Citation
Water quantity and quality impacts of the proposed Everglades agricultural area storage reservoir

Material Information

Title:
Water quantity and quality impacts of the proposed Everglades agricultural area storage reservoir : Phase 1
Alternate Title:
Water quantity and quality impacts of the proposed Everglades agricultural area storage reservoir: Phase One
Creator:
Reisinger, David ( Dissertant )
Heaney, James P. ( Thesis advisor )
Sansalone, John J. ( Reviewer )
Knight, Robert L. ( Reviewer )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Copyright Date:
2006
Language:
English
Physical Description:
xi, 97 p. ; ill.

Subjects

Subjects / Keywords:
Arithmetic mean ( jstor )
Canals ( jstor )
Design engineering ( jstor )
Lakes ( jstor )
Modeling ( jstor )
Natural reservoirs ( jstor )
Reaction kinetics ( jstor )
Simulations ( jstor )
Water quality ( jstor )
Water quantity ( jstor )
Dissertations, Academic -- Environmental Engineering Sciences -- UF ( local )
Miami metropolitan area ( local )

Notes

Abstract:
A model was developed to estimate water quality in the Everglades Agricultural Area Storage Reservoir (EAASR) and Stormwater Treatment Area (STA) 3/4 as part of the dynamic and challenging design process associated with the Everglades Restoration. The University of Florida Water Quality Design Tool is a steady-state mass balance model using the KC* model to simulate the total phosphorus (TP) concentrations of the outflows from the proposed EAASR and downstream STA 3/4. Performance under longterm average conditions was determined to be the most appropriate level of sophistication needed to provide key insights for the rapidly evolving EAASR designs. Water quantity boundary conditions were provided from output from the South Florida Water Management Model, a complex regional model that simulates daily flows and stages over a 36 year period. Water quantity and quality inputs to the model were calculated for the base and STA scenarios for each configuration. The hydraulic residence time (HRT) of the EAASR ranged from 39 to 107.5 days for various scenarios. The HRT of STA 3/4 was calculated to be 29 days for all scenarios. Inflow TP concentrations varied from 0.067 mg/L to 0.133 mg/L depending on the reservoir configuration. Lake Istokpoga was found to be the best comparable for parameter estimation. Data from 16 STAs were used as comparables to STA 3/4. The background TP concentration of the EAASR was determined to be 0.025 mg/L and 0.019 mg/L for the STA. The reaction rate calibration determined a rate of 0.016 per day of TP for the reservoir and 0.127 per day for the STA. Output concentrations from the EAASR shared a significant treatment effect: concentrations reduced to 0.028 to 0.068 mg/L depending on the configuration. Variability in performance was greatly dampened in the STA 3/4 outflow with the outflow concentrations all in the 0.020-0.021 mg/L range. Results showed that compartmentalization of the reservoir and including an additional outflow structure from Compartment 2 to STA 3/4 in the two compartment configuration can provide additional operational flexibility and increase water quality improvements. Results suggest that simulated inflows to STA 3/4 could be increased to improve water quality even more in the EAASR/STA 3/4 storage/treatment train. The reservoir in its current simulation is not used for flow equalization. Actual performance of the EAASR/STA system could vary widely from our predictions for several reasons: • Inflow quantities and water quality can be expected to vary over the next 50 years • The SFWMM estimates of inflows and operations are not based on any kind of optimization for the EAASR/STA system but represent an estimate of their role in a regional water management scenario. • Behavior and performance of the EAASR/STA can be expected to vary depending on how it is operated for flood control and water supply purposes as well as water quality enhancement.
Thesis:
Thesis--University of Florida
Bibliography:
Includes bibliograhical references.
General Note:
Title from title page of source document.
General Note:
Document formatted into pages; contains xi, 97 pages.
General Note:
Includes vita.
General Note:
Text (Electronic Thesis) in PDF format.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Reisinger, David. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
7/24/2006

Downloads

This item has the following downloads:


Full Text












WATER QUANTITY AND QUALITY IMPACTS OF THE PROPOSED
EVERGLADES AGRICULTURAL AREA STORAGE RESERVOIR: PHASE 1
















By

DANIEL L. REISINGER


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Daniel L. Reisinger















ACKNOWLEDGMENTS

I would like to thank my family for their support and Dr. James P. Heaney for his

invaluable guidance and contributions to this thesis. Additionally, Mr. Scott Knight

worked on an earlier version of this analysis. Dr. Joong G. Lee and Dr. Robert Knight

provided valuable review and suggestions. I would like to acknowledge the cooperation

from Mr. Cary White and Dr. Jaime Graulau-Santiago in preparing various SFWMM

scenarios under tight deadlines. I would like to thank my supervisory committee

members (Dr. John J. Sansalone and Dr. Robert L. Knight) for their valuable review of

this thesis.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ......... .................................................................................... iii

LIST OF TABLES ........ ........ ................................... ................. ............ vi

LIST OF FIGURES ............................................................. ......... viii

A B STR A C T ................................................. ..................................... .. x

CHAPTER

1 IN T R O D U C T IO N ............................................................................. .............. ...

2 BACKGROUND AND PREVIOUS WORK............................................................3

The Comprehensive Everglades Restoration Plan.................................... .................3
Everglades Agricultural Area Storage Reservoir ........................................ ..............4
Introduction and Previous W ork ........................................ ....................... 4
EAASR Planning and Design Challenges....................................................6
Engineering Design and the Systems Engineering Approach................. .............
Reservoir M modeling .................. .......................... .. ..... ....... .............. .. 11
W ater Quality M odels ............................. .... ..... .. ........ .............. ... 11
Reservoir and Lake Modeling in South Florida ...............................................12
Sum m ary and Conclusions ............................................................ ............... 13

3 DEVELOPMENT OF EAASR CONCEPTUAL DESIGN ALTERNATIVES ........15

Sizing C configurations ................................................................... .. ................ .. 15
Compartmentalization Configurations............................................... .................. 17
Single Com part ent EA A SR ................................... ........................... .. ........ 19
Tw o C om partm ent E A A SR ...................................................................... .....20
Four Com partm ent EA A SR ........................................ .......................... 22
Summary and Conclusion......................................... 24

4 SIMULATION MODEL FORMULATION AND INPUT....................................26

University of Florida Water Quality Design Tool............. ........... ............... 26
M ass B balance .......... .......... ... ...... .......................................... 26
Water Quantity Characterization............................ ......... 27









Water Quality Characterization.................. ........ ......................... 28
Inflow concentration ............................................................................. 28
R em ov al rate ............................................................2 8
M odel Rationale .................... .. ........... ... .......... ................ 29
Spreadsheet M odel ........................................ .... .. ....... ........ ......30
Water Quantity of EAASR and STA 3/4 .............. .............................................32
Water Quantity Data Source and Information.........................................32
Water Balance of the FC Base EAASR and STA 3/4.......................................33
Single Compartment Reservoir Configuration Scenarios .................................34
Two Compartment Reservoir Configuration Scenarios ....................................37
Four Compartment Configuration Scenarios...........................................41
Flow E equalization ...................... .. .. ....... .. ................. .. ......... .. ...... ... 44
Comparison of STAs and Lakes in Southern Florida.......................................45
W ater Quality for the EAASR and STA 3/4 ................................... .................48
Inflow W ater Quality.......................................................... ............... 48
B background Concentration........................................................ ............... 51
Reservoir TP Overall Reaction Rate ....................................... ............... 51
Long-term average evaluation of comparable systems ......................................52
More detailed analysis of selected comparable .......................................53
Selection of the appropriate period of record for the chosen comparable
sy ste m ................................................... ................ 5 4
R ate constant calibration ........................................ ......................... 58
EAASR reaction rate sensitivity analysis................................ .... ..... 61
STA TP R action R ate ................................................................................. 65
Predicted Performance of EAASR Removal of TP...........................................66
Predicted Performance of STA 3/4 Removal of TP........................................67
Sum m ary and C onclu sions .............................................................. .....................70

5 ESTIMATED WATER QUALITY CHANGES IN THE EAASR AND STA 3/4
SY S T E M S ............................................................................ 72

A analysis of Long-Term Averages ........................................ ......................... 72
Single Com partm ent R eservoir ........................................ ....................... 72
Tw o Com partm ent R eservoir ........................................ ......................... 74
Four Com part ent R eservoir.................. .............. ........................... .. ......... 76
Annual Variability in Performance for the SC STA Scenario............................ 77
Sum m ary and C onclu sions .............................................................. .....................84

6 SUMMARY AND CONCLUSIONS.....................................................................87

LIST OF REFEREN CES ............................................................................. 93

BIO GRAPH ICAL SK ETCH .................................................. ............................... 97







v
















LIST OF TABLES


Table p

3-1 Conceptual Configuration of Reservoir for Original Alternative 3 ......................17

4-1 Thirty Six Year Water Balance for EAASR and STA 3/4....................................34

4-2 Compartment A Water Quantity for the Four Compartment EAASR
C o n fig u ratio n ...................................................... ................ 4 2

4-3 Compartment B Water Quantity for the Four Compartment EAASR
C o n fig u ratio n ...................................................... ................ 4 2

4-4 Compartment C Water Quantity for the Four Compartment EAASR
C o n fig u ratio n ...................................................... ................ 4 2

4-5 Compartment D Water Quantity for the Four Compartment EAASR
C o n fig u ratio n ...................................................... ................ 4 2

4-6 Mean, Standard Deviation, and COV of Single Compartment Scenarios ..............45

4-7 Comparative HLRs and HRTs for 37 Reactors in Southeast Florida (Data from
W alker and K adlec 2005b)......................................................... .............. 47

4-8 NNR Basin Mean, Coefficient of Variation, and Count of Water Quality
P aram eters ........................................................................... 4 9

4-9 LOKNNR Canal Mean, Coefficient of Variation, and Count of Water Quality
P aram eters ............................................................................ 4 9

4-10 Miami Basin Mean, Coefficient of Variation, and Count of Water Quality
P aram eters ............................................................................ 50

4-11 LOKMiami Mean, Coefficient of Variation, and Count of Water Quality
P aram eters ........................................................................... 50

4-12 Decision Rankings for Comparable Systems Ranked Five and Better ..................53

4-13 Selected Results of the EAASR and STA 3/4 for Varying Reaction Rates.............64

4-14 Characteristics of 16 STAs in Southeast Florida (Data from Walker and Kadlec
2 0 0 5 a) ......................................................... .................. 6 6









4-15 Effect of Initial TP Concentration on TP at 50 Days and Percent Control for the
E A A S R ........................................................ ................ 6 7

4-16 Effect of Initial TP Concentration on TP at 15 Days and Percent Control for the
S T A s ...................................... .................................................... 7 0

5-1 Single Compartment EAASR Results for Total Phosphorus with Variable
D epths..................... ...................... ........ .........................73

5-2 STA 3/4 Results for the Single Compartment EAASR for Total Phosphorus with
V variable D epth s .............................. ......... ...... ............................. ............. 73

5-3 Compartment 1 Result for the Two Compartment EAASR Configuration ............75

5-4 Compartment 2 Result for the Two Compartment EAASR Configuration ............76

5-5 STA 3/4 Result for the Two Compartment EAASR Configuration.......................76

5-6 TP Removal by M ass in Kilograms per Day ................. .................................. 76

5-7 Compartment A Result for the Four Compartment EAASR Configuration ............76

5-8 Compartment B Result for the Four Compartment EAASR Configuration ............77

5-9 Compartment C Result for the Four Compartment EAASR Configuration ............77

5-10 Compartment D Result for the Four Compartment EAASR Configuration ............77

5-11 STA 3/4 Result for the Four Compartment EAASR Configuration ......................77

5-12 TP Removal by EAASR and STA 3/4 in Mass in Kilograms per Day ..................77

5-13 Annual EAASR Variability for the Single Compartment STA Scenario ................79

5-14 Annual STA 3/4 Variability for the Single Compartment STA Scenario ..............80

5-15 Summary Results for the One, Two, and Four Compartment Scenarios .................85
















LIST OF FIGURES


Figure page

2-1 Accelerated CERP Projects (Central and Southern Florida Project 2005, Page
32) ....... .... ........................ ................................... ......... ..... .. 4

2-2 EAASR and Stormwater Treatment Area 3/4 Vicinity Map (USACE and
SFW M D 2002, Page 7) ........................................... ................................. 6

3-1 Conceptual Configuration of Reservoir for original Alternative 3 (USACE and
SFWMD, electronic correspondence, April 5, 2005)................................... 16

3-2 General Layout of EAASR and STA 3/4...........................................................18

3-3 Single Compartment EAASR and STA 3/4 Flow Diagram................................20

3-4 Two compartment EAASR and STA 3/4 Configurations.......................................21

3-5 Two Compartment EAASR Configuration Flow Diagram.................................22

3-6 Four Compartment EAASR and STA 3/4 Configurations................... ..............23

3-7 Four Compartment EAASR and STA 3/4 Flow Diagram.............................. 24

4-1 Block Mass-Flow Diagram of Modeled EAASR ..............................................27

4-2 Spreadsheet Interface for UF W QDT.................................... ....................... 31

4-3 Single Compartment Base Scenario Flow Diagram............... ................... 35

4-4 Single Compartment STA Scenario Flow Diagram ...........................................36

4-5 Two Compartment Base Scenario Flow Diagram.................. ......................... 38

4-6 Two Compartment STA Scenario Flow Diagram..............................................39

4-7 Two Compartment Miami Scenario Flow Diagram..............................................40

4-8 Two Compartment Miami STA Scenario Flow Diagram .....................................41

4-9 Four Compartment Base Scenario Flow Diagram ............................................. 43









4-10 Four Compartment STA Scenario Flow Diagram.................................................44

4-11 Lake Istokpoga and M ass Balance Locations .................................. ............... 56

4-12 S-68 TP Data and Adjusted POR ........................................................... 57

4-13 Arbuckle Creek TP Data and Adjusted POR ................................. ............... 57

4-14 Water Quality Sampling Stations in Lake Istokpoga.............................................59

4-15 Median TP Values at Lake Istokpoga Sampling Stations..................................60

4-16 Lake Istokpoga TP Removal in Arbuckle Creek Flow Path..................................61

4-17 Sensitivity Analysis for EAASR Parameters k and C*................. .........................62

4-18 Outflow Concentration of the EAASR and STA 3/4 for Varying Reaction Rates ..63

4-19 Water Balance and Location of the EAASR Phase 1 (SFWMD 2006). ...............65

4-20 Effect of Initial Concentration and Residence Time on Outflow TP
C concentration for the E A A SR .................................................................... .. .. ....67

4-21 Effect of Initial Concentration and Residence Time on Outflow TP
C concentration for the STA s........................................................... ............... 69

4-22 Figure 4-21 Rescaled to Residence Times Up to 50 Days.................. ............69

5-1 M ean Annual Inflows from SC STA Scenario............................................. 81

5-2 M ean Annual Depths from SC STA Scenario ............................... ............... .82

5-3 M ean Annual HRT from SC STA Scenario.................................. .....................83

5-4 Mean Annual Inflow and Outflow Concentrations from SC STA Scenario............84















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Engineering

WATER QUANTITY AND QUALITY IMPACTS OF THE PROPOSED
EVERGLADES AGRICULTURAL AREA STORAGE RESERVOIR: PHASE 1

By

Daniel L. Reisinger

May 2006

Chair: James P. Heaney
Major Department: Environmental Engineering Sciences

A model was developed to estimate water quality in the Everglades Agricultural

Area Storage Reservoir (EAASR) and Stormwater Treatment Area (STA) 3/4 as part of

the dynamic and challenging design process associated with the Everglades Restoration.

The University of Florida Water Quality Design Tool is a steady-state mass balance

model using the KC* model to simulate the total phosphorus (TP) concentrations of the

outflows from the proposed EAASR and downstream STA 3/4. Performance under long-

term average conditions was determined to be the most appropriate level of sophistication

needed to provide key insights for the rapidly evolving EAASR designs. Water quantity

boundary conditions were provided from output from the South Florida Water

Management Model, a complex regional model that simulates daily flows and stages over

a 36 year period.

Water quantity and quality inputs to the model were calculated for the base and

STA scenarios for each configuration. The hydraulic residence time (HRT) of the









EAASR ranged from 39 to 107.5 days for various scenarios. The HRT of STA 3/4 was

calculated to be 29 days for all scenarios. Inflow TP concentrations varied from 0.067

mg/L to 0.133 mg/L depending on the reservoir configuration. Lake Istokpoga was

found to be the best comparable for parameter estimation. Data from 16 STAs were used

as comparable to STA 3/4. The background TP concentration of the EAASR was

determined to be 0.025 mg/L and 0.019 mg/L for the STA. The reaction rate calibration

determined a rate of 0.016 per day of TP for the reservoir and 0.127 per day for the STA.

Output concentrations from the EAASR shared a significant treatment effect:

concentrations reduced to 0.028 to 0.068 mg/L depending on the configuration.

Variability in performance was greatly dampened in the STA 3/4 outflow with the

outflow concentrations all in the 0.020-0.021 mg/L range. Results showed that

compartmentalization of the reservoir and including an additional outflow structure from

Compartment 2 to STA 3/4 in the two compartment configuration can provide additional

operational flexibility and increase water quality improvements. Results suggest that

simulated inflows to STA 3/4 could be increased to improve water quality even more in

the EAASR/STA 3/4 storage/treatment train. The reservoir in its current simulation is not

used for flow equalization. Actual performance of the EAASR/STA system could vary

widely from our predictions for several reasons:

* Inflow quantities and water quality can be expected to vary over the next 50 years

* The SFWMM estimates of inflows and operations are not based on any kind of
optimization for the EAASR/STA system but represent an estimate of their role in a
regional water management scenario.

* Behavior and performance of the EAASR/STA can be expected to vary depending
on how it is operated for flood control and water supply purposes as well as water
quality enhancement.














CHAPTER 1
INTRODUCTION

The Everglades Agricultural Area Storage Reservoir (EAASR) is jointly operated

for water supply, flood control, and to support water quality management. The multiple

and sometimes competitive purposes of the reservoir provide significant design and

operation challenges. Methods and guidelines for the design of reservoirs and levees

with respect to water supply and flood control are well established. However, methods

for incorporating water quality are less developed.

The United States Army Corps of Engineers (USACE) tasked the University of

Florida, as a subcontractor to Water and Air Research Inc. (WAR), to develop and use a

model for estimating of reservoir water quality. This model and prediction of water

quality is to serve as one section of a larger report by WAR, who was tasked to create an

Environmental Impact Statement for the EAASR.

The EAASR is one of the first and potentially the largest single Comprehensive

Everglades Restoration Plan component to be included in the Acceler8 program. The

Acceler8 program allows a dual-track process of design. The South Florida Water

Management District (SFWMD) completes detailed design in parallel with the USACE

meeting their prescribed requirements (South Florida Water Management District,

SFWMD 2006). As a result, challenges have arisen throughout the project. The

Acceler8 design evolved quickly, preempting USACE alternatives, and resulting in an

iterative design formulation process that included multiple simulation and analysis efforts

(Knight et al. 2006). Different modeling assumptions (between the project sponsors and









updates to the simulation model) used to predict the water quantity created uncertainty

when comparing current and previous results. To date, no Comprehensive Everglades

Restoration Plan reservoir has been constructed. Therefore comparable existing lake or

reservoir systems must be used to predict the performance of the proposed reservoirs.

Choosing among these comparable systems is a significant challenge because of the wide

range of operations.

Our aim was to create and use a model to estimate water quality in the EAASR

and Stormwater Treatment Area (STA) 3/4. The model needed to be able to function in

the dynamic and challenging EAASR design process. We hoped to gain key insights on

the most current EAASR and STA 3/4 configurations.














CHAPTER 2
BACKGROUND AND PREVIOUS WORK

The Comprehensive Everglades Restoration Plan

The Comprehensive Everglades Restoration Plan (CERP) will deploy 63 water

resource projects to help restore the Everglades, and provide water supply and flood

control. CERP creates a partnership of USACE (the federal sponsor) and the SFWMD

(the local sponsor) to restore, protect and preserve the South Florida ecosystem. The plan

calls for cost-sharing between the state and federal sponsors. It will take approximately

30 years to implement all of the proposed CERP projects. The projects are focusing on

getting the water right, which includes the quality, quantity, timing, and distribution of

flows. This goal will be accomplished by all 63 projects working in unison (Central and

Southern Florida Project 2005). Water quality benefits are an important part of CERP

projects; however they are not the primary purpose of such projects.

Eight CERP projects where chosen by the SFWMD, including the Everglades

Agricultural Area Storage Reservoir (EAASR), for accelerated funding, design and

construction. This streamlined plan was termed Acceler8. Acceler8 allows a dual track

process of design, where the SFWMD completes detailed design in parallel with the

USACE meeting their prescribed requirements. The accelerated schedule allows the

benefits to the system to be realized sooner and more cost-effectively (Figure 2-1)

(SFWMD 2006a). The Project Implementation Report (PIR) details the reconnaissance,

feasibility analysis, and selection of the preferred conceptual design.


























Figure 2-1. Accelerated CERP Projects (Central and Southern Florida Project 2005, Page
32)

Everglades Agricultural Area Storage Reservoir

Introduction and Previous Work

CERP uses extensive water storage systems, including surface and sub-surface

reservoirs and aquifer storage and recovery (Central and South Florida Project 2006).

These storage systems will be jointly operated for water supply, flood control, and water

quality management. The multiple and sometimes competitive purposes of the CERP

reservoirs provide significant design and operation challenges. Well-established methods

and guidelines for the design of reservoirs and levees with respect to water supply and

flood control purposes have been developed. However, methods for incorporating water

quality are less developed.

The EAASR, a CERP project, aims to create a reservoir on land in the southern

portion of the Everglades Agricultural Area (EAA) (Figure 2-2). The main goals and

objectives of the EAASR Phase 1 project as stated in the 2002 Project Management Plan


Design


Desig









(United States Army Corps of Engineers and South Florida Water Management District,

USACE and SFWMD 2002) are as follows:

* Reduction of Lake Okeechobee regulatory releases to the estuaries and reduction of
backpumping from the EAA into Lake Okeechobee by sending the water to the
south and into the reservoirs.

* Improved environmental releases through the storage of water and release to the
Everglades during the dry season demand.

* Flow equalization and optimization of treatment performance of Stormwater
Treatment Area (STA)-2, STA-3/4, STA-5, and STA-6 by capturing peak storm-
event discharges within the reservoirs for subsequent release to the STAs.

* Improved flood control and regional water supply for the agricultural community
currently served by the EAA canals and other areas served by Lake Okeechobee.

The USACE and SFWMD developed the 2002 Project Management Plan for the

EAASR. Details on the background, purpose, scope, and initial planning of the EAASR

are included in this document. The Conceptual Alternative report by USACE and

SFWMD (2004a) presents 22 alternative designs for the EAASR, using any combination

of Components A, B, and C in Figure 2-2. The area, depth, volume, water control

structures, and additional features of each alternative are detailed. The Screening of

Conceptual Alternatives report also by USACE and SFWMD (2004b) details the

screening criteria and decision matrix for the initial conceptual alternatives. The

screening results found five alternatives for further analyses. These analyses led to the

development of conceptual alternatives utilizing only Component A, which where

documented in the USACE and SFWMD (2005) Integrated Project Implementation

Report and Environmental Impact Assessment (PIR/EIS). The PIR/EIS selects a single

alternative, the tentatively selected plan (TSP), and provides detailed design information.

An earlier version of the results and discussion presented in this thesis are also included

in Appendix F of the PIR/EIS. SFWMD produced the Basis of Design Report (BODR)









(2006b) for the first phase of EAASR. The BODR provides greater design detail for

phase 1 and an alternative water quality analysis.


Figure 2-2. EAASR and Stormwater Treatment Area 3/4 Vicinity Map (USACE and
SFWMD 2002, Page 7)

EAASR Planning and Design Challenges

The planning and design of the EAASR is one of the first and potentially the

largest single CERP component to be accelerated. As a result, challenges have arisen









throughout the project. Knight et al. (2006) reviewed the EAASR planning process in

order to establish a cohesive approach to the design. They focus on the PIR/EIS where

"the parallel process leads to the USACE examining and evaluating multiple alternatives

while the SFWMD is beginning detailed design on a specific alternative" (Knight et al.

2006, Page 7). The USACE is attempting to incorporate the Acceler8 alternative, while

meeting statutory requirements to evaluate a wide range of alternatives. The matter was

complicated by USACE analyzing the entire 360,000 acre-feet of storage, while Acceler8

planned only the first phase, a 190,000 acre-foot reservoir. The resulting challenge arose

as the Acceler8 design evolved quickly, preempting USACE alternatives that were in the

processes of analysis for the PIR/EIS. The PIR/EIS alternatives were reformulated to

meet the new conditions and provide a range of viable alternatives. The result was an

iterative screening process for alternatives that included multiple simulation and analysis

efforts. The iterative nature of the process is documented in the discussion EAASR

configurations provided in Chapter 3.

Alternatives were simulated using the South Florida Management Model Version

5.4 (SFWMM). The SFWMM is a regional scale model used to simulate the hydrology

and water management of the SFWMD area from Lake Okeechobee to Florida Bay. This

7,600 square mile area was partitioned into 1,900 two mile by two mile squares (2,560

acres) called cells in which surface water, groundwater, and their interactions are

modeled. The model simulates the daily movement of water through the study area for

36 years from January 1, 1965 to December 31, 2000. The SFWMM is accepted as the

best tool to model this area, because it incorporates the complex operating rules that

govern the behavior of the system (SFWMD 2005a). Thus, a major simulation effort was









required to run 36 years of daily activity for 1,900 grid cells in order to evaluate one

project with 10 to 23 grid cells. As a regional model, changes to any component in the

model could affect the behavior of the EAASR. Throughout the planning process, the

model was continually updated, which improved the reliability of the results. USACE

and Acceler8 alternatives were simulated on different networks and using different

assumptions. For example, Acceler8 evaluated water quantity for the 2010 and 2015 land

use projections, while USACE used 2050 land use projections (SFWMD 2006 and

USACE and SFWMD 2005). These factors created uncertainty when comparing current

and previous results, as well as results between agencies. The SFWMM can only be run

by the sponsoring agencies and was not available for direct use by the study team. Thus,

each run of the SFWMM required a requisition to the sponsoring agencies.

The EAASR may be actively operated in a wide range of hydraulic and

hydrologic conditions, which affects the water quality of the reservoir. Empirical

parameter estimates are required as part of the modeling exercise. No CERP reservoir

has been constructed; therefore comparable existing lake and/or reservoir systems must

be used to predict the likely performance of the proposed reservoirs. The choice of these

comparable systems poses a significant challenge, due to the wide range of attributes of

these lakes and reservoirs.

Engineering Design and the Systems Engineering Approach

Planning and design approaches exist to help meet the described challenges. This

section provides an introduction to a systems engineering approach to engineering design.

According to Hazelrigg (1996), engineering design is a decision-making process. He

defines a decision as an irrevocable allocation of resources; hence the selection of design

parameters by an engineer is a decision-making process. This outlook is a departure from









traditional engineering design that is largely viewed as an exercise in problem solving. It

creates the distinction that engineering is aimed at creating information, which is related

to a specific decision, instead of knowledge, which is a set of agreed upon facts.

Considering design in this manner is commonly called systems engineering (Hazelrigg

1996).

To develop a systems engineering approach independent of domain, Braha and

Maimon (1998) performed an extensive literature review that found engineering design to

share the following common properties:

* Design begins with an acknowledgement of an unmet need and a call for action to
meet this need.

* Designing an artifact is used to transition from concepts and ideas to concrete
descriptions.

* The designer is constantly faced with the problem of bounded rationality, i.e., the
designer has limitations on his cognitive and information processing capabilities.

* The design specifications tend to evolve as part of the design process.

* Traditional engineering design methods tend to rely on satisfying rather than
finding the true optimal solution.

* Alternatives and design solutions evolve as part of the design process.

Braha and Maimon (1998) view design as a sequential process with feedback.

This process goes from general concept to preliminary and detailed design, production

planning, production, operation, and final disposal. Hazelrigg (1996, Page 8) viewed the

design process as

three distinct activities: the identification of options, development of
expectations on outcomes for each option, and use of values to select the option
that has the range of outcomes and associated probabilities that are most desired.

A key concept of Hazelrigg's view is the need to produce information that

provides a prediction of the accuracy or reliability of the design. Uncertainty is therefore









intrinsic in information. Such uncertainty is commonly accounted for by conservative

parameter estimations, factors of safety, or statistical design.

The general formulation of the engineering design problems and therefore models

consists of two main parts; the objection function and constraints (Heaney, unpublished

manuscript, 2006). Heaney (unpublished manuscript, 2006) defines the parts, where

decision variables are one-time parameter decisions and/or operating rules, as:

* Objective function: Maximize or minimize some stated objectives) by selecting
the best values of the decision variables.

* Constraints: Physical, chemical, and/or biological process relationships and/or
operational and regulatory constraints on the variables.

Traditionally design relies on constraining the system to separate the design into

manageable and domain specific parts. Traditional design leads to a reductionism

approach, where the outcome of the design is a function of the constraints. The systems

engineering approach divides the design into disciplinary models, which are incorporated

into a whole system model. This leads to a design that is less focused on constraints and

may produce more optimal designs (Hazelrigg 1996).

Lee et al. (2005) document an approach to optimize the design urban stormwater

storage-release systems. Urban approaches may also be applied to reservoir modeling as

they are fundamentally both storage-release systems. Lee et al. approach uses cost as the

objective for a design based on continuous simulation of water quantity and quality.

Spreadsheets are utilized to link powerful optimization tools to transparent process

models. Unlike traditional approaches, design may be optimized for 3 or more

parameters.









Reservoir Modeling

Mathematical models are relied upon heavily for engineering design. They

provide a prediction of the systems behavior for a given set of design or decision

variables. These models strive to answer what if questions of the designer and if

optimized can answer the question of what is best. Models are used to answer three

fundamentally different types of questions (Hazelrigg 1996):

* Will the system work as designed?
* Which of the system alternatives are better?
* Do I properly understand the system?

To develop a model for the EAASR that focuses on the second question, which of

the system alternatives is better, a review of water quality models was performed. The

review included reservoir and lake models that have been developed and calibrated for

the South Florida region. Results of the review are documented next.

Water Quality Models

The relatively shallow, actively controlled EAASR is more comparable to a lake

or shallow reservoir system than a traditional reservoir. The general framework for both

water quantity and quality models for lake or shallow reservoirs are well developed

(Chapra 1996). Chapra and Auer (1999) reviewed management models to evaluate

phosphorus loads in lakes. They classified phosphorus models in three general

categories; empirical models, simple budget models, and nutrient food-web models.

Empirical models can be divided into phosphorus loading plots and trophic

parameter correlations. Phosphorus loading plots are used to estimate the trophic level of

the lake, or when linked to a simple balance model can predict in-lake total phosphorus

concentrations. Trophic parameter correlations normally relate two trophic parameters or

can be used in tandem with phosphorus loading plots. The main advantage of empirical









models is the ease of use. These models are most accurate when calibrated for an

individual lake.

Simple budget models focus on the mass-balance of a lake system. The most

basic of these models uses inflow, outflow and one-way removal to characterize the

system. These models provide a temporal response from the lake and can be easily

adapted for different lake dynamics. Simple budget models are highly sensitive to the

quality of input data. Therefore, they are best suited for long term trends or high quality

data sets.

Nutrient and food web models are more complex models that aim to characterize

the temporal and physical aspects of matter throughout a lake or reservoir (Chapra and

Auer 1999, Page ?). These models require large amounts of data or assumptions and can

provide a more detailed analysis than previous models.

Reservoir and Lake Modeling in South Florida

Empirical and simple balance models have been used in modeling CERP

reservoirs and the EAASR. USACE and SFWMD (2003) reviewed 16 water quality

models that predicted the uptake of phosphorus in lakes and reservoirs. DMSTA was at

the top of the review's shortlist of models that met the reviewed criteria. DMSTA 2

(Walker and Havens 2005a), an improved version of DMSTA, is a mass balance model

that is calibrated for CERP reservoirs. The model incorporates a first order kinetic model

with a background concentration or a water column and sediment transfer model.

Wetland Solutions Inc. (WSI) developed water quality models to simulate 15

parameters of interest for a general CERP reservoir (WSI 2004). Each model was

calibrated for Florida lakes and reservoirs and can be applied in a spreadsheet. The

models include Eutromod (Reckhow 1979; Reckhow et al. 1992), the Vollenweider









Eutrophication Model (Vollenweider 1969, Kane 1999), and the U.S. Corps of Engineers

Bathtub Model (Walker 2004). Additionally, WSI developed three regression models

from Bums and McDonnell (2004b) based on comparable lakes and reservoirs. An

earlier version of the University of Florida Water Quality Design Tool, which assesses

TP uptake in the EAASR, is presented in the PIR/EIS (USACE and SFWMD 2005).

