HomeMy WebLinkAboutNC6170022580_20070701_Camp Lejeune Military Res._FRCERCLA SPD_Analyses of Groundwater Flow Contamination Fate and Transport and Distribution of Drinking Water at Tawara Terrace and Vicinity-OCR1,000
~
~ ~ 100 ~
~ ~ 10
~ 1l ,
a z " g 0.1
~ z
8 0.01
~
0.001
Jan Jae
1950 1955
• TT-45
1-ATSOR health study----"1
~ ~ell TT-26
,i; Wnll TT-23
I W~II TT-25
Finished water from lwellTT-~ water treatment plant
------------------we,1n61-Max11T111m co1itaminant level / l i I
~W,IITT31 I
I I
I o Finished I
I
water I sample
I from water
I treatment
plant
Jae Jan J;rn .J:,n Ja, Jan Jan Jan
1960 1965 1970 1975 1980 1985 1990 1995
-----
Atlanta, Georgia-July 2007
Front cover: Historical reconstruction process using data, information sources,
and water-modeling techniques to estimate historical exposures
Maps: U.S. Marine Corps Base Camp Lejeune, North Carolina; Tarawa Terrace area
showing historical water-supply wells and site of ABC One-Hour Cleaners
Photographs on left: Ground storage tank STT-39 and four high-lift pumps used to deliver
finished water from tank STT-39 to Tarawa Terrace water-distribution system
Photograph on right: Equipment used to measure flow and pressure at a hydrant
during field test of the present-day (2004) water-distribution system
Graph: Reconstructed historical concentrations of tetrachloroethylene (PCE) at selected
water-supply wells and in finished water at Tarawa Terrace water treatment plant
Analyses of Groundwater Flow, Contaminant Fate and Transport,
and Distribution of Drinking Water at Tarawa Terrace and Vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina:
Historical Reconstruction and Present-Day Conditions
Chapter A: Summary of Findings
By Morris L. Maslia, Jason B. Sautner, Robert E. Faye, Rene J. Suarez-Soto, Mustafa M. Aral,
Walter M. Grayman, Wonyong Jang, Jinjun Wang, Frank J. Bove, Perri Z. Ruckart,
Claudia Valenzuela, Joseph W. Green, Jr., and Amy L. Krueger
Agency for Toxic Substances and Disease Registry
U.S. Department of Health and Human Services
Atlanta, Georgia
July 2007
ii
Authors
Morris L. Maslia, MSCE, PE, D.WRE, DEE
Research Environmental Engineer and Project Officer
Exposure-Dose Reconstruction Project
Agency for Toxic Substances and Disease Registry
Atlanta, Georgia
Jason B. Sautner, MSCE, EIT
Environmental Health Scientist
Division of Health Assessment and Consultation
Agency for Toxic Substances and Disease Registry
Atlanta, Georgia
Robert E. Faye, MSCE, PE
Hydrologist
Robert E. Faye and Associates
Consultantto Eastern Research Group, Inc.
Lexington, Massachusetts
Rene J. Suarez-Soto, MSCE, EIT
Environmental Health Scientist
Division of Health Assessment and Consultation
Agency for Toxic Substances and Disease Registry
Atlanta, Georgia
Mustafa M. Aral, PhD, PE, Phy
Director and Professor
Multimedia Environmental Simulations Laboratory
School of Civil and Environmental Engineering
Georgia Institute of Technology
Atlanta, Georgia
Walter M. Grayman, PhD, PE
Consulting Engineer
W.M. Grayman Consulting Engineer
Cincinnati, Ohio
Wonyong Jang, PhD
Post Doctoral Fellow
Multimedia Environmental Simulations Laboratory
School of Civil and Environmental Engineering
Georgia Institute of Technology
Atlanta, Georgia
Jinjun Wang, MSCE
Ph.D. Candidate
Multimedia Environmental Simulations Laboratory
School of Civil and Environmental Engineering
Georgia Institute of Technology
Atlanta, Georgia
Frank J. Bove, ScD
Senior Epidemiologist
Division of Health Studies
Agency for Toxic Substances and Disease Registry
Atlanta, Georgia
Perri Z. Ruckart, MPH
Epidemiologist and Principal Investigator
Division of Health Studies
Agency for Toxic Substances and Disease Registry
Atlanta, Georgia
Claudia Valenzuela, MSCE
Post Graduate Research Fellow
Oak Ridge Institute for Science and Education
Agency for Toxic Substances and Disease Registry
Atlanta, Georgia
Joseph W. Green, Jr., MA
Post Graduate Research Fellow
Oak Ridge Institute for Science and Education
Agency for Toxic Substances and Disease Registry
Atlanta, Georgia
Amy L. Krueger, MPH
Post Graduate Research Fellow
Oak Ridge Institute for Science and Education
Agency for Toxic Substances and Disease Registry
Atlanta, Georgia
For additional information write to:
Project Officer
Exposurc~Dose Rcconstructit)ll Pn)jcct
Division of Health Assessment and Consultati(m
Agency for Toxic Substances and Disease Registry
1600 Clifton Road, Mail Stop E-32
Atlanta. Georgia 30333
Suggested citation: Maslia ML, Sautner JB, Faye RE, Suarez-Soto RJ, Aral MM, Grayman WM, Jang W,
Wang J, Bove FJ, Ruckart PZ, Valenzuela C, Green JW Jr, and Krueger AL. Analyses of Groundwater Flow,
Contaminant Fate and Transport, and Distribution of Drinklng Water at Tarawa Terrace and Vicinity, U.S.
Marine Corps Base Camp Lejeune, North Carolina: Historical Reconstruction and Present-Day Conditions-
Chapter A: Summary of Findings. Atlanta, GA: Agency for Toxic Substances and Disease Registry; 2007.
Foreword
The Agency for Toxic Suhstances and Disease Registry (ATSDR), an agency of the
U.S. Department or Health and Human Services. is conducting an epidemiological study
to evaluate whdher in utero and infant (up to I year or age) exposures to volatile organic
compounds in contaminated drinking water at U.S. Marine Corps Base Camp Lejeune.
North Carolina, were associated with spcc.:ific birth dcfccts and childhood canccrs. The study
includes births ticcurring during the peri<id 1968-1985 hl women who were pregnant while
they resided in family housing at the base. During 2004. the study protocol received approval
from the Centers for Disease Control and Prevention Institutional Review Board and the
U.S. Orl1cc or Management and Budget.
Historical exposure data needed for the epidemiological case-control study arc limited.
To obtain estimates of historical exposure. ATS DR is using water-modeling techniques and
the process of historical reconstruction. These methods are used to quantify concentrations
or particular contaminants in finished water and to compute the level and duration of human
exposure to contaminated drinking water.
Final interpretive results for Tarawa Terrace and vicinity-based on information gather-
ing. data interpretations. and water-modeling analyses-arc presented as a series of ATS DR
reports. These reports provide comprehensive descriptions of inf<irmation. data analyses
and interpretations, and modeling results used to reconstruct historical contaminant levels in
drinking water at Tarawa Terrace and vicinity. Each topical subject within the water-modeling
analysis and historical reconstruction process is assigned a chapter letter. Spccilic topics for
each chapter report are listed helow:
Chapter A: Summary or Findings
Chapter B: Gcohydrologic Framework of the Castle Hayne Aquifer System
Chapter C: Simulation of Groundwater Flow
• Chapter D: Properties and Degradation Pathways of Common Organic C()mpounds
in Groundwater
Chapter E: Occurrence of Contaminants in Groundwater
Chapter F: Simulation of the Fate and Transp<lrl ofTetrachh)rocthylcne (PCE)
in Groundv,1ater
Chapter G: Simulation (lf Three-Dimensional Multispecies. Multiphase Mass
Transport of Tetrachloroethylcne (PCE) and Associated Degradation By-Products
Chapter H: Effect of Groundwater Pumping Schedule Variation on Arrival of
Tetrachloroethylene (PCE) at Water-Supply Wells and the Water Treatment Plant
Chapter I: Parameter Sensitivity, Uncertainty, and Variahility Associated with
Model Simulations of Groundwater Flow. Contaminant Fate and Transport. and
Distribution of Drinking Water
Chapter J: Field Tests, Data Analyses. and Simulation of the Distribution
of Drinking Water
Chapter K: Supplemental Information
Electronic versions of these reports and their supp{1rting information and data will be
made available on the ATS DR Camp Lejeune Wch site at http://www.at.wh:cdc.grw/sites/
lejew1e/index.ht111I.
iii
Contents
Authors ......•••.......••••......•••.....•••.....••• 11
Foreword ......•••......•••.......•••.......••••.....•••....••• 111
Glossary and Abbreviations
Abstract
Introduction ........ .
Previous Studies and Purpose of the Current Investigation ...
Tarawa Terrace Chapter Reports
External Peer Review ......... .
Chlorinated Solvents and Volatile Organic Compounds IVOCs)
Naming Conventions ............ .
Maximum Contaminant Levels ....
Historical Background ..
Water-Distribution Investigation
Models Used for Water-Distribution Investigation ..
Data Needs and Availability
Chronology of Events ......... .
Occurrence of Contaminants in Groundwater ......... .
......... x
... Al
............ A2
. .......................................... A4
..................... A5
.. A5
... AB
.................... AB
. ......... A9
.. AlO
........................... All
. ................... All
............. A14
. ........... A15
. ........... A15
Relation of Contamination to Water Supply, Production, and Distribution ............ A17
Hierarchical Approach for Quantifying Exposure .... .
Conceptual Description of Model Calibration ... .
Quantitative Assessment of Model Calibration ......... .
Selected Simulation Results ......... .
Distribution of Tetrachloroethylene IPCE) in Groundwater
January 1958 .. .
January 1968 .. .
December 1984
December 1994 ...... .
Concentration of Tetrachloroethylene (PCE) in Finished Water
Analysis of Degradation By-Products ..
Confidence in Simulation Results.
Water-Supply Well Scheduling Analysis .....
Sensitivity Analysis .......... .
Probabilistic Analysis .............. .
Field Tests and Analyses of the Water-Distribution System
Summary and Conclusions
Availability of Input Data Files, Models, and Simulation Results ....
Acknowledgments ..... .
References ...
. .................. A22
. ................. A22
................. A24
. ................ A32
................ A32
............... A32
. ............. A32
. ........................................... A32
. .......................... A36
.................... A40
. ... A41
.......... A47
. ......... A47
.. A50
. ................ A52
......... A61
........................................... A67
. ..... A70
..A71
........................ A72
Appendix A 1.
Appendix A2.
Summaries of Tarawa Terrace Chapter Reports .......................... A77
Simulated PCE and PCE Degradation By-Products in Finished Water,
Tarawa Terrace Water Treatment Plant, January 1951-March 1987
Appendix A3. Questions and Answers ..
.......... A81
.. A95
Digital video discs IDVO) ........... . . ........... Inside back cover
Disc 1. Camp Lejeuene Water files, Miscellaneous files,
CERCLA Administrative Record files 00001-01499
Disc 2. CERCLA Administrative Record files 01500-02299
Oise 3. CERCLA Administrative Record files 02300-03744,
Calibrated model input files, IMODFLOW-96, MT3DMS, EPANET 2)
V
VI
Figures
Plate 1 Map showing location of wells and boreholes, groundwater-flow model boundary,
and present-day 12004) water-distribution systems serving Tarawa Terrace,
Holcomb Boulevard, and Hadnot Point and vicinity, U.S. Marine Corps Base
Camp Lejeune, North Carolina ... Inside back cover
Al. Map showing selected base housing and historical water-supply areas,
Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina ................ A3
A2-A3. Diagrams showing-
A2. Relation among Chapter A report !Summary of Findings), Chapters B-K reports, historical
reconstruction process, and the ATSDR epidemiological case-control study, Tarawa
_ Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina .............. A7
A3. Chronology of events related to supply and contamination of drinking
water, Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune,
North Carolina....... . ................ A 16
A4. Map showing location of groundwater-flow and contaminant fate and transport
modeling areas and water-supply facilities used for historical reconstruction analyses,
Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina .... A18
A5-A8. Graphs showing-
A5. Historical operations of water-supply wells, 1952-87, Tarawa Terrace and
vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina.... . ............ A 19
A6. Total annual groundwater pumpage at water-supply wells, 1952-1987, Tarawa
Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina ................ A20
A7. Estimated monthly delivery of finished water to the Tarawa Terrace water-distribution
system, 2000-2004, U.S. Marine Corps Base Camp Lejeune, North Carolina ................... A21
AS. Measured diurnal pattern 124 hours) of delivered finished water during field test,
September 22-0ctober 12, 2004, Tarawa Terrace water-distribution system,
U.S. Marine Corps Base Camp Lejeune, North Carolina........ . ......... A22
A9. Venn diagrams showing hierarchical approach of model calibration used to estimate
concentration of finished water: la) predevelopment groundwater flow, lb) transient
groundwater flow, I c) con ta min ant fate and transport, and Id) water-supply well mixing,
Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina ......... A23
A10-A12. Graphs showing-
A13-A15.
AlO. Observed and simulated water levels, model layer 1, and calibration targets
for la) predevelopment (steady-state) conditions and lb)transient conditions,
1951-1994, Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune,
North Carolina........................................................................ . .. A24
A 11. Comparison of observed and nondetect tetrachloroethylene sample data with
calibration targets and simulated concentrations at water-supply wells,
Tarawa Terrace, U.S. Marine Corps Base Camp Lejeune, North Carolina......... . ...... A30
A12. Comparison of observed and nondetecttetrachloroethylene sample data with
calibration targets and simulated concentrations at the water treatment plant,
Tarawa Terrace, U.S. Marine Corps Base Camp Lejeune, North Carolina ....................... A31
Maps showing simulated la) water level and direction of groundwater flow and
lb) distribution of tetra ch loroethylene, model layer 1, Tarawa Terrace and vie inity,
U.S. Marine Corps Base Camp Lejeune, North Carolina-
A13. January 1958 ..... . . .... A33
A14. January 1968 ......................... . . .................... A34
A15. Oecember1984 ............. . . ....... A35
A16. Diagram showing perspective views of the simulated distributions oftetrachloroethylene,
model layers 1, 3, and 5, December 1984, Tarawa Terrace and vicinity, U.S. Marine Corps
Base Camp Lejeune, North Carolina....... . .................... A37
A17. Maps showing simulated /a/water level and direction of groundwater flow and
lb/distribution oftetrachloroethylene, model layer 1, December 1994, Tarawa Terrace
and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina. .. .......... A38
A18-A19. Graphs showing-
A 18. Concentration of tetra c hloroethylene: simulated at selected water-supply wells
and in finished water at the water treatment plant, and measured in finished
water at the water treatment plant, Tarawa Terrace, U.S. Marine Corps Base
Camp Lejeune, North Carolina....... ... A39
A19. Simulated concentration oftetrachloroethylene IPCE) and degradation
by-products trichloroethylene ITCE), trans-1,2-dichloroethylene 11,2-tDCE),
and vinyl chloride IVC) (a) at water-supply well TT-26 and (b) in finished water
from water treatment plant, Tarawa Terrace, U.S. Marine Corps Base
Camp Lejeune, North Carolina......... .. .......... A43
A20. Map showing simulated distribution of vapor-phase tetrachloroethylene to a depth of
10 feet below land surface, (a) December 1984 and (b) December 1994, Tarawa Terrace
and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina ....... A45
A21. Graph showing sensitivity of tetrachloroethylene concentration in finished water
at the water treatment plant to variation in water-supply well operations, Tarawa Terrace,
U.S. Marine Corps Base Camp Lejeune, North Carolina ..... A48
A22. Diagram showing conceptual framework for (a) deterministic analysis and
(b) probabilistic analysis........... .. .. A53
A23-A24. Histograms showing-
A23. Probability density functions for la) recharge rate, lb) mass loading rate !source
concentration), and le) dispersivity used to conduct probabilistic analyses. .. .. A55
A24. Probability of occurrence oftetrachloroethylene contamination in
finished water at the water treatment plant derived from probabilistic analysis
using Monte Carlo simulation for /a/January 1958, lb/January 1968, /c/January 1979,
and Id/January 1985, Tarawa Terrace, U.S. Marine Corps Base Camp Lejeune,
North Carolina.. .. ............... A57
A25-A26. Graphs showing-
A25. Probabilities of exceeding tetrachloroethylene concentrations in finished
water at the water treatment plant derived from probabilistic analysis using
Monte Carlo simulation for /a/selected years, 1958-1985, and lb/selected months,
January 1985-February 1987, Tarawa Terrace, U.S. Marine Corps Base
Camp Lejeune, North Carolina.... .. A59
A26. Concentrations of tetrachloroethylene in finished water at the water
treatment plant derived from probabilistic analysis using Monte Carlo simulation,
Tarawa Terrace, U.S. Marine Corps Base Camp Lejeune, North Carolina ...................... A60
A27. Map showing locations of continuous recording water-quality monitoring equipment
and present-day 12004) Tarawa Terrace and Holcomb Boulevard water-distribution
systems used for conducting a fluoride tracer test, September 22-0ctober 12, 2004,
U.S. Marine Corps Base Camp Lejeune, North Carolina............ .. ........................... A62
A28-A30. Graphs showing-
A28. Measured water-level data from the Camp Lejeune SCADA system for
controlling elevated storage tank STT-40, September 22-0ctober 12, 2004,
U.S. Marine Corps Base Camp Lejeune, North Carolina. .. ........... A64
A29. Calibrated and measured diurnal pattern 124 hours) of delivered finished water
during field test, September 22-0ctober 12, 2004, Tarawa Terrace water-distribution
system, U.S. Marine Corps Base Camp Lejeune, North Carolina.......... .. ........... A65
A30. Measured and simulated fluoride concentrations at four monitoring locations
(a) F02, (b) F05, (c) F07, and /d) F09 in the Tarawa Terrace water-distribution system,
September 22-0ctober 12, 2004, U.S. Marine Corps Base Camp Lejeune,
North Carolina........... .. ........... A66
vii
viii
Tables
Al. Summary oftrichloroethylene and tetrachloroethylene study
characteristics and results.................. .. ...... A4
A2. Summary of ATS DR chapter reports on topical subjects of water-modeling
analyses and the historical reconstruction process, Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina ...... A6
A3. Names and synonyms of selected volatile organic compounds
detected in groundwater ..................... A9
A4. Analyses and simulation tools !models) used to reconstruct historical contamination
events at Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune,
North Carolina........................ .. ........... A 12
A5. Computed volume and mass of tetrachloroethylene in the unsaturated and
saturated zones, Tarawa Terrace and vicinity, U.S. Marine Corps Base
Camp Lejeune, North Carolina ..... .. .................. A15
A6. Historic al operations for water-supply wells, 1952-1987, Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina ........... A 19
A7. Estimated monthly delivery of finished water to the Tarawa Terrace water-distribution
system, 2000-2004, U.S. Marine Corps Base Camp Lejeune, North Carolina.. .. ...... A21
AS. Summary of calibration targets and resulting calibration statistics for simulation models
used to reconstruct historical contamination events at Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina.. .. ... A26
A9. Summary of model-derived values and observed data of tetrachloroethylene at
water-supply wells, Tarawa Terrace, U.S. Marine Corps Base Camp Lejeune,
North Carolina.................... .. .......... A27
A10. Summary of model-derived values and observed data oftetrachloroethylene at
the water treatment plant, Tarawa Terrace, U.S. Marine Corps Base Camp Lejeune,
North Carolina....... .. .... A28
Al 1. Calibrated model parameter values used for simulating groundwater flow and
contaminant fate and transport, Tarawa Terrace and vicinity, U.S. Marine Corps Base
Camp Lejeune, North Carolina ..................... A29
A12. Summary statistics for simulated tetrachloroethylene contamination of selected
water-supply wells and the water treatment plant based on calibrated model
simulation, Tarawa Terrace, U.S. Marine Corps Base Camp Lejeune, North Carolina .......... A39
A13. Summary statistics for simulated tetrachloroethylene and degradation by-product
contamination of selected water-supply wells and the water treatment plant based
on three-dimensional multispecies and multiphase model simulation, Tarawa Terrace,
U.S. Marine Corps Base Camp Lejeune, North Carolina..... .. ...... A44
A14. Summary of selected sensitivity analyses conducted on calibrated groundwater-flow
and contaminant fate and transport model parameters, Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina................... .. ...... A51
A15. Model parameters subjected to probabilistic analysis, Tarawa Terrace and vicinity,
U.S. Marine Corp Base Camp Lejeune, North Carolina..... .. ................... A54
A16. Description of locations equipped with continuous recording water-quality
monitoring equipment used to conduct a fluoride tracer test of the Tarawa Terrace
and Holcomb Boulevard water-distribution systems, September 22-Dctober 12, 2004,
U.S. Marine Corps Base Camp Lejeune, North Carolina .............................. .. .. .... A63
I
I
Conversion Factors
Multiply
im:h
foot (fl)
mile (mi)
gallon (gal)
gallon (gal)
million gallons (MG)
fool per day (fl/<l)
million gallons per day (MGD)
inch per year (in/yr)
fool per day (fl/d)
By
Length
2.54
0 3048
1.609
Volume
3.7X5
0.0113785
3.785
Flow rate
0.31148
0.114381
25.4
Hydraulic conductivity
11.3048
Concentration Conversion Factors
Unit
microgram per liter
(11g/L)
microgram per liter
(11g/L)
microgram per lita
(µg/L)
parts per billion by volume
(ppbv)
To convert to
milligram per liter
(mg/L)
milligram per cubic meter
(mg/m·1)
microgram per cubic meter
(11g/m.l)
parts per million hy volume
(ppmv)
To obtain
centimeter (cm)
meter (rn)
kilometer (km)
liter (Li
cubic meter (m·\)
cubic meter (m\)
meter per day (m/d)
cubic meter per second (mJ/s)
millimeter per year (mm/yr)
meter per day (rn/d)
Multiply by
0.001
1.11011
1.000
Vertical coordinate information is referenced to the National Geodetic Vertical Datum of 1929 INGVD 29).
Horizontal coordinate information is referenced to the North American Datum of 1983 (NAO 83).
Altitude, as used in this report, refers to distance above the vertical datum.
ix
X
Glossary and Abbreviations
Ddiniti{)llS of terms and ahhreviati(rns used thniugh(HII
this report arc listed below.
A
aerobic conditions Conditions for growth or metabolism
in which the organism is sufficiently supplied with oxygen
(IUPAC 2006)
anaerobic process A biologically-mediated process or
condition notrequiring molecular or free oxygen (IUPAC 2006)
ATSDR Agency for Toxic Substances and Disease Registry
B
biodegradation Transformation of substances into new
compounds through biochemical reactions or the actions
of microorganisms, such as bacteria. Typically expressed
in terms of a rate constant or half-life (USEPA 2004). The
new compounds are referred to as degradation by-products
(for example, TCE, 1,2-tDCE, and VC are degradation
by-products of PCE)
BTEX Benzene, toluene, ethyl benzene, and xylene; a group
of VO Cs found in petroleum hydrocarbons, such as gasoline,
and other common environmental contaminants
C
calibration See model calibration
CERCLA The Comprehensive Environmental Response, Com-
pensation, and Liability Act of 1980, also know as Superfund
CRWOME Continuous recording water-quality monitoring
equipment; equipment that can be connected to hydraulic
devices such as hydrants to continuously record water-
quality parameters such as temperature, pH, and fluoride. For
the Camp Lejeune analyses, the Horiba W-23XD continuous
recording, dual probe ion detector data logger was used
D
DCE 1,1-dichloroethylene or 1,1-dichloroethene
1,2-DCE cis-1,2-dichloroethylene or trans-1,2-dichloroethylene
1,2-cDCE cis-1,2-die hloroethylene or cis-1,2-dichloroethene
1,2-tDCE trans-1,2-dichloroethylene or trans-1,2-dichloroethene
degradation See biodegradation
degradation by-product See biodegradation
density The mass per unit volume of material, expressed in
terms of kilograms per cubic meter or grams per cubic centimeter
direct measurement or observation A method of obtaining
data that is based on measuring or observation of the param-
eter of interest
diurnal pattern The temporal variations in water usage
for a water system that typically follow a 24-hour cycle
(Haestad Methods et al. 2003)
DNAPL Dense nonaqueous phase liquids; a class of envi-
ronmental contaminants that have a specific gravity greater
than water (Huling and Weaver 1991 ). Immiscible (nonmixing)
DNAPLs exit in the subsurface as a separate fluid phase in
the presence of air and water. DNAPLs can vaporize into air
and slowly dissolve into flowing groundwater. Examples of
DNAPLs include chlorinated solvents, creosote, coal tar,
and PCBs (Kueper et al. 2003)
DVD Digital video disc
E
EPANET 2 A water-distribution system model developed
by US EPA
epidemiological study A study to determine whether a
relation exists between the occurrence and frequency of
a disease and a specific factor such as exposure to a toxic
compound found in the environment
EPS Extended period simulation; a simulation method used
to analyze a water-distribution system that is characterized
by time-varying demand and operating conditions
exposure Pollutants or contaminants that come in contact
with the body and present a potential health threat
F
fate and transport Also known as mass transport; a process
that refers to how contaminants move through, and are trans-
formed in, the environment
finished water Groundwater that has undergone treatment
at a water treatment plant and is delivered to a person's home.
For this study, the concentration of treated water at the water
treatment plant is considered the same as the concentration
of water delivered to a person's home
ft Foot or feet
G
gal Gallon or gallons
gal/min Gallons per minute
H
historical reconstruction A diagnostic analysis used to
examine historical characteristics of groundwater flow,
contaminant fate and transport, water-distribution systems,
and exposure
interconnection The continuous flow of water in a pipeline
from one water-distribution system to another
inverse distance weighting A process of assigning values
to unknown points by using values from known points; a
method used to contour data or simulation results
IUPAC International Union of Pure and Applied Chemistry
K
K" Organic carbon partition coefficient
K,. Octanol-water partition coefficient
M
MCL Maximum contaminant level; a legal threshold limit set
by the US EPA on the amount of a hazardous substance that
is allowed in drinking water under the Safe Drinking Water
Act; usually expressed as a concentration in milligrams or
micrograms per liter. Effective dates for MC Ls are as follows:
trichloroethylene (TCE) and vinyl chloride (VG), January 9, 1989;
tetra c hloroethylene (PCE) and trans-1,2-dichloroethylene
(1,2-tDCE), July 6, 1992 (40 CFR, Section 141.60, Effective
Dates, July 1, 2002, ed.)
MCS Monte Carlo simulation; see Monte Carlo analysis
MESL Multimedia Environmental Simulations Labora-
tory, School of Civil and Environmental Engineering,
Georgia Institute of Technology, Atlanta, Georgia; an
ATSDR cooperative agreement partner
pg/L Microgram per liter; 1 part per billion, a unit
of concentration
MG Million gallons
MGD Million gallons per day
mg/L Milligram per liter; 1 part per million (ppm), a unit
of concentration
ml Milliliter; 1/1000th of a liter
model calibration The process of adjusting model input pa-
rameter values until reasonable agreement is achieved between
model-predicted outputs or behavior and field observations
M0DFL0W-96 A three-dimensional groundwater-flow model,
1996 version, developed by the U.S. Geological Survey
M0DFL0W-2K A three-dimensional groundwater-flow model,
2000 version, developed by the U.S. Geological Survey
Monte Carlo analysis Also referred to as Monte Carlo simula-
tion; a computer-based method of analysis that uses statistical
sampling techniques to obtain a probabilistic approximation to
the solution of a mathematical equation or model (USEPA 1997)
xi
MTJDMS A three-dimensional mass transport, multispecies
model developed by C. Zheng and P. Wang on behalf of the
U.S. Army Engineer Research and Development Center in
Vicksburg, Mississippi
N
NPL National Priorities List; the USEPA's official list of
uncontrolled hazardous waste sites which are to be cleaned
up under the Superfund legislation
p
paired data point A location with observed data (for example,
water level or concentration) that is associated with a model loca-
tion for the purpose of comparing observed data with model results
PCE Tetrachloroethene, tetrachloroethylene, 1, 1,2,2-tetrachloro-
ethylene, or perchloroethylene; also known as PERC® or PERK®
PDF Probability density function; also known as the
probability function or the frequency function. A mathematical
function that expresses the probability of a random variable
falling within some interval
PHA Public health assessment; an evaluation conducted by
ATS DR of data and information on the release of hazardous
substances into the environment in order to assess any past,
present, or future impact on public health
potentiometric level A level to which water will rise in a
tightly cased well
potentiometric surface An imaginary surface defined by
the levels to which water will rise in a tightly cased wells.
The water table is a particular potentiometric surface
probabilistic analysis An analysis in which frequency (or
probability) distributions are assigned to represent variability
(or uncertainty) in quantities. The output of a probabilistic
analysis is a distribution (Cullen and Frey 1999)
pseudo-random number generator A deterministic algorithm
used to generate a sequence of numbers with little or no discern-
able pattern in the numbers except for broad statistical properties
PS0pS A pumping schedule optimization system simulation
tool used to assess impacts of unknown and uncertain histori-
cal groundwater well operations. The simulation tool was
developed by the Multimedia Environmental Simulations Labo-
ratory at the Georgia Institute of Technology, Atlanta, Georgia
Q
qualitative description A method of estimating data
that is based on inference
quantitative estimate A method of estimating data that
is based on the application of computational techniques
XII
R
rank-and-assign method An optimization method uniquely
developed for the pumping schedule optimization system
IPSOpS) simulation tool. This procedure updates the pumping
schedule for maximum and minimum contaminant concentra-
tion levels in finished water of the WTP based on derivative,
pumping capacity, and total pumping demand information
RMS Root-mean-square; a statistical measure of the
magnitude of a varying quantity
s
saturated zone Zone at or below the water table
SCADA Supervisory control and data acquisition; a comput-
erized data collection system used to collect hydraulic data
and information in water-distribution systems at specified
time intervals such as every 1, 5, 15, etc., minutes
sensitivity analysis An analysis method used to ascertain
how a given model output !for example, concentration) de-
pends upon the input parameters !for example, pumping rate,
mass loading rate). Sensitivity analysis is an important method
for checking the quality of a given model, as well as a powerful
tool for checking the robustness and reliability of its analysis
sequential biodegradation Degradation of a volatile organic
compound as a result of a biological process that occurs
in a progression, for example, the biodegradation of
PCE-TCE-1,2-tDCE-VC
SGA Small for gestational age; a term used to describe when
an infant's weight is very low given their gestational week of birth
SGS Sequential Gaussian simulation; a process in which
a field of values !such as hydraulic conductivity) is obtained
multiple times assuming the spatially interpolated values
follow a Gaussian !normal) distribution
skeletonization The reduction or aggregation of a water-
distribution system network so that only the major hydraulic
characteristics need be represented by a model. Skeletoniza-
tion is often used to reduce the computational requirements
of modeling an all-pipes network
SR Highway or state route
standard deviation Square root of the variance or the root-
mean-square IRMS) deviation of values from their arithmetic mean
T
TCE 1, 1,2-trichloroethene, or 1, 1,2-trichloroethylene,
or trichloroethylene
TechFlowMP A three-dimensional multispecies, multiphase
mass transport model developed by the Multimedia Environ-
mental Simulations Laboratory at the Georgia Institute of
Technology, Atlanta, Georgia
trihalomethane A chemical compound in which three of the
four hydrogen atoms of methane ICH,) are replaced by halogen
atoms. Many trihalomethanes are used in industry as solvents
or refrigerants. They also are environmental pollutants, and
many are considered carcinogenic
u
uncertainty The lack of knowledge about specific factors,
parameters, or models !for example, one is uncertain about
the mean value of the concentration of PCE atthe source)
unsaturated zone Zone or area above the water table;
also known as the vadose zone
USEPA U.S. Environmental Protection Agency
USGS U.S. Geological Survey
V
variability Observed differences attributable to heterogeneity
or diversity in a model parameter, an exposure parameter, or
a population
VC Vinyl chloride or chloroethene
Venn diagram A diagram that shows the mathematical or
logical relationship between different groups or sets; the dia-
gram shows all the possible logical relations between the sets
venturi meter A device used to measure the flow rate or
velocity of a fluid through a pipe
VOC Volatile organic compound; an organic chemical
compound I chlorinated solvent) that has a high enough vapor
pressure under normal circumstances to significantly vaporize
and enter the atmosphere. VO Cs are considered environmental
pollutants and some may be carcinogenic
w
water-distribution system A water-conveyance network
consisting of hydraulic facilities such as wells, reservoirs,
storage tanks, high-service and booster pumps, and a network
of pipelines for delivering drinking water
water table Also known as the phreatic surface; the surface
where the water pressure is equal to atmospheric pressure
WTP Water treatment plant
Use of trade names and co111111crcial sources is for idcntilication only and docs
not imply endorsement hy the Agency for Toxic Substances and Disease Registry or
the U.S. Department of Health and Human Services.
Analyses of Groundwater Flow, Contaminant Fate and Transport,
and Distribution of Drinking Water at Tarawa Terrace and Vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina:
Historical Reconstruction and Present-Day Conditions
Chapter A: Summary of Findings
By Morris L. Maslia,1 Jason B. Sautner,' Robert E. Faye,' Rene J. Suarez-Soto,' Mustafa M. Aral,3
Walter M. Grayman,' Wonyong Jang,3 Jinjun Wang,3 Frank J. Bove,1 Perri Z. Ruckart,1
Claudia Valenzuela,5 Joseph W. Green, Jr.,5 and Amy L. Krueger 5
Abstract
Two of three watcr-distrihution systems that have
historically supplied drinking water to family housing at
U.S. Marine Corps Base Camp Lejeune. North Carolina.
were contaminated with volatile organic compounds
(VOCs). Tarawa Terrace was contaminated mostly with
tctrachloroethylene (PCE). and Hadnot Point was con-
taminated mostly with trichloroethylenc (TCE). Because
scientific data relating lo the harmful effects of voes Oil
a child or fetus arc limited. the Agency for Toxic Sub-
stances and Disease Registry (ATSDR). an agency of
the U.S. Department of Health and Human Services. is
conducting an epidemiological study to evaluate potential
associations between in utcro and infant (up to I year of
age) exposures to VOCs in contaminated drinking water
at Camp Lejeune and specilic birth defects and childhood
cancers. The study includes births occurring during the
period 1968-1985 to women who were pregnant while
they resided in family housing at Camp Lejeune. Because
1 Agency for Toxic Std1\tanccs and Discasc Rcgi~try. Atlanta. Georgia.
:-Ruhn\ E. Faye and As~ociaies, consull:1111 10 Easti.:rn Re~carch Group.
Inc.. Lexing1011. Massachusell~.
-' School (lf Civil and Envinmmentil l~ngineering. Georgia ln\li!Ute of
Technology, Atlanta. Georgia.
1 W.M. Graym:111 Consulting l:ngineer. Ci11cinn:1ti. Ohio.
·1 Oak Ridge Institute rm Science and Educa1io11. Oak Ridge. Tennessee.
Chapter A: Summary of Findings
limited measurcmenls of contaminant and exposure
data are available to support the epidemiological study.
ATSDR is using modeling techniques lo reconstruct
historical conditions of groundwater llow. contaminant
fate and transport, and the distribution of drinking water
contaminated with VOCs delivered lo family housing
areas. The analyses and results presented in this Sum-
mary of Findings. and in reports described herein. refer
solely to Tarawa Terrace and vicinity. Future analyses.
and reports will present information and data about con-
tamination of the Hadnot Point water-distribution system.
Models and methods used as part of the historical
reconstruction process for Tarawa Terrace and vicinity
included: (I) MODFLOW-96. used for simulating steady-
state (predcvclopment) and transient groundwater flow:
(2) MT3DMS. used for simulating three-dimensional.
single-specie contaminant fate and transport: (3) a
materials mass balance model (simple mixing) used to
compute the flow-weighted average concentration of
PCE assigned to the linished waler at the Tarawa Terrace
water treatment plant (WTP); (4) TechFlowMP, used for
simulating three-dimensional. multispccies. multiphase
mass transport; (5) PSOpS. used for simulating the
impacts of unknown and uncertain historical well opera-
tions: (6) Monte Carlo simulation and sequential Gauss-
ian simulation used to conduct probabilistic analyses
lo assess uncertainty and variability of concentrations
A1
Introduction-------------------------------------------
of PCE-contaminatcd groundwater and drinking water:
and (7) EPANET 2. used to conduct extended-period
hydraulic and water-quality simulations of the Tarawa
Terrace water-distribution system. Through historical
reconstruction. monthly concentrations of PCE in
groundwater and in finished water distributed from
the Tarawa Terrace WTP to residents or Tarawa Terrace
were determined.
Based on field data. modeling results. and the his-
torical reconstruction process. the following conclusions
arc made:
Simulated PCE concentrations exceeded the current
maximum contaminant level (MCL) of 5 micro-
grams per liter (1tg/L) al water-supply well TT-26
for 333 months-January 1957-.lanuary 1985.
• The maximum simulated PCE concentration at
well TT-26 was 85 I ttg/L during July I 984;
the maximum measured PCE concentration
was 1.580 ttg/L during January I 985.
• Simulated PCE concentrations exceeded the
current MCL of 5 11g/L in linishcd waler al
the Tarawa Terrace WTP for 346 months-
November I 957-February I 987.
The maximum simulalcd PCE conccnlralion in
finished water al the Tarawa Terrace WTP was
183 ftg/L during March I 984: the maximum
measured PCE conccnlralion was 215 pg/L
during February I 985.
• Simulation of PCE degradation by-products-TCE.
1ra11s-I .2-dichloroethylene ( 1.2-tDCE). and vinyl
chloridc-imlicalcd that maximum concentrations
of the degradation by-products generally were in the
range of 10-1001tg/L al water-supply well TT-26:
measured concentrations ofTCE and 1.2-tDCE on
January I 6. I 985. were 57 and 92 ttg/L. respectively.
Maximum concentrations of degradation by-
products in linished water al the Tarawa Terrace
WTP generally were in the range of2-15 ftg/L;
measured concentrations of TCE and 1.2-tDCE on
February 11, I 985. were 8 and 12 1tg/L. respectively.
Based on waler-supply well scheduling analyses,
linished waler exceeding the current MCL for
PCE (5 11g/L) al the Tarawa Terrace WTP could
have heen delivered as early as December I 956
and no later than June I 960.
• Based on probabilistic analyses. the most likely
dates that finished water first exceeded the current
MCL for PCE ranged from October I 957 lo
August 1958 (95 percent probability). with an
average first exceedance date of November I 957.
• Exposure to drinking water contaminated with
PCE and PCE degradation by-products ceased
after February 1987 when the Tarawa Terrace
WTP was closed.
Introduction
The Agency for Toxic Substances and Disease
Registry (ATSDR), an agency of the U.S. Depart men I
of Health and Human Services, is conducting an epide-
miological study to evaluate whether in utero and infatll
(up to I year of age) exposures to drinking water con-
taminated with volatile organic compounds (VOCs) al
U.S. Marine Corps Base Camp Lejeune. North Carolina
(Plate I), were associated with speci fie birth defects and
childhood cancers. The study includes births occurring
during the period 1968-1985 to women who resided in
family housing at Camp Lejeune. The first year of the
study, 1968. was chosen because North Carolina com-
puterized its birth certificates starting that year. The last
year of the study, 1985. was chosen because the most
contaminated water-supply wells were removed from
regular service that year. ATS DR is using water-modeling
techniques to provide the epidemiological study with
quantitative estimates of monthly contaminant concentra-
tions in finished drinking water6 because contaminant
concentration data and exposure information arc limited.
Results obtained by using water-modeling techniques.
along with information from the mother on her water use.
can be used by the epidemiological study 10 estimate the
level and duration of exposures to the mother during her
pregnancy and to the infatll (up 10 I year of age). Using
water-modeling techniques in such a process is referred
to as historical reconstruction (Maslia el al. 200 I).
Three water-distribution systems have historically
supplied drinking water to family housing at U.S. Marine
Corps Base Camp Lejeune-Tarawa Terrace. Holcomh
Boulevard. and Hadnot Point (Plate I. Figure A I).
" For this study. finished drinking water is defined as groundwater that has
undergone treatment at a water treatment plant and is delivered to a person's
home. The concentration of contaminants in treated waler al thi.: water trcalmcnl
plant is considered the same a.'i the concenlralions in the water deliven.:d to a
person's home. This a-;st1mption is tested and vcrilied in the Chapter J report
(Sautner ct aL In press 2007). llcreaftcr, the term "finished waler" will he used.
A2 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
------------------------------------------Introduction
NORTH CAROLINA
~,w-,i=·,1
~'&~:ti; c5{'\l>:k-I :s.~><y: -
34°45'
34°43'30"
Montford
Point
Note; Camp Knox served
by Montford Point end
Tarawa Terrace water
supplies et various
historical times
<..,,-J
77°24'
Paradise
Point
Base from U.S. Marine Corps and
U.S. Geological Survey digital data files
0 5 MILES
1-l------.-1-1
0 5 KILOMETERS
Base from Camp Lejeune GIS Ottice, June 2003
77°22'30" 11°2r
0.5 1 MILE
EXPLANATION 0.5 1 KILOMETER
Historical water-supply area
D Montford Point ~ Tarawa Terrace D Holcomb Boulevard D Hadnot Point
n -26• Water-supply well and identification
Figure A1. Selected base housing and historical water-supply areas, Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina.
Chapter A: Summary of Findings A3
Introduction -------------------------------------------
Two of the water-distribution systems were contami-
nated w ith VOCs. Tarawa Terrace was contaminated
mostly w ith tetrachloroethy lene (PCE), and Hadnot
Point was contaminated mostly with trichloroethylene
(TCE). Historical information and data have indicated
that one source of contamination-ABC One-Hour
Cleaners (Figure A I )-was responsible for contaminat-
ing Tarawa Terrace water-supply wells (Shiver 1985).
Water-supply data and operational information indicate
that Tarawa Terrace wells supplied water solely to the
Tarawa Terrace water treatment plant (WTP). Addi-
tionally. the Tarawa Terrace water-distribution system
was operated independently of the other two water-
distribution systems (Holcomb Boulevard and Hadnot
Point). Therefore, analyses presented in this Summary
of Findings and in reports described herein refer solely
to Tarawa Terrace and vicinity. Future analyses and
reports w ill present information and data about contami-
nation of the Hadnot Point water-distribution system.
Previous Studies and Purpose of the
Current Investigation
Only a small number of studies have evaluated the
risk of birth defects and ch ildhood cancers from expo-
sures to drinking water contaminated with VOCs. These
include, for example, studies by Cohn cl al. ( 1994).
