United States
Environmental
Protection Agency
Office of Water
(4601)
EPA815-C-99-002
September 1999
ANALYSIS OF GAC EFFLUENT
BLENDING DURING THE
ICR TREATMENT STUDIES

-------
This page intentionally left blank.

-------
Foreword

Between 1997  and 1999, 98  public water  systems conducted treatment studies to evaluate
disinfection  byproduct (DBF) precursor removal  performance of granular activated  carbon
(GAC)  and  membranes.  The treatment study requirement was  a  part of the Information
Collection Rule (ICR) for Public Water Systems, Subpart M of the National Primary Drinking
Water Regulations, § 141.141(e).

Sixty-two public water systems evaluated GAC. The ICR required that DBF precursor removal
be evaluated in the effluent of a single contactor as a function  of run time,  to  assess the
breakthrough of DBF precursors as the GAC was exhausted.  In practice, full-scale  plants can
reduce carbon usage  rates by blending the effluents of multiple parallel  contactors prior to
disinfection. When the treatment objective is reached in the blended effluent, the contactor with
the "oldest" GAC is  taken  off-line  and replaced  by a contactor  with fresh  GAC, and  this
replacement occurs at regular intervals.  GAC effluent blending extends the  service life of each
contactor because  water from contactors that exceed  the treatment objective is blended with
water from contactors with fresher GAC that have  effluent concentrations below the treatment
objective.  The treatment objective must only be maintained in the blended contactor effluent.

A primary goal during analysis of the treatment study results by the USEPA will be to estimate
blended contactor run times to meet target regulatory treatment objectives.  This information can
be used to  estimate GAC treatment costs that reflect  full-scale effluent blending.  This study
provides the background and foundation for the analytical tools used to analyze treatment study
data, assessing the applicability and limitations of these tools.

One objective  of  this study was to evaluate mathematical models for representing  single
contactor breakthrough data.  From a data management perspective, model  parameters will be
easier to manage than the entire experimental data sets comprising 8,000 to 9,000 breakthrough
curves generated by the 62 GAC treatment studies.  A best-fit curve  also facilitates interpolation
and extrapolation of the  experimental data.  Furthermore, a  function that describes  the  single
contactor experimental data set is a prerequisite for calculating the integral breakthrough curve, a
tool for predicting blended contactor run times.

A second objective of this study was to evaluate and compare two approaches for predicting the
integral  breakthrough curve.   The first approach,  based  on application of the average value
function to the single contactor breakthrough curve, has been presented by previous researchers.
The second  approach evaluated was  a  new computationally-simpler method developed by the
treatment study technical work group (TS-TWG).

These two objectives were applied to experimental results from bench-scale  GAC runs on eight
water sources. Analytes evaluated included DBF surrogates, DBF class sums, and DBF species,
yielding an  extensive experimental  matrix  for a  thorough  evaluation of  model results.   In
addition, GAC effluent blending was assessed experimentally to test model predictions of the
integral breakthrough curve.

-------
Disclaimer

Mention  of  trade  names or  commercial  products  does  not constitute  endorsement  or
recommendation for use.

-------
ANALYSIS OF GAC EFFLUENT BLENDING DURING THE ICR
                   TREATMENT STUDIES
                         Prepared for:
    UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
       OFFICE OF GROUND WATER AND DRINKING WATER
         STANDARDS AND RISK MANAGEMENT DIVISION
                    Technical Support Center
                  26 W. Martin Luther King Drive
                     Cincinnati, Ohio 45268
                         Prepared by:
             INTERNATIONAL CONSULTANTS, INC.
                  4134 Linden Avenue, Suite 200
                      Dayton, Ohio 45432
                  SUMMERS & HOOPER, INC.
                       6 Knollcrest Drive
                     Cincinnati, Ohio 45237
            Under ICI Contract 68-C-98-051, Task Order 4

-------
This page intentionally left blank.
           -VI-

-------
Acknowledgments

This document was prepared by the United States Environmental Protection Agency, Office of
Ground Water  and  Drinking  Water, Standards and Risk  Management Division,  Technical
Support Center.  The Task  Order Project  Officer was Steven Allgeier.  The Contract Project
Officer was Phyllis Branson.

Technical consultants played a significant role in the research performed and the preparation of
this document.  This task was conducted  jointly by International  Consultants,  Inc.  (ICI) and
Summers & Hooper, Inc. (S&H), under ICI Contract 68-C-98-051, Task Order 4.   Timothy
Soward and Christopher Hill acted as Project Managers for ICI.  The S&H project team was led
by Stuart Hooper, who acted as the Technical Project Leader.  Mr.  Hooper was responsible for
technical direction and led the data analysis effort.  Vanessa Hatcher (S&H) led  and conducted
the associated experimental  work, and was assisted by Kristina Trenkamp and Carrie Wyrick.
Jeff Welge and Rick Song of ICI performed mathematical modeling and statistical analyses, and
provided statistical guidance when necessary. Technical review of the final report was provided
by  James Westrick  (USEPA-TSC),  Thomas Speth  (USEPA-RREL), and R. Scott Summers
(University of Colorado-Boulder).

The experimental portion of this project was performed in conjunction with treatment studies
required under  the Information Collection Rule.  The following  public water systems were
involved: Charleston Commissioners of Public Works, City of Aurora, City of Escondido, City
of Greensboro, City of Topeka,  Iowa-American Water Company, Miami-Dade Water and Sewer
Department, and Sweetwater Authority.  This work  could not have  been completed without the
use of the data generated as part of the treatment studies funded by these public water systems.
                                      -Vll-

-------
This page intentionally left blank.
           -Vlll-

-------
Executive Summary

When granular activated carbon (GAC) contactors are utilized for disinfection byproduct (DBF)
precursor control, multiple contactors can be used more efficiently when operated in parallel
with staggered GAC replacement cycles, with the effluents  of all  contactors  blended prior to
disinfection.  By doing so, individual contactors can be operated past a point at which the
effluent exceeds a given  treatment objective, because the treatment objective must only  be
maintained in the blended effluent of all contactors.

The design of this study incorporated two main goals.  The primary objectives were to evaluate
the ability  of the logistic  function to model  single contactor breakthrough curve data and to
evaluate  the  success and  limitations  of predictive models used to  determine  the integral
breakthrough curve,  a relationship between single  contactor run time and blended  contactor
water quality.  The  secondary objective of this study was to evaluate the applicability of these
models and predictive methods in the context of the Information Collection Rule (ICR) GAC
treatment study data analysis.

Full-, pilot-, and  bench-scale GAC treatment studies were performed by 62 utilities in fulfillment
of ICR requirements.  Regardless of scale, the ICR required that the effluent of single GAC
contactors be analyzed for DBF surrogate and formed DBF  breakthrough as a function of run
time, to assess the breakthrough of DBF precursors as the GAC was exhausted.  Bench-scale
GAC studies  typically examined two empty-bed  contact times (EBCTs) of 10 and 20 minutes
during each of four quarterly studies to account for seasonal variability in source water quality.
Pilot-scale  GAC studies were typically composed of one to  two  sessions including 10 and  20
minute EBCT contactors. Thus, a large amount of data was generated and will  be analyzed:  the
62 GAC treatment studies performed will yield a total of 8,000 to  9,000 individual breakthrough
curves.

The logistic function has  previously been used to model GAC breakthrough curves and is a
suitable  model  due  to the characteristic  'S' shape  of most breakthrough  curves.   Three
modifications to  the logistic function were developed to improve  its performance for  modeling
single contactor data.  Curve fitting involved determining which model was applicable based on
characteristics of  the breakthrough  curve, and  applying  the appropriate  model to  the
breakthrough curve for each parameter.  These enhanced forms of the logistic function model
were  able  to  successfully fit single  contactor  breakthrough  curve data for all parameters,
including DBF surrogates,  DBF sum class parameters, and individual trihalomethane (THM) and
haloacetic acid (HAA) species. A method was also employed to detect outlier data points and to
limit the influence of these deviant observations on the parameter estimates.

Two predictive approaches were compared for developing determining the integral breakthrough
curve, a relationship between operation time of each individual contactor and water quality in the
blended contactor effluent:  the direct integration (DI) method  and the surrogate correlation
approach (SCA).  The  DI method is  a time-normalized integration of the  logistic function
calculated using  the  average value function that yields the integral breakthrough curve.  The
SCA method first utilizes  the DI method to establish  an integral breakthrough  curve for total
organic carbon (TOC).  Then data points on both the single contactor and integral breakthrough
curves at a given TOC concentration  are  mapped,  and all  other water  quality parameters
                                       -IX-

-------
associated with the single contactor effluent data set at that TOC concentration are applied to the
blended effluent curve.   The SCA method is especially applicable to the  ICR data analysis
requirements because it minimizes the computations necessary to estimate blended contactor run
times to treatment objectives. An assessment of the concentration of other DBFs at any given
treatment objective will be performed as part of the data analysis effort, and the SCA procedure
is also suited to this task.  The  SCA procedure requires  that GAC breakthrough curves for all
measured parameters be represented by the logistic function model curve fit. By doing so, a
smaller amount of data are needed to represent the entire breakthrough curve experimental  data
set.  This procedure inherently relies on  the  assumption that the  relationship between  TOC
concentration and the other water quality parameters established in the single contactor effluent
is maintained in the blended contactor effluent. This study verified this assumption and found
that it was valid between TOC and other DBF surrogates, DBF class summation parameters, and
individual DBF species.  In  addition, the correlation between  TOC and bromine incorporation
factors for THMs and HAAs was  shown to be consistent between the single contactor effluent
and experimental blended effluent.

The results of the DI and SCA model predictions were compared to experimental data for eight
GAC runs  performed on  eight  water sources,  with varying pretreatments, influent  TOC
concentrations, bromide concentrations, and simulated distribution system (SDS) chlorination
conditions to evaluate DBF formation.  An analysis of the model results across all waters and
analytes showed that the prediction error for the two models was equivalent.  Both models were
biased negative, indicating a tendency to underpredict the experimental data.  The SCA model
had a slightly higher negative bias than did the DI model. Since the SCA method simplifies and
reduces the  amount of computations necessary to  estimate the integral breakthrough curve, its
use is recommended to estimate blended contactor water quality during the ICR treatment study
data analysis.

An analysis of model results for individual parameters showed that the SCA method was more
successful in predicting the integral breakthrough curve of brominated DBF species, while the DI
method was a better predictor of non-brominated DBF species breakthrough. Prediction of the
breakthrough of individual DBF species in the blended effluent is important since individual
DBFs of potential health concern will be considered during analysis of the ICR treatment study
data.  For sum parameters such  as total THM (TTHM) and the sum of five HAAs (HAAS), the
SCA method yielded results that were comparable to or superior to the DI method predictions.

Both predictive methods rely on the assumption that an infinite number of contactors are on-line
and operated in parallel-staggered mode.  This study examined this assumption and found that
the error incurred when applying run time estimates based on the infinite contactor assumption to
run times for finite numbers of contactors is  impacted  by  the number of contactors and the
magnitude  of the treatment objective  examined  in relation to the asymptotic concentration
approached by the single contactor breakthrough curve. Based on the logistic function  model of
the GAC effluent breakthrough profile, the infinite  contactor assumption will yield estimated run
times within 10  percent  of  actual run times  for  13 or  more contactors  operated in parallel-
staggered mode.  For 10 contactors on-line, the infinite contactor assumption will yield run time
estimates within 12 percent of the actual run times.  In all cases, run time estimates based on the
infinite contactor assumption are longer than those for a  finite  number of contactors,  thus
providing a best case scenario for GAC performance. The applicability of the infinite contactor
assumption in this model to finite numbers of contactors is especially important for small plants.

                                       -x-

-------
Extrapolation of the integral breakthrough curve was also examined during this study.  During
the ICR treatment study data analysis, extrapolation of some TOC integral breakthrough curves
may be necessary to estimate GAC run times when target regulatory values for DBFs or their
surrogates are exceeded upon application of the SCA method.  Extrapolation of runs performed
on two waters were compared to the same runs without extrapolation, yielding a 3 percent error
in the predicted blended effluent TOC concentration for a 21 percent run time extrapolation and
an 8  percent error in the predicted blended effluent TOC  concentration for a 61  percent
extrapolation.   The impact of extrapolation on the  SCA  procedure estimates of the  integral
breakthrough curves for other DBF surrogates and formed  DBFs was small: the mean error at
the end  of the extrapolated integral breakthrough  curve was 5  percent for a 21  percent
extrapolation and 9 percent for a 61 percent extrapolation.  Therefore, based on the waters
examined in this study, data sets that do not exceed  a given treatment objective  may be
extrapolated, and the error in predicted blended contactor water quality incurred by extrapolation
up to 50 percent of the original run time should average less than 10 percent.
                                       -XI-

-------
This page intentionally left blank.
           -Xll-

-------
Table of Contents




List of Tables	xvii



List of Figures	xix



List of Abbreviations	xxv



1   Background and Method Development	1



  1.1  Optimization of GAC Operation	2



  1.2  Modeling the Operation of Multiple Contactors Operated in Parallel-Staggered Mode .... 2



  1.3  Single Contactor Breakthrough Curve Models	6



  1.4  Direct Integration Approach	8



  1.5  Surrogate Correlation Approach	9



  1.6  Impact of Bromide Concentration on  GAC Effluent Blending Models	11



  1.7  Effluent Blending Modeling of Fewer than 10 Contactors	12



  1.8  GAC Breakthrough Curve Extrapolation	12



  1.9  ICR GAC Treatment Study Data Analysis Context	13



  1.10 Appropriateness of Model Assumptions to Full-Scale GAC Effluent Blending	14



2   Study Objectives and Approach	15



3   Materials and Methods	17



  3.1  Experimental Approach	17



    3.1.1   Rapid Small-Scale Column Test	17



      3.1.1.1 GAC Preparation Procedures	17



      3.1.1.2 RSSCT Column Setup	19



      3.1.1.3 Batch Influent Preparation	20



      3.1.1.4 RSSCT Monitoring	20



    3.1.2   Bench-Scale Blended Water Quality Assessment Approach	20



    3.1.3   DBP Formation Assessment	21



    3.1.4   Assessment of the Impact of Sampling on the Integral Breakthrough Curve	22



  3.2  Data Analysis and Modeling Approach	23



    3.2.1   Logistic Function Models	23





                                      -xiii-

-------
    3.2.2  Outlier Methods	25

    3.2.3  Direct Integration Approach	26

    3.2.4  Surrogate Correlation Approach	27

    3.2.5  Comparison of  Methods  for Predicting the Performance of  GAC  Contactors
    Operated in Parallel-Staggered Mode	28

    3.2.6  Comparison of Single Contactor and Blended Effluent DBF Bromine Incorporation.
           	28

    3.2.7  Breakthrough Curve Extrapolation	29

  3.3   Waters Examined	29

    3.3.1  Pretreatment and Water Quality	29

    3.3.2  Simulated Distribution System Chlorination Conditions	31

  3.4  Analytical  Methods	31

  3.5   Experimental QA/QC  Summary	32

4   Results and Discussion	35

  4.1   Overview	35

  4.2   Correlation between Surrogates and DBFs in Single  Contactor and Blended Contactor
  Effluents	37

    4.2.1  Correlation between Surrogate Concentration and DBF Formation	37

    4.2.2  Correlation between Surrogates and DBF Speciation	38

  4.3   Assessment of Logistic Function Fit to Single Contactor Breakthrough Curve Data	67

    4.3.1  Surrogates and Class Sum Logistic Function Curve Fits	67

    4.3.2  DBF Species Logistic Function Curve Fits	68

  4.4   Comparison of SCA  and DI Methods Used to Predict the Blended Contactor Integral
  Breakthrough Curve	89

    4.4.1  Surrogates and Class Sums	95

    4.4.2  DBF Species	96

  4.5   Analysis of Model Applicability to Finite Number of Contactors	113

  4.6   Impact of Extrapolation on Integral Breakthrough Curve Prediction	121

5   Summary and Conclusions	131

6   References	135
                                       -xiv-

-------
Appendix A: Breakthrough Curves Corrected for Impact of Sampling	137

Appendix B: SAS Code	143

Appendix C: Full- and Bench-Scale Pretreatment Schematics	147

Appendix  D:   Single Contactor and  Blended Effluent  DBF Surrogate  and Formed DBF
Correlations	157

Appendix E: Logistic Function Model Curve Fits	223

Appendix F:  Comparison of SCA Method to DI Approach for Integral Breakthrough Curve
Prediction	305

Appendix G: Logistic Function Model Best-Fit Parameters	387

Appendix H: Impact of Extrapolation on SCA Prediction of the Integral Breakthrough Curve393

Appendix I:  Impact of Extrapolation on DI Prediction of the Integral Breakthrough Curve .... 415
                                     -xv-

-------
This page intentionally left blank.
           -xvi-

-------
 List of Tables

1  Summary of RSSCT design parameters for all runs	18
2  Relationship between blended effluent sample number and RSSCT effluent sample number	21
3  Summary of pretreatment and water quality	30
4  SDS chlorination conditions	31
5  Summary of analytical methods and MRLs	32
6  Summary of laboratories conducting analyses	32
7  Summary of field duplicate precision for single contactor and blended effluent data	34
8  Frequency of logistic function model used and R2 values for all parameters and all waters	69
9  Summary of DI and SCA integral breakthrough curve prediction RSS values for Waters 1
   through 4	90
10 Summary of DI and SCA integral breakthrough curve prediction RSS values for Waters 5
   through 8	91
11 Summary of model prediction bias for Waters 1 through 4	92
12 Summary of model prediction bias for Waters 5 through 8	92
13 Summary of mean RSS, mean bias, normalized mean RSS, and normalized mean bias for all
   waters	93
14 Summary of run times to a 0.35 treatment objective	115
15 Summary of run times to a 0.50 treatment objective	115
16 Summary of run times to a 0.65 treatment objective	116
17 Summary of run times to a 0.80 treatment objective	116
                                      -xvii-

-------
This page intentionally left blank.
           -XVlll-

-------
List of Figures

1    Schematic of blending of multiple GAC contactors operated parallel-staggered mode	3
2    Effluent water quality during quasi steady-state operation of multiple contactors in parallel-
     staggered mode	3
3    Operation of multiple contactors in parallel-staggered mode to various single contactor run
     times: derivation of the blended effluent integral breakthrough curve	5
4    Graphical summary of SCA procedure used for base analysis of GAC treatment studies	10
5    Logistic function model curves	24
6    Correlations based on GAC effluent TOC concentration for single contactor and blended
     effluents for Water 1	40
7    Correlations based on GAC effluent TOC concentration for single contactor and blended
     effluents for Water 2	41
8    Correlations based on GAC effluent TOC concentration for single contactor and blended
     effluents for Water 8	42
9    THM correlations based on GAC effluent TOC concentration for single contactor and
     blended effluents for Water 2	43
10   THM correlations based on GAC effluent TOC concentration for single contactor and
     blended effluents for Water 3	44
11   THM correlations based on GAC effluent TOC concentration for single contactor and
     blended effluents for Water 8	45
12   HAA correlations based on GAC effluent TOC concentration for single contactor and
     blended effluents for Water 5	46
13   HAA correlations based on GAC effluent TOC concentration for single contactor and
     blended effluents for Water 7	47
14   HAA correlations based on GAC effluent TOC concentration for single contactor and
     blended effluents for Water 2	48
15   HAA correlations based on GAC effluent TOC concentration for single contactor and
     blended effluents for Water 4	49
16   HAA correlations based on GAC effluent TOC concentration for single contactor and
     blended effluents for Water 7	50
17   Correlation between single contactor and blended effluent TOC concentration and THM
     bromine incorporation factor (n) for Water  1	51
18   Correlation between single contactor and blended effluent TOC concentration and THM
     bromine incorporation factor (n) for Water  2	51
19   Correlation between single contactor and blended effluent TOC concentration and THM
     bromine incorporation factor (n) for Water  3	52
20   Correlation between single contactor and blended effluent TOC concentration and THM
     bromine incorporation factor (n) for Water  4	52
21   Correlation between single contactor and blended effluent TOC concentration and THM
     bromine incorporation factor (n) for Water  5	53
22   Correlation between single contactor and blended effluent TOC concentration and THM
     bromine incorporation factor (n) for Water  6	53
23   Correlation between single contactor and blended effluent TOC concentration and THM
     bromine incorporation factor (n) for Water  7	54
24   Correlation between single contactor and blended effluent TOC concentration and THM
     bromine incorporation factor (n) for Water  8	54

                                      -xix-

-------
25   Correlation between single contactor and blended effluent TOC concentration and HAA
     bromine incorporation factor (n') for Water 1	55
26   Correlation between single contactor and blended effluent TOC concentration and HAA
     bromine incorporation factor (n') for Water 2	55
27   Correlation between single contactor and blended effluent TOC concentration and HAA
     bromine incorporation factor (n') for Water 3	56
28   Correlation between single contactor and blended effluent TOC concentration and HAA
     bromine incorporation factor (n') for Water 4	56
29   Correlation between single contactor and blended effluent TOC concentration and HAA
     bromine incorporation factor (n') for Water 5	57
30   Correlation between single contactor and blended effluent TOC concentration and HAA
     bromine incorporation factor (n') for Water 6	57
31   Correlation between single contactor and blended effluent TOC concentration and HAA
     bromine incorporation factor (n') for Water 7	58
32   Correlation between single contactor and blended effluent TOC concentration and HAA
     bromine incorporation factor (n') for Water 8	58
33   Correlation between single contactor and blended effluent UV absorbance and THM
     bromine incorporation factor (n) for Water 1	59
34   Correlation between single contactor and blended effluent UV absorbance and THM
     bromine incorporation factor (n) for Water 2	59
35   Correlation between single contactor and blended effluent UV absorbance and THM
     bromine incorporation factor (n) for Water 3	60
36   Correlation between single contactor and blended effluent UV absorbance and THM
     bromine incorporation factor (n) for Water 4	60
37   Correlation between single contactor and blended effluent UV absorbance and THM
     bromine incorporation factor (n) for Water 5	61
38   Correlation between single contactor and blended effluent UV absorbance and THM
     bromine incorporation factor (n) for Water 6	61
39   Correlation between single contactor and blended effluent UV absorbance and THM
     bromine incorporation factor (n) for Water 7	62
40   Correlation between single contactor and blended effluent UV absorbance and THM
     bromine incorporation factor (n) for Water 8	62
41   Correlation between single contactor and blended effluent UV absorbance and HAA
     bromine incorporation factor (n') for Water 1	63
42   Correlation between single contactor and blended effluent UV absorbance and HAA
     bromine incorporation factor (n') for Water 2	63
43   Correlation between single contactor and blended effluent UV absorbance and HAA
     bromine incorporation factor (n') for Water 3	64
44   Correlation between single contactor and blended effluent UV absorbance and HAA
     bromine incorporation factor (n') for Water 4	64
45   Correlation between single contactor and blended effluent UV absorbance and HAA
     bromine incorporation factor (n') for Water 5	65
46   Correlation between single contactor and blended effluent UV absorbance and HAA
     bromine incorporation factor (n') for Water 6	65
47   Correlation between single contactor and blended effluent UV absorbance and HAA
     bromine incorporation factor (n') for Water 7	66

                                       -xx-

-------
48   Correlation between single contactor and blended effluent UV absorbance and HAA
     bromine incorporation factor (n') for Water 8	66
49   Single contactor and blended effluent TOC breakthrough curves for Water 1	71
50   Single contactor and blended effluent TOC breakthrough curves for Water 2	71
51   Single contactor and blended effluent TOC breakthrough curves for Water 3	72
52   Single contactor and blended effluent TOC breakthrough curves for Water 4	72
53   Single contactor and blended effluent TOC breakthrough curves for Water 5	73
54   Single contactor and blended effluent TOC breakthrough curves for Water 6	73
55   Single contactor and blended effluent TOC breakthrough curves for Water 7	74
56   Single contactor and blended effluent TOC breakthrough curves for Water 8	74
57   Single contactor and blended effluent UV254 breakthrough curves for Water 5	75
58   Single contactor and blended effluent UV254 breakthrough curves for Water 7	75
59   Single contactor and blended effluent SDS-TOX breakthrough curves for Water 4	76
60   Single contactor and blended effluent SDS-TOX breakthrough curves for Water 8	76
61   Single contactor and blended effluent SDS-TTHM breakthrough curves for Water 3	77
62   Single contactor and blended effluent SDS-TTHM breakthrough curves for Water 6	77
63   Single contactor and blended effluent SDS-HAA9 breakthrough curves for Water 1	78
64   Single contactor and blended effluent SDS-HAA9 breakthrough curves for Water 2	78
65   Single contactor and blended effluent SDS-CF breakthrough curves for Water 4	79
66   Single contactor and blended effluent SDS-BDCM breakthrough curves for Water 6	79
67   Single contactor and blended effluent SDS-BDCM breakthrough curves for Water 7	80
68   Single contactor and blended effluent SDS-BDCM breakthrough curves for Water 4	80
69   Single contactor and blended effluent SDS-BDCM breakthrough curves for Water 8	81
70   Single contactor and blended effluent SDS-BF breakthrough curves for Water 1	81
71   Single contactor and blended effluent SDS-BF breakthrough curves for Water 6	82
72   Single contactor and blended effluent SDS-DCAA breakthrough curves for Water 3	82
73   Single contactor and blended effluent SDS-TCAA breakthrough curves for Water 5	83
74   Single contactor and blended effluent SDS-DBAA breakthrough curves for Water 6	83
75   Single contactor and blended effluent SDS-BCAA breakthrough curves for Water 1	84
76   Single contactor and blended effluent SDS-DCBAA breakthrough curves for Water 2	84
77   Single contactor and blended effluent SDS-CDBAA breakthrough curves for Water 6	85
78   Single contactor and blended effluent SDS-TBAA breakthrough curves for Water 7	85
79   Single contactor and blended effluent SDS-BF breakthrough curves for Water 2	86
80   Single contactor and blended effluent SDS-CDB AA breakthrough curves for Water 5	86
81   Single contactor and blended effluent SDS-BF breakthrough curves for Water 3 (original
     step-lag logistic function model curve fit)	87
82   Single contactor and blended effluent SDS-BF breakthrough curves for Water 3 (fit to
     step-lag-peak logistic function model)	87
83   Cumulative frequency distribution plot of normalized residual sum-of-squares (RSS) for DI
     and SCA model predictions	100
84   Cumulative frequency distribution plot of normalized bias for DI and SCA model
     predictions	100
85   Comparison of DI and SCA methods for predicting the UV254 integral breakthrough curve
     for Water 1	101
86   Comparison of DI and SCA methods for predicting the UV254 integral breakthrough curve
     for Water 3	101

                                       -xxi-

-------
87   Comparison of DI and SCA methods for predicting the SDS-TTHM integral breakthrough
     curve for Water 1	102
88   Comparison of DI and SCA methods for predicting the SDS-TTHM integral breakthrough
     curve for Water 7	102
89   Comparison of DI and SCA methods for predicting the SDS-HAA5 integral breakthrough
     curve for Water 2	103
90   Comparison of DI and SCA methods for predicting the SDS-HAA5 integral breakthrough
     curve for Water 7	103
91   Comparison of DI and SCA methods for predicting the SDS-HAA6 integral breakthrough
     curve for Water 3	104
92   Comparison of DI and SCA methods for predicting the SDS-HAA6 integral breakthrough
     curve for Water 5	104
93   Comparison of DI and SCA methods for predicting the SDS-HAA9 integral breakthrough
     curve for Water 5	105
94   Comparison of DI and SCA methods for predicting the SDS-HAA9 integral breakthrough
     curve for Water 6	105
95   Comparison of DI and SCA methods for predicting the SDS-TOX integral breakthrough
     curve for Water 6	106
96   Comparison of DI and SCA methods for predicting the SDS-TOX integral breakthrough
     curve for Water 7	106
97   Comparison of DI and SCA methods for predicting the SDS-CF integral breakthrough
     curve for Water 2	107
98   Comparison of DI and SCA methods for predicting the SDS-CF integral breakthrough
     curve for Water 6	107
99   Comparison of DI and SCA methods for predicting the SDS-BDCM integral breakthrough
     curve for Water 1	108
100  Comparison of DI and SCA methods for predicting the SDS-BDCM integral breakthrough
     curve for Water 8	108
101  Comparison of DI and SCA methods for predicting the SDS-DBCM integral breakthrough
     curve for Water 3	109
102  Comparison of DI and SCA methods for predicting the SDS-DBCM integral breakthrough
     curve for Water 7	109
103  Comparison of DI and SCA methods for predicting the SDS-BF integral breakthrough
     curve for Water 1	110
104  Comparison of DI and SCA methods for predicting the SDS-BF integral breakthrough
     curve for Water 5	110
105  Comparison of DI and SCA methods for predicting the SDS-DCAA integral breakthrough
     curve for Water 1	Ill
106  Comparison of DI and SCA methods for predicting the SDS-DCAA integral breakthrough
     curve for Water 4	Ill
107  Comparison of DI and SCA methods for predicting the SDS-DBAA integral breakthrough
     curve for Water 2	112
108  Comparison of DI and SCA methods for predicting the SDS-BCAA integral breakthrough
     curve for Water 8	112
109  Integral breakthrough curves for varying numbers of contactors operated in parallel-
     staggered mode  (B=30;D=0.1)	117

                                     -xxii-

-------
110  Integral breakthrough curves for varying numbers of contactors operated in parallel-
     staggered mode (B=30; D=0.05)	117
111  Integral breakthrough curves for varying numbers of contactors operated in parallel-
     staggered mode (B=10; D=0.1)	118
112  Integral breakthrough curves for varying numbers of contactors operated in parallel-
     staggered mode (B=10; D=0.05)	118
113  Integral breakthrough curves for varying numbers of contactors operated in parallel-
     staggered mode (B=10; D=0.2)	119
114  Integral breakthrough curves for varying numbers of contactors operated in parallel-
     staggered mode (B=30; D=0.2)	119
115  Run time as a function of number of contactors in parallel, expressed as percent of run time
     for infinite n (B=30; D=0.1)	120
116  Run time as a function of number of contactors in parallel, expressed as percent of run time
     for infinite n(B=30;D=0.2)	120
117  Impact of extrapolation on DI prediction of the TOC integral breakthrough for Water 5	124
118  Impact of extrapolation on DI prediction of the TOC integral breakthrough for Water 8	124
119  Impact of extrapolation on SCA prediction of the UV254 integral breakthrough for Water 5 125
120  Impact of extrapolation on SCA prediction of the SDS-TOX integral breakthrough for
     Water5	125
121  Impact of extrapolation on SCA prediction of the SDS-TTHM integral breakthrough for
     WaterS	126
122  Impact of extrapolation on SCA prediction of the SDS-HAA5 integral breakthrough for
     WaterS	126
123  Impact of extrapolation on SCA prediction of the SDS-HAA6 integral breakthrough for
     WaterS	127
124  Impact of extrapolation on SCA prediction of the SDS-HAA9 integral breakthrough for
     WaterS	127
125  Impact of extrapolation on SCA prediction of the UV254 integral breakthrough for Water 8 128
126  Impact of extrapolation on SCA prediction of the SDS-TOX integral breakthrough for
     WaterS	128
127  Impact of extrapolation on SCA prediction of the SDS-TTHM integral breakthrough for
     WaterS	129
128  Impact of extrapolation on SCA prediction of the SDS-HAA5 integral breakthrough for
     WaterS	129
129  Impact of extrapolation on SCA prediction of the SDS-HAA6 integral breakthrough for
     WaterS	130
130  Impact of extrapolation on SCA prediction of the SDS-HAA9 integral breakthrough for
     WaterS	130
                                     -XXlll-

-------
This page intentionally left blank.
           -xxiv-

-------
List of Abbreviations
BCAA




BDCAA




BDCM




BF




BMRL




BrTOC




C(t)





C(t)





C(tf)




C(tp)




Co




CDBAA




CF





CP




DBAA




DBCM




DBF




DCAA




DCBAA




DI




EBCT
Bromochloroacetic acid




Bromodichloroacetic acid




Bromodichloromethane




Bromoform




Below the minimum reporting level




Bromide to TOC ratio




Effluent concentration




blended effluent concentration at individual contactor run time, t





Last observed data point




Measured peak concentration




Influent concentration




Chlorodibromoacetic acid




Chloroform




Logistic function model best-fit concentration at tp




Dibromoacetic acid




Dibromochloromethane




Disinfection byproduct




Dichloroacetic acid




Dichlorobromoacetic acid




Direct integration




Empty bed contact time
                                     -xxv-

-------
EPA

 7
ft
GAC
HAA
HAAS
HAA6
HAA9
ICR
K
MBAA
MCAA
MCL
MRL
n
N
NA
nBr
Nc
n'Br
NOM
PAS
United States Environmental Protection Agency
Fraction of organic matter remaining in the combined effluent

C(t)/C0
Granular activated carbon
Haloacetic acid
Sum of five haloacetic acids: MCAA, DCAA, TCAA, MBAA, DBAA
Sum of six haloacetic acids:  HAAS, BCAA
Sum of nine haloacetic acids: HAA6, DCBAA, CDBAA, TBAA
Information Collection Rule
Adsorption rate coefficient
Monobromoacetic acid
Monochloroacetic acid
Maximum contaminant level
Minimum reporting level
Freundlich isotherm parameter
Number of contactors
Not applicable
Bromine incorporation factor for THMs
Adsorption capacity coefficient
Bromine incorporation factor for HAAs
Natural organic matter
Polyaluminum sulfate
                                      -xxvi-

-------
R2

RSS

RSSCT
RTsc

SAS

SCA

SDS

SDS-BCAA

SDS-BDCAA

SDS-BDCM

SDS-BF

SDS-CDBAA

SDS-CF

SDS-DBAA

SDS-DBCM

SDS-DBPs

SDS-DCAA

SDS-HAA5
Throughput of the individual contactor when the  treatment objective is
exceeded in the single contactor

Throughput of the nth contactor at the time the  treatment objective is
exceeded in the blended effluent

Coefficient of determination

Residual sum of squares

Rapid small-scale column test

Blended contactor run time

Single contactor run time

Statistical Analysis Software

Surrogate correlation approach

Simulated distribution system

Bromochloroacetic acid evaluated under SDS conditions

Bromodichloroacetic acid evaluated under SDS conditions

Bromodichloromethane evaluated under SDS conditions

Bromoform evaluated under SDS conditions

Chlorodibromoacetic acid evaluated under SDS conditions

Chloroform evaluated under SDS conditions

Dibromoacetic acid evaluated under SDS conditions

Dibromochloromethane evaluated under SDS conditions

Disinfection byproducts evaluated under SDS conditions

Dichloroacetic acid evaluated under SDS conditions

The sum of five haloacetic acids evaluated under SDS conditions
                                    -xxvii-

-------
SDS-HAA6       The sum of six haloacetic acids evaluated under SDS conditions




SDS-HAA9       The sum of nine haloacetic acids evaluated under SDS conditions




SDS-MBAA      Monobromoacetic acid evaluated under SDS conditions




SDS-MCAA      Monochloroacetic acid evaluated under SDS conditions




SDS-TBAA       Tribromoacetic acid evaluated under SDS conditions




SDS-TCAA       Trichloroacetic acid evaluated under SDS conditions




SDS-TOX        Total organic halides evaluated under SDS conditions




SDS-TTHM       Total trihalomethanes evaluated under SDS conditions




SM              Standard Methods




t                Service time




tb                Run time at which initial breakthrough above detectable levels occurs




TBAA           Tribromoacetic acid




TCAA           Trichloroacetic acid




THM            Trihalomethane




THMs           Trihalomethanes




TOC             Total organic carbon




TOX             Total organic halide




tp                Run time at which the peak concentration occurs




TSUVA          Specific ultraviolet absorbance based on TOC




TTHM           Total trihalomethane




UV254            Ultraviolet absorbance at 254 nm




v                Linear velocity
                                    -XXVlll-

-------
Bed depth
                    -xxix-

-------
This page intentionally left blank.
           -XXX-

-------
1   Background and  Method Development

Sixty-two public water systems required by the Information Collection Rule (ICR) to perform
treatment studies for disinfection byproduct (DBF) precursor removal fulfilled this requirement
by evaluating granular activated carbon (GAC), while membranes were examined by 36 utilities.
GAC treatment studies included bench-, pilot-, and full-scale studies.  Regardless  of the scale,
the ICR required that GAC treatment studies evaluate DBF precursor removal in the effluent of a
single contactor as a function of run time, to assess the breakthrough of DBF precursors as the
GAC was exhausted.  In practice, full-scale plants can optimize GAC performance and reduce
carbon usage rates by blending the effluents of multiple contactors operated in parallel prior to
disinfection.  To optimize the blending process, replacement of the  GAC in each contactor is
staggered at regular intervals.  By doing so, the service life of each contactor is extended because
water from contactors that exceed the treatment objective is blended with water from contactors
with fresher  GAC that have effluent concentrations below the treatment objective.   A cost
analysis based on single contactor run times will overestimate the actual treatment costs  of full-
scale operation of multiple contactors  operated in parallel-staggered mode.

Previous researchers have developed mathematical relationships to characterize the water quality
in the blended effluent based on single contactor breakthrough curves. These relationships yield
a blended contactor effluent  or integral "breakthrough"  curve  that provides  a measure of the
extended run time for which each contactor can be operated to meet a treatment objective in the
blended effluent.  The integral breakthrough curve does not directly  represent blended effluent
water quality, but is a tool for determining the  GAC replacement frequency for each individual
contactor (of multiple contactors operated  in parallel-staggered mode) to maintain the blended
effluent below the treatment objective. For example, the results of a GAC study show that based
on a  single contactor breakthrough curve,  GAC  effluent formed DBF levels are maintained
below the treatment objectives for 80 days of operation.  This is the single contactor run time,
RTSc.   Since full-scale implementation of GAC will  involve  multiple contactors operated in
parallel-staggered mode, integral breakthrough curves for each parameter are developed based on
the single contactor breakthrough curves.  The integral breakthrough curves might show that the
level  of DBFs formed in the GAC blended effluent are maintained below  the treatment
objectives for 160 days of operation, indicating that each individual contactor can be operated for
160  days while maintaining the plant blended  effluent water quality below  the treatment
objective. The blended contactor run time (RTBc) is 160 days.

The basic method utilized to determine the  integral breakthrough curve has been a mathematical
or numerical integration of the single contactor breakthrough curve.  A minimal amount of
experimental verification has been  performed to verify this approach.   Some  bench-scale
experimental verification data has  been presented (Chowdhury et al.,  1996; Summers et al.,
1998),  whereby  mathematical integration  was  simulated experimentally  by continuously
collecting the effluent from a bench-scale GAC  contactor in a large reservoir, and  sampling from
this reservoir over time.  An  extensive verification study is needed to thoroughly  evaluate the
appropriateness of the integration method for predicting the integral breakthrough  curve.
                                       -1-

-------
1.1  Optimization of GAC Operation

The GAC in a contactor has to be replaced when the mass transfer zone begins to exit the column
as shown in Figure 1, and the effluent concentration exceeds the treatment objective.  However,
at this  point only part of the GAC bed is saturated and replacement of the GAC will result in
high carbon use rates (Snoeyink, 1990).  Two common methods of lowering carbon usage rates
are to  operate contactors in series or to operate multiple contactors in parallel with staggered
GAC replacement cycles.

For adsorption of micropollutants, the amount of water treated per mass GAC  can be increased
by operation of two contactors in series.   In this mode of operation, two contactors are operated
in series until the treatment objective is  exceeded in the second contactor. At this point, the
GAC in the first contactor is replaced with virgin or reactivated GAC, and valves are switched so
that the second contactor is now operated  in-line ahead  of the first contactor.   This cycle is
repeated to maintain effluent levels below the treatment objective.  For efficient  operation, the
mass transfer zone should be contained within the bed length of one contactor. This  can be
achieved using reasonable bed lengths for adsorption of micropollutants, but the mass transfer
zone for TOC removal (and therefore DBF precursor removal) is usually  too long.  For DBF
precursor control, operation of two contactors in series does not result in significantly longer run
times over single contactor operation (Sontheimer et al., 1988).

Multiple contactors operated in parallel and staggered in terms of GAC replacement times, as
shown  in Figure 1, yield blended effluent concentrations as shown in Figure 2. When the lead
contactor, which has been in operation the longest,  is taken off-line and replaced with a contactor
with fresh GAC, the blended effluent concentration decreases as shown in Figure 2. The level of
this decrease is dependent  on the number of contactors in operation.  For two contactors, the
concentration will decrease by 50 percent, while  for an infinite  number of  contactors the
decrease approaches zero.

Due  to the blending of waters from multiple  contactors  producing varying levels of effluent
water quality, individual contactors can be  operated past  the point at which the effluent water
they  are producing exceeds the treatment objective, because the  treatment objective must be
maintained in the blended effluent.  Therefore, evaluation of single contactor breakthrough curve
data  will result in overestimates  of the  carbon  usage rate  for a  full-scale system operating
multiple staggered contactors in parallel-staggered mode.  For DBF precursor control, contactor
effluents are blended prior to disinfection.

1.2  Modeling the Operation of Multiple Contactors Operated in Parallel-
     Staggered Mode

In modeling the operation of multiple contactors operated in parallel-staggered mode, the goal is
not to  simulate the actual blended  effluent water quality during normal operation (as shown in
Figure  2), but to derive the integral breakthrough  curve.  The integral breakthrough curve is a
tool used to determine the GAC replacement frequency for each individual contactor to maintain
the blended effluent below the treatment objective: it is a curve that relates single contactor run
time to blended contactor effluent  water  quality.   Multiple contactor throughput to  a treatment
objective can be estimated  from the integral breakthrough curve by determining the operation
                                       -2-

-------
 Influent water
                              0000
                                                         Disinfectant
                                     Blended effluent
                                                                      ^Finished
                                                                      water
Figure 1  Schematic of blending of multiple GAC contactors operated in parallel-staggered
mode
    Blended effluent
    concentration
Blended effluent treatment objective
                                                        New contactor
                                                        placed on-line
                                        Operation time
Figure 2 Blended effluent water quality during quasi steady-state operation of multiple
contactors in parallel-staggered mode (adapted from Summers et al., 1998)
                                     -3-

-------
time when  the  curve intersects the treatment objective, as  is done with  single contactor
breakthrough curves.

Figure 3 shows a series of graphs that describe how the integral breakthrough curve is developed.
Graphs A through  E depict eight  GAC contactors on-line  in parallel-staggered mode.   For
simplification, identical breakthrough  curves  are  assumed for each  contactor.  The  interval
between GAC replacement,  or the single contactor run time (RTSc), is increased from 16 to 99
days over these five graphs.  At each RTSc, the single  contactor effluent concentration, C(f), is
given as well as the  blended contactor effluent concentration, C(Y), which  is calculated by
averaging the effluent water quality of each of the eight contactors at RTsc (shown as a short
dotted line representing the intersection of effluent concentration at  RTSc).   The dashed  line
breakthrough curve represents the contactor that replaces the first contactor when it has reached
the end of its service  life.  As RTSc is  increased, moving from Graph  A through E, the C(t) at
RTSc increases.  Therefore, the GAC replacement interval  affects the blended contactor effluent
water quality.  Specifically, as shown graphically by the integral breakthrough curve in Graph E,
as RTSC  increases,  blended effluent water quality  declines.  Using the integral breakthrough
curve, the single contactor run time at which the treatment objective is exceeded in the blended
effluent can be determined.

Assuming a linear breakthrough curve, Westrick and Cohen  (1976) modeled the  impact of
parallel contactor operation on blended effluent  water  quality,  and derived the  following
equation:


                                                                                     <»
where TV is the number of parallel contactors, qN is the specific throughput of the Mh contactor at
the time the treatment objective is exceeded in the blended effluent, and qt is the throughput of
the individual contactor when the treatment objective is exceeded in that single contactor.  For
large N, Equation 1 shows that q^ approaches twice qt:  the run time of each contactor when the
blended effluent treatment objective is exceeded will approach twice that of a single contactor
when the treatment objective is exceeded in the  single contactor effluent. For a finite N, such as
10 contactors, q^ is a factor of 1.8 times greater than qt.

Equation 1 establishes a relationship between the specific throughput of the Mh contactor at the
time the treatment objective was exceeded in the blended effluent and the throughput of a single
contactor to that same treatment objective, assuming a linear breakthrough curve. For large N,
Equation 1 shows that the run throughput of each contactor approaches twice that for a single
contactor:

                     <7~=2
-------
                          Identical beattiroutj curves represent ffeinbp
     2  i   RTsc=16days    contactors operated in parallel-stagred mode
 o

 I
 0)
 u
 8
 o
 1
 HI
 °_.
 c
 o
 0)
                             Operation time, days
                   RTSc = 31 days
                            Operation time, davs

                            RTSc = 46 days
                            Operation time, days
                                                      C(RTSC)=   0.11 (A)
                                                     C(RTSC)= o.67

                                                     C(RTSC)= 0.31 (B)
O
c"
o
t:
                                                                           o
                                                                           O
                                                                                                                RTSC = 61 days
                                                                                     C(RTSC) =   0.99

                                                                                     crRrsc; =   0.64 (D)
                                                                                                            e            s
                                                                                                       Operation time, days
                                                                               2 -, C(RTSC)=   1.00

                                                                                   C(RTSC)=   0.80 (E)
o
1
HI
                                                                                                       Operation time, days
«,
C
o
                                                                                                                                       RTSC = 99 days
                                                                                       Single contactor
                                                                                      breakthrough curve
                                                                                                                         Blended effluent integral
                                                                                                                           breakthrough curve
                                                                                                                            n = 8 contactors)
                           Operation time, days
Figure 3  Operation of multiple contactors in parallel staggered mode to various single contactor run times: derivation of
the blended effluent integral breakthrough curve

-------
Therefore, under the linear breakthrough curve assumption, when 9 contactors are operated in
parallel-staggered mode, the throughput of each contactor when the blended effluent exceeds the
treatment objective will be 90 percent of that for an infinite number of contactors operated in
parallel-staggered mode. The analysis assumes a linear breakthrough curve and shows that N90 is
independent of the magnitude of the treatment objective.

Roberts and Summers (1982) examined the impact of contactor operation in parallel-staggered
mode on the run times of individual contactors. The authors showed that the fraction of organic
matter remaining in the combined  effluent, /, could be  estimated from a single contactor
breakthrough curve, assuming regular GAC replacement intervals:
where TV is the number of contactors and/ is the fraction, C(t)/Co, of organic matter remaining in
the effluent of the rth contactor, determined from a breakthrough curve.  A plot of the integral
breakthrough curve  over  operation time provides an  estimate  of the  service time  of  each
contactor in a multiple contactor scenario for a treatment objective.  This relationship is referred
to as the integral breakthrough curve, and the development of this curve is explained graphically
in Figure 3.

1.3  Single Contactor Breakthrough Curve Models

A model  used  to describe single  contactor effluent experimental  data  is needed for several
reasons.   From a data management  perspective,  best-fit  curve parameters that  adequately
describe experimental data are less memory intensive than storing the entire experimental data
set.  A  best-fit curve also facilitates interpolation and extrapolation necessary  to estimate run
times for given treatment objectives. Use of a best-fit model curve also provides an estimate of
the scatter in the data through the coefficient of determination,  and the model minimizes the
impact  of this  scatter  on  run time estimates.  Finally, a  function that describes the single
contactor breakthrough curve is a prerequisite for calculating the integral breakthrough curve, a
curve that relates single  contactor run time to blended  effluent water  quality under the
assumption that an infinite number of contactors are operated in parallel-staggered mode.  Run
time estimates  generated by the integral breakthrough  curve are more applicable to full-scale
GAC operation where multiple contactors are operated in parallel-staggered mode to increase the
service time of each individual contactor.

Many  researchers have applied  various forms  of the logistic  function  to predict  GAC
breakthrough curves or to fit existing  breakthrough curve data.  The  logistic function is  a
symmetric S-shaped curve  with  a  midpoint inflection.    Oulman  (1980)  describes  the
development of the Bohart-Adams equation, published  in 1920, which was used to model the
service life of activated carbon for the removal of airborne chlorine  by gas masks. The Bohart-
Adams equation is:
                                       -6-

-------
                     In
                        c(0-i
                                               (5)
where  C(f) is the effluent concentration at time t, Co  is the influent concentration, K is an
adsorption rate coefficient, Nc is an adsorption capacity  coefficient, x is the bed depth, v is the
linear velocity, and t is the  service time.  The Bohart-Adams equation is a bed depth service
model  and was derived based  on surface reaction theory (Clark, 1987).   Equation 5 can be
rewritten in the form of the logistic function:
                                1
                     C(t)   \ + e(a+bt)

where the variables a and b are defined as:
                                                                                    (6)
                     a =
                         -KNcx
                                               (7)
and
                                                                                    (8)
If the adsorption coefficients  K and Nc are  known, as well as other operational parameters
(velocity and  depth), the effluent concentration at time  t  can be predicted by Equation  6.
Because GAC breakthrough curves are generally not symmetrical, Clark (1987) proposed the use
of the generalized logistic function to model GAC breakthrough curves.  This generalized model
incorporates the Freundlich isotherm parameter and is as follows:
                              C
                                n-l
n-l
                               Ae~
                                                                                    (9)
where l/n is the Freundlich isotherm parameter, r is a constant, and A, a constant, is defined as:
                     A =
                            Cn
                                       ,-rt
                                               (10)
The derivation of this generalized logistic function is described in Clark, Symons, and Ireland
(1986).   Their approach, which builds on the work of Oulman, is an effort to predict  GAC
breakthrough curve profiles based on adsorption characteristics and influent concentration.

A predictive approach to single contactor GAC breakthrough curves was not examined as a part
of this study. Instead, a model was needed to curve fit experimental breakthrough data. Due to
the inherent ability of the logistic function to match the typical S-shaped breakthrough curve, and
the previous body of work that has utilized the logistic function to model GAC breakthrough
data, the logistic function was chosen as the equation used to fit GAC  breakthrough data in this
study.
                                       -7-

-------
Chowdhury et al. (1996) and Summers et al. (1998) applied the following form of the logistic
function to model experimental GAC breakthrough data:
where the values for A, B, and D are determined experimentally by a best-fit to the breakthrough
data.  The  parameter A represents the level to which the function approaches asymptotically.
Parameters B and D affect the shape of the curve. Equation 1 1 was found to adequately fit GAC
breakthrough curves  for three water sources and the parameters total organic carbon (TOC) and
formed total trihalomethane (TTHM). With a few modifications, as described in Section 3.2.1
below, Equation 1 1 served as a basis for the single contactor breakthrough curve modeling work
contained in this study.

1.4  Direct Integration Approach

The average value function, a mathematical integration, assumes an infinite number of parallel-
staggered contactors and replaces the numerical  integration required for solving Equation 4.
However, it is important to understand the impact of the infinite contactor assumption on model
results. Based on Equation 1, nine contactors in parallel will yield qN=9 within 10 percent of q^ .
Therefore, for 10 contactors or greater, the difference will be less than 10 percent, based on the
linear breakthrough curve assumption. Additionally, Westrick and Cohen (1976) found that the
carbon usage rate for individual contactors operated  in parallel-staggered mode will approach
half the carbon usage rate based  on meeting the  treatment objective using a single contactor.
Roberts and Summers (1982) examined integral TOC breakthrough curves applied to eight case
studies, and found that GAC run times increased by a factor of two to three over those based on
the single contactor breakthrough curve,  for a 50 percent TOC breakthrough objective.  They
concluded that staggered multiple parallel contactor operation will lower carbon usage rates even
more  than  predicted by Westrick  and  Cohen, whose  conclusions were based  on a  linear
breakthrough curve.

Chowdhury et al. (1996) and Summers et al. (1998) applied the average value function to show
that the integral breakthrough curve can be represented by:


                     C(t) = -\C(t)dt                                                 (12)
                           t o

where C(t) is the blended effluent concentration at individual contactor run time, t, and C(f) is
an equation that describes the single contactor effluent concentration as a function of time.  This
time-averaged mathematical integration  of the function that describes the breakthrough curve
yields the integral breakthrough curve, as it describes the average value of the function at any
point in time.  This procedure is referred  to here as the direct integration (DI) approach.  While
Equation 4  provides an expression of the  average blended concentration of N contactors in
parallel, Equation 12 represents a time-averaged blended  effluent concentration of  an infinite
number of  staggered parallel contactors.  Therefore, a plot of Equation 12 over operation time
represents the run time to any  given treatment objective in the blended contactor effluent for
                                       -8-

-------
each individual contactor of an infinite number of parallel-staggered contactors.  This plot is not
a direct representation of blended effluent water quality, but a tool to determine GAC run times
of each contactor operated in parallel-staggered mode.

Chowdhury  et al. (1996) and  Summers et  al. (1998)  applied Equation 12 to Equation 11 to
predict the integral breakthrough curve.  Application of the DI approach utilizing different forms
of the logistic function (Equation 11) is described below in Section 3.2.3.

1.5   Surrogate Correlation Approach

Theoretically, the direct integration approach is a reliable method for estimating blended effluent
water quality.  However, during the analysis  approach for the ICR GAC treatment study data, the
treatment study technical work group (TS-TWG) determined that it would be computationally
intensive to apply the DI approach to the large number of breakthrough curves (8,000 to  9,000)
comprising the ICR treatment study data set.  Furthermore, the DI approach may be less reliable
for breakthrough curves that do not follow the typical "S" pattern, such as "peak" curves or sharp
"S" shaped curves followed  by a  plateau.  For these reasons, another  approach for estimating
blended effluent water quality from single contactor data was developed for analysis of the  ICR
treatment study GAC data.

The surrogate  correlation approach (SCA) was developed by the TS-TWG as an alternative to
the direct integration approach for calculating the integral breakthrough curve (USEPA,  1999).
The SCA is based on the assumption that the relationship between TOC  and all other parameters
(UV254 and SDS-DBPs) in the  single contactor effluent is maintained in the blended effluent of
multiple staggered contactors.  In other words, the concentration and speciation of DBFs formed
after chlorination of the blended effluent of multiple contactors with a given TOC concentration
is the same as  that formed after chlorination of the single contactor effluent with the same TOC
concentration.  This method  of estimating blended effluent water quality from single contactor
data  requires that an integral breakthrough  curve be determined based on the single contactor
TOC breakthrough only. Once this curve is  known, along with a relationship between TOC and
all other parameters in the single contactor effluent, integral breakthrough  curves are estimated
for all other parameters.  This approach not  only allows for the evaluation  of blended contactor
run times to various breakthrough criteria, but also for  the occurrence of other DBFs at that run
time.   For example, it will  be important to evaluate  the levels of individual DBFs, such as
BDCM, at various regulatory targets under consideration. This will allow EPA to assess whether
or not regulating one DBF or DBF group can effectively control the occurrence of other DBFs or
DBF groups.

An example of the SCA procedure is described in the steps below and summarized graphically in
Figure 4.

1. Select a treatment objective (e.g., TTHM = 32 |ig/L).

2. Use the single contactor breakthrough curve for the parameter of interest and determine the
   single contactor run time  (RTSc) at which the treatment objective is exceeded.
                                       -9-

-------
         SDS-TTHM
   C(0
                 RT

         SDS-HAA6
sc   •«-
         SDS-CF
         (All measured parameters)
                         Chosen treatment
                         objective "(Step I)
                              Run time, t
                              Run time, t
                              Run time, t
                                                    Logistic function best-fit of single
                                                    contactor experimental data
Determine single contactor run time,
RTSC, to treatment objective (Step 2)
                                   Step 3 (repeat for all
                                   measured parameters)
                                                      TOC integral breakthrough curve
                                                      calculated using direct integration
                                                                approach
                              Run time, t
                                                  Run time, t
Figure 4 Graphical summary of SCA procedure used for base analysis of GAC
treatment studies fadaoted from USEPA, 1999)
                                     -10-

-------
3.  Use the RTSc from Step 2 to determine the concentrations of all other parameters from the
   single contactor breakthrough  curves.    (In  this  step, the  single  contactor  effluent
   concentrations of all parameters are "linked" through the single contactor run time, RTSc).

4.  Use the RTSc from Step 2 to determine the single contactor effluent TOC concentration that
   corresponds to the treatment objective.

5.  From the integral TOC breakthrough curve, calculated by the DI approach, determine the
   blended contactor run time (RTBc) to reach the TOC concentration calculated in step 4.  This
   is the only point in the analysis where it is necessary to apply the DI method to establish an
   integral breakthrough curve, and only the TOC  integral breakthrough curve is required.  If
   the integral breakthrough curve does not exceed the TOC concentration calculated in Step 4,
   extrapolation of the TOC integral breakthrough curve may be required.

This analysis makes the following assumptions:

1.  The logistic function model can accurately describe the breakthrough of DBF precursors and
   DBF precursor surrogates,

2.  The relationship between TOC and DBF precursors (concentration and speciation) observed
   in the single contactor effluent is maintained in the blended effluent,

3.  The TOC breakthrough curve can be extrapolated with reasonable results,

4.  The TOC integral breakthrough  curve accurately predicts the blended water  quality of an
   infinite number of multiple  contactors operated in parallel-staggered mode.  Furthermore, the
   TOC integral breakthrough curve based on an infinite number of contactors is  a reasonable
   approximation to a finite numbers of contactors.

All four of these assumptions are verified as a part of this study.

1.6  Impact of Bromide Concentration on GAC Effluent Blending Models

The second assumption listed for the SCA procedure is that the relationship between TOC and
DBF precursors (concentration and speciation) observed in the single  contactor effluent is
maintained in the blended effluent This is an  important assumption because the SCA applies
DBF formation and speciation at a given single contactor run time and TOC concentration to the
blended contactor run time at which an equivalent TOC concentration is achieved.  The relative
concentrations of TOC and bromide may influence DBF formation and speciation.

A number of researchers have discussed the impact of bromide concentration on DBF formation
and speciation.  Under constant chlorination conditions, and at a constant TOC concentration, as
the bromide concentration increases the bromide to TOC ratio increases and DBF  speciation
shifts to favor the formation of brominated species (Summers et al., 1993). GAC treatment does
not remove bromide, while TOC is adsorbed, resulting in higher GAC  effluent bromide to TOC
ratios in the GAC effluent as compared to those in the GAC influent.  Due to this increase, GAC
effluent formed DBFs  may undergo shifts in speciation  to  higher fractions  of the more
brominated DBF species.  In some cases  effluent formed  DBF species  concentrations are
measured higher than those formed  in the influent.  It is important to track the  breakthrough
                                      -11-

-------
behavior of specific DBF species, because some may be of potential health concern and a MCL
could be set for a specific DBF species.

The shift in DBF speciation for THMs can be measured by calculating the bromine incorporation
factor for THMs, nBr (Gould et al., 1983):
                                                                                  (13)
                          —
                              TTHM
where all concentrations (of species and TTHM) are expressed as molar concentrations.  The
value of nBr can range from 0 (only chloroform formed) to 3 (only bromoform formed).  Based
on the bromine incorporation factor for HAA6 (Shukairy et al., 1994), the bromine incorporation
factor, nrBr,  for HAA9 is defined as:

               1-MBAA + 1-BCAA + 1-DCBAA + 2-DBAA + 2-CDBAA + 3-TBAA
         n'B= -        (14)
           Br                             HAA9

where all concentrations (of species and HAA9) are expressed as molar concentrations.

The value of n'Br can range from  0 (only MCAA, DCAA, or TCAA formed) to 3 (only TBAA
formed).  Examining and comparing nBr and n'Br values between single contactor and blended
effluent will help determine whether the second  SCA method assumption described in Section
1.5 is valid.

1.7  Effluent Blending  Modeling of Fewer than 10 Contactors

Both the DI  and  SCA  methods  for determining integral breakthrough curves rely on the
assumption of an infinite number of contactors operated in parallel-staggered mode. For 10 or
more  contactors in parallel-staggered  operation, the  integration presented  in  Equation 12
approximates blended effluent water quality within 10 percent (Roberts and  Summers,  1982).
Equation 12 is  utilized exclusively for the DI approach, while the  SCA approach relies on
Equation 12 to establish the integral breakthrough  curve  for  TOC as part  of the  analysis.
Therefore, it is important to examine the impact of the infinite number of contactors assumption
on the  ability of these methods to predict blended contactor effluent  water quality for a finite
number of contactors.  Furthermore, for smaller plants that operate fewer than 10 contactors in
parallel-staggered mode, the approximation given by Equation 12 will be less accurate and actual
service lives will be shorter than those predicted by Equation 12.

1.8  GAC Breakthrough Curve Extrapolation

The SCA procedure is limited by  the highest TOC concentration reached by the TOC integral
breakthrough  curve, which is typically 40 to 70  percent of the highest  single contactor TOC
concentration.  Therefore,  although higher single contactor TOC concentrations are associated
with formed DBF  concentrations, these  cannot be applied  to the integral breakthrough curve
during  application of the SCA  procedure  unless the  TOC integral breakthrough  curve is
extrapolated.  It is therefore important to  establish  the impact  of extrapolation of the TOC
                                      -12-

-------
integral breakthrough curve on estimated run times to treatment objectives and blended contactor
water quality.  The sensitivity of the predicted integral breakthrough curve to extrapolation was
evaluated for two waters in this study.

1.9   ICR GAC Treatment Study Data Analysis Context

The design of this study incorporated two main goals.  The primary objectives were to evaluate
the ability  of the logistic function to model single contactor breakthrough curve data and to
evaluate  the success  and  limitations  of predictive  models  used  to  determine  the integral
breakthrough curve, a relationship between single contactor run time and blended  contactor
water quality.  The  secondary objective of this study was to evaluate the applicability of these
models and predictive  methods in the context of the ICR GAC treatment study data analysis.

A  large amount of ICR treatment study data will be analyzed:  the 62  GAC treatment studies
performed will generate  a total of 8,000 to 9,000 individual  breakthrough curves.  The SCA
method is  especially  applicable  to  this data analysis procedure  because it minimizes the
computations necessary to estimate blended contactor run times for treatment objectives.  An
assessment of the concentration of other DBFs at any given treatment objective will be required
as  part of the data analysis  effort, and the SCA procedure is also suited to this task. The SCA
procedure requires that GAC breakthrough curves for all measured parameters be represented by
the logistic  function model curve fit.  By  doing so,  a smaller  amount of data are needed to
represent breakthrough curve experimental  data.  The  following steps outline the data analysis
procedure to be utilized during the ICR treatment study data analysis effort:

1.  The logistic function model will be used to fit all water quality parameters  in the GAC
   effluent.

2.  The TOC integral  breakthrough curve will be determined  using the DI approach.  In some
   cases, this curve may be extrapolated.

3.  All logistic function model fit coefficients will be entered into a database to allow different
   breakthrough criteria to be evaluated and queried across all studies.

4.  The SCA procedure will be used to estimate run times for multiple contactors operated in
   parallel-staggered mode by determining the effluent concentrations of various water quality
   parameters  linked to a common single contactor  run time, and  calculating  the blended
   effluent  run time that corresponds to the TOC concentration.  In this manner, simultaneous
   treatment objectives, such  as THM or HAA regulatory  target  concentrations, can  be
   evaluated to determine which parameter controls the design and operation of the process.

5.  Results will be  used to evaluate blended contactor run times that meet  regulatory target
   treatment objectives, and this information can be used  to  estimate costs.   Simultaneous
   treatment objectives will also be evaluated (e.g., HAAs and THMs, TOC and THMs, THMs
   and BDCM).
                                       -13-

-------
1.10 Appropriateness of Model Assumptions to Full-Scale GAC Effluent Blending

Both the DI and SCA methods rely on the assumption that the GAC in each contactor of an array
of contactors is replaced at regular intervals, so that the service times of all contactors are equal.
They also assume that the breakthrough curve profiles of all single contactors are identical.  In a
full-scale plant, these idealized conditions will rarely occur. Variability in source water quality
may impact the run time of the contactors, depending on when they are placed in service and
DBF precursor concentrations in the GAC influent during their service life.  Variability  in
distribution system conditions, especially temperature,  may impact the  contactor service life,  as
DBF levels may increase with higher temperatures.

For a plant that operates a fixed number of contactors, water demand  changes during the year
may directly impact the EBCT of each contactor,  or the number of  contactors in  operation.
Under operation of a constant number of contactors,  a contactor that is placed on-line at the
beginning of the  summer high water demand months may be operated under a shorter EBCT  as
compared to a contactor placed on-line during the winter.  Furthermore, it is less desirable  to
remove  contactors from service to replace GAC during high demand periods. Another approach
is to increase the number of contactors on-line as water  demand increases.
                                      -14-

-------
2   Study Objectives and Approach

This study was performed in conjunction with bench-scale GAC  treatment studies that were
performed  at one laboratory in  fulfillment of ICR requirements  for  eight utilities.   It was
designed to examine the following experimental objectives:

1.  Assess the ability of the logistic function model to fit single contactor breakthrough data for
   eight GAC  runs using eight  water sources and all measured  parameters, including DBF
   surrogates, DBF class sum parameters, and DBF species.  The water sources represent a
   range  in  TOC concentrations,  DBF precursor  levels,  bromide concentrations, and  SDS
   chlorination conditions.

2.  Verify through bench-scale experiments the accuracy of the direct integration  (DI) method
   for establishing  the integral breakthrough curve, a relationship between single contactor
   operation time and blended effluent water quality.

3.  Examine the accuracy of the computationally-simpler surrogate correlation approach (SCA)
   to predict the integral breakthrough curves  of all measured parameters.  Verify a basic
   assumption  of the SCA procedure, that the relationship between DBF  formation and TOC
   concentration in the single contactor effluent and in the blended effluent is maintained.

4.  Examine the impact of GAC effluent blending on DBF speciation.  Assess the accuracy of
   models used to predict the integral breakthrough curve for individual THMs and HAAs, and
   compare bromine incorporation  into DBF formation in the single contactor  and  blended
   contactor effluents.

5.  Evaluate the impact of extrapolation of the integral breakthrough curve on blended effluent
   water quality.

6.  Verify the accuracy of the integral  breakthrough curve predictive models, which assume  an
   infinite number  of contactors, for  the prediction of the service life of finite numbers  of
   contactors.
                                      -15-

-------
This page intentionally left blank.
          -16-

-------
3   Materials and Methods


3.1  Experimental Approach


3.1.1   Rapid Small-Scale Column Test

The performance of full-scale GAC contactors was simulated using the rapid small-scale column
test (RSSCT).  Previous studies have shown that diffusion of natural organic matter (NOM) to
adsorption sites on GAC is proportional the particle size (Crittenden et al.,  1989; Sontheimer et
al., 1988). Therefore, by grinding the GAC to a smaller size, rates of adsorption are increased in
proportion to the ratio of full-scale to RSSCT GAC particle sizes, or scaling factor.  The scaling
factor also relates the RSSCT EBCT and superficial velocity to the full-scale contactor. If the
RSSCT utilized the full-scale contactor bed  length, extremely long columns requiring very high
inlet pressures would be required. However, studies have shown that adjustments to the RSSCT
Reynold's number utilized (between  0.1 and 1.0) have a negligible impact on results. Therefore,
by decreasing the RSSCT Reynold's number, a much shorter column can be designed (with
consequently shorter superficial velocities to maintain a constant EBCT).  Complete  details on
the RSSCT design for precursor removal studies can be found in the literature (USEPA, 1996;
Summers et al.,  1995; Summers et al., 1992;  Crittenden et al., 1991).

The RSSCTs designed in  this study followed the guidelines  outlined in  the GAC Precursor
Removal Studies section of the ICR Manual for Bench-  and Pilot-Scale Treatment Studies
(USEPA, 1996). A summary of the RSSCT design used for each run is given in Table 1.  For
most waters, a 20 minute full-scale EBCT contactor was simulated. However,  15 minute and 7.2
minute EBCT contactors were simulated for Waters 1 and 8,  respectively. The designs were
based on the estimated or known GAC influent TOC concentration (which can directly impact
the rate of breakthrough), the full-scale water temperature at the time of sampling, the full-scale
GAC particle size, and the full-scale EBCT simulated. The full-scale bed porosity was assumed
to be 0.45 for all runs. The minimum Reynold's number used ranged from 0.48 to 0.60.

3.1.1.1 GAC Preparation Procedures

Representative batches of Filtrasorb 300 (F-300), and Filtrasorb 400 (F-400), bituminous coal-
based  GAC, were obtained  from  the  manufacturer,  Calgon  Carbon  Corporation.   The
representative batch of the reactivated GAC/virgin GAC blend used for Water 4 was obtained
directly from the utility. Using a riffle splitter, a small (30 to 50 g) representative sample of the
GAC was obtained.  Using ajar mill, the GAC was ground to the needed mesh size.  Care was
taken to frequently remove and sieve the GAC in the jar mill.  The GAC was ground until the
entire sample passed through the upper mesh size sieve.  Usually, a recovery of 25 to 30 percent
was obtained, as defined by the amount of GAC retained between the two mesh size sieves and
divided by the total amount of GAC prior to grinding.

The ground  GAC  was  transferred  to a beaker, and covered with reagent  grade (adsorbed-
deionized) water. The GAC was washed by repeated additions and decantations of reagent grade
                                      -17-

-------
              Design parameter
                                                                                                                  Design value for each ater
                                                                                                                    45
oo
GAC manufacturer
GAC band name
GAC tpe
GAC mesh siE
Article diameter.d  c  (nm)

€kieral design parameters
Mimum avoids numbr.B      Smln  1
Full-scale operatingemperature C)
HSematic viscosity  vc (n 2s)
Bed porosity ec  J
bhsured dr)bd density     ps  fim  3)

RSSCT design parameters
BTnesh siE
Article diameter.d  s  (nm)
Siling'actor.B
                                                      Calgn Carbn Co.  Calgn Carbn Co.   Calgn Carbn Co.   i&oHrbj tend    Calgn Carbn Co.   Calgn Carbn Co.  Calgn Carbn Co.  Calgn Carbn Co.
                                                           F-8             F-8             F-8          F-E-8          F-8            F-8             F-8             F-8
                                                         Bituminous          Bituminous         Bituminous          Bituminous          Bituminous          Bituminous         Bituminous
                                                            a            e             a       m       a             B             a            a
                                                            606086              6
                                                          3032802
                                                          H-0          E-6          B-0           1-0          E-6          E-6          1-0          E-6
Bituminous
^faulic loadingate.v G (nnr) 6
Column diameter.D s (nm) 3
Flow rate.Q s (nfnin) 8
Full-scale empt^fed contact time, EBCT c (nin) 5
Estimated full-scale run time.t c T dap) a
Estimated BTun time.t G T dap) 3
tolume water reqired.V G [. Q
bfes GAC reqired.m G & 9
B"6mpt>bd contact time, EBCT s (nin) 9
Bed lengi.l s (:m) 3
2 8 S 8 2 8 8
8 09093 9
7 9 a a 9 s a
a a a a a a z
s s a 2 e 4 $
& a M a 2 a t
a B i e s a a
8968313
3989923
8 2 S 3 8 3 2
              Table 1  Summary of RSSCT design parameters for all runs

-------
water.  The reagent grade water was added at a high rate and turbulence, to stir up the GAC and
release fines.  The supernatant water containing GAC fines  was decanted after the GAC was
allowed to settle. Towards the end of the cleaning procedure, the sample was sonicated twice for
5 to 10 seconds. The sonication step helped loosen fines that were subsequently removed by the
addition and decantation of reagent grade water.

The GAC was dried in an oven at 80 to 90°C for 6 to 12 hours. The temperature was then raised
to between 100 and 110°C and the sample was  dried until it reached  a constant weight.  The
sample was removed and cooled inside a dessicator.  Once cooled, if not immediately used, it
was stored in a glass vial sealed with a lid with Teflon-lined septum until ready for use.

The dry bed density was measured using a sample of dried and cooled  GAC.  Stored GAC was
dried in an oven as described above prior to the dry bed density measurement. To measure the
dry bed density, a sample of the GAC was placed inside a 10-mL glass graduated cylinder to a
level of 5 to 9 mL.  The cylinder was tapped to pack the GAC.  A volume was measured  and
recorded.  This  GAC was then weighed on a balance.  The volume reading of the graduated
cylinder was checked and calibrated if necessary by  adding a known volume of water to it using
a 10-mL class A graduated pipette.  The GAC dry bed density  was calculated by dividing the
weight by the calibrated volume.