As stated previously, comparable lakes and reservoirs are necessary to estimate

modeling parameters. Central and South Florida lakes and reservoirs have been

relatively well studied. The restoration of the Everglades has driven recent area wide

studies and model calibration efforts, such as Burns and McDonnell (2004a) and Walker

and Kadlec (2006). Burns and McDonnell (2004a) completed a four-part project to

create and analyze a database from lakes and reservoirs that were comparable to CERP

reservoirs. They identified and acquired data for 36 potential comparable lake systems

across Florida, which were refined to eight comparable lakes and a reservoir with

sufficient data (Burns &McDonnell 2004). Walker and Kadlec (2006) developed an

extensive database of lakes for their calibration effort for DMSTA 2. The DMSTA 2

calibration uses 19 lakes and reservoirs from Walker and Havens (2003), Wetland

Solutions, Inc. (WSI) (2003), Walker (2000), and Bums and McDonnell (2004a) efforts.

Summary and Conclusions

CERP is an ambitious project to "get the water right" in the South Florida

Ecosystem and restore the Everglades. The project goal will be accomplished by all 63

projects operating together. The Acceler8 project streamlines the CERP planning,

design, and construction process to provide benefits to the system more quickly and cost-

efficiently. The EAASR, a CERP and Acceler8 project, will provide storage for water

supply, flood control, and flow equalization for water quality treatment areas. Water









supply and flood control design approaches are well developed, while water quality

approaches are less developed. The streamlining of the planning and design of the

EAASR has posed several challenges resulting in an iterative design process with

multiple conceptual alternative formulations and analyses. Due to differences in the

simulation, uncertainty exists when these multiple formulations and analyses are

compared.

The systems engineering approach to design provides a proven approach that can

meet the challenges of the EAASR planning and design process. The approach

incorporates disciplinary models that may produce a more optimal design than traditional

approaches. A model in the water quality discipline was therefore sought to assess which

of several alternatives are better. Simple empirical and mass balance models were found

that can be used for the EAASR. Comparable lake and reservoir systems were used to

provide parameter estimates for EAASR. Due to the wide range of possible operational

conditions and planning and design challenges, a water quality model developed

specifically for the EAASR will be necessary. The model must incorporate measures to

meet the planning and design challenges of the EAASR. The iterative multiple

conceptual design alternatives for the EAASR were reviewed. The model was then

formulated in the context of this chapter and the conceptual design alternatives.














CHAPTER 3
DEVELOPMENT OF EAASR CONCEPTUAL DESIGN ALTERNATIVES

The hydrology of EAASR and STA 3/4 were simulated using the SFWMM

Version 5.4. The complete water quantity dataset used in water quality modeling was

developed from the simulations. Therefore, the EAASR conceptual design alternatives

are presented in the contextual framework of the SFWMM. The physical layout of the

reservoir is termed a configuration. The combination of a configuration and particular

flow dataset is referred to as a scenario. All configurations simulated the levee walls as

vertical and of insignificant area to affect the stage-area-volume relationship. The

reservoir levees were also assumed to allow no seepage, though groundwater flow was

included.

The selection of the configuration of the EAASR was an iterative process. The

sizing of the EAASR, characterized by the area and depth, was first determined. For the

chosen size, configurations to evaluate the compartmentalization of the EAASR were

then developed. These configurations are used in the water quality analysis of this report

and are described in full detail.

Sizing Configurations

Early configurations used a combination of Components A, B, and C of Figure 2-

2 (USACE and SFWMD 2004a). A decision was made to reserve Components B and C

to expand the capacity of adjacent Stormwater Treatment Areas. Alternatives were then

developed using only the Component A location.







16


Three alternative reservoir sizes with storage capacities of 240,000, 360,000, and

480,000 acre-feet were evaluated in the early stages of Component A analysis, referred to

as the original Alternative 1, the original Alternative 3, and the original Alternative 5,

respectively (USACE and SFWMD, electronic correspondence, February 14, 2005,

March 1, 2005, March 8, 2005, and April 5, 2005). The original alternatives were

divided into two compartments of the same nominal depth. In the original alternatives,

the first compartment, called C1, was configured to provide 90,000 acre-feet of storage

for agricultural use. The second compartment, C2, varied in volume from 150,000 acre-

feet to 390,000 acre-feet and was for environmental use. Cl was able to overflow to C2

(USACE and SFWMD, electronic correspondence, April 5, 2005). A conceptual view of

the two compartment reservoir for the original Alternative 3 is provided in Figure 3-1.

Subsequently, a decision was made that the total capacity of EAASR would be 360,000

acre-feet.



MC NNRC
EARMA1 EARNH1
EARMA2 90,000 EARNH2
ac-ft 2 z

EARIN1 EARIN2





270,000
ac-ft



WCS4S +
EVBLSN


Figure 3-1. Conceptual Configuration of Reservoir for original Alternative 3 (USACE
and SFWMD, electronic correspondence, April 5, 2005)









Design depth was the next key sizing decision. Four design depths were

considered with nominal design depths of 6, 10, 12, and 14 feet. These alternatives were

named AltlR, Alt2R, Alt3R, and Alt4R, respectively (USACE and SFWMD, electronic

correspondence, April 8, 2005). The compartments in all Alt_R configurations were

termed Cl and C2, as in previous configurations. The depth and area used to achieve this

total volume were varied in each alternative; however C2's area remained fixed at 17,920

acres. The associated areas of Cl were developed to yield a capacity of 360,000 acre-

feet. The area, depth, and volume of the Alt_R configurations and the original Alt 3

simulation are presented in Table 3-1.

Table 3-1. Conceptual Configuration of Reservoir for Original Alternative 3
C1 C2 Total
Area Nominal Volume Area Nominal Volume Area Volume
Configuration (ac) Depth (ft) (ac-ft) (ac) Depth (ft) (ac-ft) (ac) (ac-ft)
Alt 3 Orig. 7,500 12 90,000 22,500 12 270,000 30,000 360,000
AltlR 40,960 6 250,000 17,920 6 110,000 58,880 360,000
Alt2R 17,920 10 180,000 17,920 10 180,000 35,840 360,000
Alt3R 12,800 12 150,000 17,920 12 210,000 30,720 360,000
Alt4R 7,680 14 110,000 17,920 14 250,000 25,600 360,000

A design depth of 12 feet was selected for the 360,000 acre-foot reservoir.

Configurations to assess the effect of compartmentalization and the source of inflows

were then developed. The configurations are presented in detail next.

Compartmentalization Configurations

Unlike previous configurations, the compartmentalization of the reservoir affects

the inflow sources to STA 3/4. Therefore, details on both the EAASR and STA 3/4 are

included for each configuration. The EAASR was modeled as 12 SFWMM cells

incorporating 30,720 acres of area and STA 3/4 was modeled as 7 SFWMM cells

incorporating 17,920 acres of area (USACE and SFWMD, electronic correspondence,









February 2 and 3, 2006). The combined total area of the EAASR and STA 3/4 is 48,640

acres and the EAASR accounts for about 63% of this total area. Figure 3-2 displays the

location of the EAASR and STA 3/4 in the SFWMM. The nominal depth of the EAASR

was 12 feet for all configurations. The EAASR was assumed to have a level pool and flat

bottom at an average elevation and therefore had a simple stage-area relationship. STA

3/4 was modeled with the same levee and pool assumptions as the reservoir.


16 17 18 19 20 21 22 23 24

Figure 3-2. General Layout of EAASR and STA 3/4

Three following configurations of the EAASR were developed from the general

layout presented in Figure 3-2:

1. Single Compartment
2. Two compartments
3. Four Compartments









A single SFWMM flow dataset (USACE and SFWMD, electronic correspondence,

February 2 and 3, 2006) was used in modeling all three configurations. However, the

quantity and source of flows to each compartment and STA 3/4 varied depending on the

configuration. Each configuration will be described in detail in the remainder of this

section.

Single Compartment EAASR

The single compartment EAASR configuration was modeled as presented in

Figure 3-2. The single compartment used all of the 30,720 acres available. All possible

inflows sources to the EAASR and STA 3/4 are represented in the configuration.

A flow diagram for the single compartment reservoir and STA 3/4 is presented in

Figure 3-3. The EAASR and STA 3/4 receive inflows via the North New River (NNR)

Canal and Miami Canal, which are shown on the east and west side of the EAASR in

Figure 3-3. The EAASR receives inflows from four external sources: NNR basin runoff,

Miami basin runoff, Lake Okeechobee (LOK) regulatory release through the NNR Canal,

and LOK regulatory releases through the Miami Canal. Water is released to the

agricultural basins or south to STA 3/4. STA 3/4 receives flow from NNR and Miami

basin runoff, LOK through the NNR and Miami Canals, and the EAASR. The STA

discharges to the water conservation Areas (WCA). Figure 3-3 and those like it represent

the aggregation by source of SFWMM flow tags. All flows are actively operated in the

SFWMM and are constrained by flow capacity and/or crest elevation.









Miami Canal NNR Canal


Miami Basin Runoff NNR Basin Runoff

Miami Basin Demands R NNR Basin Demands

LOK Miami Supply LOK NNR Supply


STA 314 Demands

Miami Basin Runoff NNR Basin Runoff


STA 3/4
LOK Miami Supply LOK NNR Supply


I STA 3/4 Outflow





Figure 3-3. Single Compartment EAASR and STA 3/4 Flow Diagram

Two Compartment EAASR

In the second configuration, the EAASR is partitioned into two compartments

(Figure 3-4). This configuration was considered to be comparable to the Tentatively

Selected Plan (TSP) in the EAASR PIR/EIS (USACE and SFWMD 2005). The eastern

compartment bordering the NNR Canal is referred to as Compartment 1 and the western

compartment bordering the Miami Canal is referred to as Compartment 2. Each

compartment was 15,360 acres and modeled as an independent level pool.









48 Miami Canal North New
River Canal

47 1 2 2

46 2 2 2 2


a -


STA 3/4 STA 3/4 STA 4


STA 3/4
UsI


42 ST 5/4 STA 3/4 STA 3/4
16 17 18 19 20 21 22 23 24
Figure 3-4. Two compartment EAASR and STA 3/4 Configurations
A flow diagram for the two compartment EAASR configuration and STA 3/4 is

presented in Figure 3-5. The compartmentalization of the EAASR altered the ability of

each compartment and the STA 3/4 to receive water from the sources presented in the

single compartment configuration. Each compartment received external flow from the

respective adjacent canal and through inter-compartmental transfers. It is important to

note that unlike early configurations water can be transferred internally between both

compartments. Water was released to the agricultural basins or south to STA 3/4 in the

reservoir. Due to the location of the compartments, only Compartment 1 was able to

release water to STA 3/4. STA 3/4 was not able to receive flow directly from the Miami

Canal; therefore the Miami Canal flows were routed through the EAASR to the STA.

The STA was able to receive inflow via NNR basin runoff, LOK through the NNR canal,

and the EAASR. STA 3/4's discharge capabilities were not altered.


B








Miami Canal NNR Canal


Miami Basin Runoff INNR Basin Runoff
Cell 1 Transfer
Miami Basin Demands NNR Basin Demands
Cell 2 Transfer
LOK Miami Supply LOK NNR Supply
I

ISTA 3/4 Demands

NNR Basin Runoff

STA 3/4
LOK NNR Supply

4 STA 3/4 Outflow

Figure 3-5. Two Compartment EAASR Configuration Flow Diagram

Four Compartment EAASR

The third configuration partitioned the EAASR into four compartments (Figure

3-6). This configuration was comparable to the Mixed/Segregated Plan (MSP) referred

to in the January Work Tasks (USACE, electronic correspondence, January 26, 2006).

The four compartments were labeled Compartments A through D. Compartment A and C

are 5,120 acres and Compartments B and D are 10,240 acres. Each compartment is

modeled as an independent level pool.

The four compartment flow diagram for the EAASR and STA 3/4 is presented in

Figure 3-7. Each of the four compartments received flow from the adjacent canal and

through inter-compartmental transfers. It is important to note that unlike early

configurations water can be transferred internally between multiple compartments. The










modeling of the four compartment configuration assumed that each compartment

received flow from a single external source. Water was released to the agricultural basins

through Compartments A or C. Due to the location of the compartments, only

Compartments B and D were able to release water to STA 3/4. STA 3/4 was able to

receive flow from both Miami and NNR Canals in this configuration. STA 3/4's

discharge capabilities were not altered.



48 ST.Aarni Canal SNorth New
River Canal


47 C C


46


45


44


43 STA 3/4 STA 3/4 STA 314 STA


42 ST 3/4 STA 34 STA 3/4


Figure 3-6. Four Compartment EAASR and STA 3/4 Configurations











Miami Canal

Miami Basin Runoff
C Transfer
Miami Basin Demands


C Transfer Transfer


LOK Miami Supply


Miami Basin Runoff


STA 314 Demands


Miami Basin Runoff


LOK Miami Supply


B Tr


STA 3/


NNR Canal

NNR Basin Runoff

NNR Basin Demands


A Transfer B Transfer


LOK NNR Supply
Snsfer

NNR Basin Runoff

SSTA 3/4 Demands




NNR Basin Runoff
4

LOK NNR Supply


j ISTA 314 Outflow


Figure 3-7. Four Compartment EAASR and STA 3/4 Flow Diagram

Summary and Conclusion

Several iterations of EAASR configurations have been developed. Early


configurations used a combination of Components A, B, and C of Figure 2-2.


Configurations were then formulated using only Component A. The sizing of the


reservoir was evaluated first. Three configurations were developed to evaluate the


volume of the reservoir. Four additional alternatives were developed to evaluate the


depth of the reservoir. A 12 foot deep, 360,000 acre-foot reservoir was subsequently


selected. Three configurations were then developed to assess the compartmentalization


of the reservoir. The configurations were developed for one, two, and four


compartments. The location of the compartments affected the source of water received









for both the EAASR and STA 3/4. Unlike early configurations, internal transfers

between the compartments occurred from multiple compartments.

The review of EAASR configurations was performed to provide the context

within which the water quality model is developed. To use these configurations, a water

quality model should include depth, area, and compartmentalization with varying area.

Due to the inter-reservoir transfers, the model must be able to simulate reservoirs in

series, with feedback loops. Additionally, the ability to include multiple inflows and

outflows would be useful. The formulation and rationale of the water quality model is

described in detail in the next section. The water quantity and quality values for the

compartmentalization configurations, with an emphasis of modeling parameters, are

detailed as well.














CHAPTER 4
SIMULATION MODEL FORMULATION AND INPUT

University of Florida Water Quality Design Tool

The EAASR is classified as a reservoir. Comparable systems in the area are

called by a variety of names including:

* Lakes
* Reservoirs
* Emergent wetlands
* Pre-existing wetlands
* Submerged aquatic vegetation systems

Nominally, reservoirs and lakes provide storage while wetlands and STAs provide

treatment. However, all of these systems can be viewed more generically as reactors that

in fact provide a blend of these functions depending on how they are designed and

operated. Accordingly, the word reactor will be used to describe the general modeling

approach for these systems. The EAASR/STA 3/4 system is viewed as a treatment train

with two reactors in series. EAASR and STA 3/4 simulations therefore incorporate both

water quantity (storage) and quality. The water quality modeling of the EAASR focuses

on Total Phosphorus (TP), the main water quality parameter of concern.

Mass Balance

A mass balance around the reactor is created to assure conservation of mass. For

the reactor, the external sources of water and pollutants are the inflow and precipitation.

The pollutant removal in the reactor is considered a final mass sink. The remaining

pollutants exit the reactor in the outflow.









In steady-state mass balance analysis, the reservoir can be modeled as having

inflows from the canals, Q,,, and outflow, Qout, which are aggregates of the daily inflows,

and outflows, and precipitation, P. A parameter of concern is assumed to enter at a

constant concentration for each inflow, C,,, and from precipitation, Cp. The pollutant is

removed as a function of detention time, td, initial concentration, C,,, and reaction rate,

kv. The pollutant exits the reactor to the surrounding system in the outflow at a

calculated concentration, Cout. The conceptual view of the mass balance (Figure 4-1) and

mathematical equation (Eqn. 4-1) also uses the concentration in each plug of water in the

reactor (C), the depth (D), and the area (A).


P
CP


10~ Reactor D
Qm A Qout
Cn Cout




Figure 4-1. Block Mass-Flow Diagram of Modeled EAASR

Q,,, C,, =Vxk, x C +Q,,, xC,, P x Cp (4-1)

Water Quantity Characterization

The water quantity characterization of the model is based on a volume balance of

the system. The volume balance of the system is represented by the total inflows minus

the total outflows equaling the change in storage (AS/At). From the reservoir alternatives

data provided by the IMC, the total inflows and outflows of the system are represented as

rainfall (P), inflows (Q,,), groundwater inflows and outflow (GWI and GWo), ET, and

outflows (Qout) (Eqn. 4-2). The change in storage per unit time, AS/At, is documented in

the SFWMM by the stage or depth of the reservoir.









ASAt = P + Q, + GW, -ET-Q,,, -GWo (4-2)

The volume balance components provide the inputs to calculate the hydraulic

residence time (HRT), which represents the water quantity of the system. The HRT is a

function of volume and flow (HRT = V/Q). For a given reservoir volume, V, HRT can be

increased by reducing the inflow rate, Q,,. The other performance measure for the

reactor is the mean operating depth, H. The above information is obtained from the

output file for the SFWMM by aggregating the cell by cell data.

Water Quality Characterization

Water quality characterization of the system is based on the inflow concentration,

Cin, the rate parameter, and the minimum concentrations. As storage and water quality

changes are inseparable, the characterization of the water quantity affects the

characterization of the water quality.

Inflow concentration

The inflow concentration, Cin, is a function of the source and quantity of inflows.

The water quantity of the system uses an aggregate inflow, Qin. Therefore, a single

representative concentration is assigned to the aggregate inflow. If more than one inflow

source exists, a flow weighted concentration is calculated from Equation 4-3 where Qi is

the flow, Ci is the concentration, and "i" designates the source.


ZQ'c,
Cn n= (4-3)
ZQ,
1=1

Removal rate

The KC* removal rate equation is used to represent pollutant concentration

profiles with time in a variety of reactor systems including wetlands (Kadlec and Knight









1996) and a wide variety of urban stormwater BMPs (CRC for Catchment Hydrology

2005). The KC* model uses a removal rate model with a background concentration (C*).

Removal is a function of the initial concentration (C,,), C* the overall rate constant, kv,

and the hydraulic residence time (HRT), td. The KC* equation is shown in Equation 4-4.

The pollutant is removed more rapidly at first and then at a decreasing rate as HRT

increases. Two-parameter calibration should be used for this model, where kv and C* are

calibrated simultaneously.

Cot = C* + (C,, C)e k_-td (4-4)

Model Rationale

The University of Florida Water Quality Design Tool (UF WQDT) was intended

from the outset to be a simple spreadsheet model used to predict the behavior of TP in the

EAASR. The storage-release framework described in Lee et al. (2005) was to be used to

provide an optimal design, which included water quality considerations. It was decided

that a basic mass balance and uptake rate model would provide a good representation of

TP behavior. Initially modeling was expected to occur at three levels; steady-state

simulation of the long-term average, frequency analysis, and daily time series analysis.

However, as the design and planning challenges became evident the scope of the model

was altered to best suit the needs of the project and the tight time schedule.

Steady-state simulation of long-term averages was deemed the most appropriate

modeling level for the dynamic conditions associated with the design of the EAASR. A

steady-state analysis can provide important insights and comparisons to design scenarios.

The simple mass-balance model is easily adapted for new configurations and scenarios,

while the steady-state nature allows for very fast run times. SFWMM updates and small









model parameter alterations generally will not create significant changes in the long-term

averages, where significant differences in the daily time series may be created. However,

large changes to the configuration or operation of the EAASR, STA 3/4, or South Florida

system will be reflected. Additionally, calibration efforts for the model could be

performed at an appropriate level of significance, which is defined by the quality of the

data. A description of the resulting model is provided in the next section.

Spreadsheet Model

The UF WQDT spreadsheet tool uses a flow and concentration calculator, a core

reactor module, and a Solver objective and constraint section if a feedback between

reactors exists. The basic UF WQDT for the SC base Scenario (Figure 4-2) does not

include feedback and all values are in US units, except TP concentrations. Values

highlighted in light blue are user entered, in white are calculated, and in orange are the

results. The flow and concentration calculator, appearing first, aggregates multiple

inflows generate a single representative flow and concentration. The core reactor module

is broken into three sections: parameters, calculations, and results. The necessary

parameters are entered into the parameter section. These values are used to calculate the

volume, HRT, and hydraulic loading rate (HLR). The HLR is an alternate representation

of the water quantity, which is nominally defined as the inflow, Qin, divided by the area,

A. Appropriate given and calculated values are used to calculate the outflow

concentration, Cout, using Equation 4-4. Percent removal of TP from the inflow

concentration is calculated as well. The flow and concentration calculator and core

reactor module are repeated for each reactor in the system (i.e. the EAASR compartments

or STA 3/4) (Figure 4-2).














University of Florida Water Quality Design Tool

Cell I Influent Flow source Variable Flow (ac-ftid) C, (mgiL)
Flow & Concentration NNR Basin Q NrMni 485 0.150
LOKNNR QLOKnCR 442 0.077
MAaim Basin QPai 8K a 470 0112
LOK Miami QLOKMni 290 0.060
Total Q 1,686 0.105

Inputs Parameter Variable Value Unit
Inflow Q, 1686 ac-f/d
Depth D 5.90 f
Area A 30,720 acres
Inflow Concentration G 0 105 mg/L
Background Concentration C 0.025 mg/L
Removal Rate kv 002 day-1
Calculations Volume 181,306 ac-ft
Hydraulic Retention Time HRT, td 107.5 days
Hydraulic Loading Rate LR, 0.05 ft/da
Output Reservoir Co.-i mg/L
Reservoir C Removal


STA 3/4 Influent Flow source Variable Flow (ac-ftid) C,, (mglL)
Flow & Concentration NNR Basin QhxMUB. 221 0150
LOKNNR QLOKNs R 219 0.077
MaimiBasin QneOJ Basi 263 0.112
LOK Miami QLosK am 317 0060
EAASR QO, 1,182 0040
Total Qin 2,201 0066

Inputs Parameter Variable Value Unit
Inflow Qn 2201 ac-f/d
Depth D 2.35 ft
Area A 17,920 acres
Inflow Concentration Cm 0.066 mg/L
Background Concentration G 0.019 mg/L
Removal Rate k 0.13 day-1
Calculations Volume V 42,043 ac-ft
Hydraulic Retention Time HRT, t 28.9 days
Hydraulic Loading Rate HLR 0.12fl/da
Output Reservoir C,,i mg/L
Reservoir C i Removal


Figure 4-2. Spreadsheet Interface for UF WQDT


The two and four compartment EAASR includes feedback between


compartments. In this configuration, the inflow concentration of one compartment is


dependent on the inflow and performance of other compartments. The inflow


concentration can be solved algebraically using a system of equations. However, this


would require programming specific equations for each compartment. The Solver tool in


Excel was used to iteratively solve for each concentration without altering the core


reactor module.









For a design with feedback flow sources, the Solver objective and constraint

section is used. Initial estimates of Cin are entered for each feedback. In the Solver

section the difference of the feedback Cin cell and the appropriate compartment Cout cell

are entered. The sum of the square differences are calculated for all feedback flows in

the Total Difference cell and used as the objective in Solver. Each feedback

concentration is constrained to the minimum achievable concentration, C*. Solver is then

used to reduce the value of the objective function to zero, which calculates the correct

concentrations. This methodology assumes full treatment of the feedback flows.

Water Quantity of EAASR and STA 3/4

The purpose of this section is to present the water quantity parameters necessary

for water quality modeling. The reported inflows and outflows to the EAASR and STA

3/4 are based on output from the SFWMM. The water quality performance of the

EAASR and STA 3/4 can be expected to vary depending on how the inflows are divided

among compartments, how the individual compartments are operated as judged by the

mean depth of storage in each compartment, and how water is transferred among

compartments before discharge from the EAASR. Performance will also depend on

whether inflows are routed to STA 3/4 directly or through the EAASR. This section

provides background information and results for the compartmentalization configurations

in Chapter 3.

Water Quantity Data Source and Information

A wide variety of options exist for directing water into the EAASR. The best

projection of the expected inflows can be obtained by using the South Florida Water

Management Model (SFWMM). The Interagency Modeling Center (IMC) modeled the

EAASR and the downstream STA 3/4 using the SFWMM Version 5.4. The IMC









provided data from their EAASR simulation to the University of Florida. All simulations

were made with the assumption of 2050 land use projections and the reservoir as the

"Next Added Increment" of the CERP project. The provided data included inflow and

outflow for each control structure, the maximum capacity for the water control structures,

rainfall, evapotranspiration (ET) and groundwater levels and the general modeling

configuration and assumptions (USACE and SFWMD, electronic correspondence,

February 2 and 3, 2006). Data were tabulated in 13,149 daily time steps for the 36 year

period, starting on January 1, 1965 and ending on December 31, 2000. The data were

provided for each of the EAASR's four square mile grid cells in the SFWMM for each

time step. The total 2x2 grid is comprised of 1,900 2x2 grid cells. Thus, a major

simulation effort was required to run 36 years of daily activity for 1,900 grid cells in

order to evaluate one project with 19 grid cells (SFWMD 2005a). The IMC output

provides the necessary information on the quantity and timing of the flows.

Water Balance of the FC Base EAASR and STA 3/4

A water balance (WB) was calculated from the SFWMM to ensure proper

modeling results for the EAASR and STA 3/4. The output data from SFWMM provided

a complete volume balance for the four compartment configuration, including stage data

for each compartment and STA 3/4. Equation 4-2 was modified to include an imbalance

term, which is attributed to ungaged flow. The imbalance term was attributed to either

ungaged inflow or ungaged outflow by its sign and included in the WB (Eqn. 4-5) where

Qinu was ungaged inflow and Qoutu was ungaged outflow.

AS / At = ,, + Q,, +P P+ GW Q,,, Q,,, -ET GW (4-5)









The long-term WB was calculated for each compartment and STA 3/4 separately.

The average daily value of each WB component for the 36 year simulation period was

used to represent the magnitude of each component. The averages are presented in units

of ac-ft/day to allow comparison between the compartments and STA 3/4. The change in

storage was calculated from the difference in volume between the first and last day of the

dataset. The long term water balances for each compartment are presented in Table 4-1.

Several key insights may be gained from inspection of the 36 year water balance.

The differences between natural WB components, rainfall, ET, and GW offset each other

for the EAASR and STA 3/4. GW and ungaged flows can be assumed to be negligible on

a long term average. The resulting simplified long term WB was used in the remainder

of the report.

Table 4-1. Thirty Six Year Water Balance for EAASR and STA 3/4
WB Term C1 C2 C3 C4 STA 3/4
Inflow (ac-ft/day) 485 470 442 290 2,201
Transfer (ac-ft/day) 105 318 83 280 NA
Ungaged (ac-ft/day) 0 0 0 2.1 0
Rainfall (ac-ft/day) 58 120 58 118 208
GWI (ac-ft/day) 0.3 0.2 0.0 0.1 0.2
Total Inflow (ac-ft/day) 648 908 583 690 2,409
Outflow (ac-ft/day) 261 695 218 488 2,177
Transfer (ac-ft/day) 318 88 296 83 NA
Ungaged (ac-ft/day) 4.0 3.6 2.5 0 13.4
ET (ac-ft/day) 64 119 64 118 216
GWo (ac-ft/day) 0.3 1.8 0.5 1.0 0.2
Total Outflow (ac-ft/day) 648 907 582 690 2,407
Storage Increase (ac-ft/day) 0.2 0.7 1.2 0.7 2.5
Net Error (ac-ft/day) 0 0 0 0 0
Average Depth (ft) 8.26 4.79 8.54 4.54 2.35

Single Compartment Reservoir Configuration Scenarios

The single compartment (SC) reservoir is the simplest scenario. Two scenarios

were performed to provide a basis for further comparison in the two and four








compartment analyses. The first scenario was the SC base scenario (Figure 4-3). The

second scenario routed STA 3/4 inflows directly from the canal through the single

compartment reservoir before entering the STA. The second scenario is referred to as the

SC STA scenario (Figure 4-4).

Miami Canal NNR Canal


442 ac-ft/d 485 ac-ftWd
EAASR
218 ac-t/d 261 ac-ft/d
A = 30,720 acres
290 ac-t/d 470 ac-ft/d


S1,182 ac-ft/d


219 ac-ft/d 221 ac-ft/d


31 a ac-rIa
------------


STA 3/4


A = 17,920 acres


S2,177 ac-ftd


b03 ac-r/a
I


Figure 4-3. Single Compartment Base Scenario Flow Diagram

The SC base Scenario's long-term water balance for the 36 year average of the

area, inflows, and outflows were calculated (Figure 4-3). The total average inflow to the

EAASR was 1,686 ac-ft/day. Of this total, 1,182 ac-ft/day was released to STA 3/4. The

average depth of the EAASR was 5.9 feet. Thus, the mean hydraulic residence time was

107.5 days or about three and a half months. The long term HRT of 29 days was









calculated for STA 3/4 from the arithmetic mean of the annual HRT presented in Table

4-15. This value is 33 percent higher than if directly calculated from the long term

average flow and depth. An HRT of 29 days was used for all subsequent long-term STA

3/4 analysis.

The SC STA Scenario's long-term water balance for the 36-year average of the

area, inflows, and outflows were calculated (Figure 4-4). The total average inflow to the

EAASR was 2,705 ac-ft/day. Of this total, 2,201 ac-ft/day was released to STA 3/4. The

average depth of the EAASR was 5.9 feet. Thus, the mean HRT was 67.5 days, slightly

more than two months. STA 3/4 results were equal to the single compartment base

scenario.

Miami Canal NNR Canal



661 ac-ftld 705 ac-ftld
EAASR
218 ac-ftld 261 ac-ftld
A = 30,720 acres
606 ac-ft/d 733 ac-ftld
----- > < -------


Figure 4-4. Single Compartment STA Scenario Flow Diagram









Two Compartment Reservoir Configuration Scenarios

The EAASR was portioned into two compartments (TC) for further analysis. The

two compartment configuration is the most similar to the current design of the EAASR.

The total external inflows and outflow for the EAASR are the same as in the single

compartment case. The key difference is in how the water moves between the two

compartments. Four scenarios were evaluated. The first scenario, which was considered

to be the TSP scenario, was referred as the TC base scenario and is presented in Figure

4-5. The second scenario routed STA 3/4 inflows directly from the canals through the

respective compartments before entering the STA. The second scenario is referred to as

the TC STA scenario (Figure 4-6).

The TC base Scenario's long-term water balance for the 36 year average of the

area, inflows, and outflow were calculated (Figure 4-5). The total average inflow to

Compartment 1 was 2,011 ac-ft/day, including inter-reservoir transfers. The total average

inflow to Compartment 2 was 1,284 ac-ft/day, including inter-reservoir transfers. Of this

total, 1,717 ac-ft/day was released to STA 3/4. The average depth of the EAASR was 5.9

feet in both compartments. Thus, the mean HRT for Compartment 1 was 45 days and 70

days for Compartment 2. STA 3/4 results were equal to the single compartment base

scenario.










Miami Canal



661 ac-ftd
2 1,056 ac-ft/d

218 ac-ftfd
17 ac-ftd
A = 15,360 acres
606 ac-ftd


.1


1,717 ac-ft/d


NNR Canal



485 ac-fWd


261 ac-ft/d


470 ac-fJd






221 ac-ft/d


STA 3/4


A = 17,920 acres 263 ac-Wd


2,177 ac-ft/d




Figure 4-5. Two Compartment Base Scenario Flow Diagram

The TC STA Scenario's long-term water balance for the 36 year average of the

area, inflows, and outflows were calculated (Figure 4-6). The total average inflow to

Compartment 1 was 2,495 ac-ft/day, including inter-reservoir transfers. The total average

inflow to Compartment 2 was 1,284 ac-ft/day, including inter-reservoir transfers. Of this

total, 2,201 ac-ft/day was released to STA 3/4. The average depth of the EAASR was 5.9

feet in both compartments. Thus, the mean HRT for Compartment 1 was 37 days and 70

days for Compartment 2. STA 3/4 results were equal to the single compartment base

scenario.









Miami Canal NNR Canal


661 ac-lfd
S 2 1,056 ac-ftd
218 ac-ftd
17 ac-lWd
A =15,360 acres
606 ac-lWd
I


1i


705 ac-tUd

261 ac-ftd

733 ac-Wd


S2,201 ac-ftd



STA 3/4


A = 17,920 acres


S2,177 ac-ftd


Figure 4-6. Two Compartment STA Scenario Flow Diagram

The HRT of Compartment 1 in both the TC base and TC STA scenarios was

relatively low. Low removal rates were expected from this Compartment. It was

hypothesized that an outflow from Compartment 2 to STA 3/4 would alleviate the heavy

loading to Compartment 1 and improve treatment. Two scenarios for were generated to

test the hypothesis, called TC Miami Scenario and TC Miami STA Scenario. The TC

Miami Scenario's long-term water balance for the 36 year average of the area, inflows,

and outflows were calculated (Figure 4-7). The total average inflow to Compartment 1

was 988 ac-ft/day, including inter-reservoir transfers. The total average inflow to

Compartment 2 was 1,284 ac-ft/day, including inter-reservoir transfers. Of this total,

2,201 ac-ft/day was released to STA 3/4. The average depth of the EAASR was 5.9 feet

in both compartments. Thus, the mean hydraulic residence time for Compartment 1 was










92 days and 70 days for Compartment 2. STA 3/4 results were equal to the single

compartment base scenario.


