected)

Determines how particle size dis-tribution varies with flaw rate.(one representative distributionis maintained throughout)

Control variable defining the sedi-ment distribution during inflow tothe basin. (sediment load is com-pletely mixed with storm inflow)

Simulates a lag between peak inflowrate and peak inflow sediment con-centration. (effect is neglected)

Volume of stored flow and/or sedi-ment bypassed during routing, inac.-ft. (volume set equal to 60%of calculated 3-yr. sediment storage)

1.0

0.0

0.0

0.0

.20

MP Number of outflow distribution points 18

Number of inflow hydrograph values >2(inflow hydrograph constructed from2 values)

Page 128: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

Variable

NS

MS

LS

PERCNT

SIZE

Definition

Number of stage-area and stage dis-charge values

Number of particle size fractions

Number of outflow hydrograph points(set equal to maximum allowable)

Number of values required to fillthe permanent pool volume (pool isfilled internally by the program)

Percent finer values inputed forparticle size distribution

Particle sizes corresponding toinputed percent finer values inmillimeters

118

Value

18

5

400

0

0.0, 28.2,63.3, 85,8,100.0

0.0, .002,.063, .125,1.00

STGI

AREA

DISCHB

INFLOW

Stage values at the riser usedfor rating curves

Area values corresponding to in-puted stage values

Discharge values corresponding toinputed stage values

Inflow hydrograph values for thestorm

Page 129: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

PROGRAM SEDIMT (INPUT , OUTPUT,TAPE5.INPUT,TAPE6.OUTPUT)

* **** ********* ***** ************** **************** ***** ***** **** ***** ******THE DEPOSITS PROGRAM APRIL 1979

THE DEPOSITS PROGRAM WAS DEVELOPED AT THE AGRICULTURAL ENGINEERINGDEPARTMENT AND THE INSTITUTE FOR MINING AND MINERALS RESEARCH AT THEUNIVERSITY OF KENTUCKY, LEXINGTON KENTUCKY. THE UNIVERSITY OF KENTUCKYASSUMES NO RESPONSIBILITY FOR ANY RESULTS OBTAINED WITH THE MODEL.THE DEPOSITS COMPUTER PROGRAM IS A SIMULATION MODEL FOR ESTIMATING THEPERFORMANCE OF A SEDIMENT DETENTION BASIN. THE MCDEL WILL DETERMINE THEBASIN TRAP EFFICIENCY, CHANGE IN BASIN GEOMETRY DUE TO SEDIMENT DEPOSITSAND THE EFFLUENT SEDIMENT CONCENTRATIONS FOR A GIVEN STORM EVENT.************************************************** ***** ********* ******* ***

****** **** ** * ******* *** ***** ****** ******* ******* ********** ******* ***** ****GLOSSARY OF TERMS

ACINFL • ACCUMULATED INFLOW VOLUME (ACRE-FEET)ACOUT • ACCUMULATED DISCHARGE FROM THE RESERVOIR. (ACRE-FEET)ACT CONTROL FLAG USED TO TERMINATE SIMULATION. (ACRE-FEET)AR AREA CONTROL VALUE USED IN DEVELOPING NEW AREA CURVE.ARA • AREA CONTROL VALUE USED IN DEVELOPING NEW AREA CURVE.ARC • AREA CONTROL VALUE USED IN DEVELOPING NEW AREA CURVE.ARE = AREA CONTROL VALUE USED IN DEVELOPING NEW AREA CURVE.AREAA • SURFACE AREA OF SURFACE PLUG LAYER. (ACRES)AREA • BASIN SURFACE AREA AT EACH STAGE POINT. (ACRES)AREAC SURFACE AREA OF THIRD PLUG LAYER. (ACRES)AREAB • SURFACE AREA OF SECOND PLUG LAYER. (ACRES)AREAD • SURFACE AREA OF BED PLUG LAYER. (ACRES)AREAS • DESIGN BASIN SURFACE AREA AT EACH STAGE POINT. (ACRES)AROLD • SURFACE AREA AT EACH STAGE POINT PRIOR TO DEPOSITION. (ACRES)AVDEP • AVERAGE DEPTH OF FLOW AT EACH INFLOW TIME STEP. (FEET)AVDPTH • AVERAGE DEPTH AT EACH STAGE POINT. (FEET)AVETME • DETENTION TIME OF FLOW CONTAINING SEDIMENT. (HRS)AVSTG • AVERAGE DEPTH AT EACH INFLOW TIME. (FEET)AVTME • SUM OF THE PRODUTS OF THE PLUG VOLUMES TIMES THE PLUG DETENTION

TIMES. (ACRE-HRS)SPOOL • VOLUME OF INFLOW USED TO FILL THE PERMANENT POOL. (ACRE-FEET)CAP = VOLUME OF SEDIMENT DEPOSITED BELOW EACH STAGE VALUE.(ACRE-FEET)CAPAC • DESIGN CAPACITY OF THE BASIN AT EACH STAGE VALUE. (ACRE-FEET)CAPACA • BASIN CAPACITY AT EACH INFLOW TIME. (ACRE-FEET)CAPCH • VOLUME CONTROL VALUE USED TO DEVELOP NEW STAGE CAPACITY CURVE.

119

Page 130: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

120

C CAPCO • DESIGN CAPACITY OF THE BASIN AT EACH STAGE VALUE. (ACRE—FEET)C CAPC • VOLUME CONTROL VALUE USED TO DEVELOP NEW STAGE CAPACITYC CURVE, (ACRE—FEET)C CAPMAX • MAXIMUM VOLUME OF RESERVOIR BASIN. (ACRE—FEET)C CAPNW • RESERVOIR VOLUME AT EACH STAGE POINT FOLLOWING DEPOSITION.C (ACRE—FEET)C CAPOOL • VOLUME OF THE PERMANENT POOL. (ACRE—FEET)C CAPO • CALCULATED CAPACITY VALUES BEFORE SMOOTHING. (ACRE—FEET)C CAPREM . TOTAL VOLUME OF SEDIMENT DEPOSITED BY STORM. (ACRE—FEET)C CAPSAV • VOLUME OF THE BASIN AT THE PEAK STAGE. (ACRE—FEET)C CENTME • DETENTION TIME FROM HYDROGRAPH CENTERS. (HRS)C CHECK • AREA CONTROL VALUE USED TO DEVELOP NEW AREA CURVE. (ACRES)C CONCED • INFLOW SEDIMENT CONCENTRATIONS. (MG/L)C CONSED • CONTROL VARIABLE DETERMINING THE INPUT OF INFLUENT SEDIMENTC CONCENTRATIONS. (MG/L)C COR • CONTROL VALUE USED TO SMOOTH NEW STAGE—CAPACITY CURVE.C DEAD • DEAD STORAGE VOLUME. (ACRE—FEET)C DELPLG • TIME INCREMENT OF EACH OUTFLOW PLUG. (HRS)C DELTAT a INFLOW HYDROGRAPH TIME INCREMENTS. (HOURS)C DENSTY • DENSITY OF THE SEDIMENT DEPOSITS.C DEP • VOLUME OF SEDIMENT DEPOSITED DURING EACH DISCHARGEC INCREMENT. (ACRE—FEET)C DEPCAP . CHANGE IN STORAGE CAPACITY DUE TO SEDIMENT DEPOSITS.C (ACRE—FEET)C DEPI • CHANGE IN STORAGE CAPACITY FOLLOWING SMOOTHINGC COMPUTATIONS. (ACRE—FEET)C DEPOC . SMOOTHING CONTROL VARIABLE.C DEPO • CONTROL VARIABLE USED IN SMOOTHING OF NEW BASIN GEOMETRYC CURVES.C DEPOST . CONTROL VARIABLE DETERMINING USE OF THE DEPOSITION OPTION.C DEPTH • AVERAGE DEPTH DURING DETENTION OF EACH PLUG. (FEET)C DEPTH1 • DEPTH OF THE SECOND PLUG LAYER. (FEET)C DEPTH2 • DEPTH OF THE THIRD PLUG LAYER. (FEET)C DEPTH3 • DEPTH Of THE BOTTOM PLUG LAYER. (FEET)C DETAVE = AVERAGE DETENTION TIME OF DISCHARGED FLOW. (HRS)C DETTME • DETENTION TIME OF EACH PLUG. (HOURS)C DIAMTR • PARTICLE SIZE WITH A FALL VELOCITY OF VELOC. (MM)C DIFF • CONTROL VALUE USED TO SUBDIVIDE THE TOP LAYER.C DISCH • DISCHARGE RATE AT EACH STAGE VALUE. (CFS)C DISCHA . DESIGN DISCHARGE RATE AT EACH STAGE VALUE. (CES)C DISCHB • DISCHARGE VALUE FOR EACH STAGE POINT. (CES)C DPTH • STAGE DEPTH FOR EACH DISCHARGE AND AREA VALUE. (FEET)C DPTH • DEPTH VALUES ON THE OUTFLOW DISTRIBUTION CURVE. (FEE)C EFLNT • EFFLUENT CONCENTRATION FOR EACH OUTFLOW INCREMENT. (MG/L)C ERROR • ERROR IN DETERMINING THE NEW BASIN CAPACITY. (ACRE—FEET)C FALL • REQUIRED DEPTH OF SETTLING. (FEET)C FILTER • CONTROL VARIABLE DETERMINING THE USE OF AN OUTLET FILTER.C FIX • CORRECTION FACTOR TO MODIFY THE FALL DEPTH.C FLOWIN • FLOW RATE AT WHICH THE SEDIMENT DISTRIBUTION WAS DETERMINED.C (CFS)C FLOW • CONTROL VARIABLE DETERMINING THE INPUT OF AN OUTFLOW DEPTHC DISTRIBUTION.C FRACTN • IF FRACTN GREATER THAN 0.1 FLOW OCCURS AS A DENSITY CURRENT.C INFLOW . INFLOW RATE AT EACH INFLOW TIME. (CES)C LS • NO OF INFLOW VALUES USED TO FILL THE PERMANENT POOL.C LLS • NO OF INFLOW VALUES USED TO FILL THE PERMANENT POOL. (LLS=LS)C MASS • SEDIMENT INFLOW LOAD. (TONS)C M • NUMBER OF INFLOW VALUES.C MS • NUMBER OF OUTFLOW ROUTING VALUES.C MS . NUMBER OF OUTFLOW ROUTING VALUES.C N • NUMBER OF STAGE VALUES.C NFLNT • THE INFLUENT CONCENTRATIONS AT EACH INFLOW TIME PCINT. (MG/L)C NS • NUMBER OF PARTICLE SIZE DISTRIBUTION VALUES.