The calculated mass of GAC for each  RSSCT was  weighed, placed inside a clean beaker,  and
covered with reagent grade water.   The wetted GAC was usually allowed to sit for 12 to 24
hours, followed by placement in a vacuum for at least 1 hour to displace  the air within the pores.

3.1.1.2 RSSCT Column Setup

The ground GAC used for the RSSCT was packed in glass chromatography columns.  Due to the
range of GAC influent TOC concentrations, which correlates to the rate of TOC breakthrough,
columns with inner diameters ranging from 8.0 to 12.6 mm were utilized.  The 8.0 and 11.0 mm
inner diameter columns were standard.  Other column diameters (9.0, 10.0, and 12.6 mm) were
custom-ordered  and the GAC required additional  support to ensure the  GAC was  within the
effective length  of the column. The GAC support for these special order columns consisted of a
stainless steel screen (60 or 100 mesh size), Teflon beads, glass wool, and two stainless  steel
screens. The support for 8.0 and 11.0 mm  inner diameter columns consisted of two stainless
steel screens placed on top of the  Teflon fitting.  The mesh size of the screens utilized were
based on the ground GAC mesh size.  For 100x200 GAC, a 100 mesh screen and a 200 mesh
screen were used.  For 140x230 GAC, a 200 mesh screen and  a 325 mesh screen were used.  For
all column inner diameter sizes, a small amount of glass wool was placed inside the Teflon
fitting, supported by a 60 mesh size stainless steel screen.

The GAC  was added to the  columns as  a slurry and packed  by repeatedly tapping the column
sides. The 20 minute full-scale equivalent EBCT RSSCTs were packed into two columns of the
same inner diameter placed  in series.  Only reagent grade water was used during the packing
process.
                                     -19-

-------
3.1.1.3 Batch Influent Preparation

Prior to RSSCT testing, all water samples were filtered through a 1.0-jim nominal pore size glass
fiber cartridge filter. The cartridge filter was pre-rinsed with deionized water. Dilute solutions
of sulfuric acid and sodium hydroxide were used to maintain the influent pH within 0.1 pH units
of the target pH during operation of the RSSCTs.

3.1.1.4 RSSCT Monitoring

The effluent flow rates were monitored constantly to ensure that the flow rates were maintained
within 5 percent of the design flow rate. The calibration of the effluent flow rate control system
was checked at least three times daily and adjusted when  a flow rate differed by more than 3
percent from the design flow rate.  The system pressure was monitored daily.  The effluent TOC
concentration was monitored frequently so that samples could be taken at the required 5  to 8
percent increments of the average influent TOC concentration.

3.1.2  Bench-Scale Blended Water Quality Assessment Approach

To simulate the integral breakthrough curve obtained by blending multiple full-scale contactors
operated  in parallel-staggered mode,  the  entire  effluent  from  a  single GAC  contactor  was
collected in a reservoir and sampled over time.  In this study, the entire effluent from the RSSCT
was collected in a clean 30 or 55 gallon drum. The only effluent water not collected in the drum
was that  required  for monitoring and sampling of the RSSCT  effluent.   This  included TOC
monitoring and  sampling, UV254 sampling, and  SDS  chlorination.   Over time, the blended
effluent drum was sampled  and analyzed for TOC and UV254.  Samples were  also taken and
chlorinated under the same target SDS conditions as those applied to the single contactor study.
During every run,  TTHM and HAA9 were analyzed, and  during two runs, TOX was  also
analyzed, in addition to TTHM and HAA9. Ten percent of blended effluent samples taken were
sampled in duplicate and all analyses were conducted on  the duplicate  samples (field sample
duplicates).

The first discrete  sample taken from the RSSCT effluent also constituted the first blended
sample. Subsequently, seven samples were taken from the blended effluent drum, evenly spaced
over the  course  of the  run.   The RSSCTs  were operated until at least  70  percent  TOC
breakthrough was reached, and the last blended effluent sample was taken at that time. Since the
ICR required that twelve samples be taken from the RSSCT effluent at even increments of TOC
breakthrough, the  blended effluent sampling  schedule followed the RSSCT sampling.   The
second blended effluent sample was taken when  the third RSSCT effluent sample was taken.
The remaining blended effluent sample points were based on the RSSCT effluent sample number
as described in Table 2.  In practice, the blended effluent samples were taken from the drum just
prior to the addition of the RSSCT effluent sample listed in Table 2.  This way, the use of the
RSSCT effluent sample water for SDS chlorination had a smaller impact on the blended effluent
sample.  All blended samples were chlorinated under the same SDS conditions as the RSSCT
effluent samples.  For two  of the  waters,  (Waters 6 and 8) the run was  extended for  five
laboratory days after 70 percent breakthrough had occurred.   In these instances, eight blended
                                     -20-

-------
effluent samples were taken with the sixth at 70 percent breakthrough, the seventh 2.5 days later,
and the eighth five days later.
   Blended effluent     RSSCT effluent sample
    sample number             number
1
2
3
4
5
6
7
8
1
3
6
8
9
10
11
12
Table  2  Relationship between blended  effluent sample number  and RSSCT  effluent
sample number
3.1.3  DBP Formation Assessment

The blended effluent samples were chlorinated under the same target conditions as the RSSCT
effluent.  To minimize the volume  of water sampled from  the blended effluent reservoir, the
chlorine dose used was based on the single column effluent chlorine demand data, by relating
chlorine demand to TOC concentration. The sampling volume required was minimized based on
the DBFs to be analyzed.  For 6 of the 8 waters, the blended effluent sample volume was 800
mL,  sufficient for TOC, UV254, and SDS chlorination for the analysis of HAA9 and TTHM.
When TOX was also analyzed after chlorination, a 1,300 mL  sample was required.

For single contactor effluent  samples, chlorine demand  studies  were performed on the first
effluent sample and the influent water.  For each sample, three  125-mL  chlorine demand-free
amber glass bottles were used. A combined phosphate/borate buffer solution was added to each
bottle (2.0 mL/L) to maintain a constant target pH during incubation. Dilute solutions of sulfuric
acid  or sodium hydroxide were used to adjust the pH prior to chlorination if necessary.  Three
chlorine doses were selected based on the TOC  concentration of the water  and the results of a 5-
minute chlorine demand study (providing a relative measure of inorganic chlorine demand).
Each bottle was filled to 80 to 90 percent of capacity. Using a pipette, a measured amount of a
standardized chlorine solution (using Standard Methods 4500-C1  B) was  added to each bottle.
The bottle was then filled and capped head-space free.  The bottles were placed in  a constant
temperature  bath in the dark  for the  duration of the target incubation period.  A titrimetric
procedure (StandardMethods 4500-C1  F) was used to measure the free chlorine residual after the
holding time.  The data generated by the chlorine demand study was  used to estimate the
chlorine dose  to  achieve the  target  residual  for  all effluent  samples  by  correlating  TOC
                                      -21-

-------
concentration to chlorine demand.  When TTHM and HAA9 were analyzed, a clean chlorine
demand-free 500 mL amber glass bottle was used.  When TOX was also analyzed, a 1000 mL
bottle was used.   Caps with Teflon-lined septa were used.   The order of DBF sampling was
THMs, HAAs, and TOX.  DBF samples were taken in duplicate in  prepared  bottles  with
appropriate preservatives and quenching agents based on the analytical method.

3.1.4  Assessment of the Impact of Sampling on the Integral Breakthrough Curve

Discrete sampling from the RSSCT effluent and sampling from the blended effluent reservoir
was required to obtain water quality data to generate single contactor and integral breakthrough
curves. However, the impact of both types of sampling on the integral breakthrough curve were
unknown.  A model was developed to simulate the experimental procedure in an effort to assess
the impact of sampling on the integral breakthrough curve. Based on the model results, RSSCT
design and sampling procedures  were  optimized to minimize the impact of sampling on the
integral breakthrough curve.

The model was based on sampling the RSSCT effluent continuously in 3-L aliquots. It assumed
that these aliquots were sampled for TOC and UV254 to monitor breakthrough as needed.  Based
on the TOC concentration of samples  taken, effluent samples and duplicates  at ICR-required
TOC increments were identified and chlorinated under SDS conditions.

All samples were added to the reservoir in the order of sampling, and at even increments over the
course of the run, samples were taken  from the blended effluent reservoir for TOC and UV254
analysis  and  SDS  chlorination.   The results of these  analyses  comprised  the integral
breakthrough curve. The sample volume required was minimized, and any unchlorinated sample
remaining after analysis was complete was reintroduced to the reservoir.

Based  on these  experimental procedures,  a model  was  developed to assess the  impact  of
sampling on the integral breakthrough curve, prior to laboratory experiments.  The model inputs
were influent TOC concentration, empty-bed contact time (EBCT), column inner diameter, full-
scale GAC mesh size, RSSCT GAC  mesh size,  minimum  Reynold's number, and full-scale
operating temperature.  Based  on a correlation between influent TOC concentration and bed
volumes to 50 percent TOC breakthrough (Summers et  al., 1994; Hooper  et al., 1996), and
adjustments to the correlation to estimate run times to different levels of TOC breakthrough, a
TOC breakthrough curve was predicted.

From the estimated single contactor breakthrough curve, two blended effluent breakthrough
curves were generated. A theoretical integral breakthrough curve assumed no column effluent or
blended effluent  sampling occurred.  The experimental integral breakthrough curve assumed
typical discrete RSSCT effluent sampling, as described in the ICR Manual for Bench- and Pilot-
Scale Treatment  Studies,  and also included the sampling required from the blended effluent
reservoir to characterize the blended effluent water quality as described in Section 3.1.2.

Increasing the RSSCT column  inner diameter (and  thus increasing the ratio of volume water
passed to water required for analyses) decreased the impact of RSSCT effluent  sampling on the
integral breakthrough  curve.  The model also  demonstrated that  sampling  from the blended
effluent reservoir had the greatest impact on the difference  between theoretical and experimental
integral breakthrough  curves.   Thus,  by increasing the volume of water passed through the
                                      -22-

-------
RSSCT (by increasing the  column inner diameter or minimum Reynold's number) this effect
could be minimized.

After each GAC  run was complete, the experimental TOC integral breakthrough curve was
compared to an integral curve from which the impact of actual single contactor and blended
effluent  sampling  was  subtracted  (corrected  integral  breakthrough  curve).    For  TOC
breakthrough of each water, the two curves are compared in Appendix A.  Overall, the impact of
sampling on the integral breakthrough curve was very small, yielding a mean difference in TOC
concentration at the end of the integral breakthrough curve (where the difference between the
two curves was typically maximized) of 3 percent.  The difference ranged from 1 to 6 percent,
with shorter runs yielding higher percent differences.  In all cases, the TOC concentration at the
end of the experimental integral curve was higher than that in the corrected integral breakthrough
curve.

3.2   Data Analysis and  Modeling Approach


3.2.1  Logistic Function Models

In this study, the logistic function was used to fit GAC breakthrough data from 8 water sources
and GAC runs, and up  to 20 measured GAC effluent parameters (including DBF surrogates,
formed  DBF class sum parameters,  and formed DBF species).  In some cases, the logistic
function presented in Equation 1 1 did not satisfactorily fit the breakthrough curve data.  Three
modifications were made to the logistic function model to enhance the ability of the model to fit
breakthrough curve data. The models are described in Figure 5.  The first modification was to
include an additional parameter, improving the curve fit for breakthrough curves with moderate
to high immediate breakthrough levels. The step logistic function is as follows:
where C(i), B, D, and t are as defined previously in Section 1.3, Equation 11.  The term A0
represents the step applied to the logistic  function to match moderate  to  high immediate
breakthrough levels.   The term A represents the  asymptote to  which the logistic function is
approaching.  In this  study, the step logistic function model was used to curve fit all single
contactor TOC  breakthrough curves.  The step logistic function model is also adequate for
modeling the GAC effluent chlorine demand breakthrough curve, which typically has relatively
high levels of immediate breakthrough.  The addition of the step term also allows the curve to
have a  negative y-intercept,  which is necessary  for incorporation into the  step-lag logistic
function model described below.

Because many breakthrough curves (especially for SDS-DBPs) result in a relatively long initial
interval  where effluent concentrations are below reportable limits prior to  breakthrough,  a lag
was incorporated into the function to shift the logistic function outward, allowing for a better fit
of experimental breakthrough data. The run time at which initial breakthrough  above detectable
levels occurs is termed 4, and the step-lag logistic function is as follows:

                           0                          t
-------
                 O


                 o
                  8

                  o
                 o
C(f) = -
                                                            a. Logistic function
                                                             O
                                                             O
                                                             o
                                                                                            Ctf) =
                                                                                                                       c. Step-lag logistic function
                                                                                                                 Run time, t
to
                                            Run time,
                 O


                 c"
                 O
                  I
                  o
                 o
                                                        b.  Step logistic function
                                                             O

                                                             c"
                                                             g
                                                             u-i
                                                             re
                                                             o
                                                             o
                                                                                                = o
                                                                                                                    (tp,cp)
                                                                                                                  d. Step-lag-peak logistic function
                                                                                                                 Run time, t
                                            Run time, t
                Figure 5 Logistic function model curves

-------
In many cases, a best-fit of the data will yield a negative value for A0 (a negative y-intercept). In
essence, this shifts the logistic function downward, so that the beginning of the curve is negative
(as shown by the dotted line on curve C in Figure 5).  However, this occurs when t  < tb, so by
Equation 16, the result is set to zero. For ICR treatment study data analysis, the result will be set
to 50 percent of the MRL for DBF surrogates and species, and will be set to zero for DBF  sum
parameters.  This modification improves the ability of the logistic  function to fit breakthrough
data that are not symmetrical.

Under  certain conditions,  some brominated  DBF  species  exhibit increasing and  decreasing
breakthrough curves.  These  "peak"  curves can be modeled using the logistic function to the
maximum concentration. After this point, effluent concentrations decrease, and this  decrease is
modeled using a simple linear function.   Prior to the  point  of  peak concentration, C(i)  is
described by the step-lag logistic function model. The step-lag-peak logistic function model  is as
follows:

                    C(0 = 0                          ttp                          (20)

where Cp is the logistic function model best-fit concentration at tp, the run time  at which the peak
occurs, and S is the slope of the linear best-fit curve.  The following  algorithm was used to detect
"peak" breakthrough curves:

1. The measured peak concentration, C(tp), is at least 20 percent greater than  the concentration
   at the last observed data point, C(tf).

2. The run time tp is less than 80 percent of the run time of the last observed data point, tf.

3. The data point corresponding to tp is located prior to the penultimate observed data point.

In all cases, a best-fit of the logistic function model to the data was generated by least squares
minimization approach.  The coefficient of determination, R2,  was computed for all best-fits.
The  logistic function models were fit by non-linear least-squares using PROC NLIN in version
6.12 of the SAS system (Littell et al., 1996).  Additional details and SAS code are included in
Appendix B.

3.2.2  Outlier Methods

The  large amount of data that will be processed during the ICR treatment studies data analysis
necessitates the use  of automated curve fitting procedures. To ensure that the fitted  models are
robust  to extreme values, an outlier adjustment methodology that uses all experimental  data
                                       -25-

-------
points but limits the influence of deviant observations on the parameter estimates was developed
and used for this analysis.

Indiscriminate deletion of deviant observations can cause goodness of fit measures (i.e., R2) to be
unrealistically high.  Rather than deleting potential outliers, suspect observations were replaced
by less extreme values by the following procedure:

1.  Fit the logistic function  model to the observed data and determine approximate 95 percent
   prediction limits on the observations.

2.  Observations that exceed the threshold Predicted + (U95 - L95)*K are adjusted to Predicted
   + (U95 - L95)*K, where U95 is the upper 95 percent confidence limit and L95 is the lower
   95  percent confidence limit.  The constant K determines the magnitude of the adjustment,
   with larger values of K corresponding to fewer declared outliers and smaller adjustments to
   those that are detected.  The value for the constant K  was set to 1/3 based on simulation
   results that indicated a good balance between false alarms and power to identify substantial
   outliers (>3 standard deviations from the best-fit prediction). Approximately 2.5 percent of
   the data in the present study were identified as outliers and adjusted.

3.  Refit the logistic function model using adjusted values.

3.2.3  Direct Integration Approach

The  use of the integrated logistic function  model for a  given parameter to predict the integral
breakthrough curve  for that parameter is  termed  the  direct integration (DI) approach.  The
average value function  described by Equation 12 was applied to the expressions used to describe
the breakthrough curves presented in Section 3.2.1.  An  expression for the average value of the
step  logistic function (Equation 15) is:
                                                                                     (21)


Application of the average value function to the step-lag logistic function model (Equations 16
and 17) yields the following equations that represent the integral breakthrough curve:

                     C(t) = 0                                      t 4 is
more complex due to the use of two functions to describe the experimental data.  The application
of the average value function to  the three functions used to describe the experimental data over
the range [0, t] is given by the following equation:
                                       -26-

-------
                     C(t) = -\C(t)dt = -
                           t J         t
               16         P         '
               Jq(0^ + Jc2(0^ + jQ
(24)
where Ci(f), C^), and C3(/) represent Equations 18, 19, and 20, respectively, the three equations
used to model the experimental data.  For t < tb, the integral breakthrough curve is represented
by:
                                                                 t< tb
(25)
For tb tp, Equation 19 is evaluated from  tb to  tp,  and Equation 20 is evaluated from tp to t.
Combining the two equations yields:
^
D
                                 •Be~Dt^
                                + Be
                                    -Dtb
                                                                            t>tn
(27)
The logistic function integral approach to determine the integral breakthrough curve assumes an
infinite number of contactors operated in parallel-staggered mode.  This assumption was verified
for finite N = 2,  3, 4, 6, 10, and 20 contactors.  For finite numbers of contactors, numerical
integration of Equations 15 through 20 was performed as described by Equation 4.

3.2.4  Surrogate Correlation Approach

The surrogate correlation approach (SCA) is a procedure that simplifies and reduces the amount
of computations necessary to estimate DBF  formation in the effluent of staggered multiple
contactors  (blended effluent).  The SCA procedure relies  on a constant relationship between
TOC concentration  and  DBF formation in both the single contactor effluent and the blended
effluent.  To apply the SCA method to a GAC run, the single contactor TOC,  UV254, SDS-DBP
class sum  parameters, and SDS-DBP species breakthrough  curves are fit to  the appropriate
logistic  function  model.  The  SCA  procedure  steps are summarized  in  Section  1.5 and
represented schematically in Figure 4.

Typically, a linear or second order polynomial relationship exists between GAC effluent TOC
concentration and any other parameter.  The SCA method  does not determine or use these
correlations directly but instead assumes they exist and are constant in both the single contactor
effluent and blended effluent.  This relationship is exploited by linking the concentration of all
water quality parameters at a common single contactor run time. Using Equation 21, the integral
breakthrough  curve  is calculated for the TOC single contactor breakthrough curve only.  Then
the single contactor TOC concentration, to which the concentrations of all other water quality
                                       -27-

-------
parameters are linked, is applied to the run time at which an equivalent TOC concentration is
reached in the TOC integral breakthrough curve.

The SCA procedure can be performed to determine blended contactor run times for a given GAC
run and blended effluent treatment objective.  In  this study, the procedure was verified by
comparing the SCA integral breakthrough curve against experimental data.

3.2.5  Comparison of Methods for Predicting the Performance of GAC Contactors
       Operated in Parallel-Staggered Mode

During this study, the methods described in Sections 3.2.3 and 3.2.4 for predicting the integral
breakthrough curve were compared to the experimentally derived integral breakthrough curve for
eight GAC runs.  Using the logistic function models, a best-fit of the  experimental  blended
effluent data was derived.    The  logistic  function  models  were applied to the  integral
breakthrough curve data because the blended effluent data curves were similar in shape to those
encountered in the single contactor effluent.

Predictions obtained  by the SCA  method, the DI method,  and  the best-fitting model were
compared  using the calculated residual sums of squares (RSS)  from the experimental  blended
effluent data. For each parameter prediction, the prediction bias was determined by averaging
the calculated residuals for each model. An evaluation of the  bias shows whether the model
tended to overpredict or underpredict the observed data. For some parameters, a limited amount
of data were measured above the minimum reporting level (MRL).  In cases  where fewer than
six points above the MRL were present, curve  fitting was not performed.

3.2.6  Comparison of Single Contactor and Blended Effluent DBP Bromine
       Incorporation

To verify  the assumption  that the  impact of TOC on DBP  bromine incorporation would be
similar in  the single contactor and blended effluents, the bromine incorporation factors nBr and
n'Br were modeled  as  a polynomial function  of TOC and UV254 concentrations for all waters
simultaneously with two regression models:

                    Y = Wi + (3iX + |32X2                                            (28)

                    Y = Wi + (3iX + p2X2 +  yZ,                                       (29)

where Y is nBr or n 'Br and  X is TOC or UV254. The intercept W; is allowed to be different for
each water, to reflect natural differences in bromine incorporation across waters. In the second
model, different intercepts are fitted for single contactor (Z=0) and blended (Z=l) data.  The
additional  parameter y represents the change  in average bromine incorporation associated with
blending.  Since the first model is  a special  case of the second with y=0, the assumption that
blending will not impact average bromine incorporation would be  supported if the goodness of
fit of the two models is similar.  Models which allowed the linear and quadratic parameters (3i
and p2 to differ for single contactor and blending were considered in order to demonstrate that
blended and single contactor  scenarios are comparable in terms of the shape  of the bromine
incorporation profile as well as mean bromine  incorporation (i.e., the intercepts).  Although these
                                      -28-

-------
models  make  efficient use  of all of the data simultaneously  in  order  to  reach  a general
conclusion, the presence of the water-specific random effects W; makes ordinary least-squares
inappropriate.  SAS  PROC  MIXED was therefore used  to  estimate the  models via general
likelihood methods (Littell, et. al., 1996).

3.2.7  Breakthrough Curve Extrapolation

For two waters, the sensitivity  of blended  effluent  water quality to a  breakthrough  curve
extrapolation procedure was verified.  To verify the extrapolation approach,  two GAC runs were
operated for five laboratory days beyond 70 percent TOC  breakthrough. Based on the scaling
factor of each run, the five days were equivalent to 49 and 69 full-scale days for Waters 5 and 8,
respectively.  The appropriate logistic function model was applied to the abbreviated data  set,
which reached 70 percent TOC breakthrough.  Next, the appropriate logistic function model was
applied to the entire available data set for each parameter. The model fit for the abbreviated data
set  was  extrapolated to the total column run time,  and  the blended effluent water quality
predicted based on extrapolation of the abbreviated data set was compared to that using the  full
experimental data set.

During  the  ICR treatment  studies, only  the TOC integral breakthrough  curve will  be
extrapolated.  The  SCA method relies on projecting water quality data associated with  single
contactor effluent TOC concentration to the integral TOC breakthrough curve. Since the integral
TOC  breakthrough  curve typically reaches  only  40 to  70 percent  of the  single  contactor
breakthrough  curve,  DBF data  associated  with higher  TOC concentrations would not be
included.  By  extrapolating the integral  TOC breakthrough curve, the benefit of the data set is
increased, because DBF formation associated with higher GAC effluent TOC concentrations will
be included in the analysis.

3.3  Waters  Examined
3.3.1  Pretreatment and Water Quality

GAC runs on eight water sources were included in this study.  Pretreatment schematics for each
water source are shown in Appendix C. The influent to GAC TOC concentration for these water
sources ranged from 2.0 to 5.6 mg/L.  The specific UV absorbance for TOC (TSUVA, defined as
100*UV254/TOC) ranged from 1.6 to 2.3 L/mg-m.  A wide range of bromide concentrations were
measured, ranging  from 28 to 300 |ig/L.  The GAC influent  bromide to TOC ratio  (Br:TOC)
ranged from 10 to  68.  Water 1, received from Miami, Florida, was a groundwater.  Water 2,
received from Aurora,  Illinois, was a mixture of groundwater and surface  water.  The other 6
waters  were all surface waters. A summary of the source water, pretreatment, and treated water
quality is shown in  Table 3.
                                      -29-

-------
Water
1


2

3
4
5

6

7

8

Water
source
Miami -
Dade
County,
Florida
Aurora,
Illinois
Topeka,
Kansas
Davenport,
Iowa
Escondido,
California
Charleston,
S. Carolina
Sweetwater,
California
Greensboro,
N. Carolina
Pretreatment Treated water quality (GAC influent)
TOC UV254 TSUVA pH Alkalinity Total Bromide
(mg/L) (I/cm) (L/mg-m) (mg/L as hardness (ng/L)
CaCO3) (mg/L as
CaCO3)
Lime- 4.5 0.094 2.1 9.2 23 54 115
softening

Lime- 2.6 0.055 2.1 9.4 58 131 105
softening
Two-stage 2.4 0.048 2.0 9.0 30 133 160
softening
Conventional 3.0 0.065 2.2 7.1 127 217 29
(PAS*)
Conventional 3.1 0.051 1.7 7.4 109 225 70
(alum)
Conventional 2.6 0.060 2.3 6.3 9 29 140
(alum)
Conventional 5.6 0.109 2.0 7.6 138 221 300
(ferric)
Conventional 2.0 0.033 1.6 7.6 23 32 28
(alum)

BrTOC
(Hg/mg)
25


40

68
10
23

53

54

14

* PAS: polyaluminum sulfate
Table 3 Summary of pretreatment and water quality

-------
3.3.2  Simulated Distribution System Chlorination Conditions

Table 4 summarizes the SDS chlorination conditions for each water.  The SDS conditions were
chosen to reflect site-specific distribution system conditions at the time the water was sampled.
SDS incubation times ranged from 6 to 48 hours; incubation pH ranged from 7.4 to 9.2; target
free chlorine residual ranged from 0.75 to 1.50 mg/L; incubation temperatures ranged from 18 to
26°C.  All samples were buffered to  maintain the target incubation pH constant  during the
incubation period.
Water
1
2
3
4
5
6
7
8
Date sampled

July 24, 1998
September 22, 1998
September 1, 1998
September 23, 1998
January 7, 1999
October 12, 1998
October 22, 1998
October 6, 1998
Target SDS chlorination conditions
Incubation
time (hrs)
6
24
48
24
24
28
24
24
Incubation
temperature
(°C)
26
20
26
20
18
20
24
20
Free chlorine
residual
(mg/L)
0.75
0.80
0.80
0.75
0.80
1.50
0.75
1.00
pH
9.1
9.1
9.2
7.4
7.4
8.5
8.0
7.7
Table 4 SDS chlorination conditions

3.4  Analytical Methods

A list of all analytical methods and minimum reporting levels (MRLs) used during the study is
shown in Table 5.  All analyses were conducted at Summers & Hooper, Inc.  or Montgomery
Watson Laboratories as outlined in Table 6.
                                      -31-

-------
Analyte
Alkalinity
Bromide
Calcium hardness
Chlorine dose (solution
standardization)
Chlorine residual
HAA (DCAA, TCAA, MBAA,
DBAA, BCAA, BDCAA)
HAA (MCAA, CDBAA)
HAA (TBAA)
HAA (DCAA, TCAA, MBAA,
DBAA, BCAA, BDCAA)
HAA (MCAA, CDBAA)
HAA (TBAA)
pH
Temperature
Total hardness
Total organic carbon (TOC)
Total organic halide (TOX)
THM (CF, BDCM, DBCM, BF)
UV absorbance at 254 nm (UV254)
SM: Standard Methods
NA: Not applicable
Table 5 Summary of analytical
Analyses performed
Alkalinity, chlorine dose, chlorine
residual, pH, temperature, TTHM,
TOC, TOX, UV254
HAA9
Bromide, calcium hardness, total
hardness
HAA9
Waters
All
All
All
All
All
1-4,6-8
1-4,6-8
1-4,6-8
5
5
5
All
All
All
All
All
All
All


methods





Method
SM 2320 B
EPA 300.0 A
EPA 200.7
SM 4500-C1 B
SM 4500-C1 F
EPA 552.2
EPA 552.2
EPA 552.2
SM6251B
SM6251B
SM6251B
4500-H+ B
SM2550B
SM 2340 B
SM5310C
SM 5320 B
EPA 551.1
SM5910B


and MRLs
Waters
All
1-4,6-8
All
5
Minimum reporting level (MRL)
5 mg/L as CaCO3
20^g/L
5 mg/L as CaCO3
NA
0.2 mg/L as C12
1.0 ng/L (each analyte)
2.0 ng/L (each analyte)
4.0 ng/L
1.0 (ig/L (each analyte)
2.0 ng/L (each analyte)
4.0 ng/L
NA
NA
5 mg/L as CaCO3
0.50 mg/L
25 ng/L as Cl"
1.0 ng/L (each analyte)
0.009 cm"1



Laboratory
Summers & Hooper, Inc.
Summers & Hooper, Inc.
Montgomery Watson Laboratories
Montgomery Watson Laboratories
Table 6 Summary of laboratories conducting analyses

3.5  Experimental QA/QC Summary

As a part of this study, field duplicates were performed on 10 percent of samples analyzed in the
blended effluent. The field duplicates for DBFs were generated by duplicate SDS chlorination of
split GAC effluent samples.   The single contactor effluent samples were duplicated at a higher
rate, following ICR guidelines, which required that three field duplicates be collected from the
                                      -32-

-------
effluent of each RSSCT (25 percent duplication).  The results of all field duplicate analyses are
summarized in Table 7.
                                       -33-

-------
Analyte
  Count
Relative percent difference (RB)
                                           50th
                                        percentile
                                   fifen
                         Standard
                           deviation
(DC
\A
BSDX
B3NA5
B3NA6
B3NA9
             3
             2
             9
                 6
  2
  5
  5
THM Species
BSDF
BSDCM
BSDBCM
2
2
HAA Species
BSGIAA
BSDCAA
BSCAA
8
0
                                      NA
                          2           NA
                             8
BS)BAA
BSCAA
BXDBAA
BS)CBAA
BSAA
2
2
2
2
4
S
B
4
Q
8
S
01
S
%
Q
8
2
B
3
2
B: relative percent difference
NA: not applicate

Table 7  Summary of field duplicate precision for single contactor and blended effluent data
                                     -34-

-------
4   Results and Discussion
4.1  Overview

A substantial portion of the results presented attempt  to verify the underlying assumptions
behind the surrogate correlation approach (SCA) and examine the limitations of the application
of this method  for determining the integral breakthrough curve, a  relationship between single
contactor run time and blended contactor effluent water quality assuming contactor operation in
parallel-staggered mode.  The SCA procedure requires that all GAC breakthrough data are fit to
a logistic function model. From the perspective of data analysis of the ICR GAC treatment study
data, there are  several advantages to using  a model to fit experimental data.  Best-fit curve
parameters that adequately describe experimental data are less memory intensive than storing the
entire experimental  data set.  A best-fit  curve also facilitates interpolation and extrapolation to
estimate run times for given treatment objectives.  Use of a  best-fit model curve provides an
estimate  of the scatter in the data through the coefficient  of determination, and the  model
minimizes the impact of this scatter on run time estimates.  Finally,  a function that describes the
single contactor experimental data set is  a prerequisite for determining the integral breakthrough
curve.  Run time estimates generated by the integral breakthrough curve  are more applicable to
full-scale GAC operation where multiple contactors are operated in parallel-staggered mode to
increase the service time of each individual contactor.

The  SCA procedure (1) links  single contactor TOC concentration to  single contactor DBF
formation at a given run time (RTSc), (2) uses the direct integration (DI)  approach to model the
integral TOC breakthrough curve,  (3) determines the blended contactor effluent run time (RTBc)
at the TOC concentration used, and (4) applies the linked parameters to estimate DBF formation
at RTBc-  This approach allows the blended contactor run time to be determined for all correlated
parameters based on  the TOC integral breakthrough curve.  The SCA procedure assumes that:

1. The correlation between TOC and DBF precursors (concentration and speciation) observed
   in the single  contactor effluent  is maintained  in the  blended  effluent.   A  discussion of
   comparisons made to evaluate the consistency of these correlations is presented in Section
   4.2.

2. The three forms of the logistic function model developed can accurately describe the range of
   breakthrough of DBF precursors and DBF precursor surrogates evaluated.  The results of
   logistic function  model fits applied  to all parameters  and all GAC  runs are discussed in
   Section 4.3.

3. The TOC integral breakthrough curve calculated by the DI approach is an accurate predictor
   of blended effluent TOC concentration when multiple contactors  are operated  in parallel-
   staggered mode.  TOC integral breakthrough curves are compared to experimental  data in
   Section 4.3, while an evaluation  of the applicability of the infinite  contactor  assumption
   inherent by application of the DI procedure is contained in Section 4.5.

4. The integral TOC breakthrough  curve can  be extrapolated with reasonable  results.  This
   assumption was verified for two GAC runs, as discussed in Section 4.6.
                                       -35-

-------
Section  4.4  compares the integral breakthrough curves obtained  by both the SCA and DI
methods against experimental data for the eight GAC runs and all measured parameters.
                                     -36-

-------
4.2  Correlation between Surrogates and DBFs in Single Contactor and Blended
     Contactor Effluents
4.2.1   Correlation between Surrogate Concentration and DBP Formation

The correlation assumption is an important underlying foundation for the SCA procedure used to
model blended effluent water quality.  Verifying the assumption would demonstrate that DBP
formation is constant at a given TOC concentration in both the single contactor and blended
contactor effluent.  Since bromide is not adsorbed by GAC, the bromide to TOC ratio will also
be constant, at a given TOC concentration, so DBP speciation should not change between single
contactor and blended contactor effluents.

To verify the correlation assumption, plots of paired data were generated between TOC and other
parameters, such as SDS-TTHM,  SDS-HAA6,  SDS-CF,  SDS-DCAA,  etc.  Both summed DBP
classes and individual  compounds were examined.  For each water, two experimental data sets
were plotted for comparison:  single contactor effluent and blended effluent. The entire graph set
is  included in Appendix D, which includes correlations based  on both  TOC and UV254  as
surrogates. Overall,  both parameters  served well as surrogates for  DBP formation, and the
correlation between single contactor and blended effluent data was approximately equivalent.  A
representative sample of the results obtained are discussed in this section.

Overall, the correlations observed between TOC and formed DBFs in the blended effluent were
very similar to those observed in the single contactors for each water.   Figures 6 and 7 show the
correlation of UV254,  SDS-TTHM, SDS-HAA6, and SDS-TOX to TOC for both  the  single
contactor and blended effluent of Waters 1  and 2 (with the exception of SDS-TOX, for  which
blended effluent samples were not analyzed in these waters).  The correlation between TOC and
UV254 in the blended effluent was very similar to that in the  single contactor  effluent for these
two waters, which was typical of that observed in most cases.  The greatest difference in blended
effluent and single contactor correlations for TOC and SDS-TTFDVI occurred for Water 8 (Figure
8) which showed a  similar disparity between  the blended effluent and  single  contactor
correlations for TOC and UV254. Still, at a given TOC concentration,  SDS-TTHM formation in
the blended effluent was within 5 |ig/L of that in the single contactor effluent.

For the THM species, correlations between single  contactor and blended effluent TOC and
formed concentrations are shown in Figures  9 and 10 for Waters 2 and 3, respectively.  For all
four species, a very good agreement existed between single contactor and blended effluent TOC
correlations. This includes SDS-BF, which yielded a "peak" curve for these two runs.  The peak
also occurred in the blended effluent, and the single contactor and blended effluent yielded very
similar curves when  SDS-BF concentrations were  plotted against TOC.   The correlation  of
formed THM compounds between single contactor and blended effluent was  very good  for all
waters.  The  largest difference between correlations seen in the single contactor and blended
effluent were observed for Water 8, shown  in Figure 11, for which the formed THM species
concentrations were low.

Four of the predominant HAA species formed (DCAA, TCAA, DBAA, and BCAA) correlated to
TOC concentration for both single contactor and blended effluents are shown in Figures 12 and
                                      -37-

-------
13 for Waters 5 and 7, respectively.  For Water 5 both single contactor and blended effluent
formed HAAs matched very well.  For Water 7, the correlations for  SDS-DBAA and SDS-
BCAA also matched very well.  The single contactor and blended effluent correlations between
TOC and SDS-DCAA and SDS-TCAA showed slight differences, not exceeding 3 ug/L.  The
formed levels of these two compounds were low for this water.

Correlations between single contactor and blended  effluent for SDS-DCBAA,  SDS-CDBAA,
SDS-HAA5, and SDS-HAA9 are shown for Waters 2 and 4 in Figures 14 and 15, respectively.
Water 2 is shown as it yielded the largest difference in the correlation between single contactor
and blended effluent for these parameters, although formed DCBAA and CDBAA levels were
low.  Water  4 is more representative  of what  was typically observed, with the  correlation
between TOC and these parameters in the blended effluent matching that observed in the single
contactor very well.  For SDS-TBAA, only Water 7 yielded measured levels above the MRL (4
ug/L).  The correlation between TOC and formed TBAA in the blended effluent agreed well
with that observed in the single contactor effluent, as shown in Figure 16.

The use of UV254 as  a surrogate for these correlations  instead of TOC was investigated.  In a few
cases, UV254  as a surrogate yielded better correlations between  single contactor and  blended
effluent formed DBFs than did TOC.  However, in other instances, and approximately the same
number of cases, the correlations  using TOC  were superior.   Because the results for the
correlations using UV254 did not improve over those using TOC, the use of TOC as the surrogate
in the SCA procedure was continued.  Appendix D also summarizes the correlations observed
based on UV254.

In summary, this  analysis shows  that the  correlation  between  DBF  formation and TOC
concentration in the single contactor and blended contactor  effluents is constant.  This find is
significant because it verifies one of the underlying assumptions behind the SCA procedure, that
TOC concentration can be correlated to DBF formation in the single  contactor effluent,  and that
this correlation can then be applied  to blended contactor effluent.  Section 4.3 will address how
single contactor effluent data can be modeled  as a function of run time so that this correlation
can be utilized to predict blended contactor effluent DBF formation.

4.2.2  Correlation  between Surrogates and DBP  Speciation

For all four  models  relating  bromine  incorporation  for  THMs  and HAA9  (nBr and  n'Br,
respectively) to TOC and UV254, the polynomial models described in Section 3.2.6 were found to
be appropriate.  The model variance over all four combinations of nBr and n'Br as a function of
TOC and UV254 yielded R2 values  greater than 0.92 in each case.  Although the effect (y) of
blending on mean bromine incorporation was found to be statistically significant, including this
effect improved R2  values trivially  (less than  1 percent for  all four cases) by.  There was no
evidence that different linear or quadratic terms were needed to describe the single contactor and
blended effluent data. For THMs, Figures 17 through 24 show the results of fitting second order
polynomials to the relationship between nBr and TOC, combining single and  blended contactor
data for each water.   For HAA9, similar results for the relationship  between n'Br and TOC are
shown in Figures 25 through 32.  The second order polynomial curve fits for the correlation
between nBr and UV254 and n'Br and UV254 for both single and blended contactor data are shown
in Figures 33  through 48.
                                      -38-

-------
This analysis shows that there is no significant difference between DBF bromine incorporation
between single contactor and blended contactor effluents at equivalent TOC or UV254 values.
This conclusion supports the use of the SCA procedure, especially for individual DBF species,
because the SCA assumes that DBF formation and speciation in the single contactor effluent at a
given TOC concentration is equivalent to that in the blended contactor effluent at the same TOC
concentration.
                                      -39-

-------
0.07 -,

0.06 •
0.05 •
"E
^ 0.04 •
- — -
a

-------
       0.04 -i
       0.03 •
    o
       °-02-
       0.01 •
       0.00
               n Single contactor
               • Blended effluent
n
          0.0
                        0.5
                                      1.0
                                 TOC (mg/L)
                                                   1.5
                                                                2.0
                                                                            100 -i
                                                                             75-
                                                                             50-
                                                                         w
                                                                             25'
                                                                               0.0
                              0.5            1.0
                                        TOC (mg/L)
                                                                                                                       1.5
                                                                                                                                     2.0
      30
      20 •
   w
   Q
   w
        0.0
                      0.5
                                    1.0
                                TOC (mg/L)
                                                  1.5
                                                                 2.0
                                                                            200 -
                                                                            150 -
           g
           w
                                                                            100 -
                                                                             50-
                                                                               0.0
                              0.5           1.0           1.5           2.0
                                        TOC (mg/L)
Figure 7  Correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 2

-------
                    0.03 n
                    0.02 •
a
>
                    0.01 •
                    0.00
                           n Single contactor
                           • Blended effluent
                       0.0
                    0.5           1.0

                             TOC (mg/L)
                                                               1.5           2.0
                                                                                        50 n
                                                                                        25-
                                                                                     H
                                                                                          o.O
                                                                                                        0.5           1.0

                                                                                                                 TOC (mg/L)
                                                                                                                                   1.5           2.0
to
                  20 n
                   10-
                OT
                D
                OT
                       -•-•-[•-•-
                    0.0           0.5
                                                1.0           1.5

                                            TOC (mg/L)
                                                                            2.0
                                                                       150 n
                                                                                     O 100 •
                                                                                     8
                                                                          0.0
                                                                                        0.5           1.0           1.5

                                                                                                 TOC (mg/L)
                                                                                                                                2.0
             Figure 8 Correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 8

-------
       25 n
       20 •
    0) 15 •
    LL
    o
       10-
    w
        5 •
         0.0
n Single contactor effluent


• Blended effluent
                       0.5            1.0

                                TOC (mg/L)
                                                  1.5
                                                                2.0
                                                                           30 n
                                                                           25 '
                                                                           20
                                                        o  15 •
                                                        Q
                                                        m

                                                        w
                                                        Q  10 •
                                                                             0.0
                                                                                           0.5
                                                                                        1.0


                                                                                    TOC (mg/L)
                                                                                                                      1.5
                                                                                                                                    2.0
      30
      25 •
   0  15 •
   m

   9
   w
   Q  10 •
   w
       5 •
        0.0
                      0.5            1.0


                                TOC (mg/L)
                                                  1.5
                                                                2.0
                                                                           15 -i
                                                                           10-
                                                            5 -
                                                                             0.0
                                                                           0.5           1.0


                                                                                    TOC (mg/L)
                                                                                                                      1.5
                                                                                                                                    2.0
Figure 9  THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 2

-------
20 i


15 •
1
LL 10-
Q
5 •
0 •
0
40 -1
n Single contactor effluent
35 •
• Blended effluent
n 30 .
n ^
n S25'
0 20 •
D
in
OT 15 •
,_, Q
I I £f)
10 •
•D
-• — B-rQBD — i 	 i 	 1 	 i 	 1 	 i 	 1 0

fl
n
n
^°"
•
= •
•
n
0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
30 -I
25 •
a
0 15 •
m
p
OT
D 10 •
OT
5 •
0 •
35 -,
B 30 •
D
25 •
5
LL.
D CO
• § 15'
B n w •
10 ;
BH _
o •
-• 	 . 	 1 	 . 	 1 	 . 	 1 	 . 	 1 n


• „ fn
D tr
n
'
        0.0
0.5           1.0           1.5
          TOC (mg/L)
                                                             2.0
                                                                           0.0
                                                                                        0.5           1.0
                                                                                                  TOC (mg/L)
                                                                                                                   1.5           2.0
Figure 10 THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 3

-------
25 •
20 •
SDS-CF (pg/l
O Ol
5 •
0 •
0
s •
n Single contactor effluent n
• Blended effluent 4 .
n
=d
n ^ 3 •
o
. * OT1"
. n g

•
• D
n ° n n D n
• n
• n
n
0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
15 -I
5 10-
O
m
D
OT
D 5 •
OT
0 •
0
2 -i
n „
• D)
— i
D LT 1 .
• n CQ
• D OT
n OT
0
• D
1


'

-1 	 • • !_•"• 	 !_• — • 	 L«-|_l 	 rl_| 	 !• 	 r|_| 	 1_| — LJ 	 LJ — ' 	 1
0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
Figure 11 THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 8

-------
1


6 •

_i
s
J
o
Q
ch
Q
w
2 •


n .
i •
n Single contactor effluent

6 •
• Blended effluent


— i
n S
r;
o
• h- 3 •
n n w
i ^
• W 2 .
1
Q
• ' ™"° 1
-• — j~i 	 n 	 , 	 . 	 . 	 , 	 . 	 . 	 1 	 1 	 1 	 1 n .

n
n
n


n

n
n

_ •

g

D
•n n
-• — .• — m~\ — rm 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 .
                     0.0       0.5
                                        1.0       1.5       2.0
                                             TOC (mg/L)
                                                                    2.5       3.0
            0.0       0.5       1.0       1.5       2.0       2.5
                                    TOC (mg/L)
                                                                                                                                                 3.0
ON
4 •
3 •
1 •
m
Q
i
w
Q
W
1 •

n .
H •
D°D
••• "a
f a '
•n § 4'
_ DQ
• i
• w
D Q
W
2 •
D
-• — JM — • 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , n .
D
D
n n
•H"


^^

•D
rf
n
-• 	 r« 	 • 	 . 	 . 	 . 	 . 	 , 	 , 	 , 	 , 	 , 	 ,
                    0.0       0.5        1.0       1.5       2.0        2.5
                                            TOC (mg/L)
3.0          0.0       0.5
                               1.0       1.5        2.0
                                    TOC (mg/L)
                                                                                                                                        2.5       3.0
             Figure 12  HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 5

-------

• 0
OT
5
E



n
n
n n
n n
• "
D
i1 "°
• . •, — . — . — . — . — . — . — . — . — . — . — . — . — . — . — . — .
                     0.0    0.5    1.0    1.5    2.0    2.5    3.0    3.5    4.0    4.5
                                             TOC (mg/L)
0.0    0.5    1.0    1.5    2.0   2.5   3.0   3.5   4.0   4.5
                       TOC (mg/L)
             Figure 13 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 7

-------
oo
3 -


„ 	 .

0) 2-
<:
m
0
Q
W ,.
Q 1 '
W
n Single contactor effluent
• Blended effluent n
n
• • D —
™ n _i
0)2-
• n D <:
• n m '
D D
O
D £/)
D 1 '
OT


^
•

n n




_


Oi — — u — u u .,.,., u -, 	 m — m |_j 	 H _j 	 |_|-«i 	 LB 	 rl_l— LJ 	 U 	 LJ-i 	 1 	 1
0.0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
20 -, 30 -,



15 -
5"
01
in
<, 10-
I


D
OT
5 -




0 -

n
25
n
^20
n n o)
• -5
• n en
< 15
• ~r
,
n co
§10

•
5
^H

••
n
n
• D

n
i
n

1
n

• n

• D

• D
D

u -t ™ • . - . - . - . u i 	 m 	 , 	 1 	 , 	 1 	 , 	 1 	 , 	 1
0.0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
            Figure 14 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 2

-------
6 -I

"B) 4 •
^.
DQ
O
Q
ih 0
Q z '
W
0 -
0
n Single contactor effluent D
• Blended effluent
n n 5"
0)
§
n °m w
n Q
• n" • w



• •HI M •!—!•• 1— 1 1— • 1— 1 1— 1 1— 1 1— 1 1— 1
0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
40 -i

30 -
LO
<• 20-
X
ch
Q
w
10 •
0 •
0
50
n
40
D 2 30
D en
X 20
n Q
• D w
• D 10
• " n
. D

n

DD
D
• D
D °
m B n"
--a
0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
Figure 15 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 4

-------
12 -i

10 •

^
0) 8 •
^
^ c
CO D •
O
9
w
Q 4 •
w

2 •
0 •
n Single contactor effluent n
• Blended effluent

IT
d "ra
^
n <
D
• n V
n M
• 0
n OT
•n
_D
t] "
-B-~ — •, 	 •! 	 , 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 .
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
TOC (mg/L)
8 -i


6 •

"en
?4-
CQ
1 	
1
OT
D
OT
2 •



0 •
7 -,

6 •

5 •

4 •


3 •


2 •

1 .


n n
n
n
D n
u
m
a

a




a

I • ' — VI 	 MJ 	 !• — ' • 1 	 ' 	 1 	 ' 	 1 	 ' 	 1 	 ' 	 1 	 ' 	 1 	 ' 	 1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
TOC (mg/L)
80 i
D
D
• •
n
1
° 1
*t
OT
D
OT
D



• . •. — n — .• . • . — . — , — . — , — . — , — . — , — . — , — . — ,

70
60

50
40

30

20

10

n

a
a

n
D n
fm

•n
rj


D •
•
f
m I
       0.0    0.5    1.0    1.5    2.0    2.5    3.0    3.5    4.0    4.5
                               TOC (mg/L)
0.0    0.5    1.0    1.5    2.0   2.5   3.0   3.5   4.0   4.5
                       TOC (mg/L)
Figure 16 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 7

-------
   2.5
C/5
^
I
2.0 -
   1.5 -
o
Q_
O
O
CD
1.0 -
•|  0.5 -
o
m
   0.0
         Water 1
      TOC: 4.5 mg/L
      Bromide: 115 ug/L
 D Single contactor effluent

 • Blended effluent

	Best fit (RA2 = 0.948)
                                TOC concentration (mg/L)


Figure 17 Correlation between single contactor and blended effluent TOC
concentration and THM bromine incorporation factor (n) for Water 1
   3.0
   2.5 -
   2.0 -
o
CD
   1.5 -
o
Q_
O
o
1.0 -
CD
g
E
2
oo
0.5 -
   0.0
         Water 2
      TOC: 2.6 mg/L
      Bromide: 105 ug/L
     0.0
 n Single contactor effluent
 • Blended effluent
	Best fit (RA2 = 0.772)
                     0.5                1.0
                             TOC concentration (mg/L)
      1.5
2.0
Figure 18 Correlation between single contactor and blended effluent TOC
concentration and THM bromine incorporation factor (n) for Water 2
                                  -51-

-------
CO

^

I
   3.0
   2.5 -
   2.0 -
o
CD
   1.5 -
o
Q_

O
O
   1.0 -
CD
c

E
o

m
   0.5 -
   0.0
         Water 3
         TOC: 2.4 mg/L

         Bromide: 160 pg/L
     0.0
                      D Single contactor effluent

                      • Blended effluent

                     	Best fit (RA2 = 0.978)
                        0.5                1.0

                                TOC concentration (mg/L)
                            1.5
              2.0
Figure 19 Correlation between single contactor and blended effluent TOC

concentration and THM bromine incorporation factor (n) for Water 3
   1.00
to
^
I
   0.75 -
o
CD
o  0.50 -

"CD
o
Q_
o
o
o>  0.
g

E
£
oo
    .25 -
   0.00
         Water 4
          TOC: 3 mg/L

          Bromide: 29 ug/L
      0.0
                     0.5
                      D  Single contactor effluent

                      •  Blended effluent

                      	Best fit (RA2 = 0.874)
   1.0            1.5

TOC concentration (mg/L)
2.0
2.5
Figure 20  Correlation between single contactor and blended effluent TOC

concentration and THM bromine incorporation factor (n) for Water 4
                                  -52-

-------
   2.0
I- 1.5 -
 (_

£
 t_
 o


£

 o 1.0 -
"H—'
 E
 o

 e-

 8
 c
 CD
 C

 E

 2
 GO
   0.5 -
   0.0
         Water 5
TOC: 3.1 mg/L


Bromide: 70 ug/L
      0.0
        0.5
                                                      n Single contactor effluent


                                                      • Blended effluent


                                                     	Best fit (RA2 = 0.406)
                             1.0         1.5         2.0


                                TOC concentration (mg/L)
  2.5
3.0
Figure 21 Correlation between single contactor and blended effluent TOC

concentration and THM bromine incorporation factor (n) for Water 5
   2.5
 to

 1 2.0 H
£
 i_
 o
   1.5 -
 o 1.0-1


 8
 CD
00
   o.o
         Water 6
          TOC: 2.6 mg/L


          Bromide: 140 ug/L
      0.0
          0.5
                                              n  Single contactor effluent


                                              •  Blended effluent


                                             	Best fit (RA2 = 0.781)
                                  1.0             1.5


                                TOC concentration (mg/L)
2.0
2.5
Figure 22 Correlation between single contactor and blended effluent TOC

concentration and THM bromine incorporation factor (n) for Water 6
                                 -53-

-------
   2.5
   2.0 -
£
 t_
 o
   1.5 -
E
o  1.0 H



8


O>

|  0.5 H


2
GO
   0.0
          TOC: 5.6 mg/L


          Bromide: 300 ug/L
                                D Single contactor effluent


                                • Blended effluent


                               	Best fit (RA2 = 0.981)
                     1
           2             3


        TOC concentration (mg/L)
Figure 23 Correlation between single contactor and blended effluent TOC

concentration and THM bromine incorporation factor (n) for Water 7
   1.00
   0.75 -
£
 i_
 o
o  0.50 -
'-4—I


O

e-

8


'^  0.25 -


I

2
00
   0.00
          Water 8
          TOC: 2 mg/L


          Bromide: 28 ug/L
                               D Single contactor effluent


                               • Blended effluent


                              	Best fit (RA2 = 0.896)
       0.0
0.5                1.0


        TOC concentration (mg/L)
                                                           1.5
2.0
Figure 24 Correlation between single contactor and blended effluent TOC

concentration and THM bromine incorporation factor (n) for Water 8
                                 -54-

-------
   2.5
o>
<  2.0 -I
I
•§  1.5 -
o
"co
 _
 o
 o
o>
•|  0.5 -\
£
00
   0.0
         Water 1
      TOC: 4.5 mg/L
      Bromide: 115 ug/L
                                                     D  Single contactor effluent
                                                     •  Blended effluent
                                                     — Best fit (RA2 = 0.751)
      0123
                                TOC concentration (mg/L)

Figure 25 Correlation between single contactor and blended effluent TOC
concentration and HAA9 bromine incorporation factor (n1) for Water 1
   3.0
<  2-5 •
X
!  2-0 H
o
•s
§  1.5 -
o
Q.
O
O
1.0 -
o>
g
E
o
m
0.5 -
   0.0
         Water 2
      TOC: 2.6 mg/L
      Bromide: 105 ug/L
     0.0
                                                      D Single contactor effluent
                                                      • Blended effluent
                                                     	Best fit (RA2 = 0.423)
                     0.5                1.0
                             TOC concentration (mg/L)
1.5
2.0
Figure 26 Correlation between single contactor and blended effluent TOC
concentration and HAA9 bromine incorporation factor (n1) for Water 2
                                  -55-

-------
   3.0
   2.5 -
t:  2.0 -
o
•s
o

"co
t_
o
Q.

O
O
g

o>
c

E

£
m
   1.5 -
   1.0 -
   0.5 -
   0.0
         Water 3
         TOC: 2.4 mg/L

         Bromide: 160 ug/L
     0.0
                                            D  Single contactor effluent


                                            •  Blended effluent


                                           	Best fit (RA2 = 0.900)
              0.5                1.0


                      TOC concentration (mg/L)
                                                            1.5
              2.0
Figure 27 Correlation between single contactor and blended effluent TOC

concentration and HAA9 bromine incorporation factor (n1) for Water 3
   1.25
o>
•§  0.75 -
   0.50 H
 _
 o
 o
O)

•|  0.25 H

o

m
   0.00
         Water 4
                                              D  Single contactor effluent


                                              •  Blended effluent


                                             	Best fit (RA2 = 0.504)
TOC: 3 mg/L


Bromide: 29 ug/L
      0.0
           0.5
                                   1.0             1.5


                                TOC concentration (mg/L)
2.0
2.5
Figure 28  Correlation between single contactor and blended effluent TOC

concentration and HAA9 bromine incorporation factor (n1) for Water 4
                                  -56-

-------
   1.5
o>
<
<
o
"G
g

"co
t_
o
Q.


8
g

o>
g

E
O
(_
m
   o.o
           Water 5
TOC: 3.1 mg/L


Bromide: 70 ug/L
                      n  Single contactor effluent


                      •  Blended effluent


                     	Best fit (RA2 = 0.283)
     0.0
        0.5
1.0          1.5          2.0


  TOC concentration (mg/L)
   2.5
3.0
 Figure 29 Correlation between single contactor and blended effluent TOC concentratioi

 bromine incorporation factor (n1) for Water 5
   2.5
   2.0 -
 a
 o
    1.5 -
 o  1.0 -
 e-
 o
 o
   0.5 -
 E
 2
 m
   0.0
           Water 6
                Co
TOC:  2.6 mg/L


Bromide:  140 ug/L
                      n Single contactor effluent

                      • Blended effluent


                     	Best fit (RA2 = 0.926)
      0.0
           0.5
     1.0             1.5


  TOC concentration (mg/L)
2.0
2.5
Figure 30 Correlation between single contactor and blended effluent TOC concentratioi

bromine incorporation factor (n1) for Water 6
                                  -57-

-------
   2.5
 o>

 3 2.0
 I
 o
 •Q 1.5
 o
 '-4—'
 CD

 o 1.0
 t_
 o
 o
o>
I  0.5


m
   0.0
           Water 7
           TOC:  5.6 mg/L


           Bromide:  300 ug/L
                                                D Single contactor effluent

                                                • Blended effluent

                                               	Best fit (RA2 = 0.422)
                     1
                                2              3

                             TOC concentration (mg/L)
Figure 31 Correlation between single contactor and blended effluent TOC concentratioi

bromine incorporation factor (n1) for Water 7
   0.5
 o>
 <
 <
 o
 Q_

 O
 O
 o>
0.4 -
   0.3 -
   0.2 -
 .S 0.1 -
 E
 £
 00
   0.0
      0.0
           Water 8
        TOC: 2 mg/L


        Bromide: 28 ug/L
                                                  D Single contactor effluent

                                                  • Blended effluent

                                                  — Best fit (RA2 = 0.891)
                    0.5               1.0

                            TOC concentration (mg/L)
                                                           1.5
2.0
Figure 32  Correlation between single contactor and blended effluent TOC concentratioi

bromine incorporation factor (n1) for Water 8
                                 -58-

-------
   2.5
CO
^
I
2.0 -
   1.5 -
o
Q_
O
O
CD
1.0 -
•|  0.5 -
o
m
   0.0
      UV-254: 0.094 1/cm
      Bromide: 115 ug/L
                                                       D Single contactor effluent

                                                       • Blended effluent

                                                      	Best fit (RA2 = 0.962)
     0.00       0.01        0.02       0.03        0.04       0.05
                         UV absorbance at 254 nm, UV-254 (1/cm)
                                                                0.06
                                                                              0.07
Figure 33 Correlation between single contactor and blended effluent UV
absorbance and THM bromine incorporation factor (n) for Water 1
   3.0
   2.5 -
I
t_
o
   2.0 -
o
CD
5  1.5 -
CD
o
Q_
   1.0 -
f  0.6H
00
   0.0
         Water 2
         UV-254:  0.055 1/cm
         Bromide: 105 ug/L
                                                    n Single contactor effluent
                                                    • Blended effluent
                                                   	Best fit (RA2 = 0.781)
     0.00
                    0.01               0.02               0.03
                      UV absorbance at 254 nm, UV-254 (1/cm)
                                                                              0.04
Figure 34 Correlation between single contactor and blended effluent UV
absorbance and THM bromine incorporation factor (n) for Water 2
                                  -59-

-------
CO
^
I
   3.0
2.5 -
   2.0 -
o
CD
   1.5 -
o
Q_
O
O
1.0 -
CD
c
E
o
m
0.5 -
   0.0
         Water 3
      UV-254: 0.048 1/cm
      Bromide: 160 pg/L
                                                      D  Single contactor effluent
                                                      •  Blended effluent
                                                     	Best fit (RA2 = 0.974)
     0.00
                          0.01                     0.02
                      UV absorbance at 254 nm, UV-254 (1/cm)
                                                                             0.03
Figure 35 Correlation between single contactor and blended effluent UV
absorbance and THM bromine incorporation factor (n) for Water 3
   1.00
to
^
I
   0.75 -
o
CD
o
"CD
o
o.
o
o
c
CD
I
m
   0.50 -
   0.25 -
   0.00
          Water 4
          UV-254: 0.065 1/cm
          Bromide: 29 ug/L
                                                    n  Single contactor effluent
                                                    •  Blended effluent
                                                   	Best fit (RA2 = 0.900)
      0.00
                     0.01               0.02               0.03
                       UV absorbance at 254 nm, UV-254 (1/cm)
                                                                             0.04
Figure 36  Correlation between single contactor and blended effluent UV
absorbance and THM bromine incorporation factor (n) for Water 4
                                  -60-

-------
   2.0
I-  1.5 -
(_
£
t_
o

£
o  1.0 -
"H—'
E
o
e-
8
g
'CD  0.5 -

E

m
   0.0
         Water 5
          UV-254: 0.051 1/cm

          Bromide: 70 ug/L
     0.00
                                                  D  Single contactor effluent

                                                  •  Blended effluent

                                                 	Best fit (RA2 = 0.410)
                       0.01               0.02              0.03

                        UV absorbance at 254 nm, UV-254 (1/cm)
                                                                            0.04
Figure 37 Correlation between single contactor and blended effluent UV
absorbance and THM bromine incorporation factor (n) for Water 5
   2.5
to
^
I
1_
£
i_
o
   2.0 -
   1.5 -
o
e-
8
c
CD
   1.0 -
00
   o.o
         Water 6
          UV-254: 0.06 1/cm

          Bromide: 140 ug/L
     0.00
                                                   n Single contactor effluent

                                                   • Blended effluent

                                                  	Best fit (RA2 = 0.783)
                       0.01               0.02              0.03

                        UV absorbance at 254 nm, UV-254 (1/cm)
                                                                            0.04
Figure 38 Correlation between single contactor and blended effluent UV
absorbance and THM bromine incorporation factor (n) for Water 6
                                 -61-

-------
   2.5
   2.0 -
£
 t_
 o
   1.5 -
E
o  1.0-1


8
c

o>

•|  0.5 H


2
m
   o.o
          UV-254: 0.109 1/cm


          Bromide: 300 ug/L
                                         D  Single contactor effluent


                                         •  Blended effluent


                                        	Best fit (RA2 = 0.982)
     0.00       0.01        0.02       0.03       0.04       0.05


                         UV absorbance at 254 nm, UV-254 (1/cm)
                                                        0.06
                                                                            0.07
Figure 39 Correlation between single contactor and blended effluent UV

absorbance and THM bromine incorporation factor (n) for Water 7
   1.00
   0.75 -
£
 i_
 o
 5 0.50 -
 o
 e-
 8
 O)
 g

 E

 £

 00
   0.25 -
   0.00
          Water 8
UV-254: 0.033 1/cm


Bromide: 28 ug/L
                                                   n Single contactor effluent


                                                   • Blended effluent


                                                  	Best fit (RA2 = 0.880)
      0.00
                   0.01                     0.02


               UV absorbance at 254 nm, UV-254 (1/cm)
                                                                            0.03
Figure 40  Correlation between single contactor and blended effluent UV

absorbance and THM bromine incorporation factor (n) for Water 8
                                 -62-

-------
   2.5
o>
<  2.0 -I
I
•§  1.5 -
o
"co
o  1.0 H
l_
o
o
g
O>
•|  0.5 -|
£
m
   o.o
         Water 1
         UV-254:  0.094 1/cm
         Bromide: 115 ug/L
                                                        D Single contactor effluent
                                                        • Blended effluent
                                                       	Best fit (RA2 = 0.746)
     0.00       0.01        0.02       0.03        0.04       0.05
                         UV absorbance at 254 nm, UV-254 (1/cm)
                                                                   0.06
                                                                              0.07
Figure 41  Correlation between single contactor and blended effluent UV
absorbance and HAA9 bromine incorporation factor (n1) for Water 1
   2.5
o>

X
1_
£
&_
o
"G
o
Q.
O
O
o>
   2.0 -
   1.5 -
   1.0 -
I  °-5 ^
o
m
   0.0
         Water 2
         UV-254:  0.055 1/cm
         Bromide: 105 ug/L
                                                       D  Single contactor effluent
                                                       •  Blended effluent
                                                       	Best fit (RA2 = 0.435)
     0.00
                       0.01               0.02               0.03
                         UV absorbance at 254 nm, UV-254 (1/cm)
                                                                              0.04
Figure 42  Correlation between single contactor and blended effluent UV
absorbance and HAAS bromine incorporation factor (n') for Water 2
                                  -63-

-------
   3.0
   2.5 -
   2.0
o
•s
   1.5 -
o
Q.
8  1-0
O>
I  0.5
00
   0.0
         Water 3
      UV-254:  0.048 1/cm
      Bromide: 160 pg/L
     0.00
                                                      D  Single contactor effluent
                                                      •  Blended effluent
                                                     	Best fit (RA2 = 0.906)
                          0.01                     0.02
                      UV absorbance at 254 nm, UV-254 (1/cm)
                                                                             0.03
Figure 43 Correlation between single contactor and blended effluent UV
absorbance and HAA9 bromine incorporation factor (n1) for Water 3
   1.25
o>
I
1_
£
t_
o
•s
o
Q.
O
O
o>
   1.00 -
   0.75 -
0.50 -
~  0.25 H
o
m
   0.00
         Water 4
                c0l
       UV-254:  0.065 1/cm
       Bromide: 29 ug/L
                                                     D  Single contactor effluent
                                                     •  Blended effluent
                                                     	Best fit (RA2 = 0.484)
      0.00
                     0.01               0.02               0.03
                       UV absorbance at 254 nm, UV-254 (1/cm)
                                                                             0.04
Figure 44  Correlation between single contactor and blended effluent UV
absorbance and HAAS bromine incorporation factor (n') for Water 4
                                  -64-

-------
   1.5
o>
<
<
t  1.0-
o
•s
o
'-4—'
CD
t_
O
Q.

O
O
g

CD
g

E

2
m
0.5 -
   0.0
       Water 5



           D
      UV-254: 0.051 1/cm


      Bromide: 70 |jg/L
  n Single contactor effluent

  • Blended effluent

 	Best fit (RA2 = 0.269)
     0.00
                    0.01                0.02               0.03

                       UV absorbance at 254 nm, UV-254 (1/cm)
                       0.04
 Figure 45 Correlation between single contactor and blended effluent UV

 absorbance and HAA9 bromine incorporation factor (n1) for Water 5
   2.5
 en
   2.0 -
    1.5 -
 o
 Q.

 O
 o
 CD
    1.0 -
 .E 0.5 H

 o
 m
   0.0
          Water 6
                                                -e-
      UV-254: 0.06 1/cm

      Bromide: 140 ug/L
 n  Single contactor effluent

 •  Blended effluent

	Best fit (RA2 = 0.935)
      0.00
                     0.01               0.02               0.03

                      UV absorbance at 254 nm, UV-254 (1/cm)
                       0.04
 Figure 46 Correlation between single contactor and blended effluent UV

 absorbance and HAA9 bromine incorporation factor (n1) for Water 6
                                  -65-

-------
   2.5
o>
   2.0 -I
 _
 •§ 1.5
 |1.0H
 8
 O>
 i 0.5 -
GO
   0.0
         Water 7
      UV-254:  0.109 1/cm
      Bromide: 300 pg/L
                                                       n Single contactor effluent
                                                       • Blended effluent
                                                      	Best fit (RA2 = 0.410)
     0.00       0.01        0.02       0.03       0.04       0.05
                         UV absorbance at 254 nm, UV-254 (1/cm)
                                                               0.06
                                                                            0.07
Figure 47 Correlation between single contactor and blended effluent UV
absorbance and HAA9 bromine incorporation factor (n1) for Water 7
   0.5
 c
 o>~
 I
 i_
 .o
0.4 -
 o
 "G 0.3 H
 o 0.2-I
 t_
 o
 o
o>
I  °-1 H
2
CO
   0.0
          UV-254:  0.033 1/cm
          Bromide: 28 ug/L
     0.00
                                                   n  Single contactor effluent
                                                   •  Blended effluent
                                                  	Best fit (RA2 = 0.918)
                          0.01                     0.02
                      UV absorbance at 254 nm, UV-254 (1/cm)
                                                                            0.03
Figure 48  Correlation between single contactor and blended effluent UV
absorbance and HAA9 bromine incorporation factor (n1) for Water 8
                                 -66-

-------
4.3  Assessment of Logistic Function Fit to Single Contactor Breakthrough
     Curve Data

A model used to describe single contactor  effluent experimental data is needed for several
reasons.   From  a data management perspective,  best-fit curve  parameters  that adequately
describe experimental data are less memory intensive than storing  the entire experimental data
set.  A best-fit curve also facilitates interpolation and extrapolation to estimate run  times  for
given treatment objectives.  Use of a best-fit model curve provides an estimate of the scatter in
the data through the  coefficient of determination, and the model minimizes the impact of this
scatter on run time estimates. Finally, a function that describes the single contactor experimental
data set is a prerequisite for determining the integral breakthrough curve,  a curve that relates
single contactor run time to blended contactor water quality under the assumption that contactors
are operated in  parallel-staggered  mode.   Run time  estimates generated  by the integral
breakthrough curve are more applicable to full-scale GAC operation.

The appropriate logistic curve  function model (step, step-lag, or step-lag-peak, as described in
Section 3.2.1) was fit to the single contactor effluent data for eight GAC runs comprised of up to
20 parameters each.  The  GAC runs were performed using conventionally-treated water from
eight waters sources, including surface and ground waters. The parameters examined included
precursors or surrogates (TOC, UV254, and SDS-TOX), DBF class sums (TTHM, HAAS, HAA6,
and HAA9), and DBF  species (CF, BDCM, DBCM, BF, MCAA, DCAA,  TCAA, MBAA,
DBAA, BCAA, DCBAA, CDBAA,  and TBAA).  For some DBF species, samples taken during
the initial portion of the breakthrough curve were measured below the MRL. Curve fits were  not
performed if fewer than six effluent data points were reported above the MRL.

Due to the large amount of graphs generated, all plots are summarized in Appendix  E, and a
selection of plots are included together with this analysis as examples. The plots contain  the
single contactor GAC effluent parameter  concentration plotted against scaled operation time. A
line representing the  logistic function model best-fit is included. In addition, the experimental
blended effluent data points are included in each plot, along with a dotted  line representing  the
DI method prediction of the integral breakthrough curve, to demonstrate the benefit obtained by
blending the effluents of multiple contactors operated in parallel  staggered mode. An analysis of
the integral breakthrough curve predictive models is deferred until Section 4.4.

4.3.1  Surrogates and Class Sum  Logistic Function Curve Fits

The step logistic function  model was used to fit single  contactor effluent TOC breakthrough
curves for all waters,  as shown in Figures 49 through 56.  GAC run  times ranged from  76 to 287
days, and the GAC influent TOC concentration ranged from 2.0 to  5.6 mg/L.  For these waters,
the step logistic function model provided excellent curve fit approximations:  the R2 values  for
the curve fits ranged from 0.966 to 0.992.

Using the step-lag logistic function model, excellent curve fits were also obtained for single
contactor effluent UV254 breakthrough curves: the R2 values for the  curve fits ranged from 0.982
to 0.998.  The measured GAC influent UV254 ranged from 0.033 to 0.109  I/cm for the waters
                                      -67-

-------
examined.  The results for Waters 5 and 7 are shown in Figures 57 and 58. The results for the
remaining waters can be found in Appendix E.

SDS-TOX breakthrough curves were also modeled  using the step-lag logistic function with
excellent results: the R2 values for the curve fits ranged from 0.990 to 0.999.  The GAC influent
SDS-TOX ranged from 156 to 486 |ig/L as Cl".  Figures 59 and 60 show examples of the  data
obtained for Waters 4 and 8.

Both SDS-TTHM and the SDS-HAA sums (HAAS, HAA6, and HAA9) were modeled using the
step-lag logistic function model. Again, single contactor effluent data were well-represented by
the model used, as shown for SDS-TTHM in Figures 61  and 62 for Waters 3 and 6. The R2
values for SDS-TTHM curve fits ranged  from 0.977 to  0.995.  Overall, the  step-lag logistic
function model was  also successful when used to fit all three SDS-HAA species sums, with R2
values ranging from  0.952 to 0.994. Figures 63 and 64 show the step-lag logistic function model
curve fits applied to single contactor effluent SDS-HAA9 data for Waters 1 and 2.