Miami Canal


NNR Canal


661 ac-ftd

218 ac-tid

606 ac-t/d


2


A = 15,360 acres


Figure 4-7. Two Compartment Miami Scenario Flow Diagram

The TC Miami STA Scenario's long-term water balance for the 36 year average

of the area, inflows, and outflows were calculated (Figure 4-8). The total average inflow

to Compartment 1 was 1,472 ac-ft/day, including inter-reservoir transfers. The total

average inflow to Compartment 2 was 1,284 ac-ft/day, including inter-reservoir transfers.

Of this total, 2,201 ac-ft/day was released to STA 3/4. The average depth of the EAASR

was 5.9 feet in both compartments. Thus, the mean hydraulic residence time for

Compartment 1 was 62 days and 70 days for Compartment 2. STA 3/4 results were equal

to the single compartment base scenario.


34 ac-ft/d


17 ac-ftd
14 ---


485 ac-ftfd

261 ac-ftd

470 ac-ftfd





221 ac-ftRd



263 ac-ftfd


1,D23 ac-f/d 1 1,178 ac-f/d



STA 3/4


A = 17,920 acres


I 2,177 ac-f/d


1


A = 15,360 acres









Miami Canal NNR Canal


661 ac-ftld
2 34 ac-ftld
218 ac-ftld
17 ac-ftd
A =15,360 acres -
606 ac-ftd


1I


705 ac-fld

261 ac-ftld

733 ac-ftld


T1,023 ac-ftd 1,178 ac-ft/d



STA 3/4

A = 17,920 acres


S2,177 ac-ftd


Figure 4-8. Two Compartment Miami STA Scenario Flow Diagram

Four Compartment Configuration Scenarios

Two scenarios were analyzed for the four compartment (FC) EAASR

configuration. The total external inflows and outflow for the EAASR are the same as in

the single compartment case. The key difference is in how the water moves among the

four compartments. The first scenario, referred to as FC base scenario, was considered

the MSP configuration (Figure 4-9). The direct flows to STA 3/4 were routed through

Compartments B and D in the FC STA Scenario (Figure 4-10). The results for the

scenarios for each compartment are presented in Tables 4-2 through 4-5. STA 3/4 results

were equal to the single compartment base scenario.









Table 4-2. Compartment A Water Quantity for the Four Compartment EAASR
Configuration
Scenario Area (acres) Inflow (ac-ft/d) Depth (feet) HRT (days)
FC base 5,120 589 8.3 72
FC STA 5,120 589 8.3 72

Table 4-3. Compartment B Water Quantity for the Four Compartment EAASR
Configuration
Scenario Area (acres) Inflow (ac-ft/d) Depth (feet) HRT (days)
FC base 10,240 788 4.8 62
FC STA 10,240 1,272 4.8 39

Table 4-4. Compartment C Water Quantity for the Four Compartment EAASR
Configuration
Scenario Area (acres) Inflow (ac-ft/d) Depth (feet) HRT (days)
FC base 5,120 525 8.5 83
FC STA 5,120 525 8.5 83

Table 4-5. Compartment D Water Quantity for the Four Compartment EAASR
Configuration
Scenario Area (acres) Inflow (ac-ft/d) Depth (feet) HRT (days)
FC base 10,240 570 4.5 82
FC STA 10,240 1,105 4.5 42











iami Canal

442 ac-ft/d


34 ac-ft/d


218 ac-ft/d
4


263 ac-ft/d


83 ac-ft/d


NNR Canal

485 ac-ftNd

261 ac-ftNd


318 ac-ftfdj 71 ac-ft/d


290 ac-ft/d


470 ac-ftWd


17 ac-ft/d
I


488 ac-f/d H 695 ac-ftld


219 ac-ftNd


317 ac-ft/d
c


STA 3/4


2,177 ac-ft/d


Figure 4-9. Four Compartment Base Scenario Flow Diagram


221 ac-ftNd



263 ac-ftNd






44


Miami Canal NNR Canal

442 ac-ft/d 485 ac-ft/d
34 ac-ftld
C
218 ac-ft/d 261 ac-ft/d


263 ac-ft/d 83 ac-ft/d 318 ac-ft/d 71 ac-ft/d


606 ac-ft/d 733 ac-ft/d
17 ac-ft/d

219 ac-ft/d 221 ac-ft/d

1,023 ac-ft/d 1,178 ac-ftld





STA 3/4





1 2,177 ac-ft/d

Figure 4-10. Four Compartment STA Scenario Flow Diagram

Flow Equalization

One of the usual purposes of storage upstream of a treatment system is to reduce

the fluctuations in inflows to the treatment unit. The long-term average mean, standard

deviation, and coefficient of variation (COV) for the inflow and outflow to the EAASR

single compartment scenarios are shown in Table 4-6. The COV is ratio of the standard

deviation and mean values. If equalization is being accomplished, then the COV of the

output should be less than the COV of the input.









Table 4-6. Mean, Standard Deviation, and COV of Single Compartment Scenarios
Inflow Outflow
Standard Coefficient Standard Coefficient
Mean Deviation of Mean Deviation of
Scenario (ac-ft/d) (ac-ft/d) Variation (ac-ft/d) (ac-ft/d) Variation
EAASR
SC base 1,686 3,262 1.9 1,662 4,179 2.5
EAASR
SC STA 2,705 3,904 1.4 2,681 4,591 1.7

The single compartment scenarios of the EAASR do not indicate the reservoir is

used for flow equalization. In both cases, the COV is higher for the outflow than inflow.

The STA scenario displays less variation, but it is not believed that this corresponds to

more flow equalization. With residence times in the range of two months, the EAASR is

able to completely dampen fluctuation from most rainfall events and should have a

significant impact on the longer-term precipitation events during the wet season. The

EAASR is to be a multi-purpose reservoir. Thus, the ability to use it to equalize inflows

to STA 3/4 may be constrained by these other purposes.

Comparison of STAs and Lakes in Southern Florida

Quantity-related design parameters for STAs and lakes and reservoirs are shown

in Table 4-7 (Walker and Kadlec 2005b). With the exception of the Iron Bridge wetland

with an HRT of 66 days, the HRTs for wetlands and STAs are in the range of 7 to 21

days. Thus, the mean residence time of 29 days for STA 3/4 shows that STA 3/4 can

potentially be used to a greater extent. The associated median operating depth for the

comparable is about 1.7 feet. The mean operating depth of 2.3 feet for STA 3/4 is on the

high side, but not exceptional. These STAs are designed as water quality control

facilities; accordingly, there is relatively little variability in how they operate.









In stark contrast to the relatively homogeneous behavior of the 22 STAs, the 15

entries in the lake and reservoir database indicate a wide range of behavior. The four

periods studied for Lake Okeechobee have HRTs in the range of 580 to 715 days. At the

other extreme, four entries have reported HRTs of less than 10 days. Mean depths range

from 4.1 to 11.2 feet. The estimated HRT range for the EAASR was 67 to 107 days with

a mean depth of 5.9 feet. Thus, the possible Lake/Reservoir comparable in Table 4-6

can be reduced to George, Istokpoga, Harney, Jessup, Crescent, and Thonotosassa with

regard to quantity characteristics. More detailed analysis of reservoir comparable will

be presented in the next section on water quality.










Table 4-7. Comparative HLRs and HRTs for 37 Reactors in Southeast Florida (Data from
Walker and Kadlec 2005b)
Mean Max
HLR HRT Depth Depth
Number Category in/day days ft ft % Full
Emergent Wetlands
1 ENRP C1 1.1 17 1.6 2.2 74.3%
2 ENRP C2 2.1 14 2.5 2.9 87.4%
3 ENRP C3 1.1 13 1.1 1.5 75.4%
4 STA1W C1 1.8 13 2.0 3.0 67.9%
5 STA1W C2 2.9 11 2.6 3.5 74.8%
6 STA1W C3 2.6 7 1.6 2.4 66.0%
7 STA5 C1AB 2.6 8 1.7 2.2 79.4%
8 STA5 C2AB 1.5 9 1.1 2.0 57.4%
9 BoneyMarsh 0.8 23 1.5 3.0 51.2%
10 IronBridge 0.4 66 2.3 2.3 99.6%
11 WCA2A 1.4 13 1.5 4.3 35.6%
Median 1.5 13 1.6 2.4 74.3%
Pre-existing Wetlands
1 STA2 C1 1.2 22 2.1 2.7 79.7%
2 STA2 C2 1.9 14 2.2 3.3 64.4%
3 STA6 C3 2.7 7 1.6 2.5 66.0%
4 STA6 C5 1.3 15 1.6 2.3 70.1%
5 WCA2A 1.7 9 1.2 4.0 31.2%
6 WCA2A 1.8 8 1.2 4.0 31.0%
7 WCA2A 1.4 11 1.2 4.0 30.7%
Median 1.7 11 1.6 3.3 64.4%
Submerged Aquatic Vegetation
Systems
1 ENRP C4 5.0 5 2.2 2.5 87.2%
2 STA1W C4 3.7 7 2.0 2.8 72.0%
3 STA1W C5AB 3.7 7 2.2 3.3 66.0%
4 STA2 C3 2.0 17 2.9 4.3 66.2%
Median 3.7 7 2.2 3.1 69.1%
Lakes or Reservoirs
1 OKEE 7578 0.2 674 8.4 10.5 79.6%
2 OKEE 7986 0.2 704 8.8 11.0 80.4%
3 OKEE 8794 0.2 714 8.6 10.9 78.6%
4 OKEE 9599 0.2 578 9.4 11.3 83.7%
5 HELLNBLAZES 24.9 2 4.1 6.4 64.3%
6 SAWGRASS 20.7 3 4.4 6.7 65.2%
7 GEORGE 1.9 73 11.2 12.6 89.0%
8 ISTOKPOGA 0.4 170 5.2 6.8 77.1%
9 ISTOK 2 0.4 178 5.2 6.1 84.3%
10 POINSETT 4.3 8 3.0 6.1 48.2%
11 HARNEY 6.9 16 9.4 15.0 63.0%
12 HARNEY 2 8.7 9 6.7 11.1 60.4%
13 JESSUP 0.4 176 5.3 9.7 54.4%
14 CRESCENT 1.1 112 10.0 12.2 81.8%
15 THONOTO 1.1 54 5.0 7.7 64.6%
Median 1.1 112 6.7 10.5 77.1%









Water Quality for the EAASR and STA 3/4

The water quality of the EAASR and STA 3/4 were separated into the inflow

water quality and uptake parameters. The inflow water quality and each uptake rate

parameter are detailed in this section.

Inflow Water Quality

Inflow to the EAASR and STA3 /4 occurred from four external sources. A mean

concentration for each source was calculated from historical data. Pumping Stations S2

and S3 are used to control the water entering and leaving Lake Okeechobee through the

NNR and Miami Canals. The water passing through pump stations S2 (NNR Canal) and

S3 (Miami Canal) was divided into two types: to LOK from the EAA (back pumping),

and from LOK to the EAA. These were designated as Basin and LOKcanal, respectively

and correspond to the external inflows in the water quantity section. The mean,

coefficient of variation (standard deviation divided by mean), and count of data for each

parameter are presented in Tables 4-8 through 4-11. The data included measurements

from 1973 to 2004, which were taken at different frequencies depending on the

parameter.

In both the NNR and Miami Canals, water quality exiting LOK was of higher

quality than water entering LOK. The mean water quality coming to LOK from the EAA

was higher quality in the Miami Canal than in the NNR Canal. The coefficients of

variation for concentrations to LOK were similar, with the exception of iron, turbidity

and dissolved oxygen. The reverse is true for flows from LOK to the EAA, where S2 had

better water quality than S3, with the exception of iron and phosphorus, as TP. The

coefficients of variation for the flows from LOK are similar with the exception of

sodium.










Table 4-8. NNR Basin Mean, Coefficient of
Parameters


Variation, and Count of Water Quality


NNR Basin
NNR Basin Coefficient of NNR Basin Number
Parameter Mean Variation of Records
TP (mg/L) 0.150 0.53 236
TKN (mg/L) 3.69 0.34 227
NOx-N (mg/L) 1.811 1.01 227
TN (mg/L) 5.505 0.46 227
DO (mg/L) 2.47 0.59 176
pH (su) 7.19 0.05 133
Specific Conductance (uS/cm) 1191 0.21 177
Turbidity (NTU) 10.7 1.34 160
TSS (mg/L) 18.1 1.45 138
Alkalinity as CaCO3 (mg/L) 317 0.27 186
Calcium (mg/L) 109.2 0.22 88
Chloride (mg/L) 137.1 0.28 205
Sulfate (mg/L) 101.9 0.41 67
Sodium (mg/L) 97.8 0.29 88
Total Iron (ug/L) 267 0.89 57
Total Mercury (ng/L) 2.82 0.50 11
Atrazine (ug/L) 0.670 0.97 16

Table 4-9. LOKNNR Canal Mean, Coefficient of Variation, and Count of Water Quality
Parameters
LOKNNR LOKNNR Coefficient of LOKNNR Number of
Parameter Mean Variation Records
TP (mg/L) 0.077 0.46 76
TKN (mg/L) 1.79 0.35 72
NOx-N (mg/L) 0.206 2.01 72
TN (mg/L) 1.998 0.45 72
DO (mg/L) 6.57 0.26 70
pH (su) 7.78 0.04 99
Specific Conductance (uS/cm) 625 0.25 72
Turbidity (NTU) 71.0 0.87 103
TSS (mg/L) 12.9 0.96 67
Alkalinity as CaCO3 (mg/L) 138 0.33 75
Calcium (mg/L) 48.6 0.23 22
Chloride (mg/L) 82.0 0.25 74
Sulfate (mg/L) 49.4 0.27 29
Sodium (mg/L) 52.6 0.30 22
Total Iron (ug/L) 170 0.70 27
Total Mercury (ng/L) NA NA 0
Atrazine (ug/L) 0.280 0.54 26










Table 4-10. Miami Basin Mean, Coefficient of Variation, and Count of Water Quality
Parameters
Miami Basin
Miami Basin Coefficient of Miami Basin Number
Parameter Mean Variation of Records
TP (mg/L) 0.112 0.73 148
TKN (mg/L) 3.06 0.33 142
NOx-N (mg/L) 1.996 0.85 142
TN (mg/L) 5.053 0.46 142
DO (mg/L) 3.45 0.42 125
pH (su) 7.23 0.05 133
Specific Conductance (uS/cm) 962 0.27 124
Turbidity (NTU) 8.2 0.82 106
TSS (mg/L) 12.9 1.10 91
Alkalinity as CaCO3 (mg/L) 248 0.24 112
Calcium (mg/L) 106.7 0.29 55
Chloride (mg/L) 112.2 0.37 125
Sulfate (mg/L) 72.8 0.43 54
Sodium (mg/L) 70.4 0.30 55
Total Iron (ug/L) 192 0.49 44
Total Mercury (ng/L) 2.2 0.33 9
Atrazine (ug/L) 0.470 1.26 13

Table 4-11. LOKMiami Mean, Coefficient of Variation, and Count of Water Quality
Parameters
LOKMiam, LOKMiam, Coefficient LOK.Mami Number of
Parameter Mean of Variation Records
TP (mg/L) 0.060 0.50 105
TKN (mg/L) 1.85 0.41 97
NOx-N (mg/L) 0.297 1.74 97
TN (mg/L) 2.143 0.52 97
DO (mg/L) 6.35 0.30 98
pH (su) 7.8 0.06 99
Specific Conductance (uS/cm) 767 0.42 100
Turbidity (NTU) 6.9 0.84 103
TSS (mg/L) 8.2 0.70 78
Alkalinity as CaCO3 (mg/L) 164 0.44 104
Calcium (mg/L) 64.0 0.44 32
Chloride (mg/L) 101.0 0.45 104
Sulfate (mg/L) 69.5 0.74 37
Sodium (mg/L) 71.5 0.64 32
Total Iron (ug/L) 148 0.93 32
Total Mercury (ng/L) NA NA 0
Atrazine (ug/L) 0.280 2.01 26









As stated previously, TP is the major parameter of interest in the system. The

overall flow-weighted average TP concentration for the EAASR and STA 3/4 system is

0.101 mg/L. TP concentrations range from a low of 0.060 mg/L for LOK releases via the

Miami Canal to a high of 0.150 mg/L for NNR Basin.

The relationship between concentration and flow was investigated for all the

water quality parameters. A linear regression was used to evaluate the relationship. For

NNR Canal, flows for total iron to LOK showed a weak relationship with an R2 of 0.36.

All other parameters did not show a significant relationship. Thus, concentration will be

assumed to be independent of flow for all constituents.

Background Concentration

The removal of TP was modeled using the KC* equation. The background

concentration, C*, is the lowest possible concentration the reactor can reach. An accurate

value of C* is important, because the concentration approaches C* asymptotically. A

Walker and Kadlec (undated) report calibrated values of C* ranging from 0.004 to 0.020

mg/L for STAs. Walker and Havens (2003) use 0.007 mg/L as the background

concentration for precipitation. In this study, the background concentration will be

determined using calibration data with the constraint that C* >= 0.007 mg/L, the

estimated value for precipitation.

Reservoir TP Overall Reaction Rate

Several water quality models have been developed to simulate the TP kinetics in

general CERP reservoirs including Walker and Kadlec (2005a) and USACE and

SFWMD (2005). The general CERP reservoir models use large datasets of comparable

systems to find the average overall TP kinetics. A more specific set of comparable was

sought to develop a water quality model for the Everglades Agricultural Reservoir









(EAASR), a CERP reservoir. The selection of comparable systems is also documented in

Reisinger et al (2006).

Long-term average evaluation of comparable systems

Walker and Kadlec (2005a) tabulated the long term averages of a comprehensive

set of 18 comparable Florida lakes and a reservoir in the documentation of the DMSTA

v.2 calibration (Figure 4-6). The long term average values were used to determine

several closely comparable systems of the Single Compartment base EAASR Scenario.

The SC base scenario was used, because it is the most general of the scenarios. The

analysis was performed on both the water quantity and TP water quality of the

comparable dataset, as they were inseparable for parameter estimation. The analysis of

comparable was performed using three categories of decision variables: lake

characterization, water quantity, and water quality. The decision variables and weighting

factors were chosen to best represent the parameters in Equation 3-4. The selected

decision variables for lake characterization are surface area and depth. Water quantity

was represented by the HRT and HLR. Water quality was represented by the inflow TP

concentration. A weighted decision matrix was used to evaluate the dataset of

comparable. A rating of one to five was given for each decision variable for each lake,

where five was the most comparable to the EAASR. A weighting factor was applied to

each decision variable, where the sum of all weighting factors equals one. The percent of

the total points (five) was used to rate each dataset. Equation 4-6 was used to calculate

the rating of each lake or reservoir.



P = f*100= 1-1- Eqn. 4-6
m
JWR
J=1









Where i = individual lake,j = decision variable, n = number of lakes, m = number of

decision variables, P = percent of total points, f = ratio of points received, W = weighting

factor, and R = rating from one to five.

The weighting factors, ratings, percent of total points, and overall rank for lakes

ranked five and under are presented in Table 4-12. The Crescent Lake, Lake

Istokpoga 2, Lake Jessup, Lake George, and Lake Poinsett datasets were developed by

Burns and McDonnell (2004a). The Lake Istokpoga dataset was developed by Walker

and Haven (2003). Crescent Lake and Lake Istokpoga were found to be clearly the most

comparable overall to the EAASR. A sensitivity analysis was performed using a range

the weighting factors and repeatedly found Crescent Lake and Lake Istokpoga to be the

highest ranked systems. These lakes were determined to be the most comparable to

EAASR and were reviewed in further detail.

Table 4-12. Decision Rankings for Comparable Systems Ranked Five and Better
Percent
Surface Inflow of Total
Decision Variable Area Depth HLR HRT TP Conc. Points Rank
Weighting 0.1 0.1 0.1 0.35 0.35 NA NA
Crescent Lake 4 3 4 5 5 92% 1
Lake Istokpoga 5 5 5 3 5 86% 2
Lake Istokpoga 2 5 5 5 3 5 86% 3
Lake George 3 2 2 4 4 70% 4
Lake Jessup 3 5 5 3 3 68% 5
Lake Poinsett 2 4 1 2 5 63% 3

More detailed analysis of selected comparable

Next, the salient attributes of Crescent Lake and Lake Istokpoga were analyzed in

more detail by looking at time series data and the extent to which the terms in the water

and TP budgets were measured. The analysis was again divided into lake

characterization, water quantity and water quality.









Crescent Lake was found to have a large quantity and high frequency of TP

samples. However, large uncertainty exists in the mass balance due to high proportion of

ungaged flows and a large percentage of TP measurements associated with ungaged

flows. Additionally, no direct stage-area or stage-volume relationship exists. The depth

profile fluctuated regularly from 8.8 feet to 11.5 feet in depth. This more detailed

analysis found that Crescent Lake was not a very reliable data source for parameter

estimation.

Lake Istokpoga has a relatively complete water balance. TP measurements are

infrequent, but have an average of two data points per average residence time. A recent

bathymetric map provides a stage-area and stage-volume relationship for the lake

(SFWMD 2005b). The depth profile fluctuated generally over a relatively small range

from 4.9 feet to 5.6 feet. Three recent reports are available for Lake Istokpoga: Walker

and Havens (2003), Burns and McDonnell (2004a), and South Florida Water

Management District (2005). The conclusion from the more detailed analysis is that

Lake Istokpoga is the best candidate for parameter estimation.

Selection of the appropriate period of record for the chosen comparable system

Lake Istokpoga is located northwest of Lake Okeechobee, near the center of

Highland County. The inflow and outflow TP concentrations and the HRT were assumed

to be key variables for parameter estimation when using Equation 4-3. The mass balance

terms used to calculate the value of key variables were found from recent reports and

weather station records. The period of record (POR) used for parameter estimation was

selected by influent TP concentrations in the range of 0.100 mg/L and no significant

trends over time.









The inflow mass balance terms include gaged and ungaged inflows and rainfall. A

map of Lake Istokpoga and the location of major mass balance terms, except ET, are

shown in Figure 4-11. Inflows to the lake were measured at Arbuckle Creek and

Josephine Creek. TP was measured at each of the creeks. Ungaged inflows were

estimated as 17% of the gaged flow and were assumed to be from seepage (Walker and

Haven 2003). Rainfall was measured at the S-68 structure. The TP concentration in

ungaged inflows was estimated as 0.050 mg/L and rainfall concentrations as 0.007 mg/L

(Walker and Haven 2003). Outflows from the lake occur mainly from the measured S-68

structure and evapotranspiration (ET). Outflows can occur through the controlled

Istokpoga Canal, although it is seldom used and is considered to be a negligible outflow

(Walker and Haven 2003). Pan evaporation was measured approximately 30 miles

northeast of the lake at the S-65 structure. A pan coefficient of 0.76 was used to convert

the measurement to ET (Burs and McDonnell 2004a). The outflow concentration at S-

68 was sampled and it was assumed that no TP was transported in the ET. The volume in

the lake was calculated using the stage-volume relationship reported in SFWMD (2005b)

and the measured headwater stage at the S-68 structure. The removed TP mass was

calculated as the difference in inflow and outflow mass balance terms.

The POR was determined for the mass balance terms presented above. The

beginning date of the POR was determined by the TP samples, which were first taken in

February of 1988 (Burs and McDonnell 2004). Water quantity and quality data from

calendar years 1988 to 2002 were downloaded from the SFWMD DBHYDRO database

and used in the remainder of the analysis. This period reflects the longest continuous

period where all inputs were measured. Missing data were interpolated as needed.




























Figure 4-11. Lake Istokpoga and Mass Balance Locations

Quarterly and bimonthly TP sampling frequencies were found at each measured

flow location during the 1988 to 2005 POR, with the exceptions of 2001 and 2002 that

were sampled less frequently. Time series plots of the TP concentrations indicate an

increasing trend in the latter part of the time series for S-68 (Figure 4-12) and Arbuckle

Creek (Figure 4-13). The desired comparable TP concentration is 0.100 mg/L for the

Arbuckle Creek inflow and with minimal trends for Arbuckle Creek and S-68. If the

entire POR is used, then the TP concentration increases from approximately 0.060 to

0.190 mg/L for Arbuckle Creek and from approximately 0.025 mg/L to 0.085mg/L for S-

68. A portion of the POR was selected to minimize the increasing trend and the optimal

POR for S-68 was determined to be from 1988 through 1997. The outflow had a mean

TP concentration of approximately 0.035 mg/L for this period (Figure 4-12). The portion

of the total POR was reduced to yield an optimal POR for Arbuckle from 1988 through

1995 (Figure 4-13). The inflow concentration for this period was approximately 0.080

mg/L for this period.












0.16

0.14 *

0.12


0.1
E
d 0.08
o0
c. 0.06

0.04
0.04 2

'
0
88 90 92 94 96 98 00 02 04
Year


Figure 4-12. S-68 TP Data and Adjusted POR


0.7


0.6


0.5

-7
10.4
E


S0.3


0.2


0.1


0
88 90 92 94 96 98 00 02 04
Year

Figure 4-13. Arbuckle Creek TP Data and Adjusted POR


*Arbuckle
-88-05
-88-95


* S-68
-88-05
-88-97









A compromise POR from 1988 through 1997 was chosen to produce the least

temporal trends overall. This period provided the least trend in the S-68 dataset and a

slight positive trend from 0.100 mg/L to 0.125 mg/L in Arbuckle Creek. Additionally,

the selected POR was on average closer to the inflow concentration goal than the optimal

Arbuckle period. This POR included 80 TP samples of Arbuckle Creek and Josephine

Creek and 89 TP samples of S-68. The TP samples for each location were analyzed for

trends as a function of the calendar month. Arbuckle and Josephine Creeks included four

to eight samples for each calendar month and S-68 had six to 10 samples for each

calendar month. No clear monthly trend was found.

A POR of January 1, 1988 to December 31, 1997 was determined for Lake

Istokpoga from the constraints set by the TP temporal trends. A monthly averaging

period for the mass balance of Lake Istokpoga was chosen. Preliminary estimates of the

EAASR HRT were found to be between two and five months, which allow for a

parameter estimate on a monthly scale. A monthly averaging period was found to be

acceptable for Lake Istokpoga's water quantity and quality data. Excellent inflow and

stage records allow the estimation of the several short periods of missing data with little

additional error. Due to the bimonthly and quarterly TP sampling, uncertainty exists at

any water quality averaging period. A number of averaging or regression methods could

be reasonably used to estimate the missing monthly TP concentrations for the POR.

Therefore, an averaging period equal to that of the water quantity was chosen.

Rate constant calibration

All TP measurement stations were used for the TP calibration of the overall rate

constant. Data was generated for each station shown in Figure 4-14. It was assumed that

no change in the quantity or quality of the influent occurred between sampling and the









lake. The inflow measuring points, Arbuckle Creek and Josephine Creek, were therefore

relocated to their respective lake inlets (Figure 4-14).

The median value of each sampling station in tg/L was determined (Figure 4-15).

A well defined concentration gradient existed from the Arbuckle Creek inflow at the

northern end of the lake to the outflow at the southeast corner of the lake. A probable

flow path was developed for the Arbuckle and Josephine Creek flows. Due to the low

concentration, the Josephine Creek flow path was not used.


Figure 4-14. Water Quality Sampling Stations in Lake Istokpoga









































Figure 4-15. Median TP Values at Lake Istokpoga Sampling Stations

The average HRT of 197 days was developed for the lake using the mass balance

terms described in the preceding section. Knowing the HRT for the lake, it is possible to

develop a TP concentration vs. residence time curve based on the inlet, outlet, and in-lake

measurements. The resulting curve is shown in Figure 4-16. Using the boundary

conditions of the concentration at 0 days was equal to 0.082 mg/L, the concentration at

60 days was equal to 0.0475 mg/L, and the concentration is 0.028 mg/L at 197.5 days,

then the overall rate constant, kv, was calculated as 0.016 days-1 and the background

concentration, C*, is 0.0254 mg/L.












90


80


70


60
-J


50
.0
c-
8 40
a.
0
S30


20


10



0 50 100 150 200
HRT (days)

Figure 4-16. Lake Istokpoga TP Removal in Arbuckle Creek Flow Path

EAASR reaction rate sensitivity analysis

A sensitivity analysis was performed on the calibration of the EAASR TP


removal rate constant. As a two parameter calibration was performed, both the reaction


rate and background concentration were evaluated. The SC STA scenario was used for


the sensitivity analysis, which is characterized by a 67.5 day HRT and 0.101 mg/L inflow


TP concentration.


The combined effect of C* and the overall reaction rate was first analyzed. The


EAASR outflow concentration for multiple background concentrations and reaction rates

spanning two orders of magnitude were calculated (Figure 4-17). The results indicate, as











expected, an increasing trend in outflow concentration for greater background


concentrations and the greater the reaction rate the less variability in outflow occurred.


For a k of 0.01 days-', which is similar to the calibrated value, the outflow concentration


varied from about 0.055 mg/L at C* equal to 0.005 mg/L to 0.065 mg/L at C* equal to


0.030 mg/L. From this analysis, it was clear that the calibration of k and C* can


significantly affect the results of the modeling.


0.120



0.100

_1

E 0.080

S- k = 0.001
k = 0.005
S0.060 -k = 0.010
Sk = 0.050
S--k = 0.100
0
r 0.040



0.020



0.000
0.000 0.010 0.020 0.030 0.040
Background Concentration, C* (mg/L)

Figure 4-17. Sensitivity Analysis for EAASR Parameters k and C*

The effect of the overall reaction rate on both the EAASR and STA 3/4 was


evaluated. The outflow concentrations for large ranges of EAASR k values were


calculated for the EAASR and STA 3/4. This analysis used a fixed EAASR C* of 0.025


mg/L and the reported STA 3/4 parameters. A large effect on the outflow concentration


of the EAASR was observed for reaction rates less than 0.05 days-'. The STA 3/4











outflow concentration is relatively independent of the EAASR outflow concentration,


which is the only inflow to the STA, and thus reaction rate. This independence was due


to the relatively long HRT and high reaction rate. In fact, if the STA 3/4 received all


inflow without first routing through the reservoir it would still achieve an outflow


concentration of 0.021 mg/L. This result does not account for mass loading and its long


term effect on the STA. Therefore, such an operating scheme is not suggested. Data


from this analysis for selected reaction rates are presented in Table 4-13.


0.100


0.090


0.080


0.070
-J

E
= 0.060
o

I 0.050
o
C
0.040


0 0.030


0.020


0.010


0.000


EAASR
STA 3/4


0 0.05 0.1 0.15
Reaction Rate dayss)

Figure 4-18. Outflow Concentration of the EAASR and STA 3/4 for Varying Reaction
Rates









Table 4-13. Selected Results of the EAASR and STA 3/4 for Varying Reaction Rates
STA 3/4
HRT EAASR Reaction C* EAASR Cin EAASR Cout Cout
(days) Rate (days-) (mg/L) (mg/L) (mg/L) (mg/L)
67 0.001 0.025 0.101 0.096 0.021
67 0.001 0.025 0.101 0.096 0.021
67 0.010 0.025 0.101 0.064 0.021
67 0.005 0.025 0.101 0.079 0.021
67
0.010 0.025 0.101 0.064 0.021

67 0.050 0.025 0.101 0.028 0.020
67
0.100 0.025 0.101 0.025 0.020
67 0.150 0.025 0.101 0.025 0.020

The Basis of Design Report provides an alternative analysis of the EAASR TP

uptake. The parameter estimates needed to produce these results will be evaluated to

allow comparison between the simulations. The report models only compartment 1 of the

TC configuration, which is referred to as Phase 1. Phase 1 consists of a 12 foot reservoir

on 15,833 acres of Component A creating 190,551 acre-feet of possible storage (SFWMD

2006). The location and simulated water balance are shown in Figure 4-19. The water

balance was simulated from the SFWMM Version 5.4.2 using the ECP 2010 and 2015

simulations, as compared to the 2050 Next Added Increment simulation used in the

previous scenarios. An average daily inflow of 1,988 acre-feet per day was calculated

from the water balance and an average depth over the POR of 4.5 feet was reported. An

HRT of 35.8 days was calculated for Phase 1.

SFWMD (2006) reported an average TP concentration of 82 ppb entering the

reservoir. DMSTA 2 was used to simulate Phase 1 and found an outflow of 68 ppb or a

17 percent removal. These results were generated using an areal reaction rate on a daily

basis. To obtain these results using the UF WQDT with a background concentration of

00.0254 mg/L a reaction rate of 0.0033 days- would be needed. The value is almost one

order of magnitude lower than the Lake Istokpoga estimate. Due to the differences in









configuration and reactions rates the two modeling efforts should not be directly

compared.








































Figure 4-19. Water Balance and Location of the EAASR Phase 1 (SFWMD 2006).

STA TP Reaction Rate

Key parameters for 16 STAs in southeast Florida are shown in Table 4-14. The

average inflow TP concentration is 0.102 mg/L. The optimal values of C* and k can be









determined by solving the following constrained optimization problem using the Excel

Solver. The objective function is to minimize the sum of the squares of the errors

between the calculated and measured outflow concentrations for the 16 STAs. C* is

constrained to be >= 0.007 mg/L. Optimized values of two decision variables k and C*

were calculated as 0.127 days-1 and 0.0194 mg/L, respectively for an assumed average

inflow concentration of 0.102 mg/L.