Page 131: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

121

C NSTORM • CONTROL VARIABLE DETERMINING THE NUMBER OF STORM EVENTS.C OUTFL1 • OUTFLOW DISTRIBUTION FOR THE TOP PLUG LAYER. ( )C OUTFL2 • OUTFLOW DISTRIBUTION FOR THE SECOND PLUG LAYER. ( )C OUTFL3 • OUTFLOW DISTRIBUTION FOR THE THIRD PLUG LAYER. ( )C OUTFL4 . OUTFLOW DISTRIBUTION FOR THE BOTTOM PLUG LAYER. ( )C OUTPCT • FINER OF SEDIMENT IN THE POND EFFLUENT.C PCT • PERCENT OF SEDIMENT REMAINING IN SUSPENSION IN EACH LAYER.C PCTOUT • SEDIMENT LOAD FOR EACH PARTICLE SIZE IN THE EFFLUENT.C PEAKIN • PEAK INFLOW RATE. (CES)C PEAK • PEAK DISCHARGE RATE. (CFS)C PERCNT . PERCENT OF PARTICLES CAPABLE OF FALLING THE RESPECTIVEC INDICATED DEPTH DURING THE PLUG DETENTION TIME.C PFLNT • PEAK INFLOW SEDIMENT CONCENTRATION. (MG/L)C PLGCEN • THE TIME OF INFLOW OF EACH PLUG OF OUTFLOW. (HOURS)C PLGTME • THE TIME OF OUTFLOW FOR EACH PLUG. (HOURS)C PLGVOL • THE VOLUME OF EACH PLIG. (ACRE-FEET)C PSTAGO • PEAK STAGE VALUE. (FEET)C SED • PERCENT OF THE TOTAL SEDIMENT INFLOW CONTAINED IN EACH PLUG.C SEDAVE . AVERAGE EFFLUENT SEDIMENT CONCENTRATION OF FLOW CONTAININGC SEDIMENT. (MG/L)C SEDAV2 • AVERAGE EFFLUENT SEDIMENT CONCENTRATION OF ALL FLOW. (MG/L)C SEDEND • ACCUMULATIVE TOTAL PERCENT OF THE INITIAL SEDIMENT DISCHARGED.C SEDMNT • PROPORTION OF SEDIMENT ASSOCIATED WITH EACH INFLOW INCREMENT.C SEDOUT . FRACTION OF SEDIMENT CONTAINED IN EACH PLUG. ( 1C SEDPLG . PERCENT OF SEDIMENT DISCHARGED IN EACH PLUG.C SEDTOT • ACCUMULATED VOLUME OF SEDIMENT INFLOW.C SEDT • ACCUMULATED VOLUME OF SEDIMENT DISCHARGED.C SET . EXPONENT OF FLOW SEDIMENT LOAD RELATIONSHIP.C SFLNT . INFLOW SEDIMENT CONCENTRATIONS. (MG/L)C SG • SPECIFIC GRAVITY OF THE SEDIMENT PARTICLES.C SHORT . SHORT-CIRCUITING CONTROL VARIABLE.C SIZES = FINER OF INFLOW DETERMINED AT PEAK INFLOW RATE. ( )C SIZEST • INFLOW PARTICLE SIZES DETERMINED AT PEAK INFLOW RATE. (MM)C SIZE • PARTICLE SIZE ASSOCIATED WITH EACH PERCENT FINER. (MM)C SIZOUT . FINER OF EFFLUENT SEDIMENT PARTICLES.C SLAG • LAG INCREMENTS OF TIME OF FLOW PEAK BEHIND SEDIMENT PEAK.C SMOOTH • AREA SMOOTHING FUNCTION.C SMOTH2 • AREA SMOOTHING FUNCTION.C STAGE • DEPTH OF FLOW FROM THE LOWEST BED ELEVATION. (FEET)C STAGEA • STAGE AT EACH ROUTING TIME. (FEET)C STAGO • STAGE AT EACH OUTFLOW TIME. (FEET)C STAG • STAGE VALUES PRIOR TO EACH STORM EVENT. (FEET)C STAREA • AREA UNDER THE AVERAGE DEPTH-TIME CURVE. (ACRE-FEET)C STARTV • VOLUME OF INFLOW AT THE START OF THE ROUTING CYCLE. (ACRE-FEET)C STGAR . AVERAGE STAGE AFTER EACH INCREMENT OF INFLOW. (FEET)C STGIN • STAGE DURING INFLOW OF EACH PLUG. (FEET)C STGOUT • STAGE DURING THE PLUG OUTFLOW. (FEET)C STG1 • DESIGN STAGE VALUES. (FEET)C 51 G2 • STAGE CONTROL VARIABLE USED IN SMOOTHING CALCULATION.C STG3 = STAGE CONTROL VARIABLE USED IN SMOOTHING CALCULATION.C STORM • STORM INFLOW VOLUME. (ACRE-FEET)C SIP . ACCUMULATED VOLUME OF OUTFLOW. (ACRE-FEET)C STPV . ACCUMULATED INFLOW AT TIME Ti. (ACRE-FEET)C STRMOT . STORM VOLUME DISCHARGED. (ACRE-FEET)C STRMTM • DETENTION TIME INCLUDING STORED FLOW. (HRS)C SUMTME • VARIABLE USED IN EVALUATING THE STORM DETENTION TIME.C SUMVOL • VOLUME OF FLOW DISCHARGED. (ACRE-FEET)c sum:. . DEPO**2.0*(AREA(J)-AREA(J-1))+SUM2C SUM2 • DEP0*(AREA(J)-AREA(J-1))+SUM2C T • TIME TAKEN TO FILL PERMANENT POOL. (F R S)c TMEIN . TIME DURING INFLOW. (HOURS)c TOTAL • VOLUME OF DISCHARGE USED IN COMPUTATION OF CENTME.C TOTVOL . VOLUME OF DISCHARGE USED IN COMPUTATICN OF CENTME.

Page 132: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

122

C TRAP . TRAP EFFICIENCY. ( )C TRP • INPUT VALUE OF REQUIRED TRAP EFFICIENCY. ( )C Ti • TIME SINCE THE START OF INFLOW. (HRS)C VAR • VARIABLE USED IN SMOOTHING CALCULATION.C VELOC • FALL VELOCITY. (FEET/HOUR)C VISCOS • VISCOSITY OF THE FLOW. (CM.SO./SEC)C VOL • VOLUME OF EACH PLUG LAYER. (ACRE—FEET)C VOLACT • INCREMENTAL CHANGE IN BASIN VOLUME. (ACRE—FEET)C VOLA • VOLUME OF EACH PLUG. (ACRE—FEET)C VOLS • VOLUME OF FLOW BELOW THE SURFACE PLUG LAYER. (ACRE—FEET)C VOLE • VOLUME OF FLOW BELOW THE SECOND PLUG LAYER. (ACRE—FEET)C VOLD • VOLUME OF FLOW BELOW THE THIRD PLUG LAYER. (ACRE—FEET)C VOLC • VOLUME OF EACH LAYER ALLOWING SETTLING INTO THE NEXT LAYER.C VOLIN • VOLUME OF INFLOW ACCOUNTED FOR AFTER EACH PLUG DISCHARGES.C VOLOUT • FRACTION OF INFLOW SEDIMENT LOAD ROUTED AT THE END OF EACHC TIME POINT.C VOLSED • CALCULATED CAPACITY AT EACH STAGE VALUE STAGO. (ACRE —FEET)C VOLTME • AVERAGE TIME DURING THE INFLOW OF EACH INCREMENT OF FLOW.C VOLTOT • TOTAL VOLUME OF INFLOW (ACRE—FEET)C VOLUME • VOLUME OF INFLOW DURING EACH INFLOW TIME INCREMENT. (ACRE—FEET)C X1 • ROUTING VOLUME USED TO SOLVE CONTINUITY EQUATION. (ACRE—FEET)C X2 • ROUTING VOLUME USED TO SOLVE CONTINUITY EQUATION. (ACRE —FEET)C

COMMON/HOLD/INFLOW,MXPLLS/DELTATDIMENSION FLOWIN(400),SIIEST(10,400),SI2OUT(10,400),PCTOUT(10p400)DIMENSION DEPTH1(400) , DEPTH2(400),DEPTH3(400)PDEPTH(400)DIMENSION PERCNT( 4 00) , X1(400),X2(400),SEDPLG(400),SFLNT(400)DIMENSION PLGVOL( 4 00),AROLD(400)/CAPNW(400),DIFF1400),STP(400)DIMENSION ACINFL(400),VOLUME(400)pSTARTV(400),STPV(400),STAGE(400)DIMENSION STAGEA( 400),CAPACA(400),T1(400),DISCHA(400),CAPAC(400)DIMENSION DISCH(400) , NFLNT(400),EFLNT(400),CAPC0(400),CONCED(40C)DIMENSION AREAA(400),AREAB(400),AREAC(400),AREAD(400),INFLOW(400)DIMENSION VOL(4,400),SED(4,400),VELOC(4,400),FALL(4,400)DIMENSION VOLC(3,400),DEP(4,400),PCT(4,400),PERCT(5,400)DIMENSION AREAS(50)PSIZE(50),OUTFL1(50),OUTFL2(50),OUTFL3(50)DIMENSION OUTFL4(50),STG1(50),DISCHB(50),AREA(50),DPTH(50)DIMENSION SI2ES(50),VOLA(400),VOLB(400),VOLE(400),VOLD(400)DIMENSION DIAMTR(5,400),VOLSED(400),VOLACT(400),SEDMNT(400)DIMENSION AVDPTH(400),AVSTG(400),STGIN(400),STGOUT(400),STAGO(400)DIMENSION STAREA(400),STGAR(400),ACOUT(400),VOLOUT(400),TMEIN(400)DIMENSION VOLTME(400),SEDTOT(400),SEDOUT(400),OUTPCT(20)DIMENSION PLGTME(400),VOLIN(400),DETTME(400),PLGCEN(400)DIMENSION STAG(400),AVDEP(400),SEDEND(400),SEDT(400)REAL OUTFL1pOUTFL2pOUTFL3p0UTFL4REAL NFLNIPMASSPINFLOWPNSTORMREAD(5,5100)NSTORM,CONSED,DEPOST,MASS,FLOW,TRP,FILTER,DENSTY,SGPVI

1SCOSIF(NSTORM.LT.0.0) DELTAT.CONSEDIF(NSTORM.LT.0.0) LLS.0.0IF(NSTORM.LT.0.0) CALL WASHIF(NSTORM .LT. 0.0) GO TO 3950READ(5,5000) DELTAT,DELPLG,SET,SHORT,FIX,FLOWAV,FRACTN,SLAG,DEAD

5000 F0RMAT(9F8.0)READ(5,5200)MP,M,NsNS,MS,LSREAD(5,5100)(PERCNT(NL),NL*1,NS)READ(5,5100)(SIZE(NL),NL•liNS)READ(5,5100)(STG1(I),I*1,N)READ(5,5100)(AREAS(I),I*1,N )