Table 8 summarizes the R2  values measured for all curve fits, including DBF  surrogates, DBF
sum class parameters, and DBF species.  For all waters  and  all parameters the mean R2  was
0.973 ± 0.046, indicating that all breakthrough curve data were successfully fit using the logistic
function models.  For DBF  surrogates and DBF  sum  class parameters only, the mean R2 value
was 0.982 ±0.012.

4.3.2  DBP Species Logistic Function Curve Fits

THM Species.  For all waters, the application of the  step-lag logistic function  resulted in good
curve fits for SDS-CF, as shown by R2 values ranging between 0.922 and 0.998.  Typically, the
SDS-CF breakthrough curve shape was similar to that of SDS-TTHM, as shown in the example
given for Water 4 in Figure 65. For most SDS-BDCM and SDS-DBCM breakthrough curves,
the step-lag logistic  function model was used and yielded good results: R2 values ranged from
0.946 to 0.998. An  example of the SDS-BDCM breakthrough curve is shown  in Figure  66 for
Water 6 and an example of the SDS-DBCM breakthrough curve is shown in Figure 67 for Water
7.  The step-lag logistic function  model was able to adequately fit the sharp 'S' shape  (steep
breakthrough followed by flat plateau) observed for the SDS-BDCM experimental data in  Figure
66 for Water 6 (R2 = 0.979). Water 2 also exhibited this behavior, as did Waters 1 and 3 to a less
pronounced extent, but in all cases the data were successfully fit by the model. For Waters 4 and
8, a peak breakthrough curve for SDS-BDCM was detected by the curve fit algorithm.  These
curves were fit using the step-lag-peak logistic function model, as shown in Figures 68 and 69.
The curve fit procedure  successfully fit the single contactor peak curves; the R2 values were
greater than 0.94 for both waters.  For 3 out of 6 waters with SDS-BF levels measured above the
MRL, a peak curve was detected.  The step-lag-peak logistic function model also successfully fit
these peak curves, as shown in Figures 70 and 71  for Waters 1 and 6. The R2 values for SDS-BF
curve fits ranged from 0.910 to 0.984.  Summary plots of the model fits for all THM species and
all waters are included in Appendix E.

HAA Species.  Of the non-brominated HAA species, no curve fits were applied to SDS-MCAA
because only for a few samples was MCAA measured above the MRL (2.0  |ig/L) in the GAC
effluent.  The experimental breakthrough curves for  SDS-DCAA and SDS-TCAA were well
                                      -68-

-------
Analyte
TOC
UV-254
SDS-TOX
SDS-TTHM
SDS-HAA5
SDS-HAA6
SDS-HAA9
SDS-CF
SDS-BDCM
SDS-DBCM
SDS-BF
SDS-MCAA
SDS-DCAA
SDS-TCAA
SDS-MBAA
SDS-DBAA
SDS-BCAA
SDS-CDBAA
SDS-DCBAA
SDS-TBAA
Total
Number
Step
8
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
9
of each type
Step-lag
0
8
8
8
7
8
7
7
6
8
2
0
8
5
0
6
8
2
7
0
105
of curve fit used for all
Step-lag-peak
0
0
0
0
1
0
1
0
2
0
4
0
0
0
0
0
0
1
0
1
10
waters
No fit*
0
0
0
0
0
0
0
0
0
0
2
8
0
3
8
2
0
5
1
7
36
Curve fit
Mean
0.980
0.993
0.995
0.985
0.975
0.974
0.976
0.991
0.980
0.992
0.964
NA
0.975
0.973
NA
0.957
0.970
0.946
0.907
0.976

R2 value
SD
0.009
0.005
0.003
0.007
0.012
0.010
0.013
0.004
0.016
0.005
0.027
NA
0.023
0.027
NA
0.050
0.012
0.032
0.167
NA

*Curve fits were not performed on breakthrough curves with fewer than 6 points measured above the MRL
SD: Standard deviation
NA: Not applicable

Table 8  Frequency of logistic function model used and R2 values for all parameters and all waters

-------
modeled by the step-lag logistic function.  Examples of the curve fits are shown in Figure 72
(SDS-DCAA, Water 3) and Figure 73 (SDS-TCAA, Water 5). For all waters, R2 values ranged
from 0.922 to  0.999.  SDS-MBAA was not measured above the MRL (1.0 jig/L) in the GAC
effluent during any of the studies. Due to low or BMRL GAC  effluent SDS-DBAA levels for
Waters 4 and 8 curve fitting procedures were not applied.  For the remaining waters, the SDS-
DBAA breakthrough curve typically showed a sharp 'S' shape (steep breakthrough followed by
flat plateau) and was successfully fit by the step-lag logistic function model (R2 values ranging
from 0.859 to 0.992).  An example of SDS-DBAA breakthrough and model fit is shown in
Figure 74 for Water 6.  The breakthrough curves for SDS-BCAA were also well-fit by the step-
lag logistic function model (R2 values ranging from 0.952 to 0.987),  such as that for Water 1,
shown in Figure 75. Examples of curve fits for SDS-DCBAA, SDS-CDBAA, and SDS-TBAA
are shown in Figures 76, 77, and 78, respectively.  For these three compounds, R2 values ranged
from 0.913 to 0.987 except for the curve fit for SDS-DCBAA for Water 3, which yielded an R2
value of 0.531. For this run, GAC effluent SDS-DCBAA  concentrations were low, measured
between 1.0 and 2.0 |ig/L.

A peak curve was detected in 10 cases out of 126 total curve fits.  Every water examined yielded
at least one peak curve.  In general, the  step-lag-peak  logistic function model was able  to
adequately fit the peak curve  data. Additional examples of peak curve fits are shown in Figures
79 and  80. Figure 81 (SDS-BF for Water 3) shows an example of single contactor data that
qualitatively shows a peak, but the peak curve algorithm did not detect it as such.  The logistic
function under predicted experimental data at the peak by up to 4 |ig/L, and over predicted it at
the end  of the run by up to 3 |ig/L. The R2 value for this curve fit was 0.856. When refit to the
step-lag-peak logistic function model, Figure 82, the R2 value improved to 0.984.

In general, the logistic function models successfully characterized the variety of breakthrough
curves observed in this study.  Three logistic function models were utilized: step, step-lag, and
step-lag-peak.  These models were able to match not only typical S-shaped breakthrough curves
but also peak curves observed for some parameters. Table  8 also summarizes the frequency of
use for each of the three types of logistic function models used.  The step-lag logistic function
was the most commonly used, utilized in 66 percent of the  curve fits performed.  For the DBF
surrogates and class sums, the step-lag logistic function model  was utilized for 82 percent of
curve fits.  The step function was almost exclusively utilized for  TOC breakthrough curves. For
23 percent of all data sets, no curve fit was performed because fewer than six data points were
measured above the MRL. Most of the parameters for which no curve fit was performed were
HAA species.   For all data  sets that  were modeled, the mean R2 was  0.974, indicating that
overall,  the models were able to successfully fit the GAC breakthrough profiles.
                                      -70-

-------
   3 -
 o
'•a 2
 CD
 O
 O
O
   1 -
         TOO
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.966)
 O   Blended effluent
 	Dl prediction (RSS = 0.152)
                                                                   O
                                                                     EBCT = 20 min.
                                                                     c0 = 4.54 mg/L
                        20                  40
                                Scaled operation time (days)
                                                    60
                 80
Figure 49 Single contactor and blended effluent TOC breakthrough curves
for Water 1
   2.5
   2.0 -
 o
'-4—'
 CD
 o
O
   1.5 -
   1.0 H
   0.5 -
   0.0
         TOC
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.972)
 O   Blended effluent
 	Dl prediction (RSS = 0.057)
                      50
                                                                    EBCT = 20 min.
                                                                    c0 =  2.6 mg/L
                         100            150
                     Scaled operation time  (days)
200
250
Figure 50 Single contactor and blended effluent TOC breakthrough curves
for Water 2
                                    -71-

-------
   2.0
   1.5 -
 o
'•a
 O>
 o
 c
 o
O
   0.5 -
   0.0
         TOO
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.986)
 O   Blended effluent
 	Dl prediction (RSS = 0.044)
                                                                 D    D
                                                                     EBCT = 20 min.
                                                                     c0 = 2.35 mg/L
                   50
                    100          150          200
                     Scaled operation time (days)
250
300
Figure 51 Single contactor and blended effluent TOC breakthrough curves
for Water 3
   2.5
   2.0 -
 O)
   1.5 H
   1.0 H
 o
O
   0.5 -
   0.0
         TOC
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.984)
 O   Blended effluent
 	Dl prediction (RSS = 0.05)
                                                                     EBCT = 20 min.
                                                                     c0 = 2.98 mg/L
                          50                  100                 150
                                 Scaled operation time (days)
                                                                       200
Figure 52 Single contactor and blended effluent TOC breakthrough curves
for Water 4
                                    -72-

-------
   3.0
   2.5 -
   2.0 -
 o
'•a
 O>
 o
 o 1.0
O
   0.5 -
   0.0
         TOO
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.984)
 O   Blended effluent
 	Dl prediction (RSS = 0.035)
                                                            .---o"
                                                                  .---O
                                                                    EBCT = 20 min.
                                                                    c0 =  3.08 mg/L
                 50         100        150        200        250
                                 Scaled operation time (days)
                                                            300
             350
Figure 53 Single contactor and blended effluent TOC breakthrough curves
for Water 5
   2.5
   2.0 -
 O)
   1.5 H
   1.0 H
 o
O
   0.5 -
   0.0
         TOC
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.992)
 O   Blended effluent
 	Dl prediction (RSS = 0.015)
                   50
                    100          150          200
                     Scaled operation time (days)
                                                                    EBCT = 20 min.
                                                                    c0 =  2.64 mg/L
250
300
Figure 54 Single contactor and blended effluent TOC breakthrough
curves for Water 6
                                    -73-

-------
   4 -
   3 -
.g
"co
 o> 9
 o *-
 c
 o
O
   1 -
         TOO
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.979)
 O   Blended effluent
 	Dl prediction (RSS = 0.065)
                                                                    EBCT = 20 min.
                                                                    c0 =  5.58 mg/L
                  25
                   50           75           100
                     Scaled operation time (days)
125
150
Figure 55 Single contactor and blended effluent TOC breakthrough
curves for Water 7
   2.0
   1.5 -

'•a 1-0 H
 o>
 o
 c
 o
O
   0.5 -
   0.0
         TOC
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.974)
 O   Blended effluent
 	Dl prediction (RSS = 0.033)
                                                                    EBCT = 7.2 min.
                                                                    C0 =  2.02 mg/L
                         50                 100
                                 Scaled operation time (days)
                                                    150
             200
Figure 56 Single contactor and blended effluent TOC breakthrough
curves for Water 8
                                    -74-

-------
0.040

0.035 -

0.030 -

0.025 -
CD
            UV254
               D   Single contactor effluent
              	Logistic function best fit (RA2 = 0.992)
               O   Blended effluent
               	Dl prediction (RSS = 0.002)
                                                                    EBCT = 20 min.
                                                                    c0 = 0.051 1/cm
                   50
                          100        150        200        250
                               Scaled operation time (days)
  300
350
Figure 57 Single contactor and blended effluent UV254 breakthrough
curves for Water 5
   0.07
   0.06 -
   0.05 -
.o
^ 0.04 -
o>
o
c
CD
•9 0.03 -
o
   0.02 H
   0.01 -
   0.00
            UV254
            D   Single contactor effluent
           	Logistic function best fit (RA2 = 0.994)
            O   Blended effluent
            	Dl prediction (RSS = 0.0015)
                        O. -
                    25
                              50           75           100
                               Scaled operation time (days)
                                                                 EBCT = 20 min.
                                                                 c0 = 0.109 1/cm
125
150
Figure 58 Single contactor and blended effluent UV254 breakthrough
curves for Water 7
                                    -75-

-------
   175
   150 -
-T 125 -
O
   100 -
o
'-4—'
TO
    75 -
SDS-TOX

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.999)
   O   Blended effluent
 	Dl prediction (RSS = NA)
                                                                   EBCT = 20 min.
                                                                   C0 = 288 ug/L Cl-
                         50                100
                                Scaled operation time (days)
                                                   150
200
 Figure 59 Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 4
   125
O
o
'-4— '
TO
0>
O
c
o
O
   100 -
    75 -
    50 -
    25 -
           SDS-TOX
   D   Single contactor effluent
  	Logistic function best fit (RA2 = 0.99)
   O   Blended effluent
   	Dl prediction (RSS = NA)
                                                                   EBCT = 7.2 min.
                                                                   c0= 156 ug/LCI-
                         50                100
                                Scaled operation time (days)
                                                   150
200
 Figure 60 Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 8
                                   -76-

-------
   125
   100 -
    75 -
s
-I—'
§   50
c
o
O
    25 -
SDS-TTHM

  D   Single contactor effluent
	Logistic function best fit (RSS 0.23A2 = 0.986)
            O  Blended effluent
            	Dl prediction (RSS = 2.87)
                                                                  EBCT = 20 min.

                                                                  c0 = 154 ug/L
                   50
                      100          150          200
                       Scaled operation time (days)
250
300
Figure 61  Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 3
   100
    75 -
o
'«   50 H
o>
o
c
o
o
    25 -
     0 ^
         SDS-TTHM
    D   Single contactor effluent
    	Logistic function best fit (RA2 = 0.98)
    O   Blended effluent
    	Dl prediction (RSS = 6.43)
                                                                  EBCT = 20 min.

                                                                  c0= 128 ug/L
                   50
                      100          150          200
                       Scaled operation time (days)
250
300
Figure 62  Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 6
                                    -77-

-------
    30
    25 -
    20 -
 o
 |5  15 H
 -I—•
 §    "

 I  10^
     5 -
SDS-HAA9
   D   Single contactor effluent
 	Logistic function best fit  (RA2 = 0.982)
   O   Blended effluent
 	Dl prediction (RSS = 0.78)
     0 40
      0
                20                 40
                        Scaled operation time (days)
                                                                  EBCT = 20 min.
                                                                  c0 = 29 ug/L
60
 80
Figure 63  Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 1
    30
    25 -
    20 -
SDS-HAA9
 EBCT = 20 min
 c0 = 37 M9/L
                                               D  Single contactor effluent
                                               	Logistic function best fit (RA2 = 0.988)
                                               O  Blended effluent
                                               	Dl prediction (RSS = 3.64)
                      50
                           100             150
                        Scaled operation time (days)
   200
250
Figure 64 Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 2
                                    -78-

-------
   30


   25 -


§20-
3.
c
o
TS 15 H
         SDS-CF
              D   Single contactor effluent
             	Logistic function best fit  (RA2 = 0.998)
              O   Blended effluent
              	Dl prediction (RSS = 0.41)
                                                                  EBCT = 20 min.
                                                                  c0 = 50.7 ug/L
                         50                 100
                                 Scaled operation time (days)
                                                              150
             200
 Figure 65  Single contactor and blended effluent SDS-CF breakthrough
 curves for Water 4
   35
   30 -
   25 -
   20 -
o
'-4—'
cc
o>
o
o
O
   15 -
   10 -
    5 -
    0 -I
        SDS-BDCM
        EBCT = 20 min
        C0 = 27.4 ug/L
                  50
                                                                            _n
                                                                            O
                                                   Single contactor effluent
                                                   Logistic function best fit (RA2 = 0.979)
                                               O   Blended effluent
                                               	Dl prediction (RSS = 3.63)	
                              100          150          200
                                Scaled operation time (days)
250
300
 Figure 66  Single contactor and blended effluent SDS-BDCM breakthrough
 curves for Water 6
                                    -79-

-------
    50
    40 -
    30 -
 o
 '-4—'
 cc
    20-1
 o
 O
    10 -
     0 -I
SDS-DBCM

   D   Single contactor effluent

 	Logistic function best fit (RA2 = 0.996)
              O   Blended effluent

              	Dl prediction (RSS = 0.99)
                   25
                                     '  O
                                                         o-'
                     50           75           100

                      Scaled operation time (days)
                                                       '  o



                                                       EBCT = 20 min.

                                                       c0 = 66.2 ug/L
125
150
Figure 67  Single contactor and blended effluent SDS-DBCM breakthrough
curves for Water 7
   3 -
 o
T5 2 H
 o>
 o
 c
 o
O
       SDS-BDCM

        EBCT = 20 min

        c0 = 2.2 ug/L
                                               D   Single contactor effluent

                                               	Logistic function best fit (RA2 = 0.984)

                                               O   Blended effluent

                                               	Dl prediction (RSS = 0.58)
                                           100

                                Scaled operation time (days)
                                                    150
            200
Figure 68 Single contactor and blended effluent SDS-BDCM breakthrough
curves for Water 4
                                    -80-

-------
   4 -
   3 -
 g
'-4—'
 CD
 CD
 O
 c
 O
O
2 -
   1 -
   0 -O
     SDS-BDCM




    EBCT = 7.2min.

    C0 = 2.6 |jg/L
                                 O
                                   D
                                   O
                                                                            O
                                                 D


                                                O
                                         D  Single contactor effluent

                                        	Logistic function best fit (RA2 = 0.946)

                                         O  Blended effluent

                                         	Dl prediction (RSS = 1.2)
                        50                 100

                                Scaled operation time (days)
                                                           150
200
Figure 69 Single contactor and blended effluent SDS-BDCM breakthrough
curves for Water 8
                                                D   Single contactor effluent

                                                	Logistic function best fit (RA2 = 0.971)

                                                O   Blended effluent

                                                	Dl prediction (RSS = 2.76)
                                                                   EBCT = 20 min.

                                                                   C0 = 3.7
                         20                 40

                                Scaled operation time (days)
                                                            60
 80
Figure 70 Single contactor and blended effluent SDS-BF breakthrough
curves for Water 1
                                   -81-

-------
   20
    15 -
 g
 '•£  10 H
 O>
 o
 c
 o
 O
     5 -
     0 -D
      0
         SDS-BF
         EBCT = 20 min.
         c0 = 3.3 |jg/L
         50
                                                D  Single contactor effluent
                                                	Logistic function best fit (RA2 = 0.978)
                                                O  Blended effluent
                                                	Dl prediction (RSS = 3.47)
100          150          200
  Scaled operation time (days)
250
300
 Figure 71  Single contactor and blended effluent SDS-BF breakthrough
 curves for Water 6
    6 -
 g
 '•SS  4H
 o>
 o
 c
 o
 O
    2 -
SDS-DCAA

  D   Single contactor effluent
	Logistic function best fit (RA2 = 0.995)
  O   Blended effluent
	Dl prediction (RSS = 0.4)
                                      O—O—r-
                  50
                     100          150          200
                       Scaled operation time (days)
                                                                   EBCT = 20 min.
                                                                  c0=  15.7ug/L
                                      250
             300
Figure 72  Single contactor and blended effluent SDS-DCAA breakthrough
curves for Water 3
                                    -82-

-------
   6 -
SDS-TCAA

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.989)
   O   Blended effluent
 	Dl prediction (RSS = 0.28)
                                                                   EBCT = 20 min.
                                                                   C0 =  12.7 ug/L
                50
                100        150        200        250
                      Scaled operation time (days)
                                        300
             350
 Figure 73 Single contactor and blended effluent SDS-TCAA breakthrough
 curves for Water 5
   10
    6 -
o
O
    2 -
        SDS-DBAA
        EBCT = 20 min.
        c0 = 5.7 ug/L
    0 -t>
      0
                                                           O
                                                D   Single contactor effluent
                                                	Logistic function best fit (RA2 = 0.952)
                                                O   Blended effluent
                                                	Dl prediction (RSS = 1.31)
        50
100          150          200
  Scaled operation time (days)
250
300
 Figure 74 Single contactor and blended effluent SDS-DBAA breakthrough
 curves for Water 6
                                   -83-

-------
   6 -
 O>
 o
 c
 o
O
SDS-BCAA

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.981)
   O   Blended effluent
 	Dl prediction (RSS = 0.38)
                        20
                                   40
                        Scaled operation time (days)
                                                                  EBCT = 20 min.
                                                                  c0= 7 |jg/L
60
 80
Figure 75  Single contactor and blended effluent SDS-BCAA breakthrough
curves for Water 1
   3.0
   2.5 -
 O)
 o
73
 o>
 o
   2.0 -
 o 1.0 H
   0.5 -
   0.0 -I
 SDS-DCBAA
 EBCT = 20 min.
 c0 =  3 |jg/L
                     50
                                                                        D

                                                                        -O
                                                 O
                                     D   Single contactor effluent
                                    	Logistic function best fit (RA2 = 0.952)
                                     O   Blended effluent
                                     	Dl prediction (RSS = 0.48)
                            100            150
                        Scaled operation time (days)
   200
250
Figure 76 Single contactor and blended effluent SDS-DCBAA breakthrough
curves for Water 2
                                   -84-

-------
   3.5
   3.0 -
   2.5 -
 c 2.0 H
 o
 g 1-5 H
 o
 c
 o
0 1.0 H
   0.5 -
   0.0 -I
          SDS-CDBAA
 EBCT = 20 min.
c0 = 3 |jg/L
                   50
                                            D  D
                                   . •"   D   Single contactor effluent
                                      	Logistic function best fit  (RA2 = 0.913)
                                        O   Blended effluent
                                      	Dl prediction (RSS = 0.72)
                      100          150          200
                       Scaled operation time (days)
250
Figure 77  Single contactor and blended effluent SDS-CDBAA breakthrough
curves for Water 6
300
    6 -
          SDS-TBAA
           EBCT = 20 min
           C0 = BMRL
                  25
                                                 D   Single contactor effluent
                                                 	Logistic function best fit (RA2 = 0.976)
                                                 O   Blended effluent
                                                 	Dl prediction (RSS = 1.92)
                     50           75           100
                       Scaled operation time (days)
125
150
 Figure 78 Single contactor and blended effluent SDS-TBAA breakthrough
 curves for Water 7
                                    -85-

-------
    14
    12 -
    10 -
 g
 '-4—'
 CD
 CD
 O
 c
 O
 O
     6 -
     4 -
     2 -
          SDS-BF
                                        O
                                                         EBCT = 20 min.
                                                         C0 = 3.7 |jg/L
                                   D  Single contactor effluent
                                   	Logistic function best fit (RA2 = 0.91)
                                   O  Blended effluent
                                   	Dl prediction (RSS = 3.73)
                      50
                           100            150
                       Scaled operation time (days)
200
250
 Figure 79  Single contactor and blended effluent SDS-BF breakthrough
 curves for Water 2
    5 -
    4 -
 o
 •^ ^
 CD °
 t_
 CD
 O

 I2
    1 -
SDS-CDBAA
 EBCT = 20 min.
 c0 =  3.7 |jg/L
                50
                                         D   Single contactor effluent
                                         	Logistic function best fit (RA2 = 0.95)
                                         O   Blended effluent
                                         	Dl prediction (RSS = 0.88)
                 100        150        200         250
                       Scaled operation time (days)
    300
350
Figure 80 Single contactor and blended effluent SDS-CDBAA breakthrough
curves for Water 5
                                    -86-

-------
   40
   30 -
g
'•£  20 H
O>
o
c
o
O
   10 -
SDS-BF

EBCT = 20 min.
C0 = 11.8 |jg/L     n

              D
                                         o
                                      D
                                     n  n
                                                  o
                                           D   Single contactor effluent
                                           	Logistic function best fit (RA2 = 0.856)
                                           O   Blended effluent
                                           	Dl prediction (RSS = 6.3)
                  50
                     100          150          200
                       Scaled operation time (days)
250
300
Figure 81  Single contactor and blended effluent SDS-BF breakthrough
curves for Water 3 (original step-lag logistic function model curve fit)
   40
   30 -
o
'•£  20 H
o>
o
c
o
O
   10 -
SDS-BF

EBCT = 20 min.
c0=  11.8 ug/L
                                           D   Single contactor effluent
                                           	Logistic function best fit (RA2 = 0.984)
                                           O   Blended effluent
                                           	Dl prediction (RSS = 5.67)
                  50
                     100          150          200
                       Scaled operation time (days)
250
300
Figure 82 Single contactor and blended effluent SDS-BF breakthrough
curves for Water 3 (fit to step-lag-peak logistic function model)
                                   -87-

-------
This page intentionally left blank.
          -88-

-------
4.4  Comparison of SCA and Dl Methods Used to Predict the Blended Contactor
     Integral Breakthrough Curve

Two methods were used to predict the integral breakthrough curve, which is used to estimate
blended contactor run times: direct integration (DI) and surrogate correlation approach (SCA).
The DI procedure was explained in Sections 1.4 and 3.2.3, while the steps followed by the SCA
procedure were described in Sections 1.5 and 3.2.4.  The DI method has been presented in the
literature and is the traditional method for predicting the integral breakthrough curve, the curve
that relates blended contactor effluent water quality to single contactor run time.  However,
previous verification  of  the  DI  method  was limited to  DBF  surrogates  and  class sums.
Verification of the DI method here is expanded to include more GAC runs  and water sources,
and a larger experimental matrix, including application to  DBF species.   The SCA method
developed in this study is verified against experimental data and compared to results obtained for
the DI procedure.  The SCA method is  computationally simpler than the  DI procedure,  an
important consideration when selecting a predictive method for application to the ICR GAC
treatment study data set, which may require 8,000 to 9,000 curve fits.

The results of these predictive models were compared to the experimentally-obtained blended
effluent data set.  A best-fit curve was used to  describe the observed  data.   The best-fit was
derived using the  same  logistic function models used for the single contactor data, as the shapes
observed for many of the blended effluent curves were similar to those encountered during single
contactor breakthrough curve analysis. The models were evaluated by comparing the predictions
to the experimental data  (not the best-fit curve) and calculating the residual sum of squares
(RSS) and model bias.  The model bias  is defined as the  mean  of the residuals, calculated
between the model prediction and experimental results. A summary of the calculated model RSS
values is shown in Tables 9 and 10, while a summary of the calculated  bias  values is shown in
Tables 11 and  12. Note that the RSS and bias values have units equivalent to the units of the
parameter from which they are calculated; therefore, the magnitudes of these values for different
parameters may vary widely and  are not directly comparable. A summary of average RSS and
bias values is given in Table 13.   Based on direct comparisons between the RSS values for each
predictive approach, and across all parameters (110 comparisons) the two methods were equally
successful in predicting the observed data:  for 52 percent of the predictions, the SCA method
RSS value was lower than that for the DI method.

To examine the performance of each model for predicting the integral breakthrough  curve across
all water sources  and water quality parameters, cumulative frequency  distribution plots were
developed for the RSS  and bias data.  To provide a consistent basis of comparison of RSS and
bias data across different parameters,  the  data was normalized.   This was  accomplished  by
dividing  the RSS and  bias values measured  for each parameter and water by  the average
concentration of that parameter in the blended effluent during each run.  This procedure  did
result in some extremely high normalized  values, and overall, the average normalized values
were relatively high.  However, this is due  to the relatively low average effluent concentrations
of many  parameters, by which the RSS and bias values were normalized. The normalized RSS
and bias values are tools that allow comparison of the two predictive models across all available
data in this study,  and are not intended to provide an indication of model performance outside of
                                      -89-

-------
VO
o
§ra meter
TOO (ngL)
UV-254 flfcm)
TOX &L Cl ')
TTHM gL)
HAAS gL)
HAA6 gL)
HAA9 gL)
CFgL)
BDCM gL)
DBCM gL)
BFgL)
MCAA gL)
DCAA gL)
TCAA gL)
MBAA gL)
DBAA gL)
BCAA gL)
DCBAAgL)
CDBAAgL)
TBAA gL)

Dl
0.152
0.0011
NA
3.93
0.38
0.63
0.78
0.39
1.52
0.40
2.76
NA
0.35
NA
NA
0.46
0.38
NA
NA
NA
MfeM
SCA
NA
0.0017
NA
2.55
1.58
1.26
1.04
0.65
1.03
0.73
1.31
NA
1.46
NA
NA
0.39
0.29
NA
NA
NA
Residual sum
Hter2
Lower
NA
Dl
NA
SCA
Dl
Dl
Dl
Dl
SCA
Dl
SCA
NA
Dl
NA
NA
SCA
SCA
NA
NA
NA
Dl
0.057
0.0012
NA
8.49
1.69
2.50
3.64
0.57
4.35
2.18
3.73
NA
0.49
0.27
NA
1.10
0.82
0.48
NA
NA
Dl 76CA: 5
SCA
NA
0.0015
NA
4.38
1.48
2.25
3.43
1.59
1.98
1.72
2.18
NA
0.68
0.34
NA
0.60
0.78
0.35
NA
NA
Dl
Lower
NA
Dl
NA
SCA
SCA
SCA
SCA
Dl
SCA
SCA
SCA
NA
Dl
Dl
NA
SCA
SCA
SCA
NA
NA
4SCA: 10
of sqares RSS)
HterS
Dl
0.044
0.0010
NA
2.87
2.48
2.29
4.36
0.85
1.05
1.04
5.67
NA
0.40
NA
NA
1.02
0.63
0.68
NA
NA

SCA
NA
0.0010
NA
2.64
1.13
2.30
2.20
0.40
2.57
0.68
3.14
NA
0.37
NA
NA
1.56
0.63
0.72
NA
NA
Dl
Lower
NA
Dl
NA
SCA
SCA
Dl
SCA
SCA
Dl
SCA
SCA
NA
SCA
NA
NA
Dl
Dl
Dl
NA
NA
6SCA: 7
Hter4
Dl
0.050
0.0015
NA
1.39
1.24
1.34
1.70
0.41
0.58
0.59
NA
NA
0.32
0.96
NA
NA
0.31
0.49
NA
NA

SCA
NA
0.0023
NA
2.03
0.87
1.07
1.12
1.33
0.55
0.55
NA
NA
0.47
0.64
NA
NA
0.22
0.50
NA
NA
Dl
Lower
NA
Dl
NA
Dl
SCA
SCA
SCA
Dl
SCA
SCA
NA
NA
Dl
SCA
NA
NA
SCA
Dl
NA
NA
5SCA: 7
                  NA: not applicable
                  Table SSummary of Dland SCA integral breakthrough curve prediction RSS values for Hfers through 4

-------
grameter
TOO (ngL)
UV-254(lfcm)
TOXgLCI ')
TTHM gL)
HAAS gL)
HAA6 gL)
HAA9 gL)
CFgL)
BDCM gL)
DBCM gL)
BFgL)
MCAA gL)
DCAA gL)
TCAA gL)
MBAAgL)
DBAA gL)
BCAA gL)
DCBAAgL)
CD BAA gL)
TBAA gL)


Dl
0.035
0.0020
NA
1.89
0.56
0.82
1.19
0.31
0.85
0.54
0.55
NA
0.49
0.28
NA
0.47
0.40
0.40
0.88
NA

HterS
SCA
NA

Lower
NA
0.0022 Dl
NA
1.30
0.52
0.55
1.13
0.65
0.34
0.72
0.26
NA
0.59
0.39
NA
0.20
0.21
0.34
1.10
NA

NA
SCA
SCA
SCA
SCA
Dl
SCA
Dl
SCA
NA
Dl
Dl
NA
SCA
SCA
SCA
Dl
NA
Dl 6SCA: 9

Dl
0.015
0.0017
2.39
6.43
1.02
1.30
1.88
1.37
3.63
1.02
3.47
NA
0.47
NA
NA
1.31
0.43
0.34
0.72
NA

Hter6
SCA
NA
0.0026
6.26
5.91
1.54
1.92
2.41
0.74
2.31
1.84
1.86
NA
0.31
NA
NA
1.11
0.44
0.34
0.67
NA
Dl
Residual sum
Lower
NA
Dl
Dl
SCA
Dl
Dl
Dl
SCA
SCA
Dl
SCA
NA
SCA
NA
NA
SCA
Dl
Dl
SCA
NA
8SCA: 7
Dl
0.065
0.0015
6.92
15.22
2.02
2.09
4.52
0.85
5.57
0.99
11.53
NA
0.59
0.40
NA
2.21
0.35
1.13
0.25
1.92

of sqares RSS)
tffer7
SCA
NA
0.0021
9.30
9.72
1.52
2.25
4.70
0.46
3.38
1.10
5.39
NA
0.63
1.34
NA
0.97
1.10
0.85
1.02
1.84
Dl
Lower
NA
Dl
Dl
SCA
SCA
Dl
Dl
SCA
SCA
Dl
SCA
NA
Dl
Dl
NA
SCA
Dl
SCA
Dl
SCA
9SCA: 8
Dl
0.033
0.0017
NA
2.59
1.35
1.83
2.20
0.58
1.20
1.15
NA
NA
0.68
0.73
NA
NA
0.59
0.46
NA
NA

HterS
SCA
NA
0.0017
NA
3.87
1.84
1.86
1.85
1.62
0.81
1.19
NA
NA
0.75
1.00
NA
NA
0.31
0.29
NA
NA
Dl
All waters lower RSS
Lower
NA
Dl
NA
Dl
Dl
Dl
SCA
Dl
SCA
Dl
NA
NA
Dl
Dl
NA
NA
SCA
SCA
NA
NA
8SCA: 4
Dl
0
8
2
2
3
5
3
5
1
5
0
0
6
4
0
1
3
3
2
0
53
SCA
0
0
0
6
5
3
5
3
7
3
6
0
2
1
0
5
5
4
1
1
57
NA: not applicable




Table OSummary of Dland SCA integral breakthrough curve prediction RSS values forttfers 3hrough 8

-------
grameter


TOC (ngL)
UV-254flcm)
TOX gL Cl ")
TTHM gL)
HAAS gL)
HAA6 gL)
HAA9 gL)
CFgL)
BDCM gL)
DBCM gL)
BFgL)
MCAA gL)
DCAA gL)
TCAA gL)
MBAAgL)
DBAA gL)
BCAA gL)
DCBAAgL)
CDBAAgL)
TBAA gL)
Model Bias
fetter 1
Dl
8.074
-0.0007
NA
-2.54
-0.15
-0.16

SCA
NA
-0.0012
NA
-1.65
8.29
8.15
-0.18 -0.03
8.18
-1.15
8.07
-2.17 -0
NA
-0.39
-0.74
-0.48
.33
NA
8.11 8.28
NA
NA
-0.21
8.04
NA
NA
NA
NA
NA
8.22
-0.06
NA
NA
NA
fetter
Dl
9.027
-0.0004
NA
-7.69
-1.45
-2.16
-3.07
-0.15
-3.83
-1.63
-2.91
NA
-0.38
-0.04
NA
-0.90
-0.72
-0.35
NA
NA
2
SCA
NA
-0.0005
NA
-4.09
-1.20
-1.85
-2.87
-0.92
-1.61
-1.15
-1.50
NA
-0.57
9.00
NA
-0.35
-0.66
-0.24
NA
NA
fetter 3
Dl
9.038
-0.0008
NA
-2.55
1-.41
1-.48
2.82
9.57
-0.35
9.59
-5.23
NA
9.29
NA
NA
-0.03
9.36
9.58
NA
NA

SCA
NA
-0.0007
NA
9.89
9.22
9.65
1-.23
-0.22
1-.53
-0.08
-1.17
NA
-0.22
NA
NA
9.52
9.17
9.62
NA
NA
fetter 4
Dl
9.037

SCA
NA
-0.0009 -0.0016
NA
-0.81
9.56
9.44
NA
-1.40
-0.57
-0.65
9.64 -0.36
-0.05
-0.47
-0.36
NA
NA
9.04
9.36
NA
NA
-0.13
9.34
NA
NA
-0.89
-0.09
-0.38
NA
NA
-0.33
-0.44
NA
NA
-0.09
9.38
NA
NA
NA: not applicable
Table 1 Summary of model
ifameter
TOC (ngL)
UV-254flcm)
TOX gL Cl ")
TTHM gL)
HAAS gL)
HAA6 gL)
HAA9 gL)
CFgL)
BDCM gL)
DBCM gL)
BFgL)
MCAA gL)
DCAA gL)
TCAA gL)
MBAAgL)
DBAA gL)
BCAA gL)
DCBAAgL)
CDBAAgL)
TBAA gL)
prediction
felterS
Dl
8.023
-0.0013
NA
-1.29
-0.22
-0.31
-0.86 -0
-0.04
-0.62
-0.25
-0.45 -0
NA
-0.35
8.00
NA
-0.27
-0.10
-0.18
-0.28
NA
SCA
NA
-0.0015
NA
-0.89
-0.03
8.03
.44
-0.44
9.04
-0.40
.12
NA
-0.44
-0.25
NA
9.10
9.06
-0.21
-0.02
NA
bias for ttfers through
fetter
Dl
-0.001
-0.0012
-1.53
-4.83
-0.55
-0.48
-0.10
9.55
-2.30
-0.21
-2.73
NA
9.34
NA
NA
-0.92
9.07
9.18
9.34
NA
4

Model Bias
6 fetter 7
SCA
NA
-0.0019
-4.89
-4.11
-1.14
-1.44
-1.50
-0.33
-1.47
-1.27
-0.76
NA
-0.04
NA
NA
-0.81
-0.25
-0.10
9.32
NA
Dl
9.044
-0.0007
-6.24
-11.39
-1.35
-1.20
-0.54
9.59
-3.28
9.71
-9.15
NA
9.48
9.03
NA
-1.65
9.17
9.50
9.05
9.08
SCA
NA
-0.0008
-5.48
-7.42
-0.23
9.33
1-.91
-0.30
-1.98
-0.59
-4.18
NA
9.53
-0.81
NA
-0.07
9.58
9.05
9.62
1-.02

fetter 8
Dl
-0.019


SCA
NA
-0.0001 -0.0008
NA
-2.01
-0.20
-0.43
-0.43
-0.25
-0.85
-0.90
NA
NA
-0.13
-0.09
NA
NA
-0.25
9.00
NA
NA
NA
-3.15
-0.78
-0.87
-0.79
-1.29
-0.58
-1.03
NA
NA
-0.35
-0.39
NA
NA
-0.10
9.14
NA
NA
NA: not applicable



Table 1 Summary of model prediction bias for ttfers through 8
                                        -92-

-------
§ra meter
TOC (ngL)
UV-254 (l£m)
TOX &L Cl ')
TTHM &L)
HAAS &L)
HAA6 &L)
HAA9 &L)
CF&L)
BDCM &L)
DBCM &L)
BFfiL)
MCAA &L)
DCAA &L)
TCAA &L)
MBAA&L)
DBAA &L)
BCAA &L)
DC BAA &L)
CDBAA &L)
TBAA &L)
Mean RSS
Dl
0.06
0.0015
4.7
5.4
1.3
1.6
2.5
0.7
2.3
1.0
4.6
NA
0.5
0.5
NA
1.1
0.5
0.6
0.6
1.9
SCA
NA
0.0019
7.8
4.1
1.3
1.7
2.2
0.9
1.6
1.1
2.4
NA
0.7
0.7
NA
0.8
0.5
0.5
0.9
1.8
Mean normaliEd RSS Mean bias
%
Dl
7.2
18
8.5
19
26
23
29
36
29
19
38
NA
40
39
NA
28
28
119
98
87
SCA
NA
23
16
18
28
25
26
36
20
21
20
NA
44
61
NA
21
24
115
111
84
Dl
8.028
-0.0008
-3.88
-4.14
-0.24
-0.35
-0.22
8.17
-1.61
-0.25
-3.77
NA
8.05
8.05
NA
-0.66
-0.07
8.15
8.04
8.08
SCA
NA
-0.0011
-5.19
-2.73
-0.43
-0.46
-0.36
-0.60
-0.61
-0.67
-1.34
NA
-0.14
-0.38
NA
-0.07
-0.04
8.09
8.31
4.02
Mean normaliEd bias
%
Dl
3.7
-8.2
-7.4
-15
-4.4
-4.7
-2.3
6.5
-16
-4.6
-30
NA
3.2
4.1
NA
-17
-3.5
15
2.8
3.8
SCA
NA
-12
-9.8
-10
-7.9
-6.1
-3.9
-22
-6.1
-13
-11
NA
-8.9
-29
NA
-1.6
-2.2
8.4
24
46
Count
8
8
2
8
8
8
8
8
8
8
6
0
8
5
0
6
8
7
3
1
NA: not applicable

Table 3Summary of mean RSS,mean bias.normalied mean RSS,and normlied mean bias for all
waters
                                           -93-

-------
the context of the actual magnitude of the RSS and bias values.  For each parameter, the average
normalized RSS and bias across all waters are summarized in Table 13.

The cumulative frequency distribution plot  of normalized RSS values  is shown in Figure 83.
The distribution of normalized RSS values for both predictive approaches is similar, indicating
that there was little difference in the relative success of the two methods to predict experimental
data over the entire data set.  The 25th to 75th percentile range of normalized RSS values for DI
predictions was 16 to 39 percent, while that for SCA predictions was similar, 16 to 38 percent.
The 10th to 90th percentile range of normalized RSS values for both methods was also similar, at
11 to 57 percent for the SCA method and  12 to 57 percent for the DI method.

Figure 84 shows the cumulative  frequency  distribution of normalized bias values.  Across all
waters and  water  quality parameters, both predictive methods tended to underestimate the
experimental data.  The median of the distribution for the DI method was -6 percent, while that
for the SCA method was -10 percent. The 25th to 75th percentile range of the distribution was -
16 to  +3  percent for the DI method.   The SCA  method  more often underpredicted the
experimental data as indicated by a 25th to 75th percentile range of the distribution of -20 to -1
percent. The 10th to 90th percentile range of the distribution for the DI method was -25 to +30
percent, while that for the SCA method was -28 to +10 percent.

Across all waters and analytes, the distribution of prediction error was shown to be similar based
on the cumulative frequency distribution of RSS.  However, an evaluation of the results for each
parameter is still needed and is addressed below in Sections 4.4.1 and 4.4.2. By comparing the
error,  as  measured by  RSS and bias,  for each  parameter  across all  waters, the relative
performance of each model for each individual analyte can be assessed.  A summary of all model
predictions for all parameters and all waters is included in Appendix F.   For this discussion,
examples of DI and SCA model results were  selected from the eight runs.

4.4.1   Surrogates and  Class Sums

As shown in Figures 49 through 56 (Section  4.3), the DI method was able to successfully predict
the integral breakthrough curve for TOC for the eight runs examined.  The average DI model
RSS for all eight runs was 0.055 mg/L, which compares to a value of 0.025 mg/L for the average
of all eight  best-fit curves  applied to the data.  The DI method prediction average RSS  was
slightly more than twice that for a best curve fit, but both values were very low.  Expressed as a
fraction of the average TOC concentration  of the experimental blended effluent data for each
water, the  normalized RSS for the DI prediction was low, 7 percent.  The average bias for the DI
method prediction was  low and  positive,  +0.028  mg/L.   The mean normalized bias  (bias
expressed  as a fraction  of the average blended  effluent experimental  data) was also  low,  4
percent.

The success of the DI model for predicting the TOC integral breakthrough curve is an important
verification prior to the application of the  SCA method, as this method relies on the TOC integral
breakthrough curve as a basis  for predicting the  integral  breakthrough  curves of all other
parameters.  Although an excellent predictor of the TOC integral  breakthrough curve, the DI
method was not as successful in predicting the integral breakthrough curve for DBFs, especially
brominated DBF species, as shown by the examples given in Section 4.3 (Figures 61 through
                                      -94-

-------
82).  The results of integral breakthrough curves predicted by the computationally-simpler SCA
procedure will be compared to those predicted by the DI method.

Figures 85  and 86 compare the DI and SCA UV254 integral breakthrough predictions against
experimental data for Waters 1 and 3.  During both these runs, the RSS of the predictions were
very low (all less than 0.002 I/cm). For Water 1, the DI method was able to match experimental
results early in the run very well, while the SCA method underpredicted the observed data. Later
in the run,  both methods slightly underpredicted the experimental integral breakthrough curve.
The DI method yielded a closer prediction based on RSS values for Water 1. The bias for both
predictions was negative (DI:  -0.0007 I/cm; SCA: -0.0012 I/cm), indicating quantitatively that
the models slightly underpredicted the experimental data.  For Water  3,  both  methods
underpredicted blended effluent UV254 early in the run. Beyond the midpoint of the run, both
methods matched the experimental results more closely.  The RSS values for each predictive
approach were equivalent, and the bias values were both slightly negative (DI:   -0.0008 I/cm;
SCA:  -0.0007 I/cm).  Overall, the UV254 DI prediction results yielded lower RSS values for
seven of the eight runs.

The integral breakthrough curve predictions for SDS-TTHM are given in Figures 87 and 88 for
Waters 1 and 7. Based on the calculated RSS values, the SCA method yielded better predictions
of the observed data than the DI method (Water 1 DI RSS: 3.9 |ig/L; Water 1 SCA RSS:   2.6
|ig/L; Water 7 DI RSS:  15  jig/L; Water 7 SCA RSS: 9.7 |ig/L), as was the case with six of the
eight runs.  Early in the run, at the point of initial breakthrough, the SCA method underpredicted
experimental results,  while the DI method  matched  the experimental data  more closely.
However, later in the run,  at higher breakthrough levels, the SCA method was able  to better
predict  the  experimental data.   This  pattern was  repeated during most  runs.   The more
pronounced deviation early in the run is preferable over later in the  run, because the target
treatment objective for SDS-TTHM is usually  exceeded later  in  the run, where  the  SCA
prediction is closest to the experimental data.  The prediction bias for Water 1 for both methods
was negative (DI: -2.5 |J.g/L; SCA: -1.6  |ig/L).  Both predictive methods also yielded a negative
bias for Water 7  (DI: -11 |ig/L; SCA: -7.4 |ig/L), indicating that the experimental results were
underpredicted by both predictive models.  The average bias  for all  runs was negative for both
methods.

Neither predictive  method was consistently more  successful for the prediction of the integral
breakthrough curve for SDS-HAA5, SDS-HAA6 and SDS-HAA9.  For SDS-HAA5 and SDS-
HAA9, the DI method was slightly more successful, with closer matches to experimental  data
based on RSS values in five out of eight cases for each parameter.   For SDS-HAA6,  the SCA
method was a better predictor in five  out of eight  cases.  The  average bias for both  model
predictions for SDS-HAA5, SDS-HAA6, and SDS-HAA9 for all runs was negative.

Figures 89 and 90 show the  model prediction results for SDS-HAA5  for Waters 2 and 7,
respectively.  For these two waters, the SCA method yielded a better prediction of the observed
blended effluent data by comparison of calculated RSS values  (Water 2 DI RSS: 1.7 |ig/L; Water
2 SCA RSS:  1.5 |ig/L; Water 7 DI RSS: 2.0 |ig/L; Water 7 SCA RSS:  1.5 |ig/L).  For Water 2,
both methods underpredicted experimental results throughout the run, while for Water 7, the DI
method underpredicted  the  observed data throughout the run.   The SCA method initially
underpredicted experimental data,  but between the midpoint and end  of the run was a close
match to the data.  The calculated bias for both waters and both methods was negative (Water 2


                                      -95-

-------
DI bias: -1.4 ug/L; Water 2 SCA bias:  -1.2 ug/L; Water 7 DI bias: -1.3 ug/L; Water 7 SCA
bias: -0.2 ug/L).

For Waters 3 and 5, Figures 91 and 92, respectively, compare the model prediction results for
SDS-HAA6.  For Water 3, both models overpredicted blended effluent concentrations, especially
towards the end of the run (DI bias:  +1.5 ug/L; SCA bias:  +0.7 ug/L), and the RSS values were
similar (DI RSS:  2.3  ug/L;  SCA  RSS:  2.3 ug/L).  For Water 5, both  models  matched the
observed results well, with the SCA method resulting in the closest match based on RSS values
(DI RSS:  0.8;  SCA RSS:  0.6).  The bias calculated for the DI prediction was negative (-0.3
ug/L), while that for the SCA prediction was slightly positive (+0.03 ug/L).

For Waters 5 and 6, the model predictions for blended contactor effluent SDS-HAA9 are shown
in Figures 93 and 94, respectively.  The bias for both model  predictions  for both waters was
negative (Water 5 DI bias: -0.9 ug/L; Water 5 SCA bias: -0.4 ug/L; Water 6 DI bias:  -0.1 ug/L;
Water  6  SCA  bias:  -1.5  ug/L).   Although  the  SCA underpredicted  blended  effluent
concentrations early in the run, it yielded a very close match to the experimental data towards the
end of the run.   For Water 5, the  RSS value for  the  SCA method prediction (1.1  ug/L) was
slightly lower than that for the DI method prediction (1.2 ug/L).  For Water 6, the DI method
yielded a better prediction of the integral breakthrough curve than did the SCA method, based on
RSS values (DIRSS: 1.9 ug/L; SCA RSS:  2.4 ug/L).

For two waters (Waters 6 and 7), SDS-TOX was also analyzed in the blended effluent.  For these
two cases, the DI method was a better predictor of the integral breakthrough curve than the SCA
method. For Water 6 (Figure 95), the SCA method underpredicted experimental data throughout
most of the run, while for Water 7 (Figure 96), the SCA method underpredicted the  experimental
results  during the initial portion of the run. Later in the run the SCA method resulted in a very
good prediction of the integral breakthrough curve.  For both waters, the RSS  values for the DI
method were lower than those for the SCA method (Water 6 DI RSS: 2.4 ug/L Cl"; Water 6 SCA
RSS:  6.3 ug/L Cl"; Water 7 DI RSS:  6.9  ug/L Cl"; Water 7 SCA RSS:  9.3 ug/L Cl").  The bias
values  for both runs and both model predictions were negative  (Water 6 DI bias: -1.5 ug/L Cl";
Water 6 SCA bias: -4.9 ug/L Cl"; Water 7 DI bias:  -6.2 ug/L Cl"; Water 7 SCA bias: -5.5 ug/L
Cl").

4.4.2  DBP Species

With a few exceptions, the DI method was able to better predict non-brominated DBP species,
while the SCA method was a better predictor of the brominated species.  This is due in part to the
marginal ability of the DI method to predict blended effluent concentrations that result when the
single contactor curve shows  a peak, or a  steep breakthrough followed by a flat plateau.  The
influence of the bromide to TOC ratio on the single contactor breakthrough  curve is not captured
by the  DI method, while  the SCA method, which inherently correlates  DBP formation to TOC
concentration, was better able to predict the integral  breakthrough curve for  brominated DBP
species. The relationship between TOC concentration and DBP formation and speciation in both
the single and blended contactor effluents is discussed in Section 4.2.

For SDS-CF, the DI method yielded lower RSS values for experimental predictions  in five out of
eight cases.   Figures  97 and 98 show examples  of model predictions  for Waters 2 and 6,
                                      -96-

-------
respectively. For these two examples, the SCA method underpredicted the integral breakthrough
curve at the point of initial breakthrough and towards the end of the run.  The SCA prediction
bias was negative for both runs (Water 2:  -0.9 |ig/L; Water 6:  -0.3 |ig/L).  The DI method
yielded a better match to the observed for Water 2 (DI RSS: 0.6 |ig/L; SCA RSS: 1.6 |ig/L), but
the SCA method was a better predictor of the SDS-CF integral breakthrough curve for Water 6
(DI RSS:  1.4  |ig/L; SCA RSS: 0.7 |ig/L).  The DI model bias was negative for Water 2 (-0.1
|ig/L) and positive for Water 6 (+0.5  |ig/L). Overall, the average bias for all runs was slightly
positive for the DI prediction, while that for the SCA prediction was slightly negative.

Figures 99 and 100 show the DI and  SCA method predictions of SDS-BDCM experimental
results for Waters 1 and 8, respectively. Overall, the SCA method was a more successful  model
for predicting observed data, with RSS values lower than those for the DI method for six of the
eight runs.  In  most cases, and as shown in the two examples, the DI method underpredicted the
experimental blended effluent concentration throughout the entire run.  This was also reflected
by  the negative bias values for the  prediction results.   The  SDS-BDCM single  contactor
breakthrough curves typically exhibited sharp breakthrough curves followed by a flat plateau or
peak curves. For the two examples given, the SCA method prediction provided a closer match to
the experimental data (Water 1 DI RSS:  1.5 |ig/L; Water 1 SCA RSS:   1.0 |ig/L; Water 8 DI
RSS: 1.2 |ig/L; Water 8 SCA RSS: 0.8 |ig/L).  The SCA prediction bias was negative (Water 1:
-0.7 |ig/L; Water 8: -0.6 |ig/L), as was that for the DI predictions (Water 1:  -1.1 |ig/L; Water 8:
-0.8 |ig/L). For all runs, the average model bias for both models was negative, although the SCA
prediction bias was lower in magnitude as compared to that for the DI method.

For SDS-DBCM, the DI method yielded a better fit to the observed data  in five of eight cases.
Although a brominated DBF species, the breakthrough  of  SDS-DBCM  was typically  shaped
similar to that of SDS-CF:  a gradual increase in concentration  over time, without  the  "sharp
increase and plateau"  or "peak" curve shapes observed for other brominated species. Figures 101
and  102 show  examples of the results obtained for Waters 3 and 7, respectively.  For Water 3
(Figure 101), both methods slightly underpredicted effluent concentrations at the beginning of
the run. The DI method overpredicted effluent concentrations towards the end of the run, while
the SCA method closely  matched experimental data after the first third of the run.   The SCA
prediction bias was positive (+0.6 |ig/L), while that for the DI prediction was slightly negative (-
0.1 |ig/L).  The RSS value for the SCA prediction (0.7  jig/L) was lower than  that  for the DI
prediction (1.0 |ig/L).  For Water 7,  shown in Figure  102, the  SCA method  underpredicted
blended effluent  SDS-DBCM levels at the point of initial breakthrough and towards  the  end of
the run, while the DI method either overpredicted or matched effluent levels. In this case,  the DI
prediction RSS value (1.0 |ig/L) was slightly lower than those for the SCA prediction (1.1  |ig/L).
For these two examples, a positive model bias was measured for the DI prediction (+0.7  |ig/L),
while a negative model bias occurred  for the SCA prediction (-0.6 |ig/L).  For all waters, the
mean prediction bias for both methods was negative. The magnitude of the mean bias was lower
for the DI procedure.

The SCA method resulted in closer matches to the observed  data than did the DI method, based
on calculated RSS values, for all SDS-BF integral breakthrough curve predictions (six  of the
eight runs yielded data above the MRL). Several single contactor effluent SDS-BF data formed
peak curves, and some of the integral breakthrough curves were also peak curves.  The DI
method underpredicted blended effluent SDS-BF concentrations.  The SCA method was able to
better  match the peak-shaped integral breakthrough curves that were  sometimes  observed.

                                      -97-

-------
Examples of the model prediction results are given in Figures 103 and 104 for Waters 1 and 5,
respectively.  The DI method underpredicted the experimental results throughout the entire run
for both  waters.   Although  the SCA method initially  underpredicted  blended  effluent
concentrations,  at higher concentrations the  accuracy  of the  predicted results  increased,
especially relative to the DI method results.  The measured bias values for the SCA prediction
(Water 1:  -0.3 |ig/L; Water 5: -0.1  |ig/L) were lower in magnitude than those for the DI method
(Water 1:  -2.2 ng/L; Water 5:  -0.5 ng/L), and bias values were negative for both models.  The
average model bias for the six runs analyzed (DI:  -3.8 ^ g/L;  SCA:  -1.3 |ig/L) was negative for
both methods, and the magnitude of the DI method bias was  greater than that for the SCA
method by a factor of 2.8.

The DI method continued to perform well for non-brominated HAA species, while the SCA
method was superior when predicting the integral breakthrough curve for brominated  species.
Exceptions to this trend occurred for  SDS-CDBAA and SDS-TBAA, for which the DI method
more  often yielded better predictions.  However, only four total cases were  available for
evaluation of both species, due to concentrations not exceeding the MRLs during many runs.
Furthermore, analysis was complicated because formed levels were typically only slightly higher
than the MRL. Values below the MRL were assigned a value  of zero, which was not the case for
the prediction  methods. For the one case available to compare SDS-TBAA blended  effluent
concentrations, the SCA method  more closely matched effluent levels (three points) that were
measured above the MRL (4.0 |ig/L).  However, the overall calculated RSS for the SCA method
was slightly higher than that  for the DI method.  For SDS-DCBAA, comparisons were possible
during only three runs, and  the  SCA method was a more  accurate predictor of the  integral
breakthrough curve in all three cases.  No comparisons were  possible for SDS-MCAA or SDS-
MBAA.

For SDS-DCAA, the DI method resulted in lower RSS values in six of the eight runs.  Examples
of the results are shown in Figures 105 and 106 for Waters 1  and 4, respectively.  While the DI
method well-predicted the data throughout the entire run for Water 1  (RSS: 0.4 |ig/L), the SCA
method initially underpredicted and then overpredicted the observed data (RSS:  1.5 |ig/L). For
Water 4,  the  DI  method again was a successful  predictor  of  the  SDS-DCAA  integral
breakthrough curve (RSS:  0.3  |ig/L),  while  the SCA method consistently  underpredicted
experimental data (RSS:  0.5 |ig/L). For Water 1, both predictions yielded a positive bias (DI:
+0.1 ng/L; SCA: +0.3 |ig/L), while for Water 4, the bias for the SCA prediction was negative (-
0.3 |ig/L), and that for the DI prediction was low and positive  (+0.04 |ig/L).  Similar results were
obtained for SDS-TCAA  predictions based on  the two  methods:  for  four of  five possible
comparisons,  the  RSS  value for the  DI method was  lower than that for the  SCA method,
indicating a better prediction. The  average bias for the DI method prediction (+0.1 |ig/L) was
low and positive, while that for the SCA method was negative (-0.4 |ig/L).

For SDS-DBAA and SDS-BCAA, whose single contactor breakthrough curve typically exhibit
sharp breakthrough curves  followed  by a flat  plateau  or peak  curves, the  SCA  method
outperformed the DI method for integral breakthrough curve prediction. RSS values for the SCA
prediction were lower in four out six cases for SDS-DBAA, and in six  out of eight cases for
SDS-BCAA.  The average model bias was negative for both parameters.   For SDS-DBAA, the
average bias magnitude was larger for  DI predictions (DI:  -0.7 |ig/L; SCA: -0.1 |ig/L), while for
SDS-BCAA, the average bias for both methods was low  (DI:  -0.07 |ig/L; SCA:   -0.04 |ig/L).
Figure 107 shows the SDS-DBAA results obtained for Water 2 (DI RSS:  1.1 |ig/L; SCA RSS:

                                      -98-

-------
0.6 |ig/L), while Figure 108 shows the SDS-BCAA results obtained for Water 8 (DI RSS: 0.6
|ig/L; SCA RSS: 0.3 |ig/L). In both these examples, the SCA prediction was more accurate than
the DI prediction based on RSS values. The model bias for these two examples was negative for
both models, and lower in magnitude for SCA method predictions as compared to DI method
predictions .

This  analysis showed that the DI prediction  was  successful in  predicting the integral
breakthrough curve for TOC, which is an important step for use of the  SCA method.   The
cumulative frequency distribution comparisons of the SCA and DI model results showed that the
two methods were equivalent in their ability to predict the integral breakthrough curve, based on
a comparison across all waters  and  water quality  parameters.  Over the entire data set, both
methods were biased negative in their predictions  of the experimental data.  Both the DI and
SCA  predictions agreed well with  experimental  data  for surrogates and class sums.   For
chlorinated DBF species, the DI method was a better predictor of the integral breakthrough curve
than the SCA method.  However, SCA method predictions outperformed those determined by the
DI procedure for brominated DBF species integral breakthrough curves.

The main advantage of the  SCA method over the DI procedure is that predictions using the SCA
minimize the number of calculations necessary to predict blended contactor water quality as a
function of  single  contactor run time.  Based on the results of this  study, use of the  SCA
procedure for predicting the integral breakthrough curve is  recommended during  ICR  GAC
treatment study data analysis.
                                      -99-

-------
   CD
      100
       80 -
   o  60
   CD
   .Q
   O
   -I—'
   CD
   CD
   CL
      40 -
                                                       To
                                                              » Dl prediction
                                                              o SCA prediction
                                10                    100

                          Normalized residual sum-of-squares, RSS (%)
                                                       1000
Figure 83 Cumulateive frequency distribution plot of normalized residual sum-of-squares
(RSS) for Dl and SCA model predictions
      100
   CD
       80 -
   o  60
   CD
   .Q
   _
   O
   -i—'
   CD
   CD
   CL
      40 -
      20 -
        -100
-50
 0            50

Normalized Bias (%)
                                                              » Dl prediction
                                                              o SCA prediction
100
150
Figure 84 Cumulative frequency distribution plot of normalized bias for Dl and SCA model
predictions
                                    -100-

-------
      0.035
      0.030 -
      0.025 -
      0.000
      UV254

      EBCT = 15 min.
      c0 = 0.094 1/cm
                                                     •   Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = 0.0011)
                                                   	SCA prediction (RSS = 0.0017)
           0                       25                      50
                                    Scaled operation time (days)

Figure 85 Comparison of Dl and SCA methods for predicting the UV254
integral breakthrough curve for Water 1
                                                                          75
        0.020
        0.015 -
      o
      o>
      o
      c
      CD
     .Q
      O
      
-------
                                                     •  Observed data
                                                        Best fit
                                                   	Dl prediction  (RSS = 3.93)
                                                        SCA prediction (RSS = 2.55)
         0                       25                      50
                                 Scaled operation time (days)

Figure 87 Comparison of Dl and SCA methods for predicting the SDS-TTHM
integral breakthrough curve for Water 1
                                                                    75
    125
    100 -
  a  75 H
  o
  '-4—'
  S
  |  50
  o
  O
     25 -
SDS-TTHM

EBCT = 20 min.
 co = 200
                                           •  Observed data
                                         	Best fit
                                         	Dl prediction  (RSS = 15.22)
                                         	SCA prediction (RSS = 9.72)
                   25
                     50           75          100
                      Scaled operation time (days)
125
150
Figure 88 Comparison of Dl and SCA methods for predicting the SDS-
TTHM integral breakthrough curve for Water 7
                                    -102-

-------
    15
    10
 O)
 o
 "H—•
 2
 -i—'

 I
 O   c J
 O   5 H
          SDS-HAA5
          EBCT = 20 min

          c° = 24 |jg/L
                                                     •   Observed data
                                                  	Best fit
                                                  	Dl prediction (RSS = 1.69)
                                                  	SCA prediction (RSS = 1.48)
                     50
                         100            150
                      Scaled operation time (days)
200
250
Figure 89 Comparison of Dl and SCA methods for predicting the SDS-
HAA5 integral breakthrough curve for Water 2
    25
    20 -
     15 -
  CD
  O
  c
  o
  O
     5 -
SDS-HAA5

EBCT = 20 min.
CD = 65 pg/L
                                            •   Observed data
                                         	Best fit
                                         	Dl prediction (RSS = 2.02)
                                         	SCA prediction (RSS = 1.52)
                   25
                     50          75          100
                       Scaled operation time (days)
  125
150
Figure 90 Comparison of Dl and SCA methods for predicting the SDS-
HAA5 integral breakthrough curve for Water 7
                                    -103-

-------
       20
       15 -
     o
     '•a 10
     O>
     o
     o
     O
        5 -
          SDS-HAA6

          EBCT = 20 min
          co =  43 |jg/L
                                                      •  Observed data
                                                    	Best fit
                                                    	Dl prediction  (RSS = 2.29)
                                                    	SCA prediction (RSS = 2.3)
                     50
                             100         150         200
                               Scaled operation time (days)
250
300
Figure 91  Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 3
   16

   14 -

   12 -

1 1°-
o
'•a   8
   O>
   c  6
   o
   O
      4 -
      2 -
           SDS-HAA6
           EBCT = 20 min.
           c0 =  34 M9/L-
                                                   •   Observed data
                                                	Best fit
                                                	Dl prediction (RSS = 0.82)
                                                	SCA prediction (RSS = 0.55)
                  50
                         100        150       200       250
                              Scaled operation time (days)
 300
350
Figure 92 Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 5
                                    -104-

-------
    25
    20 -
    15 -\
 §  10-

 o
 O
     5 -
   SDS-HAA9


   EBCT = 20 min


   C0 = 48 |jg/L
                                             •   Observed data


                                           	Best fit


                                           	Dl prediction (RSS = 1.19)

                                           	SCA prediction (RSS = 1.13)
                50
                   100        150       200        250


                        Scaled operation time (days)
 300
350
Figure 93 Comparison of Dl and SCA methods for predicting the SDS-

HAA9 integral breakthrough curve for Water 5
   25
   20 -
 a  15 H

 c
 o
 '-4—'

 S


 §  10
 c
 o
 O
    5 -
SDS-HAA9




 EBCT = 20 min.



 Co =  61 |jg/L
                                                  Observed data


                                                  Best fit


                                            	Dl prediction (RSS = 1.88)


                                            	SCA prediction (RSS = 2.41)
                  50
                      100          150          200


                        Scaled operation time (days)
250
 300
Figure 94 Comparison of Dl and SCA methods for predicting the SDS-

HAA9 integral breakthrough curve for Water 6
                                    -105-

-------
    o
    o
    '-4—'
    CD
    CD
    O
    C
    O
    O
       100
        80 -
        60 H
        40 -
        20 -
 SDS-TOX

EBCT = 20 min.
co = 305 |jg/L Cl-
                                         •   Observed data
                                       	Best fit
                                       	Dl prediction (RSS = 2.39)
                                       	SCA prediction (RSS = 6.26)
                     50
                    100         150         200
                      Scaled operation time (days)
250
300
Figure 95 Comparison of Dl and SCA methods for predicting the SDS-TOX
integral breakthrough curve for Water 6
       200
       150 -
    O
    .2  100 -
    "CD
    CD
    o
    o
    0
        50 H
 SDS-TOX

EBCT = 20 min.
co = 486 |jg/L Cl
                                                      •   Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = 6.92)
                                                    	SCA prediction (RSS = 9.3)
                     25
                     50          75          100
                      Scaled operation time (days)
125
150
Figure 96 Comparison of Dl and SCA methods for predicting the SDS-
TOX integral breakthrough curve for Water 7
                                    -106-

-------
       15
    O)
    o
    '-4—'
    TO
    o>
    o
    c
    o
    O
       10 -
              SDS-CF
              EBCT = 20 min.
              c0= 41.9 |jg/L
                                                      •   Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = 0.57)
                                                    	SCA prediction (RSS = 1.59)
                       50
                         100           150
                      Scaled operation time (days)
                                200
250
Figure 97 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 2
       6 -
    O
    '•£  4
    o>
    o
    c
    o
    O
 SDS-CF

EBCT = 20 min
c0 = 55.3 pg/L
                    50
  •   Observed data
	Best fit
	Dl prediction (RSS = 1.37)
	SCA prediction (RSS = 0.74)
                    100          150          200
                     Scaled operation time (days)
                                  250
300
Figure 98 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 6
                                    -107-

-------
      20
      16 -
    O)
    a 12 H
    c
    o
    '-4—'
    S
    §  8
    c
    o
    O
       4 -
SDS-BDCM

EBCT = 15 min.
 c0 = 19.3|jg/L
                                          •  Observed data
                                        	Best fit
                                        	Dl prediction (RSS = 1.52)
                                        	SCA prediction (RSS = 1.03)
                                25                      50
                                 Scaled operation time (days)
                                                                   75
Figure 99 Comparison of Dl and SCA methods for predicting the SDS-
BDCM integral breakthrough curve for Water 1
        4 -
        3 -
      8 2
      o
     O
        1 -
 SDS-BDCM
 EBCT = 7.2 min.
 c0 = 2.6 |jg/L
        0 -!•-
         0
                                       •   Observed data
                                     	Best fit
                                     	Dl prediction (RSS = 1.2)
                                     	SCA prediction (RSS = 0.81)
               50               100
                     Scaled operation time (days)
150
200
Figure 100 Comparison of Dl and SCA methods for predicting the SDS-
BDCM integral breakthrough curve for Water 8
                                    -108-

-------
    o
    I
    I
    o
    O
16

14 -

12 -

10 -

 8 -

 6 -

 4 -

 2 -
            SDS-DBCM

           EBCT = 20 min
           co = 44.5 |jg/L
                                                     •   Observed data
                                                  	Best fit
                                                  	Dl prediction (RSS = 1.04)
                                                  	SCA prediction (RSS = 0.68)
                    50
                         100         150         200
                          Scaled operation time (days)
250
300
Figure 101 Comparison of Dl and SCA methods for predicting the SDS-
DBCM integral breakthrough curve for Water 3
   25
   20 -
    15 -
 §  10
 c
 o
 O
    5 -
   SDS-DBCM

   EBCT = 20 min.
   c0 = 66.2 |jg/L
                                               •  Observed data
                                             	Best fit
                                             	Dl prediction  (RSS = 0.99)
                                             	SCA prediction (RSS = 1.1]
                  25
                        50           75          100
                         Scaled operation time (days)
125
150
Figure 102 Comparison of Dl and SCA methods for predicting the SDS-
DBCM integral breakthrough curve for Water 7
                                   -109-

-------
16

14 -

12 -

10 -
  o
  '-4—'
  CD
  CD
  O
  c
  o
  O
          SDS-BF

            EBCT = 15 min
            c0=  3.7|jg/L
                                                     •  Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = 2.76)
                                                   	SCA prediction  (RSS = 1.31)
                               25                      50
                                Scaled operation time (days)
                                                                          75
 Figure 103 Comparison of Dl and SCA methods for predicting the SDS-BF
 integral breakthrough curve for Water 1
     3.0
     2.5 -
     2.0 -
  o
  '•SS  1-5 H
  o  1.0 H
     0.5 -
     0.0
        SDS-BF
        EBCT = 20 min.
        c0=  1-2M9/L
                                                •   Observed data
                                             	Best fit
                                             	Dl prediction  (RSS = 0.55)
                                             	SCA prediction (RSS = 0.26)
                  50
                      100        150        200        250
                           Scaled operation time (days)
300
350
Figure 104 Comparison of Dl and SCA methods for predicting the SDS-BF
integral breakthrough curve for Water 5
                                    -110-

-------
       10
       6 -
    o
    "CD
    O>  4 .
    O  ^ I
    c
    o
    O
       2 -
             SDS-DCAA

             EBCT = 15 min.
             c0 -  12.5 |jg/L
                                              •  Observed data
                                            	Best fit
                                            	Dl prediction  (RSS = 0.35)
                                            	SCA prediction (RSS = 1.46)
         0                      25                      50
                                 Scaled operation time (days)
Figure 105 Comparison of Dl and SCA methods for predicting the SDS-
DCAA integral breakthrough curve for Water 1
                                                                        75
   6 -
   5 -
   4 -
 o
 '-4—'
 CD
 SDS-DCAA

EBCT = 20 min.

 co = 20.3 |jg/L
                                                     •   Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = 0.32)
                                                   	SCA prediction (RSS = 0.47)
                       50
                                  100
                        Scaled operation time (days)
150
200
Figure 106 Comparison of Dl and SCA methods for predicting the SDS-
DCAA integral breakthrough curve for Water 4
                                    -111-

-------
      6 -
   o
   14H
   I
   o
   O
 SDS-DBAA

EBCT = 20 min
c0 = 5 |jg/L
                                                       •  Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = 1.1)
                                                     	SCA prediction (RSS = 0.6)
                                    100            150
                                 Scaled operation time (days)
                                                        200
  250
Figure 107 Comparison of Dl and SCA methods for predicting the SDS-
DBAA integral breakthrough curve for Water 2
       3.5
       3.0 -
       2.0 H
     O
     0  1.0 H
       0.5 -
       0.0
     SDS-BCAA
     EBCT = 7.2min.
     CD = 3 |jg/L
                                           •   Observed data
                                         	Best fit
                                         	Dl prediction (RSS = 0.59)
                                         	SCA prediction (RSS = 0.31)
                           50               100
                                  Scaled operation time (days)
                                                    150
200
Figure 108 Comparison of Dl and SCA methods for predicting the SDS-
BCAA integral breakthrough curve for Water 8
                                    -112-

-------
4.5   Analysis of Model Applicability to Finite Number of Contactors

Both  the DI and SCA  models for predicting the integral breakthrough curve,  a relationship
between single contactor run time and blended  contactor effluent water quality, rely on the
assumption that an infinite number of contactors are operated on-line in parallel-staggered mode.
The application of the DI procedure to any parameter relies on the infinite contactor assumption;
the DI prediction of the TOC integral breakthrough curve is a preliminary step utilized by the
SCA  method  for predicting the integral  breakthrough  curves  of  all  other  water quality
parameters.   The application  of these  methods to  situations involving finite  numbers  of
contactors may be limited  by this assumption.   The  following  analysis was developed  to
determine the acceptability of the error incurred by the infinite contactor assumption.