Table 4-14. Characteristics of 16 STAs in Southeast Florida (Data from Walker and
Kadlec 2005a)
Mean Inflow Outflow Percent
STA Depth HRT Conc. Conc. Removal
(ft) (days) (mg/L) (mg/L)
ENRP C1 1.6 17 0.089 0.039 56%
ENRP C2 2.5 14 0.068 0.034 49%
ENRP C3 1.1 13 0.039 0.019 50%
ENRP C4 2.2 5 0.036 0.017 53%
STA1W C1 2.0 13 0.140 0.05 64%
STA1W C2 2.6 11 0.142 0.088 38%
STA1W C3 1.6 7 0.058 0.033 43%
STA1W C4 2.0 7 0.100 0.032 68%
STA1W C5AB 2.2 7 0.153 0.055 64%
STA2 C1 2.1 22 0.100 0.012 88%
STA2 C2 2.2 14 0.110 0.021 81%
STA2 C3 2.9 17 0.128 0.015 89%
STA5 C1AB 1.7 8 0.121 0.071 42%
STA5 C2AB 1.1 9 0.203 0.125 39%
STA6 C3 1.6 7 0.071 0.019 73%
STA6 C5 1.6 15 0.079 0.016 79%
Average 2.0 11 0.104 0.042 60%

Predicted Performance of EAASR Removal of TP

The calculated outflow concentrations from the EAASR for various assumed

values of initial concentration and residence time are shown in Figure 4-20. For a

residence time of 50 days, the effect of the initial concentration at 50 days and the

associated percent removal is shown in Table 4-15. The percent removal varies from a











low of 19 % if the initial concentration is 0.025 mg/L to 42 % if the initial TP

concentration is 0.150 mg/L.


0.16


0.14


0.12


=- 0.1

o.
0.08
C

0
S 0.06


0.04


0.02


0


- Cin = 0.050
-- Cin = 0.075
--Cin = 0.100
Cin= 0.125
-- Cin = 0.150


0 25 50 75 100 125 150
HRT (days)

Figure 4-20. Effect of Initial Concentration and Residence Time on Outflow TP
Concentration for the EAASR

Table 4-15. Effect of Initial TP Concentration on TP at 50 Days and Percent Control for
the EAASR
Initial TP (mg/L) TP (mg/L) at 50 days % Control
0.050 0.04 19%
0.075 0.06 25%
0.100 0.07 28%
0.150 0.09 42%


Predicted Performance of STA 3/4 Removal of TP

The primary differences between the EAASR and STA 3/4 with regard to the


removal Equation 4-3 is that the residence times in the STAs are typically in the range of









7 to 21 days whereas the rate constant is an order of magnitude higher at 0.127 day-.

The effect of the much higher rate constant for the STAs is dramatic as shown by

comparing Figure 4-20 for the EAASR and Figure 4-21 for the STAs. An assumed rate

constant of k equals to 0.127 day-1 causes the concentration to reach the assumed

background level in 50 days over the entire range of assumed inflow concentrations.

Figure 4-22 shows the same information with the x axis rescaled to a maximum residence

time of 50 days instead of 150 days.

The results shown in Figure 4-22 suggest that using residence times in the 7 to 21

day range provide a relatively good level of performance. For a residence time of 15

days, the associated outflow concentrations and % control are shown in Table 4-16. In

this case, the percent control ranges from a low of 60% for an initial concentration of

0.025 mg/L to a high of 76% if the initial concentration is 0.150 mg/L, though all are

approaching the background concentration.

These results indicate the following key points about the EAASR/STA system:

* The STAs are much more effective water quality controls than the EAASR in that
they achieve significant removals in 7 to 21 days of residence time as dramatically
illustrated by comparing Figures 4-25 and 4-27.

* The EAASR can provide significant water quality improvements if the residence
times exceed 25 days and the initial concentrations are relatively high.

* For the same area, the residence times in the STAs can be expected to be
significantly less than for the EAASR since they are operated at much shallower
depths.

* The initial concentration has a significant effect on performance especially if
performance is measured as percent pollutant removal.














0.16


0.14



0.12



2 0.1


08

0.08
U
o 0.06


0.04



0.02



0


I \


HRT (days)


Figure 4-21. Effect of Initial Concentration and Residence Time on Outflow TP

Concentration for the STAs


016



014



012



2 01


0
0 08
C
U
0 06



0 04


---Cn = 0050
-- Cin = 0075
--Cin = 0 100
in= 0 125
-- Cin= 0 150


10 15 20 25
HRT (days)


Figure 4-22. Figure 4-21 Rescaled to Residence Times Up to 50 Days


---Cin= 0.050
---Cin = 0.075
-- Cin = 0.100
Cin = 0.125
--Cin = 0.150









Table 4-16. Effect of Initial TP Concentration on TP at 15 Days and Percent Control for
the STAs
Initial TP (mg/L) TP (mg/L) at 50 days % Control
0.050 0.020 60%
0.075 0.021 72%
0.100 0.021 79%
0.150 0.022 86%

Summary and Conclusions

The UF WQDT is a steady-state mass balance model utilizing the KC* kinetic

model and was developed to determine the TP water quality of the EAASR and STA 3/4.

The model uses a single aggregate inflow and concentration, depth, background

concentration, and a volumetric reaction rate to simulate TP. The model calculates the

HRT and HLR of each reactor, as well as the outflow concentration and percent removal.

Long-term averages were determined to be the most appropriate level of sophistication

needed in order to provide key insights for the dynamic EAASR design.

Water quantity and quality inputs to the model were calculated for the base and

STA scenarios for each configuration. The HRT of the EAASR ranged from 39 to 107.5

days. The EAASR was not used for flow equalization. The HRT of STA 3/4 was

calculated to be 29 days for all scenarios. Mean inflow concentrations of 17 parameters

of interest for the four external inflow sources to the reservoir and STA 3/4 were

reported. Parameter estimations were performed using closely comparable systems with

high quality data. Lake Istokpoga was found to be the best comparable for the general

single compartment scenario of the EAASR. Data from 16 STAs were used as

comparable to STA 3/4. The following water quality parameters estimates for the

EAASR and STA 3/4 were selected:

* C* = 0.0254 mg/L for the reservoir and 0.0194 mg/L for the STA
* k = 0.016 day'1 for the reservoir and 0.127 day'1 for the STA









A sensitivity analysis of the EAASR reaction rate and background concentration

was performed for the SC STA scenario. The calibration of each parameter was found to

significantly affect the results especially if the reaction rates are less than 0.05 days-1 for a

background concentration of 0.025 mg/L. To reproduce TP removal estimates provided

in the BODR, a reaction rate of 0.0033 days-1 for a C* of 0.025 is needed. Due to the

differences in configurations and reaction rates the modeling efforts are not directly

comparable.

An analysis of the predicted performance of the EAASR and STA 3/4 systems

was performed. STAs were found to provide significantly greater water quality treatment

than reservoirs. The STA systems will approach the background concentration

asymptotically for retention times greater than 25 days. However, since the STA is

operated at much lower depths than the reservoir, they will have much lower HRT for a

given area and inflow. The initial concentration can affect the outcome of the EAASR or

STA 3/4, especially if measured by the percent removal of TP.

The initial concentration of each scenario is calculated in Chapter 5, as the inflow

of the TC and FC scenarios are a function of their removal. The resulting concentrations

and the inputs from this chapter will be used to generate TP estimates for each scenario.














CHAPTER 5
ESTIMATED WATER QUALITY CHANGES IN THE EAASR AND STA 3/4
SYSTEMS

The results of the analysis of the SFWMM output and the water quality studies to

estimate the outflow concentrations from the EAASR and STA 3/4 are provided in this

chapter. Each scenario is evaluated on a steady-state long-term basis using the UF

WQDT. Key insights are discussed for each scenario and summarized at the conclusion

of the section. The annual variability in the SC STA scenario is evaluated to provide an

indication of the variability in the 36 year data set. This analysis was performed using a

variant of the UF WQDT that made annual steady-state evaluations. Key insights on the

variability of water quality are discussed.

Analysis of Long-Term Averages

The EAASR and STA 3/4 were evaluated on a steady-state basis for the average

behavior over the 36 years of SFWMM simulation. Results from the UF WQDT for each

scenario are presented below for the one, two, and four compartment reservoir options.

Single Compartment Reservoir

The single compartment reservoir was the simplest configuration. Two scenarios

were evaluated to provide a basis for further comparison with the two and four

compartment scenarios; the SC base Scenario (Figure 4-3) and the SC STA scenario

(Figure 4-4). The depth of each scenario was varied to 9 feet and 12 feet from 5.9 feet to

illustrate the effects of depth on the overall performance of the reservoir and STA 3/4.

STA 3/4 received water directly from the NNR and Miami Canals in the SC base









scenario. In the SC STA scenario, STA 3/4 inflows were first routed through the single

compartment reservoir (Figure 4-4), and therefore the inflow concentration is equal to the

EAASR outflow concentration. The results of the above scenarios are presented for the

EAASR and STA 3/4 in Tables 5-1 and 5-2, respectively.

Table 5-1. Single Compartment EAASR Results for Total Phosphorus with Variable
Depths
Area Inflow Depth HRT Cin Cout Percent
Scenario (acres) (ac-ft/d) (ft) (days) (mg/L) (mg/L) Removal
SC base 30,720 1,686 5.9 108 0.105 0.040 62%
SC base @ 9 ft. 30,720 1,686 9.0 164 0.105 0.031 70%
SC base @ 12 ft. 30,720 1,686 12.0 219 0.105 0.028 74%
SC STA 30,720 2,705 5.9 67 0.101 0.051 49%
SC STA @ 9 ft. 30,720 2,705 9.0 102 0.101 0.040 60%
SC STA @ 12 ft. 30,720 2,705 12.0 136 0.101 0.034 66%

Table 5-2. STA 3/4 Results for the Single Compartment EAASR for Total Phosphorus
with Variable Depths
Area Inflow Depth HRT Cin Cout Percent
Scenario (acres) (ac-ft/d) (ft) (days) (mg/L) (mg/L) Removal
SC base 17,920 2,201 2.4 29 0.066 0.021 69%
SC base @ 9 ft. 17,920 2,201 2.4 29 0.061 0.020 67%
SC base @ 12 ft. 17,920 2,201 2.4 29 0.060 0.020 66%
SC STA 17,920 2,201 2.4 29 0.051 0.020 61%
SC STA @ 9 ft. 17,920 2,201 2.4 29 0.040 0.020 50%
SC STA @ 12 ft. 17,920 2,201 2.4 29 0.034 0.020 42%

The results of varying the depth show that it may be possible to improve the

quality of the EAASR outflow by increasing the mean depth beyond 5.9 feet. The

EAASR has a maximum depth of 12 feet. If it could be kept full, then the HRT increases

from 108 days to 219 days resulting in a significant improvement in water quality.

However, it is unlikely that the EAASR could be operated full since this would conflict

with other purposes such as flood control and water supply. An intermediate increase in

depth to 9 feet also shows good improvement in water quality and would be more

feasible to attain.









Routing all STA 3/4 inflows, the SC STA scenario, through the EAASR provides

large gains in water quality when compared to the base scenario. Long-term STA 3/4

inflow concentrations were identical for the SC STA scenario and the SC base scenario at

12 feet of depth. Therefore, it may be possible to achieve similar gains to increasing the

depth of the EAASR by routing all STA inflow through the EAASR first. The

combination of increasing depth and routing STA flows achieves some additional gains

in water quality. At the low concentrations of the STA, it is difficult to achieve large

gains in water quality. Large gains are more easily made by altering the EAASR

operations.

STA 3/4 alone could achieve nearly the same water quality as the EAASR/STA

two reactor system because of the 29 day HRT. Recall from Figure 4-22, the outflow TP

concentration at an HRT of 29 for STA 3/4 is approaching the background level. This

suggests that the HRT may be under used and more inflow could be accommodated to

reduce its HRTs to about 10-15 days, without a significant decrease in load reduction.

Two Compartment Reservoir

The two compartment reservoir differs from the single compartment case mainly

due to the interaction between the two compartments and the elimination of direct flow

from the Miami Canal. The base scenario was modeled as presented (Figure 4-5). In the

TC STA Inflow scenario, the direct NNR canal flows to STA 3/4 were routed through the

reservoir before being released to the STA (Figure 4-6). The results of the analyses are

shown in Tables 5-3 through 5-5. The TC base scenario EAASR achieved good removal,

though lower than the SC base scenario. However, the STA achieved identical outflow

concentrations in the SC and TC base scenarios. The increased loading in the TC STA

scenario decreased the EAASR performance, as expected. However, the STA inflow









concentration was less than the TC base scenario due to mixing and treatment in the

EAASR. The STA outflow concentration TC STA scenario was higher than the SC STA

scenario, reinforcing the findings presented in the above section.

The higher concentrations in Compartment 1 as compared to the SC scenarios

were mainly caused by reduced treatment due to the higher loading for Compartment 2.

An outflow from C2 to STA 3/4 was included in the TC Miami and TC Miami STA

(Figures 4-7 and 4-8), and the outcome of the resulting scenario analyses are presented in

Tables 5-3 through 5-5. The resulting reduction in load increased the inflow

concentration and treatment in Compartment 1, resulting in little gains. The outflow

from C1 decreased the inflow concentration of STA 3/4 and therefore resulted in lower

outflow concentrations. From these results, it is believed that the addition of an outflow

from Compartment 2 to STA 3/4 may allow additional water quality benefits, due to

operational flexibility and mixing.

The water quality gains were evaluated on a mass basis (Table 5-6). The total

removal from the system was extremely similar between scenarios, with the exception of

TC Miami. However, routing the STA inflows through the reservoir shifted the removal

from the STA to the reservoir. These findings are consistent with those for the single

compartment scenarios. The lower removal in the case of TC Miami is largely a factor of

less mass to be removed in the STA.

Table 5-3. Compartment 1 Result for the Two Compartment EAASR Configuration
Area Inflow Depth HRT Cin Cout Percent
Scenario (acres) (ac-ft/d) (ft) (days) (mg/L) (mg/L) Removal
TC base 15,360 2,011 5.9 45 0.078 0.051 35%
TC STA 15,360 2,495 5.9 37 0.084 0.058 31%
TC Miami 15,360 988 5.9 92 0.112 0.045 60%
TC Miami STA 15,360 1,472 5.9 62 0.111 0.057 49%









Table 5-4. Compartment 2 Result for the Two Compartment EAASR Configuration
Area Inflow Depth HRT Cin Cout Percent
Scenario (acres) (ac-ft/d) (ft) (days) (mg/L) (mg/L) Removal
TC base 15,360 1,284 5.9 70 0.087 0.045 48%
TC STA 15,360 1,284 5.9 70 0.087 0.045 48%
TC Miami 15,360 1,284 5.9 70 0.087 0.045 48%
TC Miami STA 15,360 1,284 5.9 70 0.087 0.045 48%

Table 5-5. STA 3/4 Result for the Two Compartment EAASR Configuration
Area Inflow Depth HRT Cin Cout Percent
Scenario (acres) (ac-ft/d) (ft) (days) (mg/L) (mg/L) Removal
TC base 17,920 2,201 2.4 19 0.064 0.020 76%
TC STA 17,920 2,201 2.4 19 0.058 0.020 77%
TC Miami 17,920 2,201 2.4 19 0.048 0.020 77%
TC Miami STA 17,920 2,201 2.4 19 0.052 0.020 77%


Table 5-6. TP Removal by


Mass in Kilograms per Day


Scenario Cl C2 EAASR Total STA 3/4 System Total
TC base 67 67 134 119 253
TC STA 80 67 147 104 250
TC Miami 82 67 148 77 225
TC Miami STA 98 67 165 86 251

Four Compartment Reservoir

Two scenarios were analyzed for the four compartment (FC) EAASR

configuration. The FC base scenario was modeled as presented in Figure 5.9. The FC

STA scenario was modeled as presented in Figure 5-10. The results for each

compartment are presented in Tables 5-7 through 5-11. Results from the analysis are

similar to the two compartment configuration, therefore little water quality gains were

realized from the increased compartmentalization.

Table 5-7. Compartment A Result for the Four Compartment EAASR Configuration
Area Inflow (ac- Depth HRT Cin Cout %
Scenario (acres) ft/d) ft (days) (mg/L) (mg/L) Removal
FC base 5,120 589 8.3 72 0.131 0.059 55%
FC STA 5,120 589 8.3 72 0.133 0.060 55%









Table 5-8. Compartment B Result for the Four Compartment EAASR Configuration
Area Inflow (ac- Depth HRT Cin Cout %
Scenario (acres) ft/d) ft) (days) (mg/L) (mg/L) Removal
FC base 10,240 788 4.8 62 0.085 0.058 32%
FC STA 10,240 1,272 4.8 39 0.088 0.063 28%

Table 5-9. Compartment C Result for the Four Compartment EAASR Configuration
Area Inflow (ac- Depth HRT Cin Cout %
Scenario (acres) ft/d) (ft) (days) (mg/L) (mg/L) Removal
FC base 5,120 525 8.5 83 0.102 0.046 46%
FC STA 5,120 525 8.5 83 0.102 0.050 43%

Table 5-10. Compartment D Result for the Four Compartment EAASR Configuration
Area Inflow (ac- Depth HRT Cin Cout %
Scenario (acres) ft/d) (ft) (days) (mg/L) (mg/L) Removal
FC base 10,240 570 4.5 82 0.067 0.047 45%
FC STA 10,240 1,105 4.5 42 0.068 0.048 46%

Table 5-11. STA 3/4 Result for the Four Compartment EAASR Configuration
Area Inflow (ac- Depth HRT Cin Cout %
Scenario (acres) ft/d) (ft) (days) (mg/L) (mg/L) Removal
FC base 17,920 2,201 2.4 19 0.053 0.020 77%
FC STA 17,920 2,201 2.4 19 0.056 0.021 70%

The water quality gains were evaluated on a mass basis (Table 5-12). The total

removal from the system was virtually identical between scenarios (Table 5-11). Routing

the STA inflows through the reservoir shifted the majority of the removal from the STA

to the reservoir.

Table 5-12. TP Removal by EAASR and STA 3/4 in Mass in Kilograms per Day
Scenario A B C D EAASR Total STA 3/4 System Total
FC base 52 26 36 14 129 90 281
FC STA 53 39 34 27 153 95 280

Annual Variability in Performance for the SC STA Scenario

As shown in long-term analysis, the overall performance does not vary greatly for

the one, two, and four compartment cases. Also, it is desirable to direct inflows through

the EAASR before sending them to STA 3/4. Thus, the SC STA scenario (Figure 4-4)









will be used to estimate the annual variability in performance due to variability in inflow

rates and operating depths.

The annual variability in the SC STA scenario for the EAASR and for STA 3/4

are reported in Table 5-13 and Table 5-14, respectively. On an annual basis, the reservoir

varies over a large range of flows and depths. The annual average inflows varied from

695 to 5,662 acre feet/day. The mean annual depths ranged from 0.7 to 11.9 feet. The

resulting calculated HRTs are between 23 and 113 days. The inflow concentration to the

EAASR was relatively stable and varied from 0.084 to 0.135 mg/L. The calculated

outflow concentrations reflected the HRT and Cin characterizations and varied from 0.042

to 0.092 mg/L. The associated percent removals varied from as low as 24% to a high of

69%.

On an annual basis, STA 3/4 received inflows ranging from 253 to 5,540 acre feet

per day. Average depths ranged from 1.6 to 3.1 feet. The resulting calculated HRTs were

between 10 and 116 days. The inflow concentration varied widely from 0.042 to 0.092

mg/L. The calculated outflow concentrations ranged from 0.019 to 0.029 mg/L.








Table 5-13. Annual EAASR Variabilit for the Sinle Co artment STA Scenario
EAASR EAASR EAASR EAASR EAASR Percent
Year Qin HRT Depth Cin Co.t Removal
(ac-ft/d) (days) (ft) (m /L) (m /L) (%)
1965 2,104 56 3.83 0.113 0.061 46o
1966 3,898 73 9.32 0.097 0.048 51%
1967 1,641 70 3.74 0.118 0.056 53%
1968 3,471 62 6.97 0.104 0.055 47%
1969 4,236 70 9.64 0.096 0.049 50%
1970 4,942 70 11.18 0.088 0.046 48%
1971 1,616 91 4.79 0.126 0.049 61%
1972 1,294 118 4.96 0.134 0.042 69%
1973 1,074 97 3.41 0.127 0.047 63%
1974 1,789 50 2.92 0.117 0.066 43%
1975 1,755 73 4.15 0.124 0.056 55%
1976 1,220 105 4.17 0.126 0.044 65%
1977 1,564 112 5.69 0.128 0.043 67%
1978 2,608 92 7.80 0.107 0.044 59%
1979 5,111 41 6.85 0.089 0.058 34%
1980 1,749 67 3.81 0.108 0.054 50%
1981 992 46 1.50 0.135 0.077 43%
1982 3,354 63 6.93 0.107 0.055 49%
1983 5,290 57 9.75 0.088 0.051 42%
1984 3,643 62 7.39 0.088 0.048 45%
1985 1,505 62 3.04 0.122 0.061 50%
1986 1,701 113 6.27 0.124 0.041 66%
1987 1,671 66 3.57 0.114 0.056 51%
1988 1,757 44 2.53 0.099 0.061 38%
1989 937 23 0.69 0.122 0.092 24%
1990 695 54 1.22 0.135 0.072 47%
1991 3,670 53 6.35 0.095 0.055 42%
1992 3,420 59 6.60 0.097 0.053 45%
1993 3,548 68 7.87 0.093 0.048 48%
1994 3,618 63 7.45 0.105 0.054 48%
1995 3,914 93 11.91 0.090 0.040 56%
1996 4,118 75 10.09 0.086 0.044 49%
1997 2,556 78 6.45 0.100 0.047 53%
1998 5,662 38 7.09 0.084 0.057 32%
1999 3,214 66 6.95 0.093 0.049 48%
2000 2,057 84 5.60 0.101 0.045 55%
Average 2,705 67 5.90 0.101 0.051 49%
COV 0.51 0.33 0.47 0.16 0.21 NA









Table 5-14. Annual STA 3/4 Variability for the Single Compartment STA Scenario
STA 3/4 STA 3/4 STA 3/4 STA 3/4 STA 3/4 Percent
Year Qin HRT Depth Cin Cout Removal
(ac-ft/d) (days) (ft) (mg/L) (mg/L) (%)


1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000


1,062
3,686
1,250
2,818
3,612
4,726
1,042
1,149
521
1,494
1,224
687
598
2,787
4,293
1,673
441
2,224
5,161
3,054
896
1,182
1,031
1,487
532
253
3,149
2,584
3,111
2,585
3,555
4,071
1,428
5,540
2,384
1,957


1.72
2.80
2.07
2.53
2.83
3.17
2.07
2.10
1.64
2.08
1.92
1.96
1.86
2.53
2.84
2.27
1.57
2.34
3.14
2.60
2.01
2.02
2.04
2.11
1.82
1.63
2.57
2.49
2.78
2.67
3.08
3.03
2.19
3.24
2.45
2.29


0.061
0.048
0.056
0.055
0.049
0.046
0.049
0.042
0.047
0.066
0.056
0.044
0.043
0.044
0.058
0.054
0.077
0.055
0.051
0.048
0.061
0.041
0.056
0.061
0.092
0.072
0.055
0.053
0.048
0.054
0.040
0.044
0.047
0.057
0.049
0.045


0.020
0.024
0.020
0.024
0.024
0.025
0.020
0.020
0.019
0.021
0.020
0.019
0.019
0.023
0.028
0.021
0.019
0.023
0.027
0.024
0.020
0.020
0.020
0.021
0.019
0.019
0.025
0.023
0.023
0.023
0.022
0.024
0.020
0.029
0.022
0.021


67%
49%
64%
56%
50%
45%
60%
53%
59%
68%
64%
56%
54%
49%
52%
61%
75%
59%
46%
51%
68%
52%
65%
66%
79%
73%
55%
56%
52%
58%
44%
45%
57%
49%
54%
53%


Average 2,201 29 2.35 0.051 0.022 57%
COV 0.65 0.73 0.20 0.21 0.12 NA











Extensive operating information and guidelines have been developed for STAs in


South Florida. For example, the draft operations plan for STA-1E indicates an operating


depth range between 0.5 and 4.5 feet (Goforth 2006). Annual depth results (Table 5-14)


indicate that the STA is operated in this range.


The annual summaries of the operation of the EAASR and STA 3/4 over the 36


year period are summarized in cumulative density functions (CDFs) (Figure 5-1) that


indicate the % <= indicated value as a function of variables of interest. As shown in


Figure 5-1, the mean annual inflows of the EAASR and STA 3/4 follow a similar pattern,


which was expected as all STA inflows are first routed through the EAASR.


100%

90%

80%

70%
01
m 60%
"c
*i 50%
._
* 40%
II
V
v
30%

20%

10%

0%


--EAASR
STA 3/4


0 1,000 2,000 3,000 4,000 5,000 6,000
Annual inflow, aflday


Figure 5-1. Mean Annual Inflows from SC STA Scenario

A clear difference in operating depths is shown in the CDF's for EAASR and


STA 3/4 (Figure 5-2). The depth profile of the EAASR is nearly linear 2:1 slope







82


throughout the entire range of the design depth. The STA operates over a relatively

narrow range from 1.5 to 3.5 feet, as stated previously.


100%

90%

80%

70%

g 60%

S50%

.E 40%
II
v
30%

20%

10%

0%


-- EAASR
- STA 3/4


0 2 4 6 8 10 12
Mean depth, ft.


Figure 5-2. Mean Annual Depths from SC STA Scenario

The HRT profile reflects the difference in depth profiles (Figure 5-3). The

EAASR operates under a large range of HRT's. The HRT distribution shows STA 3/4


operating in the within the normal range of STAs, 14 to 21 days, for approximately 19%

of the years for <= 14 days and 50% of the years for <= 21 days. The remaining 50

percent of HRT are above this range, again indicating the STA may be able to receive

more hydraulic loading.












100%

90%

80%

70%

> 60%

i* 50%
.2
.E 40%
II
30%

20%

10%

0%


0 20 40 60 80 100 120 140
Hydraulic residence time, days


Figure 5-3. Mean Annual HRT from SC STA Scenario

The reductions of TP concentrations through the EAASR/STA 3/4 system were


observed (Figure 5-4). Median (50%) inflow TP concentration to the EAASR is about


0.110 mg/L. This TP concentration is reduced to a median of about 0.050 mg/L as it


leaves the EAASR and enters STA 3/4. It exits STA 3/4 at a TP concentration of about


0.020 mg/L. As the treatment progresses through the EAASR and STA, the variability in


concentration decreases as shown by the increasingly vertical CDF curves. This is an


indication of a functioning storage/treatment system.


--EAASR
STA 3/4











100%
90%
80%
70%
S60%
50%
S40%
v 30%
20%
10%
0%


-- EAASR Cin
- EAASR Cout
- STA 3/4 Cout


0.0 50.0 100.0 150.0
TP concentration, ppb

Figure 5-4. Mean Annual Inflow and Outflow Concentrations from SC STA Scenario

Summary and Conclusions

The overall results for the three configurations are shown in Table 5-15. The

input TP concentrations vary from 0.067 to 0.133 mg/L depending on the number of

compartments. Similarly, the output concentrations from the EAASR indicate a

significant treatment effect with the concentrations reduced to 0.028 to 0.063 mg/L

depending on the configuration. However, the variability in performance is greatly

dampened in the STA 3/4 outflow with the outflow concentrations all very similar in the

0.020-0.021 mg/L range. The EAASR has a significant water quality impact because its

hydraulic residence times exceed two months. The residence times in STA 3/4 are also

high for an STA with a calculated mean of about 29 days. Thus, the combined system

has ample time to store and treat the water. The EAASR can also provide a significant

ability to manage the inflows to STA 3/4 in order to optimize its performance. From an

operational point of view, it is desirable to maximize the amount of inflow that can be

passed through the EAASR before entering STA 3/4.









Table 5-15. Summary Results for the One, Two, and Four Compartment Scenarios
Compartments EAASR EAASR STA
Cin (mg/L) Cout (mg/L) Cout (mg/L)
1 0.101-0.105 0.028-0.051 0.020-0.21
2 0.078-0.112 0.045-0.058 0.020
4 0.067-0.133 0.046-0.063 0.020-0.021

The analysis of annual variability of the SC STA scenario found a similar inflow

pattern for the EAASR and STA 3/4. The depth profiles were clearly different, with the

EAASR fluctuating through out the entire range of design depths and STA 3/4 remaining

relatively constant. The variation in concentration decreased as the water was treated by

the EAASR/STA 3/4 system. This indicated a functioning storage/treatment system.

The multiple scenarios were compared and the following conclusions were drawn. The

compartmentalization of the reservoir can provide additional operational flexibility to

manage the mixing of influents and treatment times to achieve additional water quality

improvements. An additional outflow structure from Compartment 2 to STA 3/4 can

increase the operational flexibility of the EAASR and may increase water quality

improvements. Operating the EAASR at greater depths may further increase the water

quality treatment. The EAASR does not significantly improve the outflow water quality

from STA 3/4 as compared to using STA 3/4 only because of the relatively long

residence times in STA 3/4. This suggests that the current inflows to STA 3/4 could be

increased to improve water quality even more in the EAASR/STA 3/4 storage/treatment

train.

The actual performance of the EAASR/STA system could vary widely from the

predictions presented in this thesis for several reasons including:

* Inflow quantities and water quality can be expected to vary over the next 50 years






86


* The SFWMM estimates of inflows and operations are not based on any kind of
optimization for the EAASR/STA system but represent an estimate of their role in a
regional water management scenario.

* The behavior and performance of the EAASR/STA can be expected to vary
depending on how it is operated for flood control and water supply purposes as
well as water quality enhancement.














CHAPTER 6
SUMMARY AND CONCLUSIONS

This effort aimed to create and use a model to estimate water quality in the

EAASR and Stormwater Treatment Area (STA) 3/4. The model was designed to

function in the dynamic and challenging EAASR design process. Key insights on the

most current EAASR and STA 3/4 configurations were found.

Chapter 2 presented the background and previous work on CERP, the EAASR,

systems engineering, and reservoir modeling. CERP is an ambitious project to "get the

water right" in the South Florida Ecosystem and restore the Everglades. The project goal

will be accomplished by all 63 projects operating together. The Acceler8 project

streamlines the CERP planning, design, and construction process to provide benefits to

the system more quickly and cost-efficiently. The EAASR, a CERP and Acceler8

project, will provide storage for water supply, flood control, and flow equalization for

water quality treatment areas. Water supply and flood control design approaches are well

developed, while water quality approaches are less developed. The streamlining of the

planning and design of the EAASR has posed several challenges resulting in an iterative

design process with multiple conceptual design alternative formulations and analyses.

Due to differences in the simulation, uncertainty exists when these multiple formulations

and analyses are compared.

The systems engineering approach to design provides a proven approach that can

meet the challenges of the EAASR planning and design process. The approach

incorporates disciplinary models that may produce a more optimal design than traditional









approaches. A model in the water quality discipline was therefore sought to assess which

of several alternatives are better. Simple empirical and mass balance models were found

that can be used for the EAASR. Comparable lake and reservoir systems were found to

provide parameter estimates for EAASR. Due to the wide range of possible operational

conditions and planning and design challenges, a water quality model developed

specifically for the EAASR was be necessary. The water quality model must therefore

incorporate measures to meet the planning and design challenges of the EAASR project.

Chapter 3 presents several iterations of EAASR configurations that have been

developed throughout the project. Early configurations used a combination of

Components A, B, and C of Figure 2-2. Configurations were then formulated using only

Component A. The sizing of the reservoir was evaluated first. Three configurations were

developed to evaluate the volume of the reservoir. Four additional alternatives were

developed to evaluate the depth of the reservoir. A 12 foot deep, 360,000 acre-foot

reservoir was subsequently decided upon. Three configurations were then developed to

assess the compartmentalization of the reservoir. The configurations were developed for a

single compartment, two compartments, and four compartments. The location of the

compartments affected the source of water received for both the EAASR and STA 3/4.

Unlike early configurations, internal transfers between the compartments occurred from

multiple compartments.

The review of the EAASR configurations was performed to provide the context in

which the water quality model is developed. To use these configurations, a water quality

model should include depth, area, and compartmentalization with varying area. Due to

the inter-reservoir transfers, the model must be able to simulate reservoirs in series, with









feedback loops. Additionally, the ability to include multiple inflows and outflows would

be useful.

Chapter 4 details the water quality model, the UF WQDT, and water quantity and

quality inputs. The UF WQDT is a steady-state mass balance model utilizing the KC*

kinetic model and was developed to simulate the TP water quality of the EAASR and

STA 3/4. The model uses a single aggregate inflow and concentration, depth,

background concentration, and a volumetric reaction rate to simulate TP. The model

calculates the HRT and HLR of each reactor, as well as the outflow concentration and

percent removal. Long-term average simulations were determined to be the most

appropriate level of sophistication needed in order to provide key insights for the

dynamic EAASR design.