20 READ(5 , 5100)(DISCHB(I),I*1,N)5100 FORMAT(10F8.0)

PO 40 I*1,N

Page 133: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

AREA(I)AREAS(I)OISCH(I)=DISCH8(I)STAGE(I)=STG1(I)

40 CONTINUENNNeNSTORMDO 3900 IN=1,NNNREAD (5,5100) NSTORM , CONSED , DEPOSTJMASSPFLOW,TRP,FILTER,DENSTY,SGoVI

1SCOSREAD(5,5000) DELTAT , DELPLG , SET,SNORT,FIX,FLOWAV,FRACTNPSLAGsDEADREAD(5,5200)MP,M,NoNS,M5,LSLLS=LSIF(M.GT.LS) READ(5,5100)(INFLOW(I),I=1,M)IF(M.LE.LS.AND.LS.GT.0.0)READ(5,5100)(INFLOW(I),I=1,M)1F(M.LE.LS) CALL WASHIF(M.LE.LS)M=MXIF(CONSED.GT.1.99.AND.CONSEO.LE.2.01) GO 10 60GO TO 80

60 READ( 5, 5100) (CONCED(I),I.1,M)80 CONTINUE

DO 100 I=1,NAROLD(I)=AREA(I)STAG(I)=STAGE(I)

100 CONTINUECAPC0(1)=0.0PEAKIN=0.0DO 120 .1=1), M •IF(INFLOW(J) .GT. PEAKIN .AND. J •GT. LS) PEAKIN=INFLOW(J)

120 CONTINUEDO 140 1=1,NDPTN(I)=STG1(I)

140 CONTINUEIF(FLOW.GT.3.99.AND.FLOW.LE.4.01) GO TO 260DO 160 I=1,MPOUTFL1(I)=25.0OUTFL2(I)=25.0OUTFL3(I)=25.0OUTFL4(I)=25.0

160 CONTINUEIF(FLOW.GT.2.99.AND.FLOW.LE.3.01) GO TO 200DO 180 I=1,MPIF(FLOW.GT.0.99.AND.FLOW.LE.1.01) GO TO 240OUTFL1(I)=0.0OUTFL2(I)=0.0OUTFL3(I)=0.0OUTFL4(I)=100.0

180 CONTINUEGO TO 240

200 DO 220 I=1,MPOUTFL1(I)=100.00UTFL2(I)=0.0OUTFL3(I)=0.00UTFL4(I)=0.0

220 CONTINUE240 CONTINUE

GO TO 280260 CONTINUE

5200 FORMAT(6I8)READ(5,5100)(DPTN(I),I=1,MP)READ(5,5100)(OUTFL1(I),I=1,MP)RE4D(5,5100)(OUTFL2(I),I.1.01P)READ(5,5100)(OUTFL3(I),I=1,MP)READ(5,5100HOUTFL4(I),I.1,MP)

280 CONTINUE

123

Page 134: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

WRITE (5,5300)5300 FORMAT(1H1)

WRITE(6.5400)5400 F 0RMAT(// , 15X,"*************** ******* ********* THE DEPOSITS MODEL.

1 JANUARY 1979 «)WRITE (6,5500)

5500 FORMAT(////,45X,"***** INPUT CONTROL VARIABLES *****")WRITE (5,5500)

5600 FORMAT(//,15X,"MP",BX,"M",9X,"N",10X."NS",7X."MS",7X,"LS")WRITE(6 , 5700)MP,M.N,NS.MS,LS

5700 FORMAT(//p7X,6I10)WRITE (6,5800)

5800 FORMAT(//,15X,"NSTORM",4X,"CONSED",4X,"DEPOST",4X,"FLOW".6X,"TRP".17 X , "FILTER" ,4 X , "FIX",7X,"FRACTN",7X,"FLOWAV")WRITE (6,5900 )NSTORM , CONSED , DEPOST.FLOW,TRP,FILTER,FIX,FRACTN1 .FLOWAV

5900 FORMAT(//p9X,8F10.2,6X,F10.2)IF(FRACTN.LE.0.1) GO TO 300WRITE (5,6000)

6000 FORMAT(//,15WINFLOW SEDIMENT DISTRIBUTION",10WDENSITY CURRENT"1)GO TO 320

300 WRITE(6,6100)6100 FORMAT(//p15X,"INFLOW SEDIMENT DISTRIBUTION",10WCOMPLETE MIXING"

1)320 CONTINUE

WRITE( 5,6200)6200 FORMAT(// , 15X , "MASS",7X."VISCOS",5X,"DELTAT",4X,"DELPLG",4X

1 "DENSTY" , 6X,"SG",8X,"SET",6X,"SHORT",6Xs"SLAG",6X,"DEAD")WRITE(6,6300) MASS , VISCOS,DELTAT,DELPLG,DENSTY,SG,SET,SHORT,SLAG,

10E AD6300 FORMAT(//,11X,F10.3,2X,F10.408F10.2)

WRITE(6,6400)6400 FORMAT(//,45X,"***** DEPOSITS ERROR MESSAGES *****")

IF(VISCOS.LE.0.005) WRITE(6,6500)IF(VISCOS.GE.0.2) WRITE( 6,6500)

IF(M .GT. 400) WRITE(6,6450)6450 FORMAT(//.15X,"***** ERROR ***** • INFLOW ARRAY STORAGE EXCEEDED.

1")

6500 FORMAT(//,15X," ERROR ***** • USE DEFAULT VISCOS • 0.0114 CM.1S0./SEC.")IF(SG.LE.1.0) WRITE(6,6600)IF(SG•GE.4.0) WRITE(6,6600)IF(MS.GT•400) WRITE(6,6550)

6550 FORMAT(//,15X,"***** ERROR ***** • OUTFLOW ARRAY STORAGE EXCEEDED.1")

6600 FORMAT(//p15X,"***** ERROR ***** • USE DEFAULT SG • 2.65 •")IF(SG.LE.1.0) 5G-2.65IF(SG•GE.4.0) SG-2.65IF(VISCCS•LE.0.005) VISCOSm0.0114IF(VISCOS.GE.0.2) VISCOS .0.0114IF(N.GT•50) WRITE( 6,6550)IF(M.GT•400) Mm399IF(MS•GT•400) MS.399IF(N.GT•50) Nm49THE VALUE OF NS CANNOT EXCEED 10.IF(NS.GT•10) NSm10

6650 FORMAT(//,15X," ERROR ***** • BASIN GEOMETRY ARRAY STORAGE1 EXCEEDED.")IF(STAGE(1).GT.0.001) WRITE (5,6700)

6700 FORMAT(//p15X,"***** ERROR ***** • ELEVATION VALUES CHANGED TO STA1GE VALUES.")DO 340 I•lsN

124

Page 135: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

STAGE(I).STAGE(I)—STAGE(1)340 CONTINUE

AVDEP(1).0.0X1(1).0.0X2(1).0.0FLOWIN(1).INFLOW(1)AVTME=0.0CAPAC(1)80.0CAPNW(1).0.0AVDPTH(1)=0.0DO 360 J.2,NCAPAC(J) . (AREA(J)+AREA(.1-1))*(STA6E(J)-5TAGE(J-1))/2.04.CAPAC(J-1)CAPNW(J).CAPAC(J)IF(DISCH(J).E0.0.0) VAR-STAGE(J)X 1 (J)=CAPAC(J)—(DISCH(J)/2.0)*DELTAT*.08264X 2 (J).CAPAC(J)+(DISCH(J)/2.0)*DELTAT*.08264CAPCO(J) . (AREAS(J)+AREAS(J-1))*(STG1(J)—STG1(J-1))/2.0+CAPCO(J-1)

360 CONTINUEC IF THE PERMANENT POOL IS TO BE FILLED INTERNALLY BY THE PROGRAMC NO DUMMY VALUES SHOULD BE ENTERED TO FILL THE PREMANENT POOL.C LS MUST BE ENTERED AS ZERO.C IF YOU USE THE WASH MODEL, M MUST ALSO BE SET TO ZERO.

DO 980 J.1,NIF(DISCHB(J) • LE. 0.001) CAPOOL.CAPAC(J)

980 CONTINUEAVSTG(1)=0.0

C SEDMNTs SEDEMENT CONCENTRATION FOR EACH TIME INCREMENT (VOLUMETRIC).NFLNT(1)-0.0CAP-0.0SEDMNT(1).0.0SEDTOT(1)=0.0TOTVOL.0.0CENTME.0.0TOTAL-O.0STORM.0.0SEDOUT(1).0.0VOLSED(1).0.0ACINFL(1)=0.0MM.M+1DO 440 I.2,NSUM1.0.0SUM2.0.0DO 380 J.2,IIF(AREA(J).EQ.AREA(J —1)) GO TO 460DEPO. STAGE (I)—(STAGE(J)+STAGE(J-1))/2.0SUM1.DEPO**2.0*(AREA(J)—AREA(J-1))+SUM1SUM2.0EP0*(AREA(J)—AREA(J-1))+SUM2

380 CONTINUEIF(SUM2.LE.0.0) GO TO 400AVDPTH(I).SUM1/SUM2GO TO 420

400 AVOPTH(1).0.0420 CONTINUE440 CONTINUE

GO TO 540460 DO 520 J.2,N

IF(CAPAC(J).LE.0.0) GO TO 480AVDEP(J).(CAPAC(J-1)IAVDEP(J-1)+(CAPAC(J)—CAPAC(J-1))*(STAGE(J)+ST

1AGE(J-1))/2.0)/CAPAC(J)AVDPTH(J).(STAGE(J)—AVDEP(J))*2.0GO TO 500

480 AVDPTH(J).0.0AVDEP(J)-0.0

125

Page 136: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

500 CONTINUE520 CONTINUE540 CONTINUE

DO 560 S.MM,MSINFLOW(I).0.0

560 CONTINUEACINFL(1)=CAPOOLBPOOL.CAPOOLVOLUME(1).CAPOOLDO 580 1=2,MSACINFL ( I ). ACINFL(I-1)+C(INFLOW(I-1)+INFLOW(1))/2.0)*OELTAT*.08264VOLUME(I)=ACINFL(I)-ACINFL(I-1)

580 CONTINUESTAREA(1).0.0STP(1)=CAPOOLSTGAR(1).0.0SUMTME.0.0VOLTOT=0.0STARTV(1)=CAPOOLSTAGEA(1).0.0CAPACA(1)=0.0DISCHA(1).0.0T1(1).0.0MR.(MS)/(DELPLG/DELTAT)+.01PEAK-C. 0DO 600 I=MM,MSCONCED(I)=0.0