Bove et al. ( 1995. 2002), Costas et al. (2002), Massa-
chusetts Department of Public Health ( 1996). and the
New Jersey Department of Health and Senior Services
(2003). Five stud ies that have evaluated exposures
to TCE and PCE in drinking water and adverse birth
outcomes are summarized in Table A 1. Compared to
Table A1. Summary of trichloroethylene and tetrachloroethylene study characteristics and results.1
!OR, odd, ratio: TCE. 1rid1loroethylcnc: PCE. tetrachlorocthylcnc: SGA. ,mall for gc,tational age: LBW. low hirth weight: NTD. neural tuhe defect,:
MBW. mean birth "eight: MBWD. mean birth weight difference: VLBW. very low hinh weight: GIS. geogrnphk information ,y,tcm: =. e4ual: s. le" than
or equal to:-. negative: g. gram: yr. year I
Study site and period
Arizona
1969-1981
(Goldberg ct al. 1990)
Woburn, Massachusetts
1975-1979
(MDPH,CDC 1996)
1969-1979
orthern New Jersey
1985-1988
(Bove et al. 1995)
Camp Lejeune,
North Carolina
1968-1985
(ATSDR 1998)
Ari1.011a 1979-198 1
(high exposure)
and 1983-1985
(post exposure)
(Rodenbeck
Ct al. 2000)
'Bovl! ct al. (2002)
Outcome Number of subjects
Cardiac defect~ 365 ca~e~
SGA
preterm birth
birth defects
fetal death
LBW
SGA
preterm birth
birth dcfcc1~
fetal death
MBW
SGA
preterm birth
2,21 I births
19 fetal deaths
5,347 birth
80.938 live birth,.
594 fetal dea1h,
31 births expo ed to
TCE, 997 unex-
posed; 6,11 7 births
exposed to PCE.
5,681 births
unexposed
Exposure Results (OR)2
I'' trimester residence Prevalence ratio= 2.58
(or employment)
in area of TCE
contamination
Modeled distribution
system to estimate
monthly exposures;
address at deli very
Estimated average
monthly levels of
sol vent, ba,cd on
tap waler sample
data and addrc,s a1
delivery
Residence in a base
housing area known
to have received
contaminated water
SGA = 1.55; LBW s 1.0; preterm
delivery s 1.0: fetal death = 2.57;
NTD = 2.21; cleft palate= 2.21;
heart defects= 0.40; eye defects=
4.41; cluster of choanal atresia
TCE: SGA s 1.0: pretcrm birth = 1.02:
NTD = 2.53: oral clefl s = 2.24:
heart defect, = 1.24: fetal death s 1.0
PCE: SGA s 1.0: pretcrm birth s 1.0:
NTD = 1.16: oral clef!\= 3.54:
hcat1 defects = 1.13: fetal death s 1.0
TCE: SGA = 1.5; MBWD =-139 g;
preterm birth = 0.0;
males: SGA = 3.9; MBWD = -3 12 g
PCE: SGA = 1.2; MBWD = -24 g;
preterm birth = 1.0; women > 35 yr:
SGA = 4.0; MBWD = -205 g:
women with 2:2 fetal losses: SGA = 2.5
LBW 1.099 exposed births. Maternal residence in TCE: LBW= 0.90: VLBW = 3.30:
VLBW 877 unexposed target or cornpari-full-term LBW = 0.81
full-term LBW births ,on censu5 1racts al
delivery: GIS mod-
eling of ground-
water plume
'Result, in bold 1ypc indicate tho,e that were calculated by the reviewing au1hor, ( Bove cl al. 2002)
A4 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-------------------------------------------Introduction
the aforemcnlioned studies. ihc current study al Camp
Lejeune is unique in that it will examine the associa-
lions beiwccn well-defined. quanlitative levels of PCE
and TCE in drinking water and the risk of developing
specific birth dcfects-spina bilida. anencephaly. cleft
lip, and cleft pala1e--childhood leukemia. and non-
Hodgkin's lymphoma. The currenl study includes parent
interviews conducted to obtain residential history, infor-
mation on water consumption habits, and risk factors.
Using model-derived drinking-water concentrations
and interview data. associations between exposure to
PCE and TCE during various time periods of intcrcst-
preconceplion. trimesters, cnlire pregnancy, and infancy
(up to I year of agc)-and the risk of particular health
outcomes can be thoroughly examined.
The purpose of the analyses described in this report
and associated chapter reports is to provide epidemi-
ologists wilh historical monthly concentrations of con-
taminants in drinking water to facilitate the estimation
of exposures. Because historical contaminant C(Hlccn-
tration data arc limited. the process of historical recon-
struction-which included water-modeling analyses-
was used to synthesize information and quantify
estimates or contaminant occurrences in groundwater
and the water-distribution system al Tarawa Terrace.
Tarawa Terrace Chapter Reports
Owing to the complexity. uniqueness, and the
number of topical subjects included in the hislorical
reconstruction process. a number of reports were
prepared that provide comprehensive descriptions of
information. data, and methods used to conduct historical
and present-day analyses at Tarawa Terrace and vicinity.
Table A2 lists the 11 chapters (A-K) and chapter lilies
of reports thal compose the complete description and
details of ihc historical reconstruction process used for
the Tarawa Terrace analyses. Also included in Table A2
are listings of the authors and a topical summary of each
chapter report. Figure A2 shows the relation among the
Chapter A report (Summary of Findings-this report).
Chapters B-K reports. and the overall process of histori-
cal reconstruction as it relates to quantifying exposures
and the ATSDR case-control epidemiological study.
Reports for chapters B-K present detailed information,
data, and analyses. Summaries or results from each
chapter report arc provided in Appendix A I. Readers
interested in details of a specific topic, for example.
Chapter A: Summary of Findings
numerical model development, model-calibration pro-
cedures. synoptic maps showing groundwater migration
of PCE at Tarawa Terrace. or probabilislic analyses.
should consult the appropriate chapter report (Table A2.
Appendix A I). Also provided with the Chapter A report
is a searchable eleclronic database-on digital video disc
(DVD) format-of information and data sources used to
conduct the historical reconstruction analysis. Electronic
versions of each chapter report-summarized in Appen-
dix A I-and supporting information and data will be
made available on the ATSDR Camp Lejeune Web site
at Ir t t /J :/lwwiv. a tsd 1: cdc. gr J\1/s it es/I eje une/index. h tm I.
External Peer Review
Throughout this investigation. ATSDR has sought
independent external expert scientific input and review or
project methods, approaches, and interprclations to assure
scienlilic credibility or the analyses described in the
Tarawa Terrace reports. The review process has included
convening an expert peer review panel and submitting
individual chapter reports 10 outside experts for technical
reviews. On March 28-29, 2005, ATSDR convened an
exlernal expert panel to review the approach used in con-
ducting the historical reconstruction analysis and to provide
input and recommendations on preliminary analyses and
modeling results (Maslia 2005). The panel was composed
of experts with professional backgrounds from govern-
ment and academia, as well as the privale sector. Areas
or expertise included numerical model development and
simulation. groundwater-flow and contaminant fate and
transport analyses and model calibration. hydraulic and
water-quality analysis of water-distribution systems.
epidemiology. and puhlic healih. After reviewing data
and initial approaches and analyses provided hy ATSDR,
panel members made the following rec<)mmcn<lations:
Oma discoverv: ATSDR should expend additional
effort and resources in the area of conducting
more rigorous data discovery activities. To the
extent possible, the agency should augment,
enhance, and refine data it is relying on to
conduct water-modeling activities.
• Chmnologr of' events: ATS DR should focus efforts
on refining its understanding of chronological
events. These need to include documenting periods
of known contamination. times when water-distri-
bution systems were interconnected, and the start
of operations of the Holcomb Boulevard WTP.
A5
Introduction-------------------------------------------
Table A2. Summary of ATS DR chapter reports on topical subjects of water-modeling analyses and the historical reconstruction
process, Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina.
I ATSDR, Agency l'or Toxic Substances and Disease Kegistry: VOC. volatile organic compound: l'CE. tctrachloroethyknc: WTP. water treat men! plant]
A
ll
C
I)
E
F
c;
H
K
A6
Mas!ia ML. Sautner JB.
Faye RE, Suarez-Soto R.l.
Aral MM. Cirayman WM.
Jang W, Wang .I. Bove F.I.
Ruckart PZ. Valenzuela C.
Green JW Jr. and
Krueger AL
Faye RE
Faye RE. and Valenzuela C
Lawrence SJ
Faye RE. and Circcn JW Jr
Faye RE
Jang W. and Aral MM
Wang J. and Aral MM
tvlaslia l'v1L. Sui'ircz-Soto RJ.
Wang J. Aral MM.
Sautner JB. and
Valenzuela C
Sautner JB, Valenzuela C.
Maslia ML. and
Grayman WM
ivlaslia ML. Sautner JB.
Faye RE. Su;ircz-Soto RJ.
Aral MM. Grayman WM.
Jang W. W:rng J. Rove F.I.
Ruckan l'Z. Valenzuela C.
Green JW Jr, and
Krueger AL
rl''\:litt~·:rK,t~~'?Jj]~lij;mi.'l!.fii t( ;F~.aP._ 1er!AV~~1an~(_re1~,ren_, ~~~}ita1i;1_ , ,. , ,:®ffl, , , t•a"'ffi!lR,i'fullfflm1,
Summary of Findings: Maslia et al. 2007
(this report)
Gcohydrologic Framework of the
Castle Hayne Aquifer System:
Faye (In press 2007a)
Simulation of Groundwater Flow:
Faye and Valenwela (In press 2007)
Pniperties of Dcgradati,in Pathways or
Common Organic Compounds in
Groundwater: Lawrence (In press 2007)
Occurrence of Contaminants in Ground-
water: Faye and Green (In press 2007)
Simulation of the Fate and Transporl
of Tetrachloroethylcnc (PCE):
Faye (In press 2007b)
Simulation of Three-Dimcnsi1mal Multi-
specics. Multiphase Mass Transport {if
Tctrachloniethylcnc (PCE) and Associ-
ated Degradation By-Products: Jang and
Aral (In press 2007)
Effect of Groundwater Pumping Schedule
Variation on Arrival of Tetrachloroethyl-
cne (PCE) at Water-Supply Wells and the
Water Treatment Plant; Wang and Aral
(In press 2007)
Parameter Sensitivity. Unccnaimy. and Vari-
ability Associated with Model Simulations
of Groundwater Flow. Contaminant Fate
and Transport. and Distrihution of Drink-
ing Water: Maslia ct a!. (In press 2007h)
Field Tests. Data Analyses. and Simulation
of the Distribution of Drinking Water:
Sautner et al. (In press 2007)
Supplemental Information: Maslia et al.
(In press 2007a)
Summary of dc!:1iled technical findings
(found in Chapters B-K) focusing 011 the
historical reconstruction analysis and present-
day conditions of groundwater flow. contami-
nant fate and transport. and distrihution of
drinking water
Analyses of well and geohydrologie data used
lO develop the gcohydrologic framework of
the Castle Hayne aquifer system at Tarawa
Terrace and vicin~ty
Analyses of gnlundwater flow including devel-
oping a predevclopment (steady state) and
transient grouml\.~atcr-flow model
Describes and summarizes the properties. degra-
dation pathways. and degradation hy-products
of VOCs (non-trihalomethane) commonly
detected in groun~water
Describes the occurrence and distribution of
PCE and rdak.:d contaminants within the
Tarawa Terrace aquifer and the Upper Castle
Hayne aquircr system at and in the vicinity of
the Tarawa Tcrrac.e housing area
Historical reconstruction of the fate and transport
of PCE in groundwater from the vicinity of
ABC One-Hour Cleaners to individual water-
supply wells and the Tarawa Terrace WTP
Descriptions ahout the development ,ind :1pplica-
tion of a model capahlc of simulating threc-
dimensional. multispecies. and multiphase
transport of PCE and associated degradation
by-product:-.
Analysis of the effect of groundwater pumping
schedule variation on the arrival of PCE at
water-supply wells and the Tarawa Terrace WTP
Assessment of parameter sensitivity. uncertainty.
and variability associated with model simula-
tions of groundwater now. contaminant fate and
transport and the distribution of drinking w:1ter
Field tests, data analyses. and simulation of the
distribution of drinking water at Tarawa Ter-
race and vicinity
Additional information such as synuptic maps
showing groundwater levels. direct ions of
groundwater flow, and the distribution of PCE
based on :-.imulation: a complete list or refer-
ences: and other ancillary information and
data that were used as the basis of this study
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-------------------------------------------Introduction
I<-
'---<-
Chapter A: Summary of Findings
I
I Executive Summary I
, Published June 2007 1
L __________ I
Chapter B Chapter D Chapter E
Geohydrologic Properties and Occurrence
framework degradation pathways of contaminants of voes
Chapter C Chapter F
Simulation of Simulation of PCE
groundwater flow fate and transport
Auxilliary and enhanced analyses
Chaoter G
Simulation of
three-dimensional
multispecles, multiphase
mass transport
Chapter J
Field tests and ---> analyses of water-
distribution system
Chapter K
I<-Supplemental
information and data
Chapter H
Analyses of
water-supply well
pumping variation
l
His•oric I
const ucti n
Quantified concentrations
of PCE-contaminated
drinking water
at Tarawa Terrace
water treatment plant
t
Epidemiological
study
Quantified historical
exposures to PCE
derived from historical
reconstruction and
water-modeling analyses
Chapter I
Parameter and model
sensitivity, uncertainty,
and variability
Figure A2, Relation among Chapter A report (Summary of Findings), Chapters B-K reports, historical
reconstruction process, and the ATSDR epidemiological case-control study, Tarawa Terrace and
vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina. [VOCs, volatile organic compounds;
PCE, tetrachloroethylene]
Chapter A: Summary of Findings
'
A7
Chlorinated Solvents and Volatile Organic Compounds-------------------------
• G1nu11dwa1er modeling, Tr.11-c1111a Terrace area:
Several rcc(lmmcndations were made with respect
to groundwater modeling and associated activities
f<>r the Tarawa Terrace area. and these included:
(I) refine operational schedules of water-supply
wells. (2) conduct fate and dispersive transport
analyses. (3) conduct sensitivity and uncertainty
analyses to reline initial estimates of model
parameter values. and (4) determine sensitivity
of model to cell sizes and boundary conditions.7
Water-distribwion sysrem analyses: In light of
available data. the ATSDR water-modeling team
slloul<l consider using more simplified mixing
models (rather than complex water-distribution
system models) 10 quantify historical exposures to
drinking-water supplies. More complex modeling
might be warranted only if data discovery shows
that the water-distribution systems had a greater
frequency or interconnectivity.
The recommendations of the external expert panel
were implemented as part of the historical reconstruc-
tion analysis efforts. Results of these efforts arc presented
in conjunction with specific data needs. descriptions of
the historical reconstruction simulations. and sensitivity
analyses that are summarized in this report (Chapter A) and
discussed in detail in suhsequenl chapter reports (B-J).
Chlorinated Solvents and
Volatile Organic Compounds (VOCs)
The compoumls am! contaminants discussed in this
report and other Tarawa Terrace chapter reports belong to
a class of chemicals rcl"erred to as chlorinated solvents.
The denser-than-water characteristic of liquid chlorinated
solvents has led to their being called "dense nonaqueous
phase liquids .. (DNAPLs8) (Pankow and Cherry 1996).
The significant volatility that characterizes chlorinated
solvents also has led to these compounds being referred
lo as --volatile organic compounds•• (VOCs). It is the
property of signilicant volatility that has led to the great-
est lack of understanding of their potential for causing
1 Dctailcd di~c11s~io11~ rel:1tcd to spccitic model charactcri~tic~ .\w.:h
<IS gcomctry. cell ~i'le, hound:iry cmtditions, and more. :ire pmvided i11
Chapter C (Faye and V:1lenn1ela In press 2007) and Chapter F (Faye In
press 2007h) report~.
~ Dense nonaqueou~ phase liquid~ (DNAPLs) have a .\pecitic gravi1y
grea\cr than water(> 1.0). and arc immiscible (nonmi:1;ing) in water.
groundwater contamination (Schwille 1988). Thus.
VOCs are organic compounds that have a high enough
vapor pressure under normal circumstances to signifi-
cantly vaporize and enter the atmosphere.
In the United States. the production of chlori-
nated solvents. and more generally. synthetic organic
chemicals. was most probably a direct result of World
War I. As of 1914. PCE was manufactured as a by-
prnduct of carbon tetrachloride, and domestic produc-
tion ofTCE is reported to have begun during the 1920s
(Doherty 2000a. b). Contamination of groundwater
systems by chlorinated solvents. however. was not rec-
ognized in North America until the late l 970s.9 The late-
ness of this recognition was due in part because monitor-
ing for VOCs and nearly all other organic compounds
was not common until that time. Research into the
properties of chlorinated solvents and how their proper-
ties. such as density (DNAPLs) and significant volatility
(VOCs). were capable of leading to severe groundwater
problems was first recognized by Schwille in West
Germany during the 1970s (Schwille 1988). Thus. VOCs
arc considered environmental pollutants. and some may
be carcinogenic. Briclly described next arc naming
conventions used for VOCs and maximum contaminant
levels (MCLs) established by the U.S. Environmental
Protection Agency (USEPA) for selected VOCs.
Naming Conventions
It is common to lind a confusing variety of names
used to identify VOCs. For example. tetrachloroethene
also is known as perchloroethylene. PCE. PERC®. and
tetrachloroethylenc (Tahle A3). The variety of different
names for VOCs depends on ( 1) the hrand name under
which the product is sold. (2) the region where the com-
pound is used. (3) the type of publication referring to
the compound. (4) the popularity of the name in recently
published literature. (5) the profession of the person
using the name, or (6) a combination of all or part of
the above. As early as the late 1800s. chemists and
others recognized the need to have a consistent naming
convention for chemical compounds. The International
Union of Pure and Applied Chemistry (IUPAC) is an
organization responsible for formal naming conventions
'' Contaminants were detected in gnHrndwater sampling hy the New Jersey
Department of Environmental Protection during 1lJ78 (Cohn et al. llJlJ4) and
al Woburn. ivtassachusctts. during t-.fay 1979 (;\Jas~achu~elts Department of
Public Health 1996).
AB Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------------------Chlorinated Solvents and Volatile Organic Compounds
and corresponding names assigned to chemical com-
pounds. Table A3. obtained from Lawrence (2006). lists
the IUPAe names and synonyms (associated common.
alternate. and other possible names) for selected voes
detected in groundwater. The common or alternate
names are used in this and all of the Tarawa Terrace
reports for case of reference to, and recognition of,
previously published reports. documents. and laboratory
analyses that pertain to the Tarawa Terrace area.10
Maximum Contaminant Levels
The maximum contaminant level or MCL is a legal
threshold set by the USEPA to quantify the amount of
a hazardous substance allowed in drinking water under
the Safe Drinking Water Act. For example. the MeL
for PeE was set at S micrograms per liter (11g/L) during
1992 because. given the technology at that time. 5 11g/L
mA dct:iilcd discussion and description of selected \'Olatilc organic
compounds and a .. :-.ociatcd tfrgradation pathways i:-. prc.,cntcd in the
Chaph.:r D report (L,iwn:ncc In pn.:ss 2007).
was the lowest level that water systems could be required
to achieve. Effective dates for MeLs presented in this
report are as follows: TeE and vinyl chloride (VC).
January 9. 1989: PeE and 1ra11s-l ,2-dichloroethylene
(1.2-tDeE). July 6. 1992 (40 eFR. Section 141.60.
Effective Dates. July I. 2002. ed.). In this report and
other Tarawa Terrace chapter reports. the current MeL
for a specific voe-for example. 5 11g/L for PeE-is
used as a reference concentration to compare histori-
cally measured data and computer simulation results.
These comparisons are not intended to imply (I) that
the Mel was in effect at the time of sample measure-
ment or simulated historical time or (2) that a mea-
sured or simulated concentration above an MCL was
necessarily unsafe. Hereafter. the use o"r the term Mel
should be understood to mean the current Mel associ-
ated with a particular contaminant. A complete list of
MeLs for common voes can be found in USEPA report
EPA 8I6-F-03-016 (2003). A complete list of effective
dates for MeLs can be found in 40 eFR. Section 141.60.
Effective Dates. July I. 2002, edition.
Table A3. Names and synonyms of selected volatile organic compounds detected in groundwater.'
[IUl'AC. International Union of Pure :ind Applied Chemistry: CAS, Chemical Ahs1ract Services:-, ll(lt applic;1hlel
I IUPAC name'
hcn,cnc
1.2-diniethylbenzcne
1.3-dimethylbcnzcnc
1.4-dimethylbenzenc
cthylhcnzene
mcthylbenzenc
chl(irocthene
1.1-dichloroethcnc
cis-1.2-dichloroethcne
rm11s-l .2-dichlorocthcnc
tctrnchl(micthenc
1.1.2-trichloroethcne
Common or alternate name
(synonym)'
o-xylene
111-xylene
p-xylene
toluene
vinyl chloride
I, 1-dichloroethylcne, DCE
cis-1,2-dichlorocthylcne
rm11s-l .2-dichloroethylene
pcrchl,iroethylcnc, PCE,
I, 1,2.2-tctrachloroethylcne
1.1,2-trichloroethylene. TCE
1Lawn:nn: (modified from 200(i, In press 2007)
'International Union or l'urc and Applied Chemistry (2006)
-'USEPA ( 1995)
Chapter A: Summary of Findings
Other possible names3
The B in BTEX, coal naptha,
1.3,5-cyclohexatrienc. mineral naptha
The X in BTEX. dimethyltolucne, Xylol
The E in BTEX. Ethylbenzo!. phcnylethane
The Tin BTEX, phenyl methane. Mcthacide.
Toluol. Antisal I A
chl1in1ethylcnc. YC. monochlonlethylene,
monovinyl chloride, MVC
vinylidcnc chloride
! ,2 DCE. Z-1 ,2-dich!orocthenc
1.2 DCE, E-1.2-dichloroethene
ethylene tetrachloride. carbon dichloride,
PERC®. PERK®. tetrachloroethykne
acetylene trichlorocthylenc. trichloroethylenc
CAS number'
71-43-2
95-47-6
IOS-38-3
to6-42-3
100-41-4
I 08-88-3
75-01-4
75-35-4
I 56-59-2
156-60-2
127-18-4
79-01-6
A9
Historical Background------------------------------------
Historical Background
U.S. Marine Corps Base Camp Lejeune is located in
the Coastal Plain of North Carolina, in Onslow County,
southeast or the City or Jacksonville and about 70 miles
northeast or the City or Wilmington, North Carolina
(Figure A I). Operations began at Camp Lejeune during
the 1940s. Today. nearly 150,000 people work and live
on base. including active-duty personnel. dependents.
retirees. and civilian employees. About two-thirds or the
active-duty personnel and their dependents arc less than
25 years of age.
Camp Lejeune consists or 15 different housing
areas; families live in base housing for an average
or 2 years. During the 1970s and 1980s, family
housing areas were served by three water-distribution
systems. all of which used groundwater as the source
for drinking water-Hadnot Point, Tarawa Terrace,
and Holcomb Boulevard (Plate I). Hadnot Point was
the original water-distribution system serving the
entire base with drinking water during the 1940s. The
'forawa 'ICrracc WTP began delivering drinking water
during 1952-1953. and the Holcomb Boulevard WTP
began delivering drinking water during June 1972
(S.A. Brewer, U.S. Marine Corps Base Camp Lejeune,
written communication. September 29, 2005).
The Tarawa Terrace housing area was con-
structed during 1951 and was subdivided into housing
areas I and II (Figure A I). Originally, areas I and II con-
tained a total of 1,846 housing units and accommodated
a resident population of about 6.000 persons (Sheet 3 of
18, Map of Tarawa Terrace II Quarters, June 30, I 961:
Sheet 7 or34. Tarawa Terrace I Quarters. July 31, 1984).
The general area of Tarawa Terrace is bounded on the
cast by Northeast Creek, to the south by New River and
Northeast Creek, to the west by New River, and lo the
north hy North Carolina Highway 24 (SR 24).
The documented onset of pumping at Tarawa
Terrace is unknown but is estimated to have begun
during 1952. Water-supply well Tl~26, located about
900 feet southeast of ABC One-Hour Cleaners (Fig-
ure A I). began operations during 1952. ABC One-Hour
Cleaners-an off-base dry-cleaning facility that used
PCE in the dry-cleaning process (Melts 200 I )-is the
only documented source of PCE contamination of
groundwater resources at Tarawa Terrace (Shiver 1985).
The first occurrence of PCE contamination at a Tarawa
Terrace water-supply well probably occurred at
well Tl~26 after the onset of dry-cleaning operations at
ABC One-Hour Cleaners during 1953.
The Camp Knox trailer park area was constructed
during 1976 with 112 trailer spaces. An additional
75 spaces were added during 1989 allowing for a total
of 187 housing units, which could accommodate a
population of 629 persons (Sheet 5 of 34, Map of Knox
Trailer Park Area. July 31. 1984). The Camp Knox
trailer park area is located in the southwestern part of
the Tarawa Terrace area and is bounded on the south
by Northeast Creek (Figure A I). Camp Johnson and
Montford Point arc located to the west and southwest
of Tarawa Terrace, respectively. Historically. the Camp
Knox trailer park was served by both Tarawa Terrace
and Montford Point water supplies.
During 1989, the US EPA placed U.S. Marine Corps
Base Camp Lejeune and ABC One-Hour Cleaners on its
National Priorities List (NPL) of sites requiring environ-
mental investigation (also known as the list of Superrund
sites). During August 1990. ATSDR conducted a public
health assessment (PHA) at ABC One-Hour Cleaners.
The PHA found that PCE. detected in onsitc and o!Tsitc
wells. was the primary contaminant of concern. Other
detected contaminants included TCE. 1.2-dichloro-
ethylcne ( 1,2-DCE), 1,2-tDCE. I. 1-dichloroethylene
(DCE), VC. benzene. and toluene (ATSDR 1990).
During 1997. ATSDR completed a PHA for Camp
Lejeune which concluded that estimated exposures to
VOCs in drinking water were significantly helow the
levels shown to be of concern in animal studies. Thus.
ATSDR determined that exposure to VOCs in on-base
drinking water was unlikely to result in cancer and
noncancer health effects in adults. However. because
scientific data relating to the harmful effects of VOCs
on a child or a fetus were limited. ATS DR recommended
conducting an epidemiological study to assess the risks
to infants and children during in utero exposure to chlo-
rinated solvents (for example. PCE and TCE) contained
in on-base drinking water (ATS DR 1997).
Following this recommendation, during 1998
ATSDR published a study or adverse birth outcomes
(ATSDR 1998). ATSDR used various databases to evalu-
ate possihlc associations between maternal exposure to
contaminants contained in drinking water on the base
and mean birth weight deficit. preterm birth (less than
37 weeks gestational age), and small for gestational
age (SGA). To identiry women living in base housing
when they delivered. birth certificates were collected
A10 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
' ,,
------------------------------------Water-Distribution Investigation
for live births that occurred January I, 1968-Decem-
ber 31, 1985. The study found that exposure to PCE in
drinking water was related to an elevated risk of SGA
for mothers older than 35 years or who experienced two
or more prior fetal losses (ATSDR 1998; Sonnenfeld ct
al. 200 I). The study could not, however, evaluate child-
hood cancers and birth defects because the study relied
solely on birth certificates to ascertain adverse birth
outcomes.11 However, because this study used incorrect
information on the start-up date for the Holcomb Boule-
vard WTP. 12 errors were made in assigning exposures to
the mothers. Therefore. this study is being re-analyzed
using the results from the historical reconstruction
process and water-modeling analyses.
During 1999, ATSDR began an epidemiological
study to evaluate whether in utero and infant (up to I year
of age) exposure to YOC-contaminated drinking water
was associated with specific birth defects and childhood
cancers. The study includes births during 1968-1985
to women who resided at the base anytime during their
pregnancy. The first year of the study. 1968. was chosen
because North Carolina computerized its birth certificates
starting that year. The last year of the study, 1985, was
chosen because the most contaminated Tarawa Terrace
water-supply wells (TT-23 and TT-26, Figure A I) were
removed from regular service that year (February 1985).
The study is evaluating the central nervous system
defects known as neural tube defects (for example, spina
bifida and aneneephaly). cleft lip and cleft palate, and
childhood leukemia and non-Hodgkin's lymphoma.
The study consists of a multistep process that includes:
• a scientific literature review to identify
particular childhood cancers and birth defects
associated with exposure to VOC-contaminated
drinking water.
• a telephone survey to identify potential cases,
• a medical records search to confirm the
diagnoses of the reported cases. and
• a case-control study to interview parents
(collect information on a mother's residential
11 Birth defect:, are only poorly a:-.ccnained using hirth certificate:-.:
childhood cancers arc not included on birth certi!ica!es.
12 Curn.:111 inl"ormatinn from the Camp Lejeune Puhlic Wmks Dqx1r1111c111
Utilities Section indicates th:11 thc 1/okomb Boulcvan! WTP began ~upplying
finished water to area~ serviced hy the Holcomb Hnulcvard WTP (Plate 1)
during June 1972 (S.A. Brewer. U.S. ~brine Corp:-. Ba~e Camp Lejeune.
written C(Hnrnunicatinn. September 29. 2005).
Chapter A: Summary of Findings
history and water use as well as potential risk
factors such as a mother's occupation and ill-
nesses during pregnancy) and obtain exposure
estimates through water-modeling analyses and
the historical reconstruction process.
During 2004. the study protocol received approval from
the Centers for Disease Control and Prevention Institu-
tional Review Board and the U.S. Office of Management
and Budget.
Water-Distribution Investigation
Given the paucity of measured historical
contaminant-specific data and the lack of historical
exposure data during most of the period relevant to the
epidemiological study (January 1968-Deeember 1985),
ATS DR decided to apply the concepts of historical
reconstruction to synthesize and estimate the spatial and
temporal distributions of contaminant-specific concen-
trations in the drinking-water supply at Tarawa Terrace.
Historical reconstruction typically includes the applica-
tion of simulation tools, such as models, to recreate (or
synthesize) past conditions. For this study, historical
reconstruction included the linking of groundwater
fate and transport models with materials mass balance
(simple mixing) and water-distribution system models
(Table A4). The primary focus for the investigation of
the Tarawa Terrace historical reconstruction analyses
was the fate and transport of, and exposure to. a single
constituent-PCE. Additional and enhanced analyses
that relate to degradation by-products of PCE-TCE,
1,2-tDCE, and VC-also are presented (Figure A2).
Based on groundwater and water-quality data collection
and analyses by Shiver ( 1985). PCE originating from
the site of ABC One-Hour Cleaners is considered the
primary YOC compound responsible for contaminating
the Tarawa Terrace water-supply wells.
Models Used for Water-Distribution Investigation
Applying simulation tools or models to recon-
struct historical contamination and exposure events
at Tarawa Terrace and vicinity required the develop-
ment of databases from diverse sources of information
such as well and gcohydrologic analyses. computa-
tions of PCE mass at the ABC One-Hour Cleaners
site and within the Tarawa Terrace and Upper Castle
Hayne aquifers. and analyses and assessment of
Water-Distribution Investigation-----------------------------------
Table A4. Analyses and simulation tools (models) used to reconstruct historical contamination events at Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina.
I VOC. volatile organic mm pound: PCE. tctrachlorocthylenc: GIS. geographic infonn:nion .\ystcm: WTP. water treatment plant: TCE. trichlornethylene:
1.2-tDCE. llWH-1.2-did1lnrnc1hylcnc: VC. vinyl chloride)
Gcohy<lrologic framework
Prcdeveloprncnt ground-
water now
Transient ground-
water flow
Properties or VOCs
in groundwater
Computation of PCE mass
Fate and transport of PCE
PCE concentration in
WTP finished waler
Fate and 1ransport of
PCE and dcgrad:iti1m
by-produc1s in ground-
water and vapor phase
Early and late arrival of
PCE at WTP
Parameter uncertainty
and vuriahility
Distribution of PCE
in drinking water
A12
Detailed analyses of wdl and geohydro-
logic data used to develop framt:work
of the Castle Hayne aquifer system at
T,irawa Terrace and vicinity
Steady-state groundwater llow, occurring
prior to initiation of water-supply well
activities ( 1951) or after recovery of
water levels from cessation of pumping
activities (1994)
Unsteady-state groundwater !low occur-
ring primarily because of the initiation
and continued operation of w,.11cr-supply
wells (January 1951-December 1994)
Properties of degradation pathways of com-
mon orgunic compounds in groundwater
Estimates of mass (volume) of PCE:
(a) unsaturated zune (above water tahll:)
in vicinity or AHC One-Hour Cleaners
based on 1987-1991 data; (h) within
Tarawa Terrace and Upper Castle l·layne
aquifers based on 1991-1991 data
Simulation of the fate and migration
of PCE from its source (ABC One-
Hour Cleaners) to Tarawa Terrace
water-supply wel Is (January ! 951-
December 1994)
Computation of concenlration of PCE
in drinking waler from 1he Tarawa
Terrace WTP using results from fa!c
and transport 11H1ddi11g
Three-dimensional. muhiphase simula1ion
(lf the fate. degradation. and transport
of PCE degradation by-products: TCE,
1.2-tDCE. and VC
Analysis to assess impact of schedule
variation of water-supply well operations
on arrival of PCE al wells and !he
Tarawa Terrace WTP
Assessment of parameter sensitivity, un-
ccriainty. and vari:1hility associated with
model simulations of ground-water flow,
fate and transport. and water distribution
Simulation of hydraulics and water quality
in water-distributi1in system serving
Tarawa Terrace based u11 presen1-day
(2()04) co11dili(HJS
Data analysis
MODFLOW-96-
numerical model
MODFLOW-96-
numcrical model
Literature survey
Site investigation data.
GIS, and -"Patial analyses
MTJDMS-numerical
model
i\-latcrials mass balance
model using principles of
conservation of 111ass and
continuity-algebraic
TcehFlowMP-numerical
PSOpS-numerical:
op1imization
Faye (In press 2007a)
Harbaugh and McDonald
( 1996): Faye and Valen-
zuela (In press 2007)
Harbaugh and McDonald
( 1996 ): Faye and Valen-
zuela (In press 2007)
Lawrence (2006,
In press 2007)
Roy F. Weston, Inc. ( 1992.
1994): Pankow and Cherry
( 1996 ): Faye and Clreen
(In press 2007)
Zheng and Wang ( 1999):
Faye (In press 2007b)
i\fasters ( 1998): Faye
(In pn::ss 2007b)
fang and Aral (2005.
2007, In press 2007)
Wang and Aral (2007,
In press 2007)
PEST; Monte Carlo simula-Doherty (2005): Maslia
lion-probabilistic ct al. (In press 2007b)
EPANET 2-numerical Rossman (2000): Sautner
et al. ( In press 2007)
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
---------------------------------Water-Distribution Investigation
historical and present-day (2002) operations of the
water-distribution system serving Tarawa Terracc.13 A
complete list of analysis and simulation tools used to
reconstruct historical contamination and exposure events
at Tarawa Terrace and vicinity is provided in Table A4.
Information and data were applied to the models in the
following sequence:
I. Geohydrologic framework information, aquifer and
confining unit hydraulic data. and climatic data were
used to determine prcdevclopment (prior to 1951)
groundwater-flow charactcristics.1~ To simulate
predevel<1pmcnt groundwater-now conditions.
the public-domain code MODFLOW-96 (Har-
baugh and McDonald 1996)-a three-dimensional
groundwater-fl<1w model code-was used.
2. Transient groundwater conditions occurring primar-
ily because of the initiation and continued operation
of water-supply wells at Tarawa Terrace also were
simulated using the three-dimensional model code
MODFLOW-96: well operations were accounted for
and could vary on a monthly basis.
3. Groundwater velocities or specific discharges
derived from the transient groundwater-flow model
were used in conjunction with PCE source, fate, and
transport data to develop a fate and transport model.
To simulate the fate and transport of PCE as a single
specie from its source at ABC One-Hour Cleaners
to Tarawa Terrace water-supply wells, the public
do111ain code MT3DMS (Zheng and Wang 1999)
was used. MT3DMS is a 111odel capable of simulat-
ing three-dimensional fate and transport. Simulations
describe PCE concentrations on a monthly basis
during January 1951-December 1994.
4. The monthly concentrations of PCE assigned
to finished water at the Tarawa Terrace WTP
were determined using a materials mass balance
model (simple mixing) to compute the flow-
weighted average concentration of PCE. The
model is based on the principles of continuity and
conservation of mass (Masters 1998).
u A comprehensive list of references u~ed to gather. analy,,e, and a~semhlc
inf1Jnnatilm and data for the Tarawa Terrat.:e water-distribution investigation
is provided on the dectronit.: media (DVD) accompanying this report and in
the Chapter K rep(lrt (Maslia et al. In press 2007a_).
14 !'redevelopment or slt.:ady-statc refers to groundwaler t.:ondi1ions prim
to or after the t.:css:ilion of :111 water-M1pply well pumping activity.
Chapter A: Summary of Findings
5. To analyze the degradation of PCE into degrada-
tion by-products (TCE. 1.2-tDCE. and VC) and
to simulate the fate and transport of these contam-
inants in the unsaturated zone (zone above the
water table). a three-dimensional. multispecies, and
multiphase mass transport model was developed by
the Multimedia Simulations Laboratory (MESL)
at the Georgia Institute of Technology (Jang and
Aral 2005. 2007. In press 2007).
6. To analyze and understand the impacts of unknown
and uncertain historical pumping schedule varia-
tions of water-supply wells on arrival of PCE at
the Tarawa Terrace water-supply wells and WTP,
a pumping and schedule optimization system tool
(PSOpS) was used. This 111odel was also developed
by the MESL (Wang and Aral 2007, In press 2007).
7. To assess parameter sensitivity, uncertainty, and
variability associated with model simulations of
llow, fate and transport, and computed PCE con-
centrations in finished water at the Tarawa Terrace
WTP, sensitivity and probabilistic analyses were
conducted. Sensitivity analyses were conducted
using a one-at-a-time approach: the probabilistic
analyses applied the Monte Carlo simulation (MCS)
and sequential Gaussian si111ulation (SGS) methods
to results previously obtained using MODFLOW-96,
MT3DMS. and the drinking-water mixing model.
8. The initial approach for estimating the concentration
of PCE delivered to residences of Tarawa Terrace
used the public domain model. EPANET 2 (Ross-
man 2000)-a water-distribution system model
used to simulate street-by-street PCE concentra-
tions (Sautner et al. 2005. 2007). Based on expert
peer review of this approach (Maslia 2005) and
exhaustive reviews of historical data-including
water-supply well and WTP operational data when
available-study staff concluded that the Tarawa
Terrace WTP and water-distribution syste111 was
not interconnected with other water-distribution
syste111s at Ca111p Lejeune for any substantial ti111e
periods (greater than 2 weeks).15 Thus, all water
15 The term ••imerconnection·· is defined in this swdy a" the continuous
tlow of water in a pipeline from one water-di!->tribution sy!->lem to another
for periods excet.:ding two week~. Pipelines did connect two or more
water-distribution sy..,tems. but unle\s continuous !low wus documented.
the w:11cr-dis1rihutio11 systems were assumed not In be inten.:unnected.
A13
Water-Distribution Investigation-----------------------------------
arriving at the WTP was assumed to originate solely
from Tarawa Terrace water-supply wells (Faye and
Valenzuela In press 2007: Faye In press 2007b)
and to be completely and uniformly mixed prior to
delivery to residents of Tarawa Terrace through the
network of distribution system pipelines and stor-
age tanks. Based on these information and data,
study staff concluded that a simple mixing model
approach, based on the principles of continuity and
conservation of mass, would provide a sufficient
level of detail and accuracy to estimate monthly
PCE exposure concentrations at Tarawa Tcrrace.11,
Thus, results of the monthly now-weighted aver-
age PCE-c<rncentration computati()JlS were provided
to agency health scientists and epidemiologists to
assess population exposure to PCE.
Data Needs and Availability
The historical n.:constructi(m process required
information and data describing the functional and
physical characteristics of the groundwater-flow
system, the chemical specilic contaminant (PCE) and
its degradation by-products, and the waler-distribution
system. Required for the successful completion of
the historical reconstruction process, specific data
can he categorized into four generalized informa-
tion types that relate to: (I) aquifer geometry and
hydraulic characteristics (for example, horizontal
hydraulic conductivity, effective porosity, and disper-
sivity); (2) well-construction. capacity. and pumpage
data (for example, drilling dates, well depth, opera-
tional dates. and quantities of pumped groundwater
by month): (3) chemical properties and transport
parameters (for example, partition coefficients,
sorption rate, solubility, and biodcgradation rate); and
(4) water-distrihution system design and operation
data (ror example, monthly delivery or finished water
rrom the Tarawa Terrace WTP, network geometry
and materials of pipelines, and size and location of
storage tanks). Availability of specific data, methods of
obtaining data, assessment of the reliability of the data.
and implications with respect to model assumptions and
simulations arc discussed in detail in chapter reports B-J
(Table A2 and Appendix A I).
1
'' This assumption is tcste<l and vcriticd in the Chapter J report
(Sautner ct al. In press 2007) of this .~tudy.
Ideally, data collection in support or the historical
reconstruction process is through direct measurement and
observation. In reality, however, data collected are not
routinely available by direct measurement and must be
recreated or synthesized using generally accepted engi-
neering analyses and methods (for example, modeling
analyses). Additionally, the reliability or data obtained by
direct measurement or observation must be assessed in
accordance with methods used to obtain the data. Issues
or data sources and the methods used to obtain data that
cannot be directly measured, or arc based on methods
of less accuracy, ultimately reflect on the credibility of
simulation results. The methods for obtaining the neces-
sary data for the historical reconstruction analysis were
grouped into three categories (ATSDR 2001):
• Di reel meusureme11t or ohservathm-Data included
in this category were obtained by direct measurement
or observation of historical data and are veri liable hy
independent means, Data obtained by direct measurement
or observation still must be assessed as to the methods
used in measuring the data. For example, in the Chap-
ter C report, Faye and Valenzuela (In press 2007) discuss
that water-level data ohtained from properly constructed
monitor wells using electric-or steel-tape measurements
are more reliable than water-level data obtained from
water-supply wells using airline measurements. Of
the three data categories discussed, data obtained by
direct measurement were the most preferred in terms
of reliability and least affected by issues of uncertainty.