Water quantity and quality inputs to the model were calculated for the base and

STA scenarios for each configuration. The HRT of the EAASR ranged from 39 to 107.5

days. The EAASR was not used for flow equalization. The HRT of STA 3/4 was

calculated to be 29 days for all scenarios. Mean inflow concentrations of 17 parameters

of interest for the four external inflow sources to the reservoir and STA 3/4 were

reported. Parameter estimations were performed using closely comparable system with

high quality data. Lake Istokpoga was found to be the best comparable for the general

single compartment scenario of the EAASR. Data from 16 STAs were used as

comparable to STA 3/4. The following water quality parameters estimates for the

EAASR and STA 3/4 were selected:

* C* = 0.0254 mg/L for the reservoir and 0.0194 mg/L for the STA
* k = 0.016 day'1 for the reservoir and 0.127 day'1 for the STA




Full Text

PAGE 1

WATER QUANTITY AND QUALITY IMPACTS OF THE PROPOSED EVERGLADES AGRICULTURAL AREA STORAGE RESERVOIR: PHASE 1 By DANIEL L. REISINGER A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Daniel L. Reisinger

PAGE 3

iii ACKNOWLEDGMENTS I would like to thank my family for their support and Dr. James P. Heaney for his invaluable guidance and contributions to this thesis. Additionally, Mr. Scott Knight worked on an earlier version of this analys is. Dr. Joong G. Lee and Dr. Robert Knight provided valuable review and suggestions. I would like to acknow ledge the cooperation from Mr. Cary White and Dr. Jaime Grau lau-Santiago in prep aring various SFWMM scenarios under tight deadlines. I woul d like to thank my supervisory committee members (Dr. John J. Sansalone and Dr. Robert L. Knight) for their valuable review of this thesis.

PAGE 4

iv TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... ..x CHAPTER 1 INTRODUCTION........................................................................................................1 2 BACKGROUND AND PREVIOUS WORK...............................................................3 The Comprehensive Everglades Restoration Plan........................................................3 Everglades Agricultural Area Storage Reservoir.........................................................4 Introduction and Previous Work...........................................................................4 EAASR Planning and Design Challenges.............................................................6 Engineering Design and the Systems Engineering Approach......................................8 Reservoir Modeling....................................................................................................11 Water Quality Models.........................................................................................11 Reservoir and Lake Mode ling in South Florida..................................................12 Summary and Conclusions.........................................................................................13 3 DEVELOPMENT OF EAASR CONC EPTUAL DESIGN ALTERNATIVES........15 Sizing Configurations.................................................................................................15 Compartmentalization Configurations........................................................................17 Single Compartment EAASR..............................................................................19 Two Compartment EAASR.................................................................................20 Four Compartment EAASR................................................................................22 Summary and Conclusion...........................................................................................24 4 SIMULATION MODEL FORMULATION AND INPUT........................................26 University of Florida Water Quality Design Tool......................................................26 Mass Balance.......................................................................................................26 Water Quantity Characterization.........................................................................27

PAGE 5

v Water Quality Characterization...........................................................................28 Inflow concentration....................................................................................28 Removal rate................................................................................................28 Model Rationale..................................................................................................29 Spreadsheet Model..............................................................................................30 Water Quantity of EAASR and STA 3/4....................................................................32 Water Quantity Data Source and Information.....................................................32 Water Balance of the FC Base EAASR and STA 3/4.........................................33 Single Compartment Reservoir Configuration Scenarios...................................34 Two Compartment Reservoir Configuration Scenarios......................................37 Four Compartment Configuration Scenarios.......................................................41 Flow Equalization................................................................................................44 Comparison of STAs and Lakes in Southern Florida..........................................45 Water Quality for the EAASR and STA 3/4..............................................................48 Inflow Water Quality...........................................................................................48 Background Concentration..................................................................................51 Reservoir TP Overall Reaction Rate...................................................................51 Long-term average evaluation of comparable systems.......................................52 More detailed analysis of selected comparables..........................................53 Selection of the appropriate period of record for the chosen comparable system.....................................................................................................54 Rate constant calibration..............................................................................58 EAASR reaction rate sensitivity analysis.....................................................61 STA TP Reaction Rate........................................................................................65 Predicted Performance of EAASR Removal of TP.............................................66 Predicted Performance of STA 3/4 Removal of TP............................................67 Summary and Conclusions.........................................................................................70 5 ESTIMATED WATER QUALITY CHANGE S IN THE EAASR AND STA 3/4 SYSTEMS..................................................................................................................72 Analysis of Long-Term Averages..............................................................................72 Single Compartment Reservoir...........................................................................72 Two Compartment Reservoir..............................................................................74 Four Compartment Reservoir..............................................................................76 Annual Variability in Performa nce for the SC STA Scenario....................................77 Summary and Conclusions.........................................................................................84 6 SUMMARY AND CONCLUSIONS.........................................................................87 LIST OF REFERENCES...................................................................................................93 BIOGRAPHICAL SKETCH.............................................................................................97

PAGE 6

vi LIST OF TABLES Table page 3-1 Conceptual Configuration of Rese rvoir for Original Alternative 3.........................17 4-1 Thirty Six Year Water Balance for EAASR and STA 3/4.......................................34 4-2 Compartment A Water Quantity for the Four Compartment EAASR Configuration...........................................................................................................42 4-3 Compartment B Water Quantity for the Four Compartment EAASR Configuration...........................................................................................................42 4-4 Compartment C Water Quantity for the Four Compartment EAASR Configuration...........................................................................................................42 4-5 Compartment D Water Quantity for the Four Compartment EAASR Configuration...........................................................................................................42 4-6 Mean, Standard Deviation, and C OV of Single Compartment Scenarios...............45 4-7 Comparative HLRs and HRTs for 37 R eactors in Southeast Florida (Data from Walker and Kadlec 2005b).......................................................................................47 4-8 NNR Basin Mean, Coefficient of Va riation, and Count of Water Quality Parameters................................................................................................................49 4-9 LOKNNR Canal Mean, Coefficient of Vari ation, and Count of Water Quality Parameters................................................................................................................49 4-10 Miami Basin Mean, Coefficient of Variation, and Count of Water Quality Parameters................................................................................................................50 4-11 LOKMiami Mean, Coefficient of Variation, and Count of Water Quality Parameters................................................................................................................50 4-12 Decision Rankings for Comparable Systems Ranked Five and Better....................53 4-13 Selected Results of the EAASR a nd STA 3/4 for Varying Reaction Rates.............64 4-14 Characteristics of 16 STAs in South east Florida (Data from Walker and Kadlec 2005a).......................................................................................................................66

PAGE 7

vii 4-15 Effect of Initial TP Concentration on TP at 50 Days and Percent Control for the EAASR.....................................................................................................................67 4-16 Effect of Initial TP Concentration on TP at 15 Days and Percent Control for the STAs.........................................................................................................................70 5-1 Single Compartment EAASR Results for Total Phosphorus with Variable Depths.......................................................................................................................73 5-2 STA 3/4 Results for the Single Compar tment EAASR for Total Phosphorus with Variable Depths........................................................................................................73 5-3 Compartment 1 Result for the Two Compartment EAASR Configuration.............75 5-4 Compartment 2 Result for the Two Compartment EAASR Configuration.............76 5-5 STA 3/4 Result for the Two Co mpartment EAASR Configuration.........................76 5-6 TP Removal by Mass in Kilograms per Day............................................................76 5-7 Compartment A Result for the Four Compartment EAASR Configuration............76 5-8 Compartment B Result for the Four Compartment EAASR Configuration............77 5-9 Compartment C Result for the Four Compartment EAASR Configuration............77 5-10 Compartment D Result for the Four Compartment EAASR Configuration............77 5-11 STA 3/4 Result for the Four Compartment EAASR Configuration........................77 5-12 TP Removal by EAASR and STA 3/ 4 in Mass in Kilograms per Day....................77 5-13 Annual EAASR Variability for the Single Compartment STA Scenario................79 5-14 Annual STA 3/4 Variability for th e Single Compartment STA Scenario................80 5-15 Summary Results for the One, Two, and Four Compartment Scenarios.................85

PAGE 8

viii LIST OF FIGURES Figure page 2-1 Accelerated CERP Projects (Central and Southern Florida Project 2005, Page 32)............................................................................................................................ ...4 2-2 EAASR and Stormwater Treatment Area 3/4 Vicinity Map (USACE and SFWMD 2002, Page 7)..............................................................................................6 3-1 Conceptual Configuration of Reservoi r for original Alternative 3 (USACE and SFWMD, electronic correspondence, April 5, 2005)...............................................16 3-2 General Layout of EAASR and STA 3/4.................................................................18 3-3 Single Compartment EAASR and STA 3/4 Flow Diagram.....................................20 3-4 Two compartment EAASR and STA 3/4 Configurations........................................21 3-5 Two Compartment EAASR Co nfiguration Flow Diagram......................................22 3-6 Four Compartment EAASR and STA 3/4 Configurations.......................................23 3-7 Four Compartment EAASR and STA 3/4 Flow Diagram........................................24 4-1 Block Mass-Flow Diagram of Modeled EAASR.....................................................27 4-2 Spreadsheet Inte rface for UF WQDT.......................................................................31 4-3 Single Compartment Base Scenario Flow Diagram.................................................35 4-4 Single Compartment STA Scenario Flow Diagram.................................................36 4-5 Two Compartment Base Scenario Flow Diagram....................................................38 4-6 Two Compartment STA Scenario Flow Diagram....................................................39 4-7 Two Compartment Miami Scenario Flow Diagram.................................................40 4-8 Two Compartment Miami ST A Scenario Flow Diagram........................................41 4-9 Four Compartment Base Scenario Flow Diagram...................................................43

PAGE 9

ix 4-10 Four Compartment STA Scenario Flow Diagram....................................................44 4-11 Lake Istokpoga and Mass Balance Locations..........................................................56 4-12 S-68 TP Data and Adjusted POR.............................................................................57 4-13 Arbuckle Creek TP Data and Adjusted POR...........................................................57 4-14 Water Quality Sampling St ations in Lake Istokpoga...............................................59 4-15 Median TP Values at La ke Istokpoga Sampling Stations........................................60 4-16 Lake Istokpoga TP Removal in Arbuckle Creek Flow Path....................................61 4-17 Sensitivity Analysis for EAASR Parameters k and C*............................................62 4-18 Outflow Concentration of the EAASR and STA 3/4 for Varying Reaction Rates..63 4-19 Water Balance and Location of the EAASR – Phase 1 (SFWMD 2006)................65 4-20 Effect of Initial Concentration and Residence Time on Outflow TP Concentration for the EAASR..................................................................................67 4-21 Effect of Initial Concentration and Residence Time on Outflow TP Concentration for the STAs......................................................................................69 4-22 Figure 4-21 Rescaled to Re sidence Times Up to 50 Days.......................................69 5-1 Mean Annual Inflows from SC STA Scenario.........................................................81 5-2 Mean Annual Depths from SC STA Scenario.........................................................82 5-3 Mean Annual HRT from SC STA Scenario.............................................................83 5-4 Mean Annual Inflow and Outflow C oncentrations from SC STA Scenario............84

PAGE 10

x Abstract of Thesis Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering WATER QUANTITY AND QUALITY IMPACTS OF THE PROPOSED EVERGLADES AGRICULTURAL AREA STORAGE RESERVOIR: PHASE 1 By Daniel L. Reisinger May 2006 Chair: James P. Heaney Major Department: Environmental Engineering Sciences A model was developed to estimate wate r quality in the Everglades Agricultural Area Storage Reservoir (EAASR) and Stormwat er Treatment Area (STA) 3/4 as part of the dynamic and challenging design process asso ciated with the Everglades Restoration. The University of Florida Water Quality Design Tool is a steady-state mass balance model using the KC* model to simulate the total phosphorus (TP) c oncentrations of the outflows from the proposed EAASR and dow nstream STA 3/4. Performance under longterm average conditions was determined to be the most appropriate level of sophistication needed to provide key insights for the rapidly evolving EAASR designs. Water quantity boundary conditions were provided from output from the South Florida Water Management Model, a complex regional model that simulates daily flows and stages over a 36 year period. Water quantity and quality inputs to the m odel were calculated for the base and STA scenarios for each configuration. The hydraulic residence time (HRT) of the

PAGE 11

xi EAASR ranged from 39 to 107.5 days for vari ous scenarios. The HRT of STA 3/4 was calculated to be 29 days for al l scenarios. Inflow TP concentrations varied from 0.067 mg/L to 0.133 mg/L depending on the reservoi r configuration. Lake Istokpoga was found to be the best comparable for parameter estimation. Data from 16 STAs were used as comparables to STA 3/4. The background TP concentration of the EAASR was determined to be 0.025 mg/L and 0.019 mg/L fo r the STA. The reac tion rate calibration determined a rate of 0.016 per day of TP fo r the reservoir and 0.127 per day for the STA. Output concentrations from the EAASR shared a significant treatment effect: concentrations reduced to 0.028 to 0.068 mg/L depending on the configuration. Variability in performance was greatly da mpened in the STA 3/4 outflow with the outflow concentrations all in the 0.0200.021 mg/L range. Results showed that compartmentalization of the reservoir and in cluding an additional outflow structure from Compartment 2 to STA 3/4 in the two compar tment configuration can provide additional operational flexibility and increase water qu ality improvements. Results suggest that simulated inflows to STA 3/4 could be increas ed to improve water quality even more in the EAASR/STA 3/4 storage/treatment train. The reservoir in its current simulation is not used for flow equalization. Actual performance of the EAASR/STA system could vary widely from our predictions for several reasons: Inflow quantities and water quality can be expected to vary over the next 50 years The SFWMM estimates of inflows and ope rations are not based on any kind of optimization for the EAASR/STA system but re present an estimate of their role in a regional water management scenario. Behavior and performance of the EAASR/ STA can be expected to vary depending on how it is operated for flood control and water supply purposes as well as water quality enhancement.

PAGE 12

1 CHAPTER 1 INTRODUCTION The Everglades Agricultural Area Storag e Reservoir (EAASR) is jointly operated for water supply, flood control, and to suppor t water quality management. The multiple and sometimes competitive purposes of the reservoir provide significant design and operation challenges. Methods and guidelines for the design of reservoirs and levees with respect to water supply and flood cont rol are well established. However, methods for incorporating water quality are less developed. The United States Army Corps of Engin eers (USACE) tasked the University of Florida, as a subcontractor to Water and Air Research Inc. (WAR), to develop and use a model for estimating of reservoir water qual ity. This model and prediction of water quality is to serve as one sect ion of a larger report by WAR, who was tasked to create an Environmental Impact Statement for the EAASR. The EAASR is one of the first and potent ially the largest single Comprehensive Everglades Restoration Plan component to be included in the Ac celer8 program. The Acceler8 program allows a dual-track pro cess of design. The South Florida Water Management District (SFWMD) completes de tailed design in parallel with the USACE meeting their prescribed requirements (S outh Florida Water Management District, SFWMD 2006). As a result, challenges ha ve arisen throughout the project. The Acceler8 design evolved quickly, preempting USACE alternatives, and resulting in an iterative design formulation process that included multiple simulation and analysis efforts (Knight et al. 2006). Different modeling assu mptions (between the project sponsors and

PAGE 13

2 updates to the simulation model) used to pr edict the water quantity created uncertainty when comparing current and previous result s. To date, no Comprehensive Everglades Restoration Plan reservoir has been constructed. Therefore comparable existing lake or reservoir systems must be used to predict the performance of the proposed reservoirs. Choosing among these comparable systems is a significant challenge because of the wide range of operations. Our aim was to create and use a model to estimate water quality in the EAASR and Stormwater Treatment Area (STA) 3/4. The model needed to be able to function in the dynamic and challenging EAASR design proce ss. We hoped to gain key insights on the most current EAASR and STA 3/4 configurations.

PAGE 14

3 CHAPTER 2 BACKGROUND AND PREVIOUS WORK The Comprehensive Everglades Restoration Plan The Comprehensive Everglades Restorat ion Plan (CERP) will deploy 63 water resource projects to help restore the Ev erglades, and provide water supply and flood control. CERP creates a pa rtnership of USACE (the fe deral sponsor) and the SFWMD (the local sponsor) to restore, protect and preserve the Sout h Florida ecosystem. The plan calls for cost-sharing between the state and fede ral sponsors. It will take approximately 30 years to implement all of the proposed CE RP projects. The projects are focusing on getting the water right, which includes the quality, quantity, timing, and distribution of flows. This goal will be accomplished by all 63 projects working in unison (Central and Southern Florida Project 2005). Water quality benefits are an important part of CERP projects; however they are not the primary purpose of such projects. Eight CERP projects where chosen by the SFWMD, including the Everglades Agricultural Area Storage Reservoir (E AASR), for accelerated funding, design and construction. This streamlined plan was termed Acceler8. Acceler8 allows a dual track process of design, where the SFWMD complete s detailed design in parallel with the USACE meeting their prescribed requirement s. The accelerated schedule allows the benefits to the system to be realized s ooner and more cost-effectively (Figure 2-1) (SFWMD 2006a). The Project Implementati on Report (PIR) details the reconnaissance, feasibility analysis, and selection of the preferred conceptual design.

PAGE 15

4 Figure 2-1. Accelerated CERP Projects (Centr al and Southern Flor ida Project 2005, Page 32) Everglades Agricultural Area Storage Reservoir Introduction and Previous Work CERP uses extensive water storage syst ems, including surface and sub-surface reservoirs and aquifer storage and recovery (Central and South Fl orida Project 2006). These storage systems will be jointly operate d for water supply, flood control, and water quality management. The multiple and sometimes competitive purposes of the CERP reservoirs provide significant design and ope ration challenges. We ll-established methods and guidelines for the design of reservoirs a nd levees with respect to water supply and flood control purposes have been developed. However, methods for incorporating water quality are less developed. The EAASR, a CERP project, aims to crea te a reservoir on land in the southern portion of the Everglades Agricultural Area (EAA) (Figure 2-2). The main goals and objectives of the EAASR Phase 1 project as st ated in the 2002 Project Management Plan

PAGE 16

5 (United States Army Corps of Engineers a nd South Florida Water Management District, USACE and SFWMD 2002) are as follows: Reduction of Lake Okeechobee regulatory rel eases to the estuar ies and reduction of backpumping from the EAA into Lake Okeechobee by sending the water to the south and into the reservoirs. Improved environmental releases through th e storage of water and release to the Everglades during the dry season demand. Flow equalization and optimization of treatment performance of Stormwater Treatment Area (STA)-2, STA-3/4, STA5, and STA-6 by capturing peak stormevent discharges within th e reservoirs for subseque nt release to the STAs. Improved flood control and regional water supply for the agricultural community currently served by the EAA canals and other areas served by Lake Okeechobee. The USACE and SFWMD developed the 2002 Project Management Plan for the EAASR. Details on the background, purpose, scope, and initial planning of the EAASR are included in this document. The C onceptual Alternative report by USACE and SFWMD (2004a) presents 22 alternative design s for the EAASR, using any combination of Components A, B, and C in Figure 2-2. The area, depth, volume, water control structures, and additional features of each al ternative are detailed. The Screening of Conceptual Alternatives report also by USACE and SFWMD (2004b) details the screening criteria and decision matrix for th e initial conceptual alternatives. The screening results found five alternatives for fu rther analyses. These analyses led to the development of conceptual alternatives utilizing only Component A, which where documented in the USACE and SFWMD (200 5) Integrated Project Implementation Report and Environmental Impact Assessment (P IR/EIS). The PIR/EIS selects a single alternative, the tenta tively selected plan (TSP), and provi des detailed design information. An earlier version of the result s and discussion presented in th is thesis are also included in Appendix F of the PIR/EIS. SFWMD pr oduced the Basis of Design Report (BODR)

PAGE 17

6 (2006b) for the first phase of EAASR. The BODR provides greater design detail for phase 1 and an alternative water quality analysis. Figure 2-2. EAASR and Stormw ater Treatment Area 3/4 Vicinity Map (USACE and SFWMD 2002, Page 7) EAASR Planning and Design Challenges The planning and design of the EAASR is one of the first and potentially the largest single CERP component to be accelerated. As a result, challenges have arisen

PAGE 18

7 throughout the project. Knight et al. (2006) reviewed th e EAASR planning process in order to establish a cohesive approach to the design. They focus on the PIR/EIS where “the parallel process leads to the USACE examining and evaluating multiple alternatives while the SFWMD is beginning detailed design on a specific alternat ive” (Knight et al. 2006, Page 7). The USACE is attempting to in corporate the Acceler8 alternative, while meeting statutory requirements to evaluate a wide range of alternatives. The matter was complicated by USACE analyzing the entire 360,000 acre-feet of storage, while Acceler8 planned only the first phase, a 190,000 acre-foot reservoir. The resulting challenge arose as the Acceler8 design evolved quickly, preemp ting USACE alternatives that were in the processes of analysis for the PIR/EIS. The PIR/EIS alternatives were reformulated to meet the new conditions and provide a range of viable alternatives. The result was an iterative screening process for alternatives that included multiple simulation and analysis efforts. The iterative nature of the pro cess is documented in the discussion EAASR configurations provided in Chapter 3. Alternatives were simulated using the South Florida Management Model Version 5.4 (SFWMM). The SFWMM is a regional scale model used to simulate the hydrology and water management of the SFWMD area fr om Lake Okeechobee to Florida Bay. This 7,600 square mile area was partitioned into 1,900 two mile by two mile squares (2,560 acres) called cells in which surface water, groundwater, and their interactions are modeled. The model simulates the daily m ovement of water through the study area for 36 years from January 1, 1965 to December 31, 2000. The SFWMM is accepted as the best tool to model this area, because it in corporates the complex operating rules that govern the behavior of the system (SFWMD 200 5a). Thus, a major simulation effort was

PAGE 19

8 required to run 36 years of daily activity for 1,900 grid cells in or der to evaluate one project with 10 to 23 grid cells. As a region al model, changes to any component in the model could affect the behavior of the E AASR. Throughout the pl anning process, the model was continually updated, which improve d the reliability of the results. USACE and Acceler8 alternatives were simulated on different networks and using different assumptions. For example, Acceler8 evaluated water quantity for the 2010 and 2015 land use projections, while USACE used 2050 land use projections (SFWMD 2006 and USACE and SFWMD 2005). These factors created uncertainty when comparing current and previous results, as well as results be tween agencies. The SFWMM can only be run by the sponsoring agencies and was not availabl e for direct use by the study team. Thus, each run of the SFWMM required a requisi tion to the sponsoring agencies. The EAASR may be actively operated in a wide range of hydraulic and hydrologic conditions, which aff ects the water quality of th e reservoir. Empirical parameter estimates are required as part of the modeling exercise. No CERP reservoir has been constructed; therefore comparable existing lake and/or reservoir systems must be used to predict the likely performance of the proposed rese rvoirs. The choice of these comparable systems poses a significant challeng e, due to the wide ra nge of attributes of these lakes and reservoirs. Engineering Design and the Systems Engineering Approach Planning and design approaches exist to he lp meet the described challenges. This section provides an introducti on to a systems engineering ap proach to engineering design. According to Hazelrigg (1996), engineering de sign is a decision-making process. He defines a decision as an irrevo cable allocation of resources; hence the selection of design parameters by an engineer is a decision-maki ng process. This outlook is a departure from

PAGE 20

9 traditional engineering design that is largely view ed as an exercise in problem solving. It creates the distinction that engineering is aimed at creat ing information, which is related to a specific decision, instead of knowledge, which is a set of agreed upon facts. Considering design in this manner is comm only called systems engineering (Hazelrigg 1996). To develop a systems engineering appr oach independent of domain, Braha and Maimon (1998) performed an exte nsive literature re view that found engineering design to share the following common properties: Design begins with an acknowledgement of an unmet need and a call for action to meet this need. Designing an artifact is used to transiti on from concepts and ideas to concrete descriptions. The designer is constantly faced with th e problem of bounded rationality, i.e., the designer has limitations on his cognitive a nd information processing capabilities. The design specifications tend to evolve as part of the design process. Traditional engineering design methods te nd to rely on satisfying rather than finding the true optimal solution. Alternatives and design solutions evolve as part of the design process. Braha and Maimon (1998) view design as a sequential process with feedback. This process goes from general concept to preliminary and detailed design, production planning, production, operation, and final dispos al. Hazelrigg (1996, Page 8) viewed the design process as three distinct activities: the iden tification of options, development of expectations on outcomes for each option, a nd use of values to select the option that has the range of outcomes and associat ed probabilities that are most desired. A key concept of HazelriggÂ’s view is the need to produce information that provides a prediction of the accuracy or reliabil ity of the design. Uncertainty is therefore

PAGE 21

10 intrinsic in information. Such uncertainty is commonly accounted for by conservative parameter estimations, factors of safety, or statistical design. The general formulation of the engineering design problems and therefore models consists of two main parts; the objection function and constraint s (Heaney, unpublished manuscript, 2006). Heaney (unpublished manus cript, 2006) defines the parts, where decision variables are one-time parameter decisions and/or operating rules, as: Objective function: Maximize or minimize some stated objective(s) by selecting the best values of the decision variables. Constraints: Physical, chemical, and/or biologi cal process relationships and/or operational and regulatory cons traints on the variables. Traditionally design relies on constraining the system to separate the design into manageable and domain specific parts. Tr aditional design leads to a reductionism approach, where the outcome of the design is a function of the constraints. The systems engineering approach divides th e design into disciplinary m odels, which are incorporated into a whole system model. This leads to a design that is less fo cused on constraints and may produce more optimal designs (Hazelrigg 1996). Lee et al. (2005) document an approach to optimize the design urban stormwater storage-release systems. Urban approaches ma y also be applied to reservoir modeling as they are fundamentally both storag e-release systems. Lee et al approach uses cost as the objective for a design based on continuous si mulation of water quantity and quality. Spreadsheets are utilized to link powerful optimization tools to transparent process models. Unlike traditional approaches, design may be optimized for 3 or more parameters.

PAGE 22

11 Reservoir Modeling Mathematical models are relied upon heavily for engineering design. They provide a prediction of the systems behavi or for a given set of design or decision variables. These models strive to answer what if questions of the designer and if optimized can answer the question of what is best Models are used to answer three fundamentally different types of questions (Hazelrigg 1996): Will the system work as designed? Which of the system alternatives are better? Do I properly understand the system? To develop a model for the EAASR that focuses on the second question, which of the system alternatives is better, a review of water quality models was performed. The review included reservoir and lake models that have been developed and calibrated for the South Florida region. Results of the review are documented next. Water Quality Models The relatively shallow, actively controlled EAASR is more comparable to a lake or shallow reservoir system than a traditiona l reservoir. The gene ral framework for both water quantity and quality models for lake or shallow reservoirs are well developed (Chapra 1996). Chapra and Auer (1999) re viewed management models to evaluate phosphorus loads in lakes. They classifi ed phosphorus models in three general categories; empirical models, simple budget models, and nutrient food-web models. Empirical models can be divided into phosphorus loading plots and trophic parameter correlations. Phosphor us loading plots are used to estimate the trophic level of the lake, or when linked to a simple bala nce model can predict in-lake total phosphorus concentrations. Trophic parameter correlations normally relate two trophic parameters or can be used in tandem with phosphorus loading plots. The main advantage of empirical

PAGE 23

12 models is the ease of use. These models are most accurate when calibrated for an individual lake. Simple budget models focus on the mass-ba lance of a lake system. The most basic of these models uses inflow, outflow and one-way removal to characterize the system. These models provide a temporal re sponse from the lake and can be easily adapted for different lake dynamics. Simple budget models are highly sensitive to the quality of input data. Theref ore, they are best suited for long term trends or high quality data sets. Nutrient and food web models are more co mplex models that aim to characterize the temporal and physical aspects of matter throughout a lake or reservoir (Chapra and Auer 1999, Page ?). These models require la rge amounts of data or assumptions and can provide a more detailed analys is than previous models. Reservoir and Lake Modeling in South Florida Empirical and simple balance models have been used in modeling CERP reservoirs and the EAASR. USACE a nd SFWMD (2003) reviewed 16 water quality models that predicted the upt ake of phosphorus in lakes and reservoirs. DMSTA was at the top of the reviewÂ’s shortl ist of models that met the reviewed criteria. DMSTA 2 (Walker and Havens 2005a), an improved version of DMSTA, is a mass balance model that is calibrated for CERP reservoirs. The model incorporates a first order kinetic model with a background concentration or a water co lumn and sediment transfer model. Wetland Solutions Inc. (WSI) developed water quality models to simulate 15 parameters of interest for a general CE RP reservoir (WSI 2004). Each model was calibrated for Florida lakes and reservoirs a nd can be applied in a spreadsheet. The models include Eutromod (Reckhow 1979; Reckhow et al. 1992), the Vollenweider

PAGE 24

13 Eutrophication Model (Vollenweider 1969, Kane 1999), and the U.S. Corps of Engineers Bathtub Model (Walker 2004). Additionally, WSI developed three regression models from Burns and McDonnell (2004b) based on co mparable lakes and reservoirs. An earlier version of the University of Florida Water Quality Design Tool, which assesses TP uptake in the EAASR, is presented in the PIR/EIS (USACE and SFWMD 2005). As stated previously, comparable lakes and reservoirs are necessary to estimate modeling parameters. Central and South Fl orida lakes and reservoirs have been relatively well studied. The restoration of th e Everglades has driven recent area wide studies and model calibration efforts, such as Burns and McDonnell (2004a) and Walker and Kadlec (2006). Burns and McDonnell ( 2004a) completed a four-part project to create and analyze a database from lakes and reservoirs that were comparable to CERP reservoirs. They identified and acquired da ta for 36 potential comparable lake systems across Florida, which were refined to eight comparable lakes and a reservoir with sufficient data (Burns &McDonnell 2004). Walker and Kadlec (2006) developed an extensive database of lakes for their calibration effort for DMSTA 2. The DMSTA 2 calibration uses 19 lakes a nd reservoirs from Walker and Havens (2003), Wetland Solutions, Inc. (WSI) (2003), Walker (2000), and Burns and McDonnell (2004a) efforts. Summary and Conclusions CERP is an ambitious project to “get the water right” in the South Florida Ecosystem and restore the Everglades. The project goal will be accomplished by all 63 projects operating together. The Acceler8 project streamlines the CERP planning, design, and construction process to provide benefits to the sy stem more quickly and costefficiently. The EAASR, a CE RP and Acceler8 project, will provide storage for water supply, flood control, and flow equalization fo r water quality treatment areas. Water

PAGE 25

14 supply and flood control design approaches are well developed, while water quality approaches are less developed. The stream lining of the planning and design of the EAASR has posed several challenges resulti ng in an iterative design process with multiple conceptual alternative formulations and analyses. Due to differences in the simulation, uncertainty exists when these multiple formulations and analyses are compared. The systems engineering approach to desi gn provides a proven approach that can meet the challenges of the EAASR planni ng and design process. The approach incorporates disciplinary mode ls that may produce a more op timal design than traditional approaches. A model in the water quality di scipline was therefore sought to assess which of several alternatives are better. Simple empirical and mass balance models were found that can be used for the EAASR. Comparable lake and reservoir systems were used to provide parameter estimates for EAASR. Due to the wide range of possible operational conditions and planning and design challe nges, a water quality model developed specifically for the EAASR will be necessary. The model must incorporate measures to meet the planning and design challenges of the EAASR. The iterative multiple conceptual design alternatives for the E AASR were reviewed. The model was then formulated in the context of this chapte r and the conceptual design alternatives.

PAGE 26

15 CHAPTER 3 DEVELOPMENT OF EAASR CONC EPTUAL DESIGN ALTERNATIVES The hydrology of EAASR and STA 3/4 were simulated using the SFWMM Version 5.4. The complete water quantity dataset used in water quality modeling was developed from the simulations. Therefore, the EAASR conceptual design alternatives are presented in the contextual framework of the SFWMM. The physical layout of the reservoir is termed a configuration. The co mbination of a configur ation and particular flow dataset is referred to as a scenario. All configurations simulated the levee walls as vertical and of insignificant area to affect the stage-area-volume relationship. The reservoir levees were also assumed to a llow no seepage, though groundwater flow was included. The selection of the confi guration of the EAASR was an iterative process. The sizing of the EAASR, characterized by the ar ea and depth, was first determined. For the chosen size, configurations to evaluate the compartmentalization of the EAASR were then developed. These configurat ions are used in the water qua lity analysis of this report and are described in full detail. Sizing Configurations Early configurations used a combinati on of Components A, B, and C of Figure 22 (USACE and SFWMD 2004a). A decision was made to reserve Components B and C to expand the capacity of adjacent Stormwater Treatment Areas. Alternatives were then developed using only the Co mponent A location.