600 CONTINUEIF(CONSED.GT.1.99.AND.CONSED.LE.2.01) MASS-0.0DO 740 4=2.MSIF(CONSED.GT.1.99.AND.CONSED.LE.2.01) GO TO 620IF(SET .LT. 1.0) ST-Z.0SEDMNT(J)*(VOLUME(J)**SET)IF(LS.GE.J) SEDMNT(J)=0.0IF(SLAG.GT.O.O.AND.SLAG.LT.1.01) 5EDMNT(J-1).SEDMNT(J)IF(SLAG.GT.1.01.AND.J.GT.2) SEDMNT(J-2)=SEDMNT(J)IF(J.EO.MS.AND.SLAG.GT.0.0) SEDMNT(J).0.0IF(J.EO.MS.AND.SLAG.GT.0.0) SEDMNT(J-1)-0.0GO TO 640

620 SEDMNT(J).(CONCED(J)+CONCED(J-1))*VOLUME(J)/(SG*2000.0)MASS.MASS+0.001359*(CONCED(J)+CONCED(J-1))*VOLUME(J)/2.0

640 CONTINUEIF(J.GT.LS) STORM • STORM + VOLUME(J)SEDTOT(J)=SEDTOT(J-1)+5EDMNT(J)IF(CONSED.GT.1.99.AND.CONSED.LE.2.01) GO TO 650IF(SLAG.GT.O.O.AND.SLAG.LT.1.01) SEDTOT(J-1)=SEDTOT(J)IF(SLAG.GT.1.01.4ND.J.GT.2) SEDTOT(J-2)=SEDTOT(J)

650 STP(J).STP(J-1)+VOLUME(J)STPV(J-1)=STARTV(J-1)+VOLUME(J)

C DO AN ITERATION TO FIND STAGE FROM STPVDO 660 K.2,NIF(STPV(J-1).LT.X2(K))G0 TO 680IF(STPV(J-1).GT.X2(N)) GO 10 3860

660 CONTINUE680 STAGEA(J).STAGE(K-1)+((STPV(J-1)-X2(K-1))/(X2(K)-X2(K-1)))*(STAGE(

1K)-STAGE(K-1))AVSTG(J)=AVOPTH(K-1)+USTPV(J-1)-X2(K-1))/(X2(K)-X2(K-1)))*(AVDPTH1(K)-AVDPTH(K-1))CONTINUEDO 700 KK.2,NIFISTAGEACJI.LT.STAGE(KK)) GO 10 720IF(STAGEA(J-1).GT.STAGE(N) ) GO 10 3860

700 CONTINUE

126

Page 137: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

720 CAPACA(J).X1(KK-1)+((STAGEA(J)—STAGE(KK--1))/(STAGE(KK)...STAGE(KK-1)1))*(X1(KK)-•X1(KK-1))

C DO AN ITERATION TO FIND DISCHARGE FOR STAGEADISCHA ( .1).DISCH(KK-1)+USTAGEA(J)—STAGE(KK-1))/(STAGE(KK)—STAGE(KK

1- 1)))*(DISCH(KK)—DISCHMK--1))IF(DISCHA(J).GT.PEAK)PEAK.DISCHA(J)CONTINUESTARTV(J).CAPACA(J)IF(STARTV(J).LT.0.0) STARTV(J).0.01 1( 4 ).(J..-1)*0ELTATIF(J.GT.LS) SUMTME.SUMTME+(4-1.-LS)*DELTATSVOLUME(J)1F(J.GT.LS) VOLTOT.VOLTOT+VOLUME(J)STAREA(J) . ABSUAVSTG(J)+AVSTG(J-1))*(DELTAT/2.0))STGAR(4).STAREA(J)+STGAR(J-1)

C THIS PART OF THE PROGRAM DIVIDES THE OUTLET HYDROGRAPH INTO PLUGS OF EQUALC TIME INCREMENT DELPLG. THE PLUG IS THEN ROUTED THROUGH THE RESERVOIR ANDC THE DETENTION TIME,STAGE AT OUTFLOW, AVERAGE DEPTH AND THE VOLUME OFC THE PLUG IS DETERMINED.

740 CONTINUEPFLNT.0.0IF(CONSED.GT.1.99.AND.CONSED.LE.2.01) GO TO 860

C PFLNT WILL BE THE PEAK INFLOW CONC_DO 820 JS.2sMIF(VOLUME(JS).E0.0.0) GO TO 780NFLNT(JS).(SEDMNT(JS)*1.*MASS*735.48)/(VOLUME (4S)*SEOTOT(M))IF(NFLNT(JS).GT.PFINT) GO TO 760GO TO 800

760 PFLNT.NFLNT(JS)GO TO 800

780 NFLNT(JS).0.0800 CONTINUE820 CONTINUE

14.0DO 840 IK.2,MIF(NFLNT(IK).E0.0.0) GO TO 840IJ.IJ+1SFLNT(IJ).NFLNT(IK)

840 CONTINUEGO TO 900

860 DO 880 J 5 .1,MNFLNT(JS)=CONCED(JS)IF(NFLNT(JS).GT.PFLNT)PFLNT.NFLNT(JS)

880 CONTINUE900 CONTINUE

DO 920 J.J.MM,MSNFLNT(44).0.0

920 CONTINUEDO 940 1.1,4

SED(I,J).0.0DIAMTRUI,J).0.0VELOC(I,4).0.0FALL(I,J).0.0VOL(I,J)=0.0PCT(I,J).0.0VOLC(I,J).0.0

940 CONTINUESEDPLG(1).0.0SUMVOL.0.0STRMOT=0.0SEDEND(1).0.0DEPTH(1).0.0ACOUT(1) . 0.0

127

Page 138: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

PLGVOL(1).0.0PLGTME(1).0.0VOLOUT(1).0.0VOLIN(1).DEADTMEIN(1).0.0SEDT(1).0.0DETTME(1)=0.0PLGCEN(1).0.0VOLTME(1)=0.0AREAA(1).0.0AREAB(1).0.0AREAC(1).0.0AREAD(1).0.0VOLA(1).0.0VOLB(1).0.0VOLE(1).0.0VOLD(1).0.0DEPTH(1)80.0DEPTH2(1)=0.0DEPTH3(1).0.0DO 960 Lm2,MSACOUT(L).ACOUT(L -1)+((DISCHA(L-1)+DISCHA(L))/2.0)*DELTAT*.08264

960 CONTINUEPSTAGO.0.0PEFLNT-0.0

C PEFLNT WILL BE THE PEAK EFFLUENT CONC.C PSTAGO WILL BE THE PEAK STAGE

DEL.0.0DO 2420 NN.21MRPLGTME(NN).PLGTME(NN -1) + DELPLGLR.(DELPLG+.01)/DELTATP.LR*(NN -1)+1PLGVOL(NN).ACOUT(P)-ACOUT(P-LR)VOLIN(NN).VOLIN(NN-1)+PLGVOL(NN)*SHORTIF(VOLIN(NN).GT.CAPOOL.AND.VGLIN(NN-1).LT.CAPOOL) T.PLGTMENN)IF(T.GT.DEL) TwDELIF(SHORT.LE.1.000)SHORT.1.0PLGCEN(NN) • (PLGTME(NN) + PLGTME(NN-1)1/2.0

C DO AN ITERATION TO FIND TMEIN FROM VOLINDO 1000 NP.2..MIF(VOLIN(NN) .LT. STP(NP) .AND. VOLIN(NN) .LE. STP(NP-1)) GO TO 10

125IF(VOLIN(NN).LT.STP(NP)) GO 10 1020

1000 CONTINUEGO TO 1025

1020 TMEIN(NN)=T1(NP-1)+((VOLIN(NN)-STP(NP- 1) )/(STP(NP)-STP(NP-1)))*DEL1TATGO TO 1035

1025 TMEIN(NN)-0.01035 CONTINUE

FLOWIN(NN)=INFLOW(NP)IF(FLOWAV.GT.0.0) GO TO 1040GO TO 1120

1040 DO 1060 I•1,NSIF(FLOWIN(NN).E0.0.0.AND.FLOWIN(NN-1).EC.0.0) GO TO 1080SIZEST(I,NN).(2.*FLOWAV/(FLOWIN(NN)+FLOWIN(NN-1)))**.3*PERCNT(I)IF(SIZEST(1,NN).GT.100.) SIZEST(IpNN)=100.

1060 CONTINUEGO TO 1100

1080 SIZEST(I,NN).PERCNT(I)1100 CONTINUE1120 CONTINUE

VOLTME(NN)m(TMEIN(NN)+TMEIN(NN-I))/2.0

128

Page 139: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

DETTME(NN).PLGCEN(NN)-VOLTME(NN)IF(DETTME(MN).LT.0.0) DETTME(NN).0.0SEDTINN) . SEDTOT(NP-1)+((VOLIN(NN)-STP(NP-1))/(STP(NP)-STP(NP-1)))

1*(5EDTOT(NP)-SEDTOT(NP-1))IF((VOLIN(NN)-STP(NP -1)) .LT. 0.0) SEDTINN)80.0SEDOUT(NN).(SEDT(NN)-SEDT(NN-1))/SEDTOT(M)STGIN(1).0.0STGOUT(1).0.0STAGO(1).10.0DO 1140 II.2,MSIF(VOLTME(NN).I.T.T1(II)) GO 10 1160IF(VOLTME(NN).GE.T1(MS)) GO TO 2080

C DO AN ITERATION TO FIND DEPTH FOR VOLTME1140 CONTINUE1160 STGIN(NN).STGAR(II-1) 4. A8S(UVOLTMENN)-T1(II-1))/(71(II)-T1(II-1))

1)*(STGAR(II)-STGAR4II-1)))CONTINUEDO 1180 II.2,MSIF(PLGCEN(NN).LT.T1(II)) GO TO 1200

C DO AN ITERATION TO FIND DEPTH FOR PLGTME1180 CONTINUE1200 STGOUT(NN).STGAR(II-1)+((PLGCEN(NN)-T1(II-1))/(71(II)-71(II-1)))*(

1STGAR(II)-STGAR(II-1))STAGO(NN).STAGEACII-1)+((PLGCEN(NN)-71(II-1))/(71(II)-T1(II-1)))*(1STAGEA(II)-STAGEA(II-1))IF(DETTME(NN).LE.0.0) DETTME(NN).0.0IF(STAGO(NN).GT.PSTAGO) GO TO 1220GO TO 1240

1220 PSTAGO.STAGO(NN)1240 IF(DETTME(NN).E0.0.0) GO TO 2120

DEPTH(NN).(STGOUT(NN)-STGIN(NN))/DETTME(NN)If(DEPTH(NN).LE.0.0) DEPTH(NN)=0.0DEPTH1(NN)*0.75*DEPTH(NN)DEPTH2(NN)=0.5*DEPTH(NN)DEPTH3(NN)=0.25*DEPTH(NN)DO 1260 LM=2,NIF(DEPTH(NN).LT.STAGE(0)) GO 10 1280