Examples of such data included aquifer water levels,
PCE concentrations in water-supply wells and in linished
water at the WTP. and PCE concentration at the location
of the contaminant source (ABC One-Hour Cleaners).
• Qua111i1ative estimates-Data included in this
category were estimated or quantified using generally
accepted computational methods and analyses, for
example, monthly inliltration or recharge rates to the
Castle Hayne aquifer system and estimates of contami-
nant mass in the vicinity of ABC One-Hour Cleaners and
the Tarawa Terrace and Upper Castle Hayne aquifers.
• Qualitative description-Data included in this
category were based on inference or were synthesized
using surrogate information, for example, water-supply
well operational information, retardation factors,
and aquifer dispcrsivity. Of the three data categories
described, data derived by qualitative description were
the least preferred in terms of reliability and the most
affected by issues of uncertainty.
A14 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
----------------------------------Water-Distribution Investigation
Chronology of Events
To reconstruct historical ex posures, a reliable chro-
nology related to operations of the identified source of
the PCE contamination, ABC One-Hour Cleaners, and of
water-supply facilities (wells and the WTP) is of utmost
importance. This information has a direct impact on
the reliability and accuracy of estimates derived for the
levels and duration or exposure to contaminated drinking
water. Using a variety of information sources and refer-
ences, events related to water supply and contamination
of groundwater and drinking water at Tarawa Terrace
and vicinity are shown graphically and explained in
Figure A3. Examples of information sources and refer-
ences used to develop the chronology of events shown
in Figure A3 in clude: (I) capacity and operational
histories of Tarawa Terrace water-supply wells and the
WTP (Faye and Valenzuela In press 2007), (2) depo-
sitions from the owners of ABC One-Hour Cleaners
(Melts 200 I ), (3) identification and characterization of
the source of PCE contamination (Shiver 1985). and
(4) laboratory analyses of samples from water-supply
wells (Granger Laboratories 1982) and the WTP
(CLW 3298-3305).
One of the purposes of Figure A3 is to present, in
a graphical manner. the relation among water supply,
contamination events, exposure to contaminated drink-
ing water in family housing areas, selected simulation
results, and the time frame of the epidemiological case-
control study. For the first time. all of these different
types of information and data sources are summarized in
one document that is be lieved to be an accurate recon-
ciliation of chronological events that relate to Tarawa
Terrace and vicinity. Three events are noteworthy:
(I ) the year shown for the start of operations of ABC
One-Hour Cleaners ( 1953) is used a the starting time
for PCE contamination of groundwater in the fate and
transport modeling of PCE, (2) sampling events and PCE
concentration values of tap water are shown for 1982.
and (3) the closure of the Tarawa Terrace WTP is shown
as occurring during March 1987. Care has been taken
to assure that chronological event information and data
required for modeling analyses and the historical recon-
struction process (I) honor original data and informati on
sources. (2) are consistent and in agreement with all
Tarawa Terrace chapter reports. and (3) reflect the most
up-to-date information.
Chapter A: Summary of Findings
Occurrence of Contaminants in Groundwater17
Detailed analyses of concentrations or PCE at
groundwater sampling locations and at Tarawa Ter-
race water-supply wells during the period 1991 -1993
were sufficient to estimate the mass. or amount. of PCE
remaining in the Tarawa Terrace and Upper Castle Hayne
aquifers. Similar methods were applied to compute the
mass of PCE in the unsaturated zone (zone above the
water table) at and in the vicinity of ABC One-Hour
Cleaners using concentration-depth data determined from
soil borings during field investigations of 1987-1993.
These analyses are presented in Faye and Green (In
press 2007) and are summarized in Table AS. This infor-
mation and data were necessary to develop accurate and
reliable databases to conduct model simulations of the
fate and transport or PCE from its source-ABC One-
Hour Cleaners-to Tarawa Terrace water-supply wells
and WTP. The total mass of PCE computed in ground-
water and within the unsaturated zone during the period
1953-1985 equals about 6.000 pounds and equate~
to a volume of about 430 gallons (gal). 18 This volume
represents an average minimum loss rate of PCE to the
Table A5. Computed volume and mass of tetrachloroethylene in
the unsaturated and saturated zones, Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina.'
I PCE. tetrachlorocthylene I
Average annual
Dates of Volume, contribution of PCE,
Zone computation in gallons2 1953-1985 ---
In gallons In grams
Un,aturated' 1987-1993 190 6 J6.340
Saturated' 1991-1993 240 7 42,397
Total -BO 13 78,737
'Refer 10 Chapter E report (Faye and Green In pre,s 2007) for ,pecilic
computational detai1'
1Dcnsity of PCE i, 1.6 gram, per cubic centimeter. or about IO I pounds
per cubic foot
'Zone above water table in vicinity of ABC One-I lour Cleaner,
'Tarawa Terrace and Upper C.1,tlc Ha) nc aquifer,
., For detailed analyse, and di,cu"ion, of occurrence ol nmtammant,
in groundwater al Tarawa Terrace and vicinity. refer to the Chapter E report
( Faye and Green In pre,, 2007).
"Typically. ;,m:h volume, also arc exprc"cd in terms uf 55-gal drums. The
aforementioned volume of -BO gal of PCE i, c4uivakn1 10 7.8 drum, of PCI-.
A15
Water-Distribution Investigation ----------------------------------
A16
1940s: Building constructed
on site of ABC One-Hour Cleaners
with septic tank soil-adsorption
(ST-STAI system
I
1942-1943: Hadnot Point water treatment
plant (WTPI begins operations
-'--_____.__--'-------l.____J____L-----'--1-L.._____J
1940 1945 1949
1951-1952: Tarawa Terrace (TT)
housing constructed
I May 1951: Well TT-26 constructed 900 feet
from ABC One-Hour Cleaners site:
Well TT-27 constructed
1951-1952: Wells TT-28, TT-29,
and TT-45 constructed
___c_ 1952-1953: TT WTP
1953: ABC One-Hour Cleaners
begins operations using existing
ST-STA for disposal of wastewater
January 1957: PCE
concentration
in well TT-26,
5.2 µg/L Ii 1957: Montford Point WTP, servicing
Camp Johnson area, begins operations
I November 1957: PCE concentration ....----.--'-I --_T-___ be~gi~on~
..-------I at TT WTP. 5.4 µg/L
1950 1955 1959
1960s: ABC One-Hour Cleaners
installs floor drain to septic system I January 1968-0ecember 1985:
t
1960
1970
I 1961: Wells TT-52, TT-53, TT-54,
J and TT-55 constructed
I I
1971: Well TT-30 constructed
1965
lime frame of ATSOR case-control
epidemiological study on birth
defects and childhood cancers
1969
I November 1971: Well TT-67 constructed
r
June 1972: Holcomb Boulevard WTP begins
deliverin.g treated water to Holcomb Boulevard area
1973. Well TT-31 constructed 9 6 C K T .1 C 1 7 : amp nox ra1 er
.....---~---,----1'--~I ,....,._3_.. Park constructed
1975 1979
J I 1981. w II TT 25 d I March 1983: Well TT-23 constructed about u Y · e · constructe 1,800 feet from ABC One-Hour Cleaners
April 1982: VO Cs detected in drinking water March 1984:_ Simulated PCE concentration at TT VVTP, 183 µg/L
May 28, 1982: Tap water at TT sampled, July 1984: Simulated PCE concentration in well TT-26, 851 µg/L
July 28, 1982: Tap water at TT sampled, TT-25, trace; TT-26, 3.9 µg/L
PCE concentration, 80 µg/L ~ July 1984: TT wells sampled forTCE: TT-23, 37 µg/l;
PCE concentration, 104 µg/l;
1980
I
February 11, 1985: STT-39A at TT sampled, d1stnbut1on system sampled,
PCE concentration, 76 and 82 µg/L I PCE, 215 µg/L: TCE, 8.0 µg/L; 1,2-tDCE, 12 µg/L
I -
11
February 8, 1985: Wells TT-23 and TT-26 taken off-line
,---1985: ABC One-Hour Cleaners
discontinues use of septic tank
March 1987: TT WTP closed
1987: Montford Point WTP closed
1985
Category of event
1989 -D D D
Housing/buildings Water supply Contaminant source Sampling event Simulated event Health study
Figure A3. Chronology of events related to supply and contamination of drinking water, Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina. (STT-39A is the pump house associated with storage tank STT-39.)
[ft, foot; µg/L, mic rogram per liter; VOC, volatile organic compound; PCE, tetrachloroethylene; TCE, trichloroethylene;
1,2-tOCE, trans-1,2-dichloroethylene; current maximum contaminant levels: PCE 5 µg/L, TCE 5 µg/L, 1,2-tDCE 100 µg/L)
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
----------------------------------Water-Distribution Investigation
subsurface at ABC One-Hour Cleaners of about 13 gal-
lons per year during the period 1953-1985 . This PCE
loss rate should be considered a minimum because (I ) the
quantity of PCE removed from the aquifers at Tarawa Ter-
race water-supply wells during I 953-1985 is unknown,
(2) biodegradation of PCE to daughter products of
TCE, 1,2-tDCE, and VC was probably occurring in the
aquifers during and prior to 1991 . and (3) PCE mass
adsorbed to the sands and clays of the aquifer porous
media and was not accounted for during the PCE mass
computations. Pankow and Cherry ( 1996) indicate that
computations of contaminant mass similar to those sum-
marized here and described in detai l in Faye and Green
(In press 2007) represent only a small fraction or the
total contaminant mass in the subsurface. Comparing the
estimated volume of 430 gal of PCE (7.8 55-gal drums)
computed by Faye and Green (In press 2007) with
documented contaminant plumes in sand-gravel aqu ifers
indica tes that the contaminant mass in the subsurface
at Tarawa Terrace would have been ranked as the third
greatest vo lume of contaminant mass among seven
contamination sites in the United States listed in a table
provided in Mackay and Cherry (Table I , 1989).
Relation of Contamination to Water Supply,
Production, and Distribution
Historically, groundwater was used as the sole
source of water supply for Camp Lejeune, and in particu-
lar, Tarawa Terrace. Of critica l need in terms of historical
reconstruction analysis, was information and data on the
monthly raw water production of' supply wells (to enable
computations of flow-weighted drinking-water concen-
trations), and the distribution of finished water to family
housing areas. The supply of drinking water to Tarawa
Terrace was composed of two components: (I) the supply
of water from groundwater wells to the Tarawa Terrace
WTP and (2) the delivery of' finished water from the
WTP through the network of pipelines and storage tanks
of the water-distribution system. The placement of water-
supply wells into service and their permanent removal
from service are critical to the analysis and simulation
of contamination events. For example. water-supply
well TT-26 was constructed during May 1951, probably
placed into service during 1952, and was permanently
taken off-line (service terminated) February 8, 1985. The
Tarawa Terrace WTP began operations during 1952-1953
and was closed during March 1987 (Figure A3). All
Chapter A: Summary of Findings
groundwater wells in the Tarawa Terrace area supplied
untreated (or raw) water to a central treatment facility-
the Tarawa Terrace WTP (Figure A4). Information
pertaining to well-capacity histories, including construc-
tion, termination of service, and abandonment dates and
spatial coordinate data are described in detail in Chap-
ter C (Faye and Valenzuela In press 2007).
After treatment at the Tarawa Terrace WTP. fin-
ished water was distributed through pipelines to storage
tanks, residential housing, military facility buildings.
and shopping centers. 19 Information and data related to
the water-distribution system (Plate I; Figure A4) were
gathered as part of data discovery and field investigation
activitie. in support of the ATS DR epidemiological case-
control study. The network of pipelines and storage tanks
shown on Plate I and in Figure A4 represents present-
day (2004) conditions, described in detail in Chapter J
(Sautner et al. In press 2007). Based on a review of
historical operating and housing information, the histor-
ical water-distribution system serving Tarawa Terrace
was considered very similar and nearly identica l to the
present-day (2004) water-distribution system-the excep-
tion being two pipelines that were put into service during
1987 after the closing of the Tarawa Terrace and Camp
Johnson WTPs. One pipeline. constructed during 1984.
follows SR 24 northwest from the Holcomb Boulevard
WTP and presently is used to supply ground storage
tank STT-39 with finished water (Plate I. Figure A4).
The other pipeline, constructed during 1986, trends
east-west from the Tarawa Terrace II area to storage tank
SM -623 and presently is used to supply finished water
from Tarawa Terrace to elevated storage tank SM-623.
Historically ( 1952-1987), the Tarawa Terrace water-
distribution system was operated independently of, and
was not interconnected w ith. the M ontford Point or
Holcomb Boulevard water-distribution system~.w
Based on epidemiological considerations, hi storical
reconstruction results were provided at monthly interval~.
Ideally, these analyses require monthly groundwater
pumpage data for the historical period. However, pump-
age data were limited and were available on a monthly
basis solely for 1978 and intermittently during the period
of 1981-1985. Faye and Valenzuela (In press 2007)
" Ba,ed on an analy,i, of building I) r,:, and u,age in Tarawa Terrace.
greater than 90% of the building, were u,ed for residential hnu,ing.
"'Although the two pipeline, di,rns,cd were con,tructcd during 1984 and
1986. hi,torical record;, ,uch a, water plant operator note, indicate that the
pipeline, did not convey lini,hcd water on a continuous ba,i, prior to 1987.
A17
Water-Distribution Investigation -----------------------------------
34°45'
34°43'30"
77°24'
Note: Camp Knox served by
Montford Point end Tarawa
Terrace water supplies at
various historical times
Base from U.S. Marine Corps and
U.S. Geological Survey digital data files
Historical water-supply
areas of Camp Lejeune
Military Reservation
D Montford Point
D Tarawa Terrace
D Holcomb Boulevard
D Other areas of Camp Lejeune
Military Reservation
■ ABC One-Hour Cleaners
77°22 30" 77'21
'--1.
Area of maps in Figures Al3a-A15a and A17a
EXPLANATION
Water distribution
Tarawa Terrace water pipeline
6 SM-62J Elevated storage tank and number
•sTT-39 Ground storage tank and number
D Water treatment plant-Tarawa
Terrace and Montford Point
(MPWTP) (closed 1987)
TT-26• Water-supply well
and identification
0.5 1 MILE
0.5 1 KILOMETER
Groundwater-flow and
contaminant fate and
transport model boundaries
Domain --Active area
Boundary conditions for
groundwater-flow model
General head No flow
Drain • Specified
head
Figure A4. Location of groundwater-flow and contaminant fate and transport modeling areas and water-supply
facilities used for historical reconstruction analyses, Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp
Lejeune, North Carolina.
provide detai ls regarding groundwater pumpage includ-
ing sources and capacity hi story. Where pumpage data
were missing or incomplete, aquifer water-level and
water-supply data, in conjunction with model simulation,
were used to synthesize and reconstruct monthly water-
supply well operations. Tarawa Terrace water-suppl y well
operati ons-in terms of online dates and off-line dates for
water supply-are presented graphically in Figure AS.
Once a well was put in service, it was assumed to oper-
ate continuously for modeling purposes unti l it was
permanently taken off-line-the exception being tern-
porary shut downs for long-term maintenance. Breaks
in continuous operations, such as those for wells TT-26
and TT-S3, also are shown in Figure AS and are based on
documented information detailing periods of maintenance
for specific well s. For example, water-supply well TT-26
was shut down for maintenance during July-August 1980
and January-February 1983 (Faye and Valenzuela In
press 2007). Table A6 lists the specific month and year
for the start of service for all Tarawa Terrace water-supply
wells and the specific month and year for the end of ser-
vice. Because raw water from all groundwater wells was
A18 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
----------------------------------Water-Distribution Investigation
...J ...J ~ ::; a... a... ::::,
V) rr. UJ ~ s:
I
.,_ Well name
/
I TT-54 .,_
Well in .,_ operation
I---
i------1
Jan
1950
TT-29
I
Jan
1955
-\-
Well not in
operation
TT-27
#7
I #6
I
-
-
TT-45
I
Jan
1960
TT-28
I
I
Jan
1965
TT-55
TT-26
I
I
Jan
1970
I
TT-67
TT-54
TT-53
TT-52
I
Jan
1975
TT-30
I
TT-31
I
Jan
1980
I
TT-25 -· -· TT-23
I
Jan
1985
Jan
1990
Figure A5. Historical operations of water-supply wells, 1952-87, Tarawa Terrace and vicinity, U.S. Marine Corps
Base Camp Lejeune, North Carolina .
Table AG. Historical operations for water-supply wells, 1952-1987, Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina.1
1-. 1101 applicable I
Well In service identification
#6 January 1952
#7 January 1952
TI-23 August 1984
TI-25 January 1982
TI-26 January 1952
TI-27 January 1952
TI-28 January 1952
TI-29 January 1952
TI-30 January I 972
TI-31 January I 973
TI-45 January 1952
TI-52 January 1962
TI-53 January 1962
TI-54 January I 962
Tf-55 January 1962
TI-67 January 1967
Off-line
February 1985
July-Augus1 I 980:
January-February 1983
September 1984
June 1984
March 1986
July-August I 98 I
February-March 1984
'Refer 10 1he Chap1cr C rcpon (Faye and Valcn,ucla In pre,, 2007) for addi1ional deiail,
Service terminated
January 1962
January 1962
May 1985
March 1987
February 1985
January 1962
January 1972
July 1958
February 1985
March 1987
January 1972
March 1987
February 1984
March 1987
January I 972
March 1987
Chapter A: Summary of Findings A19
Water-Distribution Investigation -----------------------------------
mixed at the Tarawa Terrace WTP prior to treatment and
distribution to Tarawa Terrace housing areas. the tart-up
and shut-down dates of specific water-supply wells. such
as TT-26 and TT-23. were critical 10 accurately deter-
mining the concentration of eontaminams in finished
water delivered from the Tarawa Terrace WTP.
Total annual groundwater pumpage by well for all
Tarawa Terrace water-supply wells is shown graphically
in Figure A6. Refer to the Chapter C report (Faye and
Val enzuela In Press 2007) for data sources used to derive
Figure A6. This illustrat ion also shows the contribution
10 pumpage by individual wells on an annual basis. For
example. during 1978 total annual groundwater pump-
age was 327 million gallons (MG) contributed by wells
TT-26 (64.7 MG), TT-30 (25.9 MG). TT-3 1 (46.2 MG).
TT-52 (48.1 MG), TT-53 (27.7 MG), TT-54 (62.8 MG).
and TT-67 (5 1.7 MG) (Faye and Valenzuela In press
2007). Thus. well TT-26 and TT-54 contributed about
20 percent (%) each to the total annual pumpage for
1978, and well TT-30 contributed about 8%. This total
annual groundwater pumpage is in agreement with the
average rate of water delivered to the Tarawa Terrace
WTP in 1978 of0.90 million gallons per day. reported
by Henry Yon Oesen and Assoc iates Inc. ( 1979).
400 I I I
,-Water-supply well
The historical Tarawa Terrace water-distribution
system was probably nearly identical to the present-day
(2004) water-d istribution system. Operational charac-
teristics of the present-day water-distribution system
were used for historical reconstruction analyses and
were based on data gathered during field investi gations
(Sautner et al. 2005. Maslia et al. 2005). Delivery rates of
fin ished water on a monthly basis during 2000-2004 are
listed in Table A7 and shown graphically in Figure A7.
For the 5-year period 2000-2004, the mean monthly
delivery of finished water to the Tarawa Terrace water-
distribution system was estimated to be 18.5 MG.21
Monthly variations were most probably due to troop
deployments. Monthly delivery data indicate that rela-
tively high rates of finished water were delivered during
the months of April, May. June. and July of 2000 and
200 I . In addition, May and June of 2000 we re the months
of greatest delivery of tinished water to the Tarawa Ter-
race water-distribution system-an estimated 30.9 MG of
linished water during each month (Figure A 7. Table A 7).
"Since March 1987. iini,hcd water for the Tarawa Terrace water-
di,1ribu1ion ,y,tcm ha, been provided by the Holrnmb Boulevard WTP and
delivered to ground ,1orage lank STr-39 (Plate I J. See section 011 Field Test,
and Analyse, of 1hc Water-Di,tribu1io11 Sy,1cm or 1he Chapter J report
(Sautner et al. In pre,, 2007).
I I I I
-D 16 □ TT-25 □ TT-28 • TT-31 D TT-53 D TT-67
U) z 0 -' -' <(
{!)
z 0 ::::; ::::! ~
!":
uJ {!)
<(
Q.. ~ => Q..
-' <( => z z
<(
-' ~ 0 I-
A20
,-
,-
300 t--
,-
,-
,-
200 t--
100 t--
,-
,-
0
1950
□n c::J TT-26 • TT-29 D TT-45 D TT-54
• TT-23 111 TT-27 □ TT-30 D TT-52 D TT-55 -------------~---------
->-----... ... ..... -- ->--------""' ..,
r:! -L-... ._
-... ... ... ... ... .... ... ... ..... .... ... --... --L-.... --,--,_ ,_ ... I, ..... -... ... .., -... i,.,. -... ... ... ... ... --,_ ..... ... ......
I-· ... .... ...... ,_ -.._
... ... I ... '"--,__ --._ ----._ -... I;·. ... H ,_ ... .... ,-,-L-'-.... -... .... -.... ...... ,__ -t;. -
.... ,_ -._ ... I---.... ... .... .., ----,_ ... ,__ ,__ ,_ -.... ...
I;'
I I I I
1955 1960 1965 1970
-----------------...... ... -
L-... ... -,_ -.... --,_ ... ... L--._ .... '-.... .... ..... ._ -... ... I L---L-
I '--
I L-I-,_ -....
'-I I ...
,,' -
I ~ ... ... t b -ll.. ;
L L....
r t; ii ~ 11 t7"
-.... tr .... .... .... --. J ,.11 H ·· ~ h -
I I I
1975 1980 1985 1990
Figure A6. Total annual groundwater pumpage at water-supply wells, 1952-1987, Tarawa Terrace and vicinity,
U.S. Marine Corps Base Camp Lejeune, North Carolina.
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Caro lina
Water-Distribution Investigation
Table A7. Estimated monthly delivery of finished water to the Tarawa Terrace water-distribution system, 2000-2004, U.S. Marine
Corps Base Camp Lejeune, North Carolina.1·2
IMG. million gallon,: MGD. million gallon, per dayJ
Delivered finished water3
Month 2000 2001 2002 2003
MG MGD MG MGD MG MGD MG MGD MG
January 23.500 0.758 19.028 0.613 21.017 0.678 21.775 0.702 14.238
February 20.937 0.722 18.557 0.663 17.320 0.619 14.960 0.534 13.715
March 22.847 0.737 19.338 0.624 18.300 0.590 15.735 0.508 11.721
April 26.371 0.879 27.060 0.902 18.549 0.618 14.060 0.469 12.805
May 30.924 0.998 19.468 0.628 16.974 0.548 13.365 0.431 14.088
June 30.907 1.030 25.156 0.839 17.163 0.570 13.629 0.454 12.763
July 24.297 0.784 23.984 0.774 16.440 0.530 13.604 0.439 13.945
August 22.145 0.714 17.93 1 0.578 18.020 0.581 18.539 0.598 12.106
September 19.732 0.658 16.469 0.549 16.900 0.563 19.916 0.664 12.135
October 18.274 0.589 16.619 0.536 15.907 0.513 21.798 0.703 16.435
November 20.663 0.689 17.240 0.575 I 6.807 0.560 20.607 0.687 16.982
December 25.785 0.832 17.101 0.552 17.082 0.551 20.939 0.675 16.861
'Since March 1987. lini,hcd water for the Tarawa Terrm:e water-di,tribution sy,tem ha, been provided by the Holcomb Boulevard WTP
and delivered to ground ,toragc tank STT-39 (Plate I)
'Data from Joe l Hart,oc. Camp Lejeune Public Work, Depa11mcn1 Utilitie, Section. December 6. 2006
'Flow data measured al venturi meter located in building STT-39A {Tarawa Terrace pump hou,c)
V) z a -' -' <(
(!)
z a
:::i -' ~ z
35
30
25
0 2000
□ 2001
□ 2002
-0 2003
□ 2004
-
Mean monthly delivery
equals
18.5 million gallons rr.·
I.U I-~ 20 -
---- -
!~ - - ---
u.. a
>-rr. I.U > ~ 10 -
a
:'.:;
::i::
1-z a ~ -
JAN FEB MAR APR MAY JUNE JULY AUG SEPT OCT NOV DEC
Figure A7. Estimated monthly delivery of finished water to the Tarawa Terrace water-distribution system,
2000-2004, U.S. Marine Corps Base Camp Lejeune, North Carolina. [Data from Joel Hartsoe, Camp Lejeune
Public Works Department Utilities Section, December 6, 2006; flow data measured at venturi meter located
in building STT-39A (Tarawa Terrace pump house))
Chapter A: Summary of Findings
2004
MGD
0.-159
0.473
0.378
0.427
0.454
0.425
0.450
0.391
0.405
0.548
0.566
0.544
-
-
-
Hierarchical Approach for Quantifying Exposure ---------------------------
A dditional information gathered during a field
investigation of the Tarawa Terrace water-distribution
system included hourly delivery rates of finished water.
These hourly data were used in conjunction with water-
distribution system model simulation (see ection on
Field Tests and A nalyses of the Water-Distribution
System) to determ ine a diurnal pattern of water use for
Tarawa Terrace (Figure A 8). Data from the field test
show a gradually increasing demand for water occurring
during 0200-0700 hours. Peak demand occurs between
I 000-1400 hours, at 1800 hours, and at 2200 hours.
Thus, greater amounts of water were delivered (and
presumably consumed) during these time periods than
during other hours of the day.
100
Note: Measured flows are hourly
averages using 2-minute data
IR. Cheng, Camp Lejeune Environ-
mental Management Division,
written communication. January 25, 2005)
<:;,!§' ~<:;<:; ~,s, <::,'b,s, ~ .:!' ,#' ~,s, ,<o<S> #' ,o,<S> '\,~
HOUR
Figure AS. Measured diurnal pattern (24 hours)
of delivered finished water during field test,
September 22-0ctober 12, 2004, Tarawa Terrace
water-distribution system, U.S. Marine Corps Base
Camp Lejeune, North Carolina.
Hierarchical Approach for
Quantifying Exposure
A simulation or modeling approach was used to
reconstruct and estimate (quanti fy) historical concen-
trations of PCE in finished water delivered to residents
of Tarawa Terrace. In using a simulation approach, a
calibration process is used so that the combination of
various model parameters-regardless of whether a
model is simple or complex-appropriately reproduces
the behavior of real-world systems (for example, migra-
tion o f PCE) as closely as possible. The American Water
Works Association Engineeri ng Computer Applica-
tions Committee indicates that ·'true model calibration
is achieved by adjusting whatever parameter values
need adjusting until a reasonable agreement is achieved
between model-predicted behavior and actual field
behavior" (AWWA Engineering Computer Applications
Committee 1999). A model modifi ed in this manner is
called a calibrated model (Hill and Tiedeman 2007).
Calibration of models used for the Tarawa Terrace
analyses was accomplished in a hierarchical or step-wise
approac h consisting of four successive stages or levels.
Simulation results achieved fo r each calibration level
were refined by adjusting model parameter values and
comparing these results with simulation results of previ-
ous levels until results at all levels were within ranges
of preselected calibration targets or measures. T he
step-wise order of model-calibration levels consisted of
simulating (I) predevelopment (steady or nonpumping)
ground-water-flow conditions. (2) transient (time varying
or pumping) groundwater-flow conditions. (3) the fate
and transport (migration) of PCE from its source at ABC
One-Hour Cleaners to water-supply wells, and (4) the
concentration of PCE in finished water at the Tarawa
Terrace WTP-water from the Tarawa Terrace WTP
that was delivered to residents living in family housing.
Conceptual Description of Model Calibration
The hierarchical approach to estimating the concen-
tration of PCE in finished water from the Tarawa Terrace
WTP can be conceptually described in terms of Venn
or set diagrams (Borowski and Borwein 1991 ). Such
diagrams are use ful for showing logical relations between
sets or groups of like items and arc shown in Figure A9
for each hierarchical calibration level. At level I (Fig-
ure A9a), there may be a large number of combinations
of parameters that yield so lutions to predevelopment
groundwater-flow conditions. However, onl y a smaller
set-the subset of solutions indicated by circle ''A'' in
Figure A9a-yields acceptable combinations of param-
eters for a calibrated predevelopment groundwater-flow
model. For transient groundwater-flow conditions, viable
solutions are indicated by circle "B" (Figure A9b). Only
those solutions that successfu lly simulate both predeve l-
opment and tran sient groundwater-flow conditions can be
accepted and classified as resulting in ca librated transient
and predevelopment groundwater-flow models. These
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
----------------------------Hierarchical Approach for Quantifying Exposure
a. Predevelopment groundwater flow b. Transient groundwater flow
Universe of solutions
c. Contaminant fate and transport d. Water-supply well mixing
Calibration Calibration
B B
A A
Figure A9. Venn diagrams showing hierarchical approach of model calibration used to estimate concentration of
finished water: (a) predevelopment groundwater flow, (b) transient groundwater flow, (c) contaminant fate and transport,
and (d) water-supply well mixing, Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina .
select and fewer solutions are indicated by the intersec-
tion of circles "A" and "B." The transient groundwater-
flow simulations provide velocity information (specific
discharge) required to conduct a fate and transport simu-
lation. Viable solutions for the fate and transport problem
are indicated by circle "C" (Figure A9c). Only those
solutions that satisfy: (a) predevelopment groundwater-
flow, (b) transient groundwater-flow. and (c) contaminant
fate and transport calibration criteria are accepted and
classitied as resulting in a cali brated contaminant fate and
transport model. These so lutions are even fewer than for
predevelopment and transient groundwater flow and are
indicated by the intersection or circles "A."' "B."' and ·'C."
The fourth hierarchical level used to reconstruct PCE
Chapter A: Summary of Findings
concentrations in drinking water was the development
of a calibrated mixing model (using the materials mass
balance approach and mixing PCE-contaminatecl and
uncontaminated groundwater from supply wells). Viable
calibrated solutions depend on calibrated solutions for the
previous three hierarchical levels of model calibration.
thereby resulting in even fewer calibrated solutions to the
mixing problem-circle "D" in Figure A9d. Thus. only
solutions that satisfy all four levels of model calibration.
indicated by the intersection of circles ''A,'" ·'B," "C." and
"D," provide reasonable estimates for the concentration
of PCE in finished water at the Tarawa Terrace WTP.
The final calibrated models were the end product of this
hierarchical process.
A23
Hierarchical Approach for Ouantifying Exposure ----------------------------
Quantitative Assessment of Model Calibration
Speci fie detai Is of the calibration process for each
hierarchical level are described in the Chapter C report
for levels I and 2 (Faye and Valenzuela In press 2007)
and the Chapter F report for levels 3 and 4 (Faye In
press 2007b). To summarize, at each hierarchical level.
an initial calibration target or "goodness of lit" crite-
rion was selected based on the avai lability, method of
measurement or observation, and overall reliability
of field data and related information. Once model-
speci tic parameters were calibrated, statistical and
graphical analyses were conducted to determine if
selected parameters met calibration criteria targets.
Summaries of calibration targets and resulting ca libra-
tion statistics for each of the four hierarchical levels are
listed in Table AS. Graphs of observed and simulated
water levels using paired data points22 are hown in Fig-
ure A IO for predevelopment and transient ground water-
now calibrations (hierarchical levels I and 2). Of special
note are ca libration target and resulting calibration
statistics for hierarchal level 2-transient ground water
now (Figure A I Ob and Table AS). The calibration targets
were divided into those renective of monitor well data
and those reflective of water-supply well data. As listed
in Table AS, calibration targets for water-level data
derived from monitor well data were assigned a smaller
head difference (±3 ft) when compared with ca libration
targets derived from water-supply well data(± 12 ft).
This difference in the calibration targets-and resulting
calibration statistics-reflects the more accurate mea-
surement method used Lo determ ine monitor well water
leve ls (steel-tape measurements) when compared with
the method used to determine water-supply well water
levels (airline measurements). The resulting calibration
statistics and paired data point graphs also demonstrate
a betler agreement between monitor well data and model
simulation (average magnitude of head difference of
1.4 ft) than between water-supply well data and model
simulation results (average magnitude of head di fference
of 7.1 ft).23 Detailed discussion and analyses of calibra-
tion procedures and results are provided in the Chapter C
report (Faye and Valenzuela In press 2007).
"A location with observed data (for example. water level or concentration)
that i, as,nciated with a model location for the purpo,e of comparing ob,er,ed
data with model results.
1' Definition, of head difference. average magnitude of head difference.
and other calibration target, and >latistics arc provided in Table A8.
25 ,,.-,--,-,--,--,-,----,-,c--,---,---,-r---r-,-r---r-t~T""""T~-r-r.-v-r--r-r,,
20
15
10
t;:; 5 w LL
z
_j
a. Predevelopment (steady state)
/
•
/
/.
/
/
/
/• .. .,
.. / • ·/
/ . ),'. . /. . . . . . / . . /
: /
/
/
/
• /
•/
•
w > w ....J 0 "------'---'--"'-..L.......J'----'-~~.,__,__.___..._...,___L-..J._....,_-'---..L.......Jc.......l.~~ .......... __,
a: w
0 10 15 20 25
~ 30 fTT,CTT'rrn--rr,crn--rr,..,...,..,..,...,..,..,...,..,.......,."TTTTTTTTTTT"TTTT.,..,..,..TTTTTTr,-,rrr,,
3:
Cl w ~ ~ 20
~
-10
b. Transient, 1951-1994
,
,
.j.
-20 -10 0 10 20
/
/ /
OBSERVED WATER LEVEL, IN FEET ABOVE OR BELOW H
NATIONAL GEODETIC VERTICAL DATUM OF 1929
EXPLANATION
--Line of equality
- -Calibration target, monitor-well data ± 3 feet
-----Calibration target, supply-well data ± 12 feet
• Monitor well and simulated paired data poim
+ Supply well and simulated paired data point
30
Figure A10. Obs erved and simulated water levels, model
layer 1, and calibration targets for (a) predevelopment
(steady-state) conditions and (b) transient conditions,
1951-1994, Tarawa Terrace and vicinity, U.S. Marine Corps
Base Camp Lejeune, North Carolina.
A24 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
----------------------------Hierarchical Approach for Ouantifying Exposure
To assess the calibration of the fate and transport
simulation of PCE and the mixing model computations
for finished water at the Tarawa Terrace WTP (hierarchi-
cal levels 3 and 4 ). a statistic referred to as the model
bias was computed (8 , Table A8). Model bias allows m
one to test the accuracy of a model by expressing the
bias in terms of a simulated-to-observed (or measured)
ratio (Maslia ct al. 2000. Rogers et al. 1999). Model bias.
defined as the ratio of simulated PCE concentration to
observed PCE concentration (C,;,,,/c,,,, ). is character-
ized by the following properties:
when csim I c,,b5 < I. there is
underprediction by the model.
when C~im/C0b5 =I.there is exact
agreement, and
when CSilll I cob~ > I, there is
overprediction by the model.
Data used to compute model bias are spatially and
temporally disparate and arc listed in Table A9 for water-
supply wells and Table A IO for the Tarawa Terrace WTP.
The geometric bias (Bg) is the geometric mean of the
individual c_,.1111 /C,,1,.,• ratios and is a measure of model
bias (B ). Geometric bias, (B ). is computed using the
//!,! g --
fol lowing equation:
where
B Ill,/
N
N
I:1n(B,,,;)
B =exp_,;-"''---
g N (I)
is the model bias defined as the ratio
of simulated PCE concentration to
observed PCE concentration ( C,;,,, / C,,,,, ).
is the number of observation points.
In () is the Napcrian or natural logarithm, and
B is the geometric bias. g ~
The geometric bias is used hccausc the distribution of
csiml cobs ratios is skewed like a lognormal distribution.
That is, the values are restricted for undcrprediction
(0-1 ). but arc unrestricted for overprediction (anything
greater than I).
Water-supply well data included 17 of 36 samples
recorded as nondetcct (Table A9). and these samples
were not used in the computation of the geometric
bias (B)-In addition, the computation of geometric bias
was accomplished twice; an inclusive bias computation
that included all water-supply well data and a selected
bias computation that omitted data for water-supply
Chapter A: Summary of Findings
well T1°23. The inclusive geometric bias. using data
for water-supply well TT-23. was 5.9. The selected
geometric bias. omitting data for supply well TT-23.
was 3.9 (Table AS). Both results, however. indicate over-
prediction by the model. The rationale for computing the
selected geometric bias is based on data. observations.
and discussions provided in Chapter E of this report
series (Faye and Green In press 2007). Briefly, enhanced
biodegradation possibly occurred in the vicinity of
water-supply well TT-23 during I 984 and 1985. A bio-
degradation rate for PCE of 0.5/d was computed using
analytical results and sample collection dates reported
for water-supply well TT-23. This rate probably was not
representative of biodegradation occurring in contami-
nated aquifer media at other wells and was significantly
greater than the calibrated reaction rate of 5.0 x I0-4/d
(Table A I I). Such greatly enhanced biodcgradation
would result in much lower PCE concentrations in water
samples obtained from supply well TT-23. A second
reason for computing a selected geometric hias-
omitting data from water-supply well TT-23-is bias
introduced into analytical results caused by incomplete
or inadequate sampling methodology. As noted in
Table A9, four sequential sampling events took place
during March 11-12. 1985. at water-supply well TT-23.
Each sampling event resulted in increased PCE con-
centrations compared to the preceding sample. Thus,
sampling methodology at water-supply well TT-23
may not have included a sufficient volume of water
discharged from the well bore prior to sampling, and
samples obtained did not represent PCE concentration
within the entire volume of aquifer material contributing
to the well.
For the Tarawa Terrace WTP. 15 of 25 samples were
recorded as nondetect (Table A I 0). The nondetect sam-
ples were not used in the computation of the geometric
hias (B.). The resulting geometric hias computed for
measured data at the Tarawa Terrace WTP is 1.5.
which indicates a slight overpredietion by the model.
All data. measured and nondetcct, and simulated
values arc displayed in Figures A 11 and A 12 for water-
supply wells and the WTP. respectively. The sample
numbers shown on the horizontal (x-) axis of each graph
correspond to the sample numbers listed in Table A9
for water-supply wells and Table A IO for the WTP. The
data in Figures A 11 and A 12 are compared with the
corresponding PCE concentration calibration targets
for water-supply wells and the WTP listed in Table AS.
A25
Hierarchical Approach for Quantifying Exposure----------------------------
Table A8. Summary of calibration targets and resulting calibration statistics for simulation models used to reconstruct historical
contamination events at Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina.
2
3
4
!'redevelopment (no pumping)
groundwater llliw
Transient groundwater flow-
monitor wells
Transient groundwater llow-
supply wells
Conlaminant fate and transport-
supply wells
Mixing model-treated water at
water treatment plant
Magnitude of head difference: 3 feet
Magnitude of head difference: 3 feet
Magnitude of head difference: 12 feet
Concentration difference:± one-half
order of magnitude or model bias (N,,,)
ranging from 0.3 to 3
Concentration difference:± one-half
order of magnitude or model bias (B,)
ranging from 0.3 to 3
lill,l=1.9r1
a=l.Sft
I/MS= 2.1 ft
lill,I= 1.4ft
rr=0.9ft
I/MS= 1.7ft
l6hl=7.I ft
a::: 4.6 ft
RMS= 8.5 ft
Geometric bias
Geometric bias
B = 1.5 '
1 Refer to the Chapter C report (Faye and Valenzuela In pn.:ss 2007) for calibration procedures and det:iils on le\'eb 1 and 2
"Refer lo the Chapter F report (Faye In press 2{){l7h) for calibration procedures and details on levels 3 :md 4
59
263
526
.lJ lead difference is dctined as observed water level (h ,i, ) minus simulated water \c\'el (Ii ): ,\lagnitude of head difference is de tined :is:
lt-.hl= lh,,1,, -h,,ml: a C(111ecntr:1ti()Jl diff1cr1c11c1c of± (111c-h,:1r'(m!cr of 111agnitudc e(JUates t(l a ;;~(Jdcl bias 11f (l.3 !(l 3. where. R,., = 1ll(1del bias and is dclincd as:
I/.,= C,,,,,IC,,i,,· where C,,,,, is the simulated concen1r:i1iun and C,.0, i.\ the observed concentration: when H,,, = 1. the model exactly predicts the observed
corn.:entration. when II,,, > I. the model overpredicts the cnncentration. and when nm < I. the model underpredicts the concentration
-I ' t(6i,, -6i,)' ~ Avcr:1ge magnitude of head difference is de lined as: lt111j = -I]t111.j : standard deviation of head difference is ddined as: a= \l~'""'----
N ,., ' [''' ll N-1
where L1/1 is the r.ne,t(::(t:'
1
:1.·:)1c or head di.ffcre11ce: root-mean-square of head difference is defined as: !?MS= N f L~h,2 : geometric hia\, H~, is
dc1incd as· n = exp[ ,-1 . where In ( ) is the Napcrian log;irithm
" N
~ A paired data point i" dclincd as any location with observed data that is associated with a model location for the purpose of rnmp;iring observed data with
model results rm water level ur cor1Cc11tration
''I/~= 5.8 computed using all water-.\U]lply wells listed in table NJ: 01 = 3.9 computed without rnnsidering water-supply well Tl"-23-Sec text for explanation
7 Observed concentration of 17 samples recorded as nondetect (sec Table AlJ) and arc not used in computation of geometric hias
~ob~crved conce111ration of 15 samples recorded as nondetecl (sec Table A I 0) and arc nol used in comput:ltion of geometric bias
For the nondetect sample data, the upper calibration
target was selected as the detection limit for the sample
(Tables A9 and A I OJ, and the lower calib'ra1ion target
was selected as I ~tg/L. The statistical analyses sum-
marized in Table A8 and comparisons of observed
data, simulated values, and calibration targets shown in
Figures A I Oa. A I 0/J. A 11. and A 12 for the four hier-
archical levels of model analyses provide evidence that
the models of groundwater flow (predcvelopment and
transient-Figure A I 0), contaminant fate and trans-
port (Figure A 11 ), and water-supply well mixing at the
Tarawa Terrace WTP (Figure A 12) presented herein:
(I) are reasonably calibrated and (2) provide an accept-
able representation of the groundwater-now system. the
fate and transport of PCE. and the distribution of PCE-
contaminatcd finished water to residences of'ntrawa
Terrace. A listing of calibrated model parameter values
for the predevelopment (hierarchical level I), transient
(hierarchical level 2). and fate and transport (hierarchical
level 3) models is presented in Table A 11.