PAGE 27

16 Three alternative reserv oir sizes with storage ca pacities of 240,000, 360,000, and 480,000 acre-feet were evaluated in the early stages of Compone nt A analysis, referred to as the original Alternative 1, the original Alternative 3, and the original Alternative 5, respectively (USACE and SFWMD, elec tronic correspondence, February 14, 2005, March 1, 2005, March 8, 2005, and April 5, 2005) The original alternatives were divided into two compartments of the same no minal depth. In the original alternatives, the first compartment, called C1, was confi gured to provide 90,000 acre-feet of storage for agricultural use. The second compartm ent, C2, varied in volume from 150,000 acrefeet to 390,000 acre-feet and was for environmen tal use. C1 was able to overflow to C2 (USACE and SFWMD, electroni c correspondence, April 5, 2005). A conceptual view of the two compartment reservoir for the original Alternative 3 is provided in Figure 3-1. Subsequently, a decision was made that the total capacity of EAASR would be 360,000 acre-feet. EARMA1 EARIN1EARIN2 EARMA2LKRSM1EARNH2E A R S N OEARNH1LKRSN1 MC NNRC 90,000 ac-ft 270,000 ac-ft WCS4S + EVBLSN Figure 3-1. Conceptual Configur ation of Reservoir for orig inal Alternative 3 (USACE and SFWMD, electronic correspondence, April 5, 2005)

PAGE 28

17 Design depth was the next key sizing decision. Four design depths were considered with nominal design depths of 6, 10, 12, and 14 feet. These alternatives were named Alt1R, Alt2R, Alt3R, and Alt4R, respectively (USACE and SFWMD, electronic correspondence, April 8, 2005). The compartm ents in all Alt_R configurations were termed C1 and C2, as in previous configuratio ns. The depth and area used to achieve this total volume were varied in each alternative; however C2Â’s area remained fixed at 17,920 acres. The associated areas of C1 were developed to yield a capacity of 360,000 acrefeet. The area, depth, and volume of the Alt_ R configurations and the original Alt 3 simulation are presented in Table 3-1. Table 3-1. Conceptual Configuration of Reservoir for Original Alternative 3 C1 C2 Total Configuration Area (ac) Nominal Depth (ft) Volume (ac-ft) Area (ac) Nominal Depth (ft) Volume (ac-ft) Area (ac) Volume (ac-ft) Alt 3 Orig. 7,500 12 90,000 22,500 12 270,000 30,000 360,000 Alt1R 40,960 6 250,000 17,920 6 110,000 58,880 360,000 Alt2R 17,920 10 180,000 17,920 10 180,000 35,840 360,000 Alt3R 12,800 12 150,000 17,920 12 210,000 30,720 360,000 Alt4R 7,680 14 110,000 17,920 14 250,000 25,600 360,000 A design depth of 12 feet was select ed for the 360,000 acre-foot reservoir. Configurations to assess the effect of compartmentalizati on and the source of inflows were then developed. The configura tions are presented in detail next. Compartmentalization Configurations Unlike previous configurations, the compar tmentalization of the reservoir affects the inflow sources to STA 3/4. Therefore, details on both the EAASR and STA 3/4 are included for each configuration. The EAASR was modeled as 12 SFWMM cells incorporating 30,720 acres of area and STA 3/4 was modeled as 7 SFWMM cells incorporating 17,920 acres of area (USACE and SFWMD, electronic correspondence,

PAGE 29

18 February 2 and 3, 2006). The combined tota l area of the EAASR and STA 3/4 is 48,640 acres and the EAASR accounts for about 63% of this total area. Figure 3-2 displays the location of the EAASR and STA 3/4 in the SFWMM. The nominal depth of the EAASR was 12 feet for all configurations. The EAAS R was assumed to have a level pool and flat bottom at an average elevation and therefor e had a simple stage-area relationship. STA 3/4 was modeled with the same levee and pool assumptions as the reservoir. Figure 3-2. General Layout of EAASR and STA 3/4 Three following configurations of the EAASR were developed from the general layout presented in Figure 3-2: 1. Single Compartment 2. Two compartments 3. Four Compartments

PAGE 30

19 A single SFWMM flow dataset (USACE a nd SFWMD, electronic correspondence, February 2 and 3, 2006) was used in modeling all three configurations. However, the quantity and source of flows to each compar tment and STA 3/4 varied depending on the configuration. Each configuration will be desc ribed in detail in the remainder of this section. Single Compartment EAASR The single compartment EAASR configur ation was modeled as presented in Figure 3-2. The single compartment used all of the 30,720 acres available. All possible inflows sources to the EAASR and STA 3/ 4 are represented in the configuration. A flow diagram for the single compartment reservoir and STA 3/4 is presented in Figure 3-3. The EAASR and STA 3/4 receiv e inflows via the North New River (NNR) Canal and Miami Canal, which are shown on the east and west side of the EAASR in Figure 3-3. The EAASR receives inflows from four external sources: NNR basin runoff, Miami basin runoff, Lake Okeechobee (LOK) regulatory release through the NNR Canal, and LOK regulatory releases through the Mi ami Canal. Water is released to the agricultural basins or south to STA 3/4. STA 3/4 receives flow from NNR and Miami basin runoff, LOK through the NNR and Mi ami Canals, and the EAASR. The STA discharges to the water conservation Areas (W CA). Figure 3-3 and those like it represent the aggregation by source of SFWMM flow tags All flows are actively operated in the SFWMM and are constrained by flow cap acity and/or crest elevation.

PAGE 31

20 Figure 3-3. Single Compartment E AASR and STA 3/4 Flow Diagram Two Compartment EAASR In the second configuration, the EAASR is partitioned into two compartments (Figure 3-4). This configuration was consid ered to be comparable to the Tentatively Selected Plan (TSP) in the EAASR PIR/EI S (USACE and SFWMD 2005). The eastern compartment bordering the NNR Canal is referr ed to as Compartment 1 and the western compartment bordering the Miami Canal is referred to as Compartment 2. Each compartment was 15,360 acres and modele d as an independent level pool.

PAGE 32

21 Figure 3-4. Two compartment EAAS R and STA 3/4 Configurations A flow diagram for the two compartmen t EAASR configuration and STA 3/4 is presented in Figure 3-5. The compartmentali zation of the EAASR altered the ability of each compartment and the STA 3/4 to receive water from the sources presented in the single compartment configuration. Each comp artment received external flow from the respective adjacent canal and through inter-compa rtmental transfers. It is important to note that unlike early configurations water can be transferred internally between both compartments. Water was released to the agri cultural basins or s outh to STA 3/4 in the reservoir. Due to the location of the compartments, only Compartment 1 was able to release water to STA 3/4. STA 3/4 was not able to receive flow directly from the Miami Canal; therefore the Miami Canal flows were routed through the EAASR to the STA. The STA was able to receive inflow via NNR basin runoff, LOK through the NNR canal, and the EAASR. STA 3/4Â’s discharg e capabilities were not altered.

PAGE 33

22 Figure 3-5. Two Compartment EAAS R Configuration Flow Diagram Four Compartment EAASR The third configuration partitioned the E AASR into four compartments (Figure 3-6). This configuration was comparable to the Mixed/Segregated Plan (MSP) referred to in the January Work Tasks (USACE, el ectronic correspondence, January 26, 2006). The four compartments were labeled Compar tments A through D. Compartment A and C are 5,120 acres and Compartments B and D are 10,240 acres. Each compartment is modeled as an independent level pool. The four compartment flow diagram for the EAASR and STA 3/4 is presented in Figure 3-7. Each of the four compartments received flow from the adjacent canal and through inter-compartmental tr ansfers. It is important to note that unlike early configurations water can be transferred internally between multiple compartments. The

PAGE 34

23 modeling of the four compartment confi guration assumed that each compartment received flow from a single external source. Water was released to the agricultural basins through Compartments A or C. Due to the location of the compartments, only Compartments B and D were able to releas e water to STA 3/4. STA 3/4 was able to receive flow from both Miami and NNR Cana ls in this configuration. STA 3/4Â’s discharge capabilities were not altered. Figure 3-6. Four Compartment EAAS R and STA 3/4 Configurations

PAGE 35

24 Figure 3-7. Four Compartment EA ASR and STA 3/4 Flow Diagram Summary and Conclusion Several iterations of EAASR configur ations have been developed. Early configurations used a combination of Components A, B, and C of Figure 2-2. Configurations were then formulated usi ng only Component A. The sizing of the reservoir was evaluated first. Three confi gurations were developed to evaluate the volume of the reservoir. Four additional alternatives were developed to evaluate the depth of the reservoir. A 12 foot dee p, 360,000 acre-foot reservoir was subsequently selected. Three configurations were then developed to assess the compartmentalization of the reservoir. The configurations were developed for one, two, and four compartments. The location of the compartm ents affected the source of water received

PAGE 36

25 for both the EAASR and STA 3/4. Unlike ea rly configurations, internal transfers between the compartments occurr ed from multiple compartments. The review of EAASR configurations was performed to provide the context within which the water quality model is developed. To use these configurations, a water quality model should include depth, area, and compartmentalization with varying area. Due to the inter-reservoir transfers, the mode l must be able to simulate reservoirs in series, with feedback loops. Additionally, th e ability to include multiple inflows and outflows would be useful. The formulation and rationale of the water quality model is described in detail in the ne xt section. The water quantity and quality values for the compartmentalization configurations, with an emphasis of modeling parameters, are detailed as well.

PAGE 37

26 CHAPTER 4 SIMULATION MODEL FORMULATION AND INPUT University of Florida Wat er Quality Design Tool The EAASR is classified as a reservoir. Comparable systems in the area are called by a variety of names including: Lakes Reservoirs Emergent wetlands Pre-existing wetlands Submerged aquatic vegetation systems Nominally, reservoirs and lakes provide storage while wetlands and STAs provide treatment. However, all of these systems can be viewed more generically as reactors that in fact provide a blend of these functi ons depending on how they are designed and operated. Accordingly, the word reactor will be used to describe the general modeling approach for these systems. The EAASR/STA 3/4 system is viewed as a treatment train with two reactors in series. EAASR and ST A 3/4 simulations therefore incorporate both water quantity (storage) and quality. The wa ter quality modeling of the EAASR focuses on Total Phosphorus (TP), the main water quality parameter of concern. Mass Balance A mass balance around the reacto r is created to assure c onservation of mass. For the reactor, the external sources of water a nd pollutants are the infl ow and precipitation. The pollutant removal in the reactor is c onsidered a final mass sink. The remaining pollutants exit the reactor in the outflow.

PAGE 38

27 In steady-state mass balance analysis, the reservoir can be modeled as having inflows from the canals, Qin, and outflow, Qout, which are aggregates of the daily inflows, and outflows, and precipitation, P. A parame ter of concern is assumed to enter at a constant concentration for each inflow, Cin, and from precipitation, CP. The pollutant is removed as a function of detention time, td, initial concentration, Cin, and reaction rate, kV. The pollutant exits the reactor to th e surrounding system in the outflow at a calculated concentration, Cout. The conceptual view of th e mass balance (Figure 4-1) and mathematical equation (Eqn. 4-1) also uses the concentration in each plug of water in the reactor (C), the depth (D ), and the area (A). Reactor Qin Cin Qout Cout V C kV D A P C P Figure 4-1. Block Mass-Flow Diagram of Modeled EAASR p out out V in inC P C Q C k V C Q (4-1) Water Quantity Characterization The water quantity characterization of th e model is based on a volume balance of the system. The volume balance of the system is represented by th e total inflows minus the total outflows equaling the change in storage ( S/ t ). From the reservoir alternatives data provided by the IMC, the total inflows a nd outflows of the system are represented as rainfall ( P ), inflows ( Qin), groundwater inflows and outflow ( GWI and GWO), ET and outflows ( Qout) (Eqn. 4-2). The change in storage per unit time, S/ t is documented in the SFWMM by the stage or de pth of the reservoir.

PAGE 39

28 O out I inGW Q ET GW Q P t S / (4-2) The volume balance components provide th e inputs to calculate the hydraulic residence time (HRT), which represents the wa ter quantity of the system. The HRT is a function of volume and flow (HRT = V/ Q). For a given reservoir volume, V HRT can be increased by reduci ng the inflow rate, Qin. The other performance measure for the reactor is the mean operating depth, H. Th e above information is obtained from the output file for the SFWMM by aggregating the cell by cell data. Water Quality Characterization Water quality characterization of the syst em is based on the inflow concentration, Cin, the rate parameter, and the minimum con centrations. As stor age and water quality changes are inseparable, the characteriz ation of the water quantity affects the characterization of the water quality. Inflow concentration The inflow concentration, Cin, is a function of the source and quantity of inflows. The water quantity of the system uses an aggregate inflow, Qin. Therefore, a single representative concentration is assigned to th e aggregate inflow. If more than one inflow source exists, a flow weighted concentrati on is calculated from Equation 4-3 where Qi is the flow, Ci is the concentration, and “i” designates the source. n i i n i i i inQ C Q C1 1 (4-3) Removal rate The KC* removal rate equa tion is used to represen t pollutant concentration profiles with time in a variety of reactor systems including wetlands (Kadlec and Knight

PAGE 40

29 1996) and a wide variety of urban stormwater BMPs (CRC for Catchment Hydrology 2005). The KC* model uses a removal rate model with a bac kground concentration ( C* ). Removal is a function of the initial concentration ( Cin), C* the overall rate constant, kV, and the hydraulic residence time (HRT), td. The KC* equation is shown in Equation 4-4. The pollutant is removed more rapidly at fi rst and then at a de creasing rate as HRT increases. Two-parameter calibration s hould be used for this model, where kV and C* are calibrated simultaneously. d vt k in oute C C C C ) (* (4-4) Model Rationale The University of Florida Water Quality Design Tool (UF WQDT) was intended from the outset to be a simple spreadsheet model used to predict the behavior of TP in the EAASR. The storage-release framework describe d in Lee et al. (2005) was to be used to provide an optimal design, which included wate r quality considerati ons. It was decided that a basic mass balance and uptake rate m odel would provide a good representation of TP behavior. Initially modeling was expect ed to occur at thre e levels; steady-state simulation of the long-term average, frequenc y analysis, and daily tim e series analysis. However, as the design and planning challe nges became evident the scope of the model was altered to best suit the needs of the project and the tight time schedule. Steady-state simulation of long-term aver ages was deemed the most appropriate modeling level for the dynamic conditions asso ciated with the design of the EAASR. A steady-state analysis can provide important in sights and comparisons to design scenarios. The simple mass-balance model is easily adap ted for new configurations and scenarios, while the steady-state nature allows for very fast run times. SFWMM updates and small

PAGE 41

30 model parameter alterations generally will not create significant changes in the long-term averages, where significant differences in the daily time series may be created. However, large changes to the configura tion or operation of the EAASR, STA 3/4, or South Florida system will be reflected. Additionally, calibration efforts for the model could be performed at an appropriate level of signifi cance, which is defined by the quality of the data. A description of the resulting m odel is provided in the next section. Spreadsheet Model The UF WQDT spreadsheet tool uses a flow and concentration calculator, a core reactor module, and a Solver objective and constraint section if a feedback between reactors exists. The basic UF WQDT for th e SC base Scenario (Figure 4-2) does not include feedback and all values are in US units, except TP concentrations. Values highlighted in light blue are us er entered, in white are calc ulated, and in orange are the results. The flow and concentration calculator, appearing first, aggregates multiple inflows generate a single representative flow and concentration. The core reactor module is broken into three sections: parameters, calculations, and results. The necessary parameters are entered into the parameter sect ion. These values are used to calculate the volume, HRT, and hydraulic loading rate (HLR ). The HLR is an alternate representation of the water quantity, which is nom inally defined as the inflow, Qin, divided by the area, A. Appropriate given and calculated valu es are used to calculate the outflow concentration, Cout, using Equation 4-4. Percent re moval of TP from the inflow concentration is calculated as well. The flow and concentration calculator and core reactor module are repeated fo r each reactor in the system (i.e. the EAASR compartments or STA 3/4) (Figure 4-2).

PAGE 42

31 Figure 4-2. Spreadsheet Interface for UF WQDT The two and four compartment EA ASR includes feedback between compartments. In this configuration, the inflow concentration of one compartment is dependent on the inflow and performan ce of other compartments. The inflow concentration can be solved algebraically us ing a system of equations. However, this would require programming specific equations fo r each compartment. The Solver tool in Excel was used to iteratively solve for each concentration without altering the core reactor module.

PAGE 43

32 For a design with feedback flow sour ces, the Solver objective and constraint section is used. Initial estimates of Cin are entered for each feedback. In the Solver section the difference of the feedback Cin cell and the appropriate compartment Cout cell are entered. The sum of the square differen ces are calculated for all feedback flows in the Total Difference cell and used as th e objective in Solver. Each feedback concentration is constrained to the minimum ach ievable concentration, C*. Solver is then used to reduce the value of the objective f unction to zero, which calculates the correct concentrations. This methodology assumes full treatment of the feedback flows. Water Quantity of EAASR and STA 3/4 The purpose of this section is to presen t the water quantity parameters necessary for water quality modeling. The reported inflows and outflows to the EAASR and STA 3/4 are based on output from the SFWMM. The water quality performance of the EAASR and STA 3/4 can be expected to va ry depending on how the inflows are divided among compartments, how the individual comp artments are operated as judged by the mean depth of storage in each compartm ent, and how water is transferred among compartments before discharge from the EAASR. Performance will also depend on whether inflows are routed to STA 3/4 dir ectly or through the EAASR. This section provides background information and results fo r the compartmentalizat ion configurations in Chapter 3. Water Quantity Data Source and Information A wide variety of options exist for directing water into the EAASR. The best projection of the expected inflows can be obtained by using the South Florida Water Management Model (SFWMM). The Interage ncy Modeling Center (IMC) modeled the EAASR and the downstream STA 3/4 usi ng the SFWMM Version 5.4. The IMC

PAGE 44

33 provided data from their EAASR simulation to th e University of Florida. All simulations were made with the assumption of 2050 land use projections and the reservoir as the “Next Added Increment” of the CERP project The provided data included inflow and outflow for each control structure, the maximu m capacity for the water control structures, rainfall, evapotranspiration (ET) and gr oundwater levels and the general modeling configuration and assumptions (USACE and SFWMD, electronic correspondence, February 2 and 3, 2006). Data were tabulat ed in 13,149 daily time steps for the 36 year period, starting on January 1, 1965 and ending on December 31, 2000. The data were provided for each of the EAAS R’s four square mile grid cells in the SFWMM for each time step. The total 2x2 grid is comprise d of 1,900 2x2 grid cells. Thus, a major simulation effort was required to run 36 year s of daily activity fo r 1,900 grid cells in order to evaluate one project with 19 gr id cells (SFWMD 2005a). The IMC output provides the necessary information on th e quantity and timing of the flows. Water Balance of the FC Base EAASR and STA 3/4 A water balance (WB) was calculated from the SFWMM to ensure proper modeling results for the EAASR and STA 3/ 4. The output data from SFWMM provided a complete volume balance for the four comp artment configuration, including stage data for each compartment and STA 3/4. Equation 4-2 was modified to include an imbalance term, which is attributed to ungaged flow. Th e imbalance term was a ttributed to either ungaged inflow or ungaged outflow by its sign and included in the WB (Eqn. 4-5) where QinU was ungaged inflow and QoutU was ungaged outflow. out outU out in inU inGW ET Q Q GW P Q Q t S / (4-5)

PAGE 45

34 The long-term WB was calculated for each compartment and STA 3/4 separately. The average daily value of each WB compone nt for the 36 year simulation period was used to represent the magnitude of each com ponent. The averages are presented in units of ac-ft/day to allow comparis on between the compartments and STA 3/4. The change in storage was calculated from the difference in vol ume between the first and last day of the dataset. The long term water balances for each compartment are presented in Table 4-1. Several key insights may be gained from in spection of the 36 year water balance. The differences between natural WB components, rainfall, ET, and GW offset each other for the EAASR and STA 3/4. GW and ungaged flows can be assumed to be negligible on a long term average. The resu lting simplified long term WB was used in the remainder of the report. Table 4-1. Thirty Six Year Water Balance for EAASR and STA 3/4 WB Term C1 C2 C3 C4 STA 3/4 Inflow (ac-ft/day) 485 470 442 290 2,201 Transfer (ac-ft/day) 105 318 83 280 NA Ungaged (ac-ft/day) 0 0 0 2.1 0 Rainfall (ac-ft/day) 58 120 58 118 208 GWI (ac-ft/day) 0.3 0.2 0.0 0.1 0.2 Total Inflow (ac-ft/day) 648 908 583 690 2,409 Outflow (ac-ft/day) 261 695 218 488 2,177 Transfer (ac-ft/day) 318 88 296 83 NA Ungaged (ac-ft/day) 4.0 3.6 2.5 0 13.4 ET (ac-ft/day) 64 119 64 118 216 GWO (ac-ft/day) 0.3 1.8 0.5 1.0 0.2 Total Outflow (ac-ft/day) 648 907 582 690 2,407 Storage Increase (ac-ft/d ay) 0.2 0.7 1.2 0.7 2.5 Net Error (ac-ft/day) 0 0 0 0 0 Average Depth (ft) 8.26 4.79 8.54 4.54 2.35 Single Compartment Reservoir Configuration Scenarios The single compartment (SC) reservoir is the simplest scenario. Two scenarios were performed to provide a basis for further comparison in the two and four

PAGE 46

35 compartment analyses. The first scenario wa s the SC base scenario (Figure 4-3). The second scenario routed STA 3/4 inflows di rectly from the canal through the single compartment reservoir before entering the STA. The second scenario is referred to as the SC STA scenario (Figure 4-4). Figure 4-3. Single Compartment Base Scenario Flow Diagram The SC base Scenario’s long–term water balance for the 36 year average of the area, inflows, and outflows were calculated (Fig ure 4-3). The total average inflow to the EAASR was 1,686 ac-ft/day. Of this total, 1, 182 ac-ft/day was released to STA 3/4. The average depth of the EAASR was 5.9 feet. Thus, the mean hydraulic residence time was 107.5 days or about three and a half mont hs. The long term HRT of 29 days was

PAGE 47

36 calculated for STA 3/4 from the arithmetic m ean of the annual HRT presented in Table 4-15. This value is 33 percent higher than if directly calculated from the long term average flow and depth. An HRT of 29 days was used for all subsequent long-term STA 3/4 analysis. The SC STA ScenarioÂ’s long-term water balance for the 36-year average of the area, inflows, and outflows were calculated (Fig ure 4-4). The total average inflow to the EAASR was 2,705 ac-ft/day. Of this total, 2, 201 ac-ft/day was releas ed to STA 3/4. The average depth of the EAASR was 5.9 feet. Thus, the mean HRT was 67.5 days, slightly more than two months. STA 3/4 results we re equal to the single compartment base scenario. Figure 4-4. Single Compartment STA Scenario Flow Diagram

PAGE 48

37 Two Compartment Reservoir Configuration Scenarios The EAASR was portioned into two compartm ents (TC) for further analysis. The two compartment configuration is the most si milar to the current design of the EAASR. The total external inflows and outflow for the EAASR are the same as in the single compartment case. The key difference is in how the water moves between the two compartments. Four scenarios were evaluated. The first scenario, which was considered to be the TSP scenario, was referred as the TC base scenario and is presented in Figure 4-5. The second scenario routed STA 3/4 inflows directly from the canals through the respective compartments before entering the STA. The second scenario is referred to as the TC STA scenario (Figure 4-6). The TC base Scenario’s long–term water balance for the 36 year average of the area, inflows, and outflow we re calculated (Figure 4-5). The total average inflow to Compartment 1 was 2,011 ac-ft/day, including inte r-reservoir transfers. The total average inflow to Compartment 2 was 1,284 ac-ft/day, in cluding inter-reservoir transfers. Of this total, 1,717 ac-ft/day wa s released to STA 3/4. The av erage depth of the EAASR was 5.9 feet in both compartments. Thus, the mean HRT for Compartment 1 was 45 days and 70 days for Compartment 2. STA 3/4 results were equal to the single compartment base scenario.

PAGE 49

38 Figure 4-5. Two Compartment Base Scenario Flow Diagram The TC STA Scenario’s long–term water ba lance for the 36 year average of the area, inflows, and outflows were calculated (F igure 4-6). The total average inflow to Compartment 1 was 2,495 ac-ft/day, including inte r-reservoir transfers. The total average inflow to Compartment 2 was 1,284 ac-ft/day, in cluding inter-reservoir transfers. Of this total, 2,201 ac-ft/day wa s released to STA 3/4. The av erage depth of the EAASR was 5.9 feet in both compartments. Thus, the mean HRT for Compartment 1 was 37 days and 70 days for Compartment 2. STA 3/4 results were equal to the single compartment base scenario.

PAGE 50

39 Figure 4-6. Two Compartment STA Scenario Flow Diagram The HRT of Compartment 1 in both the TC base and TC STA scenarios was relatively low. Low removal rates were expected from this Compartment. It was hypothesized that an outflow fr om Compartment 2 to STA 3/ 4 would alleviate the heavy loading to Compartment 1 and improve treatment. Two scenarios for were generated to test the hypothesis, called TC Miami Scenario and TC Miami STA Scenario. The TC Miami Scenario’s long–term water balance for the 36 year average of the area, inflows, and outflows were calculated (Figure 4-7). The total aver age inflow to Compartment 1 was 988 ac-ft/day, including inte r-reservoir transfers. Th e total average inflow to Compartment 2 was 1,284 ac-ft/day, including inte r-reservoir transfers. Of this total, 2,201 ac-ft/day was released to STA 3/4. Th e average depth of the EAASR was 5.9 feet in both compartments. Thus, the mean hydrau lic residence time for Compartment 1 was

PAGE 51

40 92 days and 70 days for Compartment 2. STA 3/4 results were equal to the single compartment base scenario. Figure 4-7. Two Compartment Mi ami Scenario Flow Diagram The TC Miami STA Scenario’s long–term water balance for the 36 year average of the area, inflows, and outflows were calcul ated (Figure 4-8). The total average inflow to Compartment 1 was 1,472 ac-ft/day, includ ing inter-reservoir tran sfers. The total average inflow to Compartment 2 was 1,284 ac-f t/day, including inter-reservoir transfers. Of this total, 2,201 ac-ft/day was released to STA 3/4. The average depth of the EAASR was 5.9 feet in both compartments. Thus the mean hydraulic residence time for Compartment 1 was 62 days and 70 days for Co mpartment 2. STA 3/4 results were equal to the single compartment base scenario.

PAGE 52

41 Figure 4-8. Two Compartment Miam i STA Scenario Flow Diagram Four Compartment Configuration Scenarios Two scenarios were analyzed for the four compartment (FC) EAASR configuration. The total external inflows a nd outflow for the EAASR are the same as in the single compartment case. The key diff erence is in how the water moves among the four compartments. The first scenario, referre d to as FC base scenario, was considered the MSP configuration (Figure 4-9). The direct flows to STA 3/4 were routed through Compartments B and D in the FC STA Scenario (Figure 4-10). The results for the scenarios for each compartment are presented in Tables 4-2 through 45. STA 3/4 results were equal to the single compartment base scenario.

PAGE 53

42 Table 4-2. Compartment A Water Quantity for the Four Compartment EAASR Configuration Scenario Area (acres) Inflow (ac-ft/d) Depth (feet) HRT (days) FC base 5,120 5898.372 FC STA 5,120 5898.372 Table 4-3. Compartment B Water Quantity for the Four Compartment EAASR Configuration Scenario Area (acres) Inflow (ac-ft/d) Depth (feet) HRT (days) FC base 10,240 7884.862 FC STA 10,240 1,2724.839 Table 4-4. Compartment C Water Quantity for the Four Compartment EAASR Configuration Scenario Area (acres) Inflow (ac-ft/d) Depth (feet) HRT (days) FC base 5,120 5258.583 FC STA 5,120 5258.583 Table 4-5. Compartment D Water Quantity for the Four Compartment EAASR Configuration Scenario Area (acres) Inflow (ac-ft/d) Depth (feet) HRT (days) FC base 10,240 5704.582 FC STA 10,240 1,1054.542

PAGE 54

43 Figure 4-9. Four Compartment Base Scenario Flow Diagram

PAGE 55

44 Figure 4-10. Four Compartment STA Scenario Flow Diagram Flow Equalization One of the usual purposes of storage upstr eam of a treatment system is to reduce the fluctuations in inflows to the treatment unit. The long-term average mean, standard deviation, and coefficient of variation (COV) for the inflow and outflow to the EAASR single compartment scenarios are shown in Tabl e 4-6. The COV is ratio of the standard deviation and mean values. If equalization is being accomplished, then the COV of the output should be less than the COV of the input.

PAGE 56

45 Table 4-6. Mean, Standard Deviation, and COV of Single Compartment Scenarios Inflow Outflow Scenario Mean (ac-ft/d) Standard Deviation (ac-ft/d) Coefficient of Variation Mean (ac-ft/d) Standard Deviation (ac-ft/d) Coefficient of Variation EAASR SC base 1,686 3,262 1.9 1,662 4,179 2.5 EAASR SC STA 2,705 3,904 1.4 2,681 4,591 1.7 The single compartment scenarios of the EAASR do not indicate the reservoir is used for flow equalization. In both cases, th e COV is higher for the outflow than inflow. The STA scenario displays less variation, but it is not believed that this corresponds to more flow equalization. With residence times in the range of two months, the EAASR is able to completely dampen fluctuation from most rainfall events and should have a significant impact on the longe r-term precipitation events during the wet season. The EAASR is to be a multi-purpose reservoir. Thus, the ability to use it to equalize inflows to STA 3/4 may be constrained by these other purposes. Comparison of STAs and Lakes in Southern Florida Quantity-related design parameters for STAs and lakes and reservoirs are shown in Table 4-7 (Walker and Kadlec 2005b). With the exception of the Iron Bridge wetland with an HRT of 66 days, the HRTs for wetla nds and STAs are in the range of 7 to 21 days. Thus, the mean residence time of 29 days for STA 3/4 shows that STA 3/4 can potentially be used to a greater extent. The associated median operating depth for the comparables is about 1.7 feet. The mean opera ting depth of 2.3 feet for STA 3/4 is on the high side, but not exceptional. These ST As are designed as water quality control facilities; accordingly, there is relatively little variability in how they operate.

PAGE 57

46 In stark contrast to the relatively homogeneous behavior of the 22 STAs, the 15 entries in the lake and reser voir database indicate a wide range of behavior. The four periods studied for Lake Okeechobee have HRTs in the range of 580 to 715 days. At the other extreme, four entries have reported HRTs of less than 10 days. Mean depths range from 4.1 to 11.2 feet. The estimated HRT range for the EAASR was 67 to 107 days with a mean depth of 5.9 feet. Thus, the possibl e Lake/Reservoir comparables in Table 4-6 can be reduced to George, Istokpoga, Harney, Jessup, Crescent, and Thonotosassa with regard to quantity characteristics. More de tailed analysis of reservoir comparables will be presented in the next section on water quality.

PAGE 58

47 Table 4-7. Comparative HLRs and HRTs for 37 Reactors in Southeast Florida (Data from Walker and Kadlec 2005b) HLR HRT Mean Depth Max Depth Number Category in/day days ft ft % Full Emergent Wetlands 1 ENRP_C1 1.1 17 1.6 2.2 74.3% 2 ENRP_C2 2.1 14 2.5 2.9 87.4% 3 ENRP_C3 1.1 13 1.1 1.5 75.4% 4 STA1W_C1 1.8 13 2.0 3.0 67.9% 5 STA1W_C2 2.9 11 2.6 3.5 74.8% 6 STA1W_C3 2.6 7 1.6 2.4 66.0% 7 STA5_C1AB 2.6 8 1.7 2.2 79.4% 8 STA5_C2AB 1.5 9 1.1 2.0 57.4% 9 BoneyMarsh 0.8 23 1.5 3.0 51.2% 10 IronBridge 0.4 66 2.3 2.3 99.6% 11 WCA2A 1.4 13 1.5 4.3 35.6% Median 1.5 13 1.6 2.4 74.3% Pre-existing Wetlands 1 STA2_C1 1.2 22 2.1 2.7 79.7% 2 STA2_C2 1.9 14 2.2 3.3 64.4% 3 STA6_C3 2.7 7 1.6 2.5 66.0% 4 STA6_C5 1.3 15 1.6 2.3 70.1% 5 WCA2A 1.7 9 1.2 4.0 31.2% 6 WCA2A 1.8 8 1.2 4.0 31.0% 7 WCA2A 1.4 11 1.2 4.0 30.7% Median 1.7 11 1.6 3.3 64.4% Submerged Aquatic Vegetation Systems 1 ENRP_C4 5.0 5 2.2 2.5 87.2% 2 STA1W_C4 3.7 7 2.0 2.8 72.0% 3 STA1W_C5AB 3.7 7 2.2 3.3 66.0% 4 STA2_C3 2.0 17 2.9 4.3 66.2% Median 3.7 7 2.2 3.1 69.1% Lakes or Reservoirs 1 OKEE_7578 0.2 674 8.4 10.5 79.6% 2 OKEE_7986 0.2 704 8.8 11.0 80.4% 3 OKEE_8794 0.2 714 8.6 10.9 78.6% 4 OKEE_9599 0.2 578 9.4 11.3 83.7% 5 HELLNBLAZES 24.9 2 4.1 6.4 64.3% 6 SAWGRASS 20.7 3 4.4 6.7 65.2% 7 GEORGE 1.9 73 11.2 12.6 89.0% 8 ISTOKPOGA 0.4 170 5.2 6.8 77.1% 9 ISTOK_2 0.4 178 5.2 6.1 84.3% 10 POINSETT 4.3 8 3.0 6.1 48.2% 11 HARNEY 6.9 16 9.4 15.0 63.0% 12 HARNEY_2 8.7 9 6.7 11.1 60.4% 13 JESSUP 0.4 176 5.3 9.7 54.4% 14 CRESCENT 1.1 112 10.0 12.2 81.8% 15 THONOTO 1.1 54 5.0 7.7 64.6% Median 1.1 112 6.7 10.5 77.1%

PAGE 59

48 Water Quality for the EAASR and STA 3/4 The water quality of the EAASR and ST A 3/4 were separated into the inflow water quality and uptake parameters. The inflow water quality and each uptake rate parameter are detailed in this section. Inflow Water Quality Inflow to the EAASR and STA3 /4 occurred from four external sources. A mean concentration for each source was calculated fr om historical data. Pumping Stations S2 and S3 are used to control the water en tering and leaving Lake Okeechobee through the NNR and Miami Canals. The water passing th rough pump stations S2 (NNR Canal) and S3 (Miami Canal) was divided into two t ypes: to LOK from the EAA (back pumping), and from LOK to the EAA. These were designated as Basin and LOKCanal, respectively and correspond to the external inflows in the water quantity section. The mean, coefficient of variation (standard deviation divided by mean), and count of data for each parameter are presented in Tables 4-8 thr ough 4-11. The data included measurements from 1973 to 2004, which were taken at different frequencies depending on the parameter. In both the NNR and Miami Canals, wa ter quality exiting LOK was of higher quality than water entering LOK. The mean water quality coming to LOK from the EAA was higher quality in the Miami Canal than in the NNR Canal. The coefficients of variation for concentrations to LOK were similar, with the exception of iron, turbidity and dissolved oxygen. The reverse is true for flows from LOK to the EAA, where S2 had better water quality than S3 with the exception of iron and phosphorus, as TP. The coefficients of variation for the flows fr om LOK are similar with the exception of sodium.