Using the step logistic function model, a  series of integral  breakthrough  curves were developed
and plotted for the following numbers of contactors in parallel, N: 2, 3, 4, 6, 10, 20, and infinite.
For the  case of an infinite number of contactors Equation 12 (Section 1.4) was used, while in all
other cases the numerical integration shown in Equation 4 (Section 1.2) was used.  The single
contactor breakthrough curve from which the integral curves were developed was  also plotted.
The results of this modeling are shown in Figures 109 through 114 for varying values of B and
D, logistic function parameters that affect the shape of the curve. Values used for B ranged from
10 to 30, while values for D ranged from 0.05 to 0.20.  These value ranges for B and D were
chosen to reflect the range typically seen for fits  of GAC breakthrough curve data. Set values
were used instead of experimental data to examine the impact of B and D while A  and A0 were
held constant.  For the curve fits  performed in this study, the 25th and 75th percentiles for the
values of B were 5 and 25, respectively, while the 25th and 75th percentiles for the values  of D
were  0.02 and  0.07, respectively.  The best-fit parameters for all  curve fits performed in this
study are included in Appendix G.  The parameter values used for A0 and A were held constant
for all six series of breakthrough curves, at values of 0 and 1, respectively. (The logistic function
parameter A  represents the asymptote to which the logistic function is  approaching,  while A0
represents a  step value  given to the function.)  All the data were plotted over 100 full-scale
operation days.

The  graphs  in  Figures  109  through 114  show that as  the  number of  contactors operated in
parallel-staggered mode increases, the integral curve approaches the model results for an infinite
number of contactors.  The largest incremental benefit afforded by operating contactors  in
parallel-staggered  mode over  single  contactor  operation  occurs  when two  contactors  are
operated.  The benefit realized by adding an additional contactor  decreases as the number of
contactors on-line increases.

To quantitatively compare the results, the run time to a range of treatment objectives expressed
as a  fraction of the parameter A, the  value  to which  the breakthrough curve approaches
asymptotically, was estimated based on the  curves developed for N contactors.   These are not
percent  breakthrough values; they represent instead a fraction of the concentration that the single
contactor curve approaches asymptotically (asymptotic concentration). Values of 0.35, 0.50, and
0.65 were utilized as treatment objectives. Note that each value corresponds to a lower percent
breakthrough value than that calculated based on  the GAC influent concentration.  An extreme
case of  a 0.80 treatment objective was also examined.  This analysis is summarized in Tables 14
through 17.

                                       -113-

-------
The goal of this analysis was to establish what the smallest value for TV was for which the integral
curve (derived based on the infinite number of contactors assumption) would yield a maximum
10  percent error in the  estimated blended contactor run times, assuming this level  of error is
acceptable. The parameter N90 will represent this breakpoint and is defined as the minimum N at
which point the throughput of each contactor operated in parallel-staggered mode  is  90 percent
of that for an infinite number of contactors on-line. This is equivalent to the point at which the
run time estimated using an assumption of an infinite number of contactors on-line is within 10
percent of the actual run time of each of TV contactors.

Figures 115 and  116 establish a  relationship between integral breakthrough curve run time
estimates  based on an  infinite number  of contactors  and that  based on a finite number of
contactors. As shown in Figures  115 and 116, the minimum number of contactors operated in
parallel-staggered mode necessary so that the integral breakthrough curve for finite A/90 is  within
10  percent of the integral breakthrough curve  for infinite  contactors varies with treatment
objective. For B = 30 and D = 0.10 and treatment objectives ranging between 0.35  and 0.65, the
A/9o ranged from 7 to 13 contactors.  As the value used for the treatment objective was increased
(a less  stringent treatment objective  relative to GAC effluent concentration), A/90 increased.
Similar results were obtained for B = 30 and  D = 0.20.  As the treatment objective was  varied
between 0.35 and  0.65, the Ngo ranged from 7 to 13 contactors.  The Ngo to meet the given
treatment objectives did not vary as B and D were varied:  for treatment objectives  between 0.35
and 0.65, the A/90 varied from 7 to 13  contactors for both pairs of B  and D values modeled.
Therefore, the error between run times estimated based on an infinite number of contactors and
that estimated for N contactors is independent of the shape of the curve, although it is dependent
on the extent of breakthrough achieved.  For the extreme asymptotic concentration fraction value
of 0.80, A/9o was 22 contactors.

The assumption of an  infinite number of  contactors operated in  parallel-staggered  mode
simplifies the analysis of the integral breakthrough curve,  as does the assumption of a linear
breakthrough curve.  However, the data presented here indicate that for  breakthrough profiles
that follow the logistic function form, N90 varies between 7 and 13 contactors,  as  the treatment
objective is varied between 35 and 65 percent of the asymptotic concentration.  Thus, when
using the infinite contactor  assumption to model the integral breakthrough curve for less than 14
contactors in parallel, the magnitude of the treatment objective in relation to the single contactor
breakthrough curve is important in minimizing the error between qN, the specific throughput of a
finite number of contactors, A/, and q^, the specific throughput of each contactor assuming an
infinite number of contactors on-line.

Based on the use of the logistic function model and for the range  of treatment objectives
evaluated in this analysis, the integral breakthrough curve developed from the infinite contactor
assumption will yield estimated run times within 10 percent of actual run times for 13 or more
contactors operated in parallel-staggered mode.  For 10 contactors on-line, the infinite contactor
assumption will yield run time estimates within 12 percent of the actual run times.  In all cases,
run time estimates based on the infinite contactor assumption are longer than those for a finite
number of contactors,  and thus  yields a best case  estimate  of  GAC performance.    The
relationship developed between integral breakthrough curve run time estimates  based  on an
infinite number of contactors and that based on a finite number of contactors can  be applied to
                                      -114-

-------
ICR treatment study data to obtain GAC run time estimates that are more applicable to GAC
applications with a small number of parallel contactors.
Number of
contactors,
N


1
2
3
4
6
10
20
Run time to

5 = 30;
£> = 0.10
56
71
78
83
88
92
96
Table 14 Summary of run
Number of
contactors,
N


1
2
3
4
6
10
20
Run time to
5 = 30;
D = 0.10
50
67
75
80
86
91
95
treatment

5 = 30;
£> = 0.05
56
71
78
83
88
92
96
objective (0.35), expressed as percentage of run
infinite number of contactors

5=10;
£> = 0.10
54
70
78
82
87
92
96
times to a 0.35 treatment
treatment
5 = 30;
£> = 0.05
NA
NA
NA
NA
NA
NA
NA

5=10;
£> = 0.05
54
70
78
82
87
92
96
objective

5=10;
D = 0.2
54
70
78
82
87
92
96

objective (0.50), expressed as percentage of run
infinite number of contactors
5=10;
D = 0.10
50
67
75
80
86
91
95
5=10;
£> = 0.05
50
67
75
80
86
91
95
5=10;
D = Q.2
50
67
75
80
86
91
95
time for

5 = 30;
D = 0.2
56
71
78
83
88
92
96

time for
5 = 30;
D = 0.2
50
67
75
80
86
91
95
NA: not applicable, treatment objective not exceeded

Table 15 Summary of run times to a 0.50 treatment objective
                                     -115-

-------
Number of Run time to
contactors,
N


1
2
3
4
6
10
20
NA: not
Table 16

5 = 30;
£> = 0.10
41
59
69
74
81
88
94
treatment

5 = 30;
£> = 0.05
NA
NA
NA
NA
NA
NA
NA
objective (0.65), expressed as percentage of run
infinite number of contactors

5=10;
£> = 0.10
43
60
70
75
82
88
94

5=10;
£> = 0.05
NA
NA
NA
NA
NA
NA
NA

5=10;
D = 0.2
43
60
70
75
82
88
94
time for

5 = 30;
£> = 0.2
41
59
69
74
81
88
94
applicable, treatment objective not exceeded
Summary of run
Number of Run time to
contactors,
N


1
2
3
4
6
10
20
5 = 30;
£> = 0.10
NA
NA
NA
NA
NA
NA
NA
times to a 0.65 treatment
treatment
5 = 30;
£> = 0.05
NA
NA
NA
NA
NA
NA
NA
objective

objective (0.80), expressed as percentage of run
infinite number of contactors
5=10;
£> = 0.10
NA
NA
NA
NA
NA
NA
NA
5=10;
£> = 0.05
NA
NA
NA
NA
NA
NA
NA
5=10;
D = 0.2
31
48
58
64
73
82
90

time for
5 = 30;
£> = 0.2
28
45
56
63
71
81
89
NA: not applicable, treatment objective not exceeded



Table 17 Summary of run times to a 0.80 treatment objective
                                     -116-

-------
   1.2
   1.0 -
o
ro  0.8 -
CD
o
 _
 CD
-I—'
 CD
   0.4 -
CD
CL
   0.2 -
   0.0
     A0: 0

     A: 1
     B: 30

     D: 0.1
                   25
                               50            75

                                 Run time (days)
100
125
Figure 109 Integral breakthrough curves for varying numbers of contactors
operated in parallel-staggered mode (B = 30; D = 0.1)
   1.0
   0.8 -
CD
O
c
o
o
t_
CD
-I—'
CD


2.
CD
CL
   0.6 -
0.4 -
   0.2 -
   0.0
                   25
                               50            75

                                 Run time (days)
100
125
Figure 110 Integral breakthrough curves for varying numbers of contactors
operated in parallel-staggered mode (B = 30; D = 0.05)
                                -117-

-------
   1.2
   1.0 -
o
ro  0.8 -
CD
o

§
E
&  0.4 -
CD
CL
   0.2 -
   0.0
A0: 0

A:  1

B:  10

D:  0.1
                    25
                           50             75

                            Run time (days)
                                                                Number of

                                                                contactors
                                                                  -2

                                                                  -4

                                                                  10
                                                                -20 - - -  Inf.
100
125
Figure 111  Integral breakthrough curves for varying numbers of contactors

operated in parallel-staggered mode (B = 10; D = 0.1)
   1.0
   0.8 -
g  °-6J
o

o
o

CD
"CD  0.4
E
CD
CD
CL


   0.2 -
   0.0
A0: 0

A:  1

B:  10

D:  0.05
                    25
                           50             75

                            Run time (days)
                                                                 -20 -  - - 'Inf.
100
125
Figure 112  Integral breakthrough curves for varying numbers of contactors

operated in parallel-staggered mode (B = 10; D = 0.05)
                                 -118-

-------
   1.2
   1.0 -
o
'CD  0.8 -
CD
o
8  0.6 -
"CD
2
CD
CL
   0.4 -
   0.2 -
   0.0
                    25
                                  50             75
                                    Run time (days)
                                                               Number of
                                                               contactors
•1  -
>3  -
6  --
20 -
        	2
        '—4
        •-- 10
         - 'Inf.
100
            125
Figure 113  Integral breakthrough curves for varying numbers of contactors
operated in parallel-staggered mode (B = 10; D = 0.2)
   1.2
   1.0 -

 o
 |5 0.8 -
 CD
 O
 8 0.6 H
 CD
 CD 0.4 -
 CD
 CL
   0.2 -
   0.0
                    25
                                   50             75
                                    Run time (days)
                                                               Number of
                                                               contactors
                                                                 •1   -
                                                                 •3   -
                                                                 -6   --
                                                                 •20  -
         	2
         •—4
         •-- 10
          - 'Inf.
100
            125
Figure 114  Integral breakthrough curves for varying numbers of contactors
operated in parallel-staggered mode (B = 30; D = 0.2)
                                 -119-

-------
   100
c
O)
^   90 --
    80 -
o

"CD
t_
en
CD
£   70 -
CD
O
t_
CD
Q.

CD"
    60 -
    50 -
    40
           90 percent
                                   10             15

                                 Number of contactors, n
                                                            Treatment

                                                             objective
                                                                  0.35
                                                                	0.50

                                                                	0.65
                                                           20
25
Figure 115  Run time as a function of number of contactors in parallel,
expressed as percent of run time for infinite n (B = 30; D = 0.1)
      100
    g

    i_
    O
       90 --
       80 -
    O

   "CD
    t_
    O)
   B  70
    c
       60 -
o


I
CD
Q.
    CD
    E  50
   a:
       40
              90 percent
                                     10            15

                                  Number of contactors, n
                                                            20
25
 Figure 116  Run time as a function of number of contactors in parallel,
 expressed as percent of run time for infinite n (B = 30; D = 0.2)
                                 -120-

-------
4.6   Impact of Extrapolation on Integral Breakthrough Curve Prediction

Extrapolation of the integral TOC breakthrough curve may be necessary during ICR treatment
study data analysis because the SCA procedure is limited by the highest  TOC concentration
estimated from the direct integration of the  single contactor TOC breakthrough curve.   The
highest blended  effluent TOC concentration is typically 40 to 70 percent of the highest single
contactor TOC  concentration.  Although higher  single contactor TOC  concentrations  are
associated with formed DBF concentrations, these cannot be applied to the integral breakthrough
curve during application of the SCA method unless the TOC  integral  breakthrough curve is
extrapolated. The impact of breakthrough curve extrapolation was analyzed for two GAC runs
to determine whether the error caused by extrapolation was acceptable for ICR data analysis.

For two water sources (Waters 5 and 8), the RSSCTs were operated longer than required by the
ICR.  For Water 5, the run was extended 49 full-scale days after 70 percent TOC breakthrough
was reached, equivalent to a 21 percent extension of the required run time.  For  Water 8,  the
extension beyond the required run time was 69 days, equivalent to a 61 percent increase. During
the extended run time,  two additional single contactor GAC effluent samples were taken.  Two
additional blended effluent samples were also taken. The first 11 GAC effluent data points that
comprised  a normal ICR treatment study run reaching 70 percent TOC  breakthrough  were
modeled separately from  the  entire 13-point GAC effluent data set.   Based on the logistic
function model best-fit  of the first 11 data points, the water quality at the end  of the run (21  or 61
percent extrapolation) was predicted by extending the model fit.

Using the  DI method, the TOC integral breakthrough curve was calculated based on  the
truncated logistic function model fit and then extrapolated to predict water quality  at the end of
the entire run. The TOC integral breakthrough curve was also calculated based on the logistic
function  best-fit of the entire single contactor effluent data set.   For the other water quality
parameters,  the  SCA  prediction of blended contactor  effluent  water  quality based on  the
extrapolated data set and was compared to that based on the entire data set.  The impact of
extrapolation on the TOC integral breakthrough curve is of special interest, due to the application
of this extrapolation as a part  of the SCA procedure during data analysis of the ICR treatment
studies.

The impact of extrapolation on integral breakthrough curve predictions is shown in Figures 117
through 130. In these figures, the single contactor and integral breakthrough curves are plotted
against scaled operation time.  The data points used for the extrapolated single  contactor data set
are plotted with filled symbols, to differentiate these from the open symbols used for the first 11
data points.  For the  single contactor data, the solid best-fit line represents the logistic function
best-fit using the entire data set, while the dashed line represents the best-fit of the first 11 points
and extrapolation to the end of the run. Similarly, a dashed line also represents the extrapolated
integral breakthrough curve. The R2 value  is given for both best-fits of the single contactor data.

Figures  117 and  118  show the impact  of extrapolation  on the TOC single  contactor logistic
function  curve fit  and  the  TOC integral breakthrough curve predicted  by  the DI method  for
Waters 5 and 8, respectively.  At the end of both runs, the extrapolated logistic function curve fit
underpredicted single contactor effluent data.  Based on  the extrapolated logistic  function,  the
TOC  concentration at the end of the run  was 8 percent  lower than  that based on the logistic


                                      -121-

-------
function best-fit of all available single contactor data for Water 5.  For Water 8, the extrapolated
predicted TOC concentration at the end of the run was 12 percent lower.

This underprediction of the single contactor effluent data by the extrapolated logistic function
best-fit had  a smaller impact on  the TOC  integral breakthrough curve prediction.  The error
associated with extrapolation of the TOC integral breakthrough curve at the end of the run was 3
percent (0.05 mg/L) for Water 5.  For Water  8, the error was slightly larger,  8 percent (0.08
mg/L).

Application  of the SCA procedure to the extrapolated single contactor breakthrough curves for
all parameters yielded an average error in the  predicted blended contactor concentration at the
end of the run of 5 ± 3 percent  for Water 5.  Figures 119 through 124 show the impact of
extrapolation on  SCA prediction of the blended contactor effluent for UV254, SDS-TOX, SDS-
TTHM,  SDS-HAA5, SDS-HAA6, and SDS-HAA9.  For these DBF surrogates  and class sums,
the average  error in the SCA prediction due to extrapolation was also 5 percent, (the run time
was extrapolated by 21 percent) and in all cases, the extrapolated prediction was lower than the
prediction using the entire data set. Comparisons of all water quality parameters, including DBF
species, are given in Appendix H.

For Water 8, the  average error in the predicted blended contactor concentration at the end of the
run based on the SCA prediction of the extrapolated integral  breakthrough curve was 9 ± 5
percent. This higher mean observed error for Water 8 as compared to Water 5 can be attributed
in part to the larger percent extrapolation, 61 percent, performed on Water 8 as compared to that
for Water 5, 21  percent.  For UV254, SDS-TOX, SDS-TTHM,  SDS-HAA5, SDS-HAA6, and
SDS-HAA9  the  average error was 7 percent.   Figures  125 through 130 show the impact of
extrapolation on  the  SCA prediction of the blended  contactor concentration of these DBF
surrogates and class sums.  The blended contactor effluent concentration at the end of the  run
after extrapolation was lower than that based on the entire data set, except for SDS-HAA5.

An analysis  of Water 8 was also made using an extrapolation of 29 percent, which is similar to
the extrapolation applied to Water 5. For this shorter level of extrapolation, the mean error in the
integral breakthrough curve at the end of the run for all  parameters was 8 percent, only slightly
lower than that observed for an extrapolation of 61  percent.  For UV254, SDS-TOX, SDS-TTFDVI,
SDS-HAA5, SDS-HAA6, and SDS-HAA9 the average error was 6 percent.  Therefore,  for
Water 8, extending the extrapolation from 29 to 61 percent of the run time did not proportionally
increase the  mean observed error  in the predicted  integral breakthrough curve at the end of the
run.

Under the SCA  procedure, the error in the extrapolated integral  breakthrough curve for any
parameter (other than TOC) is dependent on the error in the extrapolated DI prediction of the
TOC  integral breakthrough curve as well as on the difference between the parameter single
contactor curve fits using the entire data set  and the truncated data set.  This difference typically
increases over the course of the run, and can be especially large in the extrapolated portion of the
curve. However, the error at the end of the single contactor curve may not impact the prediction
of the integral breakthrough curve by the SCA method because the maximum TOC concentration
at the end of the extrapolated integral breakthrough  curve  is still less than the maximum TOC
concentration at  the end of the single contactor curve.   This upper bound on  TOC limits the
                                      -122-

-------
concentrations of all other parameters to the initial and middle portions of the single contactor
curves when using the SCA procedure to predict integral breakthrough curves.

Conversely, the  DI method prediction of the integral breakthrough  curve relies on the entire
single contactor  breakthrough curve for each parameter.  Therefore, the error at the end of the
run due to extrapolation may have a larger impact on integral breakthrough curve predictions by
the DI method than predictions by the SCA procedure.  Appendix  I summarizes the integral
breakthrough curve predictions after extrapolation based on the DI method for all parameters.
The average difference at the end of the run between DI integral breakthrough curve predictions
with and without extrapolation was 3 percent  for Water 5,  which  is slightly lower than the
average obtained using the SCA method. For Water 8, the average error was slightly higher, 10
percent.

Based on the two runs examined, extrapolation up to 61 percent beyond available  experimental
data yielded  a mean 9 percent error in the predicted water quality based on the extrapolated
integral breakthrough curve as compared to that based on the complete data set.  For a shorter
extrapolation, 21 percent, the mean error was 5 percent.  An understanding and acceptance of the
error due to breakthrough curve extrapolation is  important because extrapolation may be used in
many cases during  the ICR treatment  study data analysis to gain additional  information from
GAC breakthrough data sets.   In  particular, data sets that do not  exceed  a given treatment
objective may be extrapolated by up to 50 percent in an attempt to determine a GAC run time for
the treatment objective. Based on  the two waters examined in this study, the error incurred by
extrapolation up  to 50 percent of the original run time should average less than  10 percent.
                                      -123-

-------
   3.0
   2.5 -
   2.0 -
 o
•^=  -I c
 CO  '•*•
 O>
 o
 o  1.0 H
O
   0.5 -
   0.0
   D   Single contactor effluent
   •   Extrapolation experimental data points
       Logistic function best fit - all data (RA2 = 0.984)
  - -  Extrapolated logistic function best fit  (RA2 = 0.979)
   O   Blended effluent
 	Dl prediction
 	Dl prediction - extrapolated
                                                                                TOO
                                                                         EBCT = 20 min.

                                                                         c0 = 3.08 mg/L
                   50
                   100        150         200         250

                        Scaled operation time (days)
300
350
 Figure 117  Impact of extrapolation on Dl prediction of the TOC integral
 breakthrough curve for Water 5
   2.0
   1.5 -
O)
'•   1.0 H
o>
o
c
o
O
   0.5 -
   0.0
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data  (RA2 = 0.974)
 -  - Extrapolated logistic function best fit (RA2 = 0.961)
 O   Blended effluent
	SCA prediction
	SCA prediction - extrapolated
                                                                               TOC
                                                                         EBCT = 7.2 min.

                                                                         c0 = 2.02 mg/L
                           50                  100

                                   Scaled operation time (days)
                                                         150
           200
 Figure 118  Impact of extrapolation on Dl prediction of the TOC integral
 breakthrough curve for Water 8
                                      -124-

-------
   0.035
   0.030 -
   0.025 -
o
o>
o
   0.020 -
   0.000 -i
        0
           D   Single contactor effluent
           •   Extrapolation experimental data points
          	Logistic function best fit - all data  (RA2 = 0.992)
           -  -  Extrapolated logistic function best fit (RA2 = 0.984)
           O   Blended effluent
          	SCA prediction
          	SCA prediction - extrapolated
                                                                            UV254
                                                                       EBCT = 20 min.

                                                                       c0 = 0.051 1/cm
               50         100         150        200        250

                               Scaled operation time (days)
300
350
Figure 119 Impact of extrapolation on SCA prediction of the UV254 integral
breakthrough curve for Water 5
   150
O
   125 -
   100 -
 o  75 -
 "CD
 CD
 £
 o
o
50 -
    25 -
           D   Single contactor effluent
           •   Extrapolation experimental data points
        	Logistic function best fit - all data (RA2 = 0.995)
        - - -  Extrapolated logistic function best fit (RA2 = 0.99)
           O   Blended effluent
        	SCA prediction
        	SCA prediction - extrapolated
                                                                            SDS-TOX
                                                                       EBCT = 20 min.

                                                                       C0 = 205 |jg/L Cl-
                  50         100         150        200        250

                                   Scaled operation time (days)
                                                                      300
           350
Figure 120 Impact of extrapolation on SCA prediction of the SDS-TOX
integral breakthrough curve for Water 5
                                     -125-

-------
    50
    40 -
    30 -
 o
 '-4—'
 CD
    20 -
 o
 O
    10 -
     o 4
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data  (RA2 = 0.979)
 -  - Extrapolated logistic function best fit (RA2 = 0.975)
 O   Blended effluent
	SCA prediction
	SCA prediction - extrapolated
                                                                          SDS-TTHM
                                                                       EBCT = 20 min.

                                                                       C0 = 58 ug/L
                  50
              100         150        200         250

                    Scaled operation time (days)
           300
           350
 Figure 121 Impact of extrapolation on SCA prediction of the SDS-TTHM
 integral  breakthrough curve for Water 5
   20
                                                                       EBCT = 20 mm.

                                                                       C0 = 28 ug/L
   15 -
o
'•CD  10
CD
O
c
o
O
    5 -
    Single contactor effluent
    Extrapolation experimental data points
   •Logistic function best fit - all data (RA2 = 0.965)
    Extrapolated logistic function best fit (RA2 = 0.959)
    Blended effluent
   -SCA prediction
   • SCA prediction - extrapolated
                                                                       SDS-HAA5
                            100
                         150         200

                   Scaled operation time (days)
250
300
350
 Figure 122 Impact of extrapolation on SCA prediction of the SDS-HAA5
 integral  breakthrough curve for Water 5
                                     -126-

-------
   25
   20 -
 o
'-4—'
 CD
 o
O
   15 -
   10 -
    5 -
    0 -i
           D   Single contactor effluent
           •   Extrapolation experimental data points
          	Logistic function best fit - all data (RA2 = 0.976)
           -  -  Extrapolated logistic function best fit (RA2 = 0.971)
           O   Blended effluent
          	SCA prediction
          	SCA prediction - extrapolated
                                                                       SDS-HAA6
                                                                      EBCT = 20 min.

                                                                      c0 =  34 ug/L
                 50
                         100         150        200         250

                               Scaled operation time (days)
300
350
Figure 123  Impact of extrapolation on SCA prediction of the SDS-HAA6
integral breakthrough curve for Water 5
   40
   35 -


   30 -


I 25-
c
o
'•CD 20 \
 CD
 *
 o
O
15 -


10 -


 5 -


 0 -
          D   Single contactor effluent
          •   Extrapolation experimental data points
         	Logistic function best fit - all data (RA2 = 0.99)
          -  -  Extrapolated logistic function best fit (RA2 = 0.988)
          O   Blended effluent
         	SCA prediction
         	SCA prediction - extrapolated        __^j»*» "  n
                                                                       SDS-HAA9
                                                                      EBCT = 20 min.

                                                                      C0 =  48 ug/L
                 50
                         100         150        200         250

                               Scaled operation time (days)
300
350
Figure 124  Impact of extrapolation on SCA prediction of the SDS-HAA9
integral breakthrough curve for Water 5
                                     -127-

-------
   0.025
   0.020 -
o
^ 0.015

CD
O

CD

o  0.010
   0.005 -
   0.000
  Single contactor effluent
  Extrapolation experimental data points
 •Logistic function best fit - all data (RA2 = 0.994)
  Extrapolated logistic function best fit  (RA2 =
  Blended effluent
 -SCA prediction
 • SCA prediction - extrapolated
                                                                            UV254
                                                                       EBCT = 7.2min.

                                                                       C0 = 0.033 1/cm
                            50                 100

                                   Scaled operation time (days)
                                                  150
                                                           200
 Figure 125 Impact of extrapolation on SCA prediction of the UV254 integral
 breakthrough curve for Water 8
    125
 O
 o
 '-4—'
 CD
 CD
 O
 C
 O
 O
    100 -
     75 H
     50 -
     25 -
      o -b
 Single contactor effluent
 Extrapolation experimental data points
•Logistic function best fit - all data  (RA2 = 0.99)
 Extrapolated logistic function best fit (RA2 = 0.98)_
 Blended effluent
-SCA prediction
 SCA prediction - extrapolated
                                                                    SDS-TOX
                                                                       EBCT = 7.2 min.

                                                                       c0= 156 ug/LCI-
50                  100

        Scaled operation time (days)
                                                                   150
                                                                      200
 Figure 126 Impact of extrapolation on SCA prediction of the SDS-TOX
 integral  breakthrough curve for Water 8
                                     -128-

-------
   40
   30 -
             Single contactor effluent
             Extrapolation experimental data points
             Logistic function best fit - all data  (RA2 = 0.98)
             Extrapolated logistic function best fit (RA2 = 0.966)
             Blended effluent
             SCA prediction
             SCA prediction - extrapolated
                                                                          SDS-TTHM
                                                                      EBCT = 7.2min.

                                                                      c0 =  42 ug/L
                          50                 100

                                  Scaled operation time (days)
                                                               150
                    200
Figure 127 Impact of extrapolation on SCA prediction of the SDS-TTHM
integral breakthrough curve for Water 8
  16
14 -


12 -


10 -
               Single contactor effluent
               Extrapolation experimental data points
              •Logistic function best fit - all data  (RA2 = 0.955)
               Extrapolated logistic function best fit (RA2 = 0.911)
               Blended effluent
              -SCA prediction
              • SCA prediction - extrapolated
                                                                    SDS-HAA5
                                                                      EBCT = 7.2min.

                                                                      c0 = 23 ug/L
                         50
                                          100

                               Scaled operation time (days)
150
200
Figure 128  Impact of extrapolation on SCA prediction of the SDS-HAA5
integral breakthrough curve for Water 8
                                    -129-

-------
   20
   15 -
  D  Single contactor effluent
  •  Extrapolation experimental data points
	Logistic function best fit - all data (RA2 = 0.956)
- - - Extrapolated logistic function best fit (RA2 = 0.923)
  O  Blended effluent
	SCA prediction
	SCA prediction - extrapolated
05
o
'•   10
o>
o
c
o
O
    5 H
                                                                         SDS-HAA6
                          50                   100

                                   Scaled operation time (days)
                                                           150
                     200
 Figure 129 Impact of extrapolation on SCA prediction of the SDS-HAA6
 integral breakthrough curve for Water 8
    20
    15 -
 o
 '•S3  10
 o>
 o
 c
 o
 O
     5 -
   D   Single contactor effluent
   •   Extrapolation experimental data points
  	Logistic function best fit - all data  (RA2 = 0.952)
   -  -  Extrapolated logistic function best fit (RA2 = 0.924)
   O   Blended effluent
  	SCA prediction
  	SCA prediction - extrapolated
                           50
                                                                        SDS-HAA9
                                                                        EBCT = 7.2min.

                                                                        c0 = 30 ug/L
                                      100

                          Scaled operation time (days)
150
200
 Figure 130  Impact of extrapolation on SCA prediction of the SDS-HAA9
 integral breakthrough curve for Water 8
                                       -130-

-------
5   Summary and Conclusions

The design of this study incorporated two main goals.  The primary objectives were to evaluate
the ability  of the logistic  function to model single  contactor breakthrough curve data  and to
evaluate  the  success  and limitations of  predictive models  used to determine  the integral
breakthrough curve, a relationship between single  contactor  run  time and blended contactor
water quality. The  secondary objective of this study was to evaluate the applicability of these
models and predictive methods in the context of the ICR GAC treatment study data analysis.  A
large amount of data will be analyzed: the 62 GAC treatment studies performed will generate
8,000 to 9,000 individual breakthrough curves. Experimental verification was performed on data
from eight bench-scale GAC runs with varying  water sources, pretreatments,  DBF precursor
concentrations,  bromide concentrations, and SDS  chlorination conditions.   The GAC  runs
utilized  the rapid small-scale  column test (RSSCT)  and  were performed according to  ICR
requirements.

A primary requirement for the ICR GAC treatment study data analysis procedure is to  model the
single  contactor  effluent  breakthrough data,  for  all parameters analyzed, including DBF
surrogates,  DBF  class sums, and DBF  species.   A model used  to describe single contactor
effluent experimental data  is needed for several reasons.  From a data management perspective,
best-fit curve parameters that adequately describe experimental data are less memory intensive
than storing the entire experimental data set.  A best-fit curve also facilitates interpolation and
extrapolation to estimate run times for given treatment objectives.  Use of a best-fit model curve
also provides an estimate of the scatter in the data through the coefficient of determination, and
the model minimizes the impact of this scatter on run time estimates. Finally, a function that
describes the single contactor experimental data set is a prerequisite for determining the integral
breakthrough curve, a curve that relates single contactor  run time to blended  effluent water
quality under the  assumption that contactors are operated in parallel-staggered mode.  Run time
estimates generated by the integral breakthrough curve are more applicable to full-scale GAC
operation.

Previous researchers have used various forms of the logistic function to predict and model GAC
breakthrough curve data, and the logistic function was found  to be an adequate model  in this
study, where  it was applied to data from eight GAC  runs (160  potential curve fits, but only 126
were performed because concentrations  in the GAC effluent  were below  the MRL for some
parameters).  However, poor curve fits occurred for some parameters, especially very sharp "S"
shaped breakthrough curves, and "peak" curves (breakthrough curves for brominated species that
increased and then decreased in concentration over the course of the run due to changes in the
bromide to TOC ratio). To improve the performance of the  logistic function for modeling single
contactor data,  three enhancements were made,  yielding the  step, step-lag, and  step-lag-peak
logistic function models.  The step logistic function  model  is applicable for GAC breakthrough
curves with relatively high levels of immediate  breakthrough.  The  step-lag logistic function
model incorporates an initial phase where the model  output  is set to zero, to better fit DBF
breakthrough curves with relatively long initial intervals of effluent concentrations reported as
BMRL,  prior to breakthrough.   The step-lag-peak logistic function model incorporates a linear
decay term and is applied to "peak" breakthrough curves, which sometimes occur for brominated
DBF species.
                                      -131-

-------
Curve  fitting  involved  determining which  model  was  applicable,  and  applying  it to the
breakthrough curve for each parameter.  These enhanced  forms of the logistic function model
were able to successfully  fit single contactor breakthrough curve data for all parameters.   A
method was also developed to detect outlier data points that limited the  influence of deviant
observations on the parameter estimates.  In general, the application of the three enhanced forms
of the logistic function curve to single contactor effluent breakthrough  data was successful, with
a mean R2 of 0.974.

Two predictive approaches for determining the integral breakthrough curve were examined: the
direct  integration (DI)  and the surrogate  correlation  approach (SCA).  The  results of these
procedures were compared to  experimental  results.  Experimental  data were obtained  by
collecting the entire effluent from  each of eight bench-scale GAC  experiments in separate
reservoirs and sampling from these reservoirs over time. This experimental procedure simulates
the  integral breakthrough curve  for an infinite number of  contactors operated  in parallel-
staggered mode.

The DI procedure applied the average value function  to the logistic function model fit of the
experimental single contactor data,  and  is based on the assumption of an infinite number  of
contactors operated in parallel-staggered mode with regular GAC replacement frequencies.  For
DBF surrogates (TOC, UV254, and  SDS-TOX), the DI procedure yielded excellent results  in
comparison to experimental  data.  For class sums,  such  as SDS-TTHM and SDS-HAA5,
predictions were usually  adequate.   However,  for individual DBF compounds,  especially
brominated  species,  the  DI  approach  did not always result in  accurate predictions.  The
inaccuracy of the DI approach  for the prediction of some  DBF species  is problematic since
individual DBFs of potential health concern will be considered during analysis of the treatment
study data.

The  SCA procedure was  developed to  reduce the computational requirements of estimating
blended contactor run times for any given regulatory treatment objective evaluated during the
ICR  GAC treatment study data analysis.   The  SCA procedure relies on the DI method  to
establish an integral breakthrough curve for TOC  only.   The DI method yielded excellent
predictions of the TOC integral breakthrough curve for the waters examined in this study, with a
mean error quantified by the residual sum of squares (RSS)  of 0.055 mg/L, and a mean bias  of
+0.028 mg/L.  Once the TOC integral breakthrough curve is obtained, data points on both the
single  contactor and integral breakthrough curves at a constant TOC concentration are mapped,
and all other water quality parameters associated with the single contactor effluent data set at that
TOC concentration are applied to  the blended contactor integral breakthrough curve.  The SCA
procedure requires that single contactor effluent data be adequately modeled using the logistic
function models.   For eight GAC runs  and eight  water  sources, this  study showed that the
enhanced logistic function models successfully simulated a wide variety of GAC breakthrough
curve profiles.  The SCA procedure inherently relies on the assumption  that the  relationship
between TOC and the other organic precursors, SDS DBF  class sums, and SDS DBF species
established in the single contactor  effluent is maintained in the blended contactor effluent.  This
study found that this assumption is valid for the eight waters examined.  In addition, the
correlation between TOC and bromine incorporation factors for THMs and HAAs was shown  to
be consistent between the single contactor effluent and experimental blended effluent.
                                      -132-

-------
Application of the SCA procedure to the experimental GAC breakthrough curves for eight water
sources showed that the overall  accuracy of the model was  equivalent to the DI  method for
predicting the integral breakthrough curve.  This analysis was performed by calculating the RSS
and  bias  between  model predictions  and experimental data.   The cumulative  frequency
distribution of the normalized RSS showed that across all  waters and all analytes, the prediction
error for the two models was equivalent.  Both models were  biased negative, indicating  a
tendency to underpredict the  experimental data. The SCA model had a slightly higher negative
bias than did the DI model.  Based on these results, application of the computationally-simpler
SCA procedure to ICR treatment study data is recommended.

A comparative analysis of the two predictive models was also performed for each individual
analyte. The SCA method was more accurate than the DI method when applied to SDS-TTHM,
and was equally  accurate when applied to SDS-HAA5, SDS-HAA6, and SDS-HAA9.  For the
two waters that  examined SDS-TOX, the DI method was a better predictor of experimental
results.  For the predominant  THM and HAA species (for which comparisons could be made for
six or more  of the eight runs), the SCA method outperformed the DI method when applied to
brominated DBF species, with the exception of SDS-DBCM.   The DI method generated more
accurate  predictions  of the  non-brominated DBF species (SDS-CF,  SDS-DCAA,  and  SDS-
TCAA).  The impact of changing bromide to TOC ratio on the shape of the breakthrough curve
for brominated DBF species, yielding peak curves or very sharp curves followed by a plateau,
typically resulted in  underpredictions by the DI method of the  observed data.  The SCA method
was able to better predict the integral breakthrough curves for the brominated species because it
relied on the relationship  between  TOC and DBF  formation in the  single contactor effluent,
which inherently accounts for changing bromide to TOC ratios.

An important limitation of the DI and SCA procedures is that they rely on the assumption of an
infinite  number  of  contactors operated in  parallel-staggered  mode.   Previous  work  has
maintained that for  10 or  more contactors in  operation,  actual blended effluent run times are
within  10 percent of those estimated based on the infinite contactor assumption.  This study
found that the error incurred when  applying run time estimates based on the infinite contactor
assumption to run times for finite numbers of contactors is impacted by the number of contactors
and  the  magnitude  of the  treatment  objective  examined  in  relation to  the  asymptotic
concentration approached  by the single contactor breakthrough curve.  Based on  the logistic
function  model,  the  infinite  contactor assumption  will yield estimated run  times within 10
percent of actual run times for 13 or more contactors operated in parallel-staggered  mode.  For
10 contactors on-line, the infinite contactor assumption will yield run time estimates within 12
percent of the actual run times.  In  all cases, run time estimates based on the infinite contactor
assumption are longer than those for a finite number of contactors, thus providing  a best case
scenario for GAC performance.

The  applicability of the infinite contactor assumption  in this model  to  finite numbers  of
contactors is especially important for small plants operating fewer than 10 contactors on-line.
Also evident was that the largest incremental benefit afforded by operating contactors in parallel-
staggered mode occurs when  two contactors are operated as compared to a single contactor. The
benefit realized by adding  an additional contactor decreases as the number of contactors on-line
increases.  The relationship developed between integral breakthrough curve run time estimates
based on an infinite number of contactors and that based on a finite number of contactors can be
                                      -133-

-------
applied to  ICR  treatment study data to estimate  the  performance of large and small GAC
systems.

During the total  run time for any given  single contactor TOC breakthrough curve, the blended
contactor integral  breakthrough curve  determined by the DI  approach will typically reach
concentrations ranging from 40 to 70 percent of those measured at the end of the single contactor
GAC run.  Therefore, when using the SCA procedure, the latter portion of the single contactor
breakthrough curve (consisting of higher surrogate  and formed DBF levels) will not usually be
applied to the blended effluent data set.  This may  be an issue during ICR treatment study data
analysis if a significant number of integral breakthrough curves estimated by the SCA procedure
do not reach regulatory treatment objectives.  For these runs, extrapolation of the TOC integral
breakthrough curve would increase the usefulness of the entire data set.

For  two  waters,  experimental evaluation  of the  sensitivity  predicted  water   quality  to
extrapolation of the integral  breakthrough curve prediction was performed.  For  one water,
extrapolation by  21 percent yielded a 3 percent error in the predicted integral TOC concentration
at the end of the  run and a mean 5 percent error at the end of the run for all analytes predicted by
the SCA method.   For a second water, which  was extrapolated by 61  percent, the error in
predicted TOC concentration was 8 percent at the end of the run, and the mean error at the end of
the extrapolated run was 9 percent for all analytes.

An understanding and  acceptance  of the  error due  to  breakthrough  curve extrapolation is
important because  extrapolation may be used in many cases during the ICR treatment study data
analysis to gain additional information from GAC breakthrough data sets.  In particular, data sets
that do not  exceed a given treatment objective may  be extrapolated by a reasonable extent (up to
50 percent) in an attempt to determine a GAC run time for the treatment objective. Based on the
two waters examined in this study,  the error incurred by  extrapolation up to 50 percent of the
original run time should average less than  10 percent.  This error may be  acceptable for ICR
treatment study data analysis, given the benefit afforded by the extrapolation.

The models used to simulate GAC operation of multiple  contactors in parallel-staggered mode
rely on the assumption that the GAC is replaced at  regular intervals, so that the service times of
all contactors are equal.  They also assume that the breakthrough curve profiles of all single
contactors  are identical.  In a full-scale plant,  this idealized situation  will  rarely occur.
Variability  in source water quality will impact the run time of the contactors, depending on when
they are placed in  service and the concentration of DBF precursors in the GAC influent during
their  service life.   Variability in distribution system conditions, especially temperature, will
impact the contactor  service life,  as  formed  DBF levels increase with higher temperatures.
Furthermore, for a plant that operates a fixed number of contactors, water demand changes
during the year may directly  impact the EBCT of each contactor.  A contactor that is placed on-
line at the beginning of the summer high  demand months may be operated under a shorter EBCT
as compared to a contactor placed on-line during the winter. Furthermore, it is less desirable to
remove contactors from service to replace GAC during high demand periods.  Another approach
is to  increase the  number of contactors on-line as water  demand increases.  These full-scale
issues should be considered when interpreting the results of the ICR  treatment  study data
analysis.
                                      -134-

-------
6   References

Clark, R.M.  1987.  Modeling TOC Removal by GAC:  The General Logistic Function. Jour.
      AWWA. (79:1:33).

Chowdhury,  Z.K., G. Solarik, D.M.  Owen, S.M. Hooper,  and R.S.  Summers.  1996.  NOM
      Removal by GAC Adsorption:  Implications of Blending. In Proc. of the AWWA Annual
      Conference, Toronto, Ontario, Canada.

Clark, R.M., J.M. Symons, and J.C. Ireland. Evaluating Field Scale GAC Systems for Drinking
      Water. Jour. Environ. Engr. (112:4:744).

Crittenden, J.C., P.S. Reddy, D.W Hand and H. Arora.  1989. Prediction of GAC Performance
      Using Rapid Small-Scale Column Tests. AWWARF/AWWA.

Crittenden, J.C.,  P.S.  Reddy, H.  Arora,  J. Trynoski,  D.W.  Hand, D.L. Perram,  and R.S.
      Summers.  1991.  Predicting GAC Performance with Rapid Small-Scale Column Tests.
      Jour. AWWA. (83:1:77).

Gould, J.P., L.E. Fitchorn, and E. Urheim.  1983.  Formation of Brominated  Trihalomethanes:
      Extent and Kinetics.  Water Chlorination: Environmental Impact and Health Effects.
      Vol. 4 (R.L. Jolley et al., eds.), Ann Arbor Sci. Publ., Ann Arbor, Michigan.

Hooper, S.M., R.S. Summers, G. Solarik, and S. Hong.  1996.  GAC Performance for DBF
      Control:  Effect of Influent Concentration,  Seasonal Variation, and  Pretreatment.  In
      Proc. of the AWWA Annual Conference, Toronto, Ontario, Canada.

Littell, R.C., G.A.  Milliken, W.W.  Stroup, and R.D. Wolfmger. 1996. SAS System for Mixed
      Models. Cary, NC (SAS Institute Inc.)

Oulman, C.S.  1980.  The  Logistic Curve as a Model for Carbon Bed Design.  Jour. AWWA
      (72:1:50).

Roberts, P.V. and R.S. Summers. Performance of Granular Activated Carbon  for Total Organic
      Carbon Removal.  Jour. AWWA. (74:2:113).

Shukairy, H.M., Miltner,  R.J., and Summers, R.S.  1994.  Bromide's Effect on DBF Formation,
      Speciation and Control: Part 1, Ozonation. Jour. AWWA. (86:6:72).

Snoeyink,  V.L.  1990. Adsorption of Organic Compounds.  Water Quality and Treatment: A
      Handbook for Community Supplies.  4th ed. (F.W. Pontius, ed.), AWWA and McGraw-
      Hill, Inc.

Sontheimer,  H., J.C. Crittenden, and R.S. Summers.   1988.  Activated Carbon  for Water
      Treatment.    2nd ed.,  DVGW-Forschungsstelle,  Universitat  Karlsruhe,  Karlsruhe,
      Germany.

Standard Methods for the Examination of Water and Wastewater.  1998. APHA, AWWA, and
      WEF. Washington D.C. (20th ed.)
                                     -135-

-------
Summers, R.S., D.M. Owen, Z.K. Chowdhury, S.M. Hooper, G. Solarik, and K. Gray.  1998.
      Removal of DBF Precursors by GAC Adsorption.  AWWARF and AWWA, Denver,
      Colorado.

Summers, R.S.,  S.M. Hooper, G. Solarik, D.M. Owen, and S. Hong.   1995.  Bench-Scale
      Evaluation of GAC for NOM Control. Jour. AWWA.  (87:8:69).

Summers, R.S.,  S. Hong, S.M. Hooper, and G. Solarik.  1994. Adsorption of Natural Organic
      Matter and  Disinfection  By-Product  Precursors.   In  Proc.  of the AWWA  Annual
      Conference, New York, NY.

Summers, R.S.,  M.A. Benz, H.M. Shukairy, and L. Cummings.  1993.  Effect of Separation
      Processes on the Formation of Brominated THMs.  Jour. AWWA. (85:1:88).

Summers, R.S., L. Cummings, J.  DeMarco, DJ. Hartman, D.H. Metz, E. Howe, B. MacLeod,
      and M. Simpson.  1992. Standardized Protocol for the Evaluation of GAC. AWWARF
      and AWWA, Denver, Colorado.

USEPA. 1999. GAC Base Analysis Approach for the ICR Treatment Study Database.

USEPA. 1996. ICR Manual for Bench-and Pilot-Scale Treatment Studies. EPA 814-B-96-003.
      Technical Support Division, Office of Ground Water and Drinking Water, Cincinnati,
      Ohio.

Westrick, JJ. and J.M. Cohen.  1976.  Comparative Effects of Chemical Pretreatment on Carbon
      Adsorption.
                                    -136-

-------
Appendix A: Breakthrough Curves Corrected for Impact of Sampling
                             -137-

-------
   4.0
   3.5 -
   3.0 -
   2.5 -
O)



c
o
 en
 •£ 2.0 H

 o>
 o

 § 1-5 H
 O
 O
    1.0 -




    0.5 -




    0.0
        EBCT = 15 min

        C0 = 4.54 mg/L
       0
                                                 Single contactor effluent

                                                 Blended effluent - experimental

                                                 Blended effluent - adjusted for sampling
                        25                50


                               Scaled operation time (days)
75
100
Figure A-1  Comparison of blended effluent TOC adjusted for experimental

sampling to blended effluent experimental results for Water 1
   2.0
   1.5 -
O)


c
o

1°
~  1.0


§
o
o

O
O
   0.5 -
   0.0
         EBCT = 20 min

         c0 = 2.6 mg/L
                                                  Single contactor effluent

                                                  Blended effluent - experimental

                                                  Blended effluent - adjusted for sampling
                     50
                                  100            150


                               Scaled operation time (days)
   200
250
Figure A-2 Comparison of blended effluent TOC adjusted for experimental

sampling to blended effluent experimental results for Water 2
                                      -138-

-------
   2.0
   1.5 -
co

£  1.0

CD
O
c
O
O
O
O
   0.5 -
   0.0
         EBCT = 20 min.

         C0 = 2.35 mg/L
                                                  Single contactor effluent

                                                  Blended effluent - experimental

                                                  Blended effluent - adjusted for sampling
                  50
100         150          200


 Scaled operation time (days)
250
300
Figure A-3 Comparison of blended effluent TOC adjusted for experimental

sampling to blended effluent experimental results for Water 3
   2.5
   2.0 -
   1.5 -
CD
O
O
O
   1.0 -
   0.5 -
   0.0
         EBCT = 20 min

         c0 = 2.98 mg/L
                                                    Single contactor effluent

                                                    Blended effluent - experimental

                                                    Blended effluent - adjusted for sampling
                        50                100

                                Scaled operation time (days)
                               150
            200
Figure A-4 Comparison of blended effluent TOC adjusted for experimental

sampling to blended effluent experimental results for Water 4
                                      -139-

-------
       3.0
       2.5 -
       2'0
    ~  1.5 -
    o>
    o
    c
    o
    o
    O
    O
       0.5 -
       0.0
             EBCT = 20 min.

             C0 = 3.08 mg/L
                                                      Single contactor effluent
                                                      Blended effluent - experimental
                                                      Blended effluent - adjusted for sampling
                    50
100        150        200        250

     Scaled operation time (days)
  300
350
    Figure A-5  Comparison of blended effluent TOC adjusted for experimental
    sampling to blended effluent experimental results for Water 5
      2.5
      2.0 -
    §  1.5 H
    CD
    O
    §  1.0 H
    o
    O
    O
       0.5 -
       0.0
             EBCT = 20 min.

            c0 = 2.64 mg/L
                                                       Single contactor effluent
                                                       Blended effluent - experimental
                                                       Blended effluent - adjusted for sampling
                      50
   100          150          200

     Scaled operation time (days)
250
300
Figure A-6  Comparison of blended effluent TOC adjusted for experimental
sampling to blended effluent experimental results for Water 6
                                         -140-

-------
   4.5

   4.0 -

   3.5 -
O)
E 3.0 -
o>
o
o
o
O
O
1.5 -

1.0 -

0.5 -

0.0
      EBCT = 20 min
      c0 = 5.58 mg/L
      0
                                                  Single contactor effluent
                                                  Blended effluent - experimental
                                                  Blended effluent - adjusted for sampling
               25
50          75          100
 Scaled operation time (days)
125
150
Figure A-7 Comparison of blended effluent TOC adjusted for experimental
sampling to blended effluent experimental results for Water?
   2.0
   1.5 -
O)
E
   1.0 -
0>
o
o
o
O
O
I-  0.5 -
   0.0
         EBCT = 7.2 min
         c0 = 2.02 mg/L
                                                    Single contactor effluent
                                                    Blended effluent - experimental
                                                    Blended effluent - adjusted for sampling
                        50                 100
                                Scaled operation time (days)
                                                          150
                                                 200
Figure A-8 Comparison of blended effluent TOC adjusted for experimental
sampling to blended effluent experimental results for Water 8
                                      -141-

-------
This page intentionally left blank.
             -142-

-------
Appendix B: SAS Code
SAS code to fit the step-lag-peak logistic model and perform outlier
adjustment.
  This program fits data with the following functions:

  1. Cs(t) = 0                    for        t <= tO
  2. Cs(t) = AO+A/(l+B*exp(-D*t))  for   tO < t <= Tmax
  3. Cs(t) = Cmax+S*(t-Tmax)      for Tmax < t

  where tO = (1/D)*LN(-AO*B/(AO+A)),  Cmax = Cs(Tmax),
  Tmax is the runtime at which effluent reaches its maximum.
  Bounds: -Cmax/2 <= AO <= Cmax/2, 0 < A <= 1.5*Cmax,  0
-------
     der.A =  (l+BEXP+A*BEXP*D*dA)/BEXP2;
     der.B =-A*BEXP*(l/B-D*dB)/BEXP2;
     der.D = A*BEXP*(tO+D*dD)/BEXP2;
     end;
  else if time < &xTmax then  do;
     BEXP = B*exp(-D*time);
     model &Analyte = AO+A/(1+BEXP);
     der.AO= 1;
     der.A = I/ (1 + BEXP) ;
     der.B =-A*BEXP/(B*(1+BEXP)**2);
     der.D = A*BEXP*time/(1 + BEXP)**2 ;
     end;
  else do;
     BEXPmax = B*exp(-D*&xTmax);
     model &Analyte = AO+A/(1+BEXPmax)+S*(time-&xTmax);
     der.AO= 1;
     der.A = I/(1+BEXPmax);
     der.B =-A*BEXPmax/(B*(1+BEXPmax)**2);
     der.D = A*BEXPmax*time/(1+BEXPmax)**2;
     der.S = time-ScxTmax;
     end;
  output out=outp p=pred u95=u95  195=195;
*proc print;
data outp; set outp;
  winsor=&Analyte;
  delta=(U95-195)/3 ;
  low=0;
  high=0;
  if &Analyte ne .  and &Analytepred+delta  then do;
          outlier=&Analyte; winsor=pred+delta;  high=l;  end;
  keep time &Analyte winsor outlier  low high;
proc means data=outp noprint;
  var low high;
  output out=outout sum=outlow  outhigh;
data outout; set outout;
  keep outlow outhigh;
*proc print;

PROC NLIN DATA=outp outest=outest NOPRINT;
  parms A0=0 A=&xCmax B=10 to 50  by  10 D=0.05  to 0.15  by 0.02 S=0;
  bounds -&xCmax05 <= AO <= &xCmax05,  0< A  <=  &xCmax!5,  0
-------
     der.B =-A*BEXP*(l/B-D*dB)/BEXP2;
     der.D = A*BEXP*(tO+D*dD)/BEXP2;
     end;
  else if time < &xTmax then do;
     BEXP = B*exp(-D*time);
     model &Analyte = AO+A/(1+BEXP);
     der.AO= 1;
     der.A = I/ (1 + BEXP) ;
     der.B =-A*BEXP/(B*(1 + BEXP)**2 );
     der.D = A*BEXP*time/(1+BEXP)**2 ,•
     end;
  else do;
     BEXPmax = B*exp(-D*&xTmax);
     model &Analyte = AO+A/(1+BEXPmax)+S*(time-&xTmax)
     der.AO= 1;
     der.A = I/(1+BEXPmax);
     der.B =-A*BEXPmax/(B*(1 + BEXPmax)**2) ;
     der.D = A*BEXPmax*time/(1+BEXPmax)**2;
     der.S = time-ScxTmax;
     end;
  output out=outq p=pred r=redis u95m=u95m  195m=195m;

*proc print data=outest;
proc reg data=outq outest=r2 adjrsq noprint;
  model winsor = pred / noint;
data r2; set r2;
  keep _rsq_ _adjrsq_;

data outq; set outq;
  keep time pred redis 195m u95m;
data &Analyte; merge outp outq; by time;
  if &Analyte=. then missing=pred;
  if outlier=. then winsor=.;
  if 195m<0 then 195m=0;
RUN;

data std;  set outest;
  if _type_='COVB1;
  retain std_aO std_a std_b std_d  std_s;
  if _name_='AO'  then std_aO=sqrt(aO);
  if _name_='A' then std_a=sqrt(a);
  if _name_='B' then std_b=sqrt(b);
  if _name_='D' then std_d=sqrt(d);
  if _name_='S' then std_s=sqrt(s);
  if _name_='S';
  keep std_aO std_a std_b std_d std_s;

data estimate; set outest;
  if _type_='ITER'  or _type_='FINAL';
proc sort data=estimate; by _type_;
data estimate; set estimate; by _type_;
  if last._type_;
proc means data=estimate noprint;
  var _iter_ _sse_ aO a b d s;
  output out=outest mean=iter sse  aO  a b  d  s;

data &FitInfo; merge outest std r2 outout;


                                      -145-

-------
  analyte="&Analyte";
  if _freq_=l then converge='NO ';
              else converge='YES';
  if std_aO=0 or std_a=0 or std_b=0 or std_d=0 or std_s=0 then
singular='YES';
                                              else singular='NO  ';
  if AO+A/(1+B)  >= 0 then tO=0; else t0=(1/D)*log(-AO*B/(AO+A));
  keep analyte aO a b d s to std_aO std_a std_b std_d std_s
       _rsq_ converge singular outlow outhigh;
run;
%mend;

/*
%FitF012(RawData=Raw,  Baslnfo=Bas!nfo, Analyte=BDCM, Fitlnfo=Fit!nfo)
proc print data=Fit!nfo;
run;
*/
                                     -146-

-------
Appendix C: Full- and Bench-Scale Pretreatment Schematics
                             -147-

-------
 Full-Scale
• ^^ ^^ ^^ ^^ ^^

 Bench-Scale
                                 Influent
                                                       Biscayne Aquifer
                           Full-Scale Softening
Lime
Activated alumina
                              Bench-Scale
                           Cartridge Filtration
                           Bench-Scale Mixing
Sulfuric acid
                            Bench-Scale GAC
                               Adsorption
Figure C-l Full- and bench-scale pretreatment schematic for Water 1
                                      -148-

-------
 Full-Scale

 ^^ ^^ ^^ ^^ ^^ •

 Bench-Scale
                                 Influent
                                 Influent
                                Claricone
                          (softening/clarification)
 Fox River

   PAC
   Potassium permanganate
•   Alum
   Cationic polymer

  Deep/shallow wells
                              Bench-Scale
                           Cartridge Filtration
                           Bench-Scale Mixing
                           Bench-Scale GAC
                              Adsorption
 Sulfuric acid or
 sodium hydroxide
Figure C-2 Full- and bench-scale pretreatment schematic for Water 2
                                      -149-

-------
   Full-Scale
   • ^H ^H ^H ^H ^H

   Bench-Scale
                                    Influent
Full-Scale Rapid Mix
i
r
                               Full-Scale Primary
                               Presedimentation
                              Full-Scale Secondary
                                Presedimentation
                              Full-Scale Rapid Mix
                               Full-Scale Primary
                                 Sedimentation
                              Full-Scale Secondary
                                 Sedimentation
                                                          Kansas River
                                                         Organic polymer
                                                          Organic polymer
Alum
CaO
SOC
                                  Bench-Scale
                               Cartridge Filtration
                               Bench-Scale Mixing
                                Bench-Scale GAC
                                   Adsorption
 Sulfuric acid
Figure C-3 Full- and bench-scale pretreatment schematic for Water 3
                                    -150-

-------
      Full-Scale
      ^^ ^^ ^^ ^^ ^^ •

      Bench-Scale
                                   Influent
                             Full-Scale Rapid Mix
                               Full-Scale Solids
                             Contact Clarification
-I-
                    Mississippi River

                    Potassium permanganate
                    Cationic polymer

                    Polyaluminum sulfate
                                                          Anionic Emulsion Polymer
                                 Bench-Scale
                              Cartridge Filtration
                               Bench-Scale GAC
                                  Adsorption
Figure C-4 Full- and bench-scale pretreatment schematic for Water 4
                                      -151-

-------
                                     Influent
                               Full-Scale Rapid Mix
 Lake Dixon, Lake
 Wohlford and/or
 SDCWA
Alum
Organic polymer
Potassium permanganate
                                    Full-Scale
                                   Flocculation
       Full-Scale
                                    Full-Scale
                                  Sedimentation
       Bench-Scale
                                   Bench-Scale
                                Cartridge Filtration
                                Bench-Scale GAC
                                   Adsorption
Figure C-5 Full- and bench-scale pretreatment schematic for Water 5
                                     -152-

-------
       Full-Scale
                                      Influent
                               Full-Scale Rapid Mix
                                     Full-Scale
                                   Flocculation
                                     Full-Scale
                                  Sedimentation
 Edisto River or Bushy
 Park Reservoir
Alum
 Organic polymer
       Bench-Scale
                                   Bench-Scale
                                Cartridge Filtration
                                 Bench-Scale GAC
                                    Adsorption
Figure C-6 Full- and bench-scale pretreatment schematic for Water 6
                                     -153-

-------
    Full-Scale
                                      Influent
                                   Intake Tower
                                          <-
                                     Flash Mix
                                   Flocculation
                                  Sedimentation
                                     Filtration
                           Sweetwater
                           Reservoir
                           Chlorine (shut off when
                           sampling filter effluent)

                          •  Potassium permanganate
                          •  Ferric chloride

                            Organic polymer
                          .  Ammonia
                                                            PAC
    Bench-Scale
   Bench-Scale
Cartridge Filtration
                                 Bench-Scale GAC
                                    Adsorption
Figure C-7 Full- and bench-scale pretreatment schematic for Water 7
                                     -154-

-------
                                    Influent
Full-Scale Rapid Mix
i
r
 Lake Daniel Reservoir
 and Lake Brandt
 Reservoir
- Potassium permanganate

 Alum
                                   Full-Scale
                                  Flocculation
     Full-Scale
    • ^^ ^^ ^^ ^^ ^^

     Bench-Scale
                                   Full-Scale
                                 Sedimentation
                                  Bench-Scale
                               Cartridge Filtration
                               Bench-Scale GAC
                                  Adsorption
Figure C-8 Full- and bench-scale pretreatment schematic for Water 8
                                     -155-

-------
This page intentionally left blank.
             -156-

-------
Appendix D: Single Contactor and Blended Effluent DBF Surrogate
and Formed DBF Correlations
                             -157-

-------
0.07-1

0.06-
0.05 -
f
^ 0.04 -
—
-*•
CN 0.03 -
0.02 -


0.01 -
n nn -

D Single contactor
• Blended effluent n
D
D

D

n
B °
• D
D
j"D
• 'D
                                                                                             75 -i
                                                                                          =d  50-
                                                                                          D)
                        0.0    0.5     1.0     1.5     2.0     2.5     3.0     3.5

                                                 TOC (mg/L)
                                                                                 4.0
                                                                         0.0     0.5     1.0     1.5     2.0     2.5     3.0

                                                                                                   TOC (mg/L)
                                                                                                                             3.5     4.0
oo
                   30 -
                   20-
                I
                CO
                Q
                CO
                                                                                             150 n
                                                                                          O  100 -
                                                                    X
                                                                    O

                                                                    CO
                                                                    Q
                                                                    CO
                                                                                               0-K3-,
0.0     0.5     1.0      1.5     2.0     2.5     3.0     3.5     4.0

                          TOC (mg/L)
                                                                                                0.0     0.5     1.0     1.5    2.0     2.5     3.0     3.5     4.0

                                                                                                                        TOC (mg/L)
             Figure D-1  Correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 1

-------
       0.04 n
       0.03-
    8
       0.02 -
       0.01 -
       0.00
                                                                              100 i
               D Single contactor
               • Blended effluent
                                                                               75-
                                                                           d
                                                                           D)
                                                                               50-

           0.0            0.5
                1.0
            TOC (mg/L)
                                                     1.5            2.0
                                                                                 o.O
                                                                                               0.5            1.0
                                                                                                         TOC (mg/L)
                                                                                                                            1.5            2.0
      30
   D)

   CD
   X
   CO
      20-
                                                                              200-1
                                                                              150-
                                                    O



                                                    o
                                                    CO
                                                                               50-
                                                                                    -D	r-
        0.0
0.5             1.0             1.5
           TOC (mg/L)
                                                                   2.0            o.O
                                                                                               0.5            1.0
                                                                                                         TOC (mg/L)
                                                                                                                            1.5            2.0
Figure D-2 Correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 2

-------
                    0.03 n
                 8
                   . 0.02 -
                    0.01 -
                    0.00
                            D Single contactor

                            • Blended effluent
                        0.0
0.5            1.0            1.5

          TOC (mg/L)

                                                                                           75-
                                                                                        CO
                                                                                           25-
                                                                                                                                         B
                                                                               2.0           o.O
                                                                                                            0.5
                                                                                                                          1.0

                                                                                                                      TOC (mg/L)
                                                                                                                                        1.5            2.0
Oi
O
                   30 -
                   20-
                w

                810
                     0.0             0.5
             1.0            1.5

         TOC (mg/L)
                                                                                          150-1
                                                                                        0 100-
                                                  X
                                                  O


                                                  Q   50-
                                                                                                                                         S
                                                                               2.0           0.0
                                                                                                            0.5            1.0            1.5

                                                                                                                      TOC (mg/L)
                                                                                                                                                      2.0
             Figure D-3 Correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 3

-------
0.04 -

0.03-

1"
_o
^ 0.02 -
8

0.01 -

0 00 -
su -
D Single contactor
• Blended effluent g
-
D 2-
"5)
D 5
D x 25 -
• jz

BDO ° |
• D
• D

D

D

D
D
•
D
D
1 °
M
•D
^ ^
U.UU ' I ' I ' I ' I ' I u -I 	 H-V1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1
0.0 0.5 1.0 1.5 2.0 2.5 o.O 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
50 -i
40-

.~~.
^
S 30-
i •
X
W 20-
Q
W

10-
n -
200
n
_ 150
T~
O
^j
D D a
X 10°
O

n W
D Q

D D 50
. -5
0 D
• • ^

D

D

D
D

D

D

D
D
D
D °
n
0.0 0.5 1.0 1.5 2.0 2.5 o.O 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
Figure D-4 Correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 4

-------
                     0.04 n
                     0.03-
                  8
                     0.02 -
                     0.01 -
                     0.00
                             D Single contactor
                             • Blended effluent
                        0.0      0.5
                           1.0       1.5       2.0
                                TOC (mg/L)
                                                                                            50 i
                                                                                         8
                                                                       2.5       3.0            o.O       0.5
                                                                                                                  1.0       1.5        2.0
                                                                                                                       TOC (mg/L)
                                                                                                                                              2.5       3.0
Oi
                   30 -
                   20 -
I
w
w
                   m -
                   nu
                     0.0       0.5        1.0        1.5       2.0       2.5
                                               TOC (mg/L)
                                                                                3.0
                                                                                            150 n
                                                                                         O 100 -
%   50 H
W
                                                                                                  n   D
                                                                               0.0       0.5
                                                                                                  1.0       1.5       2.0
                                                                                                       TOC (mg/L)
                                                                                                                              2.5       3.0
             Figure D-5 Correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 5

-------
ON
0.04 -
0.03-
_o
^ 0.02 -
$
0.01 -
o.c
10 -
1UU •
D Single contactor n
• Blended effluent
75-
D ^.
D j|
D 1 50-
. o t:
CO
D Q
• °
• D
_ D
••'
Kj
D
D
D
1 D D
n
D
%-"*
0.0 0.5 1.0 1.5 2.0 2.5 o.O 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
40 -|
30-
o o

-------
0.07-1
0.06-
0.05 -
'E
« 0.04-
-*•
tN 0.03 -
0.02 -
0.01 -
0.00 -

D Single contactor
• Blended effluent "-^
D
D
D
•D
• D
B *




i
;r
^
H
w
Q
w



175 -i

150 -

125 -
100-

75-


50-

25 -
n -

D

D D
m a
" *D °
D

M
D

D
•
1
                        0.0   0.5    1.0    1.5   2.0    2.5    3.0    3.5    4.0    4.5
                                                 TOC (mg/L)
                                                                               0.0    0.5    1.0    1.5    2.0    2.5    3.0    3.5    4.0    4.5
                                                                                                        TOC (mg/L)
ON
                W
                   50 i
                   40-
                   30-
                   20 -
Q
W
   10-
                                                                                            300 -i
                                                                                            250-
                                                                                         O 200-
                                                                                            150 -
W
W
                                                                                            100-
                                                                                             50-
                                                                                                                 • D
                     0.0    0.5     1.0    1.5    2.0     2.5    3.0    3.5    4.0    4.5
                                               TOC (mg/L)
                                                                               0.0    0.5    1.0    1.5    2.0    2.5    3.0    3.5    4.0    4.5
                                                                                                        TOC (mg/L)
             Figure D-7 Correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 7

-------
ON
0.03 -


—. 0.02 -
_o
si
3 0.01 -
o.c
0 -
SU -
D Single contactor
• Blended effluent
D
D 5~
D j|
^
• D X 25 -
i
"D w
• D
• D
.*B °
	 • . n


D
D°
B
• D
• D
D
" D
m n
0.0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.
TOC (mg/L) TOC (mg/L)
20 -i


"5)
^ 10 -
X
w
Q
w
0 -
0
150
D
D _
B o 100
• D ^
° D g
H
D W
_• Q 50
cr w
D
J


D
D
D
D
D
D
D
D
D
D
n D
0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.
                                           TOC (mg/L)
TOC (mg/L)
            Figure D-8 Correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 8

-------
                     15
                     10
                  O
                  Q
              D Single contactor effluent
              • Blended effluent
                       0.0     0.5     1.0     1.5     2.0     2.5    3.0    3.5     4.0
                                               TOC (mg/L)
                                                                                            25 •
                                                                                            20-
                                                                         o  15
                                                                         Q
                                                                         m
                                                                         w
                                                                         %  10
                                                                                                                                        D     D
                                                                              0.0     0.5     1.0     1.5     2.0    2.5    3.0     3.5     4.0
                                                                                                      TOC (mg/L)
a\
o\
                   25 -i
                   20-
a isH
o
m
§10
Q
W
    5 -
                                 j'1
                     0.0     0.5     1.0     1.5     2.0     2.5     3.0     3.5     4.0
                                              TOC (mg/L)
                                                                                            20-i
                                                                                            15-
                                                                                         u.  ioH
                                                                                             5-
                                                                                                                   •   e
                                                                              0.0     0.5     1.0     1.5     2.0    2.5    3.0     3.5     4.0
                                                                                                      TOC (mg/L)
             Figure D-9 THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 1

-------
ON
25
20
"S) 15
LJ_
O
g 10
w
5
0
c
D Single contactor effluent D
• Blended effluent 25 •
D
D -2°"
-3
• D 0 15 -
D m
• w
B J 5 -
• BD

•
• D Q
• D D D
D
• D
D
•D
•D
^
).0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
30 -|
25-
0 15-
m
Q
w
Q 10 -
w
5-
0 -
0.(
''I
D
D
• D
• D 5~ 10~
D)
• D ^
D g
• Q
m D « 5-
D
•D
• . 	 . n

•
D ° -D
D D
D
D
'
'•'•'•' 0 -i 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1
3 0-5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
            Figure D-10 THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 2

-------







1
Oi
OO
i



20

15
1
LL10
CO
Q
CO
5
0
C
-, 40-i
D Single contactor effluent
35-
• Blended effluent
D ^n -
D _ JU
"3> oc .
D 5
0 20 -
Q
m
CO 15-
D CO
10 :
• — • — BI-QMZI — i 	 1 	 1 	 1 	 1 	 1 	 1 o

D
D
D.
•
B •
D
.
).0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)

30 -|
25-
S" 20-
D)
^_
0 15-
m
Q
CO
s 1°-
5-
0 -
35-i
B 30 -
D
25 -
LL
D m
• CO 15-
• Q
_ D CO '
•D 10~
-• 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 n -

i 5 • • BD D
D
• n o
D f
D
•
       0.0
0.5            1.0            1.5
          TOC (mg/L)
                                                              2.0
                                                                            0.0
                                                                                         0.5           1.0           1.5
                                                                                                   TOC (mg/L)
                                                                                                                                  2.0
Figure D-11 THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 3

-------






1
Oi
VO






30
25
2-20

LL 15
W
Q
W 10
5
0
C
4 -
D Single contactor effluent n
• Blended effluent
- ° ^3"
D J? '
D 02-
• s
D W •
D W
• D 1 -
• D
•-» 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 n

D D
ODD D n
D
i i
• D
D
).0 0.5 1.0 1.5 2.0 2.5 Q.O 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)

15-,

S" 10-
D)
^_
^
O
m
Q
w
S 5"
0 -
-
D
D ^f
a ~s>
D -3
• LL-
D 9
W
H D W
•D"






3 0.5 1.0 1.5 2.0 2.5 Q.O 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
Figure D-12 THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 4

-------
       15
    O
    Q
    W  5
                 D Single contactor effluent
                 • Blended effluent
                                                                        o  10 -
                                                                        Q
                                                                        m
                                                                        w    -I
                                                                               5-
                                                                                                                            D
                                                                                                                          D  D
                                                                                                 D D
         0.0       0.5
                            1.0       1.5       2.0
                                 TOC (mg/L)
                                                     2.5       3.0            o.O       0.5
                                                                                                1.0       1.5       2.0
                                                                                                     TOC (mg/L)
                                                                                                                            2.5       3.0
      20 -
      15 -
O  10
m
Q
w
Q
W
    5
                                                                           m
                                                                              1 -
        0.0       0.5
                           1.0        1.5       2.0
                                 TOC (mg/L)
                                                     2-5        3.0           o.O       0.5
                                                                                               1.0        1.5       2.0
                                                                                                     TOC (mg/L)
                                                                                                                            2.5       3.0
Figure D-13 THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 5

-------
25


20

"S) 15

LJ_
O
g 10
w

5
n
-, 35-i
D Single contactor effluent n
30 -
• Blended effluent
_ 25 -
D 1?
E 2°"
_ Q
D CO 15-
W
Q
D W 10-
B
m D 5"
. m — rm — n-^n ........ n

D

D
• oo
D

DP


D
•fl
*"
0.0 0.5 1.0 1.5 2.0 2.5 o.O 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
35 -|
30 -


^ 25-
S
^ 20-

O
m
Q 15-
w
Q
w 10-

5-
0 -
20-i
D
n
15 -

D 5~
D)
-~i

D fe 10"
BW
Q
W
D
5 -
• D'



D
Q •

D D
•
n" D
D
n
m* Q

B
^

0.0 0.5 1.0 1.5 2.0 2.5 o.O 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
Figure D-14 THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 6

-------
to
20



15

„ — ,
— i
D)
^
u- 10
O
w
Q
W
5



o
c

50 -i
45 -

40-
3- 35-

3 30-
0 25-
CQ
§ 20~
Q
W 15 -
10 -
5 -
0 -
-, 75 n
D Single contactor effluent 70 -
n 65 -
• Blended effluent u 6Q
55 •

=d 50-
D)
3 45-

D o
Q 35-
D $ 3° '
S 25'
D 20 •
BD 15
B ^D 10-
B 5 •

).0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0
TOC (mg/L)
50-i
45 -
D
40 -
35 -
D ,-j.
n -&30-
LJ- 25 -
D 9
_ g 20-
n W
*D 15-
•
D 10 -

• i •. i — i 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 , n -

D

D D

D
I


•-i ^
B
Q

• D

B
D
HB

0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
TOC (mg/L)


g
^ n D

D
D
•
D
D

*

s
•
                     0.0    0.5    1.0
                                        1.5    2.0    2.5    3.0
                                             TOC (mg/L)
                                                                 3.5    4.0    4.5
0.0    0.5   1.0
                   1.5    2.0    2.5    3.0
                        TOC (mg/L)
                                           3.5   4.0    4.5
             Figure D-15  THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 7

-------
25
20
"S) 15
LJ_
O
g 10
w
5
0
c
D Single contactor effluent D
• Blended effluent 4 .
D _
5
D -1 3 -
D °
. S 3"
• D W
.. o°
•-•-n 	 1 	 1 	 1 	 1 	 1 	 1 	 1 n
•
• D
D D D D ° D
D
• D
D
).0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
15-,
5~ 10 -
D)
SDS-DBCIV
01
0 -
2 -|
S
• §1
. D u. 1 -
• D m
• D W
D g
• D
.8°


"

D 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
Figure D-16 THM correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 8

-------
       10
        8 -
        6 -
    o
    Q
    Q
    W
                 D Single contactor effluent
                 • Blended effluent
O
                                                                            w
                                                                            Q
                                                                               1 -
                                                                                                                     -rO-i-
         0.0
                 0.5     1.0     1.5    2.0     2.5     3.0     3.5     4.0
                                  TOC (mg/L)
    0.0     0.5     1.0
                          1.5     2.0     2.5      3.0     3.5     4.0
                              TOC (mg/L)
8 -
6 -
"5)
f 4-
Q
W
Q
W
2 -
•
0
8 -
D 6 '
D D D =?
D " -
mo m ^ 4-
• o
m
D W .
m D Q

2 -
D
•
• .• i 	 . 	 1 	 . 	 1 	 . 	 1 	 . 	 1 	 . 	 1 	 . 	 1 	 . 	 , n-
D
D
D
D
•
• °
D
• D
H D
D
-•-r^wi 	 rm— , 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 . 	 , 	 .
       0.0      0.5     1.0     1.5     2.0     2.5     3.0     3.5      4.0
                                 TOC (mg/L)
    0.0     0.5     1.0     1.5     2.0     2.5      3.0     3.5     4.0
                              TOC (mg/L)
Figure D-17 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 1

-------
10 -

8 •

3~
"5)
-3 6 -
^
o
Q
Q
W

2 •
0 •
D Single contactor effluent
• Blended effluent D


D ? 2 -
D 5
• D | .
7^
• D 'r
• D Q
w n "
• D
B ^
— • n 	 n ...... n

D

D
Q
B
D
B
D

B




0.0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
8-,
•

•
"5)
^_
1 — '
*4-
m
Q
w
Q
w
2 -


10
D 8
. D n
D D D
— I
_ "I C
H -^
5
• D S
_ D 9 ,
• W 4
Q
D W

D 2


D

D
• D

n
B D
• D

B D

B D
•D °

0.0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
Figure D-18 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 2

-------
ON
8 -

6-
"5)
f«-
Q
W
Q
W
2 -
0 -
0
D Single contactor effluent
• Blended effluent
D
D
D ? 2-
D)
n.
D 0
w
Q -I .
BD W
• • nan M • n 	 i n

D
D
D
D


• ILJBLJ • • IU 1 1 1 1 1 0 | m 	 B-rl_m_l 	 •— • 	 •!_!• — LJ— 1 	 1 	 1 	 1 	 1
0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
16 -i
14 -
12 -
[IT
]» 10 -
Q
if) 6 -
W
4 -
2 -
0 -
0
10
D
8
D
D -p
D 56
••D i
• w 4
• in w
D
2
H ^ 0


D
D o
D
DB
• •
B H D
0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
            Figure D-19 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 3

-------
16 -i

14 •

12 -
"™ 10 •
<^
O
9
W 6 •
Q
W
4 -

2 -

0 •
25
D Single contactor effluent

• Blended effluent D 20

D 5 15
D ^
o
D w 10
• D Q
W
• D
• D 5
D
0
— • •.• ......... n



D


D




D
D o'°

_ " D ™
^ ^ ^n
0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)

5-
"5)
n_
m
Q
w
Q
W

0 -
4 -
? '
"5)
5 3-
D <
D 0 '
S 2-
Q
W
D D
1 -
— • •.• — •. • ru — • . • . n . n . n . n
D
D D



• H
B n
0
^ ^ ^n
0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
Figure D-20 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 4

-------
oo
8 -
_

6-

3~
S
1 	 '
£4-
0
Q
W
Q
W
2 -

0 •
o
4 -i
'
3 -
;3"
"o>
3.
1"
Q
W
Q
W
1 -

•
0 -
D Single contactor effluent

• Blended effluent

DD „ 5 -

D S
-^ 4 .
3 .
• °

DO w
• Q
• W p .
g D
D
• • ™" 1 -
0 0.5 1.0 1.5 2.0 2.5 3.0 Q
TOC (mg/L)
8-1
D i—iD

"^~T
J" 1 '
•D 0 4"
en
S •
w
2 .