A26 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
Hierarchical Approach for Quantifying Exposure
Table A9. Summary of model-derived values and observed data of tetrachloroethylene at water-supply wells, Tarawa Terrace,
U.S. Marine Corps Base Camp Lejeune, North Carolina.'
[PCE, 1c1rachloroethylene: pg/L microgram per liter: J. estimated: ND. nondetect]
Model-derived value Observed data
Month and year PCE concentration, Sample date PCE concentration, Detection limit, Calibration tar-Sample
in µg/L ,• . in µg/L in µg/L gets', in µg/L number'
Supply well TT-23
January 1985 254 I /I 6/1985 132 10 41.7--417
February 1985 253 2/12/1985 ]7 10 11.7-117 2
February 1985 253 2/19/1985 2(1.2 2 tU-82.9 ]
February 1985 253 2/19/1985 Nil Ill 1-111 4
March 1985 265 ]/11/1985 14.9 Ill 4.7--47.1 5
~-larch 1985 265 3/11/1985 16.6 2 5.2-52.5 6
~·larch 1985 265 ]/12/1985 40.6 Ill 12.8-!28 7
March 1985 265 ]/12/1985 48.8 10 15.4-154 8
Apri! 1985 274 4/9/1985 ND 10 1-10 9
September !985 279 9/25/1985 4J 2 1.3-12.6 111
July 1991 191 7/11/1991 ND 10 1-10 11
Supply well TT-25
February 1985 7.J 2/5/1985 ND Ill 1-111 12
April 1985 9.6 4/9/1985 ND Ill 1-111 IJ
Scph.:111hcr 1985 IX.I 9/25/1985 II.HI Ill 11.14-1.4 14
October ! 985 20.4 10/29/1985 ND Ill 1-111 15
November 1985 22.8 11/4/1985 ND 111 1-1 II 16
i\'ovember l 985 22.8 11/12/1985 ND Ill 1-111 17
Decemht.:r 1985 25.5 12/3/1985 NIJ Ill 1-111 18
Jul , 1991 72.7 7/1 i/1991 23 Ill 7.J-72.7 19
Su~pjy well TT-26
January !985 8114 1/16/1985 1.5811.11 Ill 5110-4,996 20
January 1985 804 2/12/1985 J.8 Ill !.2-12 21
February 1985 798 2/19/1985 64.0 111 20.2-202 22
February 1985 798 2/ 19/l 985 55.2 Ill 17.5-175 23
April 1985 8111 4/9/1985 630.0 Ill 199-1,992 24
June 1985 799 6/24/ l 985 1,160.0 Ill ]67-3,668 25
September 1985 788 9/25/1985 1,1011.0 Ill 348-].478 26
Julv !991 670 7/1 l/1991 3511.11 Ill 111-1.1117 27
Supply well TT-30
Fehruar ' 1985 11.ll 2/6/1985 ND 10 I Ill 28
Supply well TT-31
Februarv 1985 0.17 2/6/ ! 985 ND 111 1-10 29
Supply well TT-52
Fehruar ' 1985 0.0 2/6/1985 ND Ill I Ill 30
Supply well TT-54
February 1985 6.0 2/6/1985 ND Ill 1-111 31
July !991 ]11.4 7/1 l/1991 Nil 5 1-5 32
Supply well TT-67
Februarv ! 985 4.1 2/6/1985 ND 10 I 10 33
Su~ply well RWl
~y 1991 0.11 7/12/1991 NIJ 2 1-2 34
Supply well RW2
Julv 199! 879 7/l2/l991 760 2 240-2.403 35
Supply well RW3
Jul, 1991 11.0 7/12/1991 ND 2 1-2 36
1 Model-derived values for waler-supply wells hased on simulation re.,ults ohtaincd from the fate and transport model MT]DMS (Zheng and Wang 1999):
sec the Chapter F report (Faye In pres~ 2007b) for details
!Calibration targets arc ±\12-ordcr nf magniwde for observed data: when ohscrvcd data arc indicated as ND. upper calibration target is detection limit and
lower calibration target is I pg/L
·'See Figure A 11
Chapter A: Summary of Findings A27
Hierarchical Approach for Ouantifying Exposure----------------------------
Table A10. Summary of model-derived values and observed data of tetra chloroethylene at the water treatment plant, Ta raw a Terrace,
U.S. Marine Corps Base Camp Lejeune, North Carolina.'
[l'CE. tetraehloroethylenc: 11g/L microgram per liter: i'\'D. nundetect]
~fay ! 982 148 5/27/1982
July 1982 112 7/28/1982
July 1982 112 7/28/1982
July 1982 112 7/28/1982
January 1985 176 2/5/1985
January 1985 176 2/11/1985
February I 985 3.6 2/13/1985
February 1985 3.6 2/19/1985
Fchruary 1985 3.6 2/22/1985
March 1985 8.7 3/11/1985
March 1985 8.7 3/12/1985
March 1985 8.7 3/12/1985
April 1985 8.1 4/22/1985
April 1985 8.1 4/23/1985
April 1985 8.1 4/29/1985
~foy 1985 4.8 5/15/1985
July 1985 5.5 7/1/1985
July I 985 5.5 7/8/1985
.luly1985 5.5 7/23/1985
July 1985 5.5 7/31/1985
August 1985 6.0 8/19/1985
September 1985 6.5 9/1 I /1985
September 1985 6.5 9/17/1985
Scplcmbcr 1985 6.5 9/24/1985
Oc1ohcr 1985 7.1 I0/29/1985
80
104
76
82
80
215
ND
ND
ND
ND
6.6
21.3
ND
3.7
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
10
10
10
10
10
10
10
1
2
10
10
I 0
10
10
10
10
10
10
10
10
10
10
10
10
25-253
33-329
24-240
26-259
25-253
68-6811
1-10
1-2
1-111
1-2
2.1-21
6.7-67
0.3-3.2
1-10
1.2-11.7
1-10
1-10
1-10
1-111
1-10
1-10
1-10
1-IO
1-10
1-111
. Sailiple
n·u1Jlber3
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
1 Modd-dcrived values t'or watc1· tn.:::11111..:nl pl;ml hased (Ill si111ulation results (1h1ai11ed from th..:: fate and transport llHl(!t::I MTJDMS (Zhrng a11d \Vang I()()())
and application ()f :1 material.\ mass hal:mcc ( mixing) model; sec th.: Chapter F rcpurt (Faye In press 2007b) for details
'Calibration targeb are ±½-order of magniwde for observed J:na: when observed data are indicated as ND. upper calibration target i\ detection limit and
lower calibration target is 1 ftg/L
-'See Figure A 12
A28 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
----------------------------Hierarchical Approach for Quantifying Exposure
Table A11. Calibrated model parameter values used for simulating groundwater flow and contaminant fate and transport,
Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina.
I fl.Id. fom per day: I'! 'Id. cu hie foot pa day: ft'/g. cuh1, foot per gram: gift'. gram per cubic foot: d 1• I/day: g/d. gram per day: ft. foot: ft'ld. ,qu,1rc foot per day:
-. not applicable I
Model parameter1
Model layer number2
2 3 4 5
Predevelopment groundwater-flow model (conditions prior to 1951) ---
Hori,ontal hydraulic conductivity. K11 (ft.Id) 12.2-53.4 1.0 4.3-20.0 1.0 6.4-9.0
Ratio of vertit:al to horizontal hydraulit:
conductivity. K, IK11'
I :7.3
Infiltration (recharge). IR (inche, per year) 13.2
1:10 I :8.3 1:10 I: 10
Transient groundwater-flow model, January 1951-December 1994
Specific yield. s, 0.05
Storage coefficient. S 4.0x 10 J -t.O x 10---1 4.0x 10 J 4.0x lO J
Infiltration (recharge). IR (inches per year) 6.6-19.3
Pumpage. Q, (fl 'Id) Sec footnote' Sec footnote' ()
6
1.0
1:10
➔.ox 10 J
Fate and transport of tetrachloroethylene (PCE) model, January 1951-0ecember 1994
---
Distribution coclfo.:icnt. Kd (f'I 'lg) 5.0x 10" 5.0x 10-1' 5.0x 10 6 5.0x 10 6 5.0x 10 1' 5.0x 10"
Bulk density. ri_ (gift') 77.112 77.1 12 77.112 77.112 77.112 77.112
Effective poro,ity. 111 0.2 0.2 0.2 0.2 0.2 0.2
Reaction rate. r (ti 1) 5.0x 10---1 5.0x J0-4 5.0x 10 4 5.0x 10 ·• 5.0x 10' 5.0x 10 ·1
Ma,,-loading rate'. q,C, (gld) 1.200
Longi tudinal dbpcr,ivity. (\ 1. (ft) 25 25 25 25 25 25
Tranwerse dispcrsivity, (\ r(fl) 2.5 2.5 2.5 2.5 2.5 2.5
Venical disper~ivity. <\ \ (ft) 0.25 0.25 0.25 0.25 0.25 0.25
Molecular diffusion coefficient. D* (f'i1/dJ 8.5 x JO-' 8.5x 10---1 8.S x 10---1 8.5 x 10' 8.5 x lO J 8.5 x l0 '
'Symbolic notation u,ed 10 describe model paramc1er, obrnined from Chiang and Kir11clbach (200 I)
7
I : I 0
4.0 x lO '
()
5.0x lO "
77.11 2
0.2
5.()X J{) I
25
2.5
0.25
8.S x lO J
'Rcfrr to Chap1er B (Faye In press '.!007a) and Chap1er C (Faye and Valenzuela In pre,, 2007) reports for geohydrologic framcworl. corre,pond111g 10
appropriate model la)er,: aquifer, arc model layer, I. 3. 5. and 7: ,emiconfining uni1, arc model layer, 2. 4. and 6
'For model cell, ,i111ula1ing wa1er-,upply well,. vcni,al hydraulic conduc1ivi1y { K,) equals I 00 feel per day 10 approximate 1hc gravel pack around 1hc "ell
1 Purnpagc varie, by 1110111h. year. and mod<!! layer: r~kr to Chap1cr K rcpo11 (Ma,lia el al. In press 2007a) for ,pccific pumpagc da1a
'lntroduc1ion of con1arninant nw,, hcgan Januar> 1953 and terminated December 1984
Chapter A: Summary of Findings A29
Hierarchical Approach for Quantifying Exposure ----------------------------
A30
cc w I-::::;
cc w 0..
Cl)
::E <{ cc
~ 0 cc u ~ z
z·
0 ~ cc 1-z w u z
0 u w u
0..
10,000
1,000
100
10
0.1
0
4,996 3,668
\ 3,478
1.992 I
~1,107
• '!' I!' 1 I I
500 I I ... , •••• I 154 I ■
I I 367 348
2021;5
I I 117 I
82.9
41.7
11.7
8.3
: 128 199
I
47,.1 52.5
72.7 111
• I
10 10 10 10 10
4.7 5.2
1.4
1.3
1 1 1 1 1 1 1 1.2
0.14
10 15 20 25
SAMPLE NUMBER (SEE TAB LE A9l
EXPLANATION
Observed sample data
Upper calibration target, 417
1/2-order of magnitude
higher than observed value
Nondetect sample data
10 Detection limit and upper
calibration target
I I I
I I I
I I I
I I •
• •
30
Simulated concentration of PCE
Observed concentration of PCE
Simulated concetration of PCE
Lower calibration target, 41.7
1/2-order of magnitude
lower than observed value
Lower calibration target
I
■ Simulated concentration is 0.0 µg/L,
plotted as less than 0.1 µg/L for display
and relative comparison purposes
Solid line, calibration target range
Dashed line, groups data with unique
sample identification
1 1 1
2,403
240
2 2
i i
• •
35
Figure A11 . Comparison of observed and nondetect tetrachloroethylene sample data with calibration
targets and simulated concentrations at water-supply wells, Tarawa Terrace, U.S. Marine Corps Base
Camp Lejeune, North Carolina. [PCE, tetrachloroethylene; µg/L, microgram per liter]
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
----------------------------Hierarchical Approach for Quantifying Exposure
cc w ~ ::::;
cc w 0..
(/)
~ <t: cc
(.!J
0 cc u ~ z
z·
0 ;::::
<t: cc ~ z w u z
0 u
w u 0..
329
253 240 259 253
100
67
68
33
25 24 26 25 21
11.7 10 10 10 10 10 10 10 10 10 10 10 10 10
10 ---l -!-
! 1 I I ! r ! i i T T Maximum contammant level 6.7 I -- -- --- --- - - ----_J -I
■ I 3.2
2 2
r r 2.1
1.2
0.3
0.1 '----'------'-...L..-'---'---L-'---'------'-...L..-'---'---L-'---'------1-...L.._l_ _ _j___L_1--L---1-...L.._l___.j
0 5 10 15
SAMPLE NUMBER (SEE TABLE AlO)
EXPLANATION
Observed sample data
Upper calibration target, 240
1/2-order of magnitude
higher than observed value
20
Simulated concentration of PCE
Observed concentration of PCE
Nondetect sample data
10 Detection limit and upper I "' ,,,.., , .. .,,.,
Simulated concentration of PCE
Lower calibration target, 24
1/2-order of magnitude
lower than observed value
1 Lower calibration target
Solid line, calibration target range
Dashed line, groups data with unique
sample identification
25
Figure A12. Comparison of observed and nondetect tetrachloroethylene sample data with calibration
targets and simulated concentrations at the water treatment plant, Tarawa Terrace, U.S. Marine Corps Base
Camp Lejeune, North Carolina. [PCE, tetrachloroethylene]
Chapter A: Summary of Findings A31
Selected Simulation Results -----------------------------------
Selected Simulation Results
Examples of simulation results showing the distri-
bution of PCE in groundwater and the concentration of
PCE in finished water at the Tarawa Terrace WTP are
presented in the form of maps and graphs. Maps show
simulated water levels, directions of groundwater flow,
and the areal distribution of PCE. The conce ntrations of
PCE at specific water-supply wells and in finished water
at the WTP are shown as graphs in the form of time
versus concentration.
Distribution of Tetrachloroethylene (PCE)
in Groundwater
Simulation results of groundwater flow and the fate
and transport of PCE are shown as a series of maps for
January 1958 (Figure A 13), January 1968 (Figure A 14).
December 1984 (Figure A 15), and December 1994
(Figure A 17).24 Each illustration is composed of two
maps. The upper map shows simulated potentiometric
levels (or water levels) and directions of groundwater
flow for model layer I throughout the entire active
model domain (for example, Figure A 13a). Ground-
wa ter flow is from highest to lowest potentiometric level.
The lower map (for example, Figure A 13b) shows an
enlarged area of the Tarawa Terrace housing area and the
site of ABC One-Hour Cleaners. This map shows simu-
lated potentiometric levels and the areal distribution of
PCE-contaminated groundwater. The lower maps show
simulated PCE values ranging from 5 µg/L to greater
than 1,500 µg/L. The values of PCE shown on the
maps-as igned a specific color to represent a concen-
tration range-are values of PCE that were simulated at
the center of a finite-difference cell that was part of the
numerical model's finite-difference grid.25 The simulated
PCE values shown in Figures A 13-A 17 were derived
by applying the inverse-distance weighting method to
simulated PCE-concentration values at the center of
finite-difference cells.
i, For ,ynop1ic map, of model layer I ( I 951 -1994). refer 10 1hc
Chapter K repon (Ma,lia Cl al. In press 2007a).
"Refer IO report Chap1er C (Faye and Valenzuela In press 2007) and
Chapter F (Faye In press 2007b) report, for deiail, ,pecilic 10 1he compu-
1a1ional grid and model boundarie; used 10 ,imulaie groundwater llow and
con1amina111 fa1e and 1ran,por1.
January 1958
With the onset of simulated pumping at water-
supply well TT-26 during January 1952, local cones of
depression are shown around all active supply wells. In
general, however, flow is toward Northeast Creek and
Frenchmans Creek (A 13a). Under the!>e flow condi-
tions, PCE migrated southeast from its source at the site
of ABC One-Hour Cleaners in the direction of water-
supply well TT-26 (Figure A 13b). The simulated PCE
concentration at water-supply well TT-26 during Janu-
ary 1958 was about 29 µg/L.26
January 1968
During January 1968, the designated start date of
the epidemiological case-control study (Figure A3).
groundwater flow in the northern half of the model
domain was little changed from January 1958 condi-
tions (Figure A 14a). In the immediate vicinity of the
Tarawa Terrace I housing area, groundwater now and
water levels are affected by pumpage from water-supply
wells TT-52, TT-53 and TT-54. Groundwater flow from
the vicinity of TT-26 toward well TT-54 is particularly
evident. Under these flow conditions, PCE has migrated
in a more southwardly direction from its source at the
site of ABC One-Hour Cleaners toward water-supply
we ll TT-54 (Figure A 14b) and covers a greater spatia l
extent th an during January 1958. By Janu ary 1968. the
simulated concentration of PCE in water-supply well
TT-26 was 402 µg/L.
December 1984
Groundwater pumpage increased water-level
declines during December 1984 in the vicinity of the
Tarawa Terrace I housing area and probably accelerated
the migration of PCE toward the vicinity of well TT-54
(Figure A 15a). Between January 1968 and Decem-
ber 1984, the center of mass of PCE migrated generally
southeastward from its source at the site of ABC One-
Hour Cleaners, and the arm of the PCE plume migrated
southwestward toward water-supply wells TT-23, TT-67.
and TT-54 (Figure A 15b ). The areal extent of simulated
,. Refer 10 1hc Chap1cr K report (Ma,lia ct al. In pre,s 2007a) for a
1110111hly I isling of simula1cd PCE conccmrmion, al wa1er-,uppl) wdl,
<.luring January 1952-f-cbruary 1987.
A32 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-----------------------------------Selected Simulation Results
11°23·30· 77'23 11°22·30· 77'22' 77°21 30 77'21'
77'22 30· 77°22' 77'21'30.
Base from U.S. Marine Corps and U.S. Geological Survey digital data files
EXPLANATION
Historical water-supply area
D Tarawa Terrace
D Holcomb Boulevard
--Model boundary
--Frenchmans Creek
-4 -Simulated potentiometric
contour-Shows simulated
potentiometric surface
during January 1958.
Contour interval 2 feet.
Datum is National Geodetic
Vertical Datum of 1929
► Simulated direction of
groundwater flow,
January 1958
0 ABC One-Hour Cleaners
TT-26• Pumping water-supply well
and identification
PCE concentration, in
micrograms per liter
1 to 5
Greater than 5 to 50
Greater than 50 to 500
• Greater than 500 to 1,500
• Greater than 1,500
Figure A13. Simulated (a) water level and direction of groundwater flow and (b) distribution of tetrachloroethylene, model layer 1,
January 1958, Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina. [PCE, tetrachloroethylene]
Chapter A: Summary of Findings A33
Selected Simulation Results -----------------------------------
77'23 30· 77'23' 77'22'30" 77'22' 77'21 30· 77'21
77'22'30" 77'22' 77'21'30.
EXPLANATION
Historical water-supply area
El Tarawa Terrace
D Holcomb Boulevard
--Model boundary
--Frenchmans Creek
-4 -Simulated potentiometric
contour-Shows simulated
potentiometric surface
during January 1968.
Contour interval 2 feet.
Datum is National Geodetic
Vertical Datum of 1929
Simulated direction of
groundwater flow,
January 1968
D ABC One-Hour Cleaners
r·26 Pumping water-supply well
and identification
PCE concentration, in
micrograms per liter
1 to 5
Greater than 5 to 50
Greater than 50 to 500
• Greater than 500 to 1,500
• Greaterthan 1,500
Base from U.S. Marine Corps and U.S. Geolog1cal Survey digi1al data files
Figure A14. Simulated (a/water level and direction of groundwater flow and (b)distribution of tetrachloroethylene, model layer 1,
January 1968, Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina. [PCE, tetrachloroethylene]
A34 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-----------------------------------Selected Simulation Results
77°23'30" 77°23' 77°22'30" 77°22' 77°21'30" 77'21'
2,000 4,000 FEET
• \ 500 1,000 METERS
77°22'30' 77°22' 77°21'30"
Base from U.S. Marine Corps and U.S. Geological Survey digital data files
EXPLANATION
Historical water-supply area
D Tarawa Terrace
D Holcomb Boulevard
--Model boundary
--Frenchmans Creek
-4 -Simulated potentiometric
contour-Shows simulated
potentiometric surface
during December 1984.
Contour interval 2 feet
Datum is National Geodetic
Vertical Datum of 1929
~ Simulated direction of
groundwater flow,
December 1984
D ABC One-Hour Cleaners
TT-26 Pumping water-supply well
• and identification
PCE concentration, in
micrograms per liter
1 to 5
Greater than 5 to 50
Greater than 50 to 500
• Greater than 500 to 1,500
• Greater than 1,500
Figure A15. Simulated (a) water level and direction of groundwater flow and (b) distribution of tetra chloroethylene, model layer 1,
December 1984, Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina. [PCE, tetrachloroethylene]
Chapter A: Summary of Findings A35
Selected Simulation Results ----------------------------------
PCE contam ination has increased significantly from the
areal extent of January 1958 and January 1968 (Fig-
ure!-. A 13b and A 14b, re pectively). By December 1984,
the simulated concentration of PCE in water-supply
wells TT-23, TT-25. and TT-26 was 255 µg/L. 6 µg/L.
and 805 ~tg/L. respectively. These and other water-supply
wells were pumping from model layer 3. Therefore,
simulated concentrations for these water-supply well!-.
are lower than the simulated PCE concentrations shown
in Figure A I Sb. For maps showing simulated PCE
concentration in model layer 3, refer to the Chapter F
report (Faye In Press 2007b). For information on model
layers that water-supply wells pumping from. refer to the
Chapter K report (M aslia et al. In Press 2007a).
Some water-supply wells were constructed to obtain
water from multiple water-bearing zones. Therefore, in
the model representation of these we lls, groundwater
can be withdrawn from more than one model layer.
For example, water-supply wells TT-3 1, TT-52, and
TT-54 withdraw groundwater from model layers I and
3. whereas water-supply wells TT-23. TT-25, TT-26.
TT-27, and TT-67 withdraw groundwater solely from
model layer 3 (Faye and Valenzuela In press 2007;
Maslia et al. In press 2007a). Consequently, the di tri-
bution of PCE will differ by model layer and by time,
depending on groundwater-now velocities. the number
of water-supply wells withdrawing groundwater from a
particular model layer. and the volume of groundwater
being withdrawn. An example of the multilayer distribu-
tion of PCE by model layer for December 1984 is shown
as a per pcctive diagram in Figure A 16. In this diagram.
water-supply wells are shown penetrating the model
layer or layers from which they withdraw groundwater.
Because no water-supply wells withdraw groundwater
directly from model layer 5, the distribution of PCE in
layer 5 covers a smaller area and is of lower concentra-
tion compared to model layers I and 3.
December 1994
Owino to documented PCE contamination in water 0
samples obtained from the Tarawa Terrace water-supply
wells and the WTP (Tables A9 and A I 0), wells TT-23
and TT-26 were taken off-line during February 1985.
The Tarawa Terrace WTP was clo!-.cd and pumping at
all Tarawa Terrace water-supply wells was discontinued
during March 1987 (Figures A3 and AS. Table A6).
As a result, potentiometric levels began to recover. By
December 1994. the simulated potentiometric levels
(Figure A 17a), were nearly identica l to predcvelopment
cond itions of 195 1 (Faye and Valenzuela In press 2007).
Groundwater flow was from the north and northwest
to the south and cast, di. charging to Northeast Creek.
Groundwater discharge also occurs to French mans
Creek in the westernmost area of the model domain
(Figure A 17a). Water-supply wells shown in Figure A 17
were not operating during December 1994, but are
shown on this illustration for reference purposes.
A graph showing simulated concentrations of
PCE at Tarawa Terrace water-supply wells from the
beg inning of operations at ABC One-Hour Cleaners
throuoh the closure of the wells and the WTP is shown I:>
in Figure A 18. Simulated PCE concentrations in water-
supply well TT-26 exceeded the current MCL of 5 µg/L
during January 1957 (simulated value is 5.2 ~lg/L) and
reached a maximum simulated value of 85 1 µg/L during
July 1984. The mean simulated PCE concentration in
water-supply well TT-26 for its entire period of operation
was 351 µg/L. The mean simulated PCE concentration
for the period exceeding the current MCL of 5 µg/L-
January 1957 to January 1985-wa!-. 4 14 µg/L. Thi~
represents a duration of 333 months (27 .7 years). These
results are summarized in Table A 12 along with simu-
lated results for water-supply we lls TT-23 and TT-25.
It should be noted that although !-.imulation results indi-
cate several water-supply wells were contaminated with
PCE (wells TT-23, TT-25, TT-3 1. TT-54, and TT-67),
by far, the highest concentration or PCE and the longest
duration or contamination occurred in water-supply
well TT-26 (Figure A 18).
A36 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-----------------------------------Selected Simulation Results
Model layer 1
Model layer 3
Model layer 5
EXPLANATION
Map Historical water-□ ABC One-Hour Cleaners PCE concentration, in area
supply area micrograms per liter
D Montford Point TT-54 ~ Water-supply well and 1 to 5
E:l Tarawa Terrace identification-Ground-Greater than 5 to 50
D Holcomb Boulevard water pumped from Greater than 50 to 500
model layer or layers • Greater than 500 to 1,500
Model boundary penetrated by well • Greater than 1,500
Figure A16. Diagram showing perspective views of the simulated distribution of tetrachloroethylene,
model layers 1, 3, and 5, December 1984, Tarawa Terrace and vi cinity, U.S. Marine Corps Base Camp
Lejeune, North Carolina. [PCE, tetrachloroethylene; thickness and vertical separation of layers notto scale]
Chapter A: Summary of Findings A37
7
Selected Simulation Results -----------------------------------
77°23'30" 77'23'
77°22'30" 77°22' 77°21'30"
EXPLANATION
Historical water-supply area
D Tarawa Terrace
D Holcomb Boulevard
--Model boundary
--Frenchmans Creek
-4 -Simulated potentiometric
contour-Shows simulated
potentiometric surface
during December 1994.
Contour interval 2 feet.
Datum is National Geodetic
Vertical Datum of 1929
Simulated direction of
groundwater flow,
December 1994
D ABC One-Hour Cleaners
r -26 Water-supply well
(not pumping!
and identification-
Shown for locational
reference only
PCE concentration, in
micrograms per liter
1 to 5
Greater than 5 to 50
Greater than 50 to 500
• Greater than 500 to 1,500
• Greater than 1,500
Base from U.S. Manne Corps and U.S. Geological Survey digital data files
Figure A17. Simulated (a) water level and direction of groundwater flow and (b) distribution of tetra chloroethylene, model layer 1,
December 1994, Tarawa Terrac e and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina. [PCE, tetrachloroethylene]
A38 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune. North Carolina
-----------------------------------Selected Simulation Results
a: UJ f-::::;
a: UJ a..
(/)
~ <! a:
C!)
0 a: u
~
z
z
0
~ a: f-
100
10
~ 0.1
u z
0 u
UJ u a..
0.01
Jan 1957
(5.2 µg/L)
AB
I I
Maximum contaminant level
Nov 1957
(5.4 µg/L)
o Finished water sample from
water treatment plant
A Well TT-26 begins operations
B ABC One-Hour Cleaners
begins operations
C Well TT-26 not ,n operation
July August 1980
D Well TT-26 not 111 operation
January-February 1983
E WTP closed March 1987
I I I
I I
I
I
I I
I
I
I
I
I I
I
I I
WellTT-26
r,' Well TT 23 0
E
,.-Well TT ~4
Well TT67
0
/Well TT-31
0.001 L...1._,_..L-I-L-'-J.......J._,_..L-1_.__._ ........ _,_..L-l_._-'-........ _..__...._.,._.__._._........_..L-l-'----'-L...J..-'-.l........l.-'-..L-IL...J.....J.......J
Jan
1950
Jan
1955
Jan
1960
Jan
1965
Jan
1970
Jan
1975
Jan
1980
Jan
1985
Jan
1990
Jan
1995
Figure A18. Concentration of tetrachloroethylene: simulated at selected water-supply wells
and in finished water at the water treatment plant, and measured in finished water at the
water treatment plant, Tarawa Terrace, U.S. Marine Corps Base Camp Lejeune, North Carolina.
[PCE, tetra chloroethylene; µg/L, microgram pe r liter]
Table A12. Summary statistics for simulated tetrachloroethylene contamination of selected water-supply wells
and the water treatment plant based on calibrated model simulation, Tarawa Terrace, U.S. Marine Corps Base
Camp Lejeune, North Carolina.
IMCL. maximum contaminant level: 11g/L. microgram per liter: WTP, water treatment plant: PCE. tctrachloroethylenel
Month and year and Month and year of maximum value Average
Water supply duration exceeding MCL 1 and maximum concentration, concentration.2
in pg/L in pg/L
TT-23 Augu,t I 984-April I 985 April I 985 252 8 months' 274
TT-25 July 1984-February 1987 February 1987 27 32 months 69
TT-26 January I 957-January I 985 July 1984 414 333 months" 85 1
WTP November 1957-February 1987 March 1984 70 346 months 183
1 Current MCL for PCE is 5 r1g/L. effective date July 6. 1992 (40 CFR. Section 141.60. Effective Date,. July I. 2()()2. ed.)
1For pcrio<l, exceeding 5 ftg/L when 11ater-,uppl) well 11a, operating
'Watcr-,upply well "17"-23 wa, 1101 operating during February 1985
'Water-supply well TT-26 wm, not operating July-Augu,1 1980 and January-February 1983
Chapter A: Summary of Findings A39
Selected Simulation Results ----------------------------------
Concentration of Tetrachloroethylene (PCE) in
Finished Water
Figure A 18 shows simulated PCE concentrations
in finished water delivered by the Tarawa Terrace
WTP. A monthly listing of simulated PCE concen-
trations also is provided in Appendix A2. PCE concen-
trations for the water-supply wells depicted in Fig-
ure A 18 are based on simulated monthly results for the
period of well operations (Figure AS, Table A6). PCE
contamination of water-supply well TT-26 was the
primary contributor to contamination in the finished
water of the WTP. When water-supply well TT-26
was temporarily shut down during July-August 1980
and January-February 1983 , the PCE concentra-
tion in finished water at the WTP was significantly
lower (Figure A 18). For example, during June 1980,
the simulated PCE concentration in finished water at
the Tarawa Terrace WTP was 126 µg/L, but during
July-A ugust 1980, the simulated PCE concentration
in finished water at the Tarawa Terrace WTP did not
exceed 0.8 µg/L. Furthermore, during December 1982,
the simulated PCE concentration in finished water at
the Tarawa Terrace WTP was 115 µg/L, but during
January-Febru ary 1983, the simulated PCE concentra-
tion in finished water at the Tarawa Terrace WTP was
1.3 µg/L. The PCE concentration of finished water at
the Tarawa Terrace WTP is less than the PCE concen-
tration of water-supply we ll TT-26 because the mix-
ing model uses water supplied to the WTP from all
wells--contaminated and uncontaminated.
For any given month during the historical recon-
struction period, the PCE concentration of finished
water at the Tarawa Terrace WTP was computed
using the following equations:
(2)
and
NWP
I:c,Q, c,v,p = , I (3)
QT
where
NWP
Q,
C ,
is the number of water-supply wells
simulated as operating (pumping)
during the month of interest.
is the simulated groundwater pumping
rate of water-supply well i,
is the total simulated groundwater
pumping rate from all operating
water-supply wells during the
month of interest,
is the simulated concentration for
water-supply well i , and
is the concentration of finished wa ter
delivered from the Tarawa Terrace
WTP for the month of interest.
Equation 2 is known as the continuity equation, and
Equation 3 describes the conservation of mass.
The simulated concentration of PCE in finished
water delivered by the Tarawa Terrace WTP first
exceeded the current MCL of 5 µg/L during Novem-
ber 1957-10 months after the PCE concentration
in water-supply well TT-26 exceeded the M CL
(Figure A 18). Using simulated water-supply well
concentrations and mixing model co mputations
(Equations 2 and 3), exposure to PCE-contaminated
drinking water that exceeded the current MCL
of 5 µg/L occurred for a duration of 346 months
(28.8 years)-November 1957-February 1987. A
summary o f dates and durations of PCE concentrations
at selected water-supply wells and in finished water
at the Tarawa TerTace WTP is provided in Table A 12.
Simulated values of PCE concentration in finished wa ter
of the WTP compare well w ith avai lable measured data
shown in Figures A 12 and A 18 and I isted in Table A I 0.
A40 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vic inity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------------------------Analysis of Degradation By-Products
Analysis of Degradation By-Products
Although exposure to contaminated drinking
water was eliminated after February 1987 due to the
closure of the Tarawa Terrace WTP during March 1987
(Figures A3 and A 18: Table A 12). measurable quantities
of PCE remained in the subsurface-at the source
(ABC One-Hour Dry Cleaners) and distributed within
the aquifer (Figure A 17b). For example, during
July 199 1, the PCE concentrations in water samples
obtained from off-line water-supply wells TT-25
and TT-26 were 23 µg/L and 350 µg/L, respectively
(Table A9). This mass of PCE in the subsurface con-
tinued to migrate and undergo transformation through
physical and biochemical processes such as volatili-
zation and biodegradation. As such, the potential for
exposure to PCE and its degradation by-products TCE/7
1.2-tDCE, and VC from a route other than ingestion
and inhalation of drinking water-such as inhalation of
soil vapors-continued beyond cessation of exposure to
drinking water after the closure of the Tarawa Terrace
WTP in March 1987 (Figure A3 ). To quantify histori-
cal concentrations of PCE degradation by-products in
groundwater and in soi l (vapor phase) requires a model
capable of simulating multiphase flow and multispe-
cies mass transport. For PCE. this complex analysis is
summarized herein.28
The degradation of VOCs in groundwater is a
transformation process from a parent compound (for
example, PCE) to degradation by-products such as TCE.
1,2-tDCE, and VC (Lawrence 2006, In press 2007).
Evidence of the transformation of PCE to degradation
by-products of TCE and 1.2-tDCE can be found in water
samples obtained January 16. 1985, from Tarawa Ter-
race water-supply wells TT-23 and TT-26. Laboratory
analyses of the water samples indicated concentrations
of PCE, TCE, and 1,2-tDCE of 132, 5.8. and 11 .0 µg/L,
respectively, for water-supply well TT-23 and concen-
trations of PCE. TCE. and 1,2-tDCE of 1,580. 57.0. and
92.0 µg/L, respectively, for water-supply well TT-26
(Faye and Green In press 2007). The simulation of the
"TCE abo i, used in ,omc dry-cleaning pnx;c"e,. However. ha,ed on
the dcpo.,ition from the owner or ABC One-I lour Cleaner, (Melts 200 I).
only PCE wa, u,ed at ABC One-Hour Cleaner,. Therefore. any TCE detected
in Tarawa TerTace water-,upply well;, or in WTP tini,hed water occurred
bccau,e of the degradation of PCE.
. ''Fora detailed di,cu,,ion of the anal) ,i, and ,imulation of PCE degrada-
11on by-product, at Tarawa Terrace and vicinity. refer to the Chapter G report
(Jang and Aral In pre" 2007).
Chapter A: Summary of Findings
fate and transport of PCE in groundwater, described in
the Chapter F report (Faye In press 2007b), accounted
for the degradation of PCE by applying a biodegrada-
tion rate to PCE du ring the simulation process. (The
biodegradation rate was determined from field data and
the calibration process !Faye In press 2007b].) This
transformation process typically is expressed in terms
of a rate constant or half-life. For example, in the fate
and transport simulations of PCE. the calibrated bio-
degradation (or reaction) rate for PCE was 5.0 x IO 4/
day (Table A 11 ). It is important 10 note, however, that the
basic chemical reaction package that is contained i n the
MT3DMS model was used to simulate a single-specie
and single-phase system (Zheng and Wang 1999). Thus.
as described in Faye (In press 2007b). MT3DMS was
used to simulate the transport and fate (biodegradation)
solely of PCE. To account for sequential biodegradation
of VOCs. parent-daughter chain reactions must be taken
into account in a multiphase environment (Zheng and
Bennett 2002). For example, in a four-species system, the
source (ABC One-Hour Cleaners) contains only a single
specie-PCE. As PCE migrates from the source. it under-
goes decay, and the decay product is TCE. TCE in turn
undergoes decay, and the decay product can be 1,2-tDCE.
1.2-tDCE is again biologically transformed into VC (Law-
rence 2006, In press 2007).29 T hus. to account for and to
simulate (I) parent-daughter chain reactions, (2) multi-
phase environments (water and vapor), and (3) fate and
transport in the unsaturated (above the water table) and
saturated (in groundwater) zones, a multispecies, multi-
phase modeling approach was required. For this purpose.
the TechFlowMP model code was used to simulate the
sequential biodegradation and transport of PCE and its
associated daughter by-products (TCE. 1,2-tDCE. and
V C) at Tarawa Terrace and vicinity.Jo
Using TechFlowMP. th ree-dimensional multispe-
cies, and multiphase simulations were conducted 10
quantify the fate and transport of PCE and its deg-
radation by-products from the source of the PCE
contamination-ABC One-Hour Cleaners. The same
model domain used for the MODFLOW-96 and
MT3DMS model simulations (Faye and Valenzuela
In press 2007, Faye In press 2007b) was used for the
,. Degradation pathways arc very complex procc"c~ that depend on
availability of microorganisrm :ind c111 ironmental condition,. Detail, arc
provided in LaMcnce (2006 In pre" 2007) .
"Tech Flow MP i;, a three-dimensional multi,pccie,. multipha~e ma" 1ran,-
po11 model developed by the Multimedia Environmental Simulation, Laborator}
at the Georgia ln,ti tutc of Technology. Atlanta. Georgia (Jang and Aral 2005).
A41
Analysis of Degradation By-Products -------------------------------
TechFlowMP model. Contaminants simulated using
this more complex model formulation were PCE and
its degradation by-products TCE, I,2-tDCE. and YC.
Parameter values calibrated using the MODFLOW-96
and MT3DMS models (for example, water-supply well
pumping rates, infiltration [recharge I rate, porosity.
dispersivity. and PCE biodegradation [reaction I rate)
were used in the TechFlowMP model simulations
(Table A 11 ). However, owing to the more complex set
of mathematical equations approximated by this model,
and because the contaminant ource was applied to both
the unsaturated and saturated zones (zones above and
below the water table, respectively), additiona l model
parameters were determined and assigned. Exa mples of
these parameters include: moi ture content; partitioning
coefficients for TCE, I,2-tDCE, and VC; and aerobic
(unsaturated zone) and anaerobic (saLUrated zone)
biodegradation rates for PCE. TCE, 1,2-tDCE. and YC.
Details on specific TechFlowMP model parameters and
their calibrated values are described in the Chapter G
report (Jang and A ral In press 2007).
Results obtained by conducting three-dimensional,
multispecies, and multiphase simulations are presented
herein in terms of (I) graphs of time versus concentration
of PCE and its degradation by-products (Figure A 19).
(2) a table listing summary statistics for PCE and its
degradation by-products (Table A 13), (3) maps show-
ing the distribution of vapor-phase PCE (Figure A20),
and (4) a table listing monthly PCE and PCE degrada-
tion by-products in finished water at the Tarawa Ter-
race WTP (Appendix A2). Figure A 19 shows graphs
of simulated concentrations of PCE and its degrada-
tion by-products-obtained by using the TechFlowM P
model-at water-supply well TT-26 and at the Tarawa
Terrace WTP. A lso shown on the graphs is the concen-
tration of PCE simulated using the MT3DMS single-
specie and single-phase model (compare Figure A 18
and Figure A 19). Simulated concentrations of PCE at
water-supply well TT-26 obtained using the TechFlowMP
model are slightly lower in value than PCE concentra-
tions obtained using the MT3DM S model (Figure A I9a).
This is to be expected because the TechFlowMP simu-
lations take into account flow and transport in both
the unsaturated zone (zone above the water tab le) and
saturated zone (zone at and below the water table) and
loss of PCE into the vapor pha e, whereas the MOD-
FLOW-96 and MT3DMS models consider groundwater
fl ow and contaminant fate and transport solely in the
saturated zone and in the water phase. Given the same
total ma s of PCE loaded into each of these models.
the PCE concentration at water-supply well TT-26 (and
other water-supply wells) will be simulated as a lesser
amount in the saturated zone by the TechFlowMP model
because a fraction of the mass is allocated to the unsatu-
rated zone, as well as being pa11itioned into the vapor
phase. Because water-supply well TT-26 was the primary
contributor of PCE contamination in fi nished water at
the Tarawa Terrace WTP (Figure A 18). the resulting
PCE concentrations in finished water at the Tarawa Ter-
race WTP computed using results from the TechFlowMP
model also were lower (Figure A 19b and Appendix A2).
Based on the TechFlowMP model simulations.
TCE. I.2-tDCE, and VC concentrations at water-supply
well TT-26 generally ranged from about 10 µg/L to
I 00 µg/L (Figure A 19a). Simulated concentrations of
TCE, I,2-tDCE. and YC in finished water at the Tarawa
Terrace WTP generally ranged from about 2 µg/L to
15 ~tg/L (Figure A I9b and Appendix A2). Comparison
or the simu lated concentrations of PCE degradation by-
products in finished water at the Tarawa Terrace WTP
indicate the following (Figure A I 9b):
I . TCE was below the current MCL va lue of 5 µg/L 11
for nearly the entire historical period except during
January I984-January 1985 when it ranged between
5 and 6 µg/L;
2. I,2-tDCE was below the current M CL value or
I 00 µg/L 31 for the entire historical period;
3. Y C was at or above the current MCL value of
2 µg/U1 from M ay 1958 through February 1985 at
which time water-supply well TT-26 was shut down.
Simulated concentration val ues of TCE in water-
supply well TT-26 and in finished water delivered by
the Tarawa Terrace WTP are less than simulated con-
centrations of YC and I,2-tDCE. This is in agreement
with measu red data obtained from water samples in well
TT-26 which shows a TCE concentration less than that
of I ,2-tDCE. Summary statistics of PCE and degrada-
tion by-product contamination of se lected water-supply
wells (TT-23, TT-25. and TT-26) and at the Tarawa Ter-
race WTP derived from simulations of the Tech Flow MP
model (based on three-dimensional mu ltispecies and
mu ltiphase simulation) are listed in Table A 13.