1260 CONTINUE1280 VOLA(NN). CAPAC(LM-1)+((DEPTH(NN)-STAGE(LM- 1))/(STAGE(LM)-STAGE(LM

1-1)))*(CAPAC(LM)-CAPAC(LM -1))AREAA(NN)=AREA(LM-1)+((DEPTH(NN)-STAGE(LM-1))/(STAGE(LM)-STAGE(LM -

11)))*(AREA(LM)-AREA(LM-1))CONTINUE

C CAPSAV WILL COMPUTE THE CAPACITY AT PEAK STAGEDO 1300 I.1.2,NIF(PSTAGO.LE.STG1(IJ)) GO TO 1320

1300 CONTINUEGO TO 1340

1320 CAPSAV.CAPAC(IJ-1)+ ((PSTAGO-STG1(IJ-1))/(STG1(IJ)-STG1(IJ-1)))1*(CAPAC(IJ)-CAPACCIJ-1))

1340 DO 1360 LM=2,NIFIDEPTH1(NN).LT.STAGE(LM)) GO 10 1380

1360 CONTINUECONTINUE

1380 VOLUNN)=CAPAC(LM-1)+C(DEPTH1(NN)-ST4GE(0-1))/(STAGE(LM)-STAGE(LM1-1)))*(CAPACUM)-CAPAC (LM-1)/AREAB(NN)RAREAUM-1)+C(DEPTH1(NN)-STAGE(LM-1)1/(STAGE(LM)-STAGE(LM

1-1)))*(AREA(LM)-AREA(LM-1),CONTINUEDO 1400 LM=2,NIF(DEPTH2(NN).LT.STAGE(LM)) GO TO 1420

1400 CONTINUE1420 VOLE(NN). CAPAC(LM-1)+((DEPTH2(NN)-STAGE(LM-1,)/(STAGE(LM)-STAGE(1.

129

Page 140: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

1M-1)))*(CAPAC(LM)—CAPACUM —1))AR E AC(NN)*AR EA ( LM-1) + ( lDEPTH 2 (NN) —STAGE(LA-1))/(STAGE(LM)—STAGEILM

1 -13))*(AREA(01)4.AREA(M-1))CONTINUEDO 1440 LM=2,NIF(DEPTH3(NN).LT.STAGE(LM)) GO TO 1460

1440 CONTINUE1460 VOLD(NN)* CAPAC(LM-1)+((DEPTH3(NN)—STAGE(LR-1))/(STAGEUM)—STAGEIL

1M-1)))*(CAPAC(LM)—CAPACILM-1))AREAD ( NN )= AREA(LM-1 )+C(DEPTH3(NN)—STAGE(LM-1))/(STAGE(LM)—STAGE(LM

1 .41)))*(AREAILM).4AREA(01-1))CONTINUEVOL(1,NN).VOLA ( NN)—VOL8(NN)VOL(2 , NN).VOLB(NN)—VOLE(NN)VOL(3,NN).VOLE(NN)—VOLD(NN)VOL(4,NN).VOLD(NN).FALL(1,NN).0.875*DEPTH(NN)*FIXfALL(2,NN)=0.625*DEPTHINN)*FIXFALL(3,NN).0.375*DEPTH(NN)*FIXFALL(4,NN).0.125*DEPTH(NN)*FIXIF(PLGVOL(NN).1T..00001) GO TO 2080VELOC(1,NN).FALL(1,NN)/(DETTME(NN))VELOC(2,NN)-FALL(2)NN)/(DETTME(NN))VELOC(3,NN).FALL(3,NN)/IDETTMECNN))VELOC(4,NN)=FALL(4,NN)/(DETTME(NN))DIAMTR(1 , NN).30RT(VELOC(1,NN)*VISCOS/(51.5*(SG-1)))DIAMTR(2 , NN).SORTIVELOC(2,NN)*VISCOS/(51.5*(SG-1)))DIAMTR(3 , NN)=SORTCVELOC(3,NN)*VISCOS/(51.5*(SG-1)))DIAMTR( 4, NN)=SORT(VELOC(4,NN)*VISCOS/(51.5*(SG-1)))CONTINUEIF(FLOWAV.GT.0.0) GO TO 1640DO 1480 LP.2,NSIF(DIAMTR(1,NN).L.T.SIZE(LP)) GO TO 1500

1480 CONTINUE1500 PERCT(1 , NN).PERCNTUP -1)+( (DIAMTR(1,NN)—SI2E(LP-1))/(SIZE(LP)—SIZE

1(LP —1)))*(PERCNT ( LP)—PERCNT(LP —1))DO 1520 LP*2,NSIF(DIAMTR(2/NN).LT.SI2E(LP)) GO TO 1540

1520 CONTINUE1540 PERCT(2,NN)=PERCNT(LP-..1)+((DIAMTR(2,NN)—SIZE(LP-1))/(SI2E(LP)—SIZE

1(LP*1))).0(PERCNTUP) —PERCNT(LP —1))DO 1560 LPs2,NSIF(DIAMTR(3pNN).LT.SIZE(LP)) GO TO 1580

1560 CONTINUE1580 PERCT(3,NN).PERCNT(0-1)+I(DIAMTR(3,NN)—SIZE(LP-1)3/(5IZEILP)—SIZE

1(LP-1)))*(PERCNT(LP)—PERCNT(0-1))DO 1600 LP.2,NSIF(DIAMTR(4,NN).I.T.SIZE(LP)) GO TO 1620

1600 CONTINUE1620 PERCT(4,NN)=PERCNT(LP-1)+((DIAMTR(4,NN)—SIZE(LP-1))/(SIZE(LP)—SI2E

l(LP-1)))*(PERCNT(LPC—PERCNT(LP-1)/GO TO 1820

1640 DO 1660 LP.2/N5IF(DIAMTR(1,NN).LT.SIZE(LP)) GO TO 1680

1660 CONTINUE1680 PERCT(1,NN).SIZEST(LP-1,NN) .0(IDIAMTR(1,NN)—SIZE(LP-1))/(SIZE(LP)

1—SIZE( LP-1)))*(SIZEST(LP,NN)—SIZEST(LP-1,NN))DO 1700 LP.2,NSIftDIAMTR(2,NN).I.T.SIZE(LP)) GO TO 1720

1700 CONTINUE1720 PERCT(2,NN).SI2EST(LP-1,NN) +((DIAMTR(2,NN)—SIZE(LP-1))/(SIZE(0)

1 —SIZE( 0-1 )))*(SIZEST(LP,NN)—SIZEST(LP-1,NN))00 1740 LP=2,NS

130

Page 141: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

IF(DIAMTR(3,NN).LT.SIZE(LP)) GO TO 17601740 CONTINUE1760 PERCT(3,NN).SIZEST(LP-1,NN) +((DIAMTR(3,NN)—SIZE(LP1))/(SIZE(LP)

1—SIZE( LP• 1 )))*(SIZEST(LP,NN)—SIZEST(LP-....1,NN))DO 1780 LF*2,NSIF(DIAMTR(41,NN).LT.SI7E(LP)) GO TO 1800

1780 CONTINUE1800 PERCT( 4, NN).SIIEST(LP1,NN) +((DIA)TR(4 , NN)SIZE(LP-1))/(SIIE(LP)

1 SIZE(LP1)))*(SIZEST(LP,NN)-4SIZEST(LP1,NN))1820 CONTINUE

VOLC( 1, NN).V0L(1 , NN)*(2.0*AREAB(NN)/(AREAA(NN)+4REA8(NN)))VOLC( 2, NN) . V01(2 , NN)*(AREAC(NN)*2.0/(AREAC(NN)+AREA8(NN)))VOLC(3 , NN).VOL(3,NN)*(2.0*AREAD(NN)/(AREAC(NN)+AREAD(NN)))IF(VOL(2,NN).LE.0.0) GO TO 2080IF(FR#CTN.GT.0.0j) GO TO 2000PCT(1,NN) .PERCT(4,NN)0IFF(NN).(STAGO(NN)...-VAR)4.2.0IF(DIFF(NN).GT.0.0.AND.DIFF(NN).LT..125*DEPTH(NN)) GO TO 1840GO TO 1980

1840 DIAMTR(5 , NN)*SORT(DIFF(NN)*VISCOS/(51.5*(SG-1)*DETTME(NN)))IF(FLOWAV.GT.0.0) GO TO 1900DO 1860 LP=2,NSIF(DIAMTR(5,NN).LT.SIZE(LP)) GO TO 1880

1860 CONTINUE1880 PERCT(5,NN).PERCNT(Lp1)+((OIAPjTR(5,NN)—SIZE(1.11-.- 1) )/(SIZE(LP)SIZE

l(LP ..1)))*(PERCNT(LP)PERCNT(LP —1))GO TO 1960

1900 DO 1920 LP.2,NSIF(DIAMTR(5,NN).LT.SIZE(LP)) GO TO 1940

1920 CONTINUE1940 PERCT(5,NN)=SIZEST(LP-.-1,NN) +((DIAMTR(5,NN)—SIIE(LP-1))/(SIZE(LP)

1 —SIZE(LP1)))*(SIZEST(LPoNN)—SIZEST(LP...1,NN))1960 CONTINUE

IF(DIFF(NN).1.7.0.125*DEPTH(NN))PCT(1,NN).PERCT(5,NN)1980 CONTINUE

IF(PERCT(1,NN).I.T.O.0)PERCT(1,NN).0.0IF(PERCT(2,NN).LT.0.0)PERCT(2,NN).0.0IF(PERCT(3,NN).L7.0.0)PERCT(3,NN).0.0IF(PERCT(4,NN).LT.0.0)PERCT(4,NN).0.0IF(PERCT(1,NN).GT.100.)PERCT(1,NN).100.IF(PERCT(2,NN).GT.100.)PERCT(2,NN)=100.IF(PERCT(3,NN).GT.100.)PERCT(3,NN).100.IF(PERCT(4,NN).GT.100.)PERCT(4,NN).100.PCT(2,NN).(VOL(2,NN)*PERCT(4,NN)+VOLC(1,NN)*(PERCT(3,NN)..-PERCT(4,N

1N)))/VOL(2,NN)PCT(3,NN).(VOL(3,NN)*PERCT(4,NN)+VOLC(2,NN)*(PERCT(2,NN)—PERCT(4,N

1N)))/VOL(3,NN)PCT(4,NN).(VOL(4,NN)*PERCT(4,NN)+VOLC(3,NN)*(PERCT(1,NN)—PERCT(4,N

1N)))/VOL(4,NN)GO TO 2020

2000 PCT(1,NN)•0.0PCT(2,NN)80.0PCT(3,NN).0.0PCT(4,NN).PERCT(4,NN)

2020 CONTINUEDO 2040 LM-2,NIF(STAGO(NN).LT.DPTH(LM)) GO TO 2060

2040 CONTINUE2060 SED(1,NN).PCT(1/NN)*SE0OUT(NN)*(OUTFL1(LM-1)+((STAGO(NN)—OPTH( L M-1

1))/(OPTH(LM) — OPTH(LM-11))*(OUTFL1(LM)—OUTFL1(01-1 ) ))SED(2,NN)=PCT(2,NN)SEEDOUT(NN)*(OUTFL2(LM-1)+(ISTAGO(NN)—DPTH(LM-11))/(DFTH(LM) — OPTH(LM .-1)))*(OUTFL2(LM)-0UTFL2(04-1)))SED(3 , NN).PCT(3JINN)*SEDOUT(NN)*(0UTFL3(LM..-1)+((STAGO(NN).-..DPTH(L-1