PAGE 60

49 Table 4-8. NNR Basin Mean, Coefficient of Variation, and Count of Water Quality Parameters Parameter NNR Basin Mean NNR Basin Coefficient of Variation NNR Basin Number of Records TP (mg/L) 0.150 0.53 236 TKN (mg/L) 3.69 0.34 227 NOx-N (mg/L) 1.811 1.01 227 TN (mg/L) 5.505 0.46 227 DO (mg/L) 2.47 0.59 176 pH (su) 7.19 0.05 133 Specific Conductance (uS/cm) 1191 0.21 177 Turbidity (NTU) 10.7 1.34 160 TSS (mg/L) 18.1 1.45 138 Alkalinity as CaCO3 (mg/L) 317 0.27 186 Calcium (mg/L) 109.2 0.22 88 Chloride (mg/L) 137.1 0.28 205 Sulfate (mg/L) 101.9 0.41 67 Sodium (mg/L) 97.8 0.29 88 Total Iron (ug/L) 267 0.89 57 Total Mercury (ng/L) 2.82 0.50 11 Atrazine (ug/L) 0.670 0.97 16 Table 4-9. LOKNNR Canal Mean, Coefficient of Variat ion, and Count of Water Quality Parameters Parameter LOKNNR Mean LOKNNR Coefficient of Variation LOKNNR Number of Records TP (mg/L) 0.077 0.46 76 TKN (mg/L) 1.79 0.35 72 NOx-N (mg/L) 0.206 2.01 72 TN (mg/L) 1.998 0.45 72 DO (mg/L) 6.57 0.26 70 pH (su) 7.78 0.04 99 Specific Conductance (uS/cm) 625 0.25 72 Turbidity (NTU) 71.0 0.87 103 TSS (mg/L) 12.9 0.96 67 Alkalinity as CaCO3 (mg/L) 138 0.33 75 Calcium (mg/L) 48.6 0.23 22 Chloride (mg/L) 82.0 0.25 74 Sulfate (mg/L) 49.4 0.27 29 Sodium (mg/L) 52.6 0.30 22 Total Iron (ug/L) 170 0.70 27 Total Mercury (ng/L) NA NA 0 Atrazine (ug/L) 0.280 0.54 26

PAGE 61

50 Table 4-10. Miami Basin Mean, Coefficient of Variation, and Count of Water Quality Parameters Parameter Miami Basin Mean Miami Basin Coefficient of Variation Miami Basin Number of Records TP (mg/L) 0.112 0.73 148 TKN (mg/L) 3.06 0.33 142 NOx-N (mg/L) 1.996 0.85 142 TN (mg/L) 5.053 0.46 142 DO (mg/L) 3.45 0.42 125 pH (su) 7.23 0.05 133 Specific Conductance (uS/cm) 962 0.27 124 Turbidity (NTU) 8.2 0.82 106 TSS (mg/L) 12.9 1.10 91 Alkalinity as CaCO3 (mg/L) 248 0.24 112 Calcium (mg/L) 106.7 0.29 55 Chloride (mg/L) 112.2 0.37 125 Sulfate (mg/L) 72.8 0.43 54 Sodium (mg/L) 70.4 0.30 55 Total Iron (ug/L) 192 0.49 44 Total Mercury (ng/L) 2.2 0.33 9 Atrazine (ug/L) 0.470 1.26 13 Table 4-11. LOKMiami Mean, Coefficient of Variation, and Count of Water Quality Parameters Parameter LOKMiami Mean LOKMiami Coefficient of Variation LOKMiami Number of Records TP (mg/L) 0.060 0.50 105 TKN (mg/L) 1.85 0.41 97 NOx-N (mg/L) 0.297 1.74 97 TN (mg/L) 2.143 0.52 97 DO (mg/L) 6.35 0.30 98 pH (su) 7.8 0.06 99 Specific Conductance (uS/cm) 767 0.42 100 Turbidity (NTU) 6.9 0.84 103 TSS (mg/L) 8.2 0.70 78 Alkalinity as CaCO3 (mg/L) 164 0.44 104 Calcium (mg/L) 64.0 0.44 32 Chloride (mg/L) 101.0 0.45 104 Sulfate (mg/L) 69.5 0.74 37 Sodium (mg/L) 71.5 0.64 32 Total Iron (ug/L) 148 0.93 32 Total Mercury (ng/L) NA NA 0 Atrazine (ug/L) 0.280 2.01 26

PAGE 62

51 As stated previously, TP is the major parameter of interest in the system. The overall flow-weighted average TP concentra tion for the EAASR and STA 3/4 system is 0.101 mg/L. TP concentrations range from a low of 0.060 mg/L for LOK releases via the Miami Canal to a high of 0.150 mg/L for NNR Basin. The relationship between concentration a nd flow was investigated for all the water quality parameters. A linear regression wa s used to evaluate the relationship. For NNR Canal, flows for total iron to LOK showed a weak relationship with an R2 of 0.36. All other parameters did not show a significan t relationship. Thus, concentration will be assumed to be independent of flow for all constituents. Background Concentration The removal of TP was modeled usi ng the KC* equation. The background concentration, C* is the lowest possible concentrati on the reactor can reach. An accurate value of C* is important, because the concentration approaches C* asymptotically. A Walker and Kadlec (undated) report calibrated values of C* ranging from 0.004 to 0.020 mg/L for STAs. Walker and Havens (2003) use 0.007 mg/L as the background concentration for precipitation. In this study, the background concentration will be determined using calibration data with the constraint that C* >= 0.007 mg/L, the estimated value for precipitation. Reservoir TP Overall Reaction Rate Several water quality models have been de veloped to simulate the TP kinetics in general CERP reservoirs including Wa lker and Kadlec (2005a) and USACE and SFWMD (2005). The general CERP reservoir m odels use large datasets of comparable systems to find the average overall TP kinetics A more specific set of comparables was sought to develop a water quality model fo r the Everglades Agricultural Reservoir

PAGE 63

52 (EAASR), a CERP reservoir. The selection of comparable systems is also documented in Reisinger et al (2006). Long-term average evaluation of comparable systems Walker and Kadlec (2005a) tabulated the long term averages of a comprehensive set of 18 comparable Florida lakes and a re servoir in the documentation of the DMSTA v.2 calibration (Figure 4-6). The long term average values were used to determine several closely comparable systems of the Single Compartment base EAASR Scenario. The SC base scenario was used because it is the most gene ral of the scenarios. The analysis was performed on both the water quantity and TP water quality of the comparable dataset, as they were inseparabl e for parameter estimation. The analysis of comparables was performed using three ca tegories of decision variables: lake characterization, water quantity, and water qua lity. The decision variables and weighting factors were chosen to best represent the parameters in Equation 3-4. The selected decision variables for lake characterization are surface area and depth. Water quantity was represented by the HRT and HLR. Water quality was represented by the inflow TP concentration. A weighted decision matrix was used to evaluate the dataset of comparables. A rating of one to five was given for each decision variable for each lake, where five was the most comparable to the EAASR. A weighting f actor was applied to each decision variable, where the sum of all we ighting factors equals one. The percent of the total points (five) was used to rate each dataset. Equation 4-6 was used to calculate the rating of each lake or reservoir. m j j j n i m j j i j i i iR W R W f P1 11 ,100 Eqn. 4-6

PAGE 64

53 Where i = individual lake, j = decision variable, n = number of lakes, m = number of decision variables, P = percent of total points, ƒ = ratio of points r eceived, W = weighting factor, and R = rating from one to five. The weighting factors, ratings, percent of total points, and overall rank for lakes ranked five and under are presented in Ta ble 4-12. The Crescent Lake, Lake Istokpoga_2, Lake Jessup, Lake George, and Lake Poinsett datasets were developed by Burns and McDonnell (2004a). The Lake Is tokpoga dataset was developed by Walker and Haven (2003). Crescent Lake and Lake Is tokpoga were found to be clearly the most comparable overall to the EAASR. A sensitivity analysis was performed using a range the weighting factors and repe atedly found Crescent Lake and Lake Istokpoga to be the highest ranked systems. These lakes were determined to be the most comparable to EAASR and were reviewed in further detail. Table 4-12. Decision Rankings for Comparable Systems Ranked Five and Better Decision Variable Surface Area Depth HLRHRT Inflow TP Conc. Percent of Total Points Rank Weighting 0.1 0.1 0.1 0.35 0.35 NA NA Crescent Lake 4 3 4 5 5 92% 1 Lake Istokpoga 5 5 5 3 5 86% 2 Lake Istokpoga_2 5 5 5 3 5 86% 3 Lake George 3 2 2 4 4 70% 4 Lake Jessup 3 5 5 3 3 68% 5 Lake Poinsett 2 4 1 2 5 63% 3 More detailed analysis of selected comparables Next, the salient attributes of Crescent La ke and Lake Istokpoga were analyzed in more detail by looking at time se ries data and the extent to which the terms in the water and TP budgets were measured. The an alysis was again divided into lake characterization, water quan tity and water quality.

PAGE 65

54 Crescent Lake was found to have a larg e quantity and high frequency of TP samples. However, large uncertainty exists in the mass balance due to high proportion of ungaged flows and a large percentage of TP measurements associated with ungaged flows. Additionally, no direct stage-area or stage-volume relationshi p exists. The depth profile fluctuated regularly from 8.8 feet to 11.5 feet in depth. This more detailed analysis found that Crescent Lake was not a very reliable data source for parameter estimation. Lake Istokpoga has a relatively complete water balance. TP measurements are infrequent, but have an average of two data points per average residence time. A recent bathymetric map provides a stage-area a nd stage-volume relationship for the lake (SFWMD 2005b). The depth profile fluctuated generally over a re latively small range from 4.9 feet to 5.6 feet. Th ree recent reports are availabl e for Lake Istokpoga: Walker and Havens (2003), Burns and McDonne ll (2004a), and South Florida Water Management District (2005). The conclusion from the more detailed analysis is that Lake Istokpoga is the best can didate for parameter estimation. Selection of the appropriate period of record for the chosen comparable system Lake Istokpoga is located northwest of Lake Okeechobee, near the center of Highland County. The inflow and outflow TP concentrations and the HRT were assumed to be key variables for parameter estimation when using Equation 4-3. The mass balance terms used to calculate the value of key variables were found from recent reports and weather station records. The period of reco rd (POR) used for parameter estimation was selected by influent TP concentrations in the range of 0.100 mg/L and no significant trends over time.

PAGE 66

55 The inflow mass balance terms include gaged and ungaged inflows and rainfall. A map of Lake Istokpoga and the location of major mass balance terms, except ET, are shown in Figure 4-11. Inflows to the lake were measured at Arbuckle Creek and Josephine Creek. TP was measured at each of the creeks. Ungaged inflows were estimated as 17% of the gaged flow and were assumed to be from seepage (Walker and Haven 2003). Rainfall was measured at the S68 structure. The TP concentration in ungaged inflows was estimated as 0.050 mg/L a nd rainfall concentrations as 0.007 mg/L (Walker and Haven 2003). Outflows from the lake occur mainly from the measured S-68 structure and evapotranspiration (ET). Ou tflows can occur through the controlled Istokpoga Canal, although it is seldom used and is considered to be a negligible outflow (Walker and Haven 2003). Pan evaporati on was measured approximately 30 miles northeast of the lake at the S-65 structure. A pan coefficient of 0.76 was used to convert the measurement to ET (Burns and McDonnell 2004a). The outflow concentration at S68 was sampled and it was assumed that no TP was transported in the ET. The volume in the lake was calculated using the stage-vol ume relationship reporte d in SFWMD (2005b) and the measured headwater stage at the S-68 structure. The removed TP mass was calculated as the difference in inflow and outflow mass balance terms. The POR was determined for the mass balance terms presented above. The beginning date of the POR was determined by th e TP samples, which were first taken in February of 1988 (Burns and McDonnell 2004). Water quantity and quality data from calendar years 1988 to 2002 were downloade d from the SFWMD DBHYDRO database and used in the remainder of the analysis. This period reflects the longest continuous period where all inputs were measured. Miss ing data were interpolated as needed.

PAGE 67

56 Figure 4-11. Lake Istokpoga and Mass Balance Locations Quarterly and bimonthly TP sampling fr equencies were found at each measured flow location during the 1988 to 2005 POR, with the exceptions of 2001 and 2002 that were sampled less frequently. Time series plots of the TP concentrations indicate an increasing trend in the latter part of the tim e series for S-68 (Figure 4-12) and Arbuckle Creek (Figure 4-13). The desired comparab le TP concentration is 0.100 mg/L for the Arbuckle Creek inflow and with minimal tr ends for Arbuckle Creek and S-68. If the entire POR is used, then the TP concentr ation increases from approximately 0.060 to 0.190 mg/L for Arbuckle Creek and from a pproximately 0.025 mg/L to 0.085mg/L for S68. A portion of the POR was selected to mi nimize the increasing trend and the optimal POR for S-68 was determined to be from 1988 through 1997. The outflow had a mean TP concentration of approximately 0.035 mg/L for this period (Figur e 4-12). The portion of the total POR was reduced to yield an optimal POR for Arbuckle from 1988 through 1995 (Figure 4-13). The inflow concentra tion for this period was approximately 0.080 mg/L for this period.

PAGE 68

57 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 889092949698000204 YearTP Conc. (mg/L) S-68 88-05 88-97 Figure 4-12. S-68 TP Data and Adjusted POR 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 889092949698000204 YearTP Conc. (mg/L) Arbuckle 88-05 88-95 Figure 4-13. Arbuckle Creek TP Data and Adjusted POR

PAGE 69

58 A compromise POR from 1988 through 1997 was chosen to produce the least temporal trends overall. Th is period provided the least tr end in the S-68 dataset and a slight positive trend from 0.100 mg/L to 0.125 mg/L in Arbuckle Creek. Additionally, the selected POR was on averag e closer to the inflow concentration goal than the optimal Arbuckle period. This POR included 80 TP samples of Arbuckle Creek and Josephine Creek and 89 TP samples of S-68. The TP samples for each location were analyzed for trends as a function of the calendar month. Arbuckle and Josephine Creeks included four to eight samples for each calendar month and S-68 had six to 10 samples for each calendar month. No clear monthly trend was found. A POR of January 1, 1988 to December 31, 1997 was determined for Lake Istokpoga from the constraints set by the TP temporal trends. A monthly averaging period for the mass balance of Lake Istokpoga was chosen. Preliminary estimates of the EAASR HRT were found to be between tw o and five months, which allow for a parameter estimate on a monthly scale. A monthly averaging period was found to be acceptable for Lake IstokpogaÂ’s water quantity and quality data. Excellent inflow and stage records allow the estimation of the severa l short periods of missi ng data with little additional error. Due to the bimonthly and quarterly TP sampling, uncertainty exists at any water quality averaging pe riod. A number of averaging or regression methods could be reasonably used to estimate the missi ng monthly TP concentrations for the POR. Therefore, an averaging period equal to that of the water quantity was chosen. Rate constant calibration All TP measurement stations were used fo r the TP calibration of the overall rate constant. Data was generated for each station shown in Figure 4-14. It was assumed that no change in the quantity or quality of the influent occurred between sampling and the

PAGE 70

59 lake. The inflow measuring points, Arbuckl e Creek and Josephine Creek, were therefore relocated to their respective lake inlets (Figure 4-14). The median value of each sampling station in g/L was determined (Figure 4-15). A well defined concentration gradient existe d from the Arbuckle Creek inflow at the northern end of the lake to the outflow at th e southeast corner of the lake. A probable flow path was developed for the Arbuckle and Josephine Creek flows. Due to the low concentration, the Josephine Creek flow path was not used. Figure 4-14. Water Quality Sampli ng Stations in Lake Istokpoga

PAGE 71

60 Figure 4-15. Median TP Values at Lake Istokpoga Sampling Stations The average HRT of 197 days was devel oped for the lake using the mass balance terms described in the preceding section. Know ing the HRT for the lake, it is possible to develop a TP concentration vs. residence time curve based on the inlet, outlet, and in-lake measurements. The resulting curve is s hown in Figure 4-16. Using the boundary conditions of the concentra tion at 0 days was equal to 0.082 mg/L, the concentration at 60 days was equal to 0.0475 mg/L, and the concentration is 0.028 mg/L at 197.5 days, then the overall rate constant, kV, was calculated as 0.016 days-1 and the background concentration, C*, is 0.0254 mg/L.

PAGE 72

61 0 10 20 30 40 50 60 70 80 90 050100150200 HRT (days)TP Concentration (ug/L) Figure 4-16. Lake Istokpoga TP Rem oval in Arbuckle Creek Flow Path EAASR reaction rate sensitivity analysis A sensitivity analysis was performed on the calibration of the EAASR TP removal rate constant. As a two paramete r calibration was performed, both the reaction rate and background concentration were evalua ted. The SC STA scenario was used for the sensitivity analysis, whic h is characterized by a 67.5 day HRT and 0.101 mg/L inflow TP concentration. The combined effect of C* and the over all reaction rate was first analyzed. The EAASR outflow concentration for multiple b ackground concentrations and reaction rates spanning two orders of magnitude were calcula ted (Figure 4-17). The results indicate, as

PAGE 73

62 expected, an increasing tr end in outflow concentra tion for greater background concentrations and the greater the reaction rate the less variability in outflow occurred. For a k of 0.01 days-1, which is similar to the calibrated value, the outflow concentration varied from about 0.055 mg/L at C* equal to 0.005 mg/L to 0.065 mg/L at C* equal to 0.030 mg/L. From this analysis, it was clea r that the calibration of k and C* can significantly affect the results of the modeling. 0.000 0.020 0.040 0.060 0.080 0.100 0.120 0.0000.0100.0200.0300.040 Background Concentration, C* (mg/L)EAASR Outflow Concentraion (mg/L) k = 0.001 k = 0.005 k = 0.010 k = 0.050 k = 0.100 Figure 4-17. Sensitivity Analysis for EAASR Parameters k and C* The effect of the overall reaction rate on both the EAASR and STA 3/4 was evaluated. The outflow concentrations fo r large ranges of EAASR k values were calculated for the EAASR and STA 3/4. This analysis used a fixed EAASR C* of 0.025 mg/L and the reported STA 3/4 parameters. A large effect on the outflow concentration of the EAASR was observed for reaction rates less than 0.05 days-1. The STA 3/4

PAGE 74

63 outflow concentration is rela tively independent of the EAASR outflow concentration, which is the only inflow to the STA, and t hus reaction rate. This independence was due to the relatively long HRT and high reaction rate. In fact, if the STA 3/4 received all inflow without first routing through the re servoir it would still achieve an outflow concentration of 0.021 mg/L. This result doe s not account for mass loading and its long term effect on the STA. Therefore, such an operating scheme is not suggested. Data from this analysis for selected reac tion rates are presented in Table 4-13. 0.000 0.010 0.020 0.030 0.040 0.050 0.060 0.070 0.080 0.090 0.100 00.050.10.15 Reaction Rate (days-1)Outflow Concentration (mg/L) EAASR STA 3/4 Figure 4-18. Outflow Concentration of th e EAASR and STA 3/4 for Varying Reaction Rates

PAGE 75

64 Table 4-13. Selected Results of the EAAS R and STA 3/4 for Varying Reaction Rates HRT (days) EAASR Reaction Rate (days-1) C* (mg/L) EAASR Cin (mg/L) EAASR Cout (mg/L) STA 3/4 Cout (mg/L) 67 0.001 0.025 0.101 0.096 0.021 67 0.005 0.025 0.101 0.079 0.021 67 0.010 0.025 0.101 0.064 0.021 67 0.050 0.025 0.101 0.028 0.020 67 0.100 0.025 0.101 0.025 0.020 67 0.150 0.025 0.101 0.025 0.020 The Basis of Design Report provides an alternative analysis of the EAASR TP uptake. The parameter estimates needed to produce these results will be evaluated to allow comparison between the simulations. The report models only compartment 1 of the TC configuration, which is referred to as Phase 1. Phase 1 consists of a 12 foot reservoir on 15,833 acres of Component A creating 190,551 acre-feet of possibl e storage (SFWMD 2006). The location and simulated water bala nce are shown in Figur e 4-19. The water balance was simulated from the SFW MM Version 5.4.2 using the ECP 2010 and 2015 simulations, as compared to the 2050 Next Added Increment simulation used in the previous scenarios. An average daily infl ow of 1,988 acre-feet per day was calculated from the water balance and an average dept h over the POR of 4.5 feet was reported. An HRT of 35.8 days was calculated for Phase 1. SFWMD (2006) reported an average TP concentration of 82 ppb entering the reservoir. DMSTA 2 was used to simulate Phase 1 and found an outflow of 68 ppb or a 17 percent removal. These results were gene rated using an areal reaction rate on a daily basis. To obtain these re sults using the UF WQDT with a background concentration of 00.0254 mg/L a reaction rate of 0.0033 days-1 would be needed. The value is almost one order of magnitude lower than the Lake Ist okpoga estimate. Due to the differences in

PAGE 76

65 configuration and reactions rates the two modeling efforts should not be directly compared. Figure 4-19. Water Balance and Location of the EAASR – Phase 1 (SFWMD 2006). STA TP Reaction Rate Key parameters for 16 STAs in southeas t Florida are shown in Table 4-14. The average inflow TP concentration is 0.102 mg/L. The optimal values of C* and k can be

PAGE 77

66 determined by solving the following constr ained optimization problem using the Excel Solver. The objective function is to minimi ze the sum of the squares of the errors between the calculated and meas ured outflow concentrations for the 16 STAs. C* is constrained to be >= 0.007 mg/L. Optimized values of two decision variables k and C* were calculated as 0.127 days-1 and 0.0194 mg/L, respectivel y for an assumed average inflow concentration of 0.102 mg/L. Table 4-14. Characteristics of 16 STAs in Southeast Florida (Data from Walker and Kadlec 2005a) STA Mean Depth HRT Inflow Conc. Outflow Conc. Percent Removal (ft) (days) (mg/L) (mg/L) ENRP_C1 1.6 17 0.089 0.039 56% ENRP_C2 2.5 14 0.068 0.034 49% ENRP_C3 1.1 13 0.039 0.019 50% ENRP_C4 2.2 5 0.036 0.017 53% STA1W_C1 2.0 13 0.140 0.05 64% STA1W_C2 2.6 11 0.142 0.088 38% STA1W_C3 1.6 7 0.058 0.033 43% STA1W_C4 2.0 7 0.100 0.032 68% STA1W_C5AB 2.2 7 0.153 0.055 64% STA2_C1 2.1 22 0.100 0.012 88% STA2_C2 2.2 14 0.110 0.021 81% STA2_C3 2.9 17 0.128 0.015 89% STA5_C1AB 1.7 8 0.121 0.071 42% STA5_C2AB 1.1 9 0.203 0.125 39% STA6_C3 1.6 7 0.071 0.019 73% STA6_C5 1.6 15 0.079 0.016 79% Average 2.0 11 0.104 0.042 60% Predicted Performance of EAASR Removal of TP The calculated outflow concentrations from the EAASR for various assumed values of initial concentra tion and residence time are s hown in Figure 4-20. For a residence time of 50 days, the effect of th e initial concentration at 50 days and the associated percent removal is shown in Tabl e 4-15. The percent removal varies from a

PAGE 78

67 low of 19 % if the initial concentration is 0.025 mg/L to 42 % if the initial TP concentration is 0.150 mg/L. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0255075100125150 HRT (days)Concentration (mg/L) Cin = 0.050 Cin = 0.075 Cin = 0.100 Cin = 0.125 Cin = 0.150 Figure 4-20. Effect of Initi al Concentration and Residence Time on Outflow TP Concentration for the EAASR Table 4-15. Effect of Initial TP Concentration on TP at 50 Days and Percent Control for the EAASR Initial TP (mg/L) TP (mg/L) at 50 days % Control 0.050 0.0419% 0.075 0.0625% 0.100 0.0728% 0.150 0.0942% Predicted Performance of STA 3/4 Removal of TP The primary differences between the E AASR and STA 3/4 with regard to the removal Equation 4-3 is that the residence time s in the STAs are typically in the range of

PAGE 79

68 7 to 21 days whereas the rate constant is an order of magnit ude higher at 0.127 day-1. The effect of the much higher rate consta nt for the STAs is dramatic as shown by comparing Figure 4-20 for the EAASR and Figure 4-21 for the STAs. An assumed rate constant of k equals to 0.127 day-1 causes the concentrati on to reach the assumed background level in 50 days over the entire ra nge of assumed inflow concentrations. Figure 4-22 shows the same information with th e x axis rescaled to a maximum residence time of 50 days instead of 150 days. The results shown in Figure 4-22 suggest that using residence times in the 7 to 21 day range provide a relatively good level of performance. For a residence time of 15 days, the associated outflow concentrations and % control are shown in Table 4-16. In this case, the percent control ranges from a low of 60% for an initial concentration of 0.025 mg/L to a high of 76% if the initial concentration is 0.150 mg/L, though all are approaching the background concentration. These results indicate the following ke y points about the EAASR/STA system: The STAs are much more effective water quality controls than the EAASR in that they achieve significant removals in 7 to 21 days of residence time as dramatically illustrated by comparing Figures 4-25 and 4-27. The EAASR can provide significant water quality improvements if the residence times exceed 25 days and the initial concentrations are relatively high. For the same area, the residence times in the STAs can be expected to be significantly less than for the EAASR since they are operated at much shallower depths. The initial concentration has a significan t effect on performance especially if performance is measured as percent pollutant removal.

PAGE 80

69 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0255075 HRT (days)Concentration (mg/L) Cin = 0.050 Cin = 0.075 Cin = 0.100 Cin = 0.125 Cin = 0.150 Figure 4-21. Effect of Initi al Concentration and Residence Time on Outflow TP Concentration for the STAs 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0510152025 HRT (days)Concentration (mg/L) Cin = 0.050 Cin = 0.075 Cin = 0.100 Cin = 0.125 Cin = 0.150 Figure 4-22. Figure 4-21 Rescaled to Residence Times Up to 50 Days

PAGE 81

70 Table 4-16. Effect of Initial TP Concentration on TP at 15 Days and Percent Control for the STAs Initial TP (mg/L) TP (mg/L) at 50 days % Control 0.050 0.02060% 0.075 0.02172% 0.100 0.02179% 0.150 0.02286% Summary and Conclusions The UF WQDT is a steady-state mass ba lance model utilizing the KC* kinetic model and was developed to determine the TP water quality of the EAASR and STA 3/4. The model uses a single aggregate in flow and concentration, depth, background concentration, and a volumetric reaction rate to simulate TP. The model calculates the HRT and HLR of each reactor, as well as the outflow concentration and percent removal. Long-term averages were determined to be the most appropriate level of sophistication needed in order to provide key insights for the dynamic EAASR design. Water quantity and quality inputs to the m odel were calculated for the base and STA scenarios for each configuration. Th e HRT of the EAASR ranged from 39 to 107.5 days. The EAASR was not used for flow equalization. The HRT of STA 3/4 was calculated to be 29 days for al l scenarios. Mean inflow c oncentrations of 17 parameters of interest for the four external inflow sources to the reservoir and STA 3/4 were reported. Parameter estimations were perfor med using closely comparable systems with high quality data. Lake Istokpoga was found to be the best comparable for the general single compartment scenario of the EAASR. Data from 16 STAs were used as comparables to STA 3/4. The following water quality parameters estimates for the EAASR and STA 3/4 were selected: C* = 0.0254 mg/L for the reser voir and 0.0194 mg/L for the STA k = 0.016 day-1 for the reservoir and 0.127 day-1 for the STA

PAGE 82

71 A sensitivity analysis of the EAASR react ion rate and background concentration was performed for the SC STA scenario. Th e calibration of each parameter was found to significantly affect the results especially if the reaction rates are less than 0.05 days-1 for a background concentration of 0.025 mg/L. To reproduce TP removal estimates provided in the BODR, a reaction rate of 0.0033 days-1 for a C* of 0.025 is needed. Due to the differences in configurations and reaction rates the modeling efforts are not directly comparable. An analysis of the predicted performa nce of the EAASR and STA 3/4 systems was performed. STAs were found to provide si gnificantly greater water quality treatment than reservoirs. The STA systems will approach the background concentration asymptotically for retention times greater than 25 days. However, since the STA is operated at much lower depths than the reser voir, they will have much lower HRT for a given area and inflow. The init ial concentration can affect the outcome of the EAASR or STA 3/4, especially if measured by the percent removal of TP. The initial concentration of each scenario is calculated in Chapter 5, as the inflow of the TC and FC scenarios are a function of their removal. The re sulting concentrations and the inputs from this chapter will be used to generate TP estimates for each scenario.

PAGE 83

72 CHAPTER 5 ESTIMATED WATER QUALITY CHANGE S IN THE EAASR AND STA 3/4 SYSTEMS The results of the analysis of the SFWMM output and the water quality studies to estimate the outflow concentrations from th e EAASR and STA 3/4 ar e provided in this chapter. Each scenario is evaluated on a steady-state long-term basis using the UF WQDT. Key insights are discussed for each scenario and summarized at the conclusion of the section. The annual variability in the SC STA scenario is evaluated to provide an indication of the variability in the 36 year da ta set. This analysis was performed using a variant of the UF WQ DT that made annual steady-stat e evaluations. Key insights on the variability of water quality are discussed. Analysis of Long-Term Averages The EAASR and STA 3/4 were evaluated on a steady-state basis for the average behavior over the 36 years of SFWMM simu lation. Results from the UF WQDT for each scenario are presented below for the one, tw o, and four compartment reservoir options. Single Compartment Reservoir The single compartment reservoir was the simplest configuration. Two scenarios were evaluated to provide a basis for fu rther comparison with the two and four compartment scenarios; the SC base Scenar io (Figure 4-3) and the SC STA scenario (Figure 4-4). The depth of each scenario was va ried to 9 feet and 12 feet from 5.9 feet to illustrate the effects of depth on the overall performance of the reservoir and STA 3/4. STA 3/4 received water directly from the NNR and Miami Canals in the SC base

PAGE 84

73 scenario. In the SC STA scenario, STA 3/4 inflows were first rout ed through the single compartment reservoir (Figure 44), and therefore the inflow c oncentration is equal to the EAASR outflow concentration. The results of the above scenarios are presented for the EAASR and STA 3/4 in Tables 5-1 and 5-2, respectively. Table 5-1. Single Compartment EAASR Results for Total Phosphorus with Variable Depths Scenario Area (acres) Inflow (ac-ft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) Percent Removal SC base 30,720 1,686 5.9 108 0.105 0.04062% SC base @ 9 ft. 30,720 1,686 9.0 164 0.105 0.03170% SC base @ 12 ft. 30,720 1,686 12.0 219 0.105 0.02874% SC STA 30,720 2,705 5.9 67 0.101 0.05149% SC STA @ 9 ft. 30,720 2,705 9.0 102 0.101 0.04060% SC STA @ 12 ft. 30,720 2,705 12.0 136 0.101 0.03466% Table 5-2. STA 3/4 Results for the Single Compartment EAASR for Total Phosphorus with Variable Depths Scenario Area (acres) Inflow (ac-ft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) Percent Removal SC base 17,9202,2012.4 29 0.066 0.021 69% SC base @ 9 ft. 17,9202,2012.4 29 0.061 0.020 67% SC base @ 12 ft. 17,9202,2012.4 29 0.060 0.020 66% SC STA 17,9202,2012.4 29 0.051 0.020 61% SC STA @ 9 ft. 17,9202,2012.4 29 0.040 0.020 50% SC STA @ 12 ft. 17,9202,2012.4 29 0.034 0.020 42% The results of varying the depth show that it may be possible to improve the quality of the EAASR outflow by increas ing the mean depth beyond 5.9 feet. The EAASR has a maximum depth of 12 feet. If it could be kept full, th en the HRT increases from 108 days to 219 days resulting in a si gnificant improvement in water quality. However, it is unlikely that the EAASR could be operated full since this would conflict with other purposes such as flood control and water supply. An inte rmediate increase in depth to 9 feet also shows good improveme nt in water quality and would be more feasible to attain.