'1 40 CFR. Seel ion 141.60. Effcc1ivc Dale,. July I. 2002. ed.
A42 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------------------------Analysis of Degradation By-Products
a:
UJ I-
-' a: UJ a...
Cl)
~ et a:
(.!)
0 a: u
z
0
l-et
1,000
100
10
0.1
0.01
0.001
a. Well TT-26
MCL 101 l.2-1DCE
MCL lot
PCE I I('
I
I
I
I
Contaminant Model
-PCE } MT3DMS ~ ~.~~tDCE} TechFlowMP --vc
TCE
Well TT-26 not in operation:
July-August 1980 and
January-February 1983
Notes.
l. Simulation of PCE from MTJDMS model
described in the Chapter F !Faye 2007b) report
2. Simulation of PCE and degradation
by-products TCE, 1,2-tDCE, and VC from
TechFlowMP model descnbed in the
Chapter G !Jang and Aral 2007b} report
a:
1-b. Water treatment plant (finished water) t5 1,000 i:,---,--r,----,-.r,--.--,--,--,--.--,--.,...,.--,.,--r--,-.......,.---,--r,-,-.----,-.--,--,--.--,--,--.,...,.--,r,-,-,--,---,--,---,--,,
u z 0 u
Cl
UJ l-et -' ::,
~
Cl)
100 MCllor I ,2 1DCE
0 l
0.01
0 001
Jan
1950
MCL for VC
I
!
Jan
1955
I
Jan
1960
, ... --.. ____ _
Jan
I 965
--,---..... ---;-'
Jan
1970
Jan
1975
Jan
1980
Jan
1985
Jan
1990
Jan
1995
figure A19. Simulated concentration of tetrachloroethylene (PCE) and
degradation by-products trichloroethylene (TCE), trans-1,2-dichloroethylene
(1,2-tDCE), and vinyl chloride (VC) (a) at water-supply well TT-26 and (b) in finished
water from water treatment plant, Tarawa Terrace, U.S. Marine Corps Base Camp
Lejeune, North Carolina. [MCL, maximum contaminant level)
Chapter A: Summary of Findings A43
Analysis of Degradation By-Products --------------------------------
Table A13. Summary statistics for simulated tetrachloroethylene and degradation by-product contamination of selected water-supply
wells and the water treatment plant based on three-dimensional multispecies and multiphase model simulation, Tarawa Terrace,
U.S. Marine Corps Base Camp Lejeune, North Carolina.1
I MCL. maximum con1arnina111 level: µg/L. microgram per li1cr: PCE. 1e1rachloroe1hylcne: TCE. 1richloroc1hylene: I .2-1DCE. 1m11.,-1.2-dichloroc1hylene:
VC. vinyl chloride: Aug. Augusl: Scpl. Sep1ember: Nov. November: Mar. March: Feb. February: Jan. January: WTP. wa1er 1rea1mcn1 pla111I
Month and year exceeding Maximum concentration, Average concentration.3 Duration exceeding MCL.
Water MCL.2 in pg/l in pg/l in pg/l in months
supply 1.2· 1.2-1.2-1.2-PCE TCE tOCE vc PCE TCE tDCE vc PCE TCE tDCE vc PCE TCE tDCE vc
"IT-23 Aug Sept _,, Aug 167 7 2 1 13 143 7 -• 10 8 7 -" 8
1984 1984 1984
IT-25 Mar _6 _6 July 40 2 7 5 2 1 _6 -6 4 24 _6 -6 20
1985 1985
'IT-26 Feb Nov June Nov 775 :n 107 60 332 15 105 24 332 299 3 335
1957 1959 1984 1956
WTP Jan Feb _6 May 158 7 22 12 57 6 _6 5 332 11 6 -3 1 I
1958 1984 1958
'All ,imulalions .:onducled using lhc TechFlowMP model. Sec 1ex1 and 1he Chap1er G repo11 (Jang and Aral In press 2007) for dc1aib
'Curren! MC Ls arc: PCE and TCE. 5 ftg/L: I .2-1DCE. I 00 µg/L: and VC. 2 ftg/L (USEPA. 2003): effcc1ive dale, for MCL, arc a, follow,:
TCE :111d VC. January 9. 1989: PCE and 1.2-tDCE. July 6. 1992 (40 CFR. Section 141.60. Effec1ive Da1e,. July I. 2002. ed.)
'For periods exceeding MCL when waler-supply well opcra1ing
'Wa1cr-,upply well TI-23 wa, 1101 opera1ing February 1985
'Wa1cr-,upply well TT-26 was 1101 operaling July-Augus1 1980 and January-February 1983
''MCL never exceeded during simula1ion
Maps of the area l distributions of vapor-phase PCE
for December 1984 and December 1994 are shown in
Figure A20. The maps depicL simulated vapor-phase
PCE concentrations in soi I to a depth of about IO ft.
Concentration units for the vapor-phase PCE distribu-
tions shown in Figure A20 are in micrograms per liter
of air.'2 Comparing these maps with similar maps for
dissol ved-phase PCE in groundwater for model layer I
(Figures A I5/J and A 17b, respectively) indicates that
vapor-phase concentrations are lower than dissolved-
phase PCE concentration. by about a factor of 10 -15
for December 1984 and December 1994. The following
examples are noteworthy.
"To ob1ain air conccn1ra1ion unils or micrograms per cubic mc1er (mg/m ')
1ha1 are 1ypically u,ed for indoor air ,1udies. multiply micrograrrn, per li1er by
I 000 (refer to Conver,ion Fac1ors in Con1en1, ,cc1ion of thi, rc1xi11.
I . During December 1984:
a. the maximum simulated PCE concentration in
groundwater at family housing (model layer I)
was 638 µg/L (Figure A I5/J), whereas the max i-
mum simulated vapor-phase PCE (in the top
IO ft of soil) was 20 µg/L (Figure A20a): and
b. the maximum simulated PCE concentration in
groundwater (model layer I) at the Tarawa Ter-
race elementary school was 1.418 µg/L (Fig-
ure A I Sb). whereas the max imum simulated
vapor-phase PCE (in the top 10 ft of' soil) was
137 µg/L (Figure A20a):
A44 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------------------------Analysis of Degradation By-Products
77°22'30"
•
,.:\ r
I-~ I
""I r--.l
\ -.J ~.)
I. ,._ •
~ -·• "v '/
77°22'30"
77°22
77'22'
Tarawa Terrace
Elementary School 11 /"""
Base from U.S Marine Corps and U.S. Geological Survey d1g11al data files
77'21'30"
I.ODO
250
77'21'30"
Tarawa Terrace I
amily housing)
2,000 FEET
500 METERS
EXPLANATION
Historical water-supply area
D Tarawa Terrace
D Holcomb Boulevard
--Model boundary
D ABC One-Hour Cleaners
TT-26 Pumping water-supply well
• and identification
PCE vapor-phase concentration,
in micrograms per liter of air
1 to 5
Greater than 5 to 50
Greater than 50 to 500
• Greater than 500 to 1,500
• Greater than 1,500
Figure A20. Simulated distribution of vapor-phase tetrachloroethylene to a depth of 10 feet below land surface,
(a) December 1984 and (b) December 1994, Tarawa Terrace and vi cinity, U.S. Marine Corps Base Camp Lejeune,
North Carolina. [PCE, tetrachloroethylene]
Chapter A: Summary of Findings A45
Analysis of Degradation By-Products -------------------------------
2. During December 1994:
a. the maximum simulated PCE concentration in
groundwater at family housing (model layer I)
was 688 µg/L (Figure A l7b), whereas the maxi-
mum simulated vapor-phase PCE (in the top
10 ft of soil) was 44 µg/L (Figure A20h): and
b. the maximum simulated PCE concentration
in groundwater (model layer I) at the Tarawa
Terrace elementary school was 688 µg/L (Fig-
ure A 17b), whereas the maximum simulated
vapor-phase PCE (in the top IO ft of soil) was
56 µg/L (Figure A20b).
Due to sandy soils found at Camp Lejeune (including
Tarawa Terrace), there is potential for vapors from these
plumes (for example, Figure A20) to enter buildings,
thereby providing a potential exposure pathway from
inhalation of PCE and PCE degradation by-product
vapors. At Tarawa Terrace, these buildings would
include some family housing and the elementary school.
It is important to note that historical measurements
or soil vapor (soil gas) were not available. Therefore, the
TechFlowMP model parameters related to the simulation
of vapor-phase PCE and PCE degradation by-products
cou ld not be calibrated against field conditions. For
example. an assumption was made that homogeneous
vapor exit conditions ex ist at land surface throughout the
entire Tarawa Terrace area. Realistically, housing built
on concrete slabs, streets and parking lots paved with
asphalt. bare playground area • and lawns will each have
different vapor exit conditions requiring adjustment of
model parameters to those specific conditions. This may
seem like a limitation of the reliabil ity of vapor-phase
modeling results (for example, Figure A20). However.
the focus of the current investigation is on drinking-
water contamination and the historical reconstruction of
PCE and PCE degradation by-product contamination of
groundwater (water phase) and drinking water at Tarawa
Terrace. The concentration of PCE and PCE degrada-
tion by-products in groundwater significantly impacts
the vapor-pha e simulation results. Because simulated
groundwater concentrations are based on calibrated
groundwater-flow and contaminant fate and transport
models, the results presented for vapor-phase simula-
tions should be viewed as reliable historical estimates
of generalized vapor-phase conditions in soil during
December 1984 and December 1994 at a depth of about
IO fl (Figure A20). For present-day soil-gas conditions
or to obtain a more refined historical vapor-phase
calibration for Tarawa Terrace, field studies, including
the collection of unsaturated Lone. soil gas. and indoor
air concentration data would have to be undertaken as a
separate detailed study. Details regarding the develop-
ment of the Tech Flow MP model are provided in Jang
and Aral (2005). Assumptions. parameter values spe-
cific to three-dimensional multiphase flow and multi-
species mass transport, and resulting simulations of PCE
and PCE degradation by-products in groundwater and
vapor-phase specific to Tarawa Terrace and vicinity are
provided in Jang and Aral (2007) and in the Chapter G
report (Jang and Aral In Press 2007).
A46 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
---------------------------------Confidence in Simulation Results
Confidence in Simulation Results
Models and associated calibrated parameters
described previously are inherently uncertain because
they are based on limited data. Under such circum-
stances, good modeling practice requires that evaluations
be conducted to ascertain the confidence in models by
assessing uncertainties associated with the modelin g
process and with the outcomes attributed to models
(Saltelli et al. 2000). With respect to model simulations
at Tarawa Terrace, the availabil ity of data to thoroughly
characteri ze and describe model parameters and opera-
tions of water-supply wel ls was considerabl y limited, as
described in the section on Water-Distribution Inves tiga-
tion. Such limitations give ri se to the foll owing questions:
I. Could alternative water-supply well operating
schedul es or combin ations of model parameter
values provide acceptable simulation results
when compared to observed data and previously
estab lished calibration targets?
2. What is the reliability of the hi stori cal ly
reconstructed estimates of PCE concentration
determined usin g the calibrated models (for
example, results shown in Figure A 18)?
To answer these questi ons and address the over-
arching issues of model and parameter variabi lity and
uncertainty, three analyses were conducted using the
calibrated groundwater-flow and contaminant fate and
transport models described in Faye and Valenzuela
(In press 2007) and Faye (In press 2007b), respectively.
These analyses we re: (I) an assessment of pumping
schedu le variation at Tarawa Terrace water-supply
wells with respect to contaminant arriva l times and
concentrations,'1 (2) sensiti vity analysis,14 and
(3) probabilistic analysis.34 All of the additional analyses
were conducted using PCE dissolved in groundwater
as a single specie. MODFLOW-96 and MT3DMS cali-
brated models are described in the Chapter C (Faye
and Valenzuela In press 2007) and Chapter F (Faye
In press 2007b) reports.
"A detailed de>eription and discu,,ion of the effect of watcr-,upply
\\ell ,chedulc variation on the arrival ol PCE at water-,upply well, and
the Tarawa Terrace WTP i, prc,cntcd in the Chapter H report (Wang and
Aral In pre" 2007).
"A detailed de,cription and discus,ion of ,en,i1ivity and uncertainty
analy,b. including the u,e of Monte Carlo ,imulation i, pre,entcd in the
Chapter I rcpon (Ma,lia et al. In prc,s 2007h).
Chapter A: Summary of Findings
Water-Supply Well Scheduling Analysis
The scheduling and operati on historic~ of Tarawa
Terrace water-supply wells directly affected times and
concentrations of PCE in groundwater at wells and at
the WTP during 1952-1987. Thu s. simulated water-
supply well operations could be a major cause and
cont ributor to uncertainty and variabili ty with respect
to PCE arri val and PCE concentration at water-supply
well s and in finished water at the Tarawa Terrace WTP.
To assess the impact of pumping schedule variability
and uncertainty on groundwater-flow. contaminant fa te
and transpo11, and WTP mixing models. a procedure
was developed that combined groundwater simulation
models and optimization methods. This procedure is
described in detai l in the Chapter H report (Wang and
Aral In press 2007). The simulation tool developed
for this analysis-PSOpS (Table A4)-combines the
MODFLOW-96 and MT3DMS groun dwater simul ators
with a rank-and-assign optimi zation method developed
specili cally for the Tarawa Terrace analysis. This tool
optimizes pumping (operational) schedu les to minimiLe
or ma ximize the arrival time of contaminants at water-
supply wells. Based on the optimized operati onal sched-
ules. the concentration of a contaminant is recalculated,
and the effect of pum ping schedule vari ation on con-
taminant concentration and the arrival time of ground-
water exceeding the current MCL of PCE (5 µg/L) are
evaluated. It is important to note that in this analysi~.
with the exception of pumping rates, groundwater-flow
and contaminant transport model parameters were not
varied from their cal ibrated va lues (Tabl e A 11 ; Faye and
Valenzuela I In press 2007]; Faye I In press 2007b]).
Results of analyses using the PSOpS simulati on
tool to assess the effects of water-suppl y well pu mping
variation are presented graphically as a seri es of curves
of simul ated PCE concentration in finished water at the
Tarawa Terrace WTP versus time (Fi gure A2 I).'' The
calibration curve in Figure A2 I represents the same data
presented in Figure A 18 and represents the simul ated
concentration of fini shed drinking water delivered
from the Tarawa Terrace WTP-derived from analyses
described in the Chapter F report (Faye In press 2007b).
Calibrated model results indicate that PCE exceeding the
11 In the following di,cu"ion. rcfcrcn~c i~ made to location, ,lum n in
Figure A2 I. Thc,c location, arc labeled point, A-1. Thu,. in the en,uing
di,cu,siun for the section on ··Water-Supply Well Scheduling Analy,i,."" a
reference to a ,pccific location on the graph. for example. point A. refer,
,olely to Figure A2 I.
A47
Confidence in Simulation Results ---------------------------------
current MCL of 5 µg/L in finished water was delivered
from the WTP during November 1957 (point 8 ). By
determining an optimal combination of water-supply well
pum ping in terms of on-off operations and the volumetric
pumping rate. it would have been possible for PCE al
earlier than that reported for the calibrated MT3DMS
model. The e optimized arrival times are shown as
"Earliest arrival'' in Figure A2 l and are delined as the
·'Maximum Schedule·· in the Chapter H report (Wang
and Aral In press 2007). The results show an mTival date
11 months earlier-December 1956 (point A)-than the the 5 µg/L concentration to arrive at the WTP at a date
er: w I-::;
er: 100 w 0..
Cl)
~
<( er:
(!) 0 er: 10 u ~
~ z
0
~ er:
I-z w u z 0 u w z w -' 0 1 >-:c ~ 0 er:
0 -' :c u <( 001 er: ~ I-
A48
Jan
1950
Jan
1955
Jan
1960
E
Well TT-26 simulated
as not qperating for
Minimu(ll Schedule I
Jan
1965
Jan
1970
Jan
1975
F
Jan
1980
0
Jan
1985
Maximum
contaminant
level
Jan
1990
EXPLANATION
--Calibration-As described in
Chapter F (Faye 2007bl
Arrival-As described in Chapter H
(Wang and Aral In press 2007).
Late-Defined as Minimum Schedule I,
well TT-26 not operated
January 1962-February 1976
Earliest-Defined as Maximum Schedule
Late-Defined as Minimum Schedule II,
well TT-26 operated at least
at 25 percent of capacity
o Measured at water treatment plant
Simulation time
Location Month and year on graph
A December 1956
B November 1957
C February 1960
D June 1960
E January 1962
F February 1976
G November 1977
H June 1984
I February 1987
Figure A21. Sensitivity of tetrachloroethylene concentration in finished water at the
water treatment plant to variation in water-supply well operations, Tarawa Terrace,
U.S. Marine Corps Base Camp Lejeune, North Carolina. IPCE, tetrachloroethylene;
see text for discussion of points A-II
Jan
1995
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-----------------------------------Confidence in Simulation Results
calibrated arrival date of November 1957. Also notable is
the simulated concentration for January-February 1985
of 262 11g/L. This value (262 11g/L) exceeds the observed
value of 215 11g/L by 47 11g/L compared with the cali-
brated value of 176 µg/L (Table A I 0) that underes-
timates the observed value by 39 11g/L. Overall. the
"Earliest arrival'i simulation shows a higher concentra-
tion of PCE in linished water delivered from the Tarawa
Terrace WTP with a maximum value of 305 11g/L and
an average (for concentrations exceeding S ,ug/L) of
132 11g/L. The period during which the current MCL of
S 11g/L for PCE was exceeded under the "Early arrival"
scenario was 348 months (29 years).
The PSOpS simulation tool also was used to inves-
tigate a variety of other pumping scenarios by specifying
limiting values for such well properties as the maximum
or minimum pumping rate for a specific water-supply
well or group of wells. Two additional results are pre-
sented in Figure A2 I for simulations that specify mini-
mum operating rates for water-supply well TT-26-2S%
and 0% of total capacity."' The results of these simula-
tions show that when water-supply well TT-26 operated
al least at 25% of its capacity-identified as "Mini-
mum Schedule 11" in Figure A2 I and in the Chapter H
report-the arrival of groundwater contaminated with
PCE exceeding the current MCL (S 11g/L) was delayed
hy 27 months-February 1960 (point C)-whcn com-
pared with the calibrated arrival time of November 1957
(point B). A notable result occurs, however, when water-
supply well TT-26 is simulated as being shut down for
a period of time-identified as ·'Minimum Scheduler·
in Figure A2 I and in the Chapter H report. Based on
simulation results, water-supply well TT-26 could have
been taken out of service in January 1962 (point E) and
kept out of service until February I 976 (point F) with
the remaining water-supply wells still capable of meet-
ing all of the water demand during this period for Tarawa
Terrace and vicinity. During this time, water-supply well
TT-26 was modeled as being off-line, and the resulting
simulated concentration of PCE in finished water from
the Tarawa Terrace WTP ranged from 0 to less than
2 ftg/L. After February 1976 (point F). water-supply well
TT-26 had to be simulated as operating to meet increas-
ing demand. Thus, using the PSOpS simulation tool, it
·16 Using the PSOpS simulation tool, the operation of waler-supply well
Tr-26 was simulated as being shut down for a period of time-0% capacity-
and it was allowed to operate as low as 2YY,1 of its rated capacity at times.
,\ cornplctc listing of watcr-.\upply well capacity data is provided in the
Chapter C rep011 (Faye and Va!enwcla !n press 2007).
Chapter A: Summary of Findings
was possible lo simulate the operation of water-supply
well TT-26 in such a manner that the PCE concentra-
tion of finished water delivered from the Tarawa Terrace
WTP was below 5 ftg/L from January 1962 (point E)
through February 1976 (point F). Under this simula-
tion scenario-·'Minimum Schedule J"-the current
MCL was exceeded during the period June 1960
(point D)-December 1961 and for most months during
the period November 1977-February 1987 (points G
and I. respcctively).)7 Under the "Minimum Schedule I"
scenario. the maximum PCE concentration in finished
water at the Tarawa Terrace WTP was simulated as
41 11g/L during .lune 1984 (point H).
In summary. analyses of the variation in water-
supply well scheduling demonstrate that the current
MCL for PCE (5 µg/L) could have been exceeded in
finished drinking water delivered from the Tarawa Ter-
race WTP as early as December 1956 (point A) and no
later than June 1960 (point D). Because Tarawa Terrace
WTP records indicate that water-supply well TT-26 was
most likely operated routinely. the analysis also dem-
onstrates that the earliest time that finished water at the
Tarawa Terrace WTP exceeded the current MCL for PCE
of 5 ftg/L most likely occurred between December 1956
(""Earliest arrival"· scenario, point A) and November 1957
(calibrated arrival time. point B). The most likely maxi-
mum concentration of PCE in finished water ranged
between the ·'Earliest arrival" scenario maximum of
305 11g/L and the calibrated maximum of 183 ftg/L. The
mean concentration of PCE in finished water exceeding
the current MCL of S r1g/L most likely ranged between
the "Earliest arrival" scenario mean of 13 I 11g/L and
the calibrated mean of 70 µg/L. The analyses con-
ducted using the PSOpS simulation tool provide further
evidence that drinking water contaminated with PCE
exceeding the current MCL of 5 µg/L was delivered to
residents of Tarawa Terrace for a period ranging between
the "Earliest arrival" duration of 348 months and the
calibrated model duration of 346 months. This analysis
further indicates that the concentration of PCE in fin-
ished water delivered lo residents of Tarawa Terrace,
determined from the contaminant fate and transport and
mixing model analyses (Faye In press 2007b). are rea-
s,mable estimates of historical concentrations;
17 There were 103 month~ during the period November 1977~
February 1987. For 14 different months during this period. !he PCE
conccntra1im1 in tinished water at the Tarawa Terrace WTP was below
the current MCL of 5 pg/L ranging in value from 2.3 to 4.9 ftg/L
A49
Confidence in Simulation Results-----------------------------------
Sensitivity Analysis
Sensitivity analysis is a method used to ascertain the
dependency or a given model output (ror example. water
level or concentration) upon model input parameters
(ror example. hydraulic conductivity. pumping rate, and
mass loading rate). Sensitivity analysis is important for
checking the quality of the calibration of a given model,
as well as a powcrrul tool for checking the robustness
and reliability of model simulations. Thus. sensitivity
analysis provides a method for assessing relations
between information provided as input to a model-
in the form of model input parameters-and information
produced as output rrom the model. Numerous methods
arc described in the literature for conducting sensitivity
analysis (Saltelli ct al. 2000). For the Tarawa Terrace
models, selected model parameters were varied one at a
time from their respective calibrated values (Tahle A 11 ).
and the corresponding effect of this variation on the
change in the PCE concentration of finished drinking
water at the Tarawa Terrace WTP was assessed. ~x
In conducting the sensitivity analysis, all calibrated
model parameters-with the exception of pumpage-
were increased and decreased by factors ranging from
50% to 400% of their calibrated values (Tahlc /\ 14):'''
For example. horizontal hydraulic conductivity for
model layer I was varied by 90%. 110%. 150%. and
250% or its calibrated value: dispersivity was varied hy
50%, 200%, and 400% of its calihratcd value. Ground-
water-flow model parameters that were subjected to the
sensitivity analysis were:
• horizontal hydraulic conductivity or the
aquifers (model layers I. 3. 5. and 7).
• vertical hydraulic conductivity of the semi-
confining units (model layers 2. 4. and 6).
• inliltration (recharge) rate, and
• storage coefficients (includes specific yield
ror model layer I).
.lM Thi:-, particular approach to scnsitivi1y analysis is referred to as nne-at-a-
time (OAT) dc~ign:-, or experiments: delails can be found in Saltdli d al. (2000).
_1,, Tahle /\ 14 is a lis1 tif sclcc1cd para1nc1crs varied duri11g the sensilivity
an:1lysis. For a comple!e list and discussion of all parameters \'.tried. :-,cc the
Chapter I report (ivlaslia ct al. In press 2007h).
Contaminant rate and transport model parameters that
were subjected to the sensitivity analysis were:
• distribution coenicient.
• bulk density.
effective porosity.
• reaction rate.
mass-loading rate.
longitudinal dispcrsivity. and
• molecular diffusion.
Measures of the effect of varying the groundwater-
flow and contaminant fate and transport model param-
eters were quantified in terms of live computations:
(I) the date (month and year) when finished drinking
water at the Tarawa Terrace WTP first exceeded the cur-
rent MCL for PCE (5 ftg/L). (2) the duration (in months)
that linished drinking water at the WTP exceeded the
current MCL, (3) the relative change in these durations
(percent) caused by varying the calibrated parameter
values. (4) the maximum PCE concentration in linished
water at the Tarawa Terrace WTP. and (5) the relative
change (percent) in the maximum concentration. Results
for selected sensitivity analyses arc listed in Table A 14.
Recall that for calibrated model parameters, the date
that the PCE in finished water at the WTP lirst exceeded
the current MCL was simulated as November 1957.
and the duration that finished water exceeded the MCL
ror PCE was 346 months (Figure A 18. Table A 12).
Results of the sensitivity analysis show that some
parameters arc insensitive to change. even when varied
hy factors of 10 and 20. For example, large changes in
specific yield. storage coefficient. and molecular diffu-
sion resulled in very litlle change in simulated results
(Tahle A 14). Changes in other parameters-for example,
horizontal hydraulic conductivity for model layer I and
inliltration. using values that were less than calibrated
values-resulted in wells going dry during the simu-
lation process. Generally. increasing or decreasing a
calibrated parameter value by 10% (ratio or varied to
calibrated parameter value of 0.9-1.1) resulted in changes
of 6 months or less to the date that finished water first
exceeded the MCL ror PCE (5 11g/L). Complete details
pertaining to the use of the sensitivity analysis in rela-
tion to calibrated model parameter values and results
obtained from the sensitivity analysis arc discussed in
the Chapter I report (Maslia et al. In press 2007h).
A50 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
I I
t ' I I
I
---------------------------------Confidence in Simulation Results
Table A14. Summary of selected sensitivity analyses conducted on calibrated groundwater-flow and contaminant fate and transport
model parameters, Tarawa Terrace and vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina.'
[PCE. tetr:1chlorncthykne: MCL. nwxi1num con!aminant level: r1g/L. microgram per liter; ft/d, fool per day: ft1/g, cuhic fool per gram: gift 1. grams per cuhic
root: d-I, 1/day: g/d. grams per day: IWd. square fool per day:-. no! :ipplicabk: WTP. water 1rea1111ent plant)
Groundwater-flow model parameters
HoriZ(llltal hydraulic c1mduc-12.2-53.4 0.9 -' -' -' -' -'
tivity, layer I, K11 (ft/d) I.I Aug. 1957 351 1.4 196 7.0
1.5 Oct. 1956 365 5.5 223 22.0
2.5 Oct. 1955 377 9.0 209 14.1
HoriZ(llltal hydraulic conduc-43-20.0 0.0 (kt. 1957 ]48 O.h I 84 (LS
tivity, layer 3, K11 (t'l/dl I. I Nov. 1957 345 -0.3 182 -0.5
1.5 Feh. 1958 341 -1.4 17() -2.3
2.5 Jul. 1958 339 -2.0 187 2.1
Horizontal hydraulic conduc-6.4-9.0 0.9 Oct. 1957 347 0.3 I 85 1.2
tivity, layer 5, K11 (fl/d) I.I Nov. 1957 346 0.11 I 81 -1.0
JloriwrHal hydraulic conduc-5.0 0.9 Nov. 1957 346 0.0 183 -0.1
ti\"ity. layer 7. K11 (ft/d) I.I Nov. 1957 346 0.0 183 0.1
Infillration (recharge). IR 6.6-19.3 0.75 _, -' -' -' -' (inches per year) 1.25 Dee. 1957 343 -0.9 210 14.8
Specitic yield. S,. 0.05 I().() Nov. 1957 342 -1.2 182 -0.6
20.0 Nov. 1957 338 -2.3 178 -2.6
Storage coefficient. S 4.0xl0-1 10.0 Nov. 1957 346 0.0 183 -0.2
20.0 Nov. 1957 346 ll.(I 182 -0.3
Fate and transport model parameters
Distribution coefficient. Kd 5.0xl0-6 0.5 Apr. 1956 371 7.2 214 16.7
(l't'ig) 0.9 Jul. 1957 352 1.7 191 4.2
1.5 Jun. 1959 310 -10.4 165 -10.0
2.0 Dec. 1960 286 -17.3 14] -21.7
Bulk densi1y. f\ (g/fl1) 77.112 0.9 Jul. 1957 352 1.7 191 4.2
I.I .\far. l 958 338 -2.3 180 -1.8
Effective porosity. 111_. 0.2 0.5 Dee. 1956 363 4.9 :WJ 90.9
2.0 Sep. 1959 301 -13.0 86 -53.0
Reaction rate. r (d-1) 5.0x 10--1 0.5 Oct. 1957 349 0.9 294 60.4
2.0 Jan. 1958 326 -5.8 ')4 -48.7
Ma;-,s-loading rate :1. qsC.~ 1.200 0.5 May 1958 329 -4.9 92 -50.0
(g/d) 1.5 Aug. 1957 351 1.4 275 50.0
Longitudinal dispersivity. 25 ·{l.5 Apr. llJ58 337 -2.h 184 0.3
'\. (foot) '.W Mar. 1957 356 2.9 181 -1.0
-U) Jun. 1956 367 6.1 I 7(i -3.7
Molecular diffusion cocf-8.5x I 0-1 5.0 Nov. 1957 346 0.0 183 -0.1
fieient, D* (1't2/d) 10.0 Nov. 1957 346 0.0 183 -0.1
20.0 Nov. 1957 346 0.0 182 -0.3
1 See the Chapter I report (Masli:i et al. In press 2007b) for a complete listing of parameter;-, that were ;-,uhjected tu \"ariation in the sen;-,itivity analysi.~
2 Symho!ic notation used to de;-,crihe model parameters obtained from Chiang and Kinzelbach (200 I)
.1 For calibrated modcl. date finished w;1tcr at WTI' exceeded MCL l"m PCE i;-, NovL:mher I 957. tl11ra1ion of exceeding MCL is 14<i months. and
maximum PCE concentration i~ I 83 pg/1,-sce Ta hie A 12
-1current MCL for PCE is 5 pg/L (US EPA. 2003): effective date h1r :-.-tel, is July 6. 1992 (-lO CFR. Section 141.Nl. Effecti\'e Date"· July I. 2002. ed.)
I) -I)
·1Relative change in duration (N . l of finished water al the WTP exceeding the MCL for PCE is detined as: R,, --• --11 x 1()0%, where /J is 1he /h ' n <>
calihra1ed duration in mnnths (146) and n, is the duration in months for the sen~itivity analy~is using a varied parameter 11
"Relative change in l'.oncentration (N,., ) of fini;-,hed water at Tarawa Terrace WT!' exceeding .\!CL for PCE is detined as,: R(
the calibrated concentration in ftg/L ( 1 83) and C, is the l'CE coneen1ratio11 for the ;-,erbitivity analy;-,is using a varied parameter
7 Dry wells sirnulatcd i'or this sensitivity analysis
Chapter A: Summary of Findings A51
Confidence in Simulation Results---------------------------------
Probabilistic Analysis"'
A probabilistic analysis is used to generate uncer-
tainties in model inputs (for example, hydraulic conduc-
tivity or contaminant source mass loading rate) so that
estimates of uncertainties in model outputs (for example
water level or PCE concentration in groundwater) can he
made. Although the sensitivity analysis provided some
insight into the relative importance of selected model
parameters, a prnhahilistic analysis provides quantitative
insight ahout the range and likelihood (probability) of
model outputs. Thus. one purpose of a probabilistic
analysis is to assist with understanding and characterizing
variability and uncertainty of model output (Cullen and
Frey 1999). A number of methods are available for con-
ducting a probabilistic analysis. These methods can be
grouped as follows: (I) analytical solutions for moments,
(2) analytical solutions for distributions, (3) approxima-
tion methods for moments. and (4) numerical methods.
The prnhabilistic analysis conducted on the Tarawa
Terrace models used numerical methods-Monte Carlo
simulation (MCS) and sequential Gaussian simulation
(SGS)-to assess model uncertainty and parameter
variability. Readers interested in specific details about
these methods an<l about probabilistic analysis in general
should refer to the following references: Cullen and Frey
( 1999). Deutsch and .lourncl ( 1998). Doherty (2005),
US EPA ( 1997). and Tung and Yen (2005).
Ii is important to understand the conceptual dif-
ference between the deterministic modeling analysis
approach used to calihratc model parameter values hy
Faye and Valenzuela (In press 2007) and Faye (In press
2007b) and a probabilistic analysis. As described in
Maslia and Aral (2004). with respect to the approach
referred to as a deterministic modeling analysis, single-
point values arc specilied for model input parameters
and results are obtained in terms of single-valued output.
for example, the concentration of PCE. This approach
is shown conceptually in Figure A22a. In a probabilis-
tic analysis. input parameters (all or a selected subset)
of a particular model (for example, contaminant fate
and transport) may he characterized in terms of statisti-
cal distributions that can be generated using the MCS
method (USEPA 1997. Tung and Yen 2005) or the SGS
method (Deutsch and Journel 1998. Doherty 2005).
~0 A pmhahili.,tic analysis is defined a., an analysis i.n ~·l~ich frequency. (or
probability) dis1ributi1ms arc assignc~l.11i.rcprcscn_i \:anab.1111~• (or unc~rta1n1y)
in quantities. The output of a probabil1st1c analysis 1'> a d1stnhut1on (Cullen
,md Frey 1999).
Results are obtained in terms of distrihutcd-valuc output
that can be used to assess model uncertainty and parame-
ter variability as part of the probabilistic analysis (Fig-
ure A22b). MCS is a computer-based (numerical) method
of analysis that uses statistical sampling techniques to
obtain a probabilistic approximation to the solution of
a mathematical equation or model (USEPA 1997). The
MCS method is used to simulate probability density
functions (PDFs). PDFs arc mathematical functions that
express the probability of a random variable (or model
input) falling within some interval. SGS is a process in
which a field of values (such as horizontal hydraulic
conductivity) is obtained multiple times assuming the
spatially interpolated values follow a Gaussian (normal)
distribution. Additional details pertaining to the SGS
methodology arc provided in Deutsch and Journel ( 1998)
and Doherty (2005).
For the oroundwatcr-rlow and contaminant fate and b
transport models (Faye and Valenzuela In press 2007.
Faye In press 2007b), eight parameters were assumed
to be uncertain and variable: (I) horizontal hydraulic
conductivity. (2) recharge rate, (3) effective porosity.
(4) bulk density. (5) distribution coenicient, (6) disper-
sivity, (7) reaction rate, and (8) the PCE mass loading
rate. With the exception of dispersivity, these parameters
were selected for the probabilistic analysis because the
sensitivity analysis indicated that variation from the
calibrated value of the seven parameters resulted in the
greatest percentage change in the simulated concentra-
tion of PCE in finished water at the Tarawa Terrace WTP
(Tahle A 14 ). Dispersivity was selected for the probabilis-
tic analysis because it is a characteristic aquifer property
and represents the effect of aquifer heterogeneity on the
spreading of a dissolved contaminant mass (Schwartz and
Zhang 2003). Each of the aforementioned model param-
eters can be represented by a PDF such as a normal.
lognormal, triangular, or uniform distribution (Cullen and
Frey 1999). In the current analysis. a normal distribution
was chosen to represent each uncertain parameter (or
variant) with the exception of dispersivity. This variant
was represented by a lognormal distrihution. Statislics
associated with the normal and lognormal distributions
for the variants. such as the mean, standard deviation.
minimum. and maximum. are listed in Table A 15. The
calihrated value associated with each variant-derived
from model calibrations described in Chapter C
(Faye and Valenzuela In press 2007) and Chapter F
reports (Faye In press 2007h)-was assigned as the
A52 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
---------------------------------Confidence in Simulation Results
a. Deterministic analysis
Single-value
parameter input Single-value output
Model
V
D
z 0 !;; f---•
cc
C =flV, □. R, WI f-z
R
w
~ u z
0
u~--~--
DISTANCE
b. Probabilistic analysis
Parameter probability
density function (PDF)
Distributed-value output
(uncertainty and variability)
Model
C = flV, □, R, WI
CONCENTRATION
Figure A22. Conceptual framework for (a) a deterministic analysis and (b) a probabilistic analysis
(from Maslia and Aral 2004).
mean value of the distribution associated with each
variant. Examples of PDFs generated for recharge,
mass loading rate, and dispcrsivity compared with the
appropriate theoretical distribution are shown in Fig-
ure A23a, A23/,. and A23c, respectively. Two points are
noteworthy: (I) for a normal distribution (Figure A23a
and A23b). values for the mean. mode, and median
are equal, whereas for a lognormal distribution (Fig-
ure A23c). the values for the mean, mode, and median
are not equal: and (2) because the mean value of recharge
varies yearly. the generated values of recharge associated
with the PDF also will vary yearly. but the type of PDF
will always he the same-in this case, a normal distribu-
tion (Figure A23a). These types of PDFs were gener-
ated for seven of the aforementioned variants41 with the
exception of horizontal hydraulic conductivity.
~1 Sec the Chapter I report (Maslia cl al. In press 2007b) for additional
di<;cus\ion on PDFs ror all varied parameters.
Chapter A: Summary of Findings
Horizontal hydraulic conductivity is a parameter
for which field values were spatially distributed. For
example, in model layers I. 3, and 5, there were 18,
22, and 5, respectively, spatially distributed values of
horizontal hydraulic conductivity (Faye and Valenzu-
ela In press 2007). Using these field values. spatially
distributed values of horizontal hydraulic conductivity
were generated using Shepard's inverse distance method
to approximate values throughout the entire model
domain (Chiang and Kinzelbach 200 I). This approach
resulted in cell by cell and layer by layer spatial varia-
tions of horizontal hydraulic conductivity. In this situa-
tion, an alternative method, SGS. was used to estimate
the distribution of horizontal hydraulic conductivity. The
specific code using the SGS methodology, FIELDGEN
(Doherty 2005), is advantageous in this situation because
it allows the statistical samples or realizations to be
representative of field observations. Examples of spatial
A53
Confidence in Simulation Results-----------------------------------
Table A15. Model parameters subjected to probabilistic analysis, Tarawa Terrace and vicinity, U.S. Marine Corp Base Camp Lejeune, North Carolina.'
I ft/d. foot per day: ft 1/g. cubic foot per gram: gift 1, gram per cu hie root d-1• I/day: g/J. grams per day: ft. foot SGS, sequential Gau.\sian simulation:
r-.·JCS. t\.tonte Carlo simulation: PDE prohahility density function:-. not applicable I
Horizontal hydraulic
conductivity. layer 1,
Kll (ft/d)
1-i(1riZlllllal hydraulic
conductivity. layer 3,
K11 (ft/d)
Horizontal hydraulic
conductivity. layer 5.
Kll (ft/d)
lntiltration (recharge). 111
(inches per year)
Distrihution coef!icient,
K,1 (l't'/g)
Bulk den:-.ity. p~ (g/t't-1)
Effective porosity. nr:
React ion rate. r ( d -1)
~:lass-loading rate5• qsCs
(g/cl)
Longitudinal dispersivity.
(\ I. I ft)
12.2-53.4
4.3-211.11
6.4-9.0
6.6-19.3
5.0x I 0·6
77.112
0.2
5.0x ]()·4
1.200
25
Groundwater-flow model parameters
12.2-53.4
4,3-20.11
6.4-9.0
6.6-19.3 4.4 21.9
Fate and transport model parameters
5.0xlo-•· 3.53x 10-" 2.68x 1 o-(,
77,112 (19,943 79.0114
0.2 0,1 0.3
5.0x l o-•l 2.30x l 0-4 7.70xlll-"
1.200 200 2.200
3.2189 5 125
2.2
l.77x l(J-''
1.100
0.05
l.35X l()••l
100
0.81147
SGS used to generate hydraulic
conductivity under a normal
distribution4
SGS used to gcncniti.: hydraulic
conductivity under a normal
distribution
SGS used to generate hydraulic
conductivity under a normal
distribution
MCS used to generate the PDF
using a normal distrihution: PDF
generated for each stress period
MCS used to generate the PDF
using a normal distrihution
MCS used to generate the PDF
using a normal distribution
MCS used to generate the PDF
using a normal distribution
MCS used to generate the PDF
using a normal distribution
MCS used to generate the PDF
using a normal distribution
MCS used to generate the PDF
using a log-nor11wl distrihution5
1Se-: the Chapter I report Uvlaslia et al. In press 2007h) for a C(lmplete listing of parameters that were subjected lo variation in the uncertainty analy~is
2Symbolic notation used to describe model parameters obtaim:d frmn Chiang and Kinzelbach (2001)
-1Jnput \'alues used to seed the pseudo-random number generator
~The FJELDGEN model code descrihed in Doheny (2005) was used to generate thi: random, spatially varying field~ of hydraulic condue1ivi1y
-1The mi:an valui: derived rrom In (25): ~1:mdmd deviation derived rrorn In (5)/2, where ln () is the Naperian logarithm
A54 Historical Recons1ruc1ion of Drinking-Water Con1amina1ion at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
---------------------------------Confidence in Simulation Results
50
40
>-30 u z w ::,
d w "' ~ 20
0
a. Recharge rate
' ' '
-
-
-
I
~ r/
-A
' '
' ~-► ' ' ' ' ' Theorntical mean tor 1984:
0.0028 teet per day
~ -
I -
r
II -
-
~
' ' ' 0.0008 0.0012 0.0016 0.002 0.0024 0.0028 0.0032 0.0036 0.004 0.0044 0.0048
RECHARGE RATE, IN FEET PER DAY
150
125
>-u 100 z w ::,
d w oc 75 ~
50
25
STATISTICS
Distribution
Number of realizations
Minimum
Maximum
Mean
Mode
Median
Standard deviation
c. Dispersivity
.-+ Theoretir.al mode:
13.08 feot
Theoretical
Normal
Not applicable
-Infinity
+ Infinity
0.00280
0 00280
0.00280
0 00050
,.. + Theoretical median
I 25 feet
.--+ Theoretical mean
I 34.56 leet
Monte Carlo
simulation
Normal
500
0.001
0.005
0.00280
0.00260
0.00279
0.00049
10 20 30 40 50 60 70 80 90 100 110
DISPERSIVITY, IN FEET
100
80
60
40
20
b. Mass loading rate
,. -► Theoretical mean:
1,200 grams per day
0 f-=a-4='--+-'-+-'-+-L-f---l--+-'='!=---,-----j
• • • 1~ 1,100 1a 1a IG 19 ID 1~
MASS LOADING RATE, IN GRAMS PER DAY
STATISTICS
Theoretical Monte Carlo
simulation
Distribution Normal Normal
Number of realizations Not applicable 500
Minimum -Infinity 200
Maximum + Infinity 2,200
Mean 1,200 1,196.1993
Mode 1,200 1,256.5000
Median 1,200 1,198.0200
Standard deviation 100 104.0659
STATISTICS
Theoretical Monte Carlo
simulation
Distribution Lognormal Log normal
Number of realizations Not applicable 500
Minimum 0 5
Maximum Infinity 125
Mean 34.56 31.32
Mode 13.08 Not available
Median 25 23.85
Standard deviation 32 98 23.59
Figure A23. Probability density functions for /a) recharge rate, /b) mass loading rate !source concentration), and
le) dispersivity used to conduct probabilistic analyses.[-, minus;+, plus]
Chapter A: Summary of Findings A55
Confidence in Simulation Results-----------------------------------
distributions of horizontal hydraulic conductivity derived
by using the SGS process arc discussed in greater detail
in the Chapter I report (Maslia et al. In press 2007b).