131

Page 142: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

1))/(DPTH(LM) —DPTH(LM ....1)))*(OUTFL3(LM)OUTFL3(01-+1)))SED(4,NN)*PCT(4,NN)*SEDOUT(NN)*(OUTFL4(LM-1)+((STAGO(NN) —DPTH(LM-11))/(OPTH(LM) —OPTH(LM-1)))*(OUTFL4(LM)—OUTFL4(LM-1)))SEDPLG(NN )•(SED(1,NN)+SED(2,NN)+SED( 3,NN )+SED(4oNN) ) /100.0IF(SEDPLG(NNI.GT.SEDOUT(NN)*100.)SEDPLG(NN).100.4.SEDOUT(NN)GO TO 2100

2080 SEDPLG(NN).0.0PCT(1,NN).0.0PCT(2,NN)=0.0PCT(3,NN).0.0PCT(4,NN).0.0

2100 CONTINUEGO TO 2140

2120 SEDPLG(NN).100.0*SEDOUT(NN)PCT(1,NN).100.0PCT(2yNN).100.0PCT(3,NN).100.0PCT(4,NN)*100.0IF(DEPTH(NN).LE.0.0) DEPTH(NN).0.0

2140 CONTINUESEDEND(NN).SEDEND(NN-1)+SEDPLG(NN)DO 2200 L5=1,NSIF(FLOWAV.LE.0.0) SIZEST(LS,NN).PERCNT(LS)PCTOUT(L.S,1).0.0IF(SEDPLG(NN).LE.0.0) GO TO 2160SUOUT(LSAN).100.*SEDOUT(NN)*SIIEST(LS,NN)/SEDPLG(NN)IF(SIZOUT(LS,NN).LE.0.0)SIZOUT(LS,NN).0.0GO TO 2180

2160 SIZOUTILS,NN)=SIZEST(LSPNN)SEDPLG(NN).0.0

2180 CONTINUEIF(SIZOUT(LS,NN).GT.100.),SIZOUTILS,NN).100.PCTOUT(LS,NN)=PCTOUT(LS,NN-1)+SEDPLG(NN)+SIZOUT(LS,NN)

2200 CONTINUEVOLOUT(NN)=SEDOUT(NN)+VOLOUT(NN-1)ACT.1.001*(STRMOT+CAPOOL—DEAD)IF(ACT •GT. STORM) GO TO 3440IF(PLGVOL(NN).E0.0.0) GO TO 2240EFLNT(NN)u(SEDPLG(NN)/PLGVOL(NN))*MASS*7.3548IF(EFLNT(NN).GT.PEFINT) GO TO 2220GO TO 2260

2220 PEFLNT.EFLNTINN)GO TO 2260

2240 EFLNT(NN).0.02260 CONTINUE

IF(PLGTMEINN) •GT. DEL) TOTAL.TOTAL+PLGCEN(NN)*PLGVOL(NN)IF(PLGTME(NN).GT.DEL) TOTVOL • TOTVOL +PLGVOL(NN)IF(TOTVOL.GT.0.0) CENTME.TOTAL/TOTVOL—SUMTME/VOLTOTIF(VOLIN(NN) .GT. 8POOL) AVTME.AVTME+(DETTME(NN))*PLGVOL(NN)IF(VOLIN(NN).GT.E(POOL) SUMVOL.SUMVOL+PLGVOL(NN)IF(SUMVOL.GT.0.0) DETAVE+AVTME/SUMVOLIF(VOLIN(NN).GT.BPOOL) STRMOT.STRMOT+PLGVOL(NN)IF(CENTME.LE.0.0) CENTME.DETAVESTRMTM.(CENTME*STRMOT+(DETTME(NN-1)*(STORM—STRMOT)))/STORMAVETMEs(DETAVE*STRMOT+DETTME(NN-1)*(STORM—STRMOT))/STORMIF(STRMOT.GT.STORM/AVETME.DETAVEIF(STRMOT.GT.STORM) STRMOT-STORMIF(DEPOST.LT.1.99) GO 10 2420DEP(1,NN)=MASS+0.0007360*(100.0—PCT(1,NN))*SEDCUT(NN)*(CUTFL1(LM- 11)+( CSTAGOINNi—OPTHUM-1))/(OPTH(LM)—DPTH(LM-1)))*(OUTFL1(01) —OUTFL11(LM-1)))1(10U00.0*DENSTY)DEP(2,NN).MASS*0.0007360*(100.0—PCT(2,NN))*SEDOUT(NN)*(OUTFL2(LM - 11)+USTAGO(NN)—DPTH(LM-1))/(DPTH(LM)—OPTH(LM-1)))*(OUTFL2(LM)—OUTFL

132

Page 143: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

1) +I(STAGO ( NN) -DPTH(LM- 1))/(DPTH(LM) - DPTH(L(j-1)))*(OUTFL3(LM)-OUTFL13(LM-1)))/(10000.0*DENSTY)DEP (4, NN ) =MASS* 0 . 0007360 *(100.0-PCT(4,NN))*SEDOUT(NN)*(OUTFL4(LK-1

1) +USTAGO(NN) -DPTH(LM-1))/(DPTH(LM)DPTH(LM-1)))*(OUTFL4(LM)-OUTFL14(LM-1)))/(10000.0*DENSTY)

C THIS PART OF THE PROGRAM DETERMINES THE CHANGE IN BASIN CAPACITYC DUE TO DEPOSITION.

DO 2400 I=1,N1F(AVDPTH(I).LT.DEPTH3(NN)) GO TO 2320IF(AVDPTH(I).LT.DEPTH2(NN)) GO TO 2340IF(AVOPTH(I).LT.DEPTH1(NN)) GO TO 2360IF(AVDPTH(I).L.T.DEPTH (NN)) GO TO 2380CAPNW(I)=CAPNW(I)-(DEP(4,NN)+DEP(3,NN)+DEP(2,NN)+DEP(1,NN))GO TO 2400

2320 CAPNW(I)=CAPNW(I)-DEP(4,NN)*AVDPTH(I)/DEPTH3(NN)GO TO 2400

2340 CAPNW(I)=CAPNW(I)-(DEP(4,NN)+DEP(3,NN)*(AVDPTH(I)-DEPTH3(NN))/DEPT1H2(NN))GO TO 2400

2360 CAPNW(I)=CAPNW(I)-(DEP(4,NN)+DEP(3,NN)+DEP(2,NN)..(DEP(2,NN)*(AVOPT1H(I)-DEPTH2(NN))/DEPTH1(NN)))GO TO 2400

2380 CAPNW(I)=CAPNW(I)-(DEP(4,NN)+DEP(3.NN)+DEP(2.NN)..(DEP(1.NN)*(AVOPT1H(I)-DEPTH1(NN))/DEPTH(NN)))

2400 CONTINUE .2420 CONTINUE3440 NE=NN-1

DO 3500 I=1.NSIF(NN.LT.MR) GO TO 3460IF(SEDEND(MR).LE.0.0) GO TO 3460OUTPCT(I)=PCTOUT(I,MR)/SEDEND(MR)GO TO 3480

3460 IF(SEDEND(NN).LE.0.0) OUTPCT(I)=0.0IF(SEDEND(MR).LE.0.0) OUTPCT(I)=0.0IF(SEDEND(NN).GT.0.0) OUTPCT(I)=PCTOUT(I,NN)/SEDEND(NN)

3480 CONTINUE3500 CONTINUE

TRAP.(100.0-SEDENDINN-1))SEDAV2=(SEOEND(NN-1)/(STRMOT+BPOOL-DEAD))*MASS*7.358IF(STRMOT.GT.0.0) SEDAVE=SEDEND(NN-1)*MASS*7.358/STRMOTIF(STRMOT.LE.0.0) SEDAVE=0.0WRITE (6,6900)

6900 FORMAT(1H1)WRITE (6,7000)

7000 FORMAT(45X,"***** STORM EVENT SUMMARY *****")WRITE(67100) CAPOC).

7100 FORMAT(//,15X."PERMANENT POOL CAPACITY",18X,"=",F10.2,5X,"ACRE -FT"1)WRITE(6,7200) DEAD

7200 FORMAT(//,15X,"DEAD STORAGE".29X."=",F10.2,5X,"ACRE-FT")WRITE(6,7300) STORM

7300 FORMAT(//,15X,"STORM RUNOFF VOLUME".22X,"8",F10.2,5WACRE-FT")WRITE(6,7400) STRMOT

7400 FORMAT(//.15X,"STORM VOLUME DISCHARGED ".17X."=".F10.2,5WACRE-FT1")WRITE(6.7500) CAPSAV

7500 FORMAT(//,15X,"POND VOLUME AT PEAK STAGE".16X,"=",F10.2,5X,"ACRE-FIT")WRITE(6,7800) PSTAGOWRITE(6,7600) PEAKIN

7600 FORMAT(//,15X,"PEAK INFLOW RATE",25X."="pF10.2.5X,"CFS")WRITE(6,7700) PEAK

7700 FORMAT(//,15X,"PEAK DISCHARGE RATE"..22X."="5, F10.2,5X."CFS")

133

Page 144: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

7800 FORMAT(//,15X,"PEAK STAGE",31X,".",F10.2,5X,"FT")WRITE(6,7900) PFLNT

7900 FORMAT(//015X,"PEAK INFLOW SEDIMENT CONCENTRATION",7X,"•",F10.1,5X1,"MG/L")WRITE(6,8000) PEFLNT

8000 F0RMAT(//,15X,"PEAK EFFLUENT SEDIMENT CONCENTRATION",5X,".",F10.1,15X,"MG/L")WRITE(6,8050) SEDAVE

8050 FORMAT(//,15X,"STORM AVERAGE EFFLUENT CONCENTRATION", 5X,".",1F10.1,5WMG/L")WRITE(6,8100) SEDAV2

8100 FORMAT(//,)_5X,"AVERAGE EFFLUENT SEDIMENT CONCENTRATION",2X,".",1F10.1,5X,"MG/L")WRITE(6,8200) TRAP

8200 FORMAT(//,15X,"BASIN TRAP EFFICIENCY",20X,".",F10.2,5X," ")WRITE(6,8300) AVETME

8300 FORMAT(//,15X,"DETENTION TIME OF FLOW WITH SEDIMENT",5X,".",F10.2,15X,"HRS")WRITE(6,8400)CENTME

8400 FORMAT(//,15X,"DETENTION TIME FROM HYDROGRAPH CENTERS",3X,".",F10.12,5X,"NRS")WRITE(6,8450) STRMTM

8450 FORMAT(//,15X,"DETENTION TIME INCLUDING STORED FLOW",5X,".",F10.2,15X, "MRS")WRITE(6,8500) MASS

8500 FORMAT(//,15X,"SEDIMENT LOAD",28X,".",F10.215X,"TONS")WRITE (6,9200)IF(FLOWAV.EQ.0.0) GO TO 3540WRITE(6,8600) FLOWAV

8600 FORMAT(//,15X," PARTICLE SIZE DISTRIBUTION AT INFLOW RATE1 •",F10.2,2X,"CFS *****")WRITE(6,8700)(SIZE(I),Ia1,NS)

8700 FORMAT(///,15X,"SIZE (MM)", 10F8.4)WRITE(6,8800)(PERCNT(I),I=1,NS)

8800 FORMAT(/ ,15X," FINER",1X,10F8.1)DO 3520 I.1,NSSIZES(I).(FLOWAV/PEAKIN)**.3*PERCNT(I)IF(SIZES(I).GT.100.) SIZES( 1)-100.