D
-• — r«— • 	 1 	 , 	 1 	 , 	 1 	 , 	 1 	 , 	 1 	 , n

D
D
D


D


D
• D

B
B
_
m
a
•D D
H H H~l ^^~
0 0.5 1.0 1.5 2.0 2.5 3.0
TOC (mg/L)

DD
D
D
_ D D
• B

™

•'-'
J
D
• .• — m, ..........
                   0.0        0.5        1.0       1.5        2.0       2.5
                                            TOC (mg/L)
3.0          0.0       0.5        1.0       1.5       2.0
                                    TOC (mg/L)
2.5       3.0
             Figure D-21 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 5

-------
VO
10

8
"5)
-3 6
o
Q
W 4
Q
W
2
0
C
1 6-1
D Single contactor effluent
• Blended effluent D 5 •
? 4-
D)
Jsi
D O
W
Q 2 -
D W
o B D
1 -


D

D
D
" D
• • •_••(• OH D 1 1 1 1 1 1 1 0 -|-B-B_BBIB — LJ-B— LB 	 LJ-1— LJ 	 LJ 	 1 	 1 	 1 	 1 	 1
).0 0.5 1.0 1.5 2.0 2.5 0.0 0.5 1.0 1.5 2.0 2.
TOC (mg/L) TOC (mg/L)
10 -i


8 -
3-
"5)
5 6-
m
§ 4-
Q
W
2 -
0 -
10-|

D
• D 8
° D D D ° 0-
1 6
• • m
w 4
D W
• D
• n «
D
D

D

• n
D D
.o"D
Ul 1 ' 1 ' 1 ' 1 ' 1 0 -TB-B-BBLJ 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1
D 0.5 1.0 1.5 2.0 2.5 o.O 0.5 1.0 1.5 2.0 2.
TOC (mg/L) TOC (mg/L)
            Figure D-22 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 6

-------
10
8
"5)
-3 6
o
9 4
Q
W
2

o
C

1
oo
0 20 -|


Ij~

| 10-
Q
W
Q
W
5 -
0 -
1 8-1
D Single contactor effluent n
7-
• Blended effluent
6-
D 2~ '
S5'
o o o D ° I4:
: ° .- I3;
• 2-
B1 % 1 -
• i •. 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , n
).0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0.
TOC (mg/L)

20
D
D
1
• |
B° < 10
m
D w
• 8
* 5
B
«-• 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 n

D
D D
D
. .
D

D
D 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
TOC (mg/L)




D
D
D D
D D

• "
Q °

        0.0    0.5    1.0    1.5    2.0    2.5    3.0    3.5    4.0    4.5
                                TOC (mg/L)
0.0    0.5   1.0    1.5    2.0    2.5    3.0    3.5    4.0    4.5
                        TOC (mg/L)
Figure D-23 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 7

-------
oo
8 -


6-

3~
^
—
3<-
Q
W
Q
W
2 -
•
0 •
o
D Single contactor effluent
n
• Blended effluent D

5 -
• D ?
D "ra
• — 4-
5 $ .
n o
H 3-
n C/5
• Q
D W 2-
n
^j 1 •
	 •— •— TB-Bl 	 , 	 , 	 , 	 , 	 , 	 , 	 . Q
0 0.5 1.0 1.5 2.0 o
D

D




• n
D a
D

D

D
CP
D
^ ^ P^K ^ P^K
0 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
2 -|



^^
	 i
§1
m
Q
w
Q
W
0 -
3]




=d 2 -
D)
n.
m
c/b
Q -i .
w
— •—•_!••-• — TB — m—rm-n 	 HI — IB ,n 	 n— n 	 n . r,
0.0 0.5 1.0 1.5 2.0 o

D D
D °
D
D
D

B
B
D


0 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
            Figure D-24 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 8

-------
       2 -,
    m
    o
    Q
    CO
    Q
    CO
                                                                              1 1
                 D Single contactor effluent


                 • Blended effluent
in
Q
o
co
Q
co
oo
K> 20-1


2-15"
"5)
n.
l£>
<: 10 •
X
CO
Q
CO
5-
0 -
25-i


D 20-
D
D 3.
D "5)
D D 3 15-
I;
• D Q
CO

• '• i 	 • 	 1 	 ' 	 1 	 ' 	 1 	 ' 	 1 	 ' 	 1 	 ' 	 1 	 ' 	 . 0-

D
D
D D
0°'
• D
• D
D
• .• i 	 . 	 1 	 . 	 1 	 . 	 1 	 . 	 1 	 . 	 1 	 . 	 1 	 . 	 1
        0.0     0.5     1.0     1.5     2.0     2.5

                                 TOC (mg/L)
                                                    3.0     3.5     4.0
     0.0     0.5    1.0     1.5     2.0     2.5     3.0     3.5     4.0

                              TOC (mg/L)
Figure D-25 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 1

-------
                    3n
                    2-
                 m
                 o
                 Q

                 co  .
                 Q  1
                 CO
                             D Single contactor effluent


                             • Blended effluent
                     0.0
                                   0.5
                                                 1.0


                                             TOC (mg/L)
                                                               1.5
                                                                             2.0
                                                                                         2-
in
Q
o
co
Q  1
co
                                                                                                                                        D  D
                                                                                                -•D	WO	DH
                                                                                                                      -rO-O—D	D-r-
                  05            10


                            TOC (mg/L)
                                              15
                                                            2.0
OO
                   20 -
                   15 -
                   10 -
                CO

                Q
                CO
                    5 -
                     0.0
                                                         D    D
                                   0.5            1.0


                                             TOC (mg/L)
                                                               1.5
                                                                             2.0
                                                                                         30 n
                                                                                         25-
                                                                                         20 -
                                                                                         15-
                                                                                          5-
                                                                                                                                 m      D
     0.0
                   0.5            1.0


                             TOC (mg/L)
                                               1.5
                                                            2.0
             Figure D-26  HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 2

-------
oo
3 -
-
I'"
m
o
Q
CO .
Q 1 -
CO
0 -
0
D Single contactor effluent
• Blended effluent
D D J2"
D ^
D D o
D D D • co 1 _
CO
• • r-\m M • • • n 	 i n

D

D
mii-im • • • • u i i i i i 0 1 • — m-n-imu — •— • — m-im — um-i 	 < 	 ILJ 	 < 	 1
0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
25 -i

20 -
l£>
I 10_
Q
CO
5-
0 -
0
35 -|

30
D
D
?25
5 20
Q)
^f
m "^ 15
D1 0
—
5
H ^ 0
D
D
B
D

,
D
• D IB
D
0 0.5 1.0 1.5 2.0 o.o 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
            Figure D-27 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 3

-------
                    6n
                    4 -
                 m
                 o
                 Q

                 82
                 CO
                              D Single contactor effluent


                              • Blended effluent
                                                                D  D
                     0.0         0.5
                                             1.0          1.5

                                               TOC (mg/L)
                                                                    2.0          2.5
                                           in
                                           Q
                                               Q.O         0.5
                                                                       1.0          1.5

                                                                         TOC (mg/L)
                                                                                              2.0
                                                                                                          2.5
oo
                   40 -
                   30 -
                 D)
                 n.

                l£>
                   20-
                CO
                Q
                CO
                   10 -
                     0.0
                                 0.5
1.0          1.5

  TOC (mg/L)
                                                                    2.0          2.5
                                                                                        CO
                                                                                           50 n
                                                                                           40 -
                                              30 -
                                                                                           20 -
                                                                                             Q.O         0.5
                                                                                                                    1.0          1.5

                                                                                                                      TOC (mg/L)
                                                                                                                                           2.0          2.5
             Figure D-28 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 4

-------














1
oo
Oi












10
8
3~
•3 6
m
o
9 4
CO
Q
CO
2

n
b -
D Single contactor effluent
• Blended effluent D DD 5"
D _ •
a 4"
• D D ^3-
• O
_•• o
D «o
Q 2 ~
D co
• D
D 1 -
D
• • •

D
D
D DQ
D
• i

• r%





U _ ._ _ . i . i . i . i . i u -|— « 	 _ 	 |_j 	 LB 	 LJ-, 	 , 	 1 	 , 	 1 	 , 	 1 	 , 	 1
0.0 0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0
TOC (mg/L) TOC (mg/L)


20 -|

15 -
2~
"5)
^
LO
<; 10-
X
CO
Q
CO
5-

0 -
40
35
DD 30
D 2~
n "§> 25
D 3
• ° D | 20
D X
• • co 15
• co
D BD ° 10
• 5
n
-• 	 rQ 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 n

DD
D

p.
D
B"
*
D

"D
m
-m 	 fl P 	
        0.0       0.5        1.0       1.5       2.0       2.5
                                TOC (mg/L)
                                                                3.0
0.0       0.5
                   1.0       1.5       2.0
                        TOC (mg/L)
                                              2.5       3.0
Figure D-29  HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 5

-------
oo
6 -
-
S 4"
^r
m
o
Q
82-
CO
0 -
0
4 -
D Single contactor effluent D
• Blended effluent
D _ 3-
D ~^
a g
D CO
Q
D D co
1 -


D D
D
D D
D
^ •rn^^f^ r-i ^ r^m
0 0.5 1.0 1.5 2.0 2.5 Q.O 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
25 -i
20 -
2~
l£>
5 1°-
Q
CO
5-
0 -
0
45
40
D
35
?30
D)
D cf 25
" D x 20
D D D W
CO 15
• 10
• n L ^

D
D
D
• D
D D °
•
_ •••"
Ul ' • ,.,.,., 0 -,-_ 	 u, 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 , 	 ,
0 0.5 1.0 1.5 2.0 2.5 Q.O 0.5 1.0 1.5 2.0 2.5
TOC (mg/L) TOC (mg/L)
            Figure D-30 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 6

-------
12 -
10 •

D) 8-
CD 6 -
O
9
CO .
Q 4 •
CO

2 •

0 •
0

i
OO
00 8-1

"
6 -

2~
"5)
n.
^T
m 4 "
I—
CO
Q
CO
2 -

.

0 --•
D Single contactor effluent Q
• Blended effluent

n ?
I
Q
3
• Q
D co
•D

ft m°
m i •. — m — i — i — i — i — i 	 1 — i — i — i — i — i — i — i 	 , — i
0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
JOC (mg/L)



D
D
• •
D
_ D n 3"
D "D 2
D
<
X
CO
Q
CO
D



-i-»i — n — .• i • i — i — i — i — i — i — i — i — i — i — i — i — ,
•
6 -
•
5 -
4 -

3 -

2 -
•
-1 _
i

0.



80

70
60


50

40

30
20

10

n
D D
D
D
D D
D

D '



D


3 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
JOC (mg/L)




D
D

n D
D D
JD g

g
D

D •
•
f
_ •
       0.0    0.5    1.0    1.5    2.0    2.5    3.0    3.5    4.0     4.5
                                TOC (mg/L)
0.0    0.5   1.0    1.5    2.0    2.5    3.0    3.5   4.0    4.5
                        TOC (mg/L)
Figure D-31  HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 7

-------
3 -
-
S 2"
m
o
Q
CO .
Q 1 -
CO
0 -
0
i -
D Single contactor effluent
• Blended effluent D
n 	 i
* • rj ~5)
D D <
m
1
CO
• • PB • FB 	 i n




B LB B LB 1 1 1 1 1 1 0 H 	 B-B-LB-BI 	 LB — B 	 ILB-LJ 	 BJ 	 IB 	 TLJ 	 LJ— LJ 	 LJ— " 	 1
0 0.5 1.0 1.5 2.0 o.O 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
i
oo
^P 151

? 10-
D)
i •
CO
CO 5 "

0 -
0
20
D
D
15
D ^)
B D 3
qj
r°
CO
D Q
B CO
f 5
D
• • FB • L "

D
D
" . D °
D
D
D
D
f
• LB • i i i i i i i o -i — m-m-Lm-m 	 1 	 1 	 1 	 1 	 1 	 1 	 1
0 0.5 1.0 1.5 2.0 o.o 0.5 1.0 1.5 2.0
TOC (mg/L) TOC (mg/L)
Figure D-32 HAA correlations based on GAC effluent TOC concentration for single contactor and blended effluents for Water 8

-------
                                                                                          75 -,
                                                                                       =d 50
                                                                                        D)
                                                                                       W
                                                                                       g
                                                                                                D Single contactor
                                                                                                • Blended effluent
                                                                                            0.00
                                                                                                          0.02           0.04           0.06
                                                                                                                    UV-254(1/cm)
                                                                                                                                                      0.08
VO
o
                  30 -,
                D)

                CD
Q
W
                  20 •
                    0.00
                                  0.02           0.04          0.06
                                            UV-254(1/cm)
                                                                            0.08
                                                                                          150-1
                                                                                       0 100-
                                                                                       g
                                                                                           50-
                                                                                       W
                                                                                            oo
                                                                                             0.00
                                                                                           0.02           0.04           0.06
                                                                                                    UV-254(1/cm)
                                                                                                                                                    0.08
             Figure D-33 Correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 1

-------
                                                                              100
                                                                               75 -
                                                                           X  50 •
                                                                           W
                                                                           Q
                                                                           W
                                                                               25 -
                                                                                     D Single contactor
                                                                                     • Blended effluent
                                                                                 0.00
                                                                                               0.01           0.02           0.03
                                                                                                         UV-254(1/cm)
                                                                                                                                          0.04
     30 n
   D)

   CD
   X
   W
     20 •
       0.00
                     0.01            0.02           0.03
                               UV-254(1/cm)
                                                                0.04
X
o
w
                                                                              100-
                                                                               50-
                                                                                0.00
                   0.01           0.02          0.03
                             UV-254(1/cm)
                                                                                                                                        0.04
Figure D-34 Correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 2

-------
                                                                                           125
                                                                                           100 -
                                                                                            75
                                                                                        CO
                                                                                        Q
                                                                                        CO
                                                                                            50 -
                                                                                            25 -
                                                                                                  D Single contactor
                                                                                                  • Blended effluent
                                                                                             0.00
                                                                                                                                           B
                                                                                                                0.01                0.02
                                                                                                                     UV-254(1/cm)
                                                                                                                                                      0.03
VO
3U •



5~ 20 •
D)
CD
X
CO
CO 10-

•
n .
IbU -
D
D
a
D _
0 100 -
	 I
D)
D • O
B B 'r
n Q 50-
• D co

D * °
. ...... n .

B
D

D

D
D
D
D
D
D
D
D
                    0.00
                                       0.01                0.02
                                            UV-254(1/cm)
                                                                             0.03
                                                                                             0.00
0.01                0.02
     UV-254(1/cm)
                                                                                                                                                     0.03
             Figure D-35 Correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 3

-------
                                                                                           50 -I
                                                                                        X  25-
                                                                                        w
                                                                                        Q
                                                                                        W
                                                                                                 D Single contactor
                                                                                                 • Blended effluent
                                                                                            0.00
                                                                                                           0.01
                                                                                                                         0.02
                                                                                                                     UV-254(1/cm)
                                                                                                                                        0.03
                                                                                                                                                      0.04
vo
                  40 •
                330-I
                3    1
                Q
                w
                   10 -
                    0.00
  D C
rj  •
                                  0.01           0.02           0.03
                                            UV-254 (1/cm)
                                                                             0.04
                                                                                          200 n
                                                                                           150-
                                                                                        O
                                                      X
                                                      o
                                                      w
                                                      Q
                                                         100 -
                                                                                           50 -
                                                                                            0-K3-
                                                                                             0.00
                                                                         0.01           0.02           0.03
                                                                                   UV-254 (1/cm)
                                                                                                                                                     0.04
             Figure D-36 Correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 4

-------
                                                                                          50 -,
                                                                                       X 25 -
                                                                                       W
                                                                                       Q
                                                                                       W
                                                                                                D Single contactor
                                                                                                • Blended effluent
                                                                                            0.00           0.01
                                                                                                                        0.02           0.03
                                                                                                                   UV-254(1/cm)
                                                                                                                                                    0.04
VO
                  30
                  20 -
               X
               w
               Q
               W
-in .
1U
                     ™
                                                               DD
                    0.00
                                  0.01           0.02          0.03
                                           UV-254(1/cm)
                                                                           0.04
                                                                                          150-i
                                                                                       O  100 -
                                                                                       15>
                                                                     Q  50-
                                                                          0.00           0.01          0.02           0.03
                                                                                                 UV-254(1/cm)
                                                                                                                                                   0.04
             Figure D-37 Correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 5

-------
                                                                                           100 n
                                                                                            75 -
                                                                                        X   50-
                                                                                        W
                                                                                        Q
                                                                                        W
                                                                                            25-
                                                                                                 D Single contactor
                                                                                                 • Blended effluent
                                                                                             0.00
                                                                                                           0.01           0.02           0.03
                                                                                                                    UV-254(1/cm)
                                                                                                                                                     0.04
VO

30-
5~
O>

CD
< 20-
X
w
Q
W
10-


n -
iiSU •
200-
D _
D °
=d 150-
D)
D 5
X
B D 1— 1 00 •
D D D £
W
• 50-
D


D

D

D

D
D
D B
D
i "

i"
                    0.00           0.01
                                                0.02           0.03
                                           UV-254(1/cm)
                                                                            0.04
                                                                                             0.00
0.01           0.02          0.03          0.04
         UV-254(1/cm)
             Figure D-38 Correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 6

-------
                                                                                          W
                                                                                          Q
                                                                        175

                                                                        150 -

                                                                        125

                                                                        100 -

                                                                         75

                                                                         50

                                                                         25
                                                                                                    D Single contactor
                                                                                                    • Blended effluent
                                                                                               0.00     0.01     0.02     0.03    0.04     0.05
                                                                                                                        UV-254(1/cm)
                                                                                                                                                 0.06    0.07
vo
bU •
40 •

•3-
D)
3. 30 •
CD
| '
co 20 •
Q
CO
10 •
.
n .
3UU -
D
D 25°-
D
D 0 200 -

1 . | •
^T 150 -
• s*\
D °
I —
D CO .,,-„-.
Q 100-
• CO
B 50-

i
• 	 n .
D
D

D
D
D
m
ff
m
f
a
, •
0.00     0.01      0.02     0.03     0.04    0.05     0.06     0.07
                        UV-254(1/cm)
                                                                                               0.00     0.01     0.02     0.03    0.04     0.05     0.06
                                                                                                                       UV-254(1/cm)
                                                                                                                                                        0.07
             Figure D-39  Correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 7

-------
                                                                             50 -I
                                                                          X  25 -
                                                                          W
                                                                          Q
                                                                          W
                                                                                   D Single contactor
                                                                                   • Blended effluent
                                                                              0.00
                                                                                                 0.01                0.02
                                                                                                      UV-254(1/cm)
                                                                                                                                       0.03
VO
7-1 20-1


3~
"5)
CD
5 10 •
X
w
Q
W



n <
150-i
D
D „
_ 0 100-
• _l
^0 ^-
D D 0
D. ^50-
D " W
B
D
D "



D
D
D
D
D
D
D
D
D
D
D
4
       0.00
                         0.01                0.02
                              UV-254(1/cm)
                                                              0.03
                                                                               0.00
0.01               0.02
     UV-254(1/cm)
                                                                                                                                     0.03
Figure D-40 Correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 8

-------
15 -

5-10-
3
LJ_
o
Q
W 5-

0 i
0.
D Single contactor effluent
• Blended effluent 25 •
D
|2°-
5
0 15 -
Q
0 %
Q 10 -
• D" 5-
• D D

D D
D
D D
D
D
D
D
Bi
•
DO 0.02 0.04 0.06 0.08 o.OO 0.02 0.04 0.06 O.C
UV-254 (1/cm) UV-254 (1/cm)
VO
0° 25-,

20-
;Zr
~3>
3 15-
0
m
§10-
Q
w

5-

0 1
20-i
D
D 15-
2~
"3)
D 5

D " Q •
"
• 5 •
D
D
MB 	 1 	 1 	 1 	 1 n i


a
'"U
D D
HD " D D
' D
D
D

•
"
i . . . .
       0.00
0.02           0.04           0.06
          UV-254 (1/cm)
                                                               0.08
                                                                             o.OO
                                                                                          0.02          0.04           0.06
                                                                                                    UV-254 (1/cm)
                                                                                                                                    0.08
Figure D-41  THM correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 1

-------
25 -
20 -
"S) 15 -
LJ_
O
g lO-
c/3
5-
0 -
0.
D Single contactor effluent D
• Blended effluent 25 -
D
I20:
• D 0 15 -
D m
• w
•— n — i 	 1 	 1 	 1 	 1 	 1 	 1 	 1 n

• "D ° D D
D
• D
D
ft
^
DO 0.01 0.02 0.03 0.04 0.00 0.01 0.02 0.03 O.C
UV-254 (1/cm) UV-254 (1/cm)
VO
^P 301
25-
0 15-
m
Q
w
Q 10 -
w
5-
0 -
15-
D
D
• D
• D 5~ 10~
D)
~1
f .
D 
-------
20
15
1
LL10
w
Q
W
5

0
0
-, 40-i
D Single contactor effluent
35-
• Blended effluent
° a-30:
D 1 25 '
0 20 -
Q
m
W 15 -
D w
10 :
n" " ° 5-
D •"
• • — n — m\ 	 1 	 1 	 1 	 , 	 1 Q

D
D
B
D

•
00 0.01 0.02 0.03 o.OO 0.01 0.02 0.03
UV-254(1/cm) UV-254(1/cm)
to
O
0 30-.
25-
3 .
0 15-
m
Q
w
Q 10 -
w
5-
0 -
35-i
H 30 -
D
25 -
3~
n o nri
-3 20 "
LL
D OQ
• (/) 15 -
• Q
. D W '
10;
D B 5 -
g]
• 	 1 	 , 	 1 	 , 	 1 	 , n

Da" . • ° "°
• • n
% Q
D
•
       0.00
                          0.01               0.02
                               UV-254(1/cm)
                                                               0.03
0.00
                  0.01               0.02
                       UV-254(1/cm)
                                                       0.03
Figure D-43  THM correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 3

-------
                    30
                    25 -
                    20
-  15 H


w
Q

W  10
                     5 -
                      0.00
             D Single contactor effluent


             • Blended effluent
                                    0.01
                                 0.02


                             UV-254 (1/cm)
                                                                 0.03
                                                                               0.04
                                                                                           4 -i
                                                                                         02-

                                                                                         s

                                                                                         8
                                                                                         W
                                                                                            1 -
                                                                                            o.OO
                                                                                                             D  D
                                                                                                           0.01
                                                                                                                         0.02


                                                                                                                     UV-254 (1/cm)
                                                                                                                                        0.03
                                                                                                                                                       0.04
to
                   15-,
                0
                m
                Q

                w
                Q  5-
                w
                     0.00
                                   0.01
                                 0.02


                            UV-254 (1/cm)
                                                                 0.03
                                                                               0.04
                                                                                            1 -
                                                                       W
                                                                                            o.OO
-CBrOBQ	rOB	D	D-.	D—i—

  0.01           0.02            0.03


            UV-254 (1/cm)
                                                                                                                                                       0.04
             Figure D-44 THM correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 4

-------
                    15
                 O
                 W
                 Q
                 W  5
                              D Single contactor effluent
                              • Blended effluent
                                                                  DD
                                            mS
                     o •
                      0.00
0.01           0.02           0.03
          UV-254(1/cm)
                                                                               0.04
                                                    O  10 -
                                                    Q
                                                    m
                                                    w
                                                    Q
                                                       5 -
                                                                                                         D D
                                                                                            o.OO
                                                                                                                                        D
                                                                                                                                   D      D
                                                                                                          0.01           0.02           0.03
                                                                                                                    UV-254(1/cm)
                                                                                                                                                     0.04
to
O
                   20 -
                   15 -
                O  10
                m
                Q
                w
                Q
                w
                    5-
                    0.00
                                   0.01           0.02           0.03
                                             UV-254(1/cm)
                                                                               0.04
                                                                                          3-1
                                                    LL
                                                    m
                                                    w
                                                       1 -
                                                       0.00
                                                                      0.01           0.02            0.03
                                                                                UV-254(1/cm)
                                                                                                                 0.04
             Figure D-45 THM correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 5

-------
to
25


20

3~
"S) 15

LJ_
o
g 10

w


5
n
-, 35-i
D Single contactor effluent n
30 -
• Blended effluent
_ 25 -
D S
-20-
O
Q
D op 15 -
th
Q
D W 10-
D
D • 5;
' • "• * "D
.•n-an-ji ....... r.

D

D
° ' D
D


D •



D

••
§"
^
0.00 0.01 0.02 0.03 0.04 0.00 0.01 0.02 0.03 0.04
UV-254 (1/cm) UV-254 (1/cm)
35 -|
30 -

2-25-

;§ 20-
0
m
Q 15-
w
Q
W 1Q .


5-
0 -
20-,
D
D 15-

D 5
D)
D u- 10 -

Bl
(/")
D Q
W

D
5 -
S '
•n ........ n


D D •

D D
•
° '


D
• D




^
0.00 0.01 0.02 0.03 0.04 o.OO 0.01 0.02 0.03 0.04
UV-254 (1/cm) UV-254 (1/cm)
            Figure D-46 THM correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 6

-------
20


15
3~
LL10
w
Q
W
5
o
0

to
o
fk 501
45 -
40-
3- 35-
a 30-
0 25-
cn
Q
W 15 -
10 -
5 -
0 -
0.0

-, 75 n
D Single contactor effluent 70 -
n 65 -
• Blended effluent u 60 ;
55 -
i>50'-.
a 45-
i-i ^ 40 -
D o
Q 35-
D C/5
D 20 -
_ rf 10:
• a 5:
00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 Q.
UV-254(1/cm)

50 -|
45 -
D
40-
D 3-35:
n o) 30-
3_
LJ- 25 -
D 9
W 20 -
m W 15;
D 10-
._ 1 * * 5
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 o.
UV-254(1/cm)

D
D D
D
• •


cm
a
DO 0.01 0.02 0.03 0.04 0.05 0.06 0.07
UV-254(1/cm)



[f D
• D D
D
D
D
•
D
•
1
DO 0.01 0.02 0.03 0.04 0.05 0.06 0.07
UV-254(1/cm)
Figure D-47 THM correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 7

-------
                    25
                    20 -
                  D) 15 -
                    10
                 w
                              D Single contactor effluent


                              • Blended effluent
                      0.00
 0.01                0.02


      UV-254 (1/cm)
                                                                                        0

                                                                                                                     D   D
                                                                              0.03         o.OO
                                                                                                              0.01                0.02


                                                                                                                   UV-254 (1/cm)
                                                                                                                                                     0.03
to
                   15-,
                0
                m
                Q

                w
                Q   5-
                w
                    0.00
0.01                0.02


     UV-254 (1/cm)
                                                                                          2 -i
                                                u.  1 -
                                                o
                                                8
                                                                              0.03         o.OO
          -•D	D—iD—D-

0.01                0.02


     UV-254 (1/cm)
                                                                                                                                                     0.03
             Figure D-48 THM correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 8

-------
10 -
8 -
"5)
-3 6 -
o
Q
Q
W
2 •
0 <
0.
D Single contactor effluent
• Blended effluepj D
D
D „
D ^
• 3
i1"
w
Q
. S
" D
D


D
D


30 0.02 0.04 0.06 0.08 o.OO 0.02 0.04 0.06 0.08
UV-254 (1 /cm) UV-254 (1 /cm)
to
O
ON 8 -i
6 -
"5)
I-
Q
W
Q
W
2 -
0 1
0.
8-1
D _6-
D D ° D ?
D " —
a • <: 4-
• o
m
D w
2 -
D
DO

D
D
D
D
•
D
M
BD
0.02 0.04 0.06 0.08 o.OO 0.02 0.04 0.06 0.08
UV-254 (1 /cm) UV-254 (1 /cm)
Figure D-49 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 1

-------
10 -
8 -

3~
"5)
-3 6 -
o
Q
Q
W

2 •
n .
D Single contactor effluent
• Blended effluent D

D ^ 2-
D ^
5 .
D £
w 1 "
• D
B ^
• n . n . n

D
D
D
• D"
"

0.00 0.01 0.02 0.03 0.04 o.OO 0.01 0.02 0.03 0.04
UV-254 (1 /cm) UV-254 (1 /cm)
to
O

-------
8 -

6-
"5)
O
Q
CO
Q
CO
2 -
0 -
0.
D Single contactor effluent
• Blended effluent
D
D
D 5~2-
D)
D 0
CO
Q 1 .
BD CO '
• — • — •}— D-Bfl — D 	 1 	 1 	 1 	 1 o
00 0.01 0.02 0.03 QC

D

D
D
D

0 0.01 0.02 0.03
UV-254 (1 /cm) UV-254 (1 /cm)
to
O
00 16-,
14 -
12 -
"5> 10 .
<
m
Q
CO 6-
CO
4 -

2 -
0 -
10
D
8
D
D -j-
"5)
Q 3; 6
. D <:
• °

• D H W
2
•
•• 	 1 	 1 	 1 	 1 	 1 	 , n


D
° Q
D

D
D H

       0.00
                         0.01                0.02
                              UV-254 (1/cm)
                                                              0.03
0.00
                  0.01               0.02
                       UV-254 (1/cm)
                                                      0.03
Figure D-51  HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 3

-------
16 -I

14 •

12 -
"5> 10 •

^
^ 8
O
9
W 6 •
Q
W
4 -

2 •
0 •
25
D Single contactor effluent

• Blended effluent D 20

D "B)
=1- 15
D ^

O
I 	
D W 1°
D" Q
(f)
nBD
if 5

•• ........ n



D



rj
D





D
n D"
. ^
ff D
0.00 0.01 0.02 0.03 0.04 o.OO 0.01 0.02 0.03 0.04
UV-254 (1/cm) UV-254 (1/cm)
i
to
O
vp 2-

2-
"5)
^:
5i-
m
Q
w
Q
w
•

0 -
5"
4 -
3- '
"5)
3; 3 -
D D <;
o
m
W 2-
Q
W
D '-'
1 -


D
D D


D D
B
D D
[f "
B

*

0.00 0.01 0.02 0.03 0.04 o.OO 0.01 0.02 0.03 0.04
UV-254 (1/cm) UV-254 (1/cm)
Figure D-52 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 4

-------

















1
to
o















8 -
_


6-

3~
S '
I-
Q
W
Q
W
2 -

0 •
D S\ng\e contactor effluent

• Blended effluent


DD „ 5 -

D J
D ^ •
_ O
• 1- 3 -
D D W
• <~i
J 9 -
•p
_ D
•— n — n. r,

a
D
D


D

D
0

g
B
_

D
BD
0.00 0.01 0.02 0.03 0.04 Q.OO 0.01 0.02 0.03 0.04
UV-254 (1 /cm) UV-254 (1 /cm)


4 -i
3 -
"5)
n.
?2-
m
Q
w
Q
w
-1 .
1
•
0 -
8-1
D 1 D ° ^ 6
_ • • -^
D " JL '
%
m o

• s .
D Q
W

2 -
D


D
D

g D
_
!-•

D •


8

a

0.00 0.01 0.02 0.03 0.04 Q.OO 0.01 0.02 0.03 0.04
UV-254 (1 /cm) UV-254 (1 /cm)
Figure D-53 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 5

-------
10
8
"5)
-3 6
^
O
Q
W 4
Q
W
2
0
0
b -
D Single contactor effluent
• Blended effluent D 5 -
? 4 -
D)
D n_
<
D 0
W '
Q 2 -
D W
0° '
1 -

D

D
D
" D
00 0.01 0.02 0.03 0.04 o.OO 0.01 0.02 0.03 0.04
UV-254 (1/cm) UV-254 (1 /cm)
10 -i


8 -
2~
1 6-
<
§ 4-
Q
W
2 -
0-
0.0
10

D
• D 8
°D D D ° 3-
1 6
<
• • m
w 4
D W
rf^" 2


D
D

D

- D
D° '
m
D
f
0 0.01 0.02 0.03 0.04 o.OO 0.01 0.02 0.03 0.04
                             UV-254 (1/cm)
                                                                                                UV-254 (1/cm)
Figure D-54 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 6

-------
10

8
3~
"5)
-3 6
o
9 4
Q
W
2

o
0

to
to 20-,


^5 -
^

I 10-
Q
W
Q
W
5 -
0-
1 8-1
D Single contactor effluent n
7-
• Blended effluent
6-
D 2~ '
0> 5 -
D ^
D < 4 -
D D D h-
: - .- s3:
• 2-
S J " 1-
•-•n 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 n
00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.C
UV-254 (1/cm)

20
D
D
1
• |
• o 10
m
D w
• Q
w
• 5
a
1 — i 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1 n


D
D D
D


••
D

D
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
UV-254 (1/cm)




D
D
D D
D D

• "
j—

i
       0.00    0.01     0.02
                                0.03     0.04     0.05
                               UV-254 (1/cm)
                                                        0.06     0.07
0.00     0.01     0.02     0.03     0.04     0.05     0.06    0.07
                       UV-254 (1/cm)
Figure D-55 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 7

-------
to
8 -
"
6-
^
I-
Q
W
Q
W
2 -

0
(

D Single contactor effluent
D 6-
• Blended effluent D
• D ?
D "ra
r:
D Pa-
D. Q •
D ' W 2-
o* " 1:
I ^ - - .
D
D


D
D
D
D •
•
D
100 0.01 0.02 0.03 o.oo Q.QI 0.02 0.03
UV-254 (1 /cm) UV-254 (1 /cm)
2 -|


_ 	 _
^
;|
m
Q
w
Q
w
0 )
0.
3]



? 2 -
D)
~
o
m
Q i .
w '
30 0.01 0.02 0.03 o.

D D
D ° °
D
D
D •
D "
D
30 0.01 0.02 0.03
                                         UV-254 (1/cm)
                                                                                                           UV-254 (1/cm)
            Figure D-56  HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 8

-------
                    2 -i
                 m
                 o
                 Q
                 CO
                 Q
                 CO
                     0.00
                               D Single contactor effluent


                               • Blended effluent
                                   0.02
     0.04

UV-254(1/cm)
                                                                0.06
                                                                               0.08
                                                                                           1  i
                                           CO
                                           Q
                                           o
                                           CO
                                           Q
                                           CO
                                                                                            Q.OO
                                                                                                          0.02
                                                                                                                         0.04

                                                                                                                    UV-254 (1/cm)
                                                                                                                                       0.06
                                                                                                                                                      0.08
to
^u -


15 -
"5)
n.
l£>
<; 1° -
X
CO
Q
CO
5-


0 1
2b -

D 20-
D
D -j-
D "5)
D D 3 15-
• en
^
• x
co 10 -
• O
CO
B
VD 5 -
i"
n 	 , 	 , 	 , 	 , n >
D
D

D D

rj m
D



M

•D
D

!• 	 . 	 . 	 . 	 .
                     0.00
                                   0.02            0.04           0.06

                                             UV-254 (1/cm)
                                                                               0.08
                                               0.00
                                                              0.02           0.04           0.06

                                                                       UV-254(1/cm)
                                                                                                        0.08
             Figure D-57 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 1

-------
                    3n
                 m
                 o
                 Q
                 th  .
                 Q  1
                              D Single contactor effluent


                              • Blended effluent
                     0.00
                             i-BD—Or—

                                   0.01
               0.02

          UV-254 (1/cm)
                                                                0.03
                                                                              0.04
                                                     g

                                                     O
                                                     co
                                                     Q 1
                                                     CO
                                                                                            0 00
                                                                                                                                      D   D
                                                                                                          0 01
                                                                                                                         0.02

                                                                                                                    UV-254 (1/cm)
                                                                                                                                       Q.03
                                                                                                                                                     0.04
to
                   20 -
                   15 -
                   10 -
                co
                Q
                CO
                    5 -
                     0.00
                                                     D     D
0.01           0.02           0.03

          UV-254 (1/cm)
                                                                              0.04
                                                                                           30 -i
                                                                                           25-
                                                                                           15-
                                                                                            5-
                                                                                             Q.OO
                                                                                                           0.01           0.02           0.03

                                                                                                                     UV-254 (1/cm)
                                                                                                                                                     0.04
             Figure D-58 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 2

-------
to
3 -
-
I'"
m
o
Q
CO
Q 1 -
0 -
0.
D Single contactor effluent
• Blended effluent
D D J2"
D ^
D D o
D D D • W 1 .
CO
• Bl B •• • • n n 	 i n

D

D
mi mi mm \m m i-i i \ i i o -+B — •! — mj — \-i-mm — urn m u — •!_) 	 1 	 LJ — i 	 1
00 0.01 0.02 0.03 Q.OO 0.01 0.02 0.03
UV-254 (1/cm) UV-254 (1 /cm)
25 -i

20 -
Il5-
l£>
I 1Q_
Q
CO
5-
0 -
35 -|

30
D
D
i25
-^ 20
Q)
^f
m "^ 15
m D § 1°
¥1 D H
• 5
•• 	 1 	 , 	 1 	 , 	 1 	 , n
D
D
Q
D
,
D
f\ D B
D
. ......
                    0.00
                                      0.01                0.02
                                           UV-254 (1/cm)
0.03          Q.OO
                               0.01               0.02
                                    UV-254(1/cm)
                                                                    0.03
             Figure D-59 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 3

-------
to
6 -
-
S 4"
m
o
Q
82-
CO
0 -
0.
D Single contactor effluent D
• Blended effluent
D D 5"
D)
n.
n HI
D Q
D 9
D • co
D Q
•• . 	 . n






• ' i ' i ' i ' i 0 1 ••L»T—» — wn_m_i 	 n_r« 	 u 	 i_n 	 u — i 	 u — i 	 1
00 0.01 0.02 0.03 0.04 Q.OO 0.01 0.02 0.03 0.04
UV-254 (1/cm) UV-254 (1/cm)
40 -|


30 -
"5)
n.
l£>
^ 20 -
X
CO
Q
CO
10 -
0.(
50
D
40

D 1 30
D
Q)
^f
5 20
D Q
D D
• J ' 1°
- D

D


D

D D
-••'
_Q
DO 0.01 0.02 0.03 0.04 Q.OO 0.01 0.02 0.03 0.04
UV-254 (1/cm) UV-254 (1/cm)
            Figure D-60 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 4

-------
10
8
3~
•3 6
m
o
9 4
CO
Q
CO
2
0
0
b -
D Single contactor effluent
• Blended effluent D DD 5-
D _ •
ID 1 3-
G CO
• D
D 1 -
D
• B • . 	 . n

D
D
D DQ
D


• •' 1 ' 1 ' 1 ' 1 U -fc-« — LJ-r-l_l»-l_|— i 	 1 	 1 	 1 	 1 	 1 	 1
00 0.01 0.02 0.03 0.04 0.00 0.01 0.02 0.03 O.C
UV-254 (1/cm) UV-254 (1/cm)
to
00 20 -|


15 -
"5)
n.
l£>
<; 10-
X
CO
Q
CO
5-

0-
0.0
40
35

DD 30
D 2~
D I25
• D | 20
D X
• • co 15
• co
D B D 10
• 5
,'3
0 0.01 0.02 0.03 0.04 n


D
D
D
D
"
• D
• l\
00 0.01 0.02 0.03 O.C
                              UV-254 (1/cm)
UV-254 (1/cm)
Figure D-61 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 5

-------
to
6 -
-

"^l*
H 4"

^
m
o
Q
CO _
Q 2 -
CO
n -
4 -
D Single contactor effluent D
• Blended effluent
D _ 3-
i~
"5)
n ^r
^
CO 2 •
n Q
B S •
Q
D D CO
1 -


n n
n

5 D
n






0.00 0.01 0.02 0.03 0.04 Q.OO 0.01 0.02 0.03 0.04
UV-254 (1/cm) UV-254 (1/cm)
25-i 45-1


20 -

-•— -
	 1
O)

in
<
co 10-
Q
CO
5-
n -
40

D
35

D —
=d 30
D)
n 5
05 25
• D 1 20

CO 15
• 10
-0 ° 5
• 1-1

o

n


n


• D
n
n n D

•
o
U ^1-1 • I • I • I • I u -Vl_l 	 1 	 1 	 1 	 1 	 1 	 1 	 1 	 1
0.00 0.01 0.02 0.03 0.04 Q.OO 0.01 0.02 0.03 0.04
UV-254 (1/cm) UV-254 (1/cm)
            Figure D-62 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 6

-------
12 -

10 •

2~
• — Q
D) 0
CQ 6 -
O
9
CO .
Q 4 •
CO

2 •
0 •
D Single contactor effluent g

• Blended effluent

^
D ^
—i
I
Q
5 D %
m Q
n co
§
•_• — » . . . ... . .
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
UV-254(1/cm)
i
to
to
0 8-,
D
D
6 - " "
^ D
=d D Q 2"
O) • O D)
-3 D ^
erf ^
m 4 " ^
1- X
CO . CO
CO CO
2- D

•
0 -i • • i • 	 B-i 	 1 	 1 	 1 	 1 — i 	 r-i 	 1 	 n


6 -
•
5 -

•
4 -

3 -

2 -
-
1 -

D D
D
D
D D

g
D

D



D
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
UV-254 (1 /cm)



80

70
60


50

40
30
20

10
n


D
D

_ n
D D
% •

9
D

D •
B "
•
      0.00    0.01      0.02     0.03     0.04     0.05     0.06     0.07
                               UV-254(1/cm)
0.00     0.01     0.02     0.03    0.04    0.05    0.06     0.07
                       UV-254 (1/cm)
Figure D-63 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 7

-------
                    3-1
                 m
                 o
                 Q
                 co  .
                 Q  1
                 CO
                               D Single contactor effluent


                               • Blended effluent                D


                                           D              „
                     0.00
                                        0.01                 0.02

                                             UV-254(1/cm)
                                                                                           1  i
                                                                        CO


                                                                        o

                                                                        CO
                                                                        Q
                                                                        CO
                                                               0.03          Q.OO
                                                                                               0.01
-O—iD—D-

   0.02
                                                                                                    UV-254 (1/cm)
                                                                                                                                      0.03
to
to
                   15-,
^1  10 H
D)




|   \


CO

CO   5 "
                     0.00
                                        0.01                0.02


                                             UV-254 (1/cm)
                                                                                           20 n
                                                                                           15-
                                                                                         D)



                                                                                        O5
                                                                                           10 -
                                                                                        CO
                                                                                        Q
                                                               0.03           Q.OO
                                                                                                0.01                0.02


                                                                                                    UV-254(1/cm)
                                                                                                                                      0.03
             Figure D-64 HAA correlations based on GAC effluent UV-254 absorbance for single contactor and blended effluents for Water 8

-------
This page intentionally left blank.
             -222-

-------
Appendix E: Logistic Function Model Curve Fits
                              -223-

-------
   3 -
 o
'•a 2
 o>
 o
 o
O
   1 -
         TOC
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.966)
 O   Blended effluent
 - - - - Dl prediction
                       O
                                                         EBCT = 20 min.
                                                         c0 = 4.54 mg/L
                        20                  40
                                Scaled operation time (days)
                                                    60
80
Figure E-1  Single contactor and blended effluent TOC breakthrough curves
for Water 1
   0.07
   0.06 -
   0.05 -
 o
   0.04 -
   0.00
            UV254
   D  Single contactor effluent
   	Logistic function best fit (RA2 = 0.992)
   O  Blended effluent
   - - - - Dl prediction
                                                       O. .
                                              O,---'
                          20                 40
                                  Scaled operation time (days)
                                                                    EBCT = 20 min.
                                                                    c0 = 0.094 1/cm
                                                     60
80
Figure E-2 Single contactor and blended effluent UV254 breakthrough
curves for Water 1
                                       -224-

-------
   150
O
o
'-4—'
TO
O>
O
c
o
O
   100 -
50 -
       SDS-TOX

          D   Single contactor effluent
        	Logistic function best fit (RA2 = 0.997)
          O   Blended effluent
        	Dl prediction
                                                                   EBCT = 20 min.
                                                                   C0 = 224 ug/L Cl-
                         20                 40
                                 Scaled operation time (days)
                                                           60
                   80
 Figure E-3 Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 1
   20
o>
o
c
o
O
   15 -
   10 H
    5 -
    0 +0
        SDS-CF
          D  Single contactor effluent
         	Logistic function best fit (RA2 = 0.989)
          O  Blended effluent
          - - - - Dl prediction
                         20
                                        40
                            Scaled operation time (days)
                                                                  EBCT = 20 min.
                                                                  c0 = 34.2 ug/L
60
80
Figure E-4 Single contactor and blended effluent SDS-CF breakthrough
curves for Water 1
                                      -225-

-------
    30
    25 -
    20 -
     SDS-BDCM




    EBCT = 20 min

    C0=  19.3 |jg/L
                                            D  Single contactor effluent

                                            	Logistic function best fit (RA2 = 0.984)

                                            O  Blended effluent

                                            • - - - Dl prediction
                         20                 40                 60

                                 Scaled operation time (days)
                                                                               80
 Figure E-5  Single contactor and blended effluent SDS-BDCM breakthrough
 curves for Water 1
    25
    20 -
 o
 '-4—'
 CD
 t_


 CD
 O
 c
 o
 O
    15 -
10 -
     5 -
       SDS-DBCM

          D   Single contactor effluent

        	Logistic function best fit (RA2 = 0.994)

          O   Blended effluent

        	Dl prediction
                                            40

                                 Scaled operation time (days)
                                                                  EBCT = 20 min.


                                                                  c0 = 28 |jg/L
                                                            60
80
Figure E-6  Single contactor and blended effluent SDS-DBCM breakthrough
curves for Water 1
                                       -226-

-------
    20
         SDS-BF
              D   Single contactor effluent
                 - Logistic function best fit (RA2 = 0.971)
                  Blended effluent
                         20                 40
                                 Scaled operation time (days)
                                                                   EBCT = 20 min.
                                                                   c0 = 3.7 ug/L
                                                  60
80
 Figure E-7 Single contactor and blended effluent SDS-BF breakthrough
 curves for Water 1
    80
    60 -
         SDS-TTHM
 D  Single contactor effluent
	Logistic function best fit (RA2 = 0.977)
 O  Blended effluent
 - - - - Dl prediction
                         20                 40
                                 Scaled operation time (days)
                                                                   EBCT = 20 min.
                                                                   c0 =  85 |jg/L
                                                   60
80
Figure E-8  Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 1
                                       -227-

-------
    4 -
    3 -
 g
 '-4—'
 CD
 O
 O
    1 -
    0 +O
          SDS-MCAA
             D   Single contactor effluent
            	Logistic function best fit (RA2 = NA)
             O   Blended effluent
             - - - - Dl prediction
                               Effluent concentrations were not detected
                                  above the MRL for this parameter
                                                                   EBCT = 20 min.
                                                                   C0 =  BMRL
                -rm-n-rTTi—n	m—m	QT-
                        20                  40
—i-CD-
 60
                                Scaled operation time (days)
 Figure E-9 Single contactor and blended effluent SDS-MCAA breakthrough
 curves for Water 1
                                                                                   80
    12
    10 -
o
|5   6

CD
O
     2 -
     0 +0
      0
          SDS-DCAA
            D   Single contactor effluent
          	Logistic function best fit (RA2 = 0.986)
            O   Blended effluent
          	Dl prediction
                                   P'
                                                     .-O"
                        20                  40
                                Scaled operation time (days)
                                                                   EBCT = 20 min.
                                                                   c0= 12.5 ug/L
 60
80
Figure E-10 Single contactor and blended effluent SDS-DCAA breakthrough
curves for Water 1
                                       -228-

-------
   4 -
O)
•3 3 H
c
g
'-4—'
CD
O
O
   1 -
          SDS-TCAA
    D   Single contactor effluent
   	Logistic function best fit (RA2 = NA)
    O   Blended effluent
    - - - - Dl prediction
                      Insufficient data measured above the MRL
                           to perform curve fit analysis
                                                                     EBCT = 20 min.
                                                                     c0 = 3 ug/L
   0 4o        i—rrn-n-TTT|—n	m—en	QT-
     0                  20                  40                  60                  80
                                Scaled operation time (days)

Figure E-11  Single contactor and blended effluent SDS-TCAA breakthrough
curves for Water 1
   4 -
§  2-I
c
o
O
   1  -
SDS-MBAA
   D   Single contactor effluent
 	Logistic function best fit (RA2 = NA)
   O   Blended effluent
 	Dl prediction
                               Effluent concentrations were not detected
                                  above the MRL for this parameter
                                                                    EBCT = 20 min.
                                                                   C0= BMRL
   0 -K3	1—rm-n-TTTi—n	en—rn	cp	rn	i—CD-
     0                  20                  40                  60                  80
                                Scaled operation time (days)

Figure E-12  Single contactor and blended effluent SDS-MBAA breakthrough
curves for Water 1
                                       -229-

-------
    6 -
        SDS-DBAA
    D  Single contactor effluent
   	Logistic function best fit (RA2 = 0.976)
    O  Blended effluent       ...
    	Dl prediction
                       ,O
                                             D
                                                     TT
                                                       -n
                                                         o
                                                        EBCT = 20 min.
                                                        c0 =  4 |jg/L
                        20                  40
                                Scaled operation time (days)
                                                      60
                   80
 Figure E-13  Single contactor and blended effluent SDS-DBAA breakthrough
 curves for Water 1
    20
    15 -
 o
 '•S3  10
 t_
 O>
 o
 c
 o
 O
     5 -
SDS-HAA5

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.977)
   O   Blended effluent
 	Dl prediction
     0 40
      0
                                  -O
                                             o.----
                                                                  EBCT = 20 min.
                                                                  c0 = 20 |jg/L
                20                 40
                       Scaled operation time (days)
60
80
Figure E-14 Single contactor and blended effluent SDS-HAA5 breakthrough
curves for Water 1
                                       -230-

-------
    6 -
 CD
 o
 c
 o
 O
SDS-BCAA

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.981)
   O   Blended effluent
 	Dl prediction
                                                                  EBCT = 20 min.
                                                                  c0=  7 |jg/L
                20                 40                  60
                        Scaled operation time (days)
                                                                                  80
 Figure E-15  Single contactor and blended effluent SDS-BCAA breakthrough
 curves for Water 1
    25
    20 -
 «=* 15 H
 c
 o
 '-4—'
 CD
 t_

 §  10
 c
 o
 O
     5 -
          SDS-HAA6
             D  Single contactor effluent
       - Logistic function best fit (RA2 = 0.979)
        Blended effluent
        Dl prediction          /D
                                   ,O
                                             o.
                         20                 40
                                 Scaled operation time (days)
                                                                   EBCT = 20 min.
                                                                  c0 = 27 |jg/L
                                                       60
80
Figure E-16 Single contactor and blended effluent SDS-HAA6 breakthrough
curves for Water 1
                                       -231-

-------
    4 -
    3 -
 g
 '-4—'
 CD
 O
 O
    1 -
    0 +O-
      0
SDS-DCBAA
     D   Single contactor effluent
   	Logistic function best fit (RA2 = NA)
     O   Blended effluent
   	Dl prediction
                    Insufficient data measured above the MRL
                          to perform curve fit analysis
                                                                   D
                                                                   O
                                                                    EBCT = 20 min.
                                                                    c0 = 2.3 ug/L
        -rm-n-rTTi—n—m—m-
               20
40
60
80
                                 Scaled operation time (days)
 Figure E-17 Single contactor and blended effluent SDS-DCBAA
 breakthrough curves for Water 1
    4 -
    3 -
 O
 '-4— '
 CD
 t_
 "c
 CD
 O
 c
 o
 O
    1 -
          SDS-CDBAA
    D  Single contactor effluent
    	Logistic function best fit (RA2 = NA)
    O  Blended effluent
    	Dl prediction
                    Effluent concentrations were not detected
                       above the MRL for this parameter
    0 40
      0
        -rrn-n-rTTi—n—rn—m	QT-
               20                  40
         -CD-
   EBCT = 20 min.

   c0=  BMRL
-r-CD	,	
                    60
                    80
                                 Scaled operation time (days)
Figure E-18 Single contactor and blended effluent SDS-CDBAA
breakthrough curves for Water 1
                                       -232-

-------
    4 -
    3 -
 g
 '-4—'
 CD
 O
 O
    1 -
    0 +O
          SDS-TBAA
     D   Single contactor effluent
    	Logistic function best fit (RA2 = NA)
     O   Blended effluent
     	Dl prediction
                      Effluent concentrations were not detected
                         above the MRL for this parameter
                                                                  EBCT = 20 min.
                                                                  C0 =  BMRL
        -rm-n-rTTi—n	m—m	QT-
                20                  40
—i-CD-
 60
                                                                                    80
                                 Scaled operation time (days)
 Figure E-19  Single contactor and blended effluent SDS-TBAA breakthrough
 curves for Water 1
    25
    20 -
 «=* 15 H
 c
 o
 '-4—'
 CD
 t_

 §  10
 c
 o
 O
     5 -
SDS-HAA9
   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.982)^
   O   Blended effluent
 	Dl prediction
                -r—O
                      .--6
                          20
                                    40
                        Scaled operation time (days)
                                                          EBCT = 20 min.
                                                          c0 = 29 |jg/L
 60
80
Figure E-20 Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 1
                                       -233-

-------
   2.5
   2.0 -
.g
"co
 o
O
   1.5 -
   1.0 H
   0.5 -
   0.0
         TOO
         D   Single contactor effluent
        	Logistic function best fit (RA2 = 0.972)
         O   Blended effluent
         - - - - Dl prediction
                      50
                                                                    EBCT = 20 min.
                                                                    co =2.6 mg/L
                                  100             150
                              Scaled operation time (days)
200
250
Figure E-21 Single contactor and blended effluent TOC breakthrough curves
for Water 2
   0.040
   0.030 -
 o
 o>
 o
 c
 CD
.Q
 O
 
-------
   175
   150 -
-T 125 -
O
   100 -
SDS-TOX

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.991)
   O   Blended effluent
 	Dl prediction
                                                                   EBCT = 20 min.
                                                                   c0 =220 ug/L Cl-
                     50
                         100            150
                     Scaled operation time (days)
200
250
 Figure E-23  Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 2
   25
   20 -
   15 -
o
'-4—'
TO
        SDS-CF
   D  Single contactor effluent
  	Logistic function best fit (RA2 = 0.989)
   O  Blended effluent
  - - - - Dl prediction
                                                                 EBCT = 20 min.
                                                                 c0 = 41.9 ug/L
                     50
                         100             150
                     Scaled operation time (days)
200
250
Figure E-24 Single contactor and blended effluent SDS-CF breakthrough
curves for Water 2
                                      -235-

-------
    30
    25 -
 O)
 o
 '•^
 CD
    20 -
 SDS-BDCM


EBCT = 20 min.
c0 =19.8 |jg/L
50
                                                 Cr
                                            D   Single contactor effluent
                                           	Logistic function best fit (RA2 = 0.995)
                                            O   Blended effluent
                                            - - - - Dl prediction
                                    100            150
                                Scaled operation time (days)
                                                          200
               250
 Figure E-25 Single contactor and blended effluent SDS-BDCM breakthrough
 curves for Water 2
    35
    30 -
    25 -
    20 -
  SDS-DBCM
      D  Single contactor effluent
   	Logistic function best fit (RA2 = 0.995)
      O  Blended effluent
   	Dl prediction
                      o-
                     50
                                                         EBCT = 20 min.

                                                         c0= 31.7 |jg/L
                            100             150
                        Scaled operation time (days)
200
                                                            250
Figure E-26 Single contactor and blended effluent SDS-DBCM breakthrough
curves for Water 2
                                      -236-

-------
   14
   12 -
   10 -
 g
 '-4—'
 CD
 CD
 O
 c
 O
 O
    6 -
    4 -
    2 -
         SDS-BF
                                       O
                                                        EBCT = 20 min.

                                                        c0 = 3.7|jg/L
                                   D   Single contactor effluent

                                   	Logistic function best fit (RA2 = 0.91)

                                   O   Blended effluent

                                   • - - - Dl prediction
                     50
                           100             150

                       Scaled operation time (days)
200
250
Figure E-27 Single contactor and blended effluent SDS-BF breakthrough
curves for Water 2
   100
    75 -
 o

 'CD  50 H
 t_

 CD
 O
 c
 o
 O
    25 -
SDS-TTHM




EBCT = 20 min

c0 = 97 |jg/L
                      50
                                              D   Single contactor effluent

                                              	Logistic function best fit (RA2 = 0.992)

                                              O   Blended effluent

                                              • - - - Dl prediction
                           100            150

                        Scaled operation time (days)
200
250
Figure E-28 Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 2
                                       -237-

-------
    2.0
    1.5 -
    1-0 H
 O>
 o
 c
 o
 O
    0.5 -
    o.o H
          SDS-MCAA
          EBCT = 20 min.
          c0  = BMRL
                      50
Effluent concentrations were not detected
   above the MRL for this parameter
                  D   Single contactor effluent
                 	Logistic function best fit (RA2 = NA)
                  O   Blended effluent
                  - - - - Dl prediction
                 -CD—i	CD	1	Oi	
     100            150
 Scaled operation time (days)
200
250
 Figure E-29  Single contactor and blended effluent SDS-MCAA breakthrough
 curves for Water 2
    10
     6 -
          SDS-DCAA

             D  Single contactor effluent
          	Logistic function best fit (RA2 = 0.975)
            O  Blended effluent
            - - - - Dl prediction
                                                                   EBCT = 20 min.
                                                                  c0= 14 ug/L
                     50
     100             150
 Scaled operation time (days)
200
250
Figure E-30 Single contactor and blended effluent SDS-DCAA breakthrough
curves for Water 2
                                       -238-

-------
   3.0
   2.5 -
=d  2.0 -
O)

c
g
'•S3  1-5 H
O>
o
§  1.0 H
   0.5 -
   0.0 -I
          SDS-TCAA


          EBCT = 20 min.
          c0 =5 |jg/L
                     50
                              i—O-
                                              D  Single contactor effluent
                                             	Logistic function best fit (RA2 = 0.968)
                                              O  Blended effluent
                                             - - - - Dl prediction
                                    100             150
                                 Scaled operation time (days)
200
250
Figure E-31 Single contactor and blended effluent SDS-TCAA breakthrough
curves for Water 2
   2.0
   1.5 -
o
'•§  1.0 H
-i—•

I
o
O
   0.5 -
          SDS-MBAA
          EBCT = 20 min.
          C0= BMRL
                     50
                              Effluent concentrations were not detected
                                 above the MRL for this parameter
                                                  D  Single contactor effluent
                                                 	Logistic function best fit (RA2 = NA)
                                                  O  Blended effluent
                                                  - - - - Dl prediction
                                                 -CD—,	CE	1	Oi	
                                    100             150
                                 Scaled operation time (days)
200
250
Figure E-32 Single contactor and blended effluent SDS-MBAA breakthrough
curves for Water 2
                                       -239-

-------
                                                                         o
                                                                         D
                                                -o-
                                             D  Single contactor effluent
                                             	Logistic function best fit (RA2 = 0.985)
                                             O  Blended effluent
                                             	Dl prediction
                    50
                          100             150
                       Scaled operation time (days)
200
250
 Figure E-33  Single contactor and blended effluent SDS-DBAA breakthrough
 curves for Water 2
    25
    20 -
 «=* 15 H
 c
 o
 '-4—'
 CD
 t_

 §  10
 c
 o
 O
     5 -
SDS-HAA5

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.978)
            O   Blended effluent
            - - - - Dl prediction
                                                                  EBCT = 20 min.
                                                                  c0 =  24  |jg/L
                     50
                           100            150
                       Scaled operation time (days)
200
250
Figure E-34 Single contactor and blended effluent SDS-HAA5 breakthrough
curves for Water 2
                                       -240-

-------
    10
     6 -
 g
 '-4—'
 CD
 O
 O
     2 -
         SDS-BCAA


         EBCT = 20 min.
         c0 =9 Mg/L
             D   Single contactor effluent
             	Logistic function best fit (RA2 = 0.977)
             O   Blended effluent
             	Dl prediction
                     50
    100             150
Scaled operation time (days)
200
250
 Figure E-35 Single contactor and blended effluent SDS-BCAA breakthrough
 curves for Water 2
    30
    25 -
    20 -
          SDS-HAA6
             D  Single contactor effluent
            	Logistic function best fit (RA2 = 0.98)
             O  Blended effluent
             - - - - Dl prediction
                                                                  EBCT = 20 min.
                                                                  c0 =  34
                     50
    100             150
Scaled operation time (days)
200
250
Figure E-36  Single contactor and blended effluent SDS-HAA6 breakthrough
curves for Water 2
                                      -241-

-------
   3.0
   2.5 -
   2.0 -
 O)

 c
 o
 '•^ 1
 CD  I-
 CD
 O
    .
   0.5 -
   o.o H
SDS-DCBAA
EBCT = 20 min.
co =3 |jg/L
                     50
                                                                      D

                                                                     -O
                                  D   Single contactor effluent
                                 	Logistic function best fit (RA2 = 0.952)
                                  O   Blended effluent
                                  	Dl prediction
                         100            150
                      Scaled operation time (days)
200
250
 Figure E-37 Single contactor and blended effluent SDS-DCBAA
 breakthrough curves for Water 2
^.u -

2.0 -
^j
co
5 1.5 -
o
'-4— '
CD
§ 1.0-
o
O
0.5 -
-
n n -
SDS-CDBAA 0
EBCT = 20 min.
n n
c0= BMRL

Insufficient data measured above the MRL
to perform curve fit analysis
O

D Single contactor effluent
1 — —j^i: — f. .-.-.i: — ,- UQI_I f\i /f?AO
Lugisiic Tunciion uesi TII ^i\ z
O Blended effluent
	 Dl prediction

O





- M^
- INr\j

                     50
                         100            150
                      Scaled operation time (days)
200
250
Figure E-38 Single contactor and blended effluent SDS-CDBAA
breakthrough curves for Water 2
                                     -242-

-------
    2.0
    1.5 -
    1-0 H
 O>
 o
 c
 o
 O
    0.5 -
    o.o H
          SDS-TBAA
          EBCT = 20 min.
          c0  = BMRL
                      50
                     Effluent concentrations were not detected
                        above the MRL for this parameter
                                        -en
 Single contactor effluent
- Logistic function best fit (RA2 = NA)
 Blended effluent
• Dl prediction
T	CD	1	Oi	
                            100             150
                        Scaled operation time (days)
              200
250
 Figure E-39  Single contactor and blended effluent SDS-TBAA breakthrough
 curves for Water 2
    30
    25 -
    20 -
SDS-HAA9
 EBCT = 20 min
 c0 = 37 |jg/L
                                               D   Single contactor effluent
                                               	Logistic function best fit (RA2 = 0.988)
                                               O   Blended effluent
                                               	Dl prediction
                     50
                           100             150
                        Scaled operation time (days)
              200
250
Figure E-40 Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 2
                                       -243-

-------
   2.0
   1.5 -
 o
'•a
 CD
 o
 c
 o
O
   0.5 -
   0.0
         TOO
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.986)
 O   Blended effluent
 - - - - Dl prediction
                                                                    EBCT = 20 min.
                                                                    c0 = 2.35 mg/L
                   50
                    100          150          200
                     Scaled operation time (days)
250
300
Figure E-41 Single contactor and blended effluent TOC breakthrough curves
for Water 3
   0.030
   0.025 -
 E 0.020 -
 o
 o>
 o
 CD
.Q
   0.015 -
   0.010 H
   0.005 -
   0.000
            UV254
   D  Single contactor effluent
   	Logistic function best fit (RA2 = 0.995)
   O  Blended effluent
   - - - - Dl prediction
                     50
                                                                    EBCT = 20 min.
                                                                    c0 = 0.048 1/cm
                     100          150         200
                      Scaled operation time (days)
 250
300
Figure E-42 Single contactor and blended effluent UV254 breakthrough
curves for Water 3
                                       -244-

-------
   175
   150 -
-T 125 -
O
o
'-4—'
TO
   100 -
    75 -
o>
o
3   so H
    25 -
     0 -I
SDS-TOX

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.997)
   O   Blended effluent
 	Dl prediction
                                                                   EBCT = 20 min.
                                                                   C0 = 255 ug/L Cl-
                   50
                    100          150         200
                      Scaled operation time (days)
250
300
 Figure E-43  Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 3
   20
   15 -
o
'•a  10 H
t_
o>
o
c
o
O
    5 -
    o ^
        SDS-CF
   D   Single contactor effluent
  	Logistic function best fit  (RA2 = 0.99)
   O   Blended effluent
   - - - - Dl prediction
                                         O'
                                                          O
                  50
                    100          150          200
                     Scaled operation time (days)
                                                          o
                                                       EBCT = 20 min.
                                                       c0 = 60.3 ug/L
250
300
Figure E-44 Single contactor and blended effluent SDS-CF breakthrough
curves for Water 3
                                      -245-

-------
   50
   40 -
   30 -
   20
 c
 o
O
   10 -
         SDS-BDCM

              D   Single contactor effluent
        Logistic function best fit (RA2 = 0.99)

    O   Blended effluent

        Dl prediction
                                                                  EBCT = 20 min.

                                                                  C0 = 36.2 ug/L
                  50
                    100          150         200

                      Scaled operation time (days)
250
300
 Figure E-45  Single contactor and blended effluent SDS-BDCM breakthrough
 curves for Water 3
   35
   30 -
   25 -
 c 20
 o
 '
 g 15
 o
 o

 0 10 H
    5 -
SDS-DBCM

   D   Single contactor effluent

 	Logistic function best fit (RA2 = 0.996)

   O   Blended effluent

 	Dl prediction
                                                -' O
                                                           o





                                                       EBCT = 20 min.