Once the variant PDFs and the multiple spatial
distributions of horizontal hydraulic conductivity were
generated as previously described, they were used by the
MODFLOW-2K (Harbaugh et al. 2000)42 and MT3DMS
groundwater-flow and contaminant fate and transport
models, respectively, instead of single-valued input data
used in the deterministic approach (Figure A22u). This
Probabilistic analysis results of finished water for the
"forawa Terrace WTP are shown as a series of histograms
for selected times: January 1958 (Figure A24u). Janu-
ary 1968 (A24/J). January 1979 (A24c). and January 1985
(A24d). These histograms show the probability of a range
or PCE-conccntration values occurring during a specific
month and year. For example, the probability of a PCE
concentration of about 100 r1g/L occurring in linished
water at the Tarawa Terrace WTP during January 1979
can be identified according to the following procedure:
process is shown conceptually in Figure A22b. Approxi-I. Locate the nearest concentration range that
matcly 500 realizations or Monte Carlo simulations were includes the 100 ftg/L PCE concentration value
conducted using a procedure developed specifically for
the Tarawa Terrace analyses." This procedure included
using MODFLOW-2K, MT3DMS, and mixing models
previously described. Each realization randomly selected
values from PDFs of the variants derived from MCS
and from the random distributions of horizontal hydrau-
lic conductivity derived from the SGS. Specific details
about the procedures developed to conduct the proba-
bilistic analysis using the MODFLOW-2K. MT3DMS,
and mixing models are described in the Chapter I report
(Maslia et al. In press 2007b).
12 iv1ODFLOW-2K is an updateJ versilln ol' the MODFLOW-% model cmk
developed by the U.S. Geological Survey. lkcau~e of pnigra1nrni11g require-
ments associated with conducting the MCS, it was programmatically more
eftkient to u~e the ~·1ODFLOW-2K mndel code. tvlodel parameter values for
0.IODFLOW-2K were identical and equivalent to the calibrnted model param-
eter values derived using h-IODFLOW-96 ('fable A 11: Faye and Valenzuela
In press 2007), thereby resulting in equivalent groundwater-flow simulation
results for both MODFLO\V-96 and MODFU)W-2K.
·11 Initially, 840 MCS reali1.ations were conducted. However, every
simulation did not necessarily result in a set of parameter v:ducs that yielded a
physically viable groundwater-flow or fate and transport solution. For example.
some combinations of parameter values resulted in wells drying. Therefore. out
of an initial 840 MCS realizations. 510 yielded physically viable solutions.
along the x-axis of the graph in Figure A24c.
(in this example. the different shaded histogram
bar between 96 and I 05 r1g/L)
2. Move vertically upward until intersecting the
top of the histogram bar derived from the
Monte Carlo simulation results. and
3. Move horizontally to the left until intersecting
the y-axis-for Figure A24c. about 15%.
In this example, therefore. the value on the y-axis of
Figure A24c at the point of intersection-about 15%-
is the probability that finished water at the Tarawa
Terrace WTP was contaminated with a PCE concen-
tration of about I 00 r1g/L during January 1979. As a
comparison. the same procedure described above is
used to determine the probability that finished water
was contaminated with the same concentration of PCE
( I 00 r1g/L) during January 1985 (Figure A24d). For
this situation, the probability that linished water at the
Tarawa Terrace WTP was contaminated with a PCE
concentration of about I 00 µg/L during January 1985 is
determined to be less than 2%. In other words, for condi-
tions occurring during January 1985. a PCE concentra-
tion in the range of I 00 r1g/L is on the lower end ( or
"tail") of the normal distribution curve (Figure A24c/).
A56 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
---------------------------------Confidence in Simulation Results
>-u z w
:0 d w cc ~
w > ~ ~ w cc
~
0
w
:0 ~ ;;
"' z
t: :e ::;
w I t:::
f-z w u cc w "-z
['.:-
::;
C'.J
"' C'.J
0 cc "-
20
15
10
5
25
20
a. January 1958
Mean: 4.67 pg/L
Standard deviation: 1.42 pg/L
Normal distribution
fit to Monte Carlo
simulation results
b. January 1968
Mean: 56.0 µg/L
Standard deviation: 9.78 µg/L
c. January 1979 d. January 1985
Mean: 113 pg/L Mean: 168 pg/l
Standard deviation: 22.6 µg/L Standard deviation: 37.4 µg/L
See text for
discussion
SIMULATED PCE CONCENTRATION, IN MICROGRAMS PER LITER
Figure A24. Probability of occurrence of tetrachloroethylene contamination in finished water at the water
treatment plant derived from probabilistic analysis using Monte Carlo simulation for (a)January 1958,
(b)January 1968, (c)January 1979, and (d)January 1985, Tarawa Terrace, U.S. Marine Corps Base
Camp Lejeune, North Carolina. [PCE, tetrachloroethylene; µg/L, micrograms per liter]
Chapter A: Summary of Findings A57
Confidence in Simulation Results-----------------------------------
For purposes of a health study or exposure assess-
ment, epidemiologists and health scientists are interested
in ohtaining information on the prohahility that a person
or population was exposed to a contaminant exceeding a
given health guideline or criteria. For example, the proh-
ahility that residents of Tarawa Terrace were exposed
to drinking water contaminated with PCE exceeding an
MCL of 5 11g/L. To address this issue. the MCS results
described above can be presented in the form of the com-
plementary cumulative probability function and plotted
as a series of probability '·type curves" (Figure A25). The
complementary cumulative probability function describes
the probability of exceeding a certain value or answers the
question: how often is a random variable (for example. the
concentration of PCE in llnishcd water) above a certain
value? Using results shown in figure A25. the probability
that the PCE concentration in finished water at the Tarawa
Terrace WTP exceeded a value of 5 11g/L during Jam,-
ary 1958 is determined in the following manner:
I. Locate the probabilistic type curve ror
January 1958 in Figure A25a,
2. Locate the 5 (tg/L PCE concentration along
the x-axis of the graph in Figure A25a.
3. Follow the vertical line until it intersects with the
January 1958 complementary cumulative probability
function type curve (point A. Figure A25a). and
4. Follow the horizontal line until it intersects
they-axis-for this example, 39%.
In this case. there is a prohability or 39% that the PCE
concentration in finished water at the Tarawa Terrace
WTP exceeded the current MCL of 5 r1g/L during Jam,-
ary 1958. Because the MCL does not intersect with any
other type curves on the graph (Figure A25a). this can
be interpreted that for other years shown in Figure A25a
and until water-supply well TT-26 was removed from
regular service during February 1985. the probability of
exceeding the MCL for PCE is at least 99.8%. or a
near certainty.44
As discussed previously. because of contaminated
groundwater. water-supply well TT-26 was removed
from regular service during February 1985 (Figure AS,
Table A6). This caused an immediate reduction in the
PCE concentration in finished water at the Tarawa Ter-
race WTP because of the dilution of contaminated WTP
•~ E.\ccpt during July and August 1980 and January and Fchruary 198]
when w:ik:r-supply wdl 'IT-2(1 w:1s out of .~ervice-sec Figure/\ I 8.
water with water from other water-supply wells that were
not contaminated or were contaminated with much lower
concentrations of PCE than water-supply well TT-26 (Fig-
ure A 18: Appendix A2). As a result. PCE concentrations
in finished water at the Tarawa Terrace WTP during Feb-
ruary 1985-February 1987 (when the WTP was perma-
nently closed) were signilicantly reduced compared with
January 1985 concentrations (Figure A 18: Appendix A2).
Probabilistic type curves representing the complementary
cumulative probability function for selected months during
January 1985-February I 987 shown in Figure A25b also
confirm this observation. For example. using the pro-
cedure described previously-for February 1985-the
probahility of exceeding the current MCL for PCE of
5 r1g/L is 10% (point Fin Figure A25h). compared to a
probability of 39% during January 1958 and a probabil-
ity of greater than 99.8% during January 1985.
The probahility type curves shown in Figure A25
also can be used to ascertain uncertainty and variability
associated with simulated PCE concentrations in fin-
ished water at the Tarawa Terrace WTP. For example.
referring to points Band C in Figure A25a. during Janu-
ary 1958. there is a 97.5% probahility that the concen-
tration of PCE in finished water at the Tarawa Terrace
WTP exceeded 2 ftg/L (point B). and correspondingly. a
2.5% probability that the concentration exceeded 8 ftg/L
(point C). Thus, during January I 958. 95% of MCS
results" indicate that the concentration of PCE in fin-
ished water at the Tarawa Terrace WTP was in the range
of 2-8 r1g/L. Stated in terms of uncertainty and variahil-
ity. during January 1958. the uncertainty is 5% ( I 00%
minus 95% of all MCS results), and the corresponding
variahility in PCE concentration in finished water at the
Tarawa Terrace WTP is 2-8 11g/L. As a comparison, this
same analysis is conducted for January 1968 (points D
and E). For the conditions during January 1968 (the start
of the epidemiological case-control study). 95% of MCS
results indicate that the concemration of PCE in finished
water at the Tarawa Terrace WTP was in the range of
40-80 r1g/L. Stated in terms of uncertainty and variabil-
ity. during January 1968. the uncertainty is 5% ( I 00%
minus 95% of all MCS results), and the corresponding
variability in PCE concentration in linished water at the
Tarawa Terrace WTP is 40-80 11g/L.
11 In this example, point H (Figure /\25a) rcpre~ents 97.5 percentile or
Monte Carlo ~imulations. and point C repre~l.!nt~ 2.5 percentile of tvlonlc Carlo
simulatiorl';. Thu~. the rnngc of n:'.',ult.~ repre.~enting 95 percentile of :-.Iomc
Carin simulations is obtained by subtracting the probability-axis
value or poi111 C from point B ur t)7 .5'Yr,-2.5%.
A58 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
---------------------------------Confidence in Simulation Results
f-z w u "' w O-
z
uJ u z "' 0 w w u
i'.;i
~
0
/:: ::J
<D "' <D
0
"' 0-
a. January for selected years, 1958-1985 99.9 ___ __c;_:_:.:....::.:;c..:_:c,:..:c_;..:..:~__,:~------------~-----------~-
99.8
99.5
99
98
95
90
80
70
60
50
40
30
20
10
0.5
Variability: 2-8 pg/l
Uncertainty: 5%
B _ _.
39% probability
that finished water
exceeded current MCL
during January 1958
w u 0-
A
D
L E _l
0.2 y 0.1 L_ __ _j_!__-'----'-'---'-'-'-_u ___ _L _ _J'--"----'--'-'--"--LL----'-----'--'----'---'--'-.L.l_j
b. Selected months, January 1985-February 1987
99.9 ------~---,.:..,-~----'----------~-----------~-998
99.5
99
98
95
90
80
70
60
50
40
30
20
1
0.5
0.2
0.1
10% probability
that finished water
exceeded current MCL
during February 1985
3 4 8 10 20 30 40 50 80 100 200 300 400 500
SIMULATED PCE CONCENTRATION, IN MICROGRAMS PER LITER
Figure A25. Probabilities of exceeding tetrachloroethylene concentrations in finished water at
the water treatment plant derived from probabilistic analysis using Monte Carlo simulation for
1,000
(a) selected years, 1958-1985, and (b) selected months, January 1985-February 1987, Ta raw a Terr ace,
U.S. Marine Corps Base Camp Lejeune, North Carolina (see text for discussion of points A-F).
[PCE, tetrachloroethylene; MCL, maximum contaminant level; µg/L, micrograms per liter;%, percent)
Chapter A: Summary of Findings A59
Confidence in Simulation Results-----------------------------------
The probabilistic analysis conducted using MCS
was applied to 1he entire period of opera1ion of lhe
Tarawa Terrace WTP (January 1953-February 1987).
The PCE concen1ra1ion in finished water determined
using the deterministic analysis (single-value parameter
input and output: Figure A 18) also can be expressed
I. The range of PCE concentrations derived from the
probabilistic analysis using MCS is shown as a band
of solutions in Figure A26 and represents 95% of all
possible results.
2. The current MCL for PCE (5 ftg/L) was first exceeded
in linished water during Oc1ober 1957-August 1958:
these solulions include November 1957. the date
deiermined using the calibrated fate and transporl
model (Faye In press 2007b)-a deterministic
modeling analysis approach.
and presented in terms of a range of probahilities for the
entire duration of WTP operations. Figure A26 shows
the concentration of PCE in linished water at the Tarawa
Terrace WTP in terms of the MCS results. Several results
shown on this graph arc worthy of further explanation:
AGO
cc w f-::,
cc w ~
~
2
<( cc
"' 0 cc u
2
~ z
0
~ cc f-z
100
10
~ 0.1 z
0 u
0.01
I I ' I
I '-I - - - - - -
Jae
1955
Jae
1960
Jae
1965
Jae
1970
Jae
1975
EXPLANATION
Jae
1980
I f!' ll Finished water
o,-sample from water
treatment plant
Woll TT-26
Out of service: July-August 1980
January-February 1983
Sorvice terminated: February 1985
Jae
1985
Jae
1990
Mean value of concentration derived from 97.5 percentile of Monte Carlo simulations
using MT3DMS model and Monte Carlo j
simulation in a probabilistic analysis _____ _ Calibrated concentration using
(distributed-value output, 510 realizations)-~-MT30MS model in a deterministic /-~1--analysis {single-value output}.
Range of concentrations First exceeded MCL November 1957.
representing 95 percent of See Figure A18 and Chapter F (Faye 2007b)
Monte Carlo simulations 2.5 percentile of Monte Carlo simulations
Jae
1995
Figure A26. Concentrations of tetrachloroethylene in finished water at the water treatment
plant derived from probabilistic analysis using Monte Carlo simulation, Tarawa Terrace,
U.S. Marine Corps Base Camp Lejeune, North Carolina. [PCE, tetrachloroethylene;
MCL, maximum contaminant level]
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-------------------------Field Tests and Analyses of the Water-Distribution System
3. The PCE concentration in Tarawa Terrace WTP
finished water during January 1985. simulated using
the probabilistic analysis. ranges from 110-251 µg/L
(95 percent of Monte Carlo simulations). This range
includes the maximum calibrated value of 183 µg/L
(derived without considering uncertainty and varia-
bility using MT3DMS [Faye In press 2007bl) and the
maximum measured value of215 r•g/L (Table AID).
Therefore. these probabilistic analysis results-obtained
by using Monte Carlo simulation-provide a sense of
confidence in the historically reconstructed PCE concen-
trations that were delivered to residents of Tarawa Ter-
race in finished water from the WTP.
In summary, effects of parameter uncertainty and vari-
ability have been analyzed using three approaches-water-
supply well scheduling analysis. sensitivity analysis. and
probabilistic analysis. Individually and combined, these
analyses demonstrate the high reliability of and confidence
in results determined using the calibrated MODFLOW-96
and MT3DMS models (for example, Figure A 18),
described in the Chapter C (Faye and Valenzuela In press
2007) and Chapter F (In press Faye 2007b) reports. The
probabilistic analysis. conducted using the combina-
tion of MODFLOW-2K, MT3DMS. MCS. and SGS.
provides a tool (probability type curves. Figure A25) to
address issues of parameter uncertainty and variability
with respect to the concentration of PCE in finished
water delivered from the Tarawa Terrace WTP to resi-
dents of family housing at Tarawa Terrace and vicinity.
Field Tests and Analyses of the
Water-Distribution System
As discussed previously in the section on Water-
Distribution Investigation, the initial approach for quan-
tifying the concentration of PCE delivered to residences
of Tarawa Terrace was to develop and calibrate a model
representation of the water-distribution system using
the public domain model EPANET 2 (Rossman 2000).
With this approach, street-by-street concentrations of
PCE could he simulated and reconstructed. Although
using this rigorous approach was replaced with a simpler
mixing model approach, lield studies were conducted
early in the project to gather information needed to
develop and calibrate a model of the Tarawa Terrace
water-distribution system. A summary of this informa-
tion and comparison of PCE concentration results using
the street-by-street water-distribution system model with
Chapter A: Summary of Findings
the mixing model results arc presented herein. A detailed
description and discussion of the use and application of
water-distribution system modeling with respect to the
Tarawa Terrace water-distribution system is provided in
the Chapter J report (Sautner ct al. In press 2007).
Based on reviews of historical WTP operations as
well as housing information, the authors concluded that
the historic~ll water-distribution system serving Tarawa
Terrace was nearly identical to the present-day (2004)
water-distribution system. Thus. information and data col-
lected to characterize the present-day water-distribution
system also would be useful in characterizing the histori-
cal water-distribution system. The network of pipelines
and storage tanks. shown in Figure A27 represents
the present-day water-distribution systems serving the
Tarawa Terrace and Holcornh Boulevard areas, arc nearly
identical to historical water-distribution systems serving
these areas with the following exceptions:
I. The Holcomb Boulevard WTP came online during
June 1972 (Figure A3): prior to that date, the Hol-
comb Boulevard area received linishcd water from
the Hadnot Point WTP (Plate I):
2. The Tarawa Terrace and Montford Point WTPs were
closed during 1987 (Figure A3) and presently, the
Holcomb Boulevard WTP provides finished water
to these areas:
3. A pipeline. constructed during 1984, follows SR 24
northwest from the Holcomb Boulevard WTP to
ground storage tank STT-39 and presently is used to
supply STT-39 in the Tarawa Terrace water-distribu-
tion system with linished water (Figure A27): and
4. A pipeline. constructed during 1986. trends east-
west from the Tarawa Terrace 11 area to storage
tank SM-623 and presently is used to supply the
storage tank with linishcd water.
Two types of lield tests were conducted to deter-
mine the hydraulic and water-quality parameter values
needed to develop and calibrate a water-distribution
system model for Tarawa Terrace: (I) lire-flow tests,
conducted during August 2004, in the Tarawa Terrace
and Camp Johnson areas: and (2) a fluoride tracer test,
conducted during September and October 2004, in the
Tarawa Terrace and Holcomb Boulevard areas. Detailed
descriptions of the test procedures and results of the
field tests are described in the Chapter J report (Sautner
et al. In press 2007) and in a number of related papers.
A61
Confidence in Simulation Results-----------------------------------
A62
Jr
~ " --1, "> ,.
N -7;.
i ,.
~ ~.
Base from U.S. Marine Corps and 2 MILES
U.S. Geological Survey digital data files
2 KILOMETERS
EXPLANATION
Present-day (2004) water-distribution Water pipeline-2004
system on Camp Lejeune
Military Reservation X Shut-off valve
D Tarawa Terrace
FOJE!l D Holcomb Boulevard Fluoride logger and number-CRWQME
D Hadnot Point Storage tank
D Other areas of Camp Lejeune
J,,.S2323 Elevated-Controlling
Military Reservation 2,SBJ0 Elevated-Non controlling
Holcomb Boulevard Water JiSM-623 Elevated-Intermittently controlling
Treatment Plant Service Area-and noncontrolling depending on
March 1987 to present demand conditions
■ Holcomb Boulevard Water Treatment Plant STT-39O Ground-Finished water
Figure A27. Locations of continuous recording water-quality monitoring equipment (CRWQME; F01-F09)
and present-day (2004) Tarawa Terrace and Holcomb Boulevard water-distribution systems used for
conducting a fluoride tracer test, September 22-0ctober 12, 2004, U.S. Marine Corps Base Camp Lejeune,
North Carolina.
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-------------------------Field Tests and Analyses of the Water-Distribution System
For example. fire-flow tests are described in Sautner ct
al. (2005) and Grayman ct al. (2006). A lluoridc tracer
test is described in Maslia et al. (2005) and Sautner et al.
(2005. 2007)
and (2) Tarawa Terrace." Therefore. the fluoride tracer
test included the collection of data at selected locations
within the Tarawa Terrace and Holcomb Boulevard
water-distribution systems.
The use of a fluoride tracer test to characterize a
water-distribution system is of particular importance
because results obtained from the test-the impact or
storage tank operation. travel times. and dilution rates
{)f c<mstitucnts in the water-distribution system-assist
with determining parameter values needed to calibrate
a water-distribution system model using extended
period simulation (EPS). Additionally, the movement
and distribution or lluoridc through the Tarawa Ter-
race water-distribution system would be similar to
the movement and distribution of a contaminant. such
as PCE through the water-distribution system. Since
March 1987. the Holcomb Boulevard WTP has sup-
plied finished water to two water-distribution systems at
Camp Lejeune (Figure A27): (I) Holcomb Boulevard"'
~h The Holcomb Hmtlevard \VTP pnwides finished water to the following
areas within the llokomh Boulevard water-di\lribution :-.y:-.tern: Berkeley
~fanm, Watkin~ Village. l'aradi~e Point. and Midway Park (Figure t\27).
The lluoride tracer test was conducted Septem-
ber 22-October 12. 2004. The test consisted or moni-
toring lluoride dilution and re-injection (shutoff and
startup of the sodium lluoridc feed at the Holcomb
Boulevard WTP). Nine locations in the Tarawa Terrace
and Holcomb Boulevard water-distribution systems
were equipped with continuous recording water-quality
monitoring equipment (CRWQME). Monitor loca-
tions arc shown in Figure A27 and are designated as
F0I-F09. A list of the monitoring locations and the
water-distribution system location being monitored is
provided in Table A 16. Monitoring locations included
the main transmission line from the Holcomb Boulevard
WTP to the water-distribution system (F0 I). the Tarawa
Terrace finished water reservoir (F02). two control-
ling elevated storage tanks (Paradise Point 1S23231 and
n Based on pre~ent-day operations (2004 ). the Tarawa Ten-.u . .:e wa1er-distri-
hutio11 system includes the following areas: ·1:irawa Terrace hou~ing areas I and
II. Camp Knox Trailer Park. Camp Johnson. and Montford Point (Figure A27).
Table A16. Description of locations equipped with continuous recording water-quality monitoring equipment used to conduct a
fluoride tracer test of the Tarawa Terrace and Holcomb Boulevard water-distribution systems, September 22-0ctober 12, 2004,
U.S. Marine Corps Base Camp Lejeune, North Carolina.
FOi 356478.25 2498392.43
F02 362057.78 2490580.75
F03 344823.33 2491037.83
F04 351648.84 2495750.35
F05 362270.35 2488417.94
F06 357638.42 2501665.36
F07 361760.20 2486365.30
F08 353489.91 2484738.57
F09 362945.52 2479935.36
1Sce Figure A27 for station ]()(:ation~
Holc(llllb Boulevard
Tarawa Terrace
H(1kornh Boulevard
Holcomb Boulevard.
Berkeley Manor
Tarawa Terrace.
housing area II
Htilc(>tnb Boulevard.
Midway Park
Tar,iw,1 Terrace.
housing area II
Holcomb Boulevard.
Paradise Point
Tarawa Terrace.
Camp Johnson
\Vater treatment plant. main transmission
line. fluoride source
Ground storage tank. source for Tarawa
Terrace water-distribution system
Distribution system hydrant
Distribution system hydrant and elevated
storage tank
Di:-.tribution system hydrant
Distribution system hydrant
Distribution system hydrant and elevated
storage tank
Controlling elevated storage tank
Controlling elevated s10ragc tank
1C(J(lfliinate~ are in N(irth Carolina State Plane n111rdina1c ~y.\tt.'m. NtHth American Datum 1983. Natinnal Geodetic Vertical Datum of ! 929
Chapter A: Summary of Findings A63
Field Tests and Analyses of the Water-Distribution System-------------------------
Camp Johnson [SM623[-F08 and F09. respectively).
and live hydrants located throughout housing areas (F03.
F04. F05. F06. and F07). The fluoride at the Holcomb
Boulevard WTP was shut off at 1600 hours on Septem-
ber 22. A background concentration of about 0.2 mil-
ligram per liter (mg/L) in the water-distribution system
was reached by September 28. At 1200 hours on Septem-
ber 29. the lluoridc was turned back on at the Holcomb
Boulevard WTP, and the test continued until loggers were
removed and data downloaded on October 12. In addition
to CRWQME, grab samples were collected and analyzed
for quality-assurance and quality-control purposes.
Nine rounds of water samples were collected at each
monitoring location during the test. For each round, the
Holcomb Boulevard WTP water-quality lab analyzed
25 milliliters (mL) of the sampled water. and the Federal
Occupational Health (FOH) laboratory. located in Chi-
cago. Illinois. analyzed the remaining 225 mL of water.
Storage tanks in the Tarawa Terrace and Holcomb
Boulevard water-distribution systems are categorized
as either controlling or noncontrolling. Controlling
elevated storage tanks are operated in the following
174 ~ N ~
~ 173 0
w :e > C, 171 o >-
<D "' "'c:, >-~ w"' wu IJI
~ e= z cc -w 170 _J > C wu 0 ~ >-w >-~ 0 ~w 169 "' ~ I u 0
CCC:, 0 il 2~ wO ~~ I ~o rW •o • al "' "' 168 Eo I C -,: :;/ E~ "~ -o z -~ 00 c:, ~o ~e:: e= 167 ·.::::~ I ·.:::: ~ "' g ;;, I a_ z -N 0~ ~;,, 0:
166
22 23 24 25 26 27 28 29 30
SEPTEMBER 2004
manner. Finished water is supplied to the respective
water-distribution system from the elevated controlling
storage tank in response to system demand. When the
water level in the controlling tank falls below a pre-set
water-level mark. pumps turn on and fill the tank with
finished water from a ground storage tank. When the
water level in the controlling tank reaches a pre-set high
water-level mark, the pumps are turned off. The water
level in the tank then begins to drop based on demand
unti I, once again, the water level reaches the pre-set low
water level. The fill and drain process is then repeated.
An example of water-level data collected by the Camp
Lejeune supervisory control and data acquisition
(SCADA) system for controlling storage tank STT-40
(Tarawa Terrace elevated. Figure A27) is shown in Fig-
ure A28. Two other elevated storage tanks are noncon-
1rolling tanks. These elevated storage tanks show little
water-level lluetuation because they arc not exercised
very often-they are primarily used for fire protection.
The elevated storage tanks are S830 (Berkeley Manor)
and LCH-4004 (Midway Park). both serving the Hol-
comb Boulevard water-distribution system (Figure A27).
I
6 J 8
OCTOBER 2004
10 11 12 13
Figure A28. Measured water-level data from the Camp Lejeune SCAOA system for controlling elevated
storage tank STT-40, September 22-0ctober 12, 2004, U.S. Marine Corps Base Camp Lejeune, North Carolina.
[SCADA, supervisory control and data acquisition]
A64 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-------------------------Field Tests and Analyses of the Water-Distribution System
Using results rrom the lluoride tracer test described
previously and the lire-flow test of August 2004. an
all-pipes EPS model of the Tarawa Terrace water-
distribution system was calibrated. To simplify
and reduce the computational requirements, a skcl-
etonized version of an all-pipes representation of the
water-distribution system was used for all subsequent
EPANET 2 simulations." A 24-hour diurnal pattern
based on measured llow data (delivered finished water)
1~ Skelcto11i1,:1tion i~ the reduction or aggregation ol' a water-distribution
system network so that only the major hydraulic characteristic~ need he repre-
sented by a model. Skclctoni,_ation ofte11 i:. used to reduce the cumputalional
requirements of modeling an all-pipes network.
1.6
D Calibrated demand factor -Measured flow of
1.4 r finished water -
Notes -,._
1.2
1. Demand factor calibrated using I EPANET 2 (Rossrnan 2000) I\
t-2. Measured flow are hourly averages ,.
using 2-minute data {R. Cheng,
Camp Lejeune Environmental -Management Division, written
r commun., January 25, 2005)
cc
D ,... -u :1:
D 08 z ""
r I
:, w D I 0.6 -
0.4
\ ~v --;
,__~
0.2 -
and calibrated demand factors is shown in Figure A29.49
Flow data were measured using a venturi meter located
in the Tarawa Terrace pump house (building adjacent
to STT-39 in Figure A27)."1 Calibrated demand factors
arc in reasonable agreement with measured flow data.
Details of the calibration procedure and calibration sta-
tistics arc provided in the Chapter J report (Sautner et al.
In press 2007).
4'1 Data for measured dcli\·crcd flow were pn.:viously presented and dis-
cussed in the section on Relation ofContaminmion to Water Supply,
Pnld11ction, and Distribulillll (Figure AX).
~11 A venturi meter is a device used to measure the flow rate or velocity of
a fluid through a pipe. A photograph of the Tarawa Terrat:e pump house is
shown on the front cover of this report.
600
w ,...
" -z
:i'
-500 cc w :"v--I -~
"' -. z D ~ ---~ "" '--. -"' 400 :;;; -cc
/~
w j -D w
' -
\ .
300 I
"' 2 a:
0 w cc w > -200 :::, w
D
~
' '
D ;;:
D ~ ~
-100 D w cc " ' ' "' "" . w :,
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
HOUR
Figure A29. Calibrated and measured diurnal pattern (24 hours) of delivered finished water during
field test, September 22-0ctober 12, 2004, Tarawa Terrace water-distribution system, U.S. Marine
Corps Base Camp Lejeune, North Carolina. [Flow data measured at venturi meter located in building
STT-39A (Tarawa Terrace pump house)]
Chapter A: Summary of Findings A65
Field Tests and Analyses of the Water-Distribution System-------------------------
Simulated lluoridc concentrations arc compared with
measured field data C(mccntrati(ms ohtaine<l fn)m the
CRWQME and with the grab sample measurements for
the Tarawa Terrace water-distribution system at locations
F02. F05, F07. and F09 (Figure A27). These compari-
sons are shown in the graphs of Figure A30. Note that
monitoring location F02 is used as the source of fluoride
for the Tarawa Terrace water-distribution system. Rcsuils
shown in Figure A30 along with calibration statistics
presented in the Chapter J report (Sautner ct al. In press
2007) provide evidence that the EPS model of the Tarawa
Terrace waler-distribution system is reasonably calibrated
and adequately characterizes the present-day (2004)
Tarawa Terrace water-distribution system.
:e
~
z D >= "' cc f-z w u z D u w 0
cc 0
3
1.4
1.1
0.8
0.6
0.4
01
0
1.4
1.1
LL 0.8
0.6
0.4
0.1
A66
a. Logger F02 b. Logger F05
~ e C 0 ,
0 " 0 =~ ~ ~ " , Eg m o n~
.2 ~ uO -~ E~ ~~ .2 ;g g"' .o H u::~ uC:, ·;::: a,
0 c:, ,~ "'
D
"' D
C. logger F07 d. Logger F09
D
2223242526272829301 2 3 4 5 6 7 8 S 10111213 22 23 24 25 26 27 28 29 30 1 2 3 4 5 6 7 8 9 10 11 12 13
SEPTEMBER 2004 OCTOBER 2004
Fluoride concentration,
in milligrams per liter
Measured
CRWQME
CL Lab
□ FOH Lab
--Simulated
SEPTEMBER 2004 OCTOBER 2004
EXPLANATION
Notes: CRWQME = Continuous recording water-quality
monitoring equipment
CL Lab= Holcomb Boulevard Water Treatment Plant
onsite lab analysis, Camp Lejeune, North Carolina
25 milliliter sample
FOH Lab= Federal Occupational Health Lab, Chicago, Illinois
225 milliliter sample
Figure A30. Measured and simulated fluoride concentrations at four monitoring locations (a) F02, (b) F05, (c) F07, and
(d) F09 in the Tarawa Terrace water-distribution system, September 22-0ctober 12, 2004, U.S. Marine Corps Base Camp Lejeune,
North Carolina. !See Figure A27 for monitoring locations and Table A16 for description of hydraulic device being monitored.)
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-----------------------------------Summary and Conclusions
Using the calibrated EPS model of the Tarawa
Terrace water-distribution system, conditions represent-
ing December 1984 were simulated. This was a period
of high water production and usage. The duration of
the simulation was 744 hours (31 days). The purpose
or the simulation was to test the concept that a mixing
model. based on the principles of continuity and con-
servation of mass (Equations 2 and 3). could be used to
estimate the street-by-street concentrations of a con-
taminant derived using a sophisticated numerical model
of the water-distribution system. such as EPANET 2.
The mixing model represents a condition of complete
mixing and stationary water-quality dynamics in a
water-distribution system like Tarawa Terrace where
all source water (groundwater) is mixed at the treatment
plant. Using the calibrated water-distribution system
model. for a simulation period of 744 hours (31 days)-
rcpresenting December 1984-and an initial source
concentration of 173 11g/L at the Tarawa Terrace WTP
(Figure A 18, Appendix A2), the following results
were obtained:
100% of the concentration of PCE in finished
water at the Tarawa Terrace WTP ( 173 µg/L)
reached locations F05 and F07 (Figure A27).
located in the 'forawa Terrace housing area
within 2 days,
I 00% of the concentration of PCE in finished
water at the Tarawa Terrace WTP ( 173 11g/L)
reached the Camp Johnson elevated storage
tank within 3 days, and
• I 00% of the simulated concentration of PCE
in finished water at the Tarawa Terrace WTP
( 173 11g/L) reached the Montford Point area
(farthest point from the Tarawa Terrace WTP)
within 7 days.
These results demonstrate that on a monthly basis,
the concentration of PCE at residential housing areas
throughout Tarawa Terrace would be nearly the same as
the concentration of PCE in finished water at the Tarawa
Terrace WTP. Therefore, using a mixing model based on
the principles of continuity and conservation of mass is
appropriate for determining the concentration of PCE in
linished water delivered from the Tarawa Terrace WTP.
Chapter A: Summary of Findings
Summary and Conclusions
Two of the three drinking-water systems that
served family housing at U.S. Marine Corps Base Camp
Lejeune were contaminated with VOCs. Groundwater
was the sole source of drinking-water supply. One
system. the Tarawa Terrace drinking-water system, was
mostly contaminated with PCE when water-supply wells
were contaminated hy off-base dry-cleaning operations
at ABC One-Hour Cleaners (Shiver 1985). The other
system, the Hadnot Point drinking-water system. was
contaminated mostly with TCE from 011-hase indus-
trial operations. The contaminated wells were continu-
ously used until 1985 and sporadically used until early
1987. ATSDR"s health study will try to determine if
an association exists between in utero and infant (up
to I year of age) exposures to drinking-water contami-
nants and specific hinh defects and childhood cancers.
The study includes births occurring during 1968-1985
to mothers who lived in hase family housing during
their pregnancies. Historical exposure data needed for
the epidemiological case-control study are limited. To
obtain estimates of historical exposure, ATS DR is using
water-lll()deling teclrniques and the process of histori-
cal reconstruction. These methods are used to quantify
concentrations of particular c(mtaminants in finished
water and to compute the level and duration of human
exposure to contaminated drinking water. The analyses
and results presented and discussed in this Summary of
Findings. and in reports described herein. refer solely to
Tarawa Terrace and vicinity. Future analyses and reports
will present information and data about contamination of
the Hadnot Point water-distribution system.
Based on information. data, and simulation results,
the onset of pumping at Tarawa Terrace is estimated
to have begun during 1952. Water-supply well TT-26,
located about 900 ft southeast of ABC One-Hour Clean-
ers, probahly began operations during 1952 (Figure A I,
Table A6). Additionally, the first occurrence of PCE con-
tamination at a Tarawa Terrace water-supply well prob-
ably occurred at well TT-26, following the onset of dry-
cleaning operations during 1953 (Faye In press 2007b).
Detailed analyses of PCE concentrations in ground-
water monitor wells, hydroconc sample locations,
and at Tarawa Terrace water-supply wells during the
A67
Summary and Conclusions------------------------------------
period 1991-1993 were sufficient to estimate the mass
of PCE remaining in the Tarawa Terrace and Upper
Castle Hayne aquifers. Similar methods were applied
to compute the mass of PCE in the unsaturated zone
(zone above the water table) at and in the vicinity of
ABC One-Hour Cleaners using concentration-depth
data determined from soil borings. The total mass of
PCE computed in groundwater and within the unsatu-
rated zone equals about 6.000 pounds and equates to a
volume or about 430 gallons. This volume represents an
average minimum loss rate of PCE to the subsurface at
ABC One-Hour Cleaners of about 13 gallons per year
(78.737 grams per year) for the period 1953-1985.
Pankow and Cherry ( 1996) indicate that computations
of contaminant mass similar to those summarized here
represent only a small fraction of the total contaminant
mass in the subsurface.
Calibration of the ·n.rawa Terrace models was
accomplished in a hierarchical approach consisting of
four successive stages or levels (Figure A9). Simulation
results achieved for each calibration level were iteratively
adjusted and compared to simulation results of previous
levels until results al all levels satisfactorily conformed
to pre-selected calibration targets (Table AS). In hierar-
chical order. calibration levels consisted of the simula-
tion of (I) prcdevelopmcnt groundwater-llow conditions
(Figure A I 0o). (2) transient or pumping groundwater-
llow conditions (Figure AI0/J), (3) the fate and transport
of PCE from the source al ABC One-Hour Cleaners
(Figure A 11 ), and (4) the concentration of PCE in fin-
ished water at the ·n.rawa Terrace WTP (Figure A 12).
Based on calibrated model simulations. water-
supply well TT-26 had the highest concentration of PCE-
contaminatcd groundwater and the longest duration of
PCE-contaminated groundwater with respect to any other
Tarawa Terrace water-supply well (Figure A 18). The
simulated PCE concentration in water-supply well TT-26
exceeded the current MCL of 5 µg/L during January 1957
(simulated value 5.2 r1g/L) and reached a maximum sim-
ulated value of 851 r1g/L during July 1984 (Table A 12).
The mean simulated PCE concentration during the period
exceeding the current MCL of 5 ftg/L-January 1957-
January 1985-was 414 µg/L. a duration of 333 months.
The monthly concentrations of PCE assigned to
finished water at the Tarawa Terrace WTP were deter-
mined using a materials mass balance model (simple
mixing). The model is based on the principles of
continuity and conservation of mass (Masters 1998)
and is used to compute the now-weighted average
concentration of PCE. Finished water contaminated
with PCE exceeded the current MCL of 5 ftg/L dur-
ing November 1957. Based on mixing model results.
linished water exceeded the MCL for 346 months
(29 years)-November 1957-February 1987 (Fig-
ure A 18, Table A 12)." The maximum simulated PCE
concentration in finished water was 183 ~lg/L occurring
during March 1984. The maximum observed PCE con-
centration was 215 µg/L measured on February I I. 1985
(Table A I 0). The average simulated PCE concentration
for the period exceeding the current MCL of 5 ftg/L-
November 1957-February 1987-was 70 flg/L.
The calibrated fate and transport model simulated
PCE as a single-specie contaminant dissolved in ground-
water. However. evidence of the transformation of PCE
to degradation by-products of TCE and 1.2-tDCE was
found in water samples obtained from Tarawa Ter-
race water-supply wells TT-23 and TT-26. Thus. the
simulation of PCE and its degradation by-products was
necessary. For this simulation, a model code identi-
fied as TechFlowMP. developed by the Multimedia
Environmental Simulations Laboratory (MESL) at the
Georgia Institute of Technology. was used. TechFlowMP
simulates three-dimensional multiphase, multispecies
mass transport of PCE and its associated degradation
by-products TCE. 1.2-tDCE. and YC in the unsaturated
and saturated zones at Tarawa Terrace and vicinity (that
is. the sequential biodegradation and transport of PCE).
Simulation results for finished water at the Tarawa Ter-
race WTP (Figure A 19/J). contaminated with PCE degra-
dation by-products TCE, 1.2-tDCE. and VC, show that:
(I) TCE was below the current MCL value of 5 ftg/L for
nearly the entire historical period except during Janu-
ary 1984-January 1985 when it ranged between 5 and
6 r1g/L; (2) 1,2-tDCE was below the current MCL value
of 100 µg/L for the entire historical period; and (3) YC
was at or above the current MCL value of 2 r1g/L from
May 1958 through February 1985 when water-supply
well TT-26 was shut down. As part of the degradation
by-product simulation using the TechFlowMP model.
results also were obtained for VOCs in the vapor phase
(above the water table in the unsaturated zone). Analyses
of the distribution of vapor-phase PCE indicate there is
51 This period docs not include the months of July-Augu~L 19X0 and
January-February 1983. wbcn watcr-~upply well Tr---26 was not operating.
A68 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-----------------------------------Summary and Conclusions
potential for vapors from these plumes to enter buildings
at Tarawa Terrace I. thereby providing a potential expo-
sure pathway for inhalation or PCE vapor. At Tarawa
Terrace I these buildings would include family housing
and the elementary school (Figure A20).
To address issues of model uncertainty and param-
eter variability. three types or analyses were conducted:
(I) water-supply well scheduling analysis. (2) sensitiv-
ity analysis. and (3) probabilistic analysis. All or the
additional analyses were conducted using PCE as a
single-specie contaminant dissolved in groundwater-
the calibrated models described in Chapter C (Faye and
Valenzuela In press 2007) and Chapter F (Faye In press
2007b) reports. The simulation tool. PSOpS. was used
to investigate the effects of unknown and uncertain
historical well l)perations and analyses or the variation
in water-supply well scheduling. PSOpS simulations
demonstrate that the current MCL for PCE (5 [ig/L)
would have been exceeded in finished drinking water
from the Tarawa Terrace WTP as early as Decem-
ber I 956 and no later than June 1960 (points A and D,
respectively, in Figure A2 I).