3520 CONTINUEWRITE(6,8900) PEAKIN

8900 FORMAT(//,15X," ***** PARTICLE SIZE DISTRIBUTION AT INFLOW RATE1 .",F10.2,2X,"CFS *****")WRITE(6,8700)(SIZE(I),I.1,NS)WRITE(6,8800) (SIZES(I),I.1,NS)GO TO 3560

3540 WRITE(6,9000)9000 FORMAT(//p15X,"***** PARTICLE SIZE DISTRIBUTION OF SEDIMENT INFLOW

1 WRITE(6,8700)(SIIE(I),I.1,NS)WRITE(6,8800)(PERCNT(I),I.1,NS)

3560 CONTINUEWRITE (6,9100)

9100 FORMAT(//,15X," ***** PARTICLE SIZE DISTRIBUTION OF EFFLUENT 1")WRITE(6,8700)(SIZE(I),I.1,NS)WRITE(6,8800)(OUTPCT(I),I.1,NS)

9200 FORMAT(1H1)WRITE (6,9300)

9300 FORMAT(//,32X,"***** OUTFLOW WITHDRAWAL DISTRIBUTION le)WRITE(6,9400)

9400 FORMAT( ///,15X, "STAGE",10X, "OUTFLOW 1", 10X, "OUTFLCW 2", 10X, "OUTFLO1W 3",10X,"OUTFLOW 4")WRITE (6,9500)

134

Page 145: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

9500 FORMAT(/,16X,"(FT)",13X,"( )",16X,"( )",16x,"( )",16x,"( )•)DO 3580 IR=1,MDWRITE(6,9600) DPTH(IR),OUTFL1(IR),OUTFL2(IR),OUTFL3(IR),OUTFL4(IR)

9600 FORMAT( 10X,F10.2,7X,F10.2,9X,F10.2,9X,F10.2,9X,F10.2)3580 CONTINUE

WRITE(6,9700)9700 FORMAT(1H1)

WRITE (6,9800)9800 FORMAT(//y42X," BASIN GEOMETRY *****")

IF ( DEPOST.GT.1.99.AND.MASS.GT.0.01) GO TO 3640IF(MASS.GT.0.0) GO TO 3720WRITE (6,9900)

9900 FORMAT ( //// , 15X,STAGE",10X,"AREA",7X,"AVERAGE DEPTH",5X,"DISCHARGlE",7X,"CAPACITY")WRITE (6,10000)

10000 FORMAT(/#16X,"(FT)",9X,"(ACRES)",10X,"(FT)",10X,"(CFS)",9X,"(ACRES1—FT)")DO 3600 IL•1,NW R ITE (6,10100) STG 1( IL) , AREAS(IL),AVDPTH(IL),DISCHB(IL),CAPAC(IL)

10100 FORMAT(/ ,10X,F10.4, 5 X , F1 0 .2,5X,F10.2,5X,F10.2,5X,F11.4,5X)3600 CONTINUE

WRITE (6,10200)10200 FORMAT(1H1)

WRITE(6,10300)WRITE(6,10400)

10300 FORMAT(//p45X,"***** STORM HYDROGRAPHS E SEDIMENTGRAPHS to)10400 FORMAT ( ////s 8 X , "7/ME" , 8WINFLOW",7X,"DISCHARGE",6WDETENTION TIM

1 E" , 3X , "STAGE",8X,•DEPTH",8X,"SEDIMENT")WRITE(6,10500)

10500 FORMAT(/,8X,"(HRS)",8X,"(CFS)",9X,"(CFS)",11X,"(HRS)",9X,(FT)",9X1,"(FT)",11X,"( )")DO 3620 LL=2,NEJM=(DELPLG/DELTAT)*(LL-1)+1.0WRITE (6,10600 )PLGTME(LL),INFLOW(JM),OISCHAUM),DETTME(LL),STAGO(LL1),DEPTH(LL),SEDEND(LL)

10600 FORMAT ( /y 5 X , F 7 .2 , 6X,F7.2,8X,F7.2s8X,F7.2,7X,F7.2,7X,F7.2,7X,F7.2)3620 CONTINUE

GO TO 37C03640 WRITE(6,10700)

10700 FORMAT(//,6X, "STAGE", 9X,"DEPTH",6X,"DESIGN AREA",5X,"NEW AREA",15WAVERAGE DEPTH",5X,"DISCHARGE",3X,"DESIGN CAPACITY",3X,"NEW CAP2ACITY")WRITE (6,10800)

10800 FORMAT(/, 6X , "(FT)" , 11X , "(FT)",9X,"(ACRES)",8X,"(ACRES)",10X,"(FT)1",11X,"(CFS)",7X,"(ACRE—FT)", 7X,"(ACRE—FT)")DO 3660 IL=1,NWRITE(6,10900) STG1(IL),STAGE(IL),AREAS(IL),AREA(IL),AVDPTH(IL),DI1SCH(IL),CAPCO(IL),CAPACCIL)

10900 FORMAT( F10.2,5X,F10.2,5X,F10.2,5X,F10.205X,F10.2,5X,F10.2,51X,F10.2,5X,F10.2)

3660 CONTINUEWRITE (6,11000)

11000 FORMAT(1H1)WRITE (6,11100)

11100 FORMAT(//,45X," STORM HYDROGRAPHS E SEDIMENTGRAPHS .)WRITE(6111200)

11200 FORMAT ( / ,3 X , "TIME" , 8WINFLOW",7X,"DISCHARGE",5X,"DETENTION TIME",14)( r"STAGE" ,8 X , "DEPTH" , 8X,"SEDIMENT",8X,"INFLUENT",7X,"EFFLUENT")WRITE(6,11300)

11300 FORMAT ( / ,3 X , "(HRS)" ,8 X , "(CFS)",9X,"(CFS)",11X,"(HRS)",9X,"(FT)",101X,"(FT)",10)(,"( )" , 11X , "(MG/L)",9X,"(MG/L)")DO 3680 LL=2,NEJM=(DELPLG/DELTAT)*(LL-1)+1.0

135

Page 146: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

WRITE(6,11400)PLGTME(LL),INFLOW(JM),DISCHA(JM),DETTME(LL),STAGO(LL1),DEPTH(LL),SEDEND(LL),NFINT(JM),EFLNT(Li)

11400 FORMAT( F7.2,6X,F7.2,8X,F7.2,8X,F7.2,7X,F7.2,7X,F7.2,7XpF7.2,91X,F8.1,8X1F7.1)

3680 CONTINUE3700 CONTINUE

GO TO 37803720 WRITE(6,11500)11500 FORMAT(//,15X,"STAGE",10WAREA",7WAVERAGE DEPTH",5WDISCHARGE"

1,7X, "CAPACITY")WRITE (6,11600)

11600 FORMAT(/,16X,"(FT)",9X,"(ACRES)-",10X,"(FT)",10X,"(CFS)",9X,"(ACRES1—FT)")DO 3740 Itml,NWRITE(6,11700)STG1(IL),AREAS(IL),AVDPTH( IL),DISCH8(IL),CAPACCIL)

11700 FORMAT(/ $10X,F10.4, 5X,F10.5,5X,F10.2,5X,F10.2,5X,F11.5,5X)3740 CONTINUE

WRITE (6,11800)11800 FORMAT(1H1) -

WRITE (6,11900)11900 FORMAT(//,45X,"***** STORM HYDROGRAPHS I SEDIMENTGRAPHS *****")

WRITE(6,12000)12000 FORMAT(/p3X,"TIME"s8X,"INFLOW",7X , "DISCHARGE"y 5 WDETENTION TIME",

14X,"STAGE",8X,"DEPTH",8X,"SEDIMENT",8WINFLUENT" ,7 X , "EFFLUENT" )WRITE (6,12100)

12100 FORMAT(/,3X,"(HRS)",8X,"(CFS)",9X,"(CFS3" , 11X , "(HR5)" ,9 X , "(FT)" ,101X,"(FT)",10X,"( )",11X,"(MG/L)",9X,"(MG/L)")DO 3760 LL.2,NE4MA(DELPLG/DELTAT)*(LL-1)+1.0WRITE(6,12200)PLGTME(LL),INFLOW(JM) , DISCHAUM) , DETTME(LL) , STAGO ( IL1),DEPTH(LL),SEDEND(LL),NFLNT(JM) , EFLNT(LL)

12200 FORMAT( f7.2,6X,F7.2,8X,F7.2,8X,F7.2,7X,F7.2,7X,F7.2,7X,F7.2,8lx,F8.0,8X,F7.0)

3760 CONTINUE3780 CONTINUE3800 CONTINUE

12300 FORMAT(1H1)WRITE (6,12300)IF(TRP.GT.1.0) GO TO 3820GO TO 3840

3820 IF(TRP.GT.TRAP) GO TO 203840 CONTINUE

GO TO 38803860 WRITE(6,12400)

12400 FORMAT(//,15X,"*#*** ERROR **A** • THE RESERVOIR CAPACITY IS EXCEE

1DED")3880 CONTINUE3900 CONTINUE3950 STOP

ENDSUBROUTINE WASHRETURNEND

136

Page 147: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

LIST OF REFERENCES

Auernhamer, M. E., M. M. Fogel, L. H. Hekman, Jr., and J. L. Thames.1977. Stochastic prediction of sediment yields from strip minespoils of the arid southwest. Proceedings of the 1977 Meetingsof the Arizona Section - AWRA, Vol. 7, Las Vegas, Nevada,pp. 33-40.

Blumer, S. In preparation. Master's thesis. School of RenewableNatural Resources. University of Arizona, Tucson.

Bondurant, J. A., C. E. Brockway, and M. J. Brown. 1975. Some aspectsof sedimentation pond design. Proceedings, National Symposiumon Urban Hydrology and Sediment Control, Lexington, Kentucky,pp. 117-122.