                                                       c0 = 44.5 ug/L
                  50
                    100          150          200

                      Scaled operation time (days)
250
300
Figure E-46  Single contactor and blended effluent SDS-DBCM breakthrough
curves for Water 3
                                      -246-

-------
   40
   30 -
g
'•£  20 H
O>
o
c
o
O
   10 -
SDS-BF

EBCT = 20 min.
C0 =  11.8 |jg/L
                                            D  Single contactor effluent
                                            	Logistic function best fit (RA2 = 0.984)
                                            O  Blended effluent
                                            - - - - Dl prediction
                  50
                      100          150          200
                       Scaled operation time (days)
250
300
Figure E-47 Single contactor and blended effluent SDS-BF breakthrough
curves for Water 3
   125
   100 -
    75 -
§   50
c
o
O
    25 -
 SDS-TTHM
   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.986)
   O   Blended effluent
 	Dl prediction
                                                                  EBCT = 20 min.

                                                                  c0 =  154 ug/L
                   50
                      100          150          200
                        Scaled operation time (days)
250
300
Figure E-48 Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 3
                                       -247-

-------
    2.0
    1.5 -
 g
 '•SS  1-0 H
 O>
 o
 c
 o
 O
    0.5 -
    o.o -b-
       0
          SDS-MCAA
      D  Single contactor effluent
     	Logistic function best fit (RA2 = NA)
      O  Blended effluent
      - - - - Dl prediction
          50
                     Effluent concentrations were not detected
                        above the MRL for this parameter
100          150         200
  Scaled operation time (days)
EBCT = 20 min.
C0 = BMRL
  HO—
   250
300
 Figure E-49 Single contactor and blended effluent SDS-MCAA breakthrough
 curves for Water 3
    6 -
 o
 '•SS  4H
 o>
 o
 c
 o
 O
    2 -
SDS-DCAA

  D   Single contactor effluent
	Logistic function best fit (RA2 = 0.995)
  O   Blended effluent
	Dl prediction
                                      O—O—r-
                  50
                     100          150          200
                       Scaled operation time (days)
                                                                   EBCT = 20 min.
                                                                  c0=  15.7ug/L
                                      250
                300
Figure E-50 Single contactor and blended effluent SDS-DCAA breakthrough
curves for Water 3
                                       -248-

-------
   3 -
O)
I
o
O
   1  -
   o -b-
     o
SDS-TCAA
   D   Single contactor effluent
 	Logistic function best fit (RA2 = NA)
   O   Blended effluent
 	Dl prediction
        50
                     Insufficient data measured above the MRL
                           to perform curve fit analysis
                                                            D
                                                            O
                                                                   EBCT = 20 min.
                                                                   c0 =  5 ug/L
100          150          200
  Scaled operation time (days)
250
300
 Figure E-51  Single contactor and blended effluent SDS-TCAA breakthrough
 curves for Water 3
   2.0
   1.5 -
o
'•S3  1-0 H
o>
o
c
o
O
   0.5 -
   0.0 -D-
      0
SDS-MBAA

    D   Single contactor effluent
  	Logistic function best fit (RA2 = NA)
    O   Blended effluent
  	Dl prediction
                     Effluent concentrations were not detected
                         above the MRL for this parameter
                                                                   EBCT = 20 min.
                                                                   c0= BMRL
         50
 100          150          200
   Scaled operation time (days)
 250
300
 Figure E-52 Single contactor and blended effluent SDS-MBAA breakthrough
 curves for Water 3
                                       -249-

-------
   14
   12 -
   10 -
.g
"co
 O>
 o
 c
 o
O
    6 -
    4 -
    2 -
    0 -
        SDS-DBAA


        EBCT = 20 min
        c0 = 9.7 |jg/L
                                             D   Single contactor effluent
                                             	Logistic function best fit (RA2 = 0.859)
                                             O   Blended effluent
                                             	Dl prediction
                  50
                              100          150          200
                                Scaled operation time (days)
250
300
 Figure E-53 Single contactor and blended effluent SDS-DBAA breakthrough
 curves for Water 3
   30
   25 -


§ 20 -
3.
o
'•SS 15 H
 o>
 o
 3 1(H
    5 -
    0 ^
         SDS-HAA5

            D  Single contactor effluent
         	Logistic function best fit (RA2 = 0.977)
            O  Blended effluent
         	Dl prediction                  D
                                                                D    D
                -3-
                                                                  EBCT = 20 min.
                                                                 c0 =  30 ug/L
                  50
                              100          150          200
                                Scaled operation time (days)
250
300
Figure E-54 Single contactor and blended effluent SDS-HAA5 breakthrough
curves for Water 3
                                       -250-

-------
   10
 o
O
    6 -
    4 -
    2 -
    0 -
         SDS-BCAA
             D   Single contactor effluent
            	Logistic function best fit (RA2 = 0.92
             O   Blended effluent
             - - - - Dl prediction
                  50
                              100          150          200
                                Scaled operation time (days)
                                                                  EBCT = 20 min.
                                                                  C0 = 12.7 ug/L
250
300
 Figure E-55  Single contactor and blended effluent SDS-BCAA breakthrough
 curves for Water 3
   35
   30 -
   25 -
 c 20 H
 o
%  15 H
o
o
0  10 H
    5 -
    o ^
          SDS-HAA6
           D   Single contactor effluent
         	Logistic function best fit (RA2 = 0.963P
           O   Blended effluent
          	Dl prediction
                                         O
                                                  O
                                                                D    D
                  50
                              100          150         200
                                Scaled operation time (days)
                                                                  EBCT = 20 min.

                                                                  c0 =  43 |jg/L
250
300
Figure E-56  Single contactor and blended effluent SDS-HAA6 breakthrough
curves for Water 3
                                      -251-

-------
   2.5
   2.0 -
   1.5 -
CD
o
c
o
O
   0.5 -
   0.0 -I
         SDS-DCBAA


         EBCT = 20 min.
         C0 =  4 |jg/L
                                             D   Single contactor effluent
                                            	Logistic function best fit (RA2 = 0.531)
                                             O   Blended effluent
                                             	Dl prediction
                   50
                               100          150          200
                                Scaled operation time (days)
                                      250
             300
 Figure E-57  Single contactor and blended effluent SDS-DCBAA
 breakthrough curves for Water 3
   3.0
   2.5 -
   2.0 -
CD  1-5 H
i_
-I—<
cz
CD
O

3  10H
   0.5 -
   0.0 -D-
      0
          SDS-CDBAA
           EBCT = 20 min.
          C0=  2.5|jg/L
                   50
                            Insufficient data measured above the MRL
                                  to perform curve fit analysis    D
                                                  D   Single contactor effluent
                                                 	Logistic function best fit (RA2 = NA)
                                                  O   Blended effluent
                                                  	Dl prediction
100          150         200
  Scaled operation time (days)
250
300
Figure E-58 Single contactor and blended effluent SDS-CDBAA
breakthrough curves for Water 3
                                       -252-

-------
   2.0
   1.5 -
 g
'•SS 1-0 H
 O>
 o
 c
 o
O
   0.5 -
   o.o -b-
      0
          SDS-TBAA
      D   Single contactor effluent
     	Logistic function best fit (RA2 = NA)
      O   Blended effluent
      	Dl prediction
          50
                       Effluent concentrations were not detected
                          above the MRL for this parameter
100          150         200
  Scaled operation time (days)
EBCT = 20 min.
C0 =  BMRL

   HO—
    250
300
 Figure E-59  Single contactor and blended effluent SDS-TBAA breakthrough
 curves for Water 3
   50
   40 -
   30 -
 § 20
 c
 o
 O
   10 -
    o ^
SDS-HAA9
    D  Single contactor effluent
 	Logistic function best fit (RA2 = 0.965)
    O  Blended effluent
 	Dl prediction
                   50
                                      O
                                                                 n    n
                                                         EBCT = 20 min.

                                                         c0 = 49 |jg/L
                      100          150          200
                        Scaled operation time (days)
                                      250
                 300
Figure E-60 Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 3
                                       -253-

-------
   2.5
   2.0 -
.g
"co
 o
O
   1.5 -
   1.0 H
   0.5 -
   0.0
         TOO
            D   Single contactor effluent
           	Logistic function best fit (RA2 = 0.984)
            O   Blended effluent
            - - - - Dl prediction
                                                                     EBCT = 20 min.
                                                                     c0 = 2.98 mg/L
                          50                  100                 150
                                 Scaled operation time (days)
                                                                                   200
Figure E-61 Single contactor and blended effluent TOC breakthrough curves
for Water 4
   0.040

   0.035

   0.030

P  0.025
 o>
 o
 CD
.Q
   0.020 -
   0.015 -
   0.010

   0.005

   0.000
            UV254
               D   Single contactor effluent
             	Logistic function best fit (RA2 = 0.998)
               O   Blended effluent
             	Dl prediction
                                                              9--
                                                                    EBCT = 20 min.
                                                                    c0 = 0.065 1/cm
                           50                 100                150
                                  Scaled operation time (days)
                                                                                   200
Figure E-62 Single contactor and blended effluent UV254 breakthrough
curves for Water 4
                                       -254-

-------
   175
   150 -
-T 125 -
O
   100 -
SDS-TOX

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.999)
   O   Blended effluent
 	Dl prediction
                                                                   EBCT = 20 min.
                                                                   C0 = 288 ug/L Cl-
                         50                100
                                Scaled operation time (days)
                                                   150
200
 Figure E-63  Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 4
   30
   25 -
   20 -
        SDS-CF
   D  Single contactor effluent
  	Logistic function best fit (RA2 = 0.998)
   O  Blended effluent
  - - - - Dl prediction
                                                                 EBCT = 20 min.
                                                                 c0 = 50.7 ug/L
              50                 100                150
                     Scaled operation time (days)
                                                                                 200
Figure E-64 Single contactor and blended effluent SDS-CF breakthrough
curves for Water 4
                                      -255

-------
   3 -
 g
 73 2H
 O>
 o
 c
 o
O
   1 -
        SDS-BDCM
        EBCT = 20 min
        c0 =  2.2 |jg/L
                                               D   Single contactor effluent
                                               	Logistic function best fit (RA2 = 0.984)
                                               O   Blended effluent
                                               	Dl prediction
                                 100
                     Scaled operation time (days)
                                                              150
200
 Figure E-65 Single contactor and blended effluent SDS-BDCM breakthrough
 curves for Water 4
   12
   10 -
    8 -
SDS-DBCM
   D   Single contactor effluent
 	Logistic function best fit  (RA2 = 0.993)
   O   Blended effluent
 	Dl prediction
                                                                 EBCT = 20 min.
                                                                 c0 = 15.1 ug/L
               50                 100
                      Scaled operation time (days)
                                                              150
200
Figure E-66 Single contactor and blended effluent SDS-DBCM breakthrough
curves for Water 4
                                      -256-

-------
   20
   15 -
g
'•£  10 H
 O>
 o
 c
 o
O
    5 -
         SDS-BF
    0 -fc>
      0
              D   Single contactor effluent
             	Logistic function best fit (RA2 = NA)
              O   Blended effluent
              - - - - Dl prediction
                         50
                                Effluent concentrations were not detected
                                   above the MRL for this parameter
           100
Scaled operation time (days)
    EBCT = 20 min.
    C0 = BMRL

150
200
 Figure E-67  Single contactor and blended effluent SDS-BF breakthrough
 curves for Water 4
   50
   40 -
   30 -
         SDS-TTHM
             D   Single contactor effluent
            	Logistic function best fit (RA2 = 0.995)
             O   Blended effluent
             - - - - Dl prediction
                                                                   EBCT = 20 min.
                                                                   c0 = 68 ug/L
                         50                 100
                                 Scaled operation time (days)
                                                               150
                                                  200
Figure E-68 Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 4
                                       -257-

-------
   2.0
   1.5 -
g
'•£  1.0 H
 O>
 o
 c
 o
O
   0.5 -
   o.o -k>
      0
          SDS-MCAA

             D   Single contactor effluent
           	Logistic function best fit (RA2 = NA)
             O   Blended effluent
             	Dl prediction
                               Effluent concentrations were not detected
                                  above the MRL for this parameter
                         50
           100
Scaled operation time (days)
    EBCT = 20 min.
    C0 = BMRL

150
200
 Figure E-69  Single contactor and blended effluent SDS-MCAA breakthrough
 curves for Water 4
   15 -
   10 -
          SDS-DCAA

             D  Single contactor effluent
          	Logistic function best fit (RA2 = 0.99)
             O  Blended effluent
             	Dl prediction
                                                           ,-O'
                                                                  EBCT = 20 min.

                                                                  c0 = 20.3 ug/L
                         50                 100                 150
                                Scaled operation time (days)
                                                                                  200
Figure E-70 Single contactor and blended effluent SDS-DCAA breakthrough
curves for Water 4
                                       -258-

-------
   25
   20 -
O)
a 15 H
c
.0
I
§ 10
c
o
O
    5 -
 SDS-TCAA
   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.981)
   O   Blended effluent
 	Dl prediction
                         50
                                                                   EBCT = 20 min.
                                                                   C0 = 30.7 |jg/L
                                 100
                      Scaled operation time (days)
                              150
                   200
 Figure E-71 Single contactor and blended effluent SDS-TCAA breakthrough
 curves for Water 4
   2.0
   1.5 -
o
'•S3  1-0 H
o>
o
c
o
O
   0.5 -
   0.0 -K>
      0
SDS-MBAA
   D  Single contactor effluent
	Logistic function best fit (RA2 = NA)
   O  Blended effluent
	Dl prediction
                    Effluent concentrations were not detected
                       above the MRL for this parameter
               50
           100
Scaled operation time (days)
    EBCT = 20 min.
   c0= BMRL


150
 Figure E-72 Single contactor and blended effluent SDS-MBAA breakthrough
 curves for Water 4
200
                                       -259-

-------
   4 -
   3 -
 O>
 o
 c
 o
O
   1 -
        SDS-DBAA
 D  Single contactor effluent
	Logistic function best fit (RA2 = NA)
 O  Blended effluent
 - - - - Dl prediction
                              Insufficient data measured above the MRL
                                    to perform curve fit analysis
                                                                  EBCT = 20 min.
                                                                  CD = 1 ug/L
                                                           -CD-
                        50
                               100
                   Scaled operation time (days)
                              150
200
 Figure E-73  Single contactor and blended effluent SDS-DBAA breakthrough
 curves for Water 4
   40
   30 -
 o
 '•§ 20 H
 -H-<
 c
 O>
 o
 c
 o
 O
   10 -
         SDS-HAA5

            D   Single contactor effluent
          - Logistic function best fit (RA2 = 0.982)
            O   Blended effluent
                Dl prediction
                         50
                                                                  EBCT = 20 min.
                                                                  c0 = 52 ug/L
           100                 150
Scaled operation time (days)
                                                                      200
Figure E-74 Single contactor and blended effluent SDS-HAA5 breakthrough
curves for Water 4
                                       -260-

-------
   4 -
   3 -
 O>
 O
 c
 o
O
   1 -
SDS-BCAA

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.954)
           O  Blended effluent
           	Dl prediction
                                                                 EBCT = 20 min.
                                                                 C0 =  4.3 |jg/L
                50                 100                150
                        Scaled operation time (days)
                                                                                 200
 Figure E-75 Single contactor and blended effluent SDS-BCAA breakthrough
 curves for Water 4
   50
   40 -
   30 -
          SDS-HAA6
            D  Single contactor effluent
    	Logistic function best fit (RA2 = 0.98)
     O  Blended effluent
     	Dl prediction
                                                                  EBCT = 20 min.

                                                                  c0 =  55  pg/L
                         50
                                   100
                        Scaled operation time (days)
150
200
Figure E-76  Single contactor and blended effluent SDS-HAA6 breakthrough
curves for Water 4
                                      -261-

-------
   4 -
.0
"co
O>
o
o 2 -
O   1
         SDS-DCBAA
            D  Single contactor effluent
           	Logistic function best fit (RA2 = 0.983)
            O  Blended effluent
            - - - - Dl prediction
                        50
                                                            O
                                                                              O
                                                                   EBCT = 20 min.
                                                                   C0 = 6.7 ug/L
                                           100                 150
                               Scaled operation time (days)
                                                 200
Figure E-77 Single contactor and blended effluent SDS-DCBAA
breakthrough curves for Water 4
   1.0
o
'•S3  0.5
 o>
 o
 c
 o
O
          SDS-CDBAA
   o.o -k>
      o
               D  Single contactor effluent
              	Logistic function best fit (RA2 = NA)
               O  Blended effluent
               - - - - Dl prediction
                               Effluent concentrations were not detected
                                  above the MRL for this parameter
                         50
           100
Scaled operation time (days)
   EBCT = 20 min.
   c0= BMRL

150
200
Figure E-78 Single contactor and blended effluent SDS-CDBAA
breakthrough curves for Water 4
                                       -262-

-------
   2.0
   1.5 -
 g
'•SS 1-0 H
 O>
 o
 c
 o
O
   0.5 -
   o.o -k>
      0
          SDS-TBAA
     D   Single contactor effluent
    	Logistic function best fit (RA2 = NA)
     O   Blended effluent
     	Dl prediction
                       Effluent concentrations were not detected
                          above the MRL for this parameter
                 50
           100
Scaled operation time (days)
  EBCT = 20 min.
  C0 =  BMRL


150
200
 Figure E-79  Single contactor and blended effluent SDS-TBAA breakthrough
 curves for Water 4
   50
   40 -
   30 -
 o
 '-4—'
 CD
 8 20 H
 o
 O
   10 -
SDS-HAA9
   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.981)
   O   Blended effluent
 	Dl prediction
                                            9-
                         50
                                                             o
                                                                  EBCT = 20 min.
                                                                  c0 = 61 |jg/L
           100
Scaled operation time (days)
                                                       150
                   200
Figure E-80 Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 4
                                       -263-

-------
   3.0
   2.5 -
   2.0 -
 o
'•a 1-
 O>
 o
 o 1.0 -
O
   0.5 -
   0.0
         TOO
         D   Single contactor effluent
        	Logistic function best fit (RA2 = 0.984)
         O   Blended effluent
         - - - - Dl prediction
                                   .O-"
                                        . O'
                                                    o-"
                                                               -o"
                                                                  .---O
                                                                    EBCT = 20 min.
                                                                    c0 = 3.08 mg/L
                 50         100         150        200        250
                                 Scaled operation time (days)
                                                                    300
          350
Figure E-81 Single contactor and blended effluent TOC breakthrough curves
for Water 5
0.040

0.035 -

0.030 -

0.025 -
            UV254
               D   Single contactor effluent
             	Logistic function best fit (RA2 = 0.992)
               O   Blended effluent
             	Dl prediction
                                                                    EBCT = 20 min.
                                                                    c0 = 0.051 1/cm
                   50
                          100        150       200        250
                               Scaled operation time (days)
300
350
Figure E-82 Single contactor and blended effluent UV254 breakthrough
curves for Water 5
                                       -264-

-------
O
   150
   125 -
   100 -
.9.   75 H
"CD
O>
O
c
o
O
50 -
    25 -
       SDS-TOX

          D  Single contactor effluent
       	Logistic function best fit (RA2 = 0.995)
          O  Blended effluent
       	Dl prediction
             50
                                                                   EBCT = 20 min.
                                                                   c0 = 205 ug/L Cl-
                           100        150        200        250
                                Scaled operation time (days)
300
350
 Figure E-83  Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 5
   12
   10 -
'ro   6 H
t_
o>
o
I   «H
    2 -
        SDS-CF
         D  Single contactor effluent
         	Logistic function best fit (RA2 = 0.987)
         O  Blended effluent
         - - - - Dl prediction
                                                          ,-O'
                                                               .--O
                                                                 EBCT = 20 min.
                                                                 c0 = 23.7 ug/L
                50
                       100        150        200        250
                            Scaled operation time (days)
300
350
Figure E-84 Single contactor and blended effluent SDS-CF breakthrough
curves for Water 5
                                      -265-

-------
    20
    15 -
    10 H
 O>
 o
 c
 o
 O
     5 -
         SDS-BDCM
    D  Single contactor effluent
   	Logistic function best fit (RA2 = 0.97)
    O  Blended effluent
    - - - - Dl prediction
                                                                EBCT = 20 min.

                                                                C0 = 10.8 ug/L
                 50
                 100        150        200        250
                      Scaled operation time (days)
300
350
 Figure E-85 Single contactor and blended effluent SDS-BDCM breakthrough
 curves for Water 5
    20
    15 -
 o
 '•S3  10 H
 t_
 o>
 o
 c
 o
 O
     5 -
SDS-DBCM

   D   Single contactor effluent
 	Logistic function best fit  (RA2 = 0.988)
   O   Blended effluent
 	Dl prediction
                                                                 EBCT = 20 min.

                                                                 c0 = 22.7 ug/L
                 50
                 100        150        200        250
                      Scaled operation time (days)
300
350
Figure E-86 Single contactor and blended effluent SDS-DBCM breakthrough
curves for Water 5
                                      -266-

-------
    2 -
         SDS-BF
            D   Single contactor effluent
          	Logistic function best fit (RA2 = 0.969)
            O   Blended effluent
                              n    n
          	Dl prediction
                          O
                                           ft
                                                 D
                                                 n
                                                       O
                                                           -O
                                                                  EBCT = 20 min.
                                                                  C0=  1.2|jg/L
                50
                  100        150        200        250
                       Scaled operation time (days)
300
350
Figure E-87 Single contactor and blended effluent SDS-BF breakthrough
curves for Water 5
    50
    40 -
    30 -
 o
 '-4—'
 CD
 t_

 §  20
 c
 o
 O
    10 -
SDS-TTHM
    D  Single contactor effluent
 	Logistic function best fit (RA2 = 0.979)
    O  Blended effluent
 	Dl prediction
                                                 O
                                                                  EBCT = 20 min.
                                                                  c0 =  58 |jg/L
                 50
                   100         150        200        250
                        Scaled operation time (days)
300
350
Figure E-88 Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 5
                                       -267-

-------
    4 -
    3 -
 g
 '-4—'
 CD
 O
 O
    1 -
    0 -O-
      0
         SDS-MCAA
  D   Single contactor effluent
 	Logistic function best fit (RA2 = NA)
  O   Blended effluent
  	Dl prediction
                 Insufficient data measured above the MRL
                       to perform curve fit analysis
          D       D
    50
      -r-O	rO-
100        150        200        250
     Scaled operation time (days)
EBCT = 20 min.
c0 = 2 ug/L

   300
350
 Figure E-89  Single contactor and blended effluent SDS-MCAA breakthrough
 curves for Water 5
    6 -
 CD
 O
 c
 o
 O
    2 -
    0 -(
          SDS-DCAA
             D  Single contactor effluent
	Logistic function best fit (RA2 = 0.978)
 O  Blended effluent
 - - - - Dl prediction
                                                                   EBCT = 20 min.
                                                                  c0=  10.3 ug/L
                50
               100        150        200        250
                     Scaled operation time (days)
                                            300
              350
Figure E-90 Single contactor and blended effluent SDS-DCAA breakthrough
curves for Water 5
                                       -268-

-------
   6 -
          SDS-TCAA

             D   Single contactor effluent
           	Logistic function best fit (RA2 = 0.989)
             O   Blended effluent
           	Dl prediction
                                                                    EBCT = 20 min.
                                                                    C0 = 12.7 ug/L
               50
                          100        150        200         250
                                Scaled operation time (days)
    300
350
Figure E-91  Single contactor and blended effluent SDS-TCAA breakthrough
curves for Water 5
   2.0
   1.5 -
O
'•§  1.0 H
-i—•

I
O
O
   0.5 -
0.0 -O-
   0
         SDS-MBAA
            D   Single contactor effluent
            	Logistic function best fit (RA2 = NA)
            O  Blended effluent
            - - - - Dl prediction
                               Effluent concentrations were not detected
                                  above the MRL for this parameter
                50
                           100        150        200        250
                                 Scaled operation time (days)
EBCT = 20 min.
c0 = 1 ug/L


     300
350
Figure E-92 Single contactor and blended effluent SDS-MBAA breakthrough
curves for Water 5
                                      -269-

-------
    3 -
 o
 '•^
 CO
 O>
 O
 c
 o
 O
    1 -
        SDS-DBAA
 D  Single contactor effluent
	Logistic function best fit (RA2 = 0.992)
 O  Blended effluent
                 Dl prediction
                                        O
                50
                                                    D
                                                   O
                                                                   EBCT = 20 min.
                                                                   c0 = 2 |jg/L
               100        150        200        250
                    Scaled operation time (days)
                                      300
350
 Figure E-93  Single contactor and blended effluent SDS-DBAA breakthrough
 curves for Water 5
    20
                                                                  EBCT = 20 min
                                                                 c0 =  28 ug/L
    15 -
 o
 '•S3  10
 t_
 o>
 o
 c
 o
 O
     5 -
         SDS-HAA5

            D   Single contactor effluent
          	Logistic function best fit (RA2 = 0.965)
            O   Blended effluent
            - - - - Dl prediction
      150        200        250
Scaled operation time (days)
                                                                       300
                                                                     350
Figure E-94 Single contactor and blended effluent SDS-HAA5 breakthrough
curves for Water 5
                                      -270-

-------
    6 -
 CD
 o
 c
 o
 O
    2 -
    0 •{
        SDS-BCAA


           D  Single contactor effluent

         	Logistic function best fit (RA2 = 0.987)
O   Blended effluent

	Dl prediction
                                                                  EBCT = 20 min.


                                                                  C0 =  7.3 |jg/L
                50
               100        150        200        250


                     Scaled operation time (days)
300
350
 Figure E-95  Single contactor and blended effluent SDS-BCAA breakthrough

 curves for Water 5
    25
    20 -
 «=* 15 H
 c
 o
 '-4—'
 CD
 t_


 §  10 H
 c
 o
 O
     5 -
          SDS-HAA6


            D  Single contactor effluent
	Logistic function best fit (RA2 = 0.976)

 O   Blended effluent

 - - - - Dl prediction
                                                                  EBCT = 20 min.


                                                                  c0 = 34 |jg/L
                 50
                100        150        200        250


                     Scaled operation time (days)
300
350
Figure E-96 Single contactor and blended effluent SDS-HAA6 breakthrough

curves for Water 5
                                       -271-

-------
    10
 g
 '-4—'
 CD
 CD
 O
 c
 O
 O
     6 -
4 -
     2 -
           SDS-DCBAA
             D   Single contactor effluent
           	Logistic function best fit (RA2 = 0.987)
             O   Blended effluent
           	Dl prediction
                                                                   EBCT = 20 min.
                                                                   C0 = 10.7 ug/L
                 50
                       100        150        200        250
                            Scaled operation time (days)
300
350
 Figure E-97 Single contactor and blended effluent SDS-DCBAA
 breakthrough curves for Water 5
    5 -
 O
 •^
 CD
 t_

 CD
 O
    1  -
    0 -{
     SDS-CDBAA
      EBCT = 20 min.
     c0 =  3.7 |jg/L
                50
                                             D   Single contactor effluent
                                             	Logistic function best fit  (RA2 = 0.95)
                                             O   Blended effluent
                                             - - - - Dl prediction
                      100        150        200         250
                           Scaled operation time (days)
300
350
Figure E-98  Single contactor and blended effluent SDS-CDBAA
breakthrough curves for Water 5
                                       -272-

-------
    2.0
    1.5 -
1.0 -
 O>
 o
 c
 o
 O
    0.5 -
    0.0 -O-
       0
       SDS-TBAA
          D   Single contactor effluent
        	Logistic function best fit (RA2 = NA)
          O   Blended effluent
        	Dl prediction
              50
                                Effluent concentrations were not detected
                                   above the MRL for this parameter
100        150        200        250
     Scaled operation time (days)
EBCT = 20 min.
C0 = BMRL


    300
           350
 Figure E-99 Single contactor and blended effluent SDS-TBAA breakthrough
 curves for Water 5
    40
    30 -
 o
 '•S3  20 H
 t_
 o>
 o
 c
 o
 O
    10 -
     0 -O-
      0
         SDS-HAA9
         EBCT = 20 min
         c0 = 48 |jg/L
             50
                                                  D   Single contactor effluent
                                                  	Logistic function best fit (RA2 = 0.99)
                                                  O   Blended effluent
                                                  - - - - Dl prediction
100        150        200        250
     Scaled operation time (days)
300
               350
Figure E-100 Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 5
                                       -273-

-------
   2.5
   2.0 -
.g
"co
o
O
   1.5 -
   1.0 H
   0.5 -
   0.0
         TOO
         D   Single contactor effluent
        	Logistic function best fit (RA2 = 0.992)
         O   Blended effluent
         - - - - Dl prediction
                   50
                            100          150          200
                              Scaled operation time (days)
                                                                    EBCT = 20 min.
                                                                    c0 = 2.64 mg/L
250
300
Figure E-101  Single contactor and blended effluent TOC breakthrough
curves for Water 6
   0.040
0.035 -

0.030 -

0.025 -
o>
o
            UV254
               D   Single contactor effluent
              	Logistic function best fit (RA2 = 0.998)
               O   Blended effluent
               - - - - Dl prediction
                     50
                                                                              O
                                                             o---
                                                      O-'
                                                                    EBCT = 20 min.
                                                                    c0 = 0.06 1/cm
                              100          150          200
                               Scaled operation time (days)
 250
300
Figure E-102  Single contactor and blended effluent UV254 breakthrough
curves for Water 6
                                       -274-

-------
   225

   200

   175
O
   150 -
O)
^ 125
c
o
E 100
O>
o
c
o
O
75 -
    50

    25

     0
       SDS-TOX

          D  Single contactor effluent
       	Logistic function best fit (RA2 = 0.997)
          O  Blended effluent
       	Dl prediction
                   50
                                .0-'
                           100          150          200
                            Scaled operation time (days)
                                                                           .o
EBCT = 20 min.
C0 = 305 ug/L Cl-
  250
300
 Figure E-103  Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 6
   25
   20 -
^ 15 H
c
o
'-4—'
TO
t_

§ 10 H
c
o
O
    5 -
    o ^
        SDS-CF
         D   Single contactor effluent
         	Logistic function best fit (RA2 = 0.991)
         O   Blended effluent
         - - - - Dl prediction
                                                                 EBCT = 20 min.
                                                                 c0 = 55.3 ug/L
                  50
                          100          150          200
                            Scaled operation time (days)
  250
300
Figure E-104 Single contactor and blended effluent SDS-CF breakthrough
curves for Water 6
                                      -275-

-------
   35
   30 -
   25 -
c  20 H
o
g  15 H
o
   10 -
    5 -
    o ^
        SDS-BDCM
        EBCT = 20 min.
        C0 = 27.4 |jg/L
                  50
                                                          O
                                       O  .'
                                          o-
                                    o-
                                     D   Single contactor effluent
                                    	Logistic function best fit (RA2 = 0.979)
                                     O   Blended effluent
                                     	Dl prediction
                    100          150          200

                     Scaled operation time (days)
250
300
 Figure E-105  Single contactor and blended effluent SDS-BDCM
 breakthrough curves for Water 6
    35
    30 -
    25 -
 c  20 H
 o
 o
 o
 0  10 H
     5 -
     o ^
SDS-DBCM

   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.993)
   O   Blended effluent
 	Dl prediction
                                                                 EBCT = 20 min.

                                                                 c0= 41.6ug/L
                   50
                    100          150         200
                      Scaled operation time (days)
250
300
Figure E-106 Single contactor and blended effluent SDS-DBCM
breakthrough curves for Water 6
                                      -276-

-------
   20
   15 -
g
'•   10
O>
o
c
o
O
    5 -
         SDS-BF
         EBCT = 20 min.

         C0 = 3.3 |jg/L
    0 H
      0
     50
                                                           O
                                               D   Single contactor effluent
                                               	Logistic function best fit (RA2 = 0.978)
                                               O   Blended effluent
                                               	Dl prediction
100          150         200
  Scaled operation time (days)
250
300
Figure E-107 Single contactor and blended effluent SDS-BF breakthrough
curves for Water 6
   100
    75 -
o
'CD   50 H
o>
o
c
o
O
    25 -
     0 -I
         SDS-TTHM
 D  Single contactor effluent
	Logistic function best fit (RA2 = 0.98)
 O  Blended effluent
 	Dl prediction
                   50
                                                                  EBCT = 20 min.

                                                                  c0 = 128 ug/L
                  100          150         200
                    Scaled operation time (days)
                                       250
             300
Figure E-108 Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 6
                                       -277-

-------
   2.0
   1.5 -
g

'•S3  1-0 H
 CD
 o
 c
 o
O


   0.5 -
          SDS-MCAA

             D   Single contactor effluent
   0.0 -t>

      0
            	Logistic function best fit (RA2 = NA)

             O   Blended effluent

             	Dl prediction
                   50
                               Effluent concentrations were not detected

                                  above the MRL for this parameter
                                                                   EBCT = 20 min.


                                                                   C0 =  BMRL
100          150         200

  Scaled operation time (days)
250
300
 Figure E-109 Single contactor and blended effluent SDS-MCAA

 breakthrough curves for Water 6
   10
 O)
 3.

 c
 o
'-4—'
 CD
 t_

"£
 CD
 o
 C
 O
O
    6 -
    4 -
    2 -
    0 -I
          SDS-DCAA



            D   Single contactor effluent

          	Logistic function best fit (RA2 = 0.976)

            O   Blended effluent

          	Dl prediction
                                                                   EBCT = 20 min.


                                                                   c0=  17.3ug/L
                                                   T-O-
                   50
                               100          150           200

                                Scaled operation time (days)
                                      250
             300
Figure E-110 Single contactor and blended effluent SDS-DCAA breakthrough

curves for Water 6
                                       -278-

-------
   3.0
   2.5 -
   2.0 -
g
'•SS  1-5 H
O>
o
O
   1 0 -
   LU 1
   0.5 -
   0.0 -I
          SDS-TCAA
             D  Single contactor effluent
           	Logistic function best fit (RA2 = NA)
             O  Blended effluent
           	Dl prediction
                                 Insufficient data measured above the MRL
                                       to perform curve fit analysis
                                                                              O
                                                                     EBCT = 20 min.
                                                                     C0 = 13.3 ug/L
                   50
                         -D	Ol—D—D-Oh-ID—Oh—
                                100           150
200
250
                                                                                 300
                                 Scaled operation time (days)
 Figure E-111 Single contactor and blended effluent SDS-TCAA breakthrough
 curves for Water 6
   2.0
   1.5 -
o
'•§  1.0
-i—•

I
o
O
   0.5 -
o.o -b-
   0
          SDS-MBAA
             D   Single contactor effluent
           	Logistic function best fit (RA2 = NA)
             O   Blended effluent
           	Dl prediction
                            Effluent concentrations were not detected
                                above the MRL for this parameter
                                                                     EBCT = 20 min.
                                                                    c0= BMRL
                   50
                                100          150          200
                                 Scaled operation time (days)
             250
             300
 Figure E-112 Single contactor and blended effluent SDS-MBAA
 breakthrough curves for Water 6
                                        -279-

-------
   10
 c
 CD
 O
 c
 o
O
    6 -
4 -
    2 -
    0 -
         SDS-DBAA

         EBCT = 20 min.


         c0 = 5.7 |jg/L
                                                                            O
                                    o  9,-'
                                            D   Single contactor effluent

                                            	Logistic function best fit  (RA2 = 0.952)

                                            O   Blended effluent

                                            - - - - Dl prediction
                  50
                          100          150         200

                            Scaled operation time (days)
250
300
 Figure E-113 Single contactor and blended effluent SDS-DBAA breakthrough

 curves for Water 6
   25
   20 -
   15 -
 o
'-4—'
 CD
 8 10
 c
 o
O
    5 -
    0 -I
     SDS-HAA5


       D  Single contactor effluent

     	Logistic function best fit (RA2 = 0.969)

       O  Blended effluent

     	Dl prediction
                                D  D
                           "Q   O.O-D-"
                  50
                                                         .-O"
                                                                 EBCT = 20 min.


                                                                 c0 = 36 ug/L
                          100          150          200

                            Scaled operation time (days)
250
300
Figure E-114 Single contactor and blended effluent SDS-HAA5 breakthrough

curves for Water 6
                                      -280-

-------
   10
    6 -
 I  4

 o
O
    2 -
    0 -t>

      0
         SDS-BCAA

            D  Single contactor effluent
               - Logistic function best fit (RA2 = 0.967)

               Blended effluent

               Dl prediction
                                                    o-
        50
100          150          200

  Scaled operation time (days)
                                                                   EBCT = 20 min.

                                                                   c0 =  12.3 ug/L
250
300
 Figure E-115 Single contactor and blended effluent SDS-BCAA breakthrough
 curves for Water 6
   35
   30 -
   25 -
   20 -
 o
'-4—'
 CD
 CD
 O
 C
 O
O
   15 -
   10 -
    5 -
    0 -I
SDS-HAA6

  D   Single contactor effluent

	Logistic function best fit (RA2 = 0.97)

  O   Blended effluent

	Dl prediction
                                                                  EBCT = 20 min.

                                                                  CD =  49 ug/L
                  50
                    100          150          200

                      Scaled operation time (days)
                                      250
             300
Figure E-116 Single contactor and blended effluent SDS-HAA6 breakthrough
curves for Water 6
                                       -281-

-------
   5 -
   4 -
O)
g
I  3
"c
o>
o
c
o  2
O  ^
   1  -
SDS-DCBAA
   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.975)
   O   Blended effluent
 	Dl prediction
   0 -t>
     0
                                          i-O
         50
100          150          200
  Scaled operation time (days)
                                                                   EBCT = 20 min.
                                                                   c0 = 9 |jg/L
250
300
Figure E-117 Single contactor and blended effluent SDS-DCBAA
breakthrough curves for Water 6
   3.5
   3.0 -
   2.5 -
   2.0 -
o
'-4—'
CD
CD
O
C
O
O
   1.5 -
   1.0 -
   0.5 -
   0.0 -I
          SDS-CDBAA
  EBCT = 20 min.
 c0 = 3 |jg/L
                  50
                                            D  D
                                         D   Single contactor effluent
                                        	Logistic function best fit  (RA2 = 0.913)
                                         O   Blended effluent
                                         	Dl prediction
                       100          150          200
                        Scaled operation time (days)
                                       250
             300
Figure E-118 Single contactor and blended effluent SDS-CDBAA
breakthrough curves for Water 6
                                       -282-

-------
   2.0
   1.5 -
g
'•S3  1-0 H
 CD
 o
 c
 o
O
   0.5 -
         SDS-TBAA

             D   Single contactor effluent
           	Logistic function best fit (RA2 = NA)
             O   Blended effluent
           	Dl prediction
   0.0 -t>
      0
                   50
                                Effluent concentrations were not detected
                                   above the MRL for this parameter
                                                                   EBCT = 20 min.
                                                                   C0 = BMRL
100          150          200
  Scaled operation time (days)
250
300
 Figure E-119 Single contactor and blended effluent SDS-TBAA breakthrough
 curves for Water 6
   45

   40 -

   35 -

§ 30 -
3.
I 25-
H—<
CD
£  20 H
CD
O
O  1£

   10 -

    5 -

    0 -
         SDS-HAA9
           D   Single contactor effluent
              - Logistic function best fit (RA2 = 0.966)
               Blended effluent
              • Dl prediction
                                     D  D
                                                     O
                                                                   EBCT = 20 min.
                                                                   c0 = 61 |jg/L
                   50
                               100          150          200
                                Scaled operation time (days)
                                      250
             300
Figure E-120 Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 6
                                       -283-

-------
   4 -
   3 -
.g

"co
 o> 9
 o *-
 c
 o
O
   1 -
         TOO
 D   Single contactor effluent

	Logistic function best fit (RA2 = 0.979)

 O   Blended effluent

 - - - - Dl prediction
                                                                    EBCT = 20 min.

                                                                    c0 = 5.58 mg/L
                  25
                   50           75           100

                    Scaled operation time (days)
125
150
Figure E-121  Single contactor and blended effluent TOC breakthrough

curves for Water 7
   0.07
   0.06 -
   0.05 -
.o

^ 0.04 -
o>
o
c
CD
•9 0.03 -
o
   0.02 H
   0.01 -
   0.00
            UV254
   D  Single contactor effluent

   	Logistic function best fit (RA2 = 0.994)

   O  Blended effluent

   - - - - Dl prediction
                        O.-
                    25
                                                        EBCT = 20 min.


                                                        c0 = 0.109 1/cm
                     50           75          100

                      Scaled operation time (days)
 125
150
Figure E-122  Single contactor and blended effluent UV254 breakthrough

curves for Water 7
                                       -284-

-------
   300
   250 -
O
   200 -
O)
3.

.1 150 H
-I—'
TO
O>
o
c
o
O
100 -
    50 -
     o H
        SDS-TOX

           D  Single contactor effluent
        	Logistic function best fit (RA2 = 0.994)
           O  Blended effluent
        	Dl prediction
                       O
                   25
                                                         O,---
                            50           75          100
                             Scaled operation time (days)
                                                                EBCT = 20 min.
                                                                C0 = 486 ug/L Cl-
125
150
 Figure E-123  Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 7
   20
   15 -
o
'•a  10 H
t_
o>
o
c
o
O
    5 -
    o ^
        SDS-CF
          D   Single contactor effluent
          	Logistic function best fit (RA2 = 0.996)
          O   Blended effluent
          - - - - Dl prediction
     EBCT = 20 min
     c0 = 52.4 ug/L

               25
                               50           75           100
                                Scaled operation time (days)
125
150
Figure E-124 Single contactor and blended effluent SDS-CF breakthrough
curves for Water 7
                                      -285-

-------
    80



    70



    60




    50
    40 -
 CD

 c
 o
 O
30 -
    20



    10



     0
     SDS-BDCM
          D   Single contactor effluent

         	Logistic function best fit (RA2 = 0.991)

          O   Blended effluent

          	Dl prediction
                   25
                            50           75          100

                             Scaled operation time (days)
                                                               EBCT = 20 min.

                                                               C0 = 65.3 |jg/L
125
150
 Figure E-125 Single contactor and blended effluent SDS-BDCM

 breakthrough curves for Water 7
    50
    40 -
 «=* 30 H
 c
 o
 '-4—'
 CD
 t_


 §  20 H
 c
 o
 O
    10 -
     o ^
          SDS-DBCM


              D
              Single contactor effluent

          	Logistic function best fit (RA2 = 0.996)

          O   Blended effluent

          - - - - Dl prediction
                   25
                                                         o-'
                            50           75           100

                             Scaled operation time (days)
                                                             '  o




                                                              EBCT = 20 min.


                                                             c0 =  66.2 |jg/L
125
150
Figure E-126 Single contactor and blended effluent SDS-DBCM

breakthrough curves for Water 7
                                      -286-

-------
   60
   50 -
 O)
   40 -
   30 -
 CD
 O

 o 20 J
   10 -
      SDS-BF

          D   Single contactor effluent

        	Logistic function best fit (RA2 = 0.971)

          O   Blended effluent

        	Dl prediction
                                       O
                                               O
                                                         O
                                                                            O
                                                                   EBCT = 20 min.

                                                                   c0 = 16 ug/L
                  25
                            50           75           100

                             Scaled operation time (days)
125
150
Figure E-127 Single contactor and blended effluent SDS-BF breakthrough
curves for Water 7
   200
 o
 '-4—'
 CD
 t_


 CD
 O
 c
 o
 O
   150 -
100 -
    50 -
     0 ^
         SDS-TTHM
           D  Single contactor effluent

          	Logistic function best fit (RA2 = 0.987)

           O  Blended effluent

          - - - - Dl prediction
                   25
                            50           75          100

                             Scaled operation time (days)
                                                                  EBCT = 20 min.

                                                                  c0 = 200 |jg/L
125
150
Figure E-128 Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 7
                                      -287-

-------
    2.0
    1.5 -
    1-0 H
 O>
 o
 c
 o
 O
    0.5 -
    0.0 -t>
       0
          SDS-MCAA
   D   Single contactor effluent
  	Logistic function best fit (RA2 = NA)
   O   Blended effluent
   - - - - Dl prediction
         25
                     Effluent concentrations were not detected
                        above the MRL for this parameter
                                                         EBCT = 20 min.
                                                         C0 = 5 ug/L
                                                        -O—i	
50           75           100
 Scaled operation time (days)
125
150
 Figure E-129 Single contactor and blended effluent SDS-MCAA
 breakthrough curves for Water 7
    12
    10 -
 o
 'CD   6 H
 t_
 o>
 o
     2 -
     o ^
SDS-DCAA

  D   Single contactor effluent
	Logistic function best fit (RA2 = 0.922)
             O  Blended effluent
             - - - - Dl prediction
                              .o-
                                  . - - -' "o
         -D—O—r-
                   25
                     50           75           100
                      Scaled operation time (days)
                                                         EBCT = 20 min.
                                                        c0 = 25.3 ug/L
                                      125
             150
Figure E-130 Single contactor and blended effluent SDS-DCAA breakthrough
curves for Water 7
                                       -288-

-------
   6 -
g
'•SS  4H
O>
o
c
o
O
          SDS-TCAA
             D  Single contactor effluent
          	Logistic function best fit (RA2 = 0.999)
             O  Blended effluent
          	Dl prediction
                 25
                                                         D
                              50           75           100
                                Scaled operation time (days)
                                                                   o
      .O
                                                                    EBCT = 20 min.
                                                                    C0 = 18 |jg/L
125
150
Figure E-131  Single contactor and blended effluent SDS-TCAA breakthrough
curves for Water 7
   2.0
   1.5 -
g
'•§  1.0 H
-I—•
I
o
o
   0.5 -
o.o -b-
   0
          SDS-MBAA
             D  Single contactor effluent
          	Logistic function best fit (RA2 = NA)
             O  Blended effluent
          	Dl prediction
                               Effluent concentrations were not detected
                                  above the MRL for this parameter
                   25
                               50           75           100
                                 Scaled operation time (days)
 125
150
Figure E-132  Single contactor and blended effluent SDS-MBAA
breakthrough curves for Water 7
                                       -289-

-------
    20
    15 -
    10 H
 O>
 o
 c
 o
 O
     5 -
     o H
         SDS-DBAA

          EBCT = 20 min
          C0 = 16.3 |jg/L
                   25
                                                 D   Single contactor effluent
                                                 	Logistic function best fit (RA2 = 0.976)
                                                 O   Blended effluent
                                                 - - - - Dl prediction
50           75           100
 Scaled operation time (days)
125
150
 Figure E-133 Single contactor and blended effluent SDS-DBAA breakthrough
 curves for Water 7
    40
    30 -
 o
 '•S3  20
 t_
 o>
 o
 c
 o
 O
    10 -
     o ^
         SDS-HAA5

            D   Single contactor effluent
          	Logistic function best fit (RA2 = 0.994)
            O   Blended effluent
            - - - - Dl prediction
                   25
50           75           100
 Scaled operation time (days)
                                                                  EBCT = 20 min.
                                                                 c0 = 65 ug/L
125
150
Figure E-134 Single contactor and blended effluent SDS-HAA5 breakthrough
curves for Water 7
                                      -290-

-------
   16



   14



   12

13"

1 10
c
o
'•^  ft
CD  o
 CD
 O
 c
 O
 O
    6 -



    4



    2



    0
         SDS-BCAA


            D   Single contactor effluent

          - Logistic function best fit (RA2 = 0.952)
            O   Blended effluent

                Dl prediction
                   25
                               50           75          100

                                Scaled operation time (days)
                                                                         .--o
                                                                  EBCT = 20 min.

                                                                  C0 = 23.7 |jg/L
125
150
 Figure E-135 Single contactor and blended effluent SDS-BCAA breakthrough
 curves for Water 7
    50
    40 -
    30 -
 o
 '-4—'
 CD
 t_


 §  20
 c
 o
 O
    10 -
     o ^
          SDS-HAA6

            D  Single contactor effluent

          	Logistic function best fit (RA2 = 0.988)

            O  Blended effluent

          	Dl prediction
                   25
                               50           75           100

                                Scaled operation time (days)
                                                                  EBCT = 20 min.

                                                                  c0 = 85 |jg/L
125
150
Figure E-136 Single contactor and blended effluent SDS-HAA6 breakthrough
curves for Water 7
                                       -291-

-------
    12
    10 -
    8 -
 o
 '•^  R -
 CD  D 1
 CD
 O

    2 -
    o H
SDS-DCBAA
   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.982)
   O   Blended effluent
 	Dl prediction
                  25
                     50           75          100
                      Scaled operation time (days)
                                                                  EBCT = 20 min.
                                                                  c0 = 26.7 ug/L
125
150
 Figure E-137  Single contactor and blended effluent SDS-DCBAA
 breakthrough curves for Water 7
   6 -
    5 -
   4 -
SDS-CDBAA
EBCT = 20 min
c0= 11.7 |jg/L
                  25
                                                 D   Single contactor effluent
                                                 	Logistic function best fit (RA2 = 0.976)
                                                 O   Blended effluent
                                                 • - - - Dl prediction
                    50           75           100
                     Scaled operation time (days)
125
150
Figure E-138 Single contactor and blended effluent SDS-CDBAA
breakthrough curves for Water 7
                                      -292-

-------
    6 -
          SDS-TBAA
          EBCT = 20 min
          c0 =  BMRL
                  25
                                                D   Single contactor effluent
                                                	Logistic function best fit  (RA2 = 0.976)
                                                O   Blended effluent
                                                - - - - Dl prediction
                     50           75           100
                       Scaled operation time (days)
125
150
 Figure E-139  Single contactor and blended effluent SDS-TBAA breakthrough
 curves for Water 7
    80
    60 -
 o
 '•S3  40 H
 t_
 o>
 o
 c
 o
 O
    20 -
     o ^
SDS-HAA9
   D   Single contactor effluent
 	Logistic function best fit (RA2 = 0.98)
   O   Blended effluent
 	Dl prediction
                   25
                      50           75          100
                       Scaled operation time (days)
                                                                 EBCT = 20 min.
                                                                 c0= 124 ug/L
125
150
Figure E-140 Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 7
                                      -293

-------
   2.0
   1.5 -
 o
'•a
 CD
 o
 c
 o
O
   0.5 -
   0.0
         TOO
 D   Single contactor effluent
	Logistic function best fit (RA2 = 0.974)
 O   Blended effluent
 - - - - Dl prediction
                                                                    EBCT = 7.2min.
                                                                    c0 = 2.02 mg/L
                         50                 100
                                 Scaled operation time (days)
                                                    150
200
Figure E-141  Single contactor and blended effluent TOC breakthrough
curves for Water 8
   0.025
   0.020 -
 o
^ 0.015 H
 CD
 O
 CD
.Q
 O 0.010 H
.a
   0.005 -
   0.000
            UV254
   D  Single contactor effluent
   	Logistic function best fit (RA2 = 0.994)
   O  Blended effluent
   - - - - Dl prediction
                                                                    EBCT = 7.2 min.
                                                                    c0 = 0.033 1/cm
               50                 100                150
                      Scaled operation time (days)
                                                                                   200
Figure E-142  Single contactor and blended effluent UV254 breakthrough
curves for Water 8
                                       -294-

-------
O
o
'-4—'
TO
O>
O
c
o
O
   125
   100 -
    75 -
    50 -
    25 -
     0 -to
           SDS-TOX
           D   Single contactor effluent

          	Logistic function best fit  (RA2 = 0.99)

           O   Blended effluent

           - - - - Dl prediction
                                                                   EBCT = 7.2min.

                                                                   c0= 156 ug/LCI-
                         50                 100

                                Scaled operation time (days)
                                                           150
200
 Figure E-143  Single contactor and blended effluent SDS-TOX breakthrough
 curves for Water 8
   25
   20 -
   15 -
o
'-4—'
TO
t_


O>
O
c
o
O
10 -
        SDS-CF
          D   Single contactor effluent

          	Logistic function best fit (RA2 = 0.985)

          O   Blended effluent

          - - - - Dl prediction
                                                          P--
                                                                 EBCT = 7.2min.

                                                                 c0 = 29.1Mg/L
                        50                 100

                                Scaled operation time (days)
                                                           150
200
Figure E-144 Single contactor and blended effluent SDS-CF breakthrough
curves for Water 8
                                      -295-

-------
    4 -
    3 -
 g
 '-4—'
 CD
 O
 O
    1 -
    0 -O
 SDS-BDCM



EBCT = 7.2min.

C0 = 2.6 |jg/L
                          D    D
                               O
                                 O
CD	Or—
     50
                                                                            O
                                                 D

                                                O
                                     D  Single contactor effluent
                                    	Logistic function best fit (RA2 = 0.946)
                                     O  Blended effluent
                                     - - - - Dl prediction
                                           100

                                Scaled operation time (days)
                                                       150
200
 Figure E-145  Single contactor and blended effluent SDS-BDCM
 breakthrough curves for Water 8
    12
    10 -
 o
 'CD   6
          SDS-DBCM
       D   Single contactor effluent
      	Logistic function best fit (RA2 = 0.98)
       O   Blended effluent
       - - - - Dl prediction
                   -ID
                                                              D
                                                             O
                                                                  EBCT = 7.2min.

                                                                  c0 =  10.3 |jg/L
                         50                 100
                                 Scaled operation time (days)
                                                        150
200
Figure E-146 Single contactor and blended effluent SDS-DBCM
breakthrough curves for Water 8
                                       -296-

-------
    4 -
    3 -
 g
 '-4—'
 CD
 O
 O
    1  -
    0 -O
         SDS-BF
 D   Single contactor effluent
	Logistic function best fit (RA2 = NA)
 O   Blended effluent
 	Dl prediction
                   Insufficient data measured above the MRL
                         to perform curve fit analysis
       -ii ii i k-ry-n—<"n	Oh	QI-
              50
-CD-
                                                                    EBCT = 7.2min.
                                                                    C0 =  BMRL
                                            100
             -Ch—
               150
200
                                 Scaled operation time (days)
Figure E-147  Single contactor and blended effluent SDS-BF breakthrough
curves for Water 8
    40
    30 -
         SDS-TTHM
   D   Single contactor effluent
  	Logistic function best fit (RA2 = 0.98)
   O   Blended effluent
   - - - - Dl prediction
                                                                    EBCT = 7.2min.

                                                                    C0 = 42 jjg/L.
                          50                 100
                                 Scaled operation time (days)
                                                     150
                                   200
Figure E-148 Single contactor and blended effluent SDS-TTHM breakthrough
curves for Water 8
                                       -297-

-------
    2.0
    1.5 -
1.0 -
 O>
 o
 c
 o
 O
    0.5 -
          SDS-MCAA
    o.o -lo-
       0
          D   Single contactor effluent
         	Logistic function best fit (RA2 = NA)
          O   Blended effluent
          - - - - Dl prediction
                                 Effluent concentrations were not detected
                                    above the MRL for this parameter
                                                                    EBCT = 7.2min.
                                                                    C0 = BMRL
                -ii ii i k-ry-n—m—Oh	QJ-
                      50
     -CD-
100
-OH—
  150
                                 Scaled operation time (days)
 Figure E-149 Single contactor and blended effluent SDS-MCAA
 breakthrough curves for Water 8
200
    6 -
    4 -
 o>
 o
 c
 o
 O
      SDS-DCAA

        D   Single contactor effluent
      	Logistic function best fit (RA2 = 0.974)
        O   Blended effluent
      	Dl prediction
                         50
                                                                   EBCT = 7.2min.
                                                                   c0= 10.7ug/L
                                        100
                             Scaled operation time (days)
                   150
                     200
Figure E-150 Single contactor and blended effluent SDS-DCAA breakthrough
curves for Water 8
                                       -298-

-------
   6 -
   5 -
   4 -
g
'-4—'
CD
CD
O
c
O
O
   3 -
         SDS-TCAA
            D   Single contactor effluent
            	Logistic function best fit (RA2 = 0.93)
            O   Blended effluent
            - - - - Dl prediction
                                                                   EBCT = 7.2 min.
                                                                   C0 =  12.7 |jg/L
50                 100
        Scaled operation time (days)
                                                               150
                                                                                   200
Figure E-151 Single contactor and blended effluent SDS-TCAA breakthrough
curves for Water 8
   2.0
   1.5 -
O
'•§  1.0 H
-i—•

I
O
O
   0.5 -
          SDS-MBAA
             D  Single contactor effluent
          	Logistic function best fit (RA2 = NA)
             O  Blended effluent
          	Dl prediction
   0.0 Hd-
      0
                               Effluent concentrations were not detected
                                   above the MRL for this parameter
                                                                   EBCT = 7.2 min.
                                                                   C0= BMRL
                   -ii ii i K-ry-n—m	Oh	QI-
                         50
                         -CD-
                    100
-CD-,—
  150
200
                                 Scaled operation time (days)
Figure E-152  Single contactor and blended effluent SDS-MBAA
breakthrough curves for Water 8
                                       -299-

-------
    10
     6 -
 g
 '-4—'
 CD
 O
 O
     2 -
         SDS-DBAA
     o -lo
      o
               D  Single contactor effluent
              	Logistic function best fit (RA2 = NA)
               O  Blended effluent
              	Dl prediction
                       Effluent concentrations were not detected
                          above the MRL for this parameter
                                                                  EBCT = 7.2min.
                                                                  C0 = BMRL
          -03EEKJJ-D—O\	Oh	Ol-
                50
     -CD-
100
-CDn—
  150
200
                                 Scaled operation time (days)
 Figure E-153 Single contactor and blended effluent SDS-DBAA breakthrough
 curves for Water 8
    15
    10 -
 o
 '-4—'
 CD
SDS-HAA5

   D   Single contactor effluent
 	Logistic function best fit  (RA2 = 0.955)
   O   Blended effluent
 	Dl prediction
                         50
                                                                   EBCT = 7.2min.
                                                                  c0 =  23 |jg/L
                                   100
                        Scaled operation time (days)
                   150
                     200
Figure E-154 Single contactor and blended effluent SDS-HAA5 breakthrough
curves for Water 8
                                       -300-

-------
    3 -
 o
 '•^
 CO
 O>
 O
 c
 o
 O
    1 -
    0 -O
        SDS-BCAA

           D   Single contactor effluent
         	Logistic function best fit (RA2 = 0.971)
           O   Blended effluent
           	Dl prediction
                                      -D-
O
               O
              D
                                                                  EBCT = 7.2min.
                                                                  c0 = 3 |jg/L
                                           100
                                Scaled operation time (days)
 150
200
 Figure E-155  Single contactor and blended effluent SDS-BCAA breakthrough
 curves for Water 8
    20
    15 -
          SDS-HAA6
            D   Single contactor effluent
          	Logistic function best fit  (RA2 = 0.956)
            O   Blended effluent
                                                                  EBCT = 7.2min.
                                                                  c0 =  27 |jg/L
                         50                100                150
                                Scaled operation time (days)
                    200
Figure E-156 Single contactor and blended effluent SDS-HAA6 breakthrough
curves for Water 8
                                      -301-

-------
    3.0
    2.5 -
    2.0 -
 O)

 c
 o
 '•^  1
 CD  I-
 CD
 O
     .
    0.5 -
          SDS-DCBAA
EBCT = 7.2min.
c0 = 3 |jg/L
    0.0 -lo-
       0
                                                               D
                                                              O
                                                  -D-
                                    D   Single contactor effluent
                                   	Logistic function best fit (RA2 = 0.939)
                                    O   Blended effluent
                                    	Dl prediction
                50                 100                 150
                       Scaled operation time (days)
                                      200
 Figure E-157  Single contactor and blended effluent SDS-DCBAA
 breakthrough curves for Water 8
    2.0
    1.5 -
 o
 '•CD  1.0 H
 t_
 CD
 O
 c
 o
 O
    0.5 -
           SDS-CDBAA
    o.o Ho-
       o
     D   Single contactor effluent
    	Logistic function best fit (RA2 = NA)
     O   Blended effluent
     	Dl prediction
                     Effluent concentrations were not detected
                        above the MRL for this parameter
                                                                   EBCT = 7.2 min.
                                                                   c0= BMRL
          -ini i KTy-n—m—Qh	QJ-
                50
     -CD-
-OH-
100
  150
200
                                 Scaled operation time (days)
Figure E-158 Single contactor and blended effluent SDS-CDBAA
breakthrough curves for Water 8
                                       -302-

-------
    2.0
    1.5 -
    1-0 H
 O>
 o
 c
 o
 O
    0.5 -
          SDS-TBAA
    o.o -lo-
       0
 D   Single contactor effluent
	Logistic function best fit  (RA2 = NA)
 O   Blended effluent
 	Dl prediction
                  Effluent concentrations were not detected
                      above the MRL for this parameter
                                                                  EBCT = 7.2min.
                                                                  C0 = BMRL
      -ii ii i k-ry-n—m—Oh	QJ-
            50
     -CD-
100
-OH—
  150
200
                                 Scaled operation time (days)
 Figure E-159 Single contactor and blended effluent SDS-TBAA breakthrough
 curves for Water 8
    20
    15 -
         SDS-HAA9
             D  Single contactor effluent
          	Logistic function best fit (RA2 = 0.952)
             O   Blended effluent
             - - - - Dl prediction
                                                                   EBCT = 7.2min.
                                                                   c0 = 30 |jg/L
                          50                 100                 150
                                 Scaled operation time (days)
                                                                     200
Figure E-160 Single contactor and blended effluent SDS-HAA9 breakthrough
curves for Water 8
                                       -303-

-------
This page intentionally left blank.
             -304-

-------
Appendix F:  Comparison of SCA Method to Dl Approach for Integral
Breakthrough Curve Prediction
                             -305-

-------
      2.5
      2.0 -
    O)
    o
    "co
    o
   O
      1.5 4
      1.0 H
      0.5 -
      0.0
             TOO
EBCT = 15 min.
c0 = 4.54 mg/L


                    X  •'
                   X   '
                 X
                X
                                                       •   Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = 0.152)
                                 25                      50
                                  Scaled operation time (days)
                                                                      75
Figure F-1 Dl method prediction of the TOC integral breakthrough curve for
Water 1
      0.035
      0.000
                                                     •   Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = 0.0011)
                                                   	SCA prediction (RSS = 0.0017)
           0                       25                       50
                                    Scaled operation time (days)

Figure F-2 Comparison of Dl and SCA methods for predicting the UV254
integral breakthrough curve for Water 1
                                                                       75
                                        -306-

-------
       200
       150 -
    O
    o  100 H
    o>
    o
    c
    o
    0   50 H
  SDS-TOX

  EBCT = 15 min.
  CD =  224 |jg/L Cl-
                                 SDS-TOX not analyzed in
                                   the blended effluent
                                                      •   Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = NA)
                                                    	SCA prediction (RSS = NA)
                                 25                      50
                                  Scaled operation time (days)
                                                                     75
Figure F-3 Comparison of Dl and SCA methods for predicting the SDS-TOX
integral breakthrough curve for Water 1
      5 -
      4 -
    o
    '•a
    o>
    o
   8 2
      1 -
 SDS-CF

EBCT = 15 min.
 CD = 34.2 |jg/L
                                           •  Observed data
                                         	Best fit
                                         	Dl prediction  (RSS = 0.39)
                                         	SCA prediction (RSS = 0.65)
                                25                       50
                                 Scaled operation time (days)
                                                                     75
Figure F-4 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 1
                                       -307-

-------
      20
      16 -
    O)
   a 12 H
   c
   o
   '-4— '
   S

   §  8
   c
   o
   O
       4 -
SDS-BDCM



EBCT = 15 min.


 c0 = 19.3|jg/L
                                          •  Observed data

                                        	Best fit

                                        	Dl prediction (RSS = 1.52)

                                        	SCA prediction (RSS = 1.03)
         0                       25                      50

                                 Scaled operation time (days)


Figure F-5  Comparison of Dl and SCA methods for predicting the SDS-

BDCM integral breakthrough curve for Water 1
                                                                   75
      12
      10 -
    CD  6 -\
    1_
    -I—•

    §

    I  4
       2 -
 SDS-DBCM



 EBCT = 15 min


 co = 28 |jg/L
                                          •   Observed data

                                        	Best fit

                                        	Dl prediction (RSS = 0.4)

                                        	SCA prediction  (RSS = 0.73)
         0                      25                      50

                                 Scaled operation time (days)


 Figure F-6  Comparison of Dl and SCA methods for predicting the SDS-

 DBCM integral breakthrough curve for Water 1
                                                                    75
                                       -308-

-------
16

14 -

12 -

10 -
  o
  '-4—'
  CD
  CD
  O
  c
  o
  O
          SDS-BF

            EBCT = 15 min
            c0=  3.7|jg/L
                                                    •   Observed data
                                                  	Best fit
                                                  	Dl prediction (RSS = 2.76)
                                                  	SCA prediction  (RSS = 1.31)
       0                       25                      50
                                Scaled operation time (days)

 Figure F-7 Comparison of Dl and SCA methods for predicting the SDS-BF
 integral breakthrough curve for Water 1
                                                                         75
                                                         Observed data
                                                         Best fit
                                                   	Dl prediction (RSS = 3.93)
                                                         SCA prediction (RSS = 2.55)
         0                       25                      50
                                 Scaled operation time (days)

Figure F-8  Comparison of Dl and SCA methods for predicting the SDS-TTHM
integral breakthrough curve for Water 1
                                                                           75
                                       -309-

-------
       2.0
       1.5 -
    g
    '•S3  1-0 H
    O>
    o
    c
    o
    O
       0.5 -
 SDS-MCAA

EBCT = 15 min.
c0 = BMRL
       0.0 +•-
          0
                        Blended effluent concentrations were not detected
                              above the MRL for this parameter
                                                      •   Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = NA)
                                                    	SCA prediction (RSS = NA)
                      25                      50
                       Scaled operation time (days)
75
Figure F-9 Comparison of Dl and SCA methods for predicting the SDS-MCAA
integral breakthrough curve for Water 1
       10
        6 -
    §   4H
    c
    o
    O
        2 -
             SDS-DCAA

             EBCT = 15 min.
             c0= 12.5 pg/L
                                          •   Observed data
                                        	Best fit
                                        	Dl prediction (RSS = 0.35)
                                        	SCA prediction (RSS = 1.46)
         0                       25                      50
                                  Scaled operation time (days)
Figure F-10 Comparison of Dl and SCA methods for predicting the SDS-
DCAA integral breakthrough curve for Water 1
                                                                     75
                                       -310-

-------
       2.0
       1.5 -
    g
    '•S3  1-0 H
    O>
    o
    c
    o
    O
       0.5 -
 SDS-TCAA

EBCT = 15 min.
c0 = 3 |jg/L
       0.0 +•-
          0
                            Blended effluent concentrations were not
                            detected above the MRL for this parameter
                                                        •  Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = NA)
                                                     	SCA prediction  (RSS = NA)
                     25                      50
                      Scaled operation time (days)
75
Figure F-11 Comparison of Dl and SCA methods for predicting the SDS-
TCAA integral breakthrough curve for Water 1
       2.0
       1.5 -
    O
    '•§  1.0 H
    -i—'
    O>
    o
    c
    o
    O
       0.5 -
 SDS-MBAA

 EBCT = 15 min.
 CO = BMRL
       o.o 4*-
          o
            Blended effluent concentrations were not detected
                  above the MRL for this parameter
                                                        •  Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = NA)
                                                     	SCA prediction  (RSS = NA)
                     25                      50
                      Scaled operation time (days)
75
Figure F-12 Comparison of Dl and SCA methods for predicting the SDS-
MBAA integral breakthrough curve for Water 1
                                        -311-

-------
    5 -
    4 -
    3 -
  CD
  O
  O o J
  O ^ 1
    1 -
   SDS-DBAA

    EBCT = 15 min
    co =  4 |jg/L
                                                •  Observed data
                                             	Best fit
                                             	Dl prediction (RSS = 0.46)
                                             	SCA prediction (RSS = 0.39)
      0                       25                       50
                                Scaled operation time (days)
Figure F-13 Comparison of Dl and SCA methods for predicting the SDS-
DBAA integral breakthrough curve for Water 1
                                                                         75
     15
     10 -
   o
   "CD
   CD
   o
   c
   o
   O
5 -
          SDS-HAA5

          EBCT = 15 min
          co=20 |jg/L
                                                         Observed data
                                                         Best fit
                                                         Dl prediction (RSS = 0.38)
                                                         SCA prediction  (RSS = 1.58)
        0                       25                      50
                                 Scaled operation time (days)

Figure F-14 Comparison of Dl and SCA methods for predicting the SDS-
HAA5 integral breakthrough curve for Water 1
                                                                         75
                                       -312-

-------
   4 -
 O)
 a 3 -
 c
 o
 I

 12
 o
 O
   1 -
 SDS-BCAA

 EBCT = 15 min.
 c0 = 7 |jg/L
                                              •  Observed data
                                            	Best fit
                                            	Dl prediction (RSS = 0.38)
                                            	SCA prediction (RSS = 0.29)
     0                        25                        50
                               Scaled operation time (days)
Figure F-15 Comparison of Dl and SCA methods for predicting the SDS-
BCAA integral breakthrough curve for Water 1
                                                                        75
    20
    15 -
  o
 '•S3 10 H
  o>
  o
  c
  o
 O
     5 -
SDS-HAA6

 EBCT = 15 min.
 c0 = 27 |jg/L
                                                      •   Observed data
                                                      	Best fit
                                                      	Dl prediction (RSS = 0.63)
                                                      	SCA prediction (RSS = 1.26)
       0                        25                       50
                                Scaled operation time (days)

Figure F-16 Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 1
                                                                        75
                                       -313-

-------
   2.0
   1.5 -
 o

'OB 1.0 H
 O>
 o
 c
 o
O
   0.5 -
   0.0
 SDS-DCBAA




  EBCT = 15 min.


  c0 = 2.3 ug/L
                  Insufficient data measured above the MRL to

                      compare Dl and SCA predictions
                                                        •   Observed data

                                                      	Best fit

                                                      	Dl prediction (RSS = NA)

                                                      	SCA prediction (RSS = NA)
      0                        25                        50

                                 Scaled operation time (days)


Figure F-17 Comparison of Dl and SCA methods for predicting the SDS-

DCBAA integral breakthrough curve for Water 1
                                                                          75
   2.0
   1.5 -
 o
 '•g 1.0 H
 -I—I

 §
 c
 o
 O


   0.5 -
   0.0
SDS-CDBAA




 EBCT = 15 min.


 rv= BMRL
                Blended effluent concentrations were not detected

                      above the MRL for this parameter
                                                        •   Observed data

                                                      	Best fit

                                                      	Dl prediction (RSS = NA)

                                                      	SCA prediction (RSS = NA)
      0                        25                        50

                                 Scaled operation time (days)


Figure F-18 Comparison of Dl and SCA methods for predicting the SDS-

CDBAA integral breakthrough curve for Water 1
                                                                          75
                                        -314-

-------
   2.0
   1.5 -
o
'7J  1-0 H
-i—'

I
o
O

   0.5 -
   0.0
        SDS-TBAA


          EBCT = 15 min.
          rv= BMRL
                        Blended effluent concentrations were not detected
                              above the MRL for this parameter
                                                        •  Observed data
                                                       	Best fit
                                                       	Dl prediction (RSS = NA)
                                                       	SCA prediction (RSS = NA)
      0                        25                        50
                                 Scaled operation time (days)
Figure F-19 Comparison of Dl and SCA methods for predicting the SDS-
TBAA integral breakthrough curve for Water 1
                                                                                  75
   20
   15 -
 O)
 o
 73 10
 -I— i
 §
 c
 o
 O

    5 -
       SDS-HAA9


        EBCT = 15 min.
        c0=  29
                                                        •   Observed data
                                                        	Best fit
                                                        	Dl prediction (RSS = 0.78)
                                                        	SCA prediction (RSS = 1.04)
      0                        25                        50
                                 Scaled operation time (days)
Figure F-20 Comparison of Dl and SCA methods for predicting the SDS-
HAA9 integral breakthrough curve for Water 1
                                                                                  75
                                        -315-