Sensitivity analyses were conducted using Tarawa
Terrace models. Selected model parameters were varied
one at a time from their respective calibrated values
(Table A 11 ). The effect or this variation on the change in
the PCE concentration or finished drinking water at the
Tarawa Terrace WTP was assessed. Four groundwater-
flow and seven fate and transport model parameters were
varied. Results of the sensitivity analyses showed that
some parameters-specific yield, storage coeflicient,
and molecular diffusion-were insensitive to change,
even when varied by factors or IO and 20 (Table A 14 ).
Other parameters, for example. horizontal hydraulic
conductivity for model layer I and infiltration (ground-
water recharge), were extremely sensitive to values
less than the calibrated values. Reducing the calibrated
values for these parameters resulted in wells drying up
during the simulation process. Generally. increasing or
decreasing a calibrated parameter value by I 0% (ratio of
varied to calibrated parameter value of 0.9-1.1) resulted
in changes of 6 months or less in terms or the date that
finished drinking water first exceeded the current MCL
of 5 µg/L for PCE. Results or parameter variations were
used. in parl. to assist in selecting parameters considered
for a probabilistic analysis.
Chapter A: Summary of Findings
A probabilistic analysis approach was used to inves-
tigate model uncertainty and parameter variability using
MCS and SGS. For the groundwater-llow and contami-
nant fate and transport models (Faye and Valenzuela In
press 2007. Faye In press 2007b). eight parameters were
assumed to be uncertain and variable: (I) horizontal
hydraulic conductivity. (2) recharge rate, (3) effective
porosity, (4) bulk density. (5) distribution coefficient.
(6) dispcrsivity. (7) reaction rate. and (8) the PCE mass
loading rate. With the exception or horizontal hydraulic
conductivity. PDFs were generated for the remaining
seven parameters of variation using Gaussian pseudo-
random number generators. Horizontal hydraulic con-
ductivity is a parameter for which there were spatially
distributed field values. Therefore. an alternative method.
SGS, was used to estimate the distribution of horizontal
hydraulic conductivity for model layers I, 3, and 5. The
probabilistic analyses indicated that 95% of Monte Carlo
simulations show the current MCL for PCE (5 rig/L) was
first exceeded in finished water during October 1957-
August 1958 (Figure A26); these solutions include
November 1957, the date determined l'rom the calibrated
contaminant fate and transport model (Faye In press
2007b) that was based on a deterministic (single-value
parameter input and output) approach. The PCE concen-
tration in Tarawa Terrace WTP finished water during
January 1985, simulated using the probabilistic analysis,
ranges from 110 to 251 µg/L (95 percent of Monte
Carlo simulations). This range includes the maximum
calibrated value or 183 ftg/L (derived without consider-
ing uncertainty and variability using MT3DMS) and the
maximum measured value of 215 pg/L.
As part or this investigation, field tests were
conducted on the present-day (2004) water-distribution
system serving Tarawa Terrace. Data gathered from the
investigation were used to construct a model of the water-
distribution system using the EPA NET 2 model code.
Based on reviews or historical maps and information,
the present-day (2004) water-distribution system is very
similar to the historical water-distribution system. Thus,
the operational and water-delivery patterns determined
for the present-day (2004) water-distribution system from
t"ield investigations (Sautner et al. 2005, In press 2007)
were used to characterize the historical water-distribution
system. Using a calibrated water-distribution system
model and an initial source concentration of 173 µg/L
A69
Availability of Input Data Files, Models, and Simulation Results-----------------------
al lhc Tarawa Terrace WTP (Figure A 18), an exlended
period simulalion ol' 744 hours (31 days), represenling
December 1984, indieales:
• I 00% of !he concentration of PCE in finished
water al !he Tarawa Terrace WTP ( 173 rig/L)
reached locmions F05 and F07 (Figure A27),
localed in the Tarawa Terrace housing area
within 2 days,
I 00% of !he concentration of PCE in finished
waler at !he 'forawa Terrace WTP ( 173 ftg/L)
reached the Camp Johnson clcva1ed storage tank
within 3 days, and
I 00% of the simulated concentration of PCE
in finished water at the Tarawa Terrace WTP
( 173 ftg/L) reached the Montford Point area
(farthesl poinl from the Tarawa Terrace WTP)
within 7 days.
These resulis confirm the assumplion that on a mon1hly
basis, the concen1ra1ion of PCE al residential housing
areas throughout Tarawa Terrace would be the same as
ihc concentralion of PCE in finished waler at ihc Tarawa
Terrace WTP. Therefore, using a mixing model hascd
on the principles or continuity and conservation of mass
(Equa1ions 2 and 3, respectively) was appropriale for
reconstructing the historical concentrations of PCE in
linished water delivered from ihe Tarawa Terrace WTP
In summary. based on field data. modeling rcsulls.
and the historical n.~construction process. the following
conclusions arc made with respect to drinking-water
contamination at Tarawa Terrace:
I. Simulaicd PCE concentralions exceeded the
current MCL of 5 i1g/L al water-supply well T1~26
for 333 monlhs-January 1957-.lanuary 1985:
the maximum simulated PCE concentration was
851 ~tg/L: the maximum measured PCE concen-
lration was 1,580 iig/L during January 1985.
2. Simulated PCE concentrations exceeded the current
MCL of 5 ftg/L in finished water al the Tarawa Ter-
race WTP for 346 months-November 1957-Febru-
ary 1987: ihe maximum simulaied PCE concentra-
tion in finished water was 183 p,g/L: the maximum
measured PCE concentration in finished water was
215 ftg/L during February 1985.
3. Simulation or PCE degradation by-products-TCE,
lra11s-l ,2-dichloroe1hylcne ( 1,2-IDCE), and vinyl
chlori<le-indicated that maximum concentrations
of ihe degradation by-products generally were in ihc
range of 10-100 ftg/L at water-supply well TT-26:
measured concentrations of TCE and 1,2-tDCE on
January 16, 1985, were 57 and 92 11g/L, respectively,
4. Maximum concentrations of the degradation by-
products in finished water at the Tarawa Terrace
WTP generally were in the range of2-15 ftg/L;
measured concenlrations of TCE and 1,2-tDCE on
February 11, 1985, were 8 and 12 i1g/L, respectively.
5. PCE concentrations in finished water at 1he
Tarawa Terrace WTP exceeding the currenl MCL
of 5 i1g/L could have been delivered as early as
December 1956 and no later then December 1960.
Based on probabilistic analyses, the most likely
dates 1hat finished water first exceeded ihe current
MCL ranged from Oc1ober 1957 10 August 1958
(95 percent probability), with an average first
excecdance date or November 1957.
6. Exposure to PCE and PCE degradation by-products
from contaminated drinking water ceased after
February I 987: the Tarawa Terrace WTP was
closed March 1987.
Availability of Input Data Files, Models,
and Simulation Results
Calibraicd model input data files developed for
simulating prcdcvclopmcnt groundwater flow. transient
ground-waler flow, the !'ate and transpon of PCE as a
single specie, and the distribution of water and contami-
nants in a water-distrihu1ion system are provided with
this report in a DVD l<mnal. Public domain model codes
used with these input files are available on the Internet at
the following Web sites:
• Prcdcvelopmcnt and transient groundwater flow
o Model code: MODFLOW-96
o Web site: http://watewsgs.gov/111p/
gwsojtware/nuHf/low.html
A70 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------------------------------Acknowledgments
• Fate and transport of PCE as a single specie
o Model code: MT3DMS
o Web site: h11p://hvdro.geo.11a.ec/11/
• Distribution of water and contaminants in a
water-distributi<>n system
o Model code: EPANET 2
o Web site: http://www.epa.gov/nrmr//wswrcl/
epa11et.ht111I
Specialized model codes and model input data
files were developed specilically for the Tarawa Terrace
analyses hy the MESL at the School of Civil and Envi-
ronmental Engineering. Georgia Institute of Technology.
These specialized codes and input data files were devel-
oped for simulating three-dimensional multispecics.
multiphase. mass transport (TechFlowMP) and pump-
ing schedule optimization (PSOpS) and are descrihed
in detail in the Chapter G (Jang and Aral In press 2007)
and Chapter H (Wang and Aral In press 2007) reports.
respectively. Contact information and questions related
to these codes arc provided on the Internet at the MESL
Web site at: h11p://111esl.ce.gatech.ed11.
Also included on the DVDs accompanying this
report is a lile that contains results for monthly simulated
concentrations or PCE and PCE degradation by-products
(TCE. 1.2-tDCE. and VC) in finished water at the
Tarawa Terrace WTP for January 1951-March 1987.
This lile (also provided in Appendix A2) is prepared
in Adobe® Portable Document Formal (PDF).
Readers desiring information ahout the model input
data files or the simulation results contained on the DVDs
also may contact the Project Officer of ATSDR's Exposure-
Dose Reconstruction Project at the following address:
Morris L. Maslia. MSCE, PE. D.WRE, DEE
Exposure-Dose Reconstruction Project
Division of Health Assessment and Consultation
Agency for Toxic Substances and Disease Registry
1600 Clifton Road, Mail Stop E-32
Atlanta. Georgia 30333
Telephone: (404) 498-0415
Fax: (404) 498-0069
E-mail: mmaslia@cdc.gov
Chapter A: Summary of Findings
Acknowledgments
A study of this complexity and magnitude is depen-
dent upon the assistance. input. and suggestions of many
colleagues. Thus. the authors of this report and all chapter
repons acknowledge the managers and staff of the U.S.
Geological Survey Water Science Centers in Raleigh.
North Carolina, and Atlanta. Georgia. In particular. the
contrihutions of Melinda J. Chapman. Douglas A, Harned,
and Stephen S. Howe are acknowledged for providing the
majority or well. water-level. and pumpagc data used in this
study. Keith W. McFadden is acknowledged for assistance
with spatial analyses in preparing illustrations and with
developing geodatahases. Web-hased applications, and
the querying system contained on the electronic media
accompanying Chapters A and K. Gregory C. Mayer and
Elhvard H. Martin also are acknowledged for their adminis-
trative assistance.
The authors acknowledge the staff of the Environmen-
tal Management Division. U.S. Marine Corps Base Camp
Lejeune. North Carolina. In particular. Scoll A. Brewer.
Brynn Ashton. Scott R. Williams. and Rick Cheng for
their assistance and cooperation during the course of this
study, especially for providing a large numher of technical
reports. maps, and historical documents. which summarize
the results of groundwater remedial investigations at and in
the vicinity of Tarawa Terrace. The authors also acknowl-
edge Joel Hartsoe and Danny E. Hill of the Camp Lejeune
Puhlic Works Department Utility Section.
The authors acknowledge the contributions of the
USE.PA. Region IV. Atlanta. Georgia, for providing reports
and documents summarizing the results of investigations of
groundwater contamination in the vicinity of ABC One-
Hour Cleaners and in the northern part of Tarawa Terrace.
The authors acknowledge colleagues al ATS DR. Eastern
Research Group, Inc .. the Multimedia Environmental Simu-
lations Laboratory at the Georgia Institute of Technology,
and the Oak Ridge Institute for Science and Education for
providing assistance and advice with all aspects of this study.
Thomas M. Plummer. Commander. U.S. Puhlic
Health Service. Indian Health Service. U.S. Department of
Health and Human Services, assisted with planning. field
instrumentation, and conducting tests of water-distribution
systems serving Camp Lejeune. August 18-28. 2004.
Caryl J. Wipperfurth. Bonnie J. Turcoll. Patricia L.
Nohlcs, James E. Banton, and Kimberly A. Waltenbaugh.
U.S. Geological Survey Enterprise Publishing Network,
assisted with the preparation of text. illustrations. and elec-
tronic media.
A71
References-------------------------------------------
References
Agency for Toxic Substances and Disease Registry. Prelimi-
nary Puhlic Health Assessment for ABC One-Hour Clean-
ers. Jacks<mvillc, Onslow County, North Canllina. Atlanta,
CiA: U.S. Department or Health and Human Services: 1990.
Report No.: NCD024644494.
Agency for Toxic Substances and Disease Registry. Public
Health Assessment for U.S. Marine Corps Base at Camp
Lejeune, Military Rcscrvatilrn. Camp LL'.jcunl\ Onslow
County. North Carolina. Atlanta. GA: U.S. Depart-
ment of Health and Human Services: 1997. Report No.:
NC6 I 70022580.
Agency for Toxic Substances and Disease Registry. Vola-
tile Organic Compounds in Drinking Water and Adverse
Pregnancy Outcomes. United States Marine Corps Base.
Camp Lejeune. Atlanta. GA: U.S. Department or Health
and Human Services: 1998.
Agency for Toxic Substances and Disease Registry. Historical
Reconstructi{m of the Water-Distributi{m System Serving the
Dover Township Area. Ne\v Jersey: January 1962-December
1996-Summary of Findings. Atlanta. GA: Agency for
Toxic Substances and Disease Registry: 200 I September.
AWWA Engineering Computer Applicati()llS C<irnrnittcc. Cali-
bration Guidelines for Water Distribution System Modeling.
Proceedings: 1999 AWWA Information Management and
Technology Conference (IMTech): 1999 April: New Orleans.
LA: New York: American W.iter Works Association; 1999.
Bormvski EJ. aml Borwein JM. The HarperC{1llins Dictionary
of Mathematics. New York: l-larperCollins Publishers: 1991.
Bove FJ. Fulcomer MC. Klotz JB. Esmart J. Dufficy EM. and
Savrin JE. Public Drinking Water Contamination and Birth Out-
C(1111cs. A111erican .l<iurnal <>f Epidemi<)l(igy. 1995: 14:850-862.
Bove F. Shim Y. and Zeitz P. Drinking Water Contaminants
and Adverse Pregnancy Outcomes: a Review. Environ-
mental Health Perspectives. 2002: I IO(S):61-74.
Chiang W-H. and Kinzelbaeh W. JD-Groundwater Modeling
with PMWIN: A Simulation System for Modeling Ground-
water Flow and Pollution. New York: Springer-Verlag
Berlin Heidelberg: 200 I.
CLW. Camp Lejeune Water Documents. 0001-8761 (not con-
secutively available). provided on DVD format. in Maslia
ML. Sautner .IB. Faye RE. Suarez-Soto RJ. Aral MM.
Grayman WM. Jang W. Wang J. Bove FJ. Ruckart PZ.
Valenzuela C. Green JW Jr. and Krueger AL Analyses or
Groundwater Flmv. Contaminant Fate and Transport. and
Distribution or Drinking Water at Tarawa Terrace and Vicin-
ity. U.S. Marine Corps Base Camp Lejeune. North Carolina:
Historical Reconstruction and Present-Day Conditions-
Chapter A: Summary or Findings. Atlanta. GA: Agency
f(ff T(1Xic Substances and Disease Registry: 2007.
Cohn P, Klotz J, Bove F. Bcrkmvitz M. and Pagliano J. Drink-
ing Water Contarninati<lll and the Incidence of Leukemia
and Non-Hodgkin's Lymphoma. Environmental Health
Perspectives. 1994: I 02(0-7):556-561.
Costas K. Knorr RS. and Condon SK. A Case-Control Study
of Childh<)()d Leukemia in Woburn. Massachusetts: The
Relationship Bet\veen Leukemia Incidence and Exposure
to Public Drinking Water. The Science or the Total Environ-
ment. 2002:300:23-35.
Cullen AC. and Frey HC. Probabilistic Techniques in Expo-
sure Assessment: A Handbook for Dealing with Variability
and Uncertainty in Models and Inputs. New York: Plenum
Press: 1999.
Deutsch CV. and Journel AG. GSLIB Geostatical Software
Library and User's Guide. 2nd ed. New York. NY: Oxl"ord
University Press: 1998.
Doherty J. Groundwater Data Utilities/Part B: Program
Descriptions. 2005 [ cited 2007 January 3 [: Availahlc from
h l If) :IIH wu ·. sspa. co111/1 J(;,\'l/u l i Ii ties. sh 1111 I
Doherty RE. A History or the Production and the Use or Carbon
Te1rachloride. Tetrachlorocthylene. Trichlorocthylene and
I.I.I-Trichloroethane in the Uni1ed States: Part I-Historical
Background: Carbon Tetrachloride and Tetrachlornethylene.
Journal or Environmental Forensics. 2000a; I :69-8 I.
Doherty RE. A History of the Production and the Use of
Carbon Tetrachloride. Tetrachloroethylene. Trichloro-
ethylene and 1.1, I-Trichloroethane in the United States:
Part 2-Trichlorocthylcne and I, I, I-Trichloroethane.
Journal of Environmental Forensics. 2000b; I :83-93.
Faye RE. Analyses or Groundwater Flow. Contaminant Fate
and Transport. and Distribution of Drinking \\later at Tarawa
Terrace and Vicinity. U.S. Marine Corps Base Camp Lejeune.
North Can)lina: Hist(1rical Rcconstructi(in and Present-Day
Conditions-Chapter B: Geohydrologic Framework or the
Castle Hayne Aquifer System. Atlanta. GA: Agency for
Toxic Substances and Disease Registry: In press 2007a.
Faye RE. Analyses of Groundwater Flow. Contaminant Fate
and Transport, and Distribution of Drinking Waler at 'farawa
Terrace and Vicinity. U.S. Marine Corps Base Camp Lejeune.
North Carolina: Historical Reconstruction and Present-Day
Conditions-Chapter F: Simulation or the Fate and Trans-
port of Tetrachloroethylene (PCE). Atlanta. GA: Agency for
Toxic Substances and Disease Registry: In press 2007b.
Faye RE. and Green JW Jr. Analyses or Groundwater Flow.
Contaminant Fate and Transport. and Distribution of Drink-
ing Water at Tarawa Terrace and Vicinity. U.S. Marine Corps
Base Camp Lejeune. N(ff\h Carolina: Hishirical Rcc<mstruc-
tion and Present-Day Conditions-Chapter E: Occurrence
of Contaminants in Groundwater. Atlanta. GA: Agency for
Toxic Substances and Disease Registry: In press 2007.
A72 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-------------------------------------------References
Fayt: RE. and Valenzuela C. Analyses of Groundwater Flow.
Contaminant Fate and Transport. and Distribution of Drink-
ing Water at Tarawa Terrace and Vicinity. U.S. Marine
Corps Base Camp Lejeune. North Carolina: Historical
Reconstruction and Present-Day Comlitions-Chaptcr C:
Simulathin of Groundwater Flow. Atlanta. CiA: Agency for
Toxic Suhstanccs and Disease Registry; In press 2007.
Goldberg SJ, Lebowitz MD. Graver EJ. and Hicks S. An
Association or Human Congenital Cardiac Malformations
and Drinking Water Ctintaminants. Journal of the American
College of Cardiology. 1990: 15: 155-164.
Granger Laboratories. Analyses of Samples 206 and 207 from
Site Coded 'TT" and Samples 208 and 209 from Site Coded
"HP." Lcuer from B.A. Babson to Commanding General.
Marine Corp Base Camp Lejeune. August I 0, 1982.
(Camp Lejeune Water Document [CLW[ 0592).
Grayman WM, Maslia ML, and Sautner JB. Calibrating
Distribution System Models with Fire-Flow Tests. Optlmv.
American Water Work.-; Association. 2006 April: I 0-12.
Haestad Methods. Walski TM. Chase DY. Savic DA. Gray-
man WM. Beckwith S, and Koelle E. Advanced Water
Distribution Modeling and Management. I st ed. Waterbury.
CT: Hacstad Press: 2003.
Harbaugh AW. Banta ER. Hill MC. and McDonald MG.
MODFLOW-2000, The U.S. Geological Survey Modular
Ground-Water Model: User Guide to Modularization Con-
cepts and the Ground-Water Flmv Process. Reston, VA:
U.S. Geological Survey Open-File Report 00-92: 2000.
Harhaugh AW, and McDonald MG. User's Documentation of
MODFLOW-96. an Update to the U.S. Geological Survey
Modular Finite-Difference Ground-\Vatcr Flow Model. Reston.
VA: U.S. Geological Survey Open-File Report 96-458: 1996.
Henry Yon Oesen and Associates. Inc. Study of Two Water
Plants. Tarawa Terrace-Montford Poinl. Camp Lejeune.
North Carolina: Henry Yon Ocscn and Associates. l1H.:.: 1979.
Hill MC, and Tiedeman CR. Effective Groundwater Model
Calibration. Hohoken: John Wiley & Sons. Inc.: 2007.
Huling SG. and Weaver JW. Dense Nonaqueous Phase
Liquids. \Vashington. D.C.: U.S. Environmental Protection
Agency. Office of Research and Development and Ortice or
Solid Waste and Emergency Response: 1991 March. Report
No.: EPA/540/4-91-002.
International Union of Pure and Applied Chemistry. Compen-
dium of Chemical Terminology. 2006. I cited 2006 Fehru-
ary I]: Available from htr1>:ll11·11·11·.i11pac.orglp11blicatio11sl
c( m11>endi 11111/i ndex. ht 111 I
Jang W. and Aral MM. Three-Dimensional Multiphase Flow and
Multi-Species Trausport Model. TechFlowMP. Atlanta. GA:
Multimedia Environmental Simulations Laboratory (MESL).
School of Civil and Environmental Engineering. Georgia Insti-
tute of Technology: 2005 September. Report No. MESL-02-05.
Chapter A: Summary of Findings
Jang W. and Aral MM. Simulation of Three-Dimensional
Multi-Species, Multi-Phase Mass Transport of Tetrachloro-
ethylcne (PCE) and Associated Degradation By-Products.
Atlanta. GA: Multimedia Environmental Simulation
Laboratory (MESL). School of Civil and Environmental
Engineering. Georgia Institute of Teclrnology: 2007 March.
Report No.: MESL-02-07.
Jang W. and Aral MM. Analyses of Groundwater Flow, Con-
taminant Fate and Transport. and Distribution of Drinking
Water at Tarawa Terrace and Vicinity. U.S. Marine Corps
Base Camp Lejeune. North Carolina: Historical Reconstruc-
tion and Present-Day Conditions-Chapter G: Simulation
()fThree-Dimensional Multispecies, Multiphase Mass
Transport ofTetrachll)roethylcne (PCE) and Associated
Degradation By-Products. Atlanta. GA: Agency f<ir Toxic
Substances and Disease Registry: In press 2007.
Kucper BH. Wealthall GP, Smith JWN. Leharnc SA. and
Lerner DN. 2003. An Illustrated Handbook of DNAPL
Transport and Fate in the Subsurface. Environment Agency,
R&D Publication 133. 2003 June, Bristol. United Kingdom.
Lawrence SJ. Description. Properties. and Degradation of
Selected Volatile Organic Compounds Detected in Ground
Water-A Review of Selected Literature. Atlanta. GA: U.S.
Geological Survey Open-File Report 2006-1338: 2006. A
Web-only publication al ltttJJ.i/p11hs.11sgs.g(J11/oji-/2006/I 338/.
Lawrence SJ. Analyses or Groundwater Flow. Contaminant
Fate and Transport. and Distribution of Drinking Water
at Tara\va Terrace and Vicinity, U.S. Marine Corps Base
Camp Lejeune. North Carolina: Historical Reconstruction
and Present-Day Conditions-Chapter D: Properties of
Degradation Pathways of Common Organic Compounds in
Groundwater. Atlanta. GA: Agency for Toxic Substances
and Disease Registry: In press 2007.
Mackay DM. and Cherry JA. Groundwater Contamination:
Pump-and-Treat Remcdiati<m. Environmental Science and
Technology. l 989:23(6):630-636.
Maslia ML. Expert Peer Review Panel Evaluating ATSDR's
Water-Modeling Activities in Support of the Current Study
or Childhood Birth Defects and Cancer at U.S. Marine
Corps Base Camp Lejeune. North Carolina: Atlanta. GA:
Agency for Toxic Substances and Disease Registry; 2005.
Maslia ML. and Aral MM. Analytical Contaminant Transport
Analysis System (ACTS)-Multimedia Environmental Fate
and Transport. Practice Periodical of Hazardous. Toxic. and
Radioactive Waste Management. 2004 July: 181-198.
Maslia ML. Sautner JB. and Aral MM. Analysis of the 1998
Water-Distribution System Serving the Dover Township
Arca. New Jersey: Field-Data Collection Activities and
\Yater-Distribution System Modeling. Atlanta. GA: Agency
for Toxic Substances and Disease Registry: 2000.
A73
References-------------------------------------------
Maslia ML. Sautner JB. Aral MM. Ciillig RE. Reyes JJ, and
\Villiams RC. Historical Reconstruction of the Water-Distri-
bution System Serving the Dover Township Arca. New Jer-
sey: January 1962-Dccembcr 1996. Atlanta. GA: Agency
for Toxic Suh . .;tanccs and Disease Registry: 2001 October.
Maslia ML. Sautner .Ill. Faye RE. Suarez-Soto RJ. Aral MM.
Gray1nan WM, Jang W, Wang J, Bove F.J, Ruckart PZ.
Valenzuela C. Green JW Jr. and Krueger AL. Analyses of
Groundwater Flow, Contaminant Fate and Transport. and
Distribution or Drinking Water at Tarawa Terrace and Vicin-
ity, U.S. Marine Corps Base Camp Lejeune. North Carolina:
Historical Reconstruction and Present-Day Conditions-
Chapter A: Summary of Findings. Atlanta. GA: Agency for
Toxic Substances and Disease Registry; 2007.
Maslia ML. Sautner JB. Faye RE. Su,ircz-Soto RJ. Aral MM.
Cirayman WM. Jang W. Wang J, Bove FJ. Ruckart PZ,
Valenzuela C. Green J\V Jr. and Krueger AL. Analyses of
Groundwater Flow, Contaminant Fate and Transport, and
Distribution of Drinking Water at 'fara\Va Terrace and Vicin-
ity, U.S. Marine Corps Base Camp Lejeune. North Carolina:
Historical Reconstruction and Present-Day Conditions-
Chapter K: Supplemental Information. Atlanta. GA: Agency
for Toxic Substances an<l Disease Registry: In pn.:ss 2007a.
Maslia Ml". Sautner JB. Valenzuela C. Cirayman WM. Aral MM.
and Green J\V Jr. Usc of Continuous Recording Water-Quality
Monitoring Equipment for Conducting Water-Distribution
System Traci..;r Tests: The Good, thc Bad, and the Ugly.
Proceedings: World Environmental and Water Resources
Congress: 2005 May 15-19; Anchorage, AK: 2005.
Maslia ML. Suarez-Soto RJ. Wang J. Aral MM. Sautner Jll.
and Valenzuela C. Analyses of Groundwater Flow. Con-
taminant Fate and Transport. and Distribution of Drinking
Water at Tarawa Terrace and Vicinity. U.S. Marine Corps
Base Camp Lejeune. North Carolina: Historical Reconstruc-
tion and Present-Day Conditions-Chapter I: Parameter
Sensitivity. Uncertainty, and Variability Associated with
lvlodel Simulations of Ground\vater Flow. Contaminant Fate
and Transport. and Distribution or Drinking Water. Atlanta,
GA: Agency for Toxic Substances and Disease Registry:
In press 2007b.
Massachusetts Department or Public Health. Centers for
Disease Control and Prevention. and Massachusetts Health
Research Institute. Final Report of the Woburn Environmen-
tal and Birth Study. Boston, MA: Massachusclls Depart-
ment of Public Health: 1996.
Masters GM. Introduction to Environmental Engineering and
Science. 2nd ed: Pn.:ntice Hall: 1998.
Melts VJ. Deposition, in the General Court of Justice. Superior
Court Division: 2001 April 12. Report No.: 0I-CVS-566.
New Jersey Department uf Health and Senior Services. Case-
Control Study of Childhood Cancers in Dover Tmvnship
(Ocean County), New Jersey. Trenton. NJ; New Jersey
Department of Health and Senior Services: 2003.
Pankow JF, and Cherry .IA. Dense Chlorinated Solvents and
Other DNAPLs in Groundwater: History. Behavior. and
Remediation: Waterloo Press; 1996.
Rodcnhcck SE. Sanderson LM. and Rene A. Maternal Expo-
sure to Trichloroethylene in Drinking Water and Birth-
Weight Outcomes. Archives of Environmental Health.
2000:55: 188-194.
Rogers JF. Kilough GG. Thompson SJ. Addy CL. Mc Ke-
own RE, and Cowen DJ. Estimating Environmental Expo-
sures to Sulfur Dioxide from Multiple Industrial Sources for
a Case-Control Study. Journal of Exposure Analysis
and Environmental Epidemiology. 1999:9:535-545.
Rossman LA. EPANET 2 Users Manual. Cincinnati: U.S.
Envininrnental Protecti(m Agency. Natilrnal Risk Manage-
ment Research Lahoratory: 2000 September. Report No.:
EPA/600-R-00/057.
Roy F. West(rn. Inc. Remedial Investigation Report.
ABC One-Hour Cleaners. Jacksonville. North Carolina:
Roy F. Weston. Inc.: 1992.
Roy F. Weston. Inc. Remedial Investigation. ABC One-Hour
Cleaners. Operable Unit 2. Jacksonville. North Carolina:
Roy F. Weston, Inc.; 1994.
Saltelli A. Chan K. and Scott EM. editors. Sensitivity An~ily-
sis. Chichester, England: John Wiley & Sons, Ltd; 2000.
Sautner JB. Maslia ML. and Grayman WM. Storage 'fonk
Mixing Models: Comparison of Tracer Data with Model
Simulations. World Envininmcntal and Water Resources
Congress: 2007 May 15-19: Tampa, FL: 2007.
Sautner JB. Maslia ML. Valenzuela C. Grayman WM.
Aral MM. and Green JW Jr. Field Testing or Water-Distri-
bution Systems at the U.S. Marine Corps Base. Camp Leje-
une, North Carolina. in Support or an Epidcmiologic Study.
Proceedings: World Environmental and Water Resources
Congress: 2005 May 15-19: Anchorage. AK: 2005.
Sautner JB. Valenzuela C. Maslia ML. and Grayman WM.
Analyses of Groundwater Flow. Contaminant Fate and
Transport. and Distribution of Drinking Water at Tarawa
Terrace and Vicinity. U.S. Marine Corps Base Camp
Lejeune, North Carolina: Historical Reconstruction and
Present-Day Conditions-Chapter J: Field Tests. Data
Analyses. and Simulation of the Distrihution of Drinking
Water. Atlanta. GA: Agency for Toxic Substances and
Disease Registry: In press 2007.
Schwartz FW. and Zhang H. Fundamentals of Ground Water.
New York: John Wiley & Sons, Inc.: 2003.
A74 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-------------------------------------------References
Sclnville F. Dense Chlorinated Solvents in Porous and Frac-
tured Media. Translated by Pankow. JF. Boca Raton: Lewis
Publishers (CRC Press): 1988.
Shiver R. A Groundwater Investigation to Deline the Sourcc(s)
of Tetrachlorocthylenc that Have Contaminated Three
Community Water Supply Wells at Tarawa Terrace L
Camp Lejeune Marine Corps Base. Onslow County: North
Carolina Department (if Natural Resources and Community
Development: 1985.
Sonnenfeld N. Hcrtz-Picciotto I. and Kaye WE. Tctrachlo-
rocthylcnc in Drinking Water and Birth Outcomes at the
U.S. Marine Corps Base at Camp Lejeune. North Carolina.
American Journal of Epidemiology. 200 I: 154( I 0):902-908.
Tung Y-K. and Yen B-C. Hydrosystems Engineering Uncer-
tainty Analysis. New York: McGraw-Hill: 2005.
U.S. Environmental Protection Agency. Contaminant Specific
Fact Sheets, Volatile Organic Chemicals-Technical Vasion.
EPA 81 l-F-95-004-T. 1995. [cited 2006 January 11 [: Avail-
ahlc from http://11·ww.c1)(/.gm'/.w{/'ewater/chvhlt-voc.hrl/ll
U.S. Environmental Protection Agency. Guiding Principles
for Monte Carlo Analysis. Washington. DC: United States
Environmental Protection Agency: 1997 March. Report
No.: EPA 630-R-97-001.
U.S. Environmental Protection Agency. National Primary
Drinking \Vatcr Standards. Report No.: EPA 8 I 6-F-03-0 I 6.
21103. [cited 21107 January 25 [: Available from h1tp:llwww.
qw.gov/H·atcr/
Chapter A: Summary of Findings
U.S. Environmental Protection Agency. How to Evaluate
Alternative Cleanup Technologies for Underground Storage
Tank Sites-A Guide for Corrective Action Plan Reviewers.
Washington. D.C.: U.S. Environmental Protection Agency:
2004. Report No.: EPA 510-B-95-007. Available from
ht1p://w111w. CJHl.gm/nremst I lp11h.,!tums.ht111
Wang J, and Aral MM. Effect of Groundwater Pumping
Schedule Variation on Arrival of Tetrachloroethylcne (PCE)
at Water-Supply Wells and the Water Treatment Plant.
Atlanta. GA: Multimedia Environmental Simulation Labo-
ratory (MESL). School of Civil and Environmental Engi-
lll.;ering. Georgia Institute of Technology: 2007 January.
Report No.: MESL-01-07.
Wang J. and Aral MM. Analyses of Groundwatl.;r Flow, Con-
taminant Fate and Transport. and Distribution of Drinking
Water at Tarawa Terrace and Vicinity. U.S. Marine Corps
Base Camp Lejeune. North Carolina: Historical Recon-
struction and Present-Day Conditions-Chapter H: Effect
of Groundwater Pumping Schedule Variation on Arrival
ofTetrachlorocthylenc (PCE) at Water-Supply Wells and
the Water Treatment Plant. Atlanta. GA: Agency for Toxic
Suhstanccs and Disease Registry; In press 2007.
Zheng C. and Bennett GD. Applied Contaminant Transport
Modeling. 2nd ed. New York: John Wiley & Sons, Inc.: 2002.
Zheng C. and Wang PP. MTJDMS: A Modular Three-
Dimensional Multi-Species Model for Simulation of
Advcction, Dispersion. and Chemical Reactions of Contam-
inants in Groundwater Systems: Documentation and User's
Guide. Vicksburg. MS: U.S. Anny Engineer Research and
Development Center: 1999. Report No.: SERDP-99.
A75
A76 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-----------------------Appendix Al. Summaries of Tarawa Terrace Chapter Reports
Appendix A1. Summaries of Tarawa Terrace Chapter Reports
Summaries of Tarawa Terrace chapter reports are
described below. Electronic versions of each chapter
report and their supporting information and data will be
made available on the ATSDR Camp Lejeune Web site at
Ii It p ://\ n nv. a tsd 1: cdc. g( J vlsi t es/I e je une/i nde.r.I1t111 /.
Chapter A: Summary of Findings (Maslia ct
al. 2007-this report) provides a summary or detailed
technical findings (described in Chapters B-K) focusing
on the historical reconstruction analysis and present-day
conditions of groundwater flow. contaminant fate and
transport, and distrihution of drinking water at Tarawa
Terrace and vicinity. Among the topics that this report
summarizes are: (I) methods of analyses, (2) data
sources and requirements, (3) the four-stage hierarchical
approach used ror model calibration and estimating PCE
concentrations in drinking water. (4) presentation. dis-
cussion. and implications of selected simulation results
ror PCE and its degradation by-products, and (5) quanti-
fying confidence in simulation results by varying water-
supply well historical pumping schedules and by using
sensitivity and probahilistic analyses to address issues
of uncertainty and variability in model parameters. In
addition. this report provides a searchable electronic
database-using digital video disc (DVD) format-of
information and data sources used to conduct the histori-
cal reconstruction analysis. Data were obtained from
a variety of sources. including ATSDR, USEPA. Envi-
ronmental Management Division of U.S. Marine Corps
Base Camp Lejeune, U.S. Geological Survey. private
consulting organizations, published scientific literature,
and community groups representing former marines and
their families.
Chapter A: Summary of Findings
Chapter II: Geohydrologic Framework of the
Castle Hayne Aquifer System (Faye In press 2007a)
provides detailed analyses of well and geohydrologic
data used to develop the geohydrologic framework of
the Castle Hayne aquifer system at Tarawa Terrace and
vicinity. Potentiometric levels, horizontal hydraulic con-
ductivity. and the geohydrologic fra111cwork of the Castle
Hayne aquirer system east or the New River are described
an<l quantilic<l. The geohydrologic framework is com-
posed or 11 units. 7 of which correspond to the Upper,
Middle. and Lower Castle Hayne aquifers and related
confining units. Overlying the Upper Castle Hayne
aquifer arc the Brewster Boulevard and Tarawa Terrace
aquifers and confining units. Much or the Castle Hayne
aquifer system is composed or line. rossilirerous sand.
limestone. and shell limestone. The sands are frequently
silty and contain beds and lenses or clay. Li111estone units
are probably discontinuous and occasionally cavernous.
Confining units are characterized by clays and silty clays
of significant thickness and are persistent across much of
the study area. Maximum thickness or the Castle Hayne
aquircr syste111 within the study area is ahout 300 ft. In
general, gcohydrologic units thicken from northwest to
the south and southeast. The limestones and sands of the
Castle Hayne aquirer syste111 readily yield water to wells.
Aquifer-test analyses indicate that horizontal hydraulic
conductivities of water-bearing units at supply wells
commonly range from IO to 30 reet per day. Esti111ated
predevelop111ent potentio111etric levels of the Upper and
Middle Castle Hayne aquifers indicate that groundwater-
flow directions are from highland areas north and cast of
the study area toward the major drainages of New River
and Northeast Creek.
A77
Appendix Al. Summaries of Tarawa Terrace Chapter Reports------------------------
Chapter C: Simulation of' Groundwater Flow
(Faye and Valenzuela In press 2007) provides detailed
analyses or groundwater flow at Tarawa Terrace and
vicinity, including the development of a predcvelop-
ment (steady-state) and transient groundwater-flow
nwdel using the model code MODFLOW-96 (Harbaugh
and McDonald 1996). Calibration and testing of the
model arc thoroughly described. The groundwatcr-
llow model was designed with seven layers largely
representing the Castle Hayne aquifer system. Com-
parison of 59 observed water levels representing esti-
mated predcvclopment comlitions and corrcsp(mding
simulated potentiomctric levels indicated a high degree
of similarity throughout most of the study area. The
average absolute difference between simulated am!
observed prcdevelopmcnt water levels was 1.9 ft1 and
the root-mean-square (RMS) of differences was 2.1 ft.
Transient simulations represented pumping at Tarawa
Terrace supply wells for 528 stress periods representing
528 months-January 195 I-December I 994. Assigned
pumpage at supply wells was estimated using reported
well-capacity rates and annual rates of raw water treated
at the Tarawa Terrace water treatment plant (WTP) dur-
ing 1975-1986. Calibrated model results of263 paired
water levels representing observed and simulated water
levels at monitor wells indicated an average absolute
difference hetween simulated and observed water levels
of 1.4 ft, a standard deviation of water-level difference
of 0.9 ft. and a RMS of water-level difference of 1.7 ft.
Calihrated model results of 526 paired water levels
representing observed and simulated water levels at
water-supply wells indicated an average absolute dif-
ference hetween simulated and ohserved water levels of
7.1 rt, a standard deviation of water-level difference of
4.6 rt. and a RMS of water-level difference of 8.5 ft.
Chapter D: Properties of' Degradation Pathways
of Common Organic Compounds in Groundwater
(Lawrence In press 2007) describes and summarizes
the properties. degradation pathways, and degrada-
tion by-products of VOCs (non-trihalomethanc) com-
monly detected in groundwater contamination sites
in the United States. This chapter also is published as
U.S. Geological Survey Open-File Report 2006-1338
(Lawrence 2006) and provides abridged information
describing the most salient properties and biodegrada-
tion of 27 VOCs. This report cross-references common
names and synonyms associated with VOCs with the
naming conventions supported by the IUPAC. In addi-
tion, the report describes basic physical characteristics of
those compounds such as Henry's Law constant. water
solubility, density. octanol-water partition (log K,,w ). and
organic carbon partition (log K,,,.) coeflicicnts. Descrip-
tions and illustrations are provided for natural and labo-
ratory biodegradation rates, chemical by-products. and
degradation pathways.
Chapter E: Occurrence of' Contaminants iu
Groundwater (Faye and Green In press 2007) describes
the occurrence and distribution of PCE and related
contaminants within the Tarawa Terrace aquifer and the
Upper Castle Hayne aquifer system at and in the vicin-
ity of the Tarawa Terrace housing area. The occurrence
and distribution of benzene, toluene. cthylbenzenc. and
xylene (BTEX) and related compounds also arc briefly
described. This report describes details of historical
investigations ()r VOC contamination <)r groundwater
at Tarawa Terrace with emphasis on water-supply wells
TT-23, TT-25. and TT-26 (Figure A I). Detailed analyses
of concentrations of PCE at monitor wells. at hydrocone
sample locations, and at Tarawa Terrace water-supply
wells during the period 1991-1993 were suflicient to
estimate the mass of PCE remaining in the Tarawa Ter-
race and Upper Castle Hayne aquifers. Similar methods
were applied to compute the mass of PCE in the unsatu-
rated zone (zone above the water tahle) at and in the
vicinity of ABC One-Hour Cleaners using concentration-
depth data determined from soil borings. The total mass
of PCE computed in groundwater and within the unsatu-
rated zone equals about 6,000 pounds and equates to a
volume of about 430 gallons. This volume represents an
average minimum loss rate of PCE to the suhsurface at
ABC One-Hour Cleaners of about 13 gallons per year
for the period 1953-1985.
Chapter F: Simulation of' the Fate and Transport
of Tetrachloroethylene (PCE) in Groundwater (Faye
In press 2007b) dcscrihes: (I) the fate and transport of
PCE in groundwater from the vicinity of ABC One-Hour
Cleaners to the intrusion of PCE into individual water-
supply wells (for example, TT-23 and TT-26. Figure A I).
and (2) the concentration of PCE in finished water at the
Tarawa Terrace WTP computed using a materials mass
balance model (simple mixing). The materials mass
balance model was used to compute a flow-weighted
A78 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-----------------------Appendix Al. Summaries of Tarawa Terrace Chapter Reports
average PCE concentration. which was assigned as the
linished water concentration at the Tarawa Terrace WTP
for a specilied month. The contaminant fate and trans-
pDrt simulation was conducted using the code MT3DMS
(Zheng and Wang 1999) integrated with the calihrated
groundwater-flow mDdel (Faye and Valenzuela In press
2007) based on the code MODFLOW-96. Simulated
mass loading occurred al a constant rate or 1.200 grams
per day using monthly stress periods representing the
period January 1953-Decemher 1984. The complete
simulation time was represented by the period Janu-
ary 1951-December 1994. Until 1984. the vast major-
ity Df simulated PCE-contaminated groundwater was
supplied 10 the Tarawa Terrace WTP by well TT-26.