Camp, T. R. 1946. Sedimentation and the design of settling tanks.Transactions of the American Society of Civil Engineers, 3:895-958.

Curtis, D. C. 1976. A deterministic urban storm water and sedimentdischarge model. Proceedings, National Symposium on UrbanHydrology, Hydraulics, and Sediment Control, Lexington, Ken-tucky, pp. 151-162.

Curtis, D. C., and R. H. McCuen. 1977. Design efficiency of stormwaterdetention basins. Proceedings ASCE, Journal of the Water Re-sources Planning and Management Division, 103(WR1): 125-140.

Curtis, W. R. 1974. Sediment yield from strip-mined watersheds ineastern Kentucky. Proceedings, Second Research and Applied Tech-nology Symposium on Mined-Land Reclamation, Louisville, Kentucky,pp. 88-100.

Federal Register, Surface coal mining and reclamation operations - Per-manent regulatory program, OSM, Dept. of the Interior, Vol. 44,No. 50, 3-13-79, Book 3, Part II, pp. 15311-15463.

Fischer, J. N. 1976. Simulation of hydrologic processes for surface-mined lands. Ph. D. dissertation, The University of Arizona,Tucson, 122 pages.

Flaxman, E. M. 1972. Predicting sediment yield in the Western UnitedStates. Proceedings ASCE, Journal of the Hydraulics Division,98(HY12): 2073-2086.

137

Page 148: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

138

Fogel, M. M., L. Duckstein, and A. Musey. 1976. Event-based formula-tion of watershed management. Proceedings, ASCE Specialty Con-ference on Environmental Impact of Irrigation and Drainage,Ottawa, Ontario, Canada, pp. 349-373.

Fogel, M. M., L. H. Hekman, and W. B. Vandivere. 1979. Sediment yieldprediction from Black Mesa coal spoils. Paper presented at the1979 Winter Meeting, American Society of Agricultural Engineers,New Orleans, La., 7 pages.

Fogel, M. M. 1980. Professor of Watershed Management, School of Renew-able Natural Resources, University of Arizona. Oral communica-tion.

Haan, C. T., and B. J. Barfield. 1978. Hydrology and Sedimentology ofSurface Mined Lands. Office of Continuing Education and Exten-sion, University of Kentucky, 286 pages.

Hamon, W. R. 1979. Research Hydraulic Engineer, USDA-SEA-ARS, Coshocton,Ohio. Oral communication.

Holten, H. N. 1961. A concept for infiltration estimates in watershedengineering. USDA-ARS, pp. 41-51.

Huggins, L. F., and E. J. Monke. 1966. The mathematical simulation ofthe hydrology of small watersheds. Technical Report 1, PurdueUniversity Water Resources Center, Lafayette, Indiana.

Kent, K. M. 1973. A method for estimating volume and rate of runoff insmall watersheds. Soil Conservation Service, U. S. Departmentof Agriculture, SCS-TP-149, 61 pages.

Kielliker, J. K., H. L. Manges, and R. I. Lipper. 1975. Modeling theperformance of feedlot-runoff-control facilities. Transactionsof the American Society of Agricultural Engineers, Vol. 18,No. 6, pp. 1118-1121.

Krishnamurthi, N., and J. L. Balzer. 1978. Design-storm for sedimenta-tion ponds. American Geophysical Union 1978 Annual Meeting,San Francisco, California, 12 pages.

Krumbein, W. E. 1968. Statistical methods in sedimentology. Sedimen-tology (Amsterdam) 10(1): 7-23.

McCarthy, R. E. 1977. Erosion and sediment control for coal surfacemine areas. National Symposium on Soil Erosion and Sedimenta-tion by Water, Chicago, Illinois.

Morris, H. M., and J. M. Wiggert. 1972. Applied hydraulics in engi-neering. The Ronald Press Company, New York, N.Y., pp. 268-270.

Page 149: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

139

Nadolski, J. 1980. Hydrologist, Office of Surface Mining, Denver,Colorado. Oral communication.

National Oceanic and Atmospheric Administration. 1973. Precipitation -frequency - atlas of the Western United States. U. S. Depart-ment of Commerce, Rockville, Maryland.

Oscanyan, P. C. 1975. Design of sediment basins for construction sites.Proceedings, National Symposium on Urban Hydrology and SedimentControl, Lexington, Kentucky, pp. 101-116.

Overcash, M. R., and R. L. Phillips. 1978. Dairy feedlot hydrology.Transactions of the American Society of Agricultural Engineers,Vol. 21, No. 6, pp. 1193-1198.

Parzan, E. 1962. Stochastic processes. Holden-Day, Inc., San Francisco,324 pages.

Portland Cement Association. 1964. Handbook of concrete culvert pipehydraulics. Portland Cement Association, Skokic, Illinois,pp. 87-88.

Reed, L. A. 1975. Controlling sediment from construction areas. Pro-ceedings, Third Symposium on Surface Mining and Reclamation,Volume II. Louisville, Kentucky, pp. 48-57.

Renard, K. G., and R. J. Simanton. 1975. Thunderstorm precipitationeffects on the rainfall-erosion index of the Universal Soil LossEquation. Proceedings of the 1975 Meeting of the Arizona Sec-tion, AWRA, Vol. 5, Tempe, Ariozna, pp. 47-56.

Renard, K. G., and L. J. Lane. 1975. Sediment yield as related to astochastic model of ephemeral runoff. In Present and Prospec-tive Technology for Predicting Sediment Yields and Sources,

Shirley, E. D., and L. J. Lane. 1978. A sediment yield equation froman erosion simulation model. Proceedings of the 1978 Meetingsof the Arizona Section-AWRA, Vol. 8, Flagstaff, Arizona, pp.90-96.

Smith, J. H., D. R. Davis, and M. Fogel. 1977. Determination of sedi-ment yield by transferring rainfall data. Water Resources Bul-letin, Vol. 13, No. 3, pp. 529-542.

U. S. Environmental Protection Agency. 1976. Erosion and sediment con-trol - Surface mining in the Eastern U. S., Vol. I and II.EPA-625/2-76-006, U. S. Environmental Protection Agency, Wash-ington, D. C.

Page 150: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

140

U. S. Environmental Protection Agency. 1976a. Effectiveness of surface

mine sedimentation ponds. EPA-600/2-76-117. U. S. Environmental

Protection Agency, Cincinnati, Ohio, 100 pages.

U. S. Environmental Protection Agency. 1977. Preliminary guidance forestimating erosion on areas disturbed by surface mining activitiesin the interior Western United States. EPA-908/4-77-005, U. S.Environmental Protection Agency, Denver, Colorado, 58 pages.

U. S. Weather Bureau. 1954-1955. Rainfall intensities for local drain-age design in the United States, Part I and II. U. S. Departmentof Commerce, Technical Paper No. 24.

U. S. Weather Bureau. 1956. Rainfall intensities for local drainagedesign in Western United States. U. S. Dept. of Commerce, Tech-nical Paper No. 28, 46 pages.

Vandivere, W. B., D. R. Davis, and M. M. Fogel. 1979. The effect ofuncertainty in sedimentation pond design. Paper presented atthe 1979 Winter Meeting of ASAE, New Orleans, La., 7 pages.

Ward, A. D., C. T. Haan, and B. J. Barfield. 1977a. Prediction of sedi-ment basin performance. ASAE Paper No. 77-2528, American Societyof Agricultural Engineers, St. Joseph, Michigan.

Ward, A. D., C. T. Haan, and B. J. Barfield. 1977b. Simulation of thesedimentology of sediment detention basins. Research ReportNo. 103, University of Kentucky Water Resources Research Insti-tute, Lexington, Kentucky, 133 pages.

Ward, A. D., C. T. Haan, and B. J. Barfield. 1978. The design of sedi-ment basins. ASAE Paper No. 78-2086, American Society of Agri-cultural Engineers, St. Joseph, Michigan, 32 pages.

Ward, A. D., C. T. Haan, and J. Tapp. 1979. The DEPOSITS sedimentationpond design manual. Institute for Mining and Minerals Research,University of Kentucky, Lexington, Kentucky.

Weber, J. E., M. M. Fogel, and L. Duckstein. 1976. The use of multipleregression models in predicting sediment yield. Water ResourcesBulletin, Vol. 12, No. 1, pp. 1-17.

Wensink, R. B., and J. R. Miner. 1975. A model to predict the perfor-mance of feedlot runoff control facilities at specific Oregonlocations. Transactions of the American Society of AgriculturalEngineers, Vol. 18, No. 6, pp. 1141-1145.

White, J. R., and W. T. Plass. 1974. Sediment control using modifiedmining and regrading methods and sediment control structures.Proceedings, Second Research and Applied Technology Symposiumon Mined-Land Reclamation, Louisville, Kentucky, pp. 117-123.

Page 151: N NT F TH PRFRN DNTTN PND b - Open Repositoryarizona.openrepository.com/arizona/bitstream/10150/191711/1/azu_td... · N NT F TH PRFRN F FDRLL RLTD ... nt th prpd f th trl n th ntrt

141

Williams, J. R. 1975. Sediment-yield prediction with universal equationusing runoff energy factor. In Present and Prospective Tech-nology for Predicting Sediment Yields and Sources, U. S. Dept.of Agriculture, ARS-S-40, 285 pages.

Williams, J. R., and H. D. Brendt. 1977. Determining the universal soilloss equation's length slope factor for watersheds. In, SoilErosion: Prediction and Control. Special Publication No. 21,Soil Conservation Society of America, Ankeny, Iowa, pp. 217-225.

Wilmoth, R. C., R. D. Hill, and C. E. Ettinger. 1979. Evaluation ofsedimentation pond performance. Paper presented at the 1979Winter Meeting, American Society of Agricultural Engineers, NewOrleans, Louisiana, 32 pages.

Wischmeier, W. H., and D. D. Smith. 1965. Predicting rainfall-erosionlosses from cropland east of the Rocky Mountains. U. S. Depart-ment of Agriculture, Agriculture Handbook No. 282, 47 pages.

Wischmeier, W. H., and D. D. Smith. 1978. Predicting rainfall erosionlosses - a guide to conservation planning. USDA-SEA, Agricul-ture Handbook No. 537.

Woolhiser, D. A., and P. Todovoric. 1971. A stochastic model of sedi-ment yield for ephemeral streams. Proceedings, Symposium onStatistical Hydrology, Tucson, Arizona, pp. 295-308.

Woolhiser, D. A., and P. H. Blinco. 1975. In Present and prospectivetechnology for predicting sediment yield and sources, U. S.Department of Agriculture, ARS-S-40, 285 pages.

Woolhiser, D. A., and K. G. Renard. 1978. Stochastic aspects of water-shed sediment yield. Proceedings of the Specialty Conferenceon Verification of Mathematical and Physical Models in HydraulicEngineering, ASCE, College Park, Maryland, pp. 561-567.


Recommended