-------
       1.50
       1.25 -
       1.00 -
    o
    '•a  °-75 H
    o>
    o
    o  0.50 -
    O
       0.25 -
       0.00
             TOO
EBCT = 20min.
c0 = 2.6 mg/L
                                           •   Observed data
                                         	Best fit
                                         	Dl prediction (RSS = 0.057)
                         50
                          100           150
                      Scaled operation time (days)
200
250
Figure F-21 Dl method prediction of the TOC integral breakthrough curve for
Water 2
       0.025
       0.020 -
    o
    ^ 0.015 H
    CD
    O
    CD
    .Q
    O  0.010 H
    to
       0.005 -
       0.000
 UV254
 EBCT = 20 min.
 co = 0.055 1/cm
                                         •   Observed data
                                       	Best fit
                                       	Dl prediction (RSS = 0.0012)
                                       	SCA prediction  (RSS = 0.0015)
                          50
                           100           150
                       Scaled operation time (days)
 200
 250
Figure F-22 Comparison of Dl and SCA methods for predicting the UV254
integral breakthrough curve for Water 2
                                        -316-

-------
      200
       150 -
    O
    .   100 H
    -i—'
    CD
    CD
    O
       50 H
             SDS-TOX
              EBCT = 20 min.
              CD = 220 |jg/L Cl-
                        50
SDS-TOX not analyzed in
  the blended effluent
                      •  Observed data
                   	Best fit
                   	Dl prediction  (RSS = NA)
                   	SCA prediction (RSS = NA)
    100           150
 Scaled operation time (days)
200
250
Figure F-23 Comparison of Dl and SCA methods for predicting the SDS-TOX
integral breakthrough curve for Water 2
       15
       10 -
    O
    IE
    -i—'
    CD
    O
              SDS-CF
              EBCT = 20 min
              c0= 41.9|jg/L
                                                       •   Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = 0.57)
                                                    	SCA prediction (RSS = 1.59)
                       50
    100            150
Scaled operation time (days)
200
250
Figure F-24 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 2
                                       -317-

-------
     30
     25 -
          SDS-BDCM
          EBCT = 20 min.
          c0 = 19.8 |jg/L
                                                    •   Observed data
                                                  	Best fit
                                                  	Dl prediction (RSS = 4.35)
                                                  	SCA prediction (RSS = 1.98)
                     50
                         100            150
                      Scaled operation time (days)
200
250
Figure F-25 Comparison of Dl and SCA methods for predicting the
SDS-BDCM integral breakthrough curve for Water 2
     25
     20 -
     15 -
  O
  c
  o
  O
     -in J
     ID -
     5 -
SDS-DBCM
EBCT = 20 min.
Co= 31.7M9/L
                     50
                                          •   Observed data
                                        	Best fit
                                        	Dl prediction (RSS = 2.18)
                                        	SCA prediction (RSS = 1.72)
                         100            150
                      Scaled operation time (days)
200
 250
 Figure F-26 Comparison of Dl and SCA methods for predicting the SDS-
 DBCM integral breakthrough curve for Water 2
                                       -318-

-------
    14
    12 -
    10 -
 o
 "-i—'
 2
 |   6H
 o
 °   4J
     2 -
     0
      0
     SDS-BF
     EBCT = 20 min.
     c0 = 3.7 |jg/L  |
                50
                                                •   Observed data
                                              	Best fit
                                              	Dl prediction (RSS = 3.73)
                                              	SCA prediction (RSS = 2.18)
200
                                   100            150
                               Scaled operation time (days)
Figure F-27 Comparison of Dl and SCA methods for predicting the SDS-BF
integral breakthrough curve for Water 2
250
80

70 -

60 -

50 -
  o
  IS  40 H
  o>
  g
  o
  O
30 -

20 -

10 -
     0
          SDS-TTHM
          EBCT = 20 min
          c0 = 97 |jg/L
                                                     •  Observed data
                                                   	Best fit
                                                   	Dl prediction  (RSS = 8.49)
                                                   	SCA prediction (RSS = 4.38)
                     50
                               100            150
                            Scaled operation time (days)
 200
                                                                               250
Figure F-28 Comparison of Dl and SCA methods for predicting the SDS-
TTHM integral breakthrough curve for Water 2
                                       -319-

-------
     2.0
     1.5 -
  o
  '•S  1-0 H
  o
  O
     0.5 -
     0.0
         SDS-MCAA

         EBCT = 20 min.
         co=  BMRL
                        Blended effluent concentrations were not detected
                              above the MRL for this parameter
                                                        •  Observed data
                                                     	Best fit
                                                     	Dl prediction  (RSS = NA)
                                                     	SCA prediction (RSS = NA)
                      50
                            100             150
                         Scaled operation time (days)
200
250
Figure F-29 Comparison of Dl and SCA methods for predicting the SDS-
MCAA integral breakthrough curve for Water 2
      5 -
      4 -
   o
   *=  o
   CD  J
   o
   <§2
      1 -
SDS-DCAA
EBCT = 20 min.
c0= 14 |jg/L
                                             •   Observed data
                                           	Best fit
                                           	Dl prediction (RSS = 0.49)
                                           	SCA prediction (RSS = 0.68)
                      50
                            100             150
                         Scaled operation time (days)
200
250
Figure F-30 Comparison of Dl and SCA methods for predicting the SDS-
DCAA integral breakthrough curve for Water 2
                                        -320-

-------
     2.5
     2.0 -
  O)
     1.5 -
  o
  -i—•
  CD
  8  1-0-1
  o
  O     -I
     0.5 -
     0.0
SDS-TCAA
 EBCT = 20 min.
 c0 = 5 |jg/L
                                              •   Observed data
                                            	Best fit
                                            	Dl prediction  (RSS = 0.27)
                                            	SCA prediction (RSS = 0.34)
                      50
                           100            150
                        Scaled operation time (days)
200
250
Figure F-31 Comparison of Dl and SCA methods for predicting the SDS-
TCAA integral breakthrough curve for Water 2
    2.0
    1.5 -
    1-0 H
 O>
 o
 c
 o
 O
    0.5 -
    0.0
         SDS-MBAA
         EBCT = 20 min.

         c°=  BMRL
                  Blended effluent concentrations were not
                 detected above the MRL for this parameter
                                                      •   Observed data
                                                    	Best fit
                                                    	Dl prediction  (RSS = NA)
                                                    	SCA prediction (RSS = NA)
                      50
                          100            150
                       Scaled operation time (days)
200
250
Figure F-32 Comparison of Dl and SCA methods for predicting the SDS-
MBAA integral breakthrough curve for Water 2
                                        -321-

-------
      6 -
   o
   14H
   I
   o
   O
 SDS-DBAA

EBCT = 20 min
c0 = 5 |jg/L
                                                       •  Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = 1.1)
                                                     	SCA prediction (RSS = 0.6)
                                    100            150
                                 Scaled operation time (days)
                                                        200
               250
Figure F-33 Comparison of Dl and SCA methods for predicting the SDS-
DBAA integral breakthrough curve for Water 2
    15
    10 -
 o
 '-4—'
 CD
 CD
 O
 0   «; J
 O   5 H
          SDS-HAA5
          EBCT = 20 min.

          co = 24 |jg/L
                                                     •  Observed data
                                                  	Best fit
                                                  	Dl prediction (RSS = 1.69)
                                                  	SCA prediction (RSS = 1.48)
                     50
                          100            150
                       Scaled operation time (days)
200
250
Figure F-34 Comparison of Dl and SCA methods for predicting the SDS-
HAA5 integral breakthrough curve for Water 2
                                       -322-

-------
      6 -
   o
   '•SS  4H
   O>
   o
   c
   o
   O
      2 -
          SDS-BCAA

         EBCT = 20 min.
         c0= 9|jg/L
                                                     •  Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = 0.82)
                                                   	SCA prediction (RSS = 0.78)
                      50
                            100            150
                         Scaled operation time (days)
 200
250
Figure F-35 Comparison of Dl and SCA methods for predicting the SDS-
BCAA integral breakthrough curve for Water 2
    25
    20 -
 O)
 a 15 H
 c
 o
 '-4—'
 S
 § 10
 c
 o
 O
     5 -
SDS-HAA6
EBCT = 20 min.
c0= 34
                                                Observed data
                                                Best fit
                                                Dl prediction (RSS = 2.5)
                                                SCA prediction (RSS = 2.25)
                     50
                           100             150
                        Scaled operation time (days)
200
250
Figure F-36 Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 2
                                       -323-

-------
     2.5
     2.0 -
  O)
     1.5 -
  o
  •*=
  £5
  -i—i
  §  1.0
  c
  o
  O
     0.5 -
SDS-DCBAA
EBCT = 20 min.
CD = 3 |jg/L
     0.0 -(•-
        0
                                              •  Observed data
                                           	Best fit
                                           	Dl prediction (RSS = 0.48)
                                           	SCA prediction (RSS = 0.35)
                           100            150
                         Scaled operation time (days)
200
250
Figure F-37 Comparison of Dl and SCA methods for predicting the SDS-
DCBAA integral breakthrough curve for Water 2
    3.0
    2.5 -
    2.0 -
  O)
    1-5 H
    0.5 -
    0.0
 SDS-CDBAA
 EBCT = 20 min.
 c0= BMRL
                  Insufficient data measured above the MRL
                     to compare Dl and SCA predictions
                                              •   Observed data
                                            	Best fit
                                            	Dl prediction (RSS = NA)
                                            	SCA prediction (RSS = NA)
                      50
                           100            150
                        Scaled operation time (days)
200
250
Figure F-38 Comparison of Dl and SCA methods for predicting the SDS-
CDBAA integral breakthrough curve for Water 2
                                        -324-

-------
     2.0
     1.5 -
o
'•S  1-0
   o
   o
  o
     0.5 -
          SDS-TBAA

          EBCT = 20 min.
          co=  BMRL
0.0 -(•-
   0
                            Blended effluent concentrations were not
                           detected above the MRL for this parameter
                                                     •   Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = NA)
                                                   	SCA prediction (RSS = NA)
                      50
                                  100            150
                               Scaled operation time (days)
200
250
Figure F-39 Comparison of Dl and SCA methods for predicting the SDS-
TBAA integral breakthrough curve for Water 2
    30
    25 -
    20 -
  O)
    15 H
     5 -
       SDS-HAA9
       EBCT = 20 min.
       c0 = 37
                                                     •  Observed data
                                                  	Best fit
                                                  	Dl prediction (RSS = 3.64)
                                                  	SCA prediction (RSS = 3.43)
                     50
                                  100             150
                               Scaled operation time (days)
200
250
Figure F-40 Comparison of Dl and SCA methods for predicting the SDS-
HAA9 integral breakthrough curve for Water 2
                                        -325-

-------
       1.25
       1.00 -
    O)

    -§ 0.75 -
    .g

    "co
    g  0.50
    c
    o
    O
       0.25 -
       0.00
       TOO



       EBCT = 20 min.


       c0 =  2.35 mg/L
                             X
                           X
                          X
                        X  , •


                   xx<

                  x , -'
                                                        •  Observed data

                                                      	Best fit

                                                      	Dl prediction (RSS = 0.044)
                      50
                           100         150         200

                            Scaled operation time (days)
250
300
Figure F-41  Dl method prediction of the TOC integral breakthrough curve for

Water 3
       0.020
       0.015 -
    o
    o>
    o
    c
    CD
    .Q

    O
    
-------
       200
       150 -
    O
.2  100 -
"en
    O>
    o
    c
    o
    0   50 H
          SDS-TOX

          EBCT = 20 min.
          c0 =  255 |jg/L Cl-
                      50
                                 SDS-TOX not analyzed in
                                    the blended effluent
                                                  •   Observed data
                                                	Best fit
                                                	Dl prediction (RSS = NA)
                                                	SCA prediction (RSS = NA)
                             100          150         200
                              Scaled operation time (days)
250
300
Figure F-43 Comparison of Dl and SCA methods for predicting the SDS-TOX
integral breakthrough curve for Water 3
       6 -
       5 -
       4 H
    o
    "on
    o
    O
       3 -
       2 -
       1 -
        SDS-CF

        EBCT = 20 min.
        co = 60.3 |jg/L
                                                  •   Observed data
                                                	Best fit
                                                	Dl prediction (RSS = 0.85)
                                                	SCA prediction (RSS = 0.4)
                    50
                            100          150         200
                             Scaled operation time (days)
250
300
Figure F-44 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 3
                                       -327-

-------
      35
      30 -
      25 -
    O)
      20 -
    o
    0
       10 -
       5 -
     SDS-BDCM

     EBCT = 20 min.
     c0 = 36.2 |jg/L
                                              •  Observed data
                                            	Best fit
                                            	Dl prediction (RSS = 1.05)
                                            	SCA prediction (RSS = 2.57)
                    50
                         100         150         200
                          Scaled operation time (days)
250
300
Figure F-45 Comparison of Dl and SCA methods for predicting the SDS-
BDCM integral breakthrough curve for Water 3
16

14 -

12 -

10 -
    I  6H
    O
    O
       4 -
       2 -
            SDS-DBCM

           EBCT = 20 min
           co = 44.5 |jg/L
                                              •   Observed data
                                            	Best fit
                                            	Dl prediction (RSS = 1.04)
                                            	SCA prediction (RSS = 0.68)
                    50
                         100         150         200
                          Scaled operation time (days)
 250
 300
 Figure F-46 Comparison of Dl and SCA methods for predicting the SDS-
 DBCM integral breakthrough curve for Water 3
                                       -328-

-------
           SDS-BF

          EBCT = 20 min
               11.8 |jg/L
                                                        Observed data
                                                        Best fit
                                                  	Dl prediction (RSS = 5.67)
                                                        SCA prediction (RSS = 3.14)
                    50
                        100         150         200
                          Scaled operation time (days)
250
300
Figure F-47 Comparison of Dl and SCA methods for predicting the SDS-BF
integral breakthrough curve for Water 3
    o
    O
80

70 -

60 -

50 -

40 -

30 -

20 -

10 -
        0
            SDS-TTHM

           EBCT = 20 min.
           Co= 154|jg/L
                                                     •   Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = 2.87)
                                                   	SCA prediction (RSS = 2.64)
                     50
                         100         150         200
                           Scaled operation time (days)
 250
 300
Figure F-48 Comparison of Dl and SCA methods for predicting the SDS-
TTHM integral breakthrough curve for Water 3
                                       -329-

-------
     2.0
      1.5 -
   g
   '•SS  1-0 H
O>
o
o
O
   0.5 -
      0.0
        SDS-MCAA

        EBCT = 20 min.
       co=   BMRL
                        Blended effluent concentrations were not detected
                              above the MRL for this parameter
                                                        •   Observed data
                                                     	Best fit
                                                     	Dl prediction  (RSS = NA)
                                                     	SCA prediction (RSS = NA)
                    50
                             100          150          200
                               Scaled operation time (days)
250
300
Figure F-49 Comparison of Dl and SCA methods for predicting the SDS-
MCAA integral breakthrough curve for Water 3
      3.0
      2.5 -
      2.0 -
   o
   ]3  1.5 H
   -i—'
   o>
   o
   o  1.0 H
      0.5 -
      0.0
        SDS-DCAA

        EBCT = 20 min.
        c0= 15.7
                                                        Observed data
                                                       • Best fit
                                                  	Dl prediction  (RSS = 0.4)
                                                  	SCA prediction (RSS = 0.37)
                     50
                              100          150          200
                               Scaled operation time (days)
250
300
Figure F-50 Comparison of Dl and SCA methods for predicting the SDS-
DCAA integral breakthrough curve for Water 3
                                        -330-

-------
      2.0
      1.5 -
   o
   '•SS 1-0 H
O>
o
o
O
   0.5 -
      0.0
             SDS-TCAA
            EBCT = 20 min.
        co -i
              '5 ug/L
                         Blended effluent concentrations were not detected
                               above the MRL for this parameter
                                                         •   Observed data
                                                      	Best fit
                                                      	Dl prediction (RSS = NA)
                                                      	SCA prediction  (RSS = NA)
                     50
                              100          150         200
                               Scaled operation time (days)
250
300
Figure F-51 Comparison of Dl and SCA methods for predicting the SDS-
TCAA integral breakthrough curve for Water 3
      2.0
      1.5 -
   '•§ 1.0 H
   -I—'
   o>
   o
   c
   o
   O
      0.5 -
      0.0
         SDS-MBAA

       EBCT = 20 min.
        co= BMRL
                       Blended effluent concentrations were not
                      detected above the MRL for this parameter
                                                         •   Observed data
                                                      	Best fit
                                                      	Dl prediction (RSS = NA)
                                                      	SCA prediction  (RSS = NA)
                     50
                              100          150         200
                               Scaled operation time (days)
250
300
Figure F-52 Comparison of Dl and SCA methods for predicting the SDS-
MBAA integral breakthrough curve for Water 3
                                        -331-

-------
       12
       10 -
     CD
     O
     §  4H
        2 -
SDS-DBAA

EBCT = 20 min
c0 = 9.7 pg/L
                                         •   Observed data
                                       	Best fit
                                       	Dl prediction (RSS = 1.02)
                                       	SCA prediction (RSS = 1.56)
                     50
                    100         150         200
                      Scaled operation time (days)
250
300
Figure F-53 Comparison of Dl and SCA methods for predicting the SDS-
DBAA integral breakthrough curve for Water 3
       20
       15 -
     o
     '•SS 10 H
     CD
     O
     c
     o
     O
        5 -
SDS-HAA5

EBCT = 20 min.
co =  30 |jg/L
                                                     •   Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = 2.48)
                                                   	SCA prediction (RSS = 1.13)
                     50
                    100         150         200
                      Scaled operation time (days)
250
300
Figure F-54 Comparison of Dl and SCA methods for predicting the SDS-
HAA5 integral breakthrough curve for Water 3
                                       -332-

-------
       5 -
       4 -
     o
     73 3
     O>
     o
       1 -
SDS-BCAA

EBCT = 20 min
c0=12.7|jg/L
                                         •  Observed data
                                       	Best fit
                                       	Dl prediction (RSS = 0.63)
                                       	SCA prediction  (RSS = 0.63)
                    50
                   100         150         200
                     Scaled operation time (days)
250
300
Figure F-55 Comparison of Dl and SCA methods for predicting the SDS-
BCAA integral breakthrough curve for Water 3
       20
       15 -
     o
     73 1°
     CD
     O
     c
     o
     O
        5 -
SDS-HAA6

EBCT = 20 min
co =  43 |jg/L
                                                      •  Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = 2.29)
                                                    	SCA prediction (RSS = 2.3)
                     50
                    100         150         200
                     Scaled operation time (days)
250
300
Figure F-56 Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 3
                                       -333-

-------
       2.0
       1.5 -
    o
    '"S  1.0 -\
    (D
    O
    O
    O
       0.5 -
       0.0
SDS-DCBAA

 EBCT = 20 min.
 c0=  4|jg/L
                                                       •  Observed data
                                                    	Best fit
                                                    	Dl prediction  (RSS = 0.68)
                                                    	SCA prediction (RSS = 0.72)
                     50
                    100         150         200
                      Scaled operation time (days)
250
300
Figure F-57 Comparison of Dl and SCA methods for predicting the SDS-
DCBAA integral breakthrough curve for Water 3
      2.0
      1.5 -
    o
    '•§ 1.0 H
    -i—'
    o>
    o
    c
    o
    O
      0.5 -
      0.0
SDS-CDBAA

EBCT = 20 min.
c0=2.5|jg/L
                     50
               Blended effluent concentrations were not
              detected above the MRL for this parameter
                                          •   Observed data
                                        	Best fit
                                        	Dl prediction (RSS = NA)
                                        	SCA prediction (RSS = NA)
                    100          150          200
                     Scaled operation time (days)
250
300
Figure F-58 Comparison of Dl and SCA methods for predicting the SDS-
CDBAA integral breakthrough curve for Water 3
                                        -334-

-------
     2.0
     1.5 -
  g
  '•SS  1-0 H
  CD
  o
  c
  o
  O
     0.5 -
 SDS-TBAA


EBCT = 20 min.
 co= BMRL
     0.0 -!•-
        0
                      Blended effluent concentrations were not detected above the
                                    MRL for this parameter
                                              •   Observed data
                                            	Best fit
                                            	Dl prediction (RSS = NA)
                                            	SCA prediction (RSS = NA)
           50
100         150          200
  Scaled operation time (days)
250
300
Figure F-59 Comparison of Dl and SCA methods for predicting the SDS-
TBAA integral breakthrough curve for Water 3
     25
     20 -
  o
  '-4—'
  CD
  o
  O
     15 -
     -in J
     10 -
      5 -
      o -k
SDS-HAA9

EBCT = 20 min
c0=  49
                                              •   Observed data
                                            	Best fit
                                            	Dl prediction (RSS = 4.36)
                                            	SCA prediction (RSS = 2.2)
                    50
                      100          150          200
                        Scaled operation time (days)
                                     250
            300
Figure F-60 Comparison of Dl and SCA methods for predicting the SDS-
HAA9 integral breakthrough curve for Water 3
                                        -335-

-------
      1.50
      1.25 -
      1.00 -
    c
    o
      0.75 -
    CD
    1 0.50 H
      0.25 -
      0.00
 TOC

EBCT = 20 min.
c0 = 2.98 mg/L
                                                        •   Observed data
                                                      	Best fit
                                                      	PI prediction (RSS = 0.05)
               50                100
                      Scaled operation time (days)
                                                               150
200
Figure F-61  Dl method prediction of the TOC integral breakthrough curve for
Water 4
      0.025
      0.020 -
    o
    ^ 0.015
    CD
    o 0.010
      0.005 -
      0.000
 UV254

 EBCT = 20 min.
 co=  0.065 1/cm
                                          •   Observed data
                                        	Best fit
                                        	Dl prediction (RSS = 0.0015)
                                        	SCA prediction (RSS = 0.0023)
                             50               100
                                    Scaled operation time (days)
                                                    150
 200
Figure F-62 Comparison of Dl and SCA methods for predicting the UV254
integral breakthrough curve for Water 4
                                        -336-

-------
      200
      150 -
   O
    o 100 -
    O>
    o
       50 4
  SDS-TOX


  EBCT = 20 min.

  c0 =288 |jg/L Cl-
                                SDS-TOX not analyzed in the
                                     blended effluent
                                              Observed data
                                              Best fit
                                              Dl prediction  (RSS = NA)
                                              SCA prediction (RSS = NA)
                           50                100
                                 Scaled operation time (days)
                                                  150
                                                     200
Figure F-63 Comparison of Dl and SCA methods for predicting the SDS-TOX
integral breakthrough curve for Water 4
      14
      12 -
      10 -
    o
    "co
    o>
    o
    c
    o
   O
       6 -
       4 -
       2 -
SDS-CF


  EBCT = 20 min.
                                          •   Observed data
                                        	Best fit
                                        	Dl prediction (RSS = 0.41)
                                        	SCA prediction (RSS = 1.33)
50               100
       Scaled operation time (days)
                                                             150
                                                                   200
Figure F-64 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 4
                                       -337-

-------
     3 -
  g
  •^
  CD
  CD
  O
  c
  O
  O
     1 -
SDS-BDCM



 EBCT = 20 mm.

 c0 = 2.2Mg/L
                                                        Observed data

                                                        Best fit

                                                        Dl prediction (RSS = 0.58)

                                                        SCA prediction (RSS = 0.55)
                        50                100

                               Scaled operation time (days)
                                                   150
200
Figure F-65  Comparison of Dl and SCA methods for predicting the SDS-
BDCM integral breakthrough curve for Water 4
    6 -
  o
  I 4H



  I
  o
  O

    2 -
SDS-DBCM



EBCT = 20 min.


c0= 15.1M9/L
                                                     •  Observed data

                                                   	Best fit

                                                   	Dl prediction (RSS = 0.59)

                                                   	SCA prediction (RSS = 0.55)
                        50                 100

                               Scaled operation time (days)
                                                   150
 200
Figure F-66 Comparison of Dl and SCA methods for predicting the SDS-
DBCM integral breakthrough curve for Water 4
                                       -338-

-------
   2.0
   1.5 -
 g

 '•SS 1-0 H
 O>
 o
 c
 o
 O


   0.5 -
   0.0
SDS-BF




  EBCT = 20 min.


  c0= BMRL
                      Blended effluent concentrations were not detected

                            above the MRL for this parameter
                                                      •   Observed data

                                                    	Best fit

                                                    	Dl prediction  (RSS = NA)

                                                    	SCA prediction (RSS = NA)
                         50                 100

                                Scaled operation time (days)
                                                    150
200
Figure F-67 Comparison of Dl and SCA methods for predicting the SDS-BF

integral breakthrough curve for Water 4
    25
    20 -
  O)
    15 H
  o
  '-4— '
  § 10
  c
  o
  O
     5 -
SDS-TTHM




 EBCT = 20 min.


  c0 = 68 |jg/L
                                                 Observed data

                                                 Best fit

                                                 Dl prediction (RSS = 1.39)

                                                 SCA prediction (RSS = 2.03)
                         50                100

                                 Scaled operation time (days)
                                                     150
 200
Figure F-68 Comparison of Dl and SCA methods for predicting the SDS-

TTHM integral breakthrough curve for Water 4
                                        -339-

-------
   2.0
   1.5 -
 c
 o
 ro
 o
 c
 o
 O
   1.0 -
   0.5 -
  SDS-MCAA

EBCT = 20 min.
 CD= BMRL
                    Blended effluent concentrations were not detected above the
                                  MRL for this parameter
                                                      •   Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = NA)
                                                    	SCA prediction (RSS = NA)
   0.0 4«-
      0
                 50                100
                         Scaled operation time (days)
150
200
Figure F-69 Comparison of Dl and SCA methods for predicting the SDS-
MCAA integral breakthrough curve for Water 4
   6 -
   5 -
 SDS-DCAA

EBCT = 20 min.

 co = 20.3 |jg/L
                                                      •   Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = 0.32)
                                                    	SCA prediction (RSS = 0.47)
                        50
                                   100
                        Scaled operation time (days)
150
200
Figure F-70 Comparison of Dl and SCA methods for predicting the SDS-
DCAA integral breakthrough curve for Water 4
                                        -340-

-------
   10
 O)
    6 -
 c
 o
 •*=
 £5
 -H-
 C
 0>  „
 o  4
 c
 o
 O
    2 -
           SDS-TCAA

          EBCT = 20 min.

          c0 = 30.7 |jg/L
                                               •  Observed data
                                            	Best fit
                                            	Dl prediction  (RSS = 0.96)
                                            	SCA prediction (RSS = 0.64)
                        50                 100
                                 Scaled operation time (days)
                                                     150
200
Figure F-71 Comparison of Dl and SCA methods for predicting the SDS-
TCAA integral breakthrough curve for Water 4
   2.0
    1.5 -
 c
 o
 jo  1-0 H
 -i—<
 c
 o>
 o
 c
 o
 O
   0.5 -
 SDS-MBAA

EBCT = 20 min.
co= BMRL
   0.0 -k
      0
                 Blended effluent concentrations were not detected
                       above the MRL for this parameter
                                                       •  Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = NA)
                                                     	SCA prediction  (RSS = NA)
50                 100
        Scaled operation time (days)
                                                      150
200
Figure F-72 Comparison of Dl and SCA methods for predicting the SDS-
MBAA integral breakthrough curve for Water 4
                                        -341-

-------
   2.0
    1.5 -
    1.0 H
 o
 O
   0.5 -
   0.0
SDS-DBAA

 EBCT = 20 min.
 c0 =  1 H9/L
                    Blended effluent concentrations were not detected above the
                                  MRL for this parameter
                                                       •   Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = NA)
                                                     	SCA prediction (RSS = NA)
                         50                100
                                 Scaled operation time (days)
                                                      150
200
Figure F-73 Comparison of Dl and SCA methods for predicting the SDS-
DBAA integral breakthrough curve for Water 4
                                                         Observed data
                                                         Best fit
                                                   	Dl prediction (RSS = 1.24)
                                                         SCA prediction (RSS = 0.87)
                         50                100
                                 Scaled operation time (days)
                                                     150
200
Figure F-74 Comparison of Dl and SCA methods for predicting the SDS-
HAA5 integral breakthrough curve for Water 4
                                        -342-

-------
   3.0
   2.5 -
   2.0 -
 O)
o
<§
   1-5 H
   1.0 H
   0.5 -
   0.0
        SDS-BCAA

       EBCT = 20 min.
        co = 4.3 |jg/L
                                                      •   Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = 0.31)
                                                   	SCA prediction (RSS = 0.22)
                         50                 100
                                 Scaled operation time (days)
                                                             150
 200
Figure F-75 Comparison of Dl and SCA methods for predicting the SDS-
BCAA integral breakthrough curve for Water 4
      20
      15 -
    o
    jo 10 H
    -I—I
    I
    o
    O
       5 -
           SDS-HAA6

           EBCT = 20 min.
            co = 55 |jg/L
                                                      •  Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = 1.34)
                                                    	SCA prediction  (RSS = 1.07)
                          50                100
                                 Scaled operation time (days)
                                                             150
200
Figure F-76 Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 4
                                       -343-

-------
  3.0
  2.5-
 r2.o-
c
o
 ;i.5 -
o
c
O1.0 -
  0.5-
  0.0
SDS-DCBAA


EBCT = 20 min.
c0= 6.7|jg/L
                                                 •  Observed data
                                              	Best fit
                                              	Dl prediction (RSS = 0.49)
                                              	SCA prediction  (RSS = 0.5)
                        50                 100
                                 Scaled operation time (days)
                                                       150
200
Figure F-77 Comparison of Dl and SCA methods for predicting the SDS-
DCBAA integral breakthrough curve for Water 4
    2.0
    1.5-
  O)
  o
  't5 1.0 H
  8
  c
  o
  O
    0.5 -
    0.0
         SDS-CDBAA
 EBCT = 20 min.
  CD= BMRL
                 Blended effluent concentrations were not detected
                       above the MRL for this parameter
                                                •  Observed data
                                              	Best fit
                                              	Dl prediction (RSS = NA)
                                              	SCA prediction (RSS = NA)
                          50                 100
                                 Scaled operation time (days)
                                                       150
200
Figure F-78 Comparison of Dl and SCA methods for predicting the SDS-
CDBAA integral breakthrough curve for Water 4
                                        -344-

-------
  2.0
  1.5-
o
 M.O-
0)
o
o
O
  0.5 -
  0.0
SDS-TBAA

EBCT = 20 m\n.
co=  BMRL
                  Blended effluent concentrations were not detected
                        above the MRL for this parameter
                                                 •   Observed data
                                               	Best fit
                                               	Dl prediction (RSS = NA)
                                               	SCA prediction (RSS = NA)
50                 100
         Scaled operation time (days)
                                                              150
                                                                           200
Figure F-79 Comparison of Dl and SCA methods for predicting the SDS-
TBAA integral breakthrough curve for Water 4
     25
     20-
     15 -
  c
  o
  <&  in J
  O  10 -
  C
  o
  O
     5 -
   SDS-HAA9

   EBCT = 20 min.

   cn=61 M9/L
                                                 •   Observed data
                                               	Best fit
                                               	Dl prediction (RSS = 1.7)
                                               	SCA prediction (RSS = 1.12)
                          50                 100
                                  Scaled operation time (days)
                                                        150
                                                          200
Figure F-80 Comparison of Dl and SCA methods for predicting the SDS-
HAA9 integral breakthrough curve for Water 4
                                        -345-

-------
       1.75
       1.50 -
       1.25 H
    ^ 1.00 H
    o
    u-»
    TO
    c 0.75 -
    
-------
   84

   72 -

O  60 -
      CD
      O

      I
   48 -

   36 -
        12 -

         0
0
                SDS-TOX
                EBCT = 20 min.
                c0 = 205 |jg/L Cl-
                                 SDS-TOX not analyzed in
                                   the blended effluent
                                                 •  Observed data
                                               	Best fit
                                               	Dl prediction  (RSS = NA)
                                               	SCA prediction (RSS = NA)
                    50        100       150       200       250
                                  Scaled operation time (days)
                                                               300
          350
Figure F-83 Comparison of Dl and SCA methods for predicting the SDS-TOX
integral breakthrough curve for Water 5
       4 -
       3 -
    c
    o
    CD
    O
    c
    o
    O
       1 -
        SDS-CF
        EBCT = 20 min.
        c0 =23.7 pg/L
                                                •   Observed data
                                              	Best fit
                                              	Dl prediction (RSS = 0.31)
                                              	SCA prediction (RSS = 0.65)
                  50
                       100        150        200        250
                           Scaled operation time (days)
300
                                                                    350
Figure F-84 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 5
                                       -347-

-------
      12
      10 -
   O)
   c
   g
   'CD   6
   O>
   o
   3   4
       2 -
SDS-BDCM
EBCT = 20 min
c0 = 10.8 |jg/L
                                        •  Observed data
                                      	Best fit
                                      	Dl prediction  (RSS = 0.85)
                                      	SCA prediction (RSS = 0.34)
                  50
               100       150       200       250
                    Scaled operation time (days)
300
350
Figure F-85 Comparison of Dl and SCA methods for predicting the SDS-
BDCM integral breakthrough curve for Water 5
      10
       6 -
   o
   "co
   o
   O
       2 -
              SDS-DBCM
              EBCT = 20 min.
              C0 = 22.7 M9/L
                                         •   Observed data
                                         	Best fit
                                         	Dl prediction (RSS = 0.54)
                                         	SCA prediction  (RSS = 0.72)
                  50
               100       150       200       250
                    Scaled operation time (days)
 300
 350
Figure F-86 Comparison of Dl and SCA methods for predicting the SDS-
DBCM integral breakthrough curve for Water 5
                                       -348-

-------
     3.0
     2.5 -
     2.0 -
  g
  '•a  1.5 H
  O>
     0.5 -
     0.0
SDS-BF
EBCT = 20 min.
c0= 1-2M9/L
                                        •  Observed data
                                      	Best fit
                                      	Dl prediction (RSS = 0.55)
                                      	SCA prediction  (RSS = 0.26)
                  50
               100        150       200       250
                    Scaled operation time (days)
300
350
Figure F-87 Comparison of Dl and SCA methods for predicting the SDS-BF
integral breakthrough curve for Water 5
      30
      25 -
      20 -
SDS-TTHM
EBCT = 20 min.
c0 =  58 |jg/L
                                                       •   Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = 1.89)
                                                    	SCA prediction (RSS = 1.3)
                  50
                100        150        200        250
                    Scaled operation time (days)
 300
 350
Figure F-88 Comparison of Dl and SCA methods for predicting the SDS-TTHM
integral breakthrough curve for Water 5
                                       -349-

-------
     2.0
     1.5 -
   g
  '•a 1.0 H
   CD
   o
   c
   o
  O
     0.5 -
     0.0
            SDS-MCAA
            EBCT = 20 min.
            c0 = 2 |jg/L
Blended effluent concentrations were not detected
      above the MRL for this parameter
                               •  Observed data
                             	Best fit
                             	Dl prediction (RSS = NA)
                             	SCA prediction  (RSS = NA)
                  50
    100       150        200        250
         Scaled operation time (days)
300
350
Figure F-89 Comparison of Dl and SCA methods for predicting the SDS-
MCAA integral breakthrough curve for Water 5
     3 -
   o
  •^
   CD
   CD
   O
   c
   o
  O
     1  -
           SDS-DCAA
           EBCT = 20 min.
           co= 10.3|jg/L
                              •   Observed data
                            	Best fit
                            	Dl prediction (RSS = 0.49)
                            	SCA prediction (RSS = 0.59)
                 50
   100         150        200        250
        Scaled operation time (days)
300
350
Figure F-90 Comparison of Dl and SCA methods for predicting the SDS-
DCAA integral breakthrough curve for Water 5
                                        -350-

-------
    3 -
 O)
 o
 *J  O
 s
 I
 o
 O
    1 -
    0 -li
          SDS-TCAA
          EBCT = 20 min.
          co= 12.7|jg/L
                                                      •   Observed data
                                                          Best fit
                                                          Dl prediction  (RSS = 0.28)
                                                          SCA prediction (RSS = 0.39)
                50
  100        150         200        250
        Scaled operation time (days)
300
 350
Figure F-91 Comparison of Dl and SCA methods for predicting the SDS-
TCAA integral breakthrough curve for Water 5
     2.0
     1.5 -
  o
  '•SS  1-0 H
  o>
  o
  c
  o
  O
     0.5 -
     0.0
            SDS-MBAA
            EBCT = 20 min.
            c0 = 1 H9/L
Blended effluent concentrations were not detected
      above the MRL for this parameter
                               •  Observed data
                             	Best fit
                             	Dl prediction (RSS = NA)
                             	SCA prediction  (RSS = NA)
                  50
   100        150        200        250
         Scaled operation time (days)
300
350
Figure F-92 Comparison of Dl and SCA methods for predicting the SDS-
MBAA integral breakthrough curve for Water 5
                                        -351-

-------
    3.5
    3.0 -
   2.5 -
 O)
 o
 O
    1.5 H
    1.0 H
    0.5 -
    0.0
SDS-DBAA
EBCT = 20 min
co = 2 |jg/L
                                             •   Observed data
                                           	Best fit
                                           	Dl prediction (RSS = 0.47)
                                           	SCA prediction (RSS = 0.2)
                 50
                  100        150        200        250
                        Scaled operation time (days)
300
350
Figure F-93 Comparison of Dl and SCA methods for predicting the SDS-
DBAA integral breakthrough curve for Water 5
     12
     10 -
      8 -
  o
  '•a
  o>
  o
  o   4 .
  O    1
      2 -
  SDS-HAA5
  EBCT = 20 min.
  c0 = 28 |jg/L
                                             •   Observed data
                                           	Best fit
                                           	Dl prediction (RSS = 0.56)
                                           	SCA prediction  (RSS = 0.52)
                 50
                   100        150       200        250
                        Scaled operation time (days)
300
350
Figure F-94 Comparison of Dl and SCA methods for predicting the SDS-
HAA5 integral breakthrough curve for Water 5
                                       -352-

-------
    4 -
    3 -
 .o
 "co
  o>
  o
  c
  o
 O
    1  -
       SDS-BCAA
        EBCT = 20 min

        c° = 7.3 |jg/L
                                                   •  Observed data
                                                 	Best fit
                                                 	Dl prediction  (RSS = 0.4)
                                                 	SCA prediction (RSS = 0.21)
                50
                       100        150        200        250
                             Scaled operation time (days)
300
350
Figure F-95 Comparison of Dl and SCA methods for predicting the SDS-
BCAA integral breakthrough curve for Water 5
   16

   14 -

   12 -
Q"
1 1°-
o
'•S3   8
   o>
   c  6
   o
   O
      4 -
      2 -
           SDS-HAA6
           EBCT = 20 min.
           c0 =  34 |jg/L
                                                   •  Observed data
                                                 	Best fit
                                                 	Dl prediction  (RSS = 0.82)
                                                 	SCA prediction (RSS = 0.55)
                  50
                         100        150       200       250
                              Scaled operation time (days)
300
350
Figure F-96 Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 5
                                       -353-

-------
   5 -
   4 •
 O)
 o
 *J
 CD
 CD
 O
 c
 0 o
 O 2
   1 -
SDS-DCBAA
EBCT = 20 min.
c0=  10.7|jg/L
                                            •  Observed data
                                          	Best fit
                                          	Dl prediction  (RSS = 0.4)
                                          	SCA prediction (RSS = 0.34)
               50
                100        150         200
                      Scaled operation time (days)
250
300
350
Figure F-97 Comparison of Dl and SCA methods for predicting the SDS-
DCBAA integral breakthrough curve for Water 5
    4 -
    3 -
  CD
  O
  c
  o
  O
     1 -
SDS-CDBAA
EBCT = 20 min.
co = 3.7 |jg/L
                                            •   Observed data
                                          	Best fit
                                          	Dl prediction (RSS = 0.88)
                                          	SCA prediction  (RSS = 1.1]
                50
                 100        150        200        250
                      Scaled operation time (days)
           300
          350
Figure F-98 Comparison of Dl and SCA methods for predicting the SDS-
CDBAA integral breakthrough curve for Water 5
                                       -354-

-------
   2.0
   1.5 -
o
'•S  1.0 H
 o>
 o
 c
 o
o
   0.5 -
   0.0
         SDS-TBAA
         EBCT = 20 min.
         c0= BMRL
                        Blended effluent concentrations were not detected
                              above the MRL for this parameter
                                                       •  Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = NA)
                                                     	SCA prediction  (RSS = NA)
                50
                           100        150        200
                                Scaled operation time (days)
250
300
350
Figure F-99 Comparison of Dl and SCA methods for predicting the SDS-
TBAA integral breakthrough curve for Water 5
    25
    20 -
  a  15 -\
  o
  '-4—I
  E
  -i—i
  |  10-|
  o
  O
     5 -
           SDS-HAA9
           EBCT = 20 min
           c0 =  48 pg/L
                                                       •  Observed data
                                                       	Best fit
                                                       	Dl prediction (RSS = 1.19)
                                                      	SCA prediction  (RSS = 1.13)
                 50
                           100        150        200        250
                                 Scaled operation time (days)
           300
           350
 Figure F-100 Comparison of Dl and SCA methods for predicting the SDS-
 HAA9 integral breakthrough curve for Water 5
                                        -355-

-------
       1.25
       1.00 -
    O)
       0.75 -
    g  0.50 H

    c
    o
    O
       0.25 -
       0.00
              TOO
EBCT = 20 min.


c0 = 2.64 mg/L
                                                        •  Observed data

                                                     	Best fit

                                                     	Dl prediction (RSS = 0.015)
                       50
                     100         150          200


                      Scaled operation time (days)
250
300
Figure F-101  Dl method prediction of the TOC integral breakthrough curve

for Water 6
       0.025
       0.020 -
    o

    ^ 0.015

    o>
    o
    c
    CD
    .Q

    o  0.010
    
-------
    o
    o
    '-4—'
    CD
    CD
    O
    C
    O
    O
       100
        80 -
        60 H
        40 -
        20 -
  SDS-TOX


 EBCT = 20 min.

 co = 305 |jg/L Cl-
                                          •  Observed data

                                        	Best fit

                                        	Dl prediction (RSS = 2.39)

                                        	SCA prediction  (RSS = 6.26)
                     50
                     100         150         200

                      Scaled operation time (days)
                                   250
300
Figure F-103 Comparison of Dl and SCA methods for predicting the SDS-TOX
integral breakthrough curve for Water 6
       6 -
    O
    '•£  4
    CD
    O
    c
    o
    O
 SDS-CF


EBCT = 20 min

c0 = 55.3 pg/L
                    50
  •   Observed data

	Best fit

	Dl prediction (RSS = 1.37)

	SCA prediction (RSS = 0.74)
                    100          150          200

                     Scaled operation time (days)
                                  250
300
Figure F-104 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 6
                                       -357-

-------
     30
    25 -
    20 -
     15 -
  CD
  O
  c
  °  m
  O  1U
     5 -
SDS-BDCM





 EBCT = 20 min.

 c0 = 27.4 |jg/L
                                               Observed data

                                               Best fit

                                               Dl prediction  (RSS = 3.63)

                                               SCA prediction (RSS = 2.31)
                   50
                      100          150         200


                       Scaled operation time (days)
250
300
Figure F-105 Comparison of Dl and SCA methods for predicting the SDS-

BDCM integral breakthrough curve for Water 6
     16




     14




     12




     10
  o

  "CD
  CD
  O
  c
  o
  O
SDS-DBCM




EBCT = 20 min.


C0=  41.6|jg/L
                                                     •  Observed data

                                                   	Best fit

                                                   	Dl prediction (RSS = 1.02)

                                                   	SCA prediction (RSS = 1.84)
                   50
                      100          150          200


                        Scaled operation time (days)
 250
 300
Figure F-106 Comparison of Dl and SCA methods for predicting the SDS-

DBCM integral breakthrough curve for Water 6
                                       -358-

-------
     20
     15 -
   g

  '•S3 10 H
   O>
   o
   c
   o
  O

      5 -
SDS-BF




 EBCT = 20 min.


 c0 = 3.3 |jg/L
                                                     I   Observed data

                                                     - - - Best fit

                                                     — Dl prediction (RSS = 3.47)

                                                     — SCA prediction (RSS = 1.86)
                   50
                     100         150         200


                       Scaled operation time (days)
250
300
Figure F-107 Comparison of Dl and SCA methods for predicting the SDS-BF

integral breakthrough curve for Water 6
      70
      60 -
      50 -
      40 -
    o

    "co
    o>
    o
    c
    o
    O
      30 -
      20 -
      10 -
 SDS-TTHM




 EBCT = 20 min.


 c0=  128|jg/L
                                            •   Observed data

                                          	Best fit

                                          	Dl prediction (RSS = 6.43)

                                          	SCA prediction (RSS = 5.91)
                    50
                      100         150         200


                        Scaled operation time (days)
 250
 300
Figure F-108 Comparison of Dl and SCA methods for predicting the SDS-

TTHM integral breakthrough curve for Water 6
                                       -359-

-------
   2.0
   1.5 -
 o
   1.0 -
 o
 c
 o
 O
   0.5 -
   0.0
         SDS-MCAA

         EBCT = 20 min.
         Co= BMRL
                       Blended effluent concentrations were not detected above the
                                     MRL for this parameter
                                                       •   Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = NA)
                                                          -SCA prediction (RSS = NA)
                                       •   •
                  50
                      100          150          200
                         Scaled operation time (days)
                         250
                                                                                 300
Figure F-109 Comparison of Dl and SCA methods for predicting the SDS-
MCAA integral breakthrough curve for Water 6
   3.5
   3.0 -
   2.5 -
SDS-DCAA

 EBCT = 20 min
 co= 17.3|jg/L
       0
          50
Observed data
Best fit
                                     	Dl prediction (RSS = 0.47)
                                           SCA prediction (RSS = 0.31)   /
                         250
                               100          150          200
                                 Scaled operation time (days)
Figure F-110 Comparison of Dl and SCA methods for predicting the SDS-
DCAA integral breakthrough curve for Water 6
300
                                        -360-

-------
   2.0
   1.5 -
 o
 IS 1.0
 o
 O
   0.5 -
   o.o
SDS-TCAA

 EBCT = 20 min.
 c0 =   13.3|jg/L
                           Insufficient data measured above the MRL to compare
                                       Dl and SCA predictions
                                                        •   Observed data
                                                      	Best fit
                                                      	Dl prediction (RSS = NA)
                                                      	SCA prediction (RSS = NA)
                   50
                       100          150          200
                          Scaled operation time (days)
                                     250
             300
Figure F-111 Comparison of Dl and SCA methods for predicting the SDS-
TCAA integral breakthrough curve for Water 6
     2.0
     1.5 -
  o
  '•&  1.0 H
  o>
  o
  c
  o
  O
     0.5 -
 SDS-MBAA

   EBCT = 20 min.
   c0= BMRL
                Blended effluent concentrations were not detected
                      above the MRL for this parameter
                                               •   Observed data
                                             	Best fit
                                             	Dl prediction  (RSS = NA)
                                             	SCA prediction (RSS = NA)
     0.0 -!•-
        0
            50
100         150          200
  Scaled operation time (days)
250
300
Figure F-112 Comparison of Dl and SCA methods for predicting the SDS-
MBAA integral breakthrough curve for Water 6
                                        -361-

-------
   10
    6 -
 o
 '-4—'
 CD
 CD  ..
 o  4

 o
 o
    2 -
        SDS-DBAA




        EBCT = 20 min.


        c0 = 5.7 |jg/L
                                          •   Observed data

                                        	Best fit

                                        	Dl prediction (RSS = 1.31)

                                        	SCA prediction  (RSS = 1.1
                  50
                  100          150          200

                     Scaled operation time (days)
250
300
Figure F-113 Comparison of Dl and SCA methods for predicting the SDS-

DBAA integral breakthrough curve for Water 6
     14
     12 -
     10 -
  o
  '-4—»
  CD
  CD
  O
  C
  O
  O
     6 -
     4 -
     2 -
SDS-HAA5



EBCT = 20 min.

c0 = 36 |jg/L
                                          •  Observed data

                                        	Best fit

                                        	Dl prediction (RSS = 1.02)

                                        	SCA prediction (RSS = 1.54)
                   50
                   100         150         200

                     Scaled operation time (days)
250
300
Figure F-114 Comparison of Dl and SCA methods for predicting the SDS-

HAA5 integral breakthrough curve for Water 6
                                       -362-

-------
   5 -
   4 -
 O)
   3 -
   1 -
   o -k
SDS-BCAA

 EBCT = 20 min.
 c0  12.3 |jg/L
                                              •   Observed data
                                            	Best fit
                                            	Dl prediction (RSS = 0.43)
                                            	SCA prediction (RSS = 0.44)
                 50
                      100          150         200
                        Scaled operation time (days)
250
300
Figure F-115 Comparison of Dl and SCA methods for predicting the SDS-
BCAA integral breakthrough curve for Water 6
    20
     15 -
  o
  '•&  10
  o>
  o
  c
  o
  O
     5 -
  SDS-HAA6

  EBCT = 20 min.
  c0 =  49 pg/L
                                              •  Observed data
                                            	Best fit
                                            	Dl prediction (RSS = 1.3)
                                            	SCA prediction (RSS = 1.92)
                   50
                       100          150         200
                         Scaled operation time (days)
250
300
Figure F-116 Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 6
                                       -363-

-------
    2.5
    2.0 -
    1.5 -
SDS-DCBAA


 EBCT = 20 min.

 co =  9 pg/L
 §  1.0 H
 o
 O
    0.5 -
                   50
       •   Observed data
     	Best fit
     	Dl prediction (RSS = 0.34)
                                         SCA prediction (RSS = 0.34)
                       100          150          200
                         Scaled operation time (days)
                                      250
 300
Figure F-117 Comparison of Dl and SCA methods for predicting the SDS-
DCBAA integral breakthrough curve for Water 6
    3.0
    2.5 -
    2.0 -
  O)
    1-5 H
    0.5 -
    0.0
  SDS-CDBAA


  EBCT = 20 min.

  c0 =  3 |jg/L
                   50
  •   Observed data
	Best fit
	Dl prediction (RSS = 0.72)
	SCA prediction  (RSS = 0.67)
                       100         150         200
                         Scaled operation time (days)
                                      250
Figure F-118 Comparison of Dl and SCA methods for predicting the SDS-
CDBAA integral breakthrough curve for Water 6
300
                                       -364-

-------
    2.0
    1.5 -
 O)
 g
 jo  1-0 H
 -i— •

 I
 o
 O
    0.5 -
    0.0
         SDS-TBAA

         EBCT = 20 min.
         c0=  BMRL
                Blended effluent concentrations were not detected above the
                              MRL for this parameter
                                                        •   Observed data
                                                      	Best fit
                                                      	Dl prediction (RSS = NA)
                                                      	SCA prediction (RSS = NA)
                   50
                       100          150          200
                         Scaled operation time (days)
250
300
Figure F-119 Comparison of Dl and SCA methods for predicting the SDS-
TBAA integral breakthrough curve for Water 6
    25
    20 -
 O)
 o
 '-4— '
 o
 O
    15 H
    10 -
     5 -
SDS-HAA9

 EBCT = 20 min.
 Co =  61 |jg/L
                                                   Observed data
                                                   Best fit
                                             	Dl prediction (RSS = 1.88)
                                             	SCA prediction (RSS = 2.41)
                   50
                       100           150          200
                         Scaled operation time (days)
250
300
Figure F-120 Comparison of Dl and SCA methods for predicting the SDS-
HAA9 integral breakthrough curve for Water 6
                                        -365-

-------
      2.5
      2.0 -
    O)
    g
    '-4— '
    CD
    8 to H
    o
    O
      0.5 -
      0.0
  TOC



EBCT = 20 min.

co = 5.58 mg/L
                                                    X-
                                                  X. '
                          x .
                        X , '
                       X , '
                                                        •   Observed data

                                                     	Best fit

                                                     	Dl prediction (RSS = 0.065)
                     25
                      50          75          100

                       Scaled operation time (days)
125
150
Figure F-121  Dl method prediction of the TOC integral breakthrough curve
for Water 7
      0.04
      0.03 -
    o
    CD
    CD
    O
    to
    .a
      0.02 H
      0.01 -
      0.00
   UV254


 EBCT = 20 min.

 CD =0.109 1/cm
                                                     •   Observed data

                                                   	Best fit

                                                   	Dl prediction (RSS = 0.0015)

                                                   	SCA prediction  (RSS = 0.0021]
                      25
                       50          75          100

                        Scaled operation time (days)
 125
 150
Figure F-122 Comparison of Dl and SCA methods for predicting the UV254
integral breakthrough curve for Water 7
                                        -366-

-------
       200
       150 -
    O
    .2  100 -
    "en
    O>
    o
    c
    o
    0   50 H
    SDS-TOX

  EBCT = 20 min.
  co = 486 |jg/L Cl
                                                      •   Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = 6.92)
                                                    	SCA prediction  (RSS = 9.3)
                     25
                       50          75          100
                        Scaled operation time (days)
                                    125
150
Figure F-123 Comparison of Dl and SCA methods for predicting the SDS-
TOX integral breakthrough curve for Water 7
      5 -
      4 -
    o>
    o
      1 -
 SDS-CF

EBCT = 20 min.
C0 = 52.4 |jg/L
  •   Observed data
	Best fit
	Dl prediction (RSS = 0.85)
	SCA prediction  (RSS = 0.46)
                   25
                     50          75          100
                       Scaled operation time (days)
                                    125
150
Figure F-124 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 7
                                       -367-

-------
    60
    50 -
    40 -
  O)
  o
  8 30
  CD
  o

  5 20
    10 -
  SDS-BDCM


  EBCT = 20 min.

  Co = 65.3 |jg/L
                                               •  Observed data

                                             	Best fit

                                             	Dl prediction (RSS = 5.57)

                                             	SCA prediction (RSS = 3.38)
                  25
                        50           75          100

                         Scaled operation time (days)
125
150
Figure

BDCM
F-125 Comparison of Dl and SCA methods for predicting the SDS-
integral breakthrough curve for Water 7
    25
    20 -
    15 -
  o
  '-4—'
  CD
  i_


  CD
  O
  c
  o
  O
     5 -
   SDS-DBCM



   EBCT = 20 min.

   c0 = 66.2 |jg/L
                                               •   Observed data

                                             	Best fit

                                             	Dl prediction (RSS = 0.99)

                                             	SCA prediction (RSS = 1.1]
                  25
                        50          75          100

                          Scaled operation time (days)
 125
 150
Figure F-126 Comparison of Dl and SCA methods for predicting the SDS-
DBCM integral breakthrough curve for Water 7
                                       -368-

-------
   50
   40 -
   30 -
.o

"co
 o>
 o
 c
 o
 O
   10 -
SDS-BF





EBCT = 20 min.


c°= 16 |jg/L
                                            •   Observed data

                                          	Best fit

                                          	Dl prediction (RSS = 11.53)

                                          	SCA prediction (RSS = 5.39)
                 25
                      50           75           100


                       Scaled operation time (days)
125
150
Figure F-127 Comparison of Dl and SCA methods for predicting the SDS-BF

integral breakthrough curve for Water 7
    125
    100 -
 a  75 H


 o
 '-4—'
 CD

 £

 8  50

 o
 O
     25 -
  SDS-TTHM





  EBCT = 20 min.


   co = 200 |jg/L
                                              •  Observed data


                                           	Best fit


                                           	Dl prediction (RSS = 15.22)


                                           	SCA prediction (RSS = 9.72)
                   25
                       50          75          100


                         Scaled operation time (days)
 125
 150
Figure F-128 Comparison of Dl and SCA methods for predicting the SDS-

TTHM integral breakthrough curve for Water 7
                                       -369-

-------
   2.0
   1.5 -
 O)
 o
 '•a 1.0
 O>
 o
 c
 o
 O
   0.5 -
   0.0
Figure

MCA A
        SDS-MCAA
         EBCT = 20 min.
                 Blended effluent concentrations were not detected above the

                               MRL for this parameter
                                                        •  Observed data

                                                      	Best fit

                                                      	Dl prediction (RSS = NA)

                                                      	SCA prediction  (RSS = NA)
                  25
                        50           75          100


                          Scaled operation time (days)
125
                                                                                 150
F-129  Comparison of Dl and SCA methods for predicting the SDS-

integral breakthrough curve for Water 7
   4 -
 O)
   3 -

 c
 o
 '-4— <
 cc
 l_
 -I— «


 § 2 -I
 c
 o
 O
   1 -
   0
 SDS-DCAA





 EBCT = 20 min


 co = 25.3 |jg/L
                                                •   Observed data


                                              	Best fit


                                              	Dl prediction  (RSS = 0.59)


                                              	SCA prediction (RSS = 0.63)
     0
           25
125
                              50           75           100

                                Scaled operation time (days)


Figure F-130 Comparison of Dl and SCA methods for predicting the SDS-

DCAA integral breakthrough curve for Water 7
150
                                        -370-

-------
   3.5
   3.0 -
   2.5 -
   2.0-
 SDS-TCAA

EBCT = 20 min.

co= 18 |jg/L
                                 •   Observed data
                                	Dl prediction (RSS = 0.4)
                                        . prediction (RSS = 1.34)
 8
 o
 0
   1.0 -
   0.5 -
                   25
50           75           100
  Scaled operation time (days)
                                                            125
150
Figure F-131 Comparison of Dl and SCA methods for predicting the SDS-
TCAA integral breakthrough curve for Water 7
     2.0
     1.5 -
  O)
  o
  IS  1.0 H
  8
  c
  o
  o
     0.5 -
     0.0
  SDS-MBAA

  EBCT = 20 min.
  C0 = BMRL
                        Blended effluent concentrations were not detected
                              above the MRL for this parameter
                                                       •  Observed data
                                                     	Best fit
                                                     	Dl prediction (RSS = NA)
                                                              prediction (RSS = NA)
                    25
                       50           75          100
                         Scaled operation time (days)
                                     125
150
Figure F-132 Comparison of Dl and SCA methods for predicting the SDS-
MBAA integral breakthrough curve for Water 7
                                        -371-

-------
                                                         Observed data
                                                         Best fit
                                                    	Dl prediction  (RSS = 2.21)
                                                         SCA prediction (RSS = 0.97)
                  25
                    50           75           100
                      Scaled operation time (days)
125
150
Figure F-133 Comparison of Dl and SCA methods for predicting the SDS-
DBAA integral breakthrough curve for Water 7
    25
    20 -
    15 -
  CD
  O
  c
  o
  O
     5 -
SDS-HAA5

EBCT = 20 min.
c0 = 65 |jg/L
                                           •   Observed data
                                         	Best fit
                                         	Dl prediction (RSS = 2.02)
                                         	SCA prediction (RSS = 1.52)
                   25
                     50          75          100
                       Scaled operation time (days)
125
150
Figure F-134 Comparison of Dl and SCA methods for predicting the SDS-
HAA5 integral breakthrough curve for Water 7
                                       -372-

-------
   10
        SDS-BCAA


        EBCT = 20 min.
        c0 = 23.7 |jg/L
                                                      •  Observed data
                                                    	Best fit
                                                    	Dl prediction (RSS = 0.35)
                                                    	SCA prediction (RSS = 1.1)
                  25
                    50           75           100
                      Scaled operation time (days)
125
150
Figure F-135 Comparison of Dl and SCA methods for predicting the SDS-
BCAA integral breakthrough curve for Water 7
    35
    30 -
    25 -
    20 -
  CD
  O
  O
  O
     15 -
     10 -
     5 -
SDS-HAA6

EBCT = 20 min.
c0 = 85 pg/L
                                           •   Observed data
                                         	Best fit
                                         	Dl prediction (RSS = 2.09)
                                         	SCA prediction (RSS = 2.25)
                   25
                     50          75          100
                       Scaled operation time (days)
125
150
Figure F-136 Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 7
                                       -373-

-------
    4 -
 O)
    3 -
 -—


 §  2
 c
 o
 O
    1 -
          SDS-DCBAA
EBCT = 20 min.


c° = 26.7 |jg/L
                                               •   Observed data

                                            	Best fit

                                            	Dl prediction (RSS = 1.13)

                                            	SCA prediction (RSS = 0.85)
                 25
                      50           75           100

                        Scaled operation time (days)
125
150
Figure F-137 Comparison of Dl and SCA methods for predicting the SDS-

DCBAA integral breakthrough curve for Water 7
    5 -
    4 -
 O)
 g
 't->  O _
 CO
 -I—'


 I   1
 o  ^
 O  "
    1 -
        SDS-CDBAA
       EBCT = 20 min.

                                              •   Observed data

                                            	Best fit

                                            	Dl prediction (RSS = 0.25)

                                            	SCA prediction  (RSS = 1.02)
                 25
                      50           75           100

                        Scaled operation time (days)
125
150
Figure F-138 Comparison of Dl and SCA methods for predicting the SDS-

CDBAA integral breakthrough curve for Water 7
                                       -374-

-------
    6 -
    5 -
 O)
    3 -
 o
 O
    2 -
    1 -
  SDS-TBAA


EBCT = 20 min.
c0= BMRL
                                              •   Observed data
                                            	Best fit
                                            	Dl prediction (RSS = 1.92)
                                            	SCA prediction  (RSS = 1.84)
                 25
                       50          75          100
                         Scaled operation time (days)
125
150
Figure F-139 Comparison of Dl and SCA methods for predicting the SDS-
TBAA integral breakthrough curve for Water 7
    60
    50 -
    40 -
 o
 |8  30 H
 -I—'

 I    1
 5  2°H
    10 -
  SDS-HAA9

 EBCT = 20 min.
 c0=  124|jg/L
                                               •   Observed data
                                             	Best fit
                                             	Dl prediction (RSS = 4.52)
                                             	SCA prediction  (RSS = 4.7)
                  25
                       50           75           100
                         Scaled operation time (days)
125
150
Figure F-140 Comparison of Dl and SCA methods for predicting the SDS-
HAA9 integral breakthrough curve for Water 7
                                       -375-

-------
     1.25
     1.00 -
  O)
  o
     0.75 -
  §  0.50 -
  o
  O
     0.25 -
     0.00
 TOO

EBCT = 7.2 min.
co = 2.02 mg/L
                                          •   Observed data
                                         	Best fit
                                         	Dl prediction  (RSS = 0.033)
               50                100               150
                      Scaled operation time (days)
                                                                               200
Figure F-141  Dl method prediction of the TOC integral breakthrough curve
for Water 8
    0.015
  E 0.010 -
  o
  o>
  o
    0.005 -
    0.000
           UV254

           EBCT = 7.2 min.
           co= 0.033 1/cm
                                                     •   Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = 0.0017)
                                                   	SCA prediction (RSS = 0.0017)
                           50                100
                                  Scaled operation time (days)
                                                   150
200
Figure F-142 Comparison of Dl and SCA methods for predicting the UV254
integral breakthrough curve for Water 8
                                        -376-

-------
         200
       :r 150 -
       o
       o 100 -
       -i—'
       CD
       CD
       O
       c
       O  50 -
       SDS-TOX

       EBCT = 7.2min.
       co= 156|jg/LCI-
                       SDS-TOX not analyzed in
                           blended effluent
                                            •  Observed data
                                          	Best fit
                                          	Dl prediction (RSS = NA)
                                          	SCA prediction (RSS = NA)
                             50               100
                                  Scaled operation time (days)
                                                     150
200
 Figure F-143 Comparison of Dl and SCA methods for predicting the SDS-
 TOX integral breakthrough curve for Water 8
14

12

10 -
     o
     '-4— '
     CD
     o
     O   4
         2 -
             SDS-CF

             EBCT = 7.2min.
             co= 29.1 |jg/L
                                             •   Observed data
                                           	Best fit
                                           	Dl prediction (RSS = 0.58)
                                           	SCA prediction (RSS = 1.62)
                   50               100
                          Scaled operation time (days)
                                                             150
 200
Figure F-144 Comparison of Dl and SCA methods for predicting the SDS-CF
integral breakthrough curve for Water 8
                                       -377-

-------
   4 -



8
a 3.
      01 2 -
      o   n


      o

     O
        1  -
             SDS-BDCM



             EBCT = 7.2min.


             c0 = 2.6 |jg/L
                                             •   Observed data


                                           	Best fit


                                           	Dl prediction (RSS = 1.2)


                                           	SCA prediction (RSS = 0.81)
                          50               100


                                Scaled operation time (days)
                                                      150
200
Figure F-145 Comparison of Dl and SCA methods for predicting the SDS-

BDCM integral breakthrough curve for Water 8
       10
     c
     o


     is
     -t—>


     §
     c
     o
     O
        6 -
   4 -
        2 -
            SDS-DBCM



            EBCT = 7.2 min.



            co=  10.3|jg/L
                                               •  Observed data


                                             	Best fit


                                             	Dl prediction  (RSS = 1.15)


                                             	SCA prediction (RSS = 1.19)
                          50               100


                                 Scaled operation time (days)
                                                       150
 200
Figure F-146 Comparison of Dl and SCA methods for predicting the SDS-

DBCM integral breakthrough curve for Water 8
                                       -378-

-------
        2.0
        1.5 -
        1.0 H
     o>
     o
     c
     o
     O
        0.5 -
        0.0
             SDS-BF
             EBCT = 7.2min.
             co= BMRL
              Insufficient data measured above the MRL to
                  compare Dl and SCA predictions
                                       •   Observed data
                                     	Best fit
                                     	Dl prediction (RSS = NA)
                                     	SCA prediction (RSS = NA)
                           50
                               100
150
200
                                 Scaled operation time (days)
Figure F-147 Comparison of Dl and SCA methods for predicting the SDS-BF
integral breakthrough curve for Water 8
        30
        25 -
        20 -
     o
     '•§  15 H
     -i—'
     o>
     o
     o  10 -
     O
         5 -
SDS-TTHM
EBCT = 7.2min.
co = 42 ug/L
                                         •  Observed data
                                      	Best fit
                                      	Dl prediction  (RSS = 2.59)
                                      	SCA prediction (RSS = 3.87)
                           50
                                100
 150
 200
                                  Scaled operation time (days)
Figure F-148 Comparison of Dl and SCA methods for predicting the SDS-
TTHM integral breakthrough curve for Water 8
                                        -379-

-------
         2.0
         1.5 -
      o
      '•a  1.0 H
CD
O
O
O
   0.5 -
              SDS-MCAA
              EBCT = 7.2min.
              co= BMRL
0.0 -!•-
   0
                        Blended effluent concentrations were not detected
                              above the MRL for this parameter
                                                     •   Observed data
                                                   	Best fit
                                                   	Dl prediction (RSS = NA)
                                                   	SCA prediction (RSS = NA)
                            50
                                       100
150
200
                                  Scaled operation time (days)
Figure F-149 Comparison of Dl and SCA methods for predicting the SDS-
MCAA integral breakthrough curve for Water 8
        5 -
        4 -
     c
     o
        3 -
     §
     §  2
     O
        1 -
      SDS-DCAA
      EBCT = 7.2min.
      co= 10.7|jg/L
                                               •   Observed data
                                             	Best fit
                                             	Dl prediction (RSS = 0.68)
                                             	SCA prediction (RSS = 0.75)
                           50                100
                                 Scaled operation time (days)
                                                        150
                 200
Figure F-150 Comparison of Dl and SCA methods for predicting the SDS-
DCAA integral breakthrough curve for Water 8
                                        -380-

-------
        4 -
        3 -
      CD 9 -
      o ^ n
      o
      o
        1 -
SDS-TCAA
EBCT = 7.2min.
c0 = 12.7 |jg/L
                                             Observed data
                                             Best fit
                                             Dl prediction (RSS = 0.73)
                                             SCA prediction (RSS = 1)
                           50               100
                                  Scaled operation time (days)
                                                 150
                 200
Figure F-151 Comparison of Dl and SCA methods for predicting the SDS-
TCAA integral breakthrough curve for Water 8
        2.0
     c
     o
     c
     CD
     O
     c
     o
     O
        1.5 -
        1.0 -
        0.5 -
        0.0
  SDS-MBAA
  EBCT = 7.2 min.
  co= BMRL
          Blended effluent concentrations were not detected above the
                         MRL for this parameter
                                         •   Observed data
                                       	Best fit
                                       	Dl prediction  (RSS = NA)
                                       	SCA prediction (RSS = NA)
                            50
                                100
150
200
                                  Scaled operation time (days)
Figure F-152 Comparison of Dl and SCA methods for predicting the SDS-
MBAA integral breakthrough curve for Water 8
                                        -381-

-------
        2.0
        1.5 -
      o
      '•S3 1-0 H
      CD
      O
      c
      o
      O
        0.5 -
        o.o -k-
           0
                SDS-DBAA
                EBCT = 7.2min.
                co= BMRL
          Blended effluent concentrations were not detected
                above the MRL for this parameter
                                        •   Observed data
                                      	Best fit
                                      	Dl prediction (RSS = NA)
                                      	SCA prediction (RSS = NA)
              50               100
                     Scaled operation time (days)
150
200
Figure F-153 Comparison of Dl and SCA methods for predicting the SDS-
DBAA integral breakthrough curve for Water 8
        12
        10 -
     c
     o
     CO
         6 -
     c
     CD
     O
     §   4
     O
         2 -
         0 -!•-
          0
SDS-HAA5
EBCT = 7.2min.
co = 23 |jg/L

                                            Observed data
                                      	Best fit
                                      	Dl prediction (RSS = 1.35)
                                      	SCA prediction (RSS = 1.84)
              50                100
                    Scaled operation time (days)
150
200
Figure F-154 Comparison of Dl and SCA methods for predicting the SDS-
HAA5 integral breakthrough curve for Water 8
                                        -382-

-------
        3.5
        3.0 -
     o>
       2.0 H
     O
     0  1.0 H
        0.5 -
        0.0
  SDS-BCAA
  EBCT = 7.2min.
  CD = 3 |jg/L
                                         •  Observed data
                                       	Best fit
                                       	Dl prediction (RSS = 0.59)
                                       	SCA prediction (RSS = 0.31)
                           50               100
                                  Scaled operation time (days)
                                                 150
200
Figure F-155 Comparison of Dl and SCA methods for predicting the SDS-
BCAA integral breakthrough curve for Water 8
       14
       12 -
       10 -
    o
    '-4—'
    CD
    CD
    O
    C
    O
    O
       6 -
       4 -
       2 -
SDS-HAA6
EBCT = 7.2min.

co = 27 |jg/L
                                         •  Observed data
                                       	Best fit
                                       	Dl prediction (RSS = 1.83)
                                       	SCA prediction  (RSS = 1.86)
                          50                100
                                 Scaled operation time (days)
                                                 150
200
Figure F-156 Comparison of Dl and SCA methods for predicting the SDS-
HAA6 integral breakthrough curve for Water 8
                                       -383-

-------
      2.5
      2.0 -
      1.5 -
   o
   "-i—'
   s
   "c
   o>
   o
   o
   O
      0.5 -
             SDS-DCBAA
             EBCT = 7.2min.
             C° ~ 3 |jg/L
      0.0 -\»-
         0
                            •   Observed data
                          	Best fit
                          	Dl prediction (RSS = 0.46)
                          	SCA prediction (RSS = 0.29)
 50                100
        Scaled operation time (days)
150
200
Figure F-157 Comparison of Dl and SCA methods for predicting the SDS-
DCBAA integral breakthrough curve for Water 8
     2.0
     1.5 -
     1.0 H
  o>
  o
  c
  o
  O
     0.5 -
            SDS-CDBAA
            EBCT = 7.2 min.
            c°=  BMRL
                           Blended effluent concentrations were not detected above
                                      the MRL for this parameter
                                                       •   Observed data
                                                     	Best fit
                                                     	Dl prediction  (RSS = NA)
                                                     	SCA prediction (RSS = NA)
     0.0 -{•-
        0
50                100
       Scaled operation time (days)
150
200
Figure F-158 Comparison of Dl and SCA methods for predicting the SDS-
CDBAA integral breakthrough curve for Water 8
                                        -384-

-------
       2.0
       1.5 -
       1-0 H
    O>
    o
    c
    o
    O
       0.5 -
       0.0
              SDS-TBAA
              EBCT = 7.2min.
              CD = BMRL
                        Blended effluent concentrations were not detected
                              above the MRL for this parameter
                                                 •  Observed data
                                              	Best fit
                                              	Dl prediction  (RSS = NA)
                                              	SCA prediction (RSS = NA)
                           50               100
                                  Scaled operation time (days)
                                                          150
                  200
  Figure F-159  Comparison of Dl and SCA methods for predicting the SDS-
  TBAA integral breakthrough curve for Water 8
   16