PAGE 85

74 Routing all STA 3/4 inflows, the SC STA scenario, through the EAASR provides large gains in water quality when compared to the base scenario. Long-term STA 3/4 inflow concentrations were identical for the SC STA scenario and the SC base scenario at 12 feet of depth. Therefore, it may be possibl e to achieve similar gains to increasing the depth of the EAASR by routing all STA inflow through the EAASR first. The combination of increasing depth and routing STA flows achieves some additional gains in water quality. At the low concentrations of the STA, it is difficult to achieve large gains in water quality. La rge gains are more easily made by altering the EAASR operations. STA 3/4 alone could achieve nearly the same water quality as the EAASR/STA two reactor system because of the 29 day HRT. Recall from Figure 4-22, the outflow TP concentration at an HRT of 29 for STA 3/4 is approaching the bac kground level. This suggests that the HRT may be under used a nd more inflow could be accommodated to reduce its HRTs to about 10-15 days, without a significant decrease in load reduction. Two Compartment Reservoir The two compartment reservoir differs from the single compartment case mainly due to the interaction between the two compar tments and the elimination of direct flow from the Miami Canal. The base scenario wa s modeled as presented (Figure 4-5). In the TC STA Inflow scenario, the direct NNR canal flows to STA 3/4 were routed through the reservoir before being released to the STA (F igure 4-6). The results of the analyses are shown in Tables 5-3 through 5-5. The TC base scenario EAASR achieved good removal, though lower than the SC base scenario. Ho wever, the STA achieved identical outflow concentrations in the SC and TC base scenar ios. The increased loading in the TC STA scenario decreased the EAASR performance, as expected. However, the STA inflow

PAGE 86

75 concentration was less than the TC base s cenario due to mixing and treatment in the EAASR. The STA outflow concentration TC STA scenario was higher than the SC STA scenario, reinforcing the findings presented in the above section. The higher concentrations in Compartment 1 as compared to the SC scenarios were mainly caused by reduced treatment due to the higher loading for Compartment 2. An outflow from C2 to STA 3/4 was incl uded in the TC Miami and TC Miami STA (Figures 4-7 and 4-8), and the outcome of the resulting scenario analyses are presented in Tables 5-3 through 5-5. The resulting re duction in load increased the inflow concentration and treatment in Compartment 1, resulting in little gains. The outflow from C1 decreased the inflow concentration of STA 3/4 and therefore resulted in lower outflow concentrations. From these results, it is believed that the a ddition of an outflow from Compartment 2 to STA 3/4 may allow additional water quality benefits, due to operational flexibility and mixing. The water quality gains were evaluated on a mass basis (Table 5-6). The total removal from the system was extremely similar between scenarios, with the exception of TC Miami. However, routing the STA infl ows through the reservoir shifted the removal from the STA to the reservoir. These findi ngs are consistent with those for the single compartment scenarios. The lower removal in the case of TC Miami is largely a factor of less mass to be removed in the STA. Table 5-3. Compartment 1 Result for the Two Compartment EAASR Configuration Scenario Area (acres) Inflow (ac-ft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) Percent Removal TC base 15,360 2,0115.9450.078 0.051 35% TC STA 15,360 2,4955.9370.084 0.058 31% TC Miami 15,360 9885.9920.112 0.045 60% TC Miami STA 15,360 1,4725.9620.111 0.057 49%

PAGE 87

76 Table 5-4. Compartment 2 Result for the Two Compartment EAASR Configuration Scenario Area (acres) Inflow (ac-ft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) Percent Removal TC base 15,360 1,2845.9700.0870.045 48% TC STA 15,360 1,2845.9700.0870.045 48% TC Miami 15,360 1,2845.9700.0870.045 48% TC Miami STA 15,360 1,2845.9700.0870.045 48% Table 5-5. STA 3/4 Result for the Tw o Compartment EAASR Configuration Scenario Area (acres) Inflow (ac-ft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) Percent Removal TC base 17,920 2,2012.4190.0640.020 76% TC STA 17,920 2,2012.4190.0580.020 77% TC Miami 17,920 2,2012.4190.0480.020 77% TC Miami STA 17,920 2,2012.4190.0520.020 77% Table 5-6. TP Removal by Mass in Kilograms per Day Scenario C1 C2 EAASR Total STA 3/4 System Total TC base 67 67134119253 TC STA 80 67147104250 TC Miami 82 6714877225 TC Miami STA 98 6716586251 Four Compartment Reservoir Two scenarios were analyzed for the four compartment (FC) EAASR configuration. The FC base scenario was modeled as pres ented in Figure 5.9. The FC STA scenario was modeled as presented in Figure 5-10. The results for each compartment are presented in Tables 5-7 th rough 5-11. Results from the analysis are similar to the two compartment configurati on, therefore little wa ter quality gains were realized from the increased compartmentalization. Table 5-7. Compartment A Result for the Four Compartment EAASR Configuration Scenario Area (acres) Inflow (acft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) % Removal FC base 5,120 5898.3720.1310.059 55% FC STA 5,120 5898.3720.1330.060 55%

PAGE 88

77 Table 5-8. Compartment B Result for the Four Compartment EAASR Configuration Scenario Area (acres) Inflow (acft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) % Removal FC base 10,240 7884.8620.0850.058 32% FC STA 10,240 1,2724.8390.0880.063 28% Table 5-9. Compartment C Result for the Four Compartment EAASR Configuration Scenario Area (acres) Inflow (acft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) % Removal FC base 5,120 5258.5830.1020.046 46% FC STA 5,120 5258.5830.1020.050 43% Table 5-10. Compartment D Result for the Four Compartment EAASR Configuration Scenario Area (acres) Inflow (acft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) % Removal FC base 10,240 5704.5820.0670.047 45% FC STA 10,240 1,1054.5420.0680.048 46% Table 5-11. STA 3/4 Result for the Four Compartment EAASR Configuration Scenario Area (acres) Inflow (acft/d) Depth (ft) HRT (days) Cin (mg/L) Cout (mg/L) % Removal FC base 17,920 2,2012.4190.0530.020 77% FC STA 17,920 2,2012.4190.0560.021 70% The water quality gains were evaluated on a mass basis (Table 5-12). The total removal from the system was virtually identi cal between scenarios (T able 5-11). Routing the STA inflows through the reservoir shifte d the majority of the removal from the STA to the reservoir. Table 5-12. TP Removal by EAASR and ST A 3/4 in Mass in Kilograms per Day Scenario A B C D EAASR Total STA 3/4 System Total FC base 52 26 361412990281 FC STA 53 39 342715395280 Annual Variability in Performa nce for the SC STA Scenario As shown in long-term analysis, the overa ll performance does not vary greatly for the one, two, and four compartment cases. Als o, it is desirable to direct inflows through the EAASR before sending them to STA 3/4. Thus, the SC STA scenario (Figure 4-4)

PAGE 89

78 will be used to estimate the annual variability in performance due to variability in inflow rates and operating depths. The annual variability in the SC STA scenario for the EAASR and for STA 3/4 are reported in Table 5-13 and Table 5-14, respectively. On an annual basis, the reservoir varies over a large range of flows and dept hs. The annual average inflows varied from 695 to 5,662 acre feet/day. The mean annual depths ranged from 0.7 to 11.9 feet. The resulting calculated HRTs are between 23 and 113 days. The inflow concentration to the EAASR was relatively stable and varied from 0.084 to 0.135 mg/L. The calculated outflow concentrations reflected the HRT and Cin characterizations and varied from 0.042 to 0.092 mg/L. The associated percent removals varied from as low as 24% to a high of 69%. On an annual basis, STA 3/4 received inflows ranging from 253 to 5,540 acre feet per day. Average depths ranged from 1.6 to 3.1 feet. The resulting cal culated HRTs were between 10 and 116 days. The inflow con centration varied widely from 0.042 to 0.092 mg/L. The calculated outflow concentra tions ranged from 0.019 to 0.029 mg/L.

PAGE 90

79 Table 5-13. Annual EAASR Variability for the Single Compartment STA Scenario Year EAASR Qin EAASR HRT EAASR Depth EAASR Cin EAASR Cout Percent Removal (ac-ft/d) (days) (ft) (mg/L) (mg/L) (%) 1965 2,104 56 3.83 0.113 0.061 46% 1966 3,898 73 9.32 0.097 0.048 51% 1967 1,641 70 3.74 0.118 0.056 53% 1968 3,471 62 6.97 0.104 0.055 47% 1969 4,236 70 9.64 0.096 0.049 50% 1970 4,942 70 11.18 0.088 0.046 48% 1971 1,616 91 4.79 0.126 0.049 61% 1972 1,294 118 4.96 0.134 0.042 69% 1973 1,074 97 3.41 0.127 0.047 63% 1974 1,789 50 2.92 0.117 0.066 43% 1975 1,755 73 4.15 0.124 0.056 55% 1976 1,220 105 4.17 0.126 0.044 65% 1977 1,564 112 5.69 0.128 0.043 67% 1978 2,608 92 7.80 0.107 0.044 59% 1979 5,111 41 6.85 0.089 0.058 34% 1980 1,749 67 3.81 0.108 0.054 50% 1981 992 46 1.50 0.135 0.077 43% 1982 3,354 63 6.93 0.107 0.055 49% 1983 5,290 57 9.75 0.088 0.051 42% 1984 3,643 62 7.39 0.088 0.048 45% 1985 1,505 62 3.04 0.122 0.061 50% 1986 1,701 113 6.27 0.124 0.041 66% 1987 1,671 66 3.57 0.114 0.056 51% 1988 1,757 44 2.53 0.099 0.061 38% 1989 937 23 0.69 0.122 0.092 24% 1990 695 54 1.22 0.135 0.072 47% 1991 3,670 53 6.35 0.095 0.055 42% 1992 3,420 59 6.60 0.097 0.053 45% 1993 3,548 68 7.87 0.093 0.048 48% 1994 3,618 63 7.45 0.105 0.054 48% 1995 3,914 93 11.91 0.090 0.040 56% 1996 4,118 75 10.09 0.086 0.044 49% 1997 2,556 78 6.45 0.100 0.047 53% 1998 5,662 38 7.09 0.084 0.057 32% 1999 3,214 66 6.95 0.093 0.049 48% 2000 2,057 84 5.60 0.101 0.045 55% Average 2,705 67 5.90 0.101 0.051 49% COV 0.51 0.33 0.47 0.16 0.21 NA

PAGE 91

80 Table 5-14. Annual STA 3/4 Variability fo r the Single Compartment STA Scenario Year STA 3/4 Qin STA 3/4 HRT STA 3/4 Depth STA 3/4 Cin STA 3/4 Cout Percent Removal (ac-ft/d) (days) (ft) (mg/L) (mg/L) (%) 1965 1,062 29 1.72 0.061 0.020 67% 1966 3,686 14 2.80 0.048 0.024 49% 1967 1,250 30 2.07 0.056 0.020 64% 1968 2,818 16 2.53 0.055 0.024 56% 1969 3,612 14 2.83 0.049 0.024 50% 1970 4,726 12 3.17 0.046 0.025 45% 1971 1,042 36 2.07 0.049 0.020 60% 1972 1,149 33 2.10 0.042 0.020 53% 1973 521 57 1.64 0.047 0.019 59% 1974 1,494 25 2.08 0.066 0.021 68% 1975 1,224 28 1.92 0.056 0.020 64% 1976 687 51 1.96 0.044 0.019 56% 1977 598 56 1.86 0.043 0.019 54% 1978 2,787 16 2.53 0.044 0.023 49% 1979 4,293 12 2.84 0.058 0.028 52% 1980 1,673 24 2.27 0.054 0.021 61% 1981 441 64 1.57 0.077 0.019 75% 1982 2,224 19 2.34 0.055 0.023 59% 1983 5,161 11 3.14 0.051 0.027 46% 1984 3,054 15 2.60 0.048 0.024 51% 1985 896 40 2.01 0.061 0.020 68% 1986 1,182 31 2.02 0.041 0.020 52% 1987 1,031 35 2.04 0.056 0.020 65% 1988 1,487 25 2.11 0.061 0.021 66% 1989 532 61 1.82 0.092 0.019 79% 1990 253 116 1.63 0.072 0.019 73% 1991 3,149 15 2.57 0.055 0.025 55% 1992 2,584 17 2.49 0.053 0.023 56% 1993 3,111 16 2.78 0.048 0.023 52% 1994 2,585 19 2.67 0.054 0.023 58% 1995 3,555 16 3.08 0.040 0.022 44% 1996 4,071 13 3.03 0.044 0.024 45% 1997 1,428 27 2.19 0.047 0.020 57% 1998 5,540 10 3.24 0.057 0.029 49% 1999 2,384 18 2.45 0.049 0.022 54% 2000 1,957 21 2.29 0.045 0.021 53% Average 2,201 29 2.35 0.051 0.022 57% COV 0.65 0.73 0.20 0.21 0.12 NA

PAGE 92

81 Extensive operating information and guideli nes have been developed for STAs in South Florida. For example, the draft opera tions plan for STA-1E indicates an operating depth range between 0.5 and 4.5 feet (Goforth 2006). Annual depth results (Table 5-14) indicate that the STA is operated in this range. The annual summaries of the operation of the EAASR and STA 3/4 over the 36 year period are summarized in cumulative de nsity functions (CDFs) (Figure 5-1) that indicate the % <= indicated value as a functi on of variables of interest. As shown in Figure 5-1, the mean annual inflows of the EAASR and STA 3/4 follow a similar pattern, which was expected as all STA inflows are first routed through the EAASR. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 01,0002,0003,0004,0005,0006,000 Annual inflow, af/day%<= indicated value EAASR STA 3/4 Figure 5-1. Mean Annual Inflows from SC STA Scenario A clear difference in operating depths is shown in the CDFÂ’s for EAASR and STA 3/4 (Figure 5-2). The depth profile of the EAASR is nearly linear 2:1 slope

PAGE 93

82 throughout the entire range of the design de pth. The STA operates over a relatively narrow range from 1.5 to 3.5 feet as stated previously. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 024681012 Mean depth, ft.%<= indicated value EAASR STA 3/4 Figure 5-2. Mean Annual Depths from SC STA Scenario The HRT profile reflects the differen ce in depth profiles (Figure 5-3). The EAASR operates under a large range of HRTÂ’s. The HRT distribution shows STA 3/4 operating in the within the nor mal range of STAs, 14 to 21 days, for approximately 19% of the years for <= 14 days and 50% of th e years for <= 21 days. The remaining 50 percent of HRT are above this range, again indicating the ST A may be able to receive more hydraulic loading.

PAGE 94

83 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 020406080100120140 Hydraulic residence time, days%<= indicated value EAASR STA 3/4 Figure 5-3. Mean Annual HRT from SC STA Scenario The reductions of TP concentrations through the EAASR/STA 3/4 system were observed (Figure 5-4). Median (50%) inflow TP concentration to the EAASR is about 0.110 mg/L. This TP concentration is redu ced to a median of about 0.050 mg/L as it leaves the EAASR and enters STA 3/4. It ex its STA 3/4 at a TP concentration of about 0.020 mg/L. As the treatment progresses through the EAASR and STA, the variability in concentration decreases as shown by the increas ingly vertical CDF curves. This is an indication of a functioning storage/treatment system.

PAGE 95

84 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0.050.0100.0150.0 TP concentration, ppb%<= indicated value EAASR Cin EAASR Cout STA 3/4 Cout Figure 5-4. Mean Annual Inflow and Outflow Concentrations from SC STA Scenario Summary and Conclusions The overall results for the three confi gurations are shown in Table 5-15. The input TP concentrations vary from 0 .067 to 0.133 mg/L depending on the number of compartments. Similarly, the output con centrations from the EAASR indicate a significant treatment effect with the c oncentrations reduced to 0.028 to 0.063 mg/L depending on the configuration. However, th e variability in perf ormance is greatly dampened in the STA 3/4 outflow with the outfl ow concentrations all very similar in the 0.020-0.021 mg/L range. The EAASR has a signif icant water quality impact because its hydraulic residence times exceed two months. The residence times in STA 3/4 are also high for an STA with a calculated mean of about 29 days. Thus, the combined system has ample time to store and treat the water. The EAASR can also provide a significant ability to manage the inflows to STA 3/4 in order to optimize its performance. From an operational point of view, it is desirable to maximize the amount of inflow that can be passed through the EAASR before entering STA 3/4.

PAGE 96

85 Table 5-15. Summary Results for the One, Two, and Four Compartment Scenarios Compartments EAASR EAASR STA Cin (mg/L) Cout (mg/L) Cout (mg/L) 1 0.101-0.105 0.028-0.051 0.020-0.21 2 0.078-0.112 0.045-0.058 0.020 4 0.067-0.133 0.046-0.063 0.020-0.021 The analysis of annual vari ability of the SC STA scen ario found a similar inflow pattern for the EAASR and STA 3/4. The depth profiles were clearly different, with the EAASR fluctuating through out the entire rang e of design depths and STA 3/4 remaining relatively constant. The vari ation in concentration decrease d as the water was treated by the EAASR/STA 3/4 system. This indicated a functioning storage/ treatment system. The multiple scenarios were compared and the following conclusions were drawn. The compartmentalization of the reservoir can pr ovide additional operational flexibility to manage the mixing of influents and treatment times to achieve additional water quality improvements. An additional outflow stru cture from Compartment 2 to STA 3/4 can increase the operational flexibility of the EAASR and may increase water quality improvements. Operating the EAASR at grea ter depths may further increase the water quality treatment. The EAASR does not sign ificantly improve the outflow water quality from STA 3/4 as compared to using STA 3/4 only because of the relatively long residence times in STA 3/4. This suggests th at the current inflows to STA 3/4 could be increased to improve water quality even mo re in the EAASR/STA 3/4 storage/treatment train. The actual performance of the EAASR/STA system could vary widely from the predictions presented in this thes is for several reasons including: Inflow quantities and water quality can be expected to vary over the next 50 years

PAGE 97

86 The SFWMM estimates of inflows and ope rations are not based on any kind of optimization for the EAASR/STA system but re present an estimate of their role in a regional water management scenario. The behavior and performance of the EAASR/STA can be expected to vary depending on how it is operated for fl ood control and water supply purposes as well as water quality enhancement.

PAGE 98

87 CHAPTER 6 SUMMARY AND CONCLUSIONS This effort aimed to create and use a model to estimate water quality in the EAASR and Stormwater Treatment Area (S TA) 3/4. The model was designed to function in the dynamic and challenging E AASR design process. Key insights on the most current EAASR and STA 3/ 4 configurations were found. Chapter 2 presented the background and previous work on CERP, the EAASR, systems engineering, and reservoir modeling. CE RP is an ambitious project to “get the water right” in the South Florida Ecosystem a nd restore the Everglades. The project goal will be accomplished by all 63 projects operating together. The Acceler8 project streamlines the CERP planning, design, and cons truction process to provide benefits to the system more quickly and cost-effici ently. The EAASR, a CERP and Acceler8 project, will provide storag e for water supply, flood contro l, and flow equalization for water quality treatment areas. Water supply and flood control design approaches are well developed, while water quality approaches ar e less developed. The streamlining of the planning and design of the EAASR has posed se veral challenges resulting in an iterative design process with multiple conceptual design alternative formulations and analyses. Due to differences in the simulation, uncertain ty exists when these multiple formulations and analyses are compared. The systems engineering approach to desi gn provides a proven approach that can meet the challenges of the EAASR planni ng and design process. The approach incorporates disciplinary mode ls that may produce a more op timal design than traditional

PAGE 99

88 approaches. A model in the water quality di scipline was therefore sought to assess which of several alternatives are better. Simple empirical and mass balance models were found that can be used for the EAASR. Comparab le lake and reservoir systems were found to provide parameter estimates for EAASR. Due to the wide range of possible operational conditions and planning and design challe nges, a water quality model developed specifically for the EAASR was be necessary. The water quality model must therefore incorporate measures to meet the planning and design challenges of the EAASR project. Chapter 3 presents several iterations of EAASR configurations that have been developed throughout the projec t. Early configurations used a combination of Components A, B, and C of Figure 2-2. Configur ations were then formulated using only Component A. The sizing of the reservoir was evaluated first. Thr ee configurations were developed to evaluate the volume of the rese rvoir. Four additional alternatives were developed to evaluate the depth of the reservoir. A 12 foot deep, 360,000 acre-foot reservoir was subsequently decided upon. Thr ee configurations were then developed to assess the compartmentalization of the reservoi r. The configurations were developed for a single compartment, two compartments, and four compartments. The location of the compartments affected the source of water received for both the EAASR and STA 3/4. Unlike early configurations, internal transfer s between the compartments occurred from multiple compartments. The review of the EAASR configurations was performed to provide the context in which the water quality model is developed. To use these configurations, a water quality model should include depth, area, and compartm entalization with varying area. Due to the inter-reservoir transfers, the model must be able to simulate reservoirs in series, with

PAGE 100

89 feedback loops. Additionally, the ability to include multiple inflows and outflows would be useful. Chapter 4 details the water quality model, the UF WQDT, and water quantity and quality inputs. The UF WQDT is a stea dy-state mass balance model utilizing the KC* kinetic model and was developed to simulate the TP water quality of the EAASR and STA 3/4. The model uses a single aggr egate inflow and concentration, depth, background concentration, and a volumetric reaction rate to simulate TP. The model calculates the HRT and HLR of each reactor, as well as th e outflow concentration and percent removal. Long-term average simulations were determined to be the most appropriate level of sophistication needed in order to provide key insights for the dynamic EAASR design. Water quantity and quality inputs to the m odel were calculated for the base and STA scenarios for each configuration. Th e HRT of the EAASR ranged from 39 to 107.5 days. The EAASR was not used for flow equalization. The HRT of STA 3/4 was calculated to be 29 days for al l scenarios. Mean inflow c oncentrations of 17 parameters of interest for the four external inflow sources to the reservoir and STA 3/4 were reported. Parameter estimations were perfor med using closely comparable system with high quality data. Lake Istokpoga was found to be the best comparable for the general single compartment scenario of the EAASR. Data from 16 STAs were used as comparables to STA 3/4. The following water quality parameters estimates for the EAASR and STA 3/4 were selected: C* = 0.0254 mg/L for the reser voir and 0.0194 mg/L for the STA k = 0.016 day-1 for the reservoir and 0.127 day-1 for the STA

PAGE 101

90 A sensitivity analysis of the EAASR react ion rate and background concentration was performed for the SC STA scenario. Th e calibration of each parameter was found to significantly affect results of the analysis. Particularly reaction ra tes less than 0.05 days-1 for a background concentrati on of 0.025 mg/L. To reproduce TP removal estimates provided in the BODR an reaction rate of 0.0033 days-1 for a C* of 0.025 is needed. Due to the differences in configur ation and reaction rate the mode ling efforts are not directly comparable. An analysis of the predicted performa nce of the EAASR and STA 3/4 systems was performed. STAs were found to provide si gnificantly greater water quality treatment than reservoirs. The STA systems will approach the background concentration asymptotically for retention times greater than 25 days. However, since the STA is operated at much lower depths than the reser voir, they will have much lower HRT for a given area and inflow. The init ial concentration can affect the outcome of the EAASR or STA 3/4, especially if measured by the percent removal of TP. Chapter 5 presented the results and conclu sions of the UF WQDT analysis of the scenarios described in Chapter 4. The overa ll results for the three configurations are shown in Table 5-15. The input TP con centrations vary from 0.067 to 0.133 mg/L depending on the number of compartments. Simi larly, the output concentrations from the EAASR indicate a significant trea tment effect with the concen trations reduced to 0.028 to 0.063 mg/L depending on the configuration. Howe ver, the variability in performance is greatly dampened in the STA 3/4 outflow with the outflow concentrations all very similar in the 0.020-0.021 mg/L range. The EAASR has a significant water quality impact because its hydraulic residence times exceed two months. The residence times in STA

PAGE 102

91 3/4 are also high for an STA with a calcu lated mean of about 29 days. Thus, the combined system has ample time to store and treat the water. The EAASR can also provide a significant ability to manage the inflows to STA 3/4 in order to optimize its performance. From an operational point of view, it is desirable to maximize the amount of inflow that can be passed through the EAASR before entering STA 3/4. The analysis of annual vari ability of the SC STA scen ario found a similar inflow pattern for the EAASR and STA 3/4. The depth profiles were clearly different, with the EAASR fluctuating through out the entire rang e of design depths and STA 3/4 remaining relatively constant. The EAASR and STA 3/4 remove roughly equal portions of the median concentrations. The variation in con centration decreased as the water was treated by the EAASR/STA 3/4 system. This indicated a functioning storag e/treatment system. The multiple scenarios were compared and the following conclusions were drawn. The compartmentalization of the reservoir can pr ovide additional operational flexibility to manage the mixing of influents and treatment times to achieve additional water quality improvements. An additional outflow stru cture from Compartment 2 to STA 3/4 can increase the operational flexibility of the EAASR and may increase water quality improvements. Operating the EAASR at grea ter depths may further increase the water quality treatment. The EAASR does not sign ificantly improve the outflow water quality from STA 3/4 as compared to using STA 3/4 only because of the relatively long residence times in STA 3/4. This suggests th at the current inflows to STA 3/4 could be increased to improve water quality even mo re in the EAASR/STA 3/4 storage/treatment train.

PAGE 103

92 The actual performance of the EAASR/STA system could vary widely from the predictions presented in this thes is for several reasons including: Inflow quantities and water quality can be expected to vary over the next 50 years The SFWMM estimates of inflows and ope rations are not based on any kind of optimization for the EAASR/STA system but re present an estimate of their role in a regional water management scenario. The behavior and performance of the EAASR/STA can be expected to vary depending on how it is operated for fl ood control and water supply purposes as well as water quality enhancement. Steady state modeling of long-term average outflow concentrations was performed to provide key insights on the co mpartmentalization and general operations of the EAASR and STA 3/4. Further work is n eeded to assess the operations EAASR and STA 3/4 in more detail. The results of this study should be calibra ted against daily time step simulations of the operation of the syst em. The daily simulation should be guided by an optimizer that directs the simulator towards optimal operating policies. This simulation/optimization strategy needs to incorporate a life cycle analysis and the existing as well as proposed constraints on multi-purpose operation that are embedded in the SFWMM. The life cycle evaluations of the performance of the EAASR/STA system should include explicit consideration of the n eed for periodic maintenance and removal of settled materials.

PAGE 104

93 LIST OF REFERENCES Braha, D. and Maimon, O. (1998). A Mathematical Theory of Design: Foundations, Algorithms, and Applications Kluwer Academic Publishers, Boston, MA. Buede, D.M. (1999). The Engineering Design of Systems: Models and Methods WileyInterscience, New York. Burns and McDonnell Compa ny Inc. (B&M) (2004a). Water Quality Impacts of Reservoirs Task 3 – Analysis of Data Set (March 30, 2005). Burns & McDonnell, Inc. (2004b). Water Quality Impacts of Reservoirs Task 2 – Identification of Analysis of Data Sites and Data Acquisition (March 30, 2005). Central and Southern Florida Project (2005). Comprehensive Everglades Restoration Plan Report to Congress 2005, (Ma rch 21, 2006). Chapra, S. C. (1996). Surface Water-Quality Modeling McGraw-Hill, Boston, MA. Chapra, S. C. and Auer, M. T. (1999). Phosphorous Biogeochemistry in Subtropical Ecosystems (Chapter 28 in Reddy, K.R., O’Conn or, G.A., and Schelske, C.L., Ed., Management Models to Evaluat e Phosphorous Loads in Lakes ), Lewis Publishers, Boca Raton. CRC for Catchment Hydrology (2005). “MUSIC Model.” Introduction, (July 30, 2005). Goforth, G. (2006). Interim Operation Plan Stormw ater Treatment Area 1 East Final Draft to SFWMD, West Palm Beach, FL. Hazelrigg, G.A. (1996). Systems Engineering: An Approach to Information-Based Design Prentice-Hall, Upper Saddle River, NJ. Kadlec, R. H. and R. L. Knight. (1996). Treatment Wetlands CRC, Boca Raton, FL.

PAGE 105

94 Kane, K. (1999). Proposed Phase II Phosphorus TMDL Calculations for Rondout Reservoir, New York City Department of Environmental Protection. Knight, S. et al. (2006). “A Cohesive Approach for the Implementation of the Comprehensive Everglades Restoration Plan.” Proc., ASCE EWRI 2006 Conference Omaha, NE, May, in press. Kusiak, A. (1999). Engineering Design: Products, Processes, and Systems, Academic Press, San Diego, CA. Lee, J.G., Heaney, J.P., and Lai, D. (2005) “Optimization of Integrated Urban Wetweather Control Strategies.” J. Water Res Pl-ASCE Vol. 131, No. 4. Reckhow, K.H. (1979). Quantitative Techniques for the Assessment of Lake Quality EPA-440/5-79-015, 146 pp. Reckhow, K.H., Coffey, S.C., Henning, M. H., Smith, K., and Banting, R. (1992). EUTROMOD: Technical Guidance and Spreads heet Models for Nutrient Loading and Lake Eutrophication Draft, Duke University, Durham, N.C. Reddy, K.R., O’Connor, G.A., and Schelske, C.L. (1999). Management Models to Evaluate Phosphorous Loads in Lakes, Lewis Publishers, Boca Raton. Reisinger et al. (2006). “Findi ng the Best Comparable Dataset for Estimating Parameters for a Water Quality Model”. Proc., ASCE EWRI 2006 Conference Omaha, NE, May, in press. South Florida Water Management District (2005a). “South Florida Water Management Model Fact Sheet,” (March 16, 2005). South Florida Water Management District (SFWMD). (2005b). Minimum Flows and Levels for Lake Istokpoga – Final Draft (January 4, 2006). South Florida Water Management District (2006a). “Overview – Acceler8 – Everglades Restoration Now,” (March 21, 2006).

PAGE 106

95 South Florida Water Management District (SFWMD) (2006b). Basis of Design Report Final (March 20, 2006). U.S. Army Corps of Engineers and South Fl orida Water Management District (2002). Project Management Plan Everglades Agricultural Area Storage Reservoirs Phase 1 (January 23, 2005). U.S. Army Corps of Engineers and South Fl orida Water Management District (2003). Water Quality Modeling, Reservoir P hosphorous Uptake Model Everglades Agricultural Area Storage Reservoirs Phase 1 (March 30, 2006). U.S. Army Corps of Engineers and South Fl orida Water Management District (2004a). Conceptual Alternatives E verglades Agricultural Area Storage Reservoirs Phase 1 (January 23, 2005). U.S. Army Corps of Engineers and South Fl orida Water Management District (2004b). Screening of Conceptual Alternatives Everglades Agricultural Area Storage Reservoirs Phase (January 23, 2005). U.S. Army Corps of Engineers and South Fl orida Water Management District (2005). Draft Integrated Project Implementa tion Report and Environmental Impact Statement (September 30, 2005). Vollenweider, R.A (1976). “Possibilities and Limits of Elementary models Concerning the Budget of Substances in lakes. ” Archives Hydrobiologia 66(1):1-36. Walker, W.W. (2000). "Estimation of a Phosphorus TMDL for Lake Okeechobee.” Datasets http://wwwalker.net/dmsta/reservoi rs/index.htm.>(Accessed January 4, 2003) Walker, W.W., and Havens, K. (2003). "Dev elopment & Application of a Phosphorus Balance Model for Lake Istokpoga, Florida" Lake & Reservoir Management, Vol. 19, No. 1, pp. 79-91.

PAGE 107

96 Walker, W.W. (2003). BATHTUB Empirical Mode ls for Lake & Reservoir Eutrophication Environmental Lab, Waterways Experiment Station, US Army Corps of Engineers. Walker, W. W. and Kadlec, R. ( 2005a). “DMSTA 2 for Reservoirs,” Overview, (June 1, 2005). Walker, W. W. and Kadlec, R. (2005b). “D MSTA 2 for Reservoirs – DMSTA Case Index,” Datasets (June 1, 2005). Walker, W.W. and Kadlec, R. ( 2006). "DMS TA 2 for Reservoirs Calibration Dataset for a Phosphorous Cycling Model,” Dataset (January 28, 2006). Wetland Solutions, Inc. (WSI) (2003). “Lake Okeechobee Watershed Project Calibration of DMSTA Model for Use North of Lake Okeechobee.” Datasets (January 4, 2003).

PAGE 108

97 BIOGRAPHICAL SKETCH On January 23, 1982, I was born and named Daniel L. Reisinger in Southern Florida. I attended public schools throughout my childhood and graduated from Cooper City High School in the year 2000. I attended the University of Florida and was awarded a Bachelor in Science degree from the College of Engineering, Environmental Engineering and Sciences department. Af ter obtaining my undergraduate degree, I remained at the University of Florida to obt ain a Master of Engineering degree from the College of Engineering, Environmental Engineering and Sciences department.