   14 -

   12 -
Q"
1 1°-
o
'•^   ft -
CD   O 1
     CD
     £  e ^
     o
     O
        4 -
        2 -

        0
               SDS-HAA9
               EBCT = 7.2min.
               co = 30 |jg/L
                                                 •   Observed data
                                              	Best fit
                                              	Dl prediction  (RSS = 2.2)
                                              	SCA prediction (RSS = 1.85)
          0
                       50                100
                             Scaled operation time (days)
150
200
Figure F-160 Comparison of Dl and SCA methods for predicting the SDS-
HAA9 integral breakthrough curve for Water 8
                                        -385-

-------
This page intentionally left blank.
             -386-

-------
Appendix G: Logistic Function Model Best-Fit Parameters
                              -387-

-------
Analyte
TOC
UV-254
SDS-TOX
SDS-TTHM
SDS-HAA5
SDS-HAA6
SDS-HAA9
SDS-CF
SDS-BDCM
SDS-DBCM
SDS-BF
SDS-MCAA
SDS-DCAA
SDS-TCAA
SDS-MBAA
SDS-DBAA
SDS-BCAA
SDS-CDBAA
SDS-DCBAA
SDS-TBAA
Type of curve
fit
Step
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag-peak
No fit
Step-lag
No fit
No fit
Step-lag
Step-lag
No fit
No fit
No fit
Ao
0.00
-0.03
-55.82
-33.46
-1.25
-10.79
-11.49
-2.12
-12.01
-7.70
-0.40
NA
-0.36
NA
NA
-0.05
-2.94
NA
NA
NA
A
3.28
0.09
186.75
100.37
15.59
30.90
33.13
20.37
35.33
31.82
13.59
NA
8.52
NA
NA
5.35
8.66
NA
NA
NA
B
11.11
3.21
4.02
3.45
180.13
12.40
9.47
19.81
5.74
5.91
200.86
NA
1 .39E+04
NA
NA
124.71
6.86
NA
NA
NA
D
0.104
0.058
0.070
0.083
0.273
0.157
0.137
0.068
0.117
0.066
0.388
NA
0.470
NA
NA
0.302
0.103
NA
NA
NA
S
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.148
NA
NA
NA
NA
NA
NA
NA
NA
NA
*P
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
28
NA
NA
NA
NA
NA
NA
NA
NA
NA
n*
13
13
12
11
11
11
11
8
11
11
11
0
9
2
0
11
9
0
3
0
NA: not applicable
'Number of observations above the MRL
Table G-1  Summary of single contactor best-fit logistic function model parameters for Water 1
Analyte
TOC
UV-254
SDS-TOX
SDS-TTHM
SDS-HAA5
SDS-HAA6
SDS-HAA9
SDS-CF
SDS-BDCM
SDS-DBCM
SDS-BF
SDS-MCAA
SDS-DCAA
SDS-TCAA
SDS-MBAA
SDS-DBAA
SDS-BCAA
SDS-CDBAA
SDS-DCBAA
SDS-TBAA
Type of curve
fit
Step
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag-peak
No fit
Step-lag
Step-lag
No fit
Step-lag
Step-lag
No fit
Step-lag
No fit
Ao
0.00
-0.02
-75.03
-38.87
-8.74
-12.84
-14.05
-9.99
-7.34
-12.70
0.06
NA
-4.27
0.00
NA
-0.56
-4.10
NA
-1.21
NA
A
1.78
0.05
218.15
113.39
24.47
35.99
40.85
35.27
31.41
37.33
11.27
NA
12.29
2.10
NA
6.65
11.54
NA
3.42
NA
B
7.48
2.03
2.32
2.50
2.89
2.89
3.06
3.20
8.73
3.51
11.03
NA
2.84
3.51 E+06
NA
11.45
2.86
NA
15.85
NA
D
0.037
0.022
0.025
0.037
0.032
0.032
0.031
0.016
0.071
0.036
0.127
NA
0.021
0.245
NA
0.073
0.030
NA
0.067
NA
S
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.036
NA
NA
NA
NA
NA
NA
NA
NA
NA
*P
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
39
NA
NA
NA
NA
NA
NA
NA
NA
NA
n*
13
13
12
12
11
11
11
10
11
11
12
0
9
6
0
11
10
2
8
0
NA: not applicable
'Number of observations above the MRL
Table G-2  Summary of single contactor best-fit logistic function model parameters for Water 2
                                             -388-

-------
Analyte
TOC
UV-254
SDS-TOX
SDS-TTHM
SDS-HAA5
SDS-HAA6
SDS-HAA9
SDS-CF
SDS-BDCM
SDS-DBCM
SDS-BF
SDS-MCAA
SDS-DCAA
SDS-TCAA
SDS-MBAA
SDS-DBAA
SDS-BCAA
SDS-CDBAA
SDS-DCBAA
SDS-TBAA
Type of curve
fit
Step
Step-lag
Step-lag
Step-lag
Step-lag-peak
Step-lag
Step-lag-peak
Step-lag
Step-lag
Step-lag
Step-lag-peak
No fit
Step-lag
No fit
No fit
Step-lag
Step-lag
No fit
Step-lag
No fit
Ao
0.00
-0.01
-24.55
-45.77
-0.33
2.79
-1.91
-0.91
-1.52
-7.80
3.92
NA
-2.92
NA
NA
0.37
0.44
NA
-0.96
NA
A
1.50
0.04
161.23
144.99
30.45
22.97
48.87
18.88
35.95
36.72
27.13
NA
9.26
NA
NA
10.77
6.91
NA
2.87
NA
B
9.36
3.15
4.95
1.92
24.34
228.87
17.53
36.59
12.00
5.97
32.31
NA
12.56
NA
NA
25.97
85.06
NA
1.75
NA
D
0.027
0.019
0.019
0.020
0.026
0.057
0.024
0.022
0.041
0.018
0.080
NA
0.024
NA
NA
0.048
0.047
NA
0.016
NA
S
NA
NA
NA
NA
0.000
NA
0.000
NA
NA
NA
-0.050
NA
NA
NA
NA
NA
NA
NA
NA
NA
*P
NA
NA
NA
NA
140
NA
140
NA
NA
NA
78
NA
NA
NA
NA
NA
NA
NA
NA
NA
n*
12
12
12
12
11
11
11
9
12
11
12
0
6
5
0
11
11
2
9
0
NA: not applicable
'Number of observations above the MRL
Table G-3  Summary of single contactor best-fit logistic function model parameters for Water 3
Analyte
TOC
UV-254
SDS-TOX
SDS-TTHM
SDS-HAA5
SDS-HAA6
SDS-HAA9
SDS-CF
SDS-BDCM
SDS-DBCM
SDS-BF
SDS-MCAA
SDS-DCAA
SDS-TCAA
SDS-MBAA
SDS-DBAA
SDS-BCAA
SDS-CDBAA
SDS-DCBAA
SDS-TBAA
Type of curve
fit
Step
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag-peak
Step-lag
No fit
No fit
Step-lag
Step-lag
No fit
No fit
Step-lag
No fit
Step-lag
No fit
Ao
0.00
-0.01
-75.14
-19.39
-10.44
-12.99
-14.39
-10.41
-1.62
-5.61
NA
NA
-5.31
-4.35
NA
NA
-2.05
NA
-1.52
NA
A
2.10
0.05
246.00
61.91
51.98
58.13
66.66
41.38
4.69
16.64
NA
NA
20.22
31.76
NA
NA
5.78
NA
8.53
NA
B
27.87
6.52
7.02
5.76
9.16
7.66
7.19
7.49
3.64E+03
6.38
NA
NA
6.76
15.48
NA
NA
8.18
NA
4.59
NA
D
0.044
0.025
0.028
0.026
0.021
0.021
0.020
0.023
0.162
0.032
NA
NA
0.022
0.022
NA
NA
0.040
NA
0.017
NA
S
NA
NA
NA
NA
NA
NA
NA
NA
-0.006
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
*P
NA
NA
NA
NA
NA
NA
NA
NA
91
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
n*
13
13
12
11
10
10
11
11
10
11
0
0
10
10
0
4
10
0
11
0
NA: not applicable
'Number of observations above the MRL
Table G-4  Summary of single contactor best-fit logistic function model parameters for Water 4
                                             -389-

-------
Analyte
TOC
UV-254
SDS-TOX
SDS-TTHM
SDS-HAA5
SDS-HAA6
SDS-HAA9
SDS-CF
SDS-BDCM
SDS-DBCM
SDS-BF
SDS-MCAA
SDS-DCAA
SDS-TCAA
SDS-MBAA
SDS-DBAA
SDS-BCAA
SDS-CDBAA
SDS-DCBAA
SDS-TBAA
Type of curve
fit
Step
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
No fit
Step-lag
Step-lag
No fit
Step-lag
Step-lag
Step-lag-peak
Step-lag
No fit
Ao
0.00
-0.01
-53.67
-22.75
-7.50
-10.90
-17.05
-2.28
-7.19
-6.98
-1.20
NA
-2.40
-3.00
NA
-1.60
-3.40
-2.60
-4.25
NA
A
2.33
0.05
196.80
68.25
22.06
32.01
50.40
16.34
20.65
27.30
3.34
NA
8.70
9.30
NA
4.65
9.96
7.06
12.75
NA
B
12.66
3.83
4.28
3.01
4.85
4.84
6.20
9.86
3.14
4.02
116.01
NA
4.51
7.63
NA
62.79
4.82
148.78
4.99
NA
D
0.023
0.013
0.014
0.015
0.021
0.021
0.023
0.013
0.021
0.013
0.090
NA
0.014
0.019
NA
0.064
0.022
0.060
0.019
NA
S
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.004
NA
NA
*P
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
158
NA
NA
n*
13
13
13
12
11
11
11
10
12
12
11
3
10
9
0
11
11
8
11
0
NA: not applicable
'Number of observations above the MRL
Table G-5  Summary of single contactor best-fit logistic function model parameters for Water 5
Analyte
TOC
UV-254
SDS-TOX
SDS-TTHM
SDS-HAA5
SDS-HAA6
SDS-HAA9
SDS-CF
SDS-BDCM
SDS-DBCM
SDS-BF
SDS-MCAA
SDS-DCAA
SDS-TCAA
SDS-MBAA
SDS-DBAA
SDS-BCAA
SDS-CDBAA
SDS-DCBAA
SDS-TBAA
Type of curve
fit
Step
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step
Step-lag
Step-lag
Step-lag-peak
No fit
Step-lag
No fit
No fit
Step-lag
Step-lag
Step-lag
Step-lag
No fit
Ao
0.00
-0.01
-77.38
-44.21
-9.22
-14.16
-18.87
0.00
-2.82
-13.11
-0.13
NA
-2.97
NA
NA
0.18
-4.62
0.04
-2.59
NA
A
1.98
0.05
283.38
131.09
30.95
44.81
57.94
24.13
31.63
44.11
15.07
NA
12.66
NA
NA
7.36
13.86
2.64
8.53
NA
B
69.54
9.28
9.39
12.29
10.99
12.39
14.38
372.30
605.70
14.75
8.23E+03
NA
17.57
NA
NA
1 .73E+05
18.59
7.20E+12
17.66
NA
D
0.033
0.017
0.018
0.027
0.020
0.022
0.023
0.031
0.062
0.022
0.099
NA
0.018
NA
NA
0.110
0.026
0.263
0.021
NA
S
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.057
NA
NA
NA
NA
NA
NA
NA
NA
NA
*P
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
131
NA
NA
NA
NA
NA
NA
NA
NA
NA
n*
13
13
11
11
10
10
10
8
11
10
11
0
7
4
0
10
9
7
7
0
NA: not applicable
'Number of observations above the MRL
Table G-6  Summary of single contactor best-fit logistic function model parameters for Water 6
                                             -390-

-------
Analyte
TOO
UV-254
SDS-TOX
SDS-TTHM
SDS-HAA5
SDS-HAA6
SDS-HAA9
SDS-CF
SDS-BDCM
SDS-DBCM
SDS-BF
SDS-MCAA
SDS-DCAA
SDS-TCAA
SDS-MBAA
SDS-DBAA
SDS-BCAA
SDS-CDBAA
SDS-DCBAA
SDS-TBAA
Type of curve
fit
Step
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
No fit
Step-lag
Step-lag
No fit
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag-peak
Ao
0.00
-0.03
-134.27
-81.69
-16.06
-23.26
-34.18
-2.35
-34.72
-15.55
1.39
NA
-4.10
-0.03
NA
-0.06
-7.20
-1.78
-2.77
-3.67
A
4.01
0.09
420.30
239.78
47.02
67.10
96.81
25.39
103.28
64.36
35.62
NA
13.83
6.60
NA
16.06
20.05
7.56
16.70
10.33
B
15.49
4.57
4.63
4.18
5.56
6.08
8.04
33.37
5.34
7.25
459.64
NA
3.28
6.68E+05
NA
136.54
7.76
262.52
9.30
471 .50
D
0.049
0.032
0.036
0.042
0.051
0.052
0.060
0.034
0.045
0.031
0.187
NA
0.026
0.193
NA
0.127
0.058
0.138
0.027
0.174
S
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
-0.034
*P
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
65
n*
13
13
12
12
11
11
11
7
11
10
12
0
10
6
0
11
10
9
9
9
NA: not applicable
'Number of observations above the MRL
Table G-7  Summary of single contactor best-fit logistic function model parameters for Water 7
Analyte
TOC
UV-254
SDS-TOX
SDS-TTHM
SDS-HAA5
SDS-HAA6
SDS-HAA9
SDS-CF
SDS-BDCM
SDS-DBCM
SDS-BF
SDS-MCAA
SDS-DCAA
SDS-TCAA
SDS-MBAA
SDS-DBAA
SDS-BCAA
SDS-CDBAA
SDS-DCBAA
SDS-TBAA
Type of curve
fit
Step
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag
Step-lag-peak
Step-lag
No fit
No fit
Step-lag
Step-lag
No fit
No fit
Step-lag
No fit
Step-lag
No fit
Ao
0.00
-0.01
-56.70
-18.32
-6.73
-8.03
-9.14
-11.78
-1.91
-5.12
NA
NA
-3.55
-3.19
NA
NA
-1.30
NA
-0.97
NA
A
1.53
0.03
162.37
52.36
20.20
22.41
25.34
35.35
4.97
14.78
NA
NA
10.03
9.56
NA
NA
3.68
NA
3.05
NA
B
27.83
5.40
6.31
5.00
5.50
9.66
14.18
4.06
1 .26E+03
9.93
NA
NA
8.52
5.04
NA
NA
2.49E+03
NA
1 .44E+04
NA
D
0.064
0.037
0.042
0.036
0.034
0.052
0.062
0.024
0.196
0.055
NA
NA
0.048
0.030
NA
NA
0.199
NA
0.236
NA
S
NA
NA
NA
NA
NA
NA
NA
NA
-0.004
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
*P
NA
NA
NA
NA
NA
NA
NA
NA
113
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
n*
13
13
13
11
10
10
10
11
11
11
0
0
10
9
0
0
10
0
9
0
NA: not applicable
'Number of observations above the MRL
Table G-8  Summary of single contactor best-fit logistic function model parameters for Water 8
                                             -391-

-------
This page intentionally left blank.
             -392-

-------
Appendix H: Impact of Extrapolation on SCA Prediction of the
Integral Breakthrough Curve
                              -393-

-------
   3.0
   2.5 -
   2.0 -
o
•^= -I c
CD  '•"
CD
o
o 1.0 H
O
   0.5 -
   0.0
 D   Single contactor effluent
 •   Extrapolation experimental data points
     Logistic function best fit - all data  (RA2 = 0.984)
 -  - Extrapolated logistic function best fit (RA2 = 0.979)
 O   Blended effluent
	Dl prediction
	Dl prediction - extrapolated
                                                                               TOO
                                                                         EBCT = 20 min.

                                                                         c0 = 3.08 mg/L
                  50
                 100         150         200        250

                       Scaled operation time (days)
300
350
 Figure H-1 Impact of extrapolation on Dl prediction of the TOC integral
 breakthrough curve for Water 5
   0.035
   0.030 -
   0.025 -
.o

^ 0.020 -
o>
o
c
CD
£ 0.015 -
o
   0.010 H
   0.005 -
   0.000
    D   Single contactor effluent
    •   Extrapolation experimental data points
   	Logistic function best fit - all data (RA2 = 0.992)
    -  -  Extrapolated logistic function best fit (RA2 = 0.984)
    O   Blended effluent
   	SCA prediction
   	SCA prediction - extrapolated
                                                                             UV254
                                                                         EBCT = 20 min.

                                                                         c0 = 0.051 1/cm
                    50         100         150         200        250

                                    Scaled operation time (days)
                                                                 300
            350
 Figure H-2  Impact of extrapolation on SCA prediction of the UV254 integral
 breakthrough curve for Water 5
                                          -394-

-------
 O
    150
    125 -
    100 -
 .2   75 -
 "oo
 O>
 £   50 H
 o
 O
     25 -
   D   Single contactor effluent
   •   Extrapolation experimental data points
       Logistic function best fit - all data (RA2 = 0.995)
- - -  Extrapolated logistic function best fit (RA2 = 0.99)
   O   Blended effluent
	SCA prediction
	SCA prediction - extrapolated
                                                                            SDS-TOX
                                                                       EBCT = 20 min.

                                                                       C0 = 205 ug/L Cl-
                   50         100        150        200         250

                                   Scaled operation time (days)
                                                              300
            350
 Figure H-3 Impact of extrapolation on SCA prediction of the SDS-TOX
 integral breakthrough curve for Water 5
    12
    10 -
 o
 '«   6 H
 o>
 o
 O
     2 -
     0 \
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data (RA2 = 0.987)
 -  - Extrapolated logistic function best fit (RA2 = 0.982)
 O   Blended effluent
	SCA prediction
	SCA prediction - extrapolated
                                                                             SDS-CF
                                                                       EBCT = 20 min.

                                                                       C0 = 23.7 ug/L
                  50
                100         150         200        250

                     Scaled operation time (days)
300
350
Figure H-4  Impact of extrapolation on SCA prediction of the SDS-CF integral
breakthrough curve for Water 5
                                         -395-

-------
   16


   14 -


   12 -

Q"

| 10-

g
'•£  8 H
o
O
            D   Single contactor effluent
            •   Extrapolation experimental data points
           	Logistic function best fit - all data  (RA2 = 0.97)
            -  - Extrapolated logistic function best fit (RA2 = 0.967)
            O   Blended effluent
           	SCA prediction
           	SCA prediction - extrapolated
                                                                      SDS-BDCM
                                                                     ---X)
                                                                       EBCT = 20 min.

                                                                       C0 = 10.8 ug/L
                 50
                            100         150        200         250

                                  Scaled operation time (days)
300
350
Figure H-5  Impact of extrapolation on SCA prediction of the SDS-BDCM
integral breakthrough curve for Water 5
   20
o>
o
c
o
o
   15 -
   10 -
    5 -
             D   Single contactor effluent
             •   Extrapolation experimental data points
            	Logistic function best fit - all data (RA2 = 0.988)
             -  - Extrapolated logistic function best fit (RA2 = 0.986)
             O   Blended effluent
            	SCA prediction
            	SCA prediction - extrapolated
                                                                      SDS-DBCM
                                                                      EBCT = 20 min.

                                                                      C0 =  22.7 ug/L
                 50
                            100         150        200         250

                                  Scaled operation time (days)
300
350
Figure H-6 Impact of extrapolation on SCA prediction of the SDS-DBCM
integral breakthrough curve for Water 5
                                         -396-

-------
   3.0



   2.5 -



§2.0-
3.
c
g
'•S3 1-5 H
CD
o
§ 1.0 -
O
   0.5 -
   0.0 4
        EBCT = 20 min.

        C0= 1.2|jg/L
                                                                             SDS-BF
                  50
                                          D  Single contactor effluent
                                          •  Extrapolation experimental data points
                                        	Logistic function best fit - all data (RA2 = 0.969)
                                        - -  - Extrapolated logistic function best fit (RA2 = 0.969)
                                          O  Blended effluent
                                        	SCA prediction
                                        	SCA prediction - extrapolated
                             100        150        200         250

                                  Scaled operation time (days)
300
350
Figure H-7  Impact of extrapolation on SCA prediction of the SDS-BF integral
breakthrough curve for Water 5
   50
   40 -
   30 -
o
'-4—'
CD
   20 -
o
O
   10 -
    o 4
               D   Single contactor effluent
               •   Extrapolation experimental data points
              	Logistic function best fit - all data  (RA2 = 0.979)
               -  -  Extrapolated logistic function best fit (RA2 = 0.975)
               O   Blended effluent
              	SCA prediction
              	SCA prediction - extrapolated
                                                                          SDS-TTHM
                                                                       EBCT = 20 min.

                                                                       C0 = 58 ug/L
                 50
                            100         150        200         250

                                  Scaled operation time (days)
300
350
Figure H-8  Impact of extrapolation on SCA prediction of the SDS-TTHM
integral breakthrough curve for Water 5
                                        -397-

-------
    4 -
    3 -
 g
 '-4—'
 CD
CD  2 -
o  ^ ~
o
o
    1 -
    0 -
               D   Single contactor effluent
               •   Extrapolation experimental data points
              	Logistic function best fit - all data  (RA2 = NA)
               •  •  Extrapolated logistic function best fit
               O   Blended effluent
              	SCA prediction
              	SCA prediction - extrapolated
                              Insufficient data measured above the MRL
                            to perform curve fit and extrapolation analysis
               -Q—Ol—i-O—Oh-

                 50         100
                                       150         200        250

                                 Scaled operation time (days)
                                                                          SDS-MCAA
EBCT = 20 min.

c0 =  2 ug/L


    300
350
Figure H-9  Impact of extrapolation on SCA prediction of the SDS-MCAA
integral breakthrough curve for Water 5
    7
 O
 '-4—'
 CD
   6 -


   5 -


   4 -


   3 -
 o
 O  2 4
    1 -
    o 4
              D  Single contactor effluent
              •  Extrapolation experimental data points
             	Logistic function best fit - all data (RA2 = 0.978)
              - - Extrapolated logistic function best fit  (RA2 = 0.963)
              O  Blended effluent
             	SCA prediction
             	SCA prediction - extrapolated
                                                                          SDS-DCAA
                 50
                                                                       EBCT = 20 min.

                                                                       C0 = 10.3 |jg/L
                           100         150         200        250

                                 Scaled operation time (days)
    300
350
 Figure H-10 Impact of extrapolation on SCA prediction of the SDS-DCAA
 integral breakthrough curve for Water 5
                                         -398-

-------
   6 -
   5 -
.g
"co
   4 -
   3 -
CD
O

O
O  2H
   1 -
   o 4
 Single contactor effluent
 Extrapolation experimental data points
•Logistic function best fit - all data (RA2 = 0.989)
 Extrapolated logistic function best fit (RA2 = 0.987)
 Blended effluent
-SCA prediction
 SCA prediction - extrapolated
                                                                       SDS-TCAA
                                                       EBCT = 20 min.

                                                       C0 = 12.7 ug/L
                50
             100         150         200         250

                  Scaled operation time (days)
                                               300
               350
Figure H-11  Impact of extrapolation on SCA prediction of the SDS-TCAA
integral breakthrough curve for Water 5
   4 -
   3 -
o
'-4—'
CD
CD
O
c
o
O
   1 -
   0 -O-

     0
    Single contactor effluent
    Extrapolation experimental data points
   •Logistic function best fit - all data (RA2 = NA)
    Extrapolated logistic function best fit
    Blended effluent
   -SCA prediction
   • SCA prediction - extrapolated
                                                                       SDS-MBAA
               Effluent concentrations were not detected
                  above the MRL for this parameter
 50
100         150        200         250

      Scaled operation time (days)
EBCT = 20 min.

c0 = 1 ug/L


    300
350
 Figure H-12 Impact of extrapolation on SCA prediction of the SDS-MBAA
 integral  breakthrough curve for Water 5
                                         -399-

-------
   3.5
   3.0 -
   2.5 -
.g
"co
   2.0 -
   1.5 -
o>
o
c
o
0  1.0 H
   0.5 -
   o.o 4
EBCT = 20 min.

c0 = 2 ug/L
50
                                                                       SDS-DBAA
                                 D   Single contactor effluent
                                 •   Extrapolation experimental data points
                              	Logistic function best fit - all data (RA2 = 0.992)
                              - - -  Extrapolated logistic function best fit (RA2 = 0.992)
                                 O   Blended effluent
                              	SCA prediction
                              	SCA prediction - extrapolated
                             100        150         200        250

                                  Scaled operation time (days)
300
                                                                             350
 Figure H-13 Impact of extrapolation on SCA prediction of the SDS-DBAA
 integral breakthrough curve for Water 5
   20
                                                                       EBCT = 20 mm.

                                                                       C0 = 28 ug/L
   15 -
o
'•S3  10
o>
o
c
o
O
    5 -
     D   Single contactor effluent
     •   Extrapolation experimental data points
    	Logistic function best fit - all data  (RA2 = 0.965)
     -  -  Extrapolated logistic function best fit (RA2 = 0.959)
     O   Blended effluent
    	SCA prediction
    	SCA prediction - extrapolated
                                                                        SDS-HAA5
                            100
                               150        200

                         Scaled operation time (days)
                                             250
300
350
 Figure H-14 Impact of extrapolation on SCA prediction of the SDS-HAA5
 integral  breakthrough curve for Water 5
                                         -400-

-------
   6 -
 g
'•S3 4H
 CD
 o
 c
 o
O
   2 -
   0 -<
   D  Single contactor effluent
   •  Extrapolation experimental data points
      Logistic function best fit - all data (RA2 = 0.987)
- - - Extrapolated logistic function best fit (RA2 = 0.982)
   O  Blended effluent
	SCA prediction
	SCA prediction - extrapolated        _^^^". D
                                                                       SDS-BCAA
                                                                      EBCT = 20 min.

                                                                      C0 = 7.3 ug/L
                50
                  100         150        200         250

                        Scaled operation time (days)
300
350
Figure H-15 Impact of extrapolation on SCA prediction of the SDS-BCAA
integral breakthrough curve for Water 5
   25
   20 -
 o
'-4—'
 CD
 o
o
   15 -
   10 -
    5 -
    0 4
     D   Single contactor effluent
     •   Extrapolation experimental data points
   	Logistic function best fit - all data (RA2 = 0.976)
    - -  Extrapolated logistic function best fit  (RA2 = 0.971)
     O   Blended effluent
   	SCA prediction
   	SCA prediction - extrapolated
                                                                       SDS-HAA6
                                                                      EBCT = 20 min.

                                                                      c0 =  34 ug/L
                 50
                   100        150         200        250

                        Scaled operation time (days)
300
350
Figure H-16  Impact of extrapolation on SCA prediction of the SDS-HAA6
integral breakthrough curve for Water 5
                                         -401-

-------
   10
O)
.0
"co
o
O
    6-\
    2 -
             D   Single contactor effluent
             •   Extrapolation experimental data points
            	Logistic function best fit - all data (RA2 = 0.987)
             -  -  Extrapolated logistic function best fit (RA2 = 0.987)
             O   Blended effluent
            	SCA prediction
            	SCA prediction - extrapolated
                                                                     SDS-DCBAA
                                                                      EBCT = 20 min.

                                                                      c0=  10.7 ug/L
                 50
                  100        150        200         250

                       Scaled operation time (days)
300
350
Figure H-17 Impact of extrapolation on SCA prediction of the SDS-DCBAA
integral breakthrough  curve for Water 5
   5 -
o
•^
CD
CD
O
o  -2\
O
   1 -
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data (R^2 = 0.95)
 -  - Extrapolated logistic function best fit (RA2 = 0.947)
 O   Blended effluent
	SCA prediction
	SCA prediction - extrapolated
                                                                     SDS-CDBAA
                                                                      EBCT = 20 min.

                                                                      c0 = 3.7 ug/L
                50
                 100         150         200         250

                       Scaled operation time (days)
300
350
 Figure H-18 Impact of extrapolation on SCA prediction of the SDS-CDBAA
 integral breakthrough curve for Water 5
                                        -402-

-------
   4 -
   3 -
.g
"co
 o>
 o
 c
 o
O
   1 -
   o -b-
     o
                                                                       SDS-TBAA
               Single contactor effluent
               Extrapolation experimental data points
              •Logistic function best fit - all data  (RA2 = NA)
               Extrapolated logistic function best fit
               Blended effluent
              -SCA prediction
               SCA prediction - extrapolated
                           Effluent concentrations were not detected
                              above the MRL for this parameter
             50
100         150        200         250

      Scaled operation time (days)
EBCT = 20 min.

C0 =  BMRL


    300
350
Figure H-19 Impact of extrapolation on SCA prediction of the SDS-TBAA
integral breakthrough curve for Water 5
   40
 o>
 £
 o
O
35 -


30 -


25 -


20 -


15 -


10 -


 5 -


 0 -
             D  Single contactor effluent
             •  Extrapolation experimental data points
            	Logistic function best fit - all data (RA2 = 0.99)
             - - Extrapolated logistic function best fit (RA2 = 0.988)
             O  Blended effluent
            	SCA prediction
            	SCA prediction - extrapolated        __^j»*» " n
                                                                    SDS-HAA9
                                                                      EBCT = 20 min.

                                                                      C0 =  48 ug/L
                 50
                         100         150        200         250

                               Scaled operation time (days)
                                               300
                350
Figure H-20  Impact of extrapolation on SCA prediction of the SDS-HAA9
integral breakthrough curve for Water 5
                                         -403-

-------
   2.0
   1.5 -
'•a 1-0 H
CD
o
c
o
O
   0.5 -
   0.0
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data  (RA2 = 0.974)
 -  - Extrapolated logistic function best fit (RA2 = 0.961)
 O   Blended effluent
	SCA prediction
	SCA prediction - extrapolated
                                                                               TOO
                                                                         EBCT = 7.2 min.

                                                                         c0 = 2.02 mg/L
                           50                  100

                                   Scaled operation time (days)
                                                        150
                    200
 Figure H-21  Impact of extrapolation on Dl prediction of the TOC integral
 breakthrough curve for Water 8
   0.025
   0.020 -
o
^ 0.015

CD
O

CD

o 0.010
   0.005 -
   0.000 -K>

         0
    D   Single contactor effluent
    •   Extrapolation experimental data points
   	Logistic function best fit - all data (RA2 = 0.994)
    -  - Extrapolated logistic function best fit (RA2 = 0.98
    O   Blended effluent
   	SCA prediction
   	SCA prediction - extrapolated
                                                                             UV254
                                                                         EBCT = 7.2 min.

                                                                         C0 = 0.033 1/cm
                 50                  100

                         Scaled operation time (days)
150
200
 Figure H-22  Impact of extrapolation on SCA prediction of the UV254 integral
 breakthrough curve for Water 8
                                          -404-

-------
 O
 o
 "co
 o>
 o
 c
 o
 O
    125
    100 -
     75 -
     50 -
     25 -
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data (RA2 = 0.99)
 -  - Extrapolated logistic function best fit (RA2 = 0.98)_
 O   Blended effluent
	SCA prediction
	SCA prediction - extrapolated
                                                                    SDS-TOX
                                                                       EBCT = 7.2 min.

                                                                       c0= 156 ug/LCI-
                           50                  100

                                   Scaled operation time (days)
                                                      150
200
 Figure H-23  Impact of extrapolation on SCA prediction of the SDS-TOX
 integral breakthrough curve for Water 8
    25
    20 -
    15 -
 o
 '-4—'
 CD
 o
 O
    10 -
   D   Single contactor effluent
   •   Extrapolation experimental data points
  	Logistic function best fit - all data (RA2 = 0.985)
  - -  Extrapolated logistic function best fit (RA2 = 0.917)
   O   Blended effluent
  	SCA prediction
  	SCA prediction - extrapolated
                                                                      SDS-CF
                                                                       EBCT = 7.2 min.

                                                                       C0 = 29.1 ug/L
                           50                 100

                                   Scaled operation time (days)
                                                      150
200
Figure H-24  Impact of extrapolation on SCA prediction of the SDS-CF
integral breakthrough  curve for Water 8
                                         -405-

-------
   4 -
   3 -
.g
"co
o>
o
c
o
O
   1 -
          D   Single contactor effluent
          •   Extrapolation experimental data points
         	Logistic function best fit - all data (RA2 = 0.946)
          -  - Extrapolated logistic function best fit (RA2 = 0.942)
          O   Blended effluent
         	SCA prediction                           u
          • • • • SCA prediction - extrapolated
                                                   O
                                                                   SDS-BDCM
                         50                  100

                                  Scaled operation time (days)
                                                                  150
200
Figure H-25  Impact of extrapolation on SCA prediction of the SDS-BDCM
integral breakthrough curve for Water 8
   12
   10 -
^  8-
o
73
o>
o
c
o
O
    6H
    4 -
    2 -
           D   Single contactor effluent
           •   Extrapolation experimental data points
         	Logistic function best fit - all data (RA2 = 0.98)
         - -  - Extrapolated logistic function best fit (RA2 = 0.978}
           O   Blended effluent                         D
         	SCA prediction
         	SCA prediction - extrapolated
                                                                      SDS-DBCM
                                                                       EBCT = 7.2 min.

                                                                       C0 =  10.3 ug/L
                          50                  100

                                  Scaled operation time (days)
                                                                  150
200
Figure H-26  Impact of extrapolation on SCA prediction of the SDS-DBCM
integral breakthrough curve for Water 8
                                         -406-

-------
   2.0
   1.5 -
g
'•SS  1-0 H
CD
O

O
O
   0.5 -
              D   Single contactor effluent
              •   Extrapolation experimental data points
           	Logistic function best fit - all data (RA2 = NA)
           - -  -  Extrapolated logistic function best fit
              O   Blended effluent
           	SCA prediction
           	SCA prediction - extrapolated
   0.0 -K>

      0
                                                                 O
                              Insufficient data measured above the MRL
                            to perform curve fit and extrapolation analysis
                        JQHXD——CBi   CD-

                          50
                                                   -CD-
                                                                              SDS-BF
                                                                       EBCT = 7.2min.

                                                                       C0 =  BMRL
100
150
200
                                  Scaled operation time (days)
Figure H-27  Impact of extrapolation on SCA prediction of the SDS-BF
integral breakthrough curve for Water 8
   40
   30 -
o
'•S3 20 H
CD
O
c
o
O
   10 -
           D  Single contactor effluent
           •  Extrapolation experimental data points
         	Logistic function best fit - all data (RA2 = 0.98)
         - - - Extrapolated logistic function best fit (RA2 = 0.966)
           O  Blended effluent
         	SCA prediction
         	SCA prediction - extrapolated
                                                                           SDS-TTHM
                                                                       EBCT = 7.2min.

                                                                       c0 = 42 ug/L
                          50                  100

                                  Scaled operation time (days)
                                                                  150
                                        200
Figure H-28 Impact of extrapolation on SCA prediction of the SDS-TTHM
integral breakthrough curve for Water 8
                                         -407-

-------
   4 -
   3 -
 g
 '-4—'
 CD
CD  2
O  ^
c
o
O
    1 -
   o -ta-
     o
                 Single contactor effluent
                 Extrapolation experimental data points
                •Logistic function best fit - all data  (RA2 = NA)
                 Extrapolated logistic function best fit
                 Blended effluent
                -SCA prediction
                 SCA prediction - extrapolated
                                                                          SDS-MCAA
                              Effluent concentrations were not detected
                                 above the MRL for this parameter
                                                                      EBCT = 7.2min.

                                                                      C0 =  BMRL
                                                  -CD-
                         50
            100

Scaled operation time (days)
-On—

  150
200
Figure H-29  Impact of extrapolation on SCA prediction of the SDS-MCAA
integral breakthrough curve for Water 8
   6 -
           D   Single contactor effluent
           •   Extrapolation experimental data points
         	Logistic function best fit - all data (RA2 = 0.974)
          - -  Extrapolated logistic function best fit  (RA2 = 0.953)
           O   Blended effluent
         	SCA prediction
         	SCA prediction - extrapolated
                                                                          SDS-DCAA
                                                                       EBCT = 7.2min.

                                                                       C0 = 10.7 ug/L
                         50
                                             100

                                 Scaled operation time (days)
                                150
                      200
 Figure H-30 Impact of extrapolation on SCA prediction of the SDS-DCAA
 integral breakthrough curve for Water 8
                                        -408-

-------
   6 -
   5 -
   4 -
g
'-4—'
CD
   3 -
    Single contactor effluent
    Extrapolation experimental data points
   •Logistic function best fit - all data (RA2 = 0.93)
    Extrapolated logistic function best fit (RA2 = 0.835)
    Blended effluent
   -SCA prediction
    SCA prediction - extrapolated
                                                                       SDS-TCAA
                                                                       EBCT = 7.2 min.

                                                                       c0= 12.7 ug/L
                         50
                                  100

                      Scaled operation time (days)
  150
                                                             200
Figure H-31  Impact of extrapolation on SCA prediction of the SDS-TCAA
integral breakthrough  curve for Water 8
   4 -
   3 -
o
'-4—'
CD
CD
O
c
o
O
   1 -
   0 -K>

     0
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data  (RA2 = NA)
 •  • Extrapolated logistic function best fit
 O   Blended effluent
	SCA prediction
	SCA prediction - extrapolated
                                                                       SDS-MBAA
                   Effluent concentrations were not detected
                      above the MRL for this parameter
                                                                       EBCT = 7.2 min.

                                                                       C0= BMRL
                                       -CD-
50                  100

        Scaled operation time (days)
•Or—

  150
                                                                           200
 Figure H-32 Impact of extrapolation on SCA prediction of the SDS-MBAA
 integral  breakthrough curve for Water 8
                                         -409-

-------
   2.0
   1.5 -
g
'•S3  1-0 H
o>
o
o
O
   0.5 -
   0.0 HO-

      0
                   Single contactor effluent
                   Extrapolation experimental data points
                  •Logistic function best fit - all data (RA2 = NA)
                   Extrapolated logistic function best fit
                   Blended effluent
                  -SCA prediction
                   SCA prediction - extrapolated
                                                                       SDS-DBAA
                              Effluent concentrations were not detected
                                 above the MRL for this parameter
                                                                       EBCT = 7.2 min.

                                                                       C0 = BMRL
                    -ii ii i urryjT—rn	Oh	O>

                          50
     -CD-
•OT-
100
  150
200
                                  Scaled operation time (days)
Figure H-33 Impact of extrapolation on SCA prediction of the SDS-DBAA
integral breakthrough curve for Water 8
   16
   14 -


   12 -


   10 -
                Single contactor effluent
                Extrapolation experimental data points
               •Logistic function best fit - all data  (RA2 = 0.955)
                Extrapolated logistic function best fit (RA2 = 0.911)
                Blended effluent
               -SCA prediction
               • SCA prediction - extrapolated
                                                                        SDS-HAA5
                                                                       EBCT = 7.2 min.

                                                                       c0 =  23 ug/L
                          50
                                              100

                                  Scaled operation time (days)
                    150
                      200
 Figure H-34  Impact of extrapolation on SCA prediction of the SDS-HAA5
 integral  breakthrough curve for Water 8
                                         -410-

-------
   3.5
   3.0 -
   2.5 -
   2.0 -
.o
"co
o>
o
c
o
O
   1.5 -
   1.0 -
   0.5 -



   0.0 -K>

      0
EBCT = 7.2min.

C0 =  3 |jg/L
                   50
                                                                          SDS-BCAA
                                                                                   O
                                            Single contactor effluent
                                        •   Extrapolation experimental data points
                                      	Logistic function best fit - all data (RA2 = 0.971)
                                      -  -  -  Extrapolated logistic function best fit (RA2 = 0.97)
                                        O   Blended effluent
                                      	SCA prediction
                                      	SCA prediction - extrapolated
            100

Scaled operation time (days)
150
200
 Figure H-35  Impact of extrapolation on SCA prediction of the SDS-BCAA
 integral breakthrough curve for Water 8
   20
   15 -
o
73 10
o>
o
c
o
O
    5 H
    D   Single contactor effluent
    •   Extrapolation experimental data points
 	Logistic function best fit - all data  (RA2 = 0.956)
 - - -  Extrapolated logistic function best fit (RA2 = 0.923)
    O   Blended effluent
 	SCA prediction
 	SCA prediction - extrapolated
                                                                          SDS-HAA6
                           50                   100

                                   Scaled operation time (days)
                                                             150
                                                     200
 Figure H-36  Impact of extrapolation on SCA prediction of the SDS-HAA6
 integral  breakthrough curve for Water 8
                                           -411-

-------
   3.0
   2.5 -
   2.0 -
o
73  1-5 H
O>
o
o  1.0 H
   0.5 -
   0.0 -O

      0
        EBCT = 7.2min.

        C0 = 3 |jg/L
                                                                  SDS-DCBAA
                                    D   Single contactor effluent
                                    •   Extrapolation experimental data points
                                       •Logistic function best fit - all data (RA2 = 0.939)
                                 - - -  Extrapolated logistic function best fit (RA2 = 0.914)
                                    O   Blended effluent
                                 	SCA prediction
                                 	SCA prediction - extrapolated
                 50
            100

Scaled operation time (days)
  150
200
 Figure H-37 Impact of extrapolation on SCA prediction of the SDS-DCBAA
 integral breakthrough curve for Water 8
   2.0
   1.5 -
o
73  1-0 H
o>
o
c
o
O

   0.5 -
  D  Single contactor effluent
  •  Extrapolation experimental data points
	Logistic function best fit - all data (RA2 = NA)
• • • Extrapolated logistic function best fit
  O  Blended effluent
	SCA prediction
	SCA prediction - extrapolated
   o.o -k>
      o
                 50
                     Effluent concentrations were not detected
                         above the MRL for this parameter
                                          -CT-
                                                                          SDS-CDBAA
                                                                         EBCT = 7.2min.

                                                                         c0 = BMRL
            100

Scaled operation time (days)
•Or—

  150
200
 Figure H-38 Impact of extrapolation on SCA prediction of the SDS-CDBAA
 integral breakthrough curve for Water 8
                                          -412-

-------
   4 -
O)


.0

I

I 2

o
O


   1 -
           D   Single contactor effluent
           •   Extrapolation experimental data points
               Logistic function best fit - all data (RA2 = NA)
           •  • Extrapolated logistic function best fit
           O   Blended effluent
          	SCA prediction
          	SCA prediction - extrapolated
                                                                        SDS-TBAA
                               Effluent concentrations were not detected
                                  above the MRL for this parameter
                                                                       EBCT = 7.2 min.

                                                                       C0= BMRL
                   -iMI i urryjT—rn	Qh	CD-

                         50
                                                   -CD-
                                              100
150
200
                                  Scaled operation time (days)
 Figure H-39 Impact of extrapolation on SCA prediction of the SDS-TBAA
 integral breakthrough curve for Water 8
   20
   15 -
o
'•S3  10
o>
o
c
o
O
    5 -
               Single contactor effluent
               Extrapolation experimental data points
              •Logistic function best fit - all data (RA2 = 0.952)
               Extrapolated logistic function best fit (RA2 = 0.924)
               Blended effluent
              -SCA prediction
              • SCA prediction - extrapolated
                                                                        SDS-HAA9
                                                                       EBCT = 7.2 min.

                                                                       c0 =  30 ug/L
                          50
                                              100

                                  Scaled operation time (days)
150
200
Figure H-40 Impact of extrapolation on SCA prediction of the SDS-HAA9
integral breakthrough curve for Water 8
                                          -413-

-------
This page intentionally left blank.
             -414-

-------
Appendix I:  Impact of Extrapolation on Dl Prediction of the Integral
Breakthrough Curve
                              -415-

-------
   3.0
   2.5 -
   2.0 -
o
•^= -I c
CO  '•*•
O>
o
o 1.0 H
O
   0.5 -
   0.0
 D   Single contactor effluent
 •   Extrapolation experimental data points
     Logistic function best fit - all data  (RA2 = 0.984)
 -  - Extrapolated logistic function best fit (RA2 = 0.979)
 O   Blended effluent
	Dl prediction
	Dl prediction - extrapolated
                                                                               TOO
                                                                         EBCT = 20 min.

                                                                         c0 = 3.08 mg/L
                  50
                 100         150         200         250

                       Scaled operation time (days)
300
350
 Figure 1-1 Impact of extrapolation on the Dl prediction of the TOC integral
 breakthrough curve for Water 5
   0.035
   0.030 -
   0.025 -
o
o>
o
   0.020 -
   0.000
    D   Single contactor effluent
    •   Extrapolation experimental data points
   	Logistic function best fit - all data (RA2 = 0.992)
    -  -  Extrapolated logistic function best fit (RA2 = 0.984)
    O   Blended effluent
   	Dl prediction
   	Dl prediction - extrapolated
                                                                              UV254
                                                                         EBCT = 20 min.

                                                                         c0 =  0.051 1/cm
                    50         100         150        200         250

                                    Scaled operation time (days)
                                                                 300
            350
 Figure I-2  Impact of extrapolation on the Dl prediction of the UV254 integral
 breakthrough curve for Water 5
                                          -416-

-------
 O
    150
    125 -
    100 -
 o   75 -
 o>
 £   50 H
 o
 O
     25 -
   D   Single contactor effluent
   •   Extrapolation experimental data points
       Logistic function best fit - all data  (RA2 = 0.995)
- - -  Extrapolated logistic function best fit (RA2 = 0.99)
   O   Blended effluent
	Dl prediction
	Dl prediction - extrapolated
                                                                             SDS-TOX
                                                                        EBCT = 20 min.

                                                                        C0 = 205 ug/L Cl-
                   50         100         150        200        250

                                   Scaled operation time (days)
                                                               300
            350
 Figure I-3  Impact of extrapolation on the Dl prediction of the SDS-TOX
 integral breakthrough curve for Water 5
    12
    10 -
 o
 '«   6 H
 o>
 o
 O
     2 -
     0 \
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data (RA2 = 0.987)
 -  - Extrapolated logistic function best fit  (RA2 = 0.982)
 O   Blended effluent
	Dl prediction
	Dl prediction - extrapolated
                                                                              SDS-CF
                                                                       EBCT = 20 min.

                                                                       C0 =  23.7 ug/L
                  50
                100         150        200         250

                      Scaled operation time (days)
300
350
Figure I-4  Impact of extrapolation on the Dl prediction of the SDS-CF integral
breakthrough curve for Water 5
                                          -417-

-------
   16


   14 -


   12 -

Q"

| 10-

g
'•£  8 H
o
O
            D  Single contactor effluent
            •  Extrapolation experimental data points
           	Logistic function best fit - all data (RA2 = 0.97)
            - - Extrapolated logistic function best fit (RA2 = 0.967)
            O  Blended effluent
           	Dl prediction
           	Dl prediction - extrapolated
                                                                       SDS-BDCM
                                                                       EBCT = 20 min.

                                                                       C0 = 10.8 ug/L
                 50
                             100        150         200         250

                                  Scaled operation time (days)
300
350
Figure I-5 Impact of extrapolation on the Dl prediction of the SDS-BDCM
integral breakthrough curve for Water 5
   20
o>
o
c
o
o
   15 -
   10 -
    5 -
             D   Single contactor effluent
             •   Extrapolation experimental data points
            	Logistic function best fit - all data (RA2 = 0.988)
             -  - Extrapolated logistic function best fit (RA2 = 0.986)
             O   Blended effluent
            	Dl prediction
            	Dl prediction - extrapolated
                                                                       SDS-DBCM
                                                                       EBCT = 20 min.

                                                                       C0 =  22.7 ug/L
                 50
                             100        150         200         250

                                  Scaled operation time (days)
300
350
Figure I-6  Impact of extrapolation on the Dl prediction of the SDS-DBCM
integral breakthrough curve for Water 5
                                         -418-

-------
   3.0



   2.5 -



§2.0-
3.
c
g
'•S3 1-5 H
CD
o
§ 1.0 -
O
   0.5 -
   0.0 4
        EBCT = 20 min.

        C0= 1.2|jg/L
                                                                              SDS-BF
                                           D   Single contactor effluent
                                           •   Extrapolation experimental data points
                                        	Logistic function best fit - all data (RA2 = 0.969)
                                        - - -  Extrapolated logistic function best fit (RA2 = 0.969)
                                           O   Blended effluent
                                        	Dl prediction
                                        	Dl prediction - extrapolated
                  50
                             100        150         200         250

                                  Scaled operation time (days)
300
350
Figure I-7  Impact of extrapolation on the Dl prediction of the SDS-BF integral
breakthrough curve for Water 5
   50
   40 -
   30 -
o
'-4—'
CD
   20 -
o
O
   10 -
    o 4
               D   Single contactor effluent
               •   Extrapolation experimental data points
              	Logistic function best fit - all data (RA2 = 0.979)
               -  - Extrapolated logistic function best fit (RA2 = 0.975)
               O   Blended effluent
              	Dl prediction
              	Dl prediction - extrapolated
                                                                           SDS-TTHM
                                                                       EBCT = 20 min.

                                                                       C0 = 58 ug/L
                 50
                             100         150         200        250

                                  Scaled operation time (days)
300
350
Figure I-8  Impact of extrapolation on the Dl prediction of the SDS-TTHM
integral breakthrough curve for Water 5
                                         -419-

-------
    4 -
    3 -
 g
 '-4—'
 CD
CD  2 -
o  ^ ~
o
o
    1 -
    0 -
               D   Single contactor effluent
               •   Extrapolation experimental data points
              	Logistic function best fit - all data (RA2 = NA)
               •  • Extrapolated logistic function best fit
               O   Blended effluent
              	Dl prediction
              	Dl prediction - extrapolated
                              Insufficient data measured above the MRL
                            to perform curve fit and extrapolation analysis
                -Di—Ol—i-O—Oh-

                 50         100
                                       150         200         250

                                 Scaled operation time (days)
                                                                           SDS-MCAA
EBCT = 20 min.

c0 =  2 ug/L


    300
350
Figure I-9  Impact of extrapolation on the Dl prediction of the SDS-MCAA
integral breakthrough curve for Water 5
    7
 O
 '-4—'
 CD
   6 -


   5 -


   4 -


   3 -
 o
 O  2 4
    1 -
    o 4
               D   Single contactor effluent
               •   Extrapolation experimental data points
             	Logistic function best fit - all data  (RA2 = 0.978)
              - -  Extrapolated logistic function best fit (RA2 = 0.963)
               O   Blended effluent
             	Dl prediction
             	Dl prediction - extrapolated
                                                                           SDS-DCAA
                 50
                            100        150         200         250

                                 Scaled operation time (days)
                                                                       EBCT = 20 min.

                                                                       C0 = 10.3 |jg/L
    300
350
 Figure 1-10 Impact of extrapolation on the Dl prediction of the SDS-DCAA
 integral  breakthrough curve for Water 5
                                         -420-

-------
   6 -
   5 -
.g
"co
   4 -
   3 -
CD
O

O
O  2H
   1 -
   o 4
 Single contactor effluent
 Extrapolation experimental data points
•Logistic function best fit - all data  (RA2 = 0.989)
 Extrapolated logistic function best fit (RA2 = 0.987)
 Blended effluent
-Dl prediction
 Dl prediction - extrapolated
                                                                        SDS-TCAA
                                                        EBCT = 20 min.

                                                        C0 =  12.7 ug/L
                50
             100        150         200

                  Scaled  operation time (days)
                                   250
    300
350
Figure 1-11  Impact of extrapolation on the Dl prediction of the SDS-TCAA
integral breakthrough curve for Water 5
   4 -
   3 -
o
'-4—'
CD
CD
O
c
o
O
   1 -
   0 -O-

     0
    Single contactor effluent
    Extrapolation experimental data points
   •Logistic function best fit - all data  (RA2 = NA)
    Extrapolated logistic function best fit
    Blended effluent
   -Dl prediction
   • Dl prediction - extrapolated
                                                                        SDS-MBAA
               Effluent concentrations were not detected
                  above the MRL for this parameter
 50
100         150         200        250

      Scaled operation time (days)
EBCT = 20 min.

c0 = 1 ug/L


    300
350
 Figure 1-12  Impact of extrapolation on the Dl prediction of the SDS-MBAA
 integral  breakthrough curve for Water 5
                                         -421-

-------
   3.5
   3.0 -
   2.5 -
.g
"co
   2.0 -
   1.5 -
o>
o
c
o
0  1.0 H
   0.5 -
   o.o 4
EBCT = 20 min

c0 = 2 |jg/L
                  50
                                                                        SDS-DBAA
                                 D   Single contactor effluent
                                 •   Extrapolation experimental data points
                                    •Logistic function best fit - all data (RA2 = 0.992)
                              -  -  -  Extrapolated logistic function best fit (RA2 = 0.992)
                                 O   Blended effluent
                              	Dl prediction
                              	Dl prediction - extrapolated
                    100        150         200         250

                         Scaled operation time (days)
            300
            350
 Figure 1-13 Impact of extrapolation on the Dl prediction of the SDS-DBAA
 integral  breakthrough curve for Water 5
   20
                                                                        EBCT = 20 mm.

                                                                        C0 = 28 ug/L
   15 -
o
'•S3  10
o>
o
c
o
O
    5 -
     D   Single contactor effluent
     •   Extrapolation experimental data points
    	Logistic function best fit - all data (RA2 = 0.965)
     -  -  Extrapolated logistic function best fit (RA2 = 0.959)
     O   Blended effluent
    	Dl prediction
    	Dl prediction - extrapolated
                                                                        SDS-HAA5
                             100
                               150         200

                         Scaled operation time (days)
250
300
350
 Figure 1-14  Impact of extrapolation on the Dl prediction of the SDS-HAA5
 integral  breakthrough curve for Water 5
                                         -422-

-------
   6 -
 g
'•S3 4
 CD
 o
 c
 o
O
   2 -
   0 -<
 D   Single contactor effluent
 •   Extrapolation experimental data points
     Logistic function best fit - all data (RA2 = 0.987)
 -  - Extrapolated logistic function best fit  (RA2 = 0.982)
 O   Blended effluent
	Dl prediction
	Dl prediction - extrapolated            _^^^". D
                                                                       SDS-BCAA
                                                                       EBCT = 20 min.

                                                                       C0 =  7.3 ug/L
                 50
                 100         150        200         250

                       Scaled operation time (days)
300
350
Figure 1-15 Impact of extrapolation on the Dl prediction of the SDS-BCAA
integral breakthrough curve for Water 5
   25
   20 -
 o
'-4—'
 CD
 o
o
   15 -
   10 -
    5 -
    0 4
   D  Single contactor effluent
   •  Extrapolation experimental data points
  	Logistic function best fit - all data (RA2 = 0.976)
   - - Extrapolated logistic function best fit (RA2 = 0.971)
   O  Blended effluent
  	Dl prediction
  	Dl prediction - extrapolated
                                                                       SDS-HAA6
                                                                       EBCT = 20 min.

                                                                       c0 =  34 ug/L
                 50
                  100        150         200         250

                       Scaled operation time (days)
300
350
Figure 1-16  Impact of extrapolation on the Dl prediction of the SDS-HAA6
integral breakthrough curve for Water 5
                                         -423-

-------
   10
O)
.0
"co
o
O
    6-\
    2 -
    0 -i
             D   Single contactor effluent
             •   Extrapolation experimental data points
            	Logistic function best fit - all data  (RA2 = 0.987)
             -  - Extrapolated logistic function best fit (RA2 = 0.987)
             O   Blended effluent
            	Dl prediction
            	Dl prediction - extrapolated
                                                                      SDS-DCBAA
                                                                       EBCT = 20 min.

                                                                       c0= 10.7 ug/L
                 50
             100         150        200        250

                   Scaled operation time (days)
                                         300
            350
Figure 1-17 Impact of extrapolation on the Dl prediction of the SDS-DCBAA
integral breakthrough curve for Water 5
   5 -
o
•^
CD
CD
O
o  -2\
O
   1 -
 Single contactor effluent
 Extrapolation experimental data points
•Logistic function best fit - all data (R^2 = 0.95)
 Extrapolated logistic function best fit (RA2 = 0.947)
 Blended effluent
-Dl prediction
• Dl prediction - extrapolated
                                                                      SDS-CDBAA
                                                                      --O
                                                                       EBCT = 20 min.

                                                                       c0 = 3.7 ug/L
                                   -O
                50
             100
      150        200        250

Scaled operation time (days)
300
350
 Figure 1-18 Impact of extrapolation on the Dl prediction of the SDS-CDBAA
 integral  breakthrough curve for Water 5
                                         -424-

-------
   4 -
   3 -
.g
"co
 o>
 o
 c
 o
O
   1 -
   o -b-
     o
                                                                        SDS-TBAA
               Single contactor effluent
               Extrapolation experimental data points
              •Logistic function best fit - all data (RA2 = NA)
               Extrapolated logistic function best fit
               Blended effluent
              -Dl prediction
               Dl prediction - extrapolated
                           Effluent concentrations were not detected
                              above the MRL for this parameter
             50
100         150        200        250

      Scaled operation time (days)
EBCT = 20 min.

C0 =  BMRL


    300
350
Figure 1-19 Impact of extrapolation on the Dl prediction of the SDS-TBAA
integral breakthrough curve for Water 5
   40
 o>
 £
 o
O
35 -


30 -


25 -


20 -


15 -


10 -


 5 -


 0 -
              D   Single contactor effluent
              •   Extrapolation experimental data points
            	Logistic function best fit - all data (RA2 = 0.99)
             - -  Extrapolated logistic function best fit (RA2 = 0.988)
              O   Blended effluent
            	Dl prediction
            	Dl prediction - extrapolated         __^j»*» " n
                                                                    SDS-HAA9
                                                                       EBCT = 20 min.

                                                                       C0 =  48 ug/L
                  50
                          100         150         200        250

                               Scaled operation time (days)
                                               300
                350
Figure I-20  Impact of extrapolation on the Dl prediction of the SDS-HAA9
integral breakthrough curve for Water 5
                                         -425-

-------
   2.0
   1.5 -
'•a 1-0 H
CD
o
c
o
O
   0.5 -
   0.0
  Single contactor effluent
  Extrapolation experimental data points
  •Logistic function best fit - all data  (RA2 = 0.974)
  Extrapolated logistic function best fit (RA2 = 0.961)
  Blended effluent
  -Dl prediction
  Dl prediction - extrapolated
                                                                               TOO
                                                                         EBCT = 7.2 min.

                                                                         c0 = 2.02 mg/L
                           50                  100

                                   Scaled operation time (days)
                                                      150
                    200
 Figure 1-21  Impact of extrapolation on the Dl prediction of the TOC integral
 breakthrough curve for Water 8
   0.025
   0.020 -
o
^ 0.015

CD
O

CD

o 0.010
   0.005 -
   0.000 -K>

         0
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data  (RA2 = 0.994)
 -  - Extrapolated logistic function best fit (RA2 = 0.98
 O   Blended effluent
	Dl prediction
	Dl prediction - extrapolated
                                                                              UV254
                                                                         EBCT = 7.2 min.

                                                                         C0 =  0.033 1/cm
              50                  100

                      Scaled operation time (days)
150
200
 Figure I-22  Impact of extrapolation on the Dl prediction of the UV254 integral
 breakthrough curve for Water 8
                                          -426-

-------
 O
 o
 "co
 O>
 o
 c
 o
 O
    125
    100 -
     75 -
     50 -
     25 -
 Single contactor effluent
 Extrapolation experimental data points
•Logistic function best fit - all data (RA2 = 0.99)
 Extrapolated logistic function best fit (RA2 = 0.98)_
 Blended effluent
-Dl prediction
 Dl prediction - extrapolated
                                                                     SDS-TOX
                                                                        EBCT = 7.2 min.

                                                                        c0=  156 ug/LCI-
                            50                 100

                                   Scaled operation time (days)
                                                  150
200
 Figure I-23  Impact of extrapolation on the Dl prediction of the SDS-TOX
 integral breakthrough curve for Water 8
    25
    20 -
    15 -
 o
 '-4—'
 CD
 o
 O
    10 -
  Single contactor effluent
  Extrapolation experimental data points
 •Logistic function best fit - all data (RA2 = 0.985)
  Extrapolated logistic function best fit (RA2 = 0.917)
  Blended effluent
 -Dl prediction
 • Dl prediction - extrapolated
                                                                       SDS-CF
                                                                       EBCT = 7.2 min.

                                                                       C0 = 29.1 ug/L
                           50                  100

                                   Scaled operation time (days)
                                                  150
200
Figure I-24 Impact of extrapolation on the Dl prediction of the SDS-CF
integral breakthrough curve for Water 8
                                          -427-

-------
   4 -
 D   Single contactor effluent
 •   Extrapolation experimental data points
	Logistic function best fit - all data (RA2 = 0.946)
 -  - Extrapolated logistic function best fit (RA2 = 0.942)
 O   Blended effluent
	Dl prediction                            D
	Dl prediction - extrapolated
                                           O
                                                                    SDS-BDCM
                         50                   100

                                  Scaled operation time (days)
                                                          150
200
Figure I-25 Impact of extrapolation on the Dl prediction of the SDS-BDCM
integral breakthrough curve for Water 8
   12
   10 -
^  8-
o
13   6
*-t
O>
o
o   4
O
    2 -
    0 -lO
  D   Single contactor effluent
  •   Extrapolation experimental data points
 	Logistic function best fit - all data (RA2 = 0.98)
  - - Extrapolated logistic function best fit (RA2 = 0.978)
  O   Blended effluent                         D
 	Dl prediction
 	Dl prediction - extrapolated
                                                                       SDS-DBCM
                                                                       EBCT = 7.2 min.

                                                                       C0 = 10.3 ug/L
                          50                  100

                                   Scaled operation time (days)
                                                          150
200
Figure I-26  Impact of extrapolation on the Dl prediction of the SDS-DBCM
integral breakthrough curve for Water 8
                                          -428-

-------
   2.0
   1.5 -
g
'•S3 1-0 H
o>
o
o
O
   0.5 -
              D   Single contactor effluent
              •   Extrapolation experimental data points
           	Logistic function best fit - all data (RA2 = NA)
           -  -  -  Extrapolated logistic function best fit
              O   Blended effluent
           	Dl prediction
           	Dl prediction - extrapolated
   o.o -lo-
       o
                                                                  O
                              Insufficient data measured above the MRL
                            to perform curve fit and extrapolation analysis
                    -ii n i urr>-n—rn—Oh	CD-

                           50
     -CD-
                                                                               SDS-BF
                                                                        EBCT = 7.2min.

                                                                        C0 = BMRL
100
150
200
                                   Scaled operation time (days)
Figure I-27 Impact of extrapolation on the Dl prediction of the SDS-BF
integral breakthrough curve for Water 8
   40
   30 -
o
'•S3 20
o>
o
o
O
   10 -
    o Ho
      o
              D   Single contactor effluent
              •   Extrapolation experimental data points
             	Logistic function best fit - all data (RA2 = 0.98)
             -  -  Extrapolated logistic function best fit  (RA2 = 0.966)
              O   Blended effluent
             	Dl prediction
             	Dl prediction - extrapolated
                                                                            SOS-^HM
                                                                        EBCT = 7.2min.

                                                                        c0 = 42 ug/L
                          50                  100

                                  Scaled operation time (days)
                    150
                    200
Figure I-28 Impact of extrapolation on the Dl prediction of the SDS-TTHM
integral breakthrough curve for Water 8
                                         -429-

-------
    4 -
    3 -
 g
 '-4—'
 CD
CD  9 -
o  ^ ~
o
o
    1 -
                 Single contactor effluent
                 Extrapolation experimental data points
                •Logistic function best fit - all data (RA2 = NA)
                 Extrapolated logistic function best fit
                 Blended effluent
                -Dl prediction
                 Dl prediction - extrapolated
                                                                           SDS-MCAA
                              Effluent concentrations were not detected
                                 above the MRL for this parameter
                                                                       EBCT = 7.2min.

                                                                       C0 =  BMRL
                   -dEEXB-O—CD	CDi	CD-

                          50
                                                  -CD-
•OT-
                                             100
  150
200
                                  Scaled operation time (days)
Figure I-29  Impact of extrapolation on the Dl  prediction of the SDS-MCAA
integral breakthrough curve for Water 8
    6 -
 o
 '•S3  4
 CD
 O
 c
 o
 O
                 Single contactor effluent
                 Extrapolation experimental data points
                 Logistic function best fit - all data (RA2 = 0.974)
                 Extrapolated logistic function best fit (RA2 = 0.953)
                 Blended effluent
                 Dl prediction
                 • Dl prediction - extrapolated
                                                                           SDS-DCAA
                                                                       EBCT = 7.2min.

                                                                       C0 = 10.7 ug/L
                          50
                                             100

                                 Scaled operation time (days)
  150
200
 Figure I-30 Impact of extrapolation on the Dl prediction of the SDS-DCAA
 integral  breakthrough curve for Water 8
                                         -430-

-------
   6 -
   5 -
.g
"co
   4 -
 Single contactor effluent
 Extrapolation experimental data points
•Logistic function best fit - all data  (RA2 = 0.93)
 Extrapolated logistic function best fit (RA2 = 0.835)
 Blended effluent
-Dl prediction
 Dl prediction - extrapolated
                                                                        SDS-TCAA
                                                                       EBCT = 7.2 min.

                                                                       C0 = 12.7 ug/L
                         50
                                100

                    Scaled operation time (days)
  150
                                                              200
Figure 1-31  Impact of extrapolation on the Dl prediction of the SDS-TCAA
integral breakthrough curve for Water 8
   4 -
   3 -
o
'-4—'
CD
CD
O
c
o
O
   1 -
   0 -K>

     0
      Single contactor effluent
      Extrapolation experimental data points
     •Logistic function best fit - all data  (RA2 = NA)
      Extrapolated logistic function best fit
      Blended effluent
     -Dl prediction
     • Dl prediction - extrapolated
                                                                        SDS-MBAA
                Effluent concentrations were not detected
                    above the MRL for this parameter
                                                                        EBCT = 7.2 min.

                                                                        C0= BMRL
                                     -CD-
50                  100

        Scaled operation time (days)
•Or—

  150
                                                                         200
 Figure I-32  Impact of extrapolation on the Dl prediction of the SDS-MBAA
 integral  breakthrough curve for Water 8
                                         -431-

-------
   2.0
   1.5 -
g
'•S3  1-0 H
CD
o
c
o
O
   0.5 -
   o.o -k>
      0
                Single contactor effluent
                Extrapolation experimental data points
               •Logistic function best fit - all data  (RA2 = NA)
                Extrapolated logistic function best fit
                Blended effluent
               -Dl prediction
                Dl prediction - extrapolated
                                                                        SDS-DBAA
                           Effluent concentrations were not detected
                               above the MRL for this parameter
                                                                        EBCT = 7.2 min.

                                                                        C0 = BMRL
                 -ii n i uoyjT—rn	Oh	O>

                       50
     -CD-
•OT-
100
  150
200
                                   Scaled operation time (days)
 Figure I-33 Impact of extrapolation on the Dl prediction of the SDS-DBAA
 integral  breakthrough curve for Water 8
   16
14 -


12 -


10 -
o
'-4—'
CD
                  Single contactor effluent
                  Extrapolation experimental data points
                  Logistic function best fit - all data (RA2 = 0.955)
                  Extrapolated logistic function best fit (RA2 = 0.911)
                  Blended effluent
                  Dl prediction
                  Dl prediction - extrapolated
                                                                        SDS-HAA5
                                                                        EBCT = 7.2 min.

                                                                        c0 = 23 ug/L
                          50
                                           100

                               Scaled operation time (days)
                    150
                      200
 Figure I-34  Impact of extrapolation on the Dl prediction of the SDS-HAA5
 integral  breakthrough curve for Water 8
                                         -432-

-------
   3.0
   2.5 -
   2.0 -
EBCT = 7.2min.

C0 =  3 |jg/L
                                                                        SDS-BCAA
                                           D   Single contactor effluent
                                           •   Extrapolation experimental data points
                                               Logistic function best fit - all data  (RA2 = 0.971)
                                        - -  -  Extrapolated logistic function best fit (RA2 = 0.97)
                                           O   Blended effluent
                                        	Dl prediction
                                        	Dl prediction - extrapolated
                           50
                                      100

                           Scaled operation time (days)
150
200
Figure I-35 Impact of extrapolation on the Dl prediction of the SDS-BCAA
integral breakthrough curve for Water 8
   20
   15 -
 o
'•S3 10
 o>
 o
 c
 o
O
    5 -
    D   Single contactor effluent
    •   Extrapolation experimental data points
   	Logistic function best fit - all data (RA2 = 0.956)
    -  - Extrapolated logistic function best fit  (RA2 = 0.923)
    O   Blended effluent
   	Dl prediction
   	Dl prediction - extrapolated
                          50
                                                                        SDS-HAA6
                                                                       EBCT = 7.2min.

                                                                       c0 =  27 ug/L
                                      100

                          Scaled operation time (days)
150
200
Figure I-36  Impact of extrapolation on the Dl prediction of the SDS-HAA6
integral breakthrough curve for Water 8
                                         -433-

-------
   3.0
   2.5 -
   2.0 -
g
'•§  1.5 H
"c
O>
o
o  1.0 -
O
   0.5 -
   0.0
             D   Single contactor effluent
             •   Extrapolation experimental data points
           	Logistic function best fit - all data  (RA2 = 0.939)
            - -  Extrapolated logistic function best fit (RA2 = 0.914)
             O   Blended effluent
           	Dl prediction      ...
           	Dl prediction - extrapolated
                           50
                                                                       SDS-DCBAA
                                                                  O
                                                                        EBCT = 7.2 min.

                                                                        c0 = 3 ug/L
                                               100

                                   Scaled operation time (days)
                    150
                      200
 Figure I-37  Impact of extrapolation on the Dl prediction of the SDS-DCBAA
 integral  breakthrough curve for Water 8
   2.0
   1.5 -
o
'•S3  1-0 H
o>
o
c
o
O
   0.5 -
   0.0 -K>

      0
               D  Single contactor effluent
               •  Extrapolation experimental data points
             	Logistic function best fit - all data (RA2 = NA)
             • • • Extrapolated logistic function best fit
               O  Blended effluent
             	Dl prediction
             	Dl prediction - extrapolated
                                                                       SDS-CDBAA
                              Effluent concentrations were not detected
                                  above the MRL for this parameter
                                                                        EBCT = 7.2 min.

                                                                        C0 = BMRL
                    -in11 M"r>-n-^"n——m,   rn-

                           50
     -CD-
100
•Or—

  150
200
                                   Scaled operation time (days)
 Figure I-38  Impact of extrapolation on the Dl prediction of the SDS-CDBAA
 integral  breakthrough curve for Water 8
                                         -434-

-------
    4 -
    3 -
 g
 '-4—'
 CD
CD  9 -
o  ^ ~
o
o
    1  -
            D  Single contactor effluent
            •  Extrapolation experimental data points
           	Logistic function best fit - all data (RA2 = NA)
            - - Extrapolated logistic function best fit
            O  Blended effluent
           	Dl prediction
           	Dl prediction - extrapolated
                              Effluent concentrations were not detected
                                  above the MRL for this parameter
                        JOHXB——Oli   CT-

                          50
                                                  -CT-
                                                                           SDS-TBAA
                                                                       EBCT = 7.2min.

                                                                       C0= BMRL
                                             100
-On—

  150
200
                                  Scaled operation time (days)
Figure I-39 Impact of extrapolation on the Dl prediction of the SDS-TBAA
integral breakthrough curve for Water 8
    20
    15 -
 o
 '•SS  10
 CD
 O
 c
 o
 O
     5 -
           D   Single contactor effluent
           •   Extrapolation experimental data points
          	Logistic function best fit - all data  (RA2 = 0.952)
           -  -  Extrapolated logistic function best fit (RA2 = 0.924)
           O   Blended effluent
          	Dl prediction
          	Dl prediction - extrapolated
                           50
                                                                            SDS-HAA9
                                                                       EBCT = 7.2min.

                                                                       c0 = 30 ug/L
                                              100

                                  Scaled operation time (days)
  150
200
Figure I-40  Impact of extrapolation on the Dl prediction of the SDS-HAA9
integral breakthrough curve for Water 8
                                         -435-

-------