Simulated breakthrough of PCE at well TT-26 at the
current MCL of 5 µg/L occurred during January 1957.
CDrresponding breakthrough al the location of well
Tl~23 occurred during December 1974: however. well
TT-23 was not operational until about August 1984.
Simulated maximum and average PCE concentrations
at well TT-26 following breakthrough were 851 ftg/L
and 414 µg/L. respectively. Corresponding maximum
and average concentrations at well TT-23 subsequent
ID the onset of operations were 274 ftg/L and 252 ftg/L.
respectively. Simulated breakthrough of PCE in linished
water at the Tarawa Terrace WTP occurred at the current
MCL concentration Df 5 µg/L during November 1957
and remained at or ahove a concentration of 40 ~tg/L
from May 1960 until the termination of pumping at
water-supply well TT-26 during February 1985. Com-
puted maximum and average PCE concentrations at the
WTP were 183 ftg/L and 70 ftg/L. respectively. during
the period November 1957-February 1985, when well
TT-26 was removed from service.
Chapter G: Simulation of' Three-Dimensional
Multispecies, Multiphase Mass Transport of' Tetra-
chloroethylcne (PCE) and Associated Degradation
By-Products (Jang and Aral In press 2007) provides
detailed descriptions and analyses or the develop-
ment and application or a three-dimensional model
(TechFlowMP) capable of simulating multispecies and
multiphase (water and vapor) transport of PCE and
associated degradation by-products-TCE, 1.2-tDCE,
and VC. The develDpment of the TechFLowMP model
is described in Jang and Aral (2005) and its application
to Tarawa Terrace and vicinity also is published as report
Chapter A: Summary of Findings
MESL-02-07 by the Multimedia Environmental Simula-
tions Laboratory in the School of Civil and Environmen-
tal Engineering, Georgia Institute of Technology (Jang
and Aral 2007). Simulation results show that the maxi-
mum concentrations of PCE degradation by-products,
TCE. 1,2-tDCE. and VC. generally ranged between
10 ftg/L and 100 ftg/L in Tarawa Terrace water-supply
well T'l'26 and between 2 ftg/L and 15 µg/L in finished
water delivered from the 'forawa Terrace WTP. As part
of the degradation by-product simulation using the
TechFlowMP model. results were ohtained for PCE and
PCE degradation by-products dissolved in groundwater
and in the vapor phase (above the water table in the
unsaturated zone). Analyses of the distribution of vapor-
phase PCE and PCE degradation by-products indicate
there is potential for vapors to enter buildings at Tarawa
Terrace, thereby providing a potential exposure pathway
from inhalation of PCE and PCE degradation by-product
vapors. At Tarawa Terrace these buildings would include
family housing and the elementary school.
Chapter H: Effect of' Groundwater Pumping
Schedule Variation on Arrival of' Tetrachloroethylene
(PCE) at Water-Supply Wells and the Water Treat-
ment Plant (Wang and Aral In press 2007) describes a
detailed analysis of the effect of groundwater pumping
schedule variation on the arrival of PCE at water-supply
wells and at the Tarawa Terrace WTP. Analyses con-
tained in this chapter used the calibrated model param-
eters described in Chapter C (Faye and Valenzuela In
press 2007) and Chapter F (Faye In press 2007b) reports
in combination with the groundwater pumping schedule
optimization system simulation tool (PSOpS) to assess
the influence of unknown and uncertain historical well
operations at Tarawa Terrace water-supply wells on PCE
concentrations at water-supply wells and at the Tarawa
Terrace WTP. This chapter also is published as report
MESL-01-07 by the Multimedia Environmental Simu-
lations Laboratory in the School of Civil and Environ-
mental Engineering, Georgia Instilule of Technology
(Wang and Aral 2007). Variation in the optimal pumping
schedules indicates that the arrival time of PCE exceed-
ing the current MCL of 5 11g/L at water-supply well
TT-26 varied between May 1956 and August 1959. The
corresponding arrival time of PCE exceeding the current
MCL of 5 11g/L at the Tarawa Terrace WTP varied
between December 1956 and June 1960.
A79
Appendix Al. Summaries ofTarawa Terrace Chapter Reports------------------------
Chapter I: Parameter Sensitivity, Uncertainty,
and Variahility Associated with Model Simulations
of Groundwater Flow, Contaminant Fate and Trans-
port, and Distribution of Drinking Water (Maslia
ct al. In press 2007b) describes the development and
application of a probabilistic analysis using Monte Carlo
and sequential Gaussian simulation analysis to quantify
uncertainty and variability of groundwater hydraulic
and transport parameters. These analyses demonstrate
quantitatively the high reliability and confidence in
results determined using the calihrated parameters
from the MODFLOW-96 and MT3DMS models. For
example, 95% or Monte Carlo simulations indicated
that the current MCL for PCE or 5 f.tg/L was exceeded
in finished water at the Tarawa Terrace WTP between
October 1957 and August 1958: the corresponding
breakthrough simulated by the calibrated fate and trans-
port model (Chapter F report. Faye lln press 2007bl)
occurred during November 1957.
Chapter .I: Field Tests, Data Analyses, and Sirnu-
lation of the Distribution of Drinking Water (Sautner
et al. In press 2007) describes field tests. data analyses,
and the simulation of drinking-water supply at Tarawa
Terrace and vicinity. Details of the development and cali-
bration of a water-distrihution system model for Tarawa
Terrace and vicinity arc dcscrihed based on applying the
model code EPANET 2 (Rossman 2000) to the study
area. Comparisons arc provided bet ween the PCE con-
centrations computed by Faye (In press 2007b) using a
simple mixing model and the more complex and detailed
approach of Sautner ct al. (In press 2007) that is based
on a numerical water-distribution system model. Results
of simulations conducted using cxtcndc<l period simula-
tion conlinn the assumption that, on a monthly basis,
the concentrations of PCE in drinking water delivered to
residential housing areas throughout Tarawa Terrace arc
the same as the concentrations of PCE in finished water
at the Tarawa Terrace WTP. Therefore. a simple mixing
model based on the principles of continuity and con-
servation of mass was an appropriate model to use for
determining the concentration of PCE in finished water
delivered from the Tarawa Terrace WTP.
Chapter K: Supplemental Information (Maslia
ct al. In press 2007a) presents additional information
such as (I) a tabular listing of water-supply well pump-
age by stress period (month and year): (2) synoptic maps
showing groundwater levels. directions of groundwater
flow, and the simulated distribution of PCE; (3) a tabular
listing of simulated monthly concentrations of PCE dis-
solved in groundwater at Tarawa Terrace water-supply
wells; (4) a tabular listing or simulated monthly con-
centrations of PCE and PCE degradation by-products-
TCE, 1,2-tDCE, and VC at the Tarawa Terrace WTP;
(5) a complete list of references used in conducting the
water-modeling analyses and historical reconstruction
process; and (6) other ancillary information and data
that were used during the water-modeling analyses and
historical reconstruction process.
ABO Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
----------Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant
Appendix A2. Simulated PCE and PCE Degradation By-Products
in Finished Water, Tarawa Terrace Water Treatment Plant,
January 1951-March 1987
Chapter A: Summary of Findings A81
Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant---------------
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 1987.1
[l'CE. tetrnchlorncthylcnc: ftg/L. microgram per liter: 1.2-tDCE. tmn.1-1.2-dichlorocthylene: TCE. trichlorocthylene: VC. \'inyl d1loridc: WTP. water treatment plant)
1-12 Jan-Dec 1951
13 Jan 1952
14 Feb 1952
15 Mar 1952
16 Apr 1952
17 May 1952
18 June 1952
19 July 1952
20 Aug 1952
21 Sept 1952
22 Oct 1952
23 Nov 1952
24 Dec 1952
25 Jan 1953
26 Feh 1953
27 Mar 1953
28 Apr 1953
29 May 1953
30 June 1953
31 July I 953
32 Aug 1953
33 Sept 1953
34 Oct 1953
35 Nov I 953
36 Dec 1953
37 Jan 1954
38 Feb 1954
39 Mar 1954
40 Apr 1954
41 May 1954
42 June 1954
43 July 1954
44 Aug 1954
45 Sept 1954
46 Oct 1954
47 Nov 1954
48 Dec 1954
A82
WTP not WTP not WTP not WTP not WTP not
opcrnting
0.00
0.00
(1.00
11.1111
(1.00
0.00
0.011
(1.110
0.00
0.110
0.00
0.1111
II.OIi
0.00
11.1111
(1.110
(1.00
11.00
11.1111
11.1111
0.110
0.110
0.00
II.OIi
0.00
0.00
0.00
().()()
0.00
0.00
0.1)0
0.1111
(1.00
0.00
0.00
0.00
01>crating operating operating operating
(I.I)() 0.00 0.00 0.00
0.00 0.00 0.00 0.00
0.00 0.00 0.00 11.1111
11.1111 0.00 0.011 11.1111
(1.110 0.00 0.00 0.00
0.00 0.00 0.00 0.00
11.110 0.00 0.00 0.00
11.00 II.OIi 0.00 II.OIi
(1.00 0.00 0.00 11.00
11.00 (l.00 0.00 0.00
().()() 0.00 0.00 II.Oil
0.011 11.00 0.00 11.1111
11.1111 0.00 0.00 11.1111
11.110 0.00 0.011 11.110
11.1111 0.00 0.1111 0.1111
11.1111 0.110 0.1111 11.1111
11.110 0.00 0.00 (I.OIi
11.1111 0.00 0.00 11.1111
II.OIi 11.110 0.00 11.1111
11.1111 11.110 (I. 1111 11.1111
11.1111 0.00 0.1111 11.1111
0.1111 11.00 0.00 11.1111
II.till 11.110 (1.1111 11.1111
11.1111 11.110 0.110 11.1111
0.00 0.00 0.00 11.110
0.110 0.00 0.00 (1.00
0.110 0.00 0.00 ().()()
0.1111 (1.00 0.1)0 11.1111
ll.00 0.00 0.00 0.110
0.00 0.00 0.00 0.00
0.00 0.00 0.00 0.00
0.1111 (1.00 0.00 11.1111
0.110 0.00 0.00 (1.110
0.00 0.00 0.00 0.00
0.1)1) (l.00 0.00 ll.00
0.110 0.00 0.011 0.1111
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 19871.-Continued
[PCE. lctrachlomc1hylcnc: ftg/L. microgram per liter: 1.2-tDCE. t1wH-! .2-dichloroethylenc: TCE. trichlorocthylene: VC vinyl chloride: \VTP. water treatment plant]
,, . .,,,, ; --~··"'"•>, ,, _,.,__.,.,, .,,,"<::O·'.• ,-... ,.,w,,.
49 Jan 1955 0.00 0.011 ().(Ill 11.1111 0.11 I
511 Fch 1955 (1.1111 II.OIi 11.0I 11.1111 0.11 I
51 Mar 1955 II.OIi II.II I II.II I 11.1111 II.II I
52 Apr 1955 II.OIi II.II I II.II I 11.1111 11.112
53 ~fay 1955 (I.Oil 0.11 I O.OI (I.OIi 11.112
54 June 1955 0.01 0.01 0.112 11.00 1).()3
55 July 1955 11.0 I ().()2 11.113 11.1111 11.113
56 Aug 1955 II.II I 11.113 11.113 11.1111 11.114
57 Sept 1955 0.02 11.04 0.114 II.Oil 11.115
58 Oct 1955 0.03 0.115 (1.05 11.110 0.117
59 Nov 1955 11.04 0.116 0.117 11.1111 11.08
61) Dec: 1955 11.06 0.118 11.118 II.II I II. 111
61 Jan 1956 0.08 0.11 0.10 0.0 I 0.12
62 Feb 1956 O.IO 0.14 0.12 0.01 0.14
63 Mar 1956 0. 13 0.17 0.15 0.0 I 0.17
64 Apr 1956 0.17 0.22 0.18 (1.01 0.20
65 May 1956 0.23 0.27 0.21 11.02 0.23
66 June 1956 0.29 0.33 0.25 0.02 0.26
67 July 1956 0.36 0.40 0.29 0.02 11.30
68 Aug 1956 0.46 0.49 0.33 ().()3 0.34
69 Sept 1956 0.57 0.59 0.38 11.113 11.39
70 Oct 1956 0.70 0.70 0.44 0.114 11.44
71 Nov 1956 0.85 0.83 0.50 (1.05 0.49
72 Dec 1956 1.04 0.97 0.57 0.06 0.55
73 Jan 1957 1.25 1.14 0.64 11.116 0.61
74 Feb 1957 1.47 1.33 11.72 0.117 0.6H
75 Mar 1957 1.74 1.52 11.79 0.118 II. 74
76 Apr 1957 2.04 1.75 11.88 0. 10 11.81
77 May 1957 2.39 2.00 0.97 II. I I 11.89
78 June 1957 2.77 2.28 1.118 11.12 11.97
79 July 1957 3.21 2.59 1.18 11.14 1.115
80 Aug 1957 3.69 2.93 !.29 II. I 6 1.13
81 Sept 1957 4.21 3.30 1.41 11.17 1.23
82 Oct 1957 4.79 3.69 1.53 11.19 1.32
83 Nov 1957 5.41 4.13 1.66 0.22 1.4 I
84 Dec I 957 6.111 4.59 1.811 11.24 1.51
Chapter A: Summary of Findings A83
Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant---------------
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 19871.-Continued
I PCE. tetrachloroc!hylenc: 11g/L. microgram per litn: 1-2-tDCE. 11w1,1-I .2-did1!nrncthylc11c: TCE. trichloroethylenc: VC. \'inyl chloride: WTP. water treat men I plant I
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
11111
1111
102
103
1114
1115
1116
1117
108
109
110
111
112
I 13
114
115
I 16
117
118
119
120
A84
Jan 1958
Feb 1958
Mar 1958
Apr 1958
May 1958
June 1958
July 1958
Aug 1958
Sept 1958
Oct 1958
Nov 1958
Dec 1958
Jan 1959
Feb 1959
Mar 1959
Apr 1959
May 195lJ
June 1959
July 1959
Aug 1959
Sept 1959
Ot:1 1959
Nov 1959
Dec 1959
Jan 1960
Fch 1960
Mar 1960
Apr 1%0
May 1960
June 1960
Ju!y 1960
Aug 1960
Sept 1960
Oct 1960
Nov 1960
Dec 1960
6.86
7.60
8.47
9.37
10.37
11.39
12.91
14.12
15.35
16.69
18.03
19.49
20.97
22.35
23.92
25.49
27.15
28.81
311.56
32.36
34.14
36.111
37.85
39.78
41.86
43.85
46.03
48.15
50.37
52.51
54.74
56.96
59.09
61.30
63.42
65.61
5.11 1.94 0.26 1.62
5.6S 2.09 0.29 1.72
6.17 2.22 0.31 1.81
6.79 2.38 0.34 1.92
7.41 2.53 0.37 2.02
8.IO 2.70 0.41 2.13
9.09 2.96 11.45 2.32
9.88 3.14 11.49 2.44
10.73 3.33 0.53 2.56
11.58 3.52 0.57 2.68
12.52 3.72 0.61 2.81
13.46 3.92 0.66 2.94
14.48 4. 13 11.71 3.117
15.54 4.34 0.76 3.21
16.54 4.54 0.80 3.33
17.70 4.77 0.85 3.48
18.84 4.99 0.91 J.61
2().()9 5.23 0.96 3.77
21.34 5.46 1.02 3.91
22.66 5.69 1.08 4.05
24.111 5.93 1.14 4.19
2S.3S 6.16 1.20 4.32
26.77 6.411 1.27 4.46
28.18 6.64 1.33 4.60
29.67 6.88 1.40 4.74
31.17 7.12 1.46 4.86
32.58 7.33 1.52 4.97
34.16 7.57 1.59 5.10
35.67 7.79 1.66 5.21
37.24 8.03 1.73 5.33
38.79 8.26 1.80 5.45
40.45 8.51 1.87 5.59
42.13 8.76 1.94 5.73
43.80 9.02 2.02 5.86
4S.57 9.28 2.09 6.0 I
47.31 9.54 2.17 6.15
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 19871.~Continued
I PCE. tctrachlorocthylenc: ftg/L microgrnm per liter: 1.2-tDCE. 11w1s-l .2-dichlorocthylene: TCE. trichlorocthy!cnc: VC. \'inyl chloride: \\'TP. water trealmcnl plant]
•.. -.,··-' ,l < .,,.,.,,,"'-<-,_Js';.,~._',;.;-· •·•··
121 Jan 1961 67.69 49.15 9.82 2.25 6.30
122 Fch 1961 69.54 51.03 10.10 2.33 6.46
123 Mar 1961 71.56 52.73 l(J.35 2.41 6.61
124 Apr 1961 73.49 54.69 10.64 2.49 6.77
125 May 1961 75.49 56.57 10.92 2.58 6.92
126 June 1961 77.39 58.53 11.20 2.66 7.07
127 July 1961 79.36 60.43 11.46 2.75 7.22
128 Aug 1961 81.32 62.42 11.74 2.83 7.36
129 Sept !961 83.19 64.40 12.111 2.92 7.51
130 Oct 1961 85.11 66.32 12.27 3.00 7.64
131 Nov 1961 86.95 68.33 12.55 3.09 7.79
132 Dec 1961 88.84 70.28 12.80 3.17 7.92
133 Jan 1962 60.88 47.74 8.63 2.15 5.32
134 Feb 1962 62.10 49.86 9.00 2.25 5.56
135 Mar 1962 62.94 51.28 9.17 2.31 5.64
136 Apr 1962 63.59 52.37 9.25 2.36 5.67
137 May 1962 64.17 53. 18 9.28 2.39 5.66
138 June 1962 64.70 53.88 9.28 2.41 5.63
139 July 1962 65.23 54.48 9.28 2.43 5.60
140 Aug 1962 65.74 55.06 9.26 2.45 5.56
141 Sept 1962 66.22 55.59 9.24 2.46 5.52
142 Oct 1962 66.71 56.07 9.22 2.48 5.47
143 Nov 1962 67. 18 56.54 9.19 2.49 5.42
144 Dec 1962 67.65 56.97 9.16 2.50 5.38
145 Jan 1963 68.06 57.40 9.13 2.51 5.33
146 Fch 1963 68.39 57.78 9.09 2.52 5.28
147 Mar 1963 68.73 58.11 9.06 2.53 5.24
148 Apr !963 69.03 58.49 9.02 2.54 5.20
149 May 1963 69.33 58.81 8.98 2.55 5.15
150 June 1963 69.62 59.14 8.94 2.56 5.11
151 July 1963 69.90 59.42 8.911 2.57 5.116
152 Aug 1963 711.17 59.70 8.86 2.57 5.02
153 Sept 1963 711.43 59.97 8.82 2.57 4.98
154 Oi.:t 1963 70.69 60.21 8.78 2.58 4.94
155 Nov 1963 70.93 60.45 8.74 2.58 4.90
156 Dec 1963 71.17 60.67 8.70 2.59 4.86
Chapter A: Summary of Findings A85
Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant---------------
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 19871.-Continued
[PCE, tctrad1lorocthylcnc: 11g/l .. microgram per liter; ! .2-tDCE. 11w1s-I .2-dich!orocthylenc: rCE. 1richlorocthylcnc: VC. vinyl chloride: WTP. water tremmem plant]
157 Jan 1964
158 Fch 1964 63.77
159 Mar 1964 63.95
160 Apr 1964 64.08
161 May 1964 64.19
162 June 1964 64.27
163 July 1964 64.34
164 Aug 1964 64.39
165 Sept 1964 64.43
166 Oct 1964 64.47
167 Nov 1964 64.49
168 Dec 1964 64.50
169 Jan 1965 64.511
170 Fch 1965 64.49
171 Mar 1965 64.47
172 Apr 1965 64.45
173 i\fay 1965 64.42
174 June 1965 64.38
175 July 1965 64.33
176 Aug 1965 64.27
177 Sept 19(15 64.211
178 Oc1 1965 64.13
179 Nov 1965 64.05
1811 Dec 1965 63.97
181 Jan 1966 63.88
182 Fch 1966 63.79
183 Mar 1966 63.68
184 Apr 1966 63.57
185 May 1966 63.46
186 June 1966 63.34
187 July 1966 63.21
188 Aug 1966 63.08
189 Sept 1966 62.94
190 Oct 1966 62.80
191 Nov 1966 62.65
192 Dec 1966 62.50
A86
,:.,?;,.;, ,,,,,. . .__,,, . ·-~• .
60.89 8.67 2.59 4.83
54.39 7.69 2.3 I 4.27
54.42 7.58 2.30 4.17
54.43 7.50 2.29 4.10
54.36 7.42 2.29 4.04
54.29 7.35 2.28 3.98
54.21 7.28 2.27 3.93
54.14 7.22 2.26 3.88
54.06 7.16 2.26 3.84
53.99 7.10 2.25 3.79
53.92 7.05 2.24 3.75
53.85 7.00 2.24 3.72
53.78 6.95 2.23 3.68
53.72 6.90 2.23 3.65
53.64 6.86 2.22 3.61
53.59 6.82 2.22 3.58
53.52 6.78 2.21 3.55
53.47 6.74 2.21 3.52
53.411 6.70 2.20 3.511
53.34 6.66 2.20 3.47
53.27 6.63 2.19 3.44
53.211 6.59 2.19 3.42
53.14 6.56 2.18 3.40
53.07 6.53 2.18 3.37
53.00 6.50 2.17 3.35
52.93 6.47 2.17 3.33
52.84 6.44 2.16 3.31
52.78 6.41 2. 16 3.29
52.70 6.38 2.15 3.27
52.63 6.35 2.15 3.25
52.54 6.33 2.14 3.23
52.46 6.30 2.14 3.21
52.38 6.27 2.13 3.20
52.28 6.25 2.13 3.18
52.20 6.22 2.12 3.16
52.11 6.19 2.12 3.14
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 1987'.-Continued
[!'CE. tctrachlorocthykne: ftg/L. microgram per liter: 1.2-tDCE. mms-1.2-dichloroethylene: TCE. trichloroethylene: VC. vinyl chloride: \\'Tl'. water tn:atment plant]
193 Jan 1967 62.25 52.02 6.17 2.11 3.13
194 Fch 1967 61.99 51.90 6.14 2.11 3.11
195 Mar 1967 61.67 51.76 6.11 2.10 3.09
196 Apr 1967 61.35 51.61 6.118 2.09 3.07
197 May 1967 61.02 51.43 6.114 2.08 3.05
198 June 1967 60.69 51.23 6.110 2.07 3.113
199 July 1967 60.37 51.02 5.96 2.06 3.011
2011 Aug 1967 611.115 50.79 5.92 2.05 2.98
2111 Sept !967 59.74 50.57 5.87 2.04 2.95
202 Oct 1967 59.43 50.34 5.83 2.03 2.92
203 Nov 1967 59.13 50.11 5.79 2.02 2.911
2114 Dec 1967 58.83 49.89 5.75 2.01 2.87
205 Jan 1968 58.41 49.66 5.70 2.00 2.85
206 Feb 1968 57.95 49.40 5.66 1.99 2.82
207 Mar 1968 57.43 49.IO 5.60 1.97 2.79
208 Apr 1968 56.94 48.77 5.55 1.96 2.76
209 May 1968 56.45 48.43 5.49 1.94 2.73
2111 June 1968 55.98 48.07 5.43 1.93 2.69
211 July 1968 55.49 47.67 5.36 1.91 2.65
212 Aug 1968 55.02 47.26 5.29 1.89 2.61
213 Sept I 968 54.58 46.84 5.23 1.87 2.57
214 Oct 1968 54.13 46.43 5. 16 1.85 2.54
215 Nov 1968 53.71 46.03 5.10 1.84 2.50
216 Dec I 968 53.28 45.63 5.114 1.82 2.46
217 Jan 1969 53.07 45.24 4.98 1.80 2.43
218 Feh 1969 52.97 44.91 4.93 1.79 2.411
219 Mar 1969 52.94 44.64 4.88 1.78 2.37
2211 Apr 1969 52.93 44.47 4.86 1.77 2.35
221 May 1969 52.93 44.32 4.83 1.76 2.34
222 June 1969 52.92 44.20 4.81 1.76 2.32
223 July 1969 52.90 44.09 4.79 1.75 2.31
224 Aug 1969 52.86 44.01 4.78 1.75 2.30
225 Sept 1969 52.81 43.92 4.77 1.75 2.29
226 Oct 1969 52.75 43.83 4.76 1.74 2.29
227 Nov 1969 55.19 45.75 4.97 1.82 2.38
228 Dec 1969 55.19 45.96 5.111 1.83 2.42
Chapter A: Summary of Findings A87
Appendix A2. Simulated PCE in Finished Water. Tarawa Terrace Water Treatment Plant---------------
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 19871.-Continued
I PCE. tc\rachlorocthylcnc: 11g/L. microgram per liter; 1.2-tDCE. 1m11s-l .2-dichloroethylcne: TCE. trichlorocthylcnc: VC. vinyl chloride: \VTP, water treatment plant]
229 .Ian 1970 55.01
230 Feh 1970 54.79
231 Mar 1970 54.49
232 Apr 1970 54.20
233 May 1970 53.90
234 June 1970 53.61
235 July 1970 53.32
236 Aug 1970 53.04
237 Sept 1970 52.78
238 Oct 1970 52.53
239 Nov 1970 52.29
240 Dec 1970 52.05
241 Jan 1971 51.96
242 Feh 1971 51.93
243 Mar 1971 51.95
244 Apr 1971 51.99
245 Mayl971 52.03
246 June 1971 52.118
247 July 1971 52.12
248 Aug 1971 52. 16
249 Sept 1971 52.20
2511 Oct 1971 52.23
251 Nov 1971 52.26
252 Dec 1971 52.29
253 .Ian I 972 49.34
254 Feb 1972 49.01
255 Mar 1972 48.68
256 Apr I 972 48.40
257 May 1972 48.14
258 June 1972 47.90
259 July 1972 47.67
260 Aug 1972 47.45
261 Sept 1972 47.25
262 Oct 1972 47.05
263 Nov 1972 46.87
264 Dec 1972 46.69
ABB
5.03 1.84
46.03 5.03 1.84 2.43
45.94 5.Q3 1.83 2.43
45.84 5.03 1.83 2.44
45.70 5.01 1.82 2.44
45.54 5.00 1.82 2.43
45.37 4.98 1.81 2.43
45.20 4.96 I.SO 2.42
45.00 4.94 1.79 2.41
44.79 4.91 1.78 2.40
44.58 4.89 1.78 2.39
44.37 4.87 1.77 2.38
44.17 4.84 1.76 2.37
43.99 4.82 1.75 2.35
43.86 4.80 1.74 2.34
43.76 4.79 1.74 2.34
43.66 4.78 1.74 2.:n
43.60 4.78 1.73 2.33
43.53 4.77 1.73 2.:n
43.47 4.77 1.73 2.33
43.41 4.77 1.73 2.33
43.35 4.77 1.72 2.33
43.31 4.77 1.72 2.33
43.26 4.77 1.72 2.34
41.02 4.53 1.63 2.22
40.49 4.44 1.6 I 2.17
40.0 I 4.37 1.58 2.13
39.51 4.311 1.56 2.119
39.03 4.24 1.54 2.06
38.55 4.17 1.52 2.112
38.11 4.11 I.SO 1.98
37.68 4.05 1.48 1.95
37.26 3.99 1.46 1.92
36.88 3.94 1.45 1.89
36.51 3.89 1.43 1.86
36.15 3.85 1.42 1.84
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 19877.-Continued
[ PCE. tctrachlorocthylcne: ftg/L. microgram per liter: 1.2-tDCE, lnms-1.2-Jichlorocthylene: TCE, trichlorocthylenc: VC vinyl chloride: WTP, water treatment plant I
'W = li'il!Iilllllili!Dl'JE7 . -~liil!filljflEm-~wliI!t!IJl
., • • , • . ,!J 'lltilllil ,. 'lltill!Dd . roooJm.llild Wllil ~llild
265 Jan l 97J 54.28 41.48 4.411 1.62 2.111
266 Fch 1973 54.19 42.32 4.57 1.67 2.21
267 Mar 1973 53.98 42.49 4.611 1.68 2.23
268 Apr 1973 53.76 42.42 4.60 1.68 2.24
269 May 1973 53.52 42.25 4.59 1.67 2.24
270 June l973 53.311 42.115 4.58 1.66 2.25
271 July ! 973 53.118 41.78 4.56 1.65 2.24
272 Aug 1973 52.87 41.53 4.53 1.64 2.23
273 Sept I 973 52.68 41.27 4.51 1.63 2.22
274 Oct 1973 52.51 41.0 I 4.48 1.62 2.21
275 Nov 1973 52.35 40.75 4.45 1.6 I 2.20
276 Dec 1973 52.20 411.48 4.42 1.60 2.19
277 Jan 1974 52.43 40.22 4.40 1.59 2.17
278 Feb 1974 52.82 40.13 4.39 1.59 2.17
279 Mar 1974 53.39 40.10 4.38 1.58 2.16
280 Apr 1974 53.99 40.20 4.40 1.59 2.17
281 May 1974 54.63 40.35 4.43 1.60 2.18
282 June 1974 55.25 40.59 4.48 1.6 I 2.21
283 July 1974 55.90 40.82 4.52 1.62 2.24
284 Aug 1974 56.53 41.08 4.57 1.63 2.27
285 Sept 1974 57. IO 41.35 4.62 1.64 2.31
286 Oct 1974 57.70 41.61 4.68 1.65 2.34
287 Nov 1974 58.30 41.91 4.74 1.67 2.39
288 Dec 1974 58.92 42.19 4.81 1.68 2.43
289 Jan 1975 61.00 43.76 5.02 1.74 2.55
290 Feh 1975 61.24 43.90 5.06 1.75 2.59
291 Mar 1975 61 .41 44.03 5.11 1.75 2.63
292 Apr 1975 61.57 44.18 5. 16 1.76 2.68
293 May 1975 61.72 44.29 5.21) 1.77 2.71
294 June 1975 61.88 44.38 5.24 1.77 2.75
295 July 1975 62.115 44.45 5.28 1.77 2.78
296 Aug 1975 62.25 44.52 5.31 1.78 2.81
297 Sept 1975 62.46 44.57 5.34 1.78 2.83
298 Oc1 1975 62.69 44.62 5.36 1.78 2.85
299 Nov 1975 62.92 44.69 5J9 1.78 2.87
300 Dec 1975 63.18 44.74 5.41 1.78 2.89
Chapter A: Summary of Findings A89
Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant---------------
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 1987'.-Continued
I PCE, lclrachloroethylene; !tg/L, microgram per liter; I ,2-1DCE, 1m11.1·-l .2-dichloroethylenc; TCE, trichlorocthylene: VC, vinyl chloride; WTI'. water trcatrncnt plant I
3111
3112
303
304
305
3116
307
308
309
310
311
312
313
314
315
316
317
318
319
3211
321
322
J23
324
325
326
327
328
329
3311
331
332
333
334
335
336
A90
Jan 1976
Feb 1976
Mar 1976
Apr 1976
May 1976
June 1976
July 1976
Aug 1976
Sept 1976
Oct 1976
Nov 1976
Dec 1976
Jan 1977
Feb 1977
Mar 1977
Apr 1977
May 1977
June 1977
July 1977
Aug 1977
Sept 1977
Oct 1977
Nov 1977
Dec I 977
Jan 1978
Feb 1978
Mar 1978
Apr 1978
May 1978
June 1978
July 1978
Aug 1978
Sept I 978
Oct 1978
Nov l 978
Dec 1978
73.96
74.94
75.97
76.97
78.00
79.02
80.07
81.13
82.17
83.25
84.31
85.41
86.61
87.711
88.91
90.10
91.32
92.53
93.75
94.99
96.20
97.42
98.62
99.84
101.18
102.77
103.04
104.31
1115.18
106.88
107.95
108.69
109.61
11 I. I 8
111.118
111.93
51.53 6.24 2.06 3.34
53.43 6.62
54.44 6.80
55.38 6.99
56.21 7.16
57.07 7.34
57.86 7.51
58.73 7.69
59.58 7.86
60.41 8.02
61.28 8.19
62.10 8.35
62.97 8.52
63.98 8.71
64.81 8.86
65.83 9.05
66.76 9.21
67.76 9.38
68.711 9.55
69.711 9.72
70.70 9.88
71.65 10.04
72.71 10.21
73.68 10.36
74.73 10.53
76.25 10.80
78.73 11.26
77.97 I 1.02
79.28 11.27
79.72 11.29
82.3 I 11.78
83.81 12.00
84.16 12.00
84.92 12.09
87.48 12.55
85.67 12.04
2.15
2.20
2.24
2.28
2.32
2.35
2.39
2.43
2.46
2.50
2.53
2.57
2.62
2.65
2.70
2.74
2.78
2.82
2.86
2.90
2.94
2.99
3.113
3.07
3.14
3.26
3.21
3.27
3.28
3.41
3.47
3.48
3.51
3.63
3.52
3.60
3.72
3.85
3.98
4.10
4.22
4.34
4.46
4.57
4.68
4.79
4.89
5.01
5.1 I
5.22
5.32
5.43
5.53
5.63
5.72
5.82
5.92
6.011
6.10
6.26
6.56
6.37
6.53
6.51
6.83
6.96
6.93
6.97
7.25
6.87
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 19871.-Continued
! PCE. tetrndiloructhylene: ftg/L. microgram per liter: 1.2-tDCE, 11w1s-J .2-dichloroethylene; TCE. trichlorocthylcnc: YC. vinyl chloride: WTP. water treatment plant]
337 Jan 1979 113.14 11.95 3.50 6.79
338 Fch 1979 114.115 86.75 12.16 3.56 6.91
:rw Mar 1979 114.98 87.55 12.23 3.60 6.93
340 Apr 1979 115.82 88.43 12.32 3.63 6.97
341 May 1979 116.68 89.21 12.411 3.66 7.00
342 June 1979 117.47 90.09 12.49 3.70 7.05
343 July 1979 I 18.29 90.82 12.56 3.73 7.07
344 Aug 1979 I 19.118 91.67 12.65 3.76 7.11
345 Sept 1979 119.82 92.44 12.72 3.79 7.14
346 Oct 1979 120.59 93.22 12.81 3.82 7.18
347 Nov 1979 121.31 94.110 12.88 3.85 7.21
348 Dec 1979 122.114 94.78 12.96 3.89 7.24
349 Jan I 980 123.28 9S.56 13.03 3.92 7.27
350 Feb 1980 122.98 98.20 13.49 4.04 7.56
351 Mar 1980 124.03 96.35 12.98 3.94 7.19
352 Apr 1980 123.90 97.86 13.28 4.01 7.39
353 May 1980 124.69 96.00 12.78 3.90 7.03
3S4 June 1980 125.83 96.23 12.80 3.91 7.03
355 July 1980 0.72 0.00 0.110 0.00 0.00
356 Aug 1980 0.75 0.00 0.00 0.00 0.00
357 Sept I 980 121.36 95.07 12.43 3.92 6.83
358 Oct 19811 121.72 91.40 11.24 3.63 5.84
359 Nov 1980 122.14 91.00 I 1.17 3.63 5.82
360 Dec 1980 122.95 90.64 11.14 3.62 5.81
361 Jan 1981 114.115 84.14 I0.41 3.37 5.46
362 Fch 1981 114.39 84.80 111.53 3.41 5.55
363 i\far 198 I 115.60 84.13 111.37 3.37 5.44
364 Apr 1981 116.S5 85.90 10.74 3.46 5.69
365 May 1981 117.311 87.53 11.02 3.54 5.87
366 June 1981 118.36 88.90 I 1.26 3.611 6.03
367 July 1981 133.29 102.10 13.12 4.17 7.09
368 Aug 1981 134.31 105.46 13.75 4.33 7.511
369 Sept 1981 1211.72 96.34 12.64 3.96 6.93
3711 Oct 1981 121.114 96.29 12.60 3.95 6.90
371 Nov 1981 121.41 96.69 12.67 J.96 6.93
372 Dec 1981 121.81 97.27 12.74 J.98 6.97
Chapter A: Summary of Findings A91
Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant---------------
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 1987'.-Continued
[ PCE. tctrnchlomcthylcnc; )tg/L. microgram per liter: 1.2-tDCE. 1m11.1·-l .2-dichloroethylene: TCE. 1richloroethylcnc: VC. vinyl chloride: \\'TP. water treatment plant]
373 Jan 1982 103.95
374 Fch I 982 105.86
375 Mar 1982 107.52
376 Apr 1982 108.83
377 May 1982 148.50
378 June 1982 110.78
379 July 1982 111.98
380 Aug 1982 113.07
381 Sept 1982 114.04
382 Oct 1982 114.60
383 Nov 1982 113.87
384 Dec 1982 115.16
385 Jan 1983 1.25
386 Fch 1983 1.29
387 Mar 1983 111.76
388 Apr 1983 112.66
389 ~fay 1983 113.97
390 June 1983 106.10
391 July !983 116.70
392 Aug 1983 117.72
:wJ Sept 1983 117.83
394 Oct 1983 117.97
JlJ5 Nov 1983 118.63
39(1 Dec 1983 120.78
397 Jan 1984 132.87
398 Fch 1984 180.39
399 i\far 1984 183.02
400 Apr 1984 151.46
401 May 1984 153.42
402 June 1984 182.13
403 July 1984 156.39
404 Aug 1984 170.47
405 Sept 1984 181.22
406 Oct 1984 173.73
407 Nov 1984 173.77
408 Dec 1984 173.18
A92
81.28 I 0.65 3.33
83.47 I 1.06 3.43
85.42 11.40 3.51
87.32 11.75 3.60
120.45 16.30 4.98
92.65 12.81 3.86
92.98 12.77 3.86
94.09 12.97 3.91
95.33 13.18 3.96
96.51 13.37 4.01
96.63 13.31 4.00
93.14 12.43 3.80
0. 10 0.()4 11.00
0.12 0.05 0.0 I
88.43 11.55 3.65
86.39 10.85 3.43
87.67 11.04 3.52
82.26 10.54 3.33
92.03 11.95 3.75
94.46 12.45 U7
96.92 12.94 3.99
96.60 12.82 3.96
95.49 12.58 3.89
95.52 12.60 3.89
111.52 15.09 4.61
145.48 19.20 5.94
155.54 21.34 6.47
132.07 18.23 5.52
132.19 18.09 5.49
158.14 21.85 6.60
140.96 19.72 5.92
118.88 16.05 4.81
149.36 19.60 6.17
136.04 17.33 5.56
131.63 16.46 5.34
128.47 15.83 5.18
5.81
6.09
6.31
6.55
9.13
7.26
7.21
7.34
7.46
7.57
7.51
6.88
0.05
0.07
6.37
5.77
5.88
5.70
6.52
6.87
7.21
7.12
6.95
6.96
8.43
10.56
11.97
10.26
l0.13
12.28
11.14
8.94
11.20
9.39
8.87
8.46
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
--------------Appendix A2. Simulated PCE in Finished Water, Tarawa Terrace Water Treatment Plant
Appendix A2. Simulated tetrachloroethylene and its degradation by-products in finished water, Tarawa Terrace water treatment plant,
January 1951-March 1987'.-Continued
I PCE. lctrachlorocthylcnc: itg/L, microgram per liter: 1.2-tDCE. 1m11s-l ,2-dichlorocthylcnc: TCE. trichloroethylene: VC. vinyl chloride: \VTP. water tn:atmcn1 plant I
409 Jan 1985 176.12 127.80 15.48 5.13 8.20
410 Feb 1985 3.64 1.10 0.29 0.05 0.22
411 Mar I 985 8.71 3.88 0.68 0.17 0.47
412 Apr 1985 8.09 3.70 0.68 0.16 0.49
413 May 1985 4.76 1.65 0.44 O.ll7 0.35
414 June 1985 5.14 1.88 0.50 (1.08 0.41
415 July 1985 5.54 2.10 0.56 (l.09 0.47
416 Aug 1985 6.01 2.34 0.63 (I. 10 0.52
417 Sept I 985 6.50 2.62 0.71 0.12 0.59
418 Oct 1985 7.06 2.91 0.79 0.13 0.65
419 Nov 1985 7.(,4 3.24 0.87 0. I 5 0. 71
420 Dec !985 8.27 3.58 0.95 0.16 0.76
421 Jan 1986 8.85 3.95 1.04 0.18 0.82
422 Feb 1986 9.42 4.24 1.08 0.19 0.83
423 Mar 1986 12.14 5.40 1.34 0.24 1.01
424 Apr 1986 10.83 4.93 1.20 0.22 0.89
425 May 1986 11.56 5.25 1.25 0.23 0.91
426 June 1986 12.28 5.61 1.30 0.25 0.92
427 July 1986 13.06 5.97 1.35 0.26 0.94
428 Aug 1986 13.84 6.36 1.39 0.28 0.96
429 Sept 1986 14.61 6.75 1.44 0.30 0.97
430 Oct 1986 15.42 7.12 1.48 0.31 0.99
431 Nov 1986 16.21 7.52 1.52 0.33 1.00
432 Dec 1986 17.03 7.89 1.56 0.34 1.01
433 Jan 1987 17.85 8.28 1.59 0.36 I.OJ
434 Feb 1987 18.49 8.71 1.64 0.38 1.03
435 Mar 1987 WTP closed WTP dosed WTPcloscd WTP closed WTP closed
1Currcnt maximum contamin:mt kn:b (iv1CL~) arc: tctr;H;h!orocthylene (PCE) and trichloroethylenc (TCE). 5 µg/L: rrn11s-1.2-dichloroethylene ( l ,2-tDCE).
JOO ftg/L: and vinyl chloride (VC). 2 ftg/L (USEPA, 2003): effective dates for r-.ICLs are as follow~: TCE and VC. January 9, 1989: PCE and 1.2-tDCE,
July 6. 1992 (40 CFR. Section 141.Ml. Effective Dates. July I. 2002. ed.)
~MTJDMS: A three-dimensional ma~s transport. multispecies model developed by C. Zheng and P. Wang ( 1999) on behalf of the U.S. Army Engineer
Research and Development Ci.:nter in Vicksburg, Mississippi (http://hydro.gl'O.ua.ed11/mt3d/)
3'J'tx:hF1nwMP: A three-dimcnsi,mal multispccics. multiphase mass transport model developed by the \1ultirnedia Envirnnrncntal Simulati1ms Lab(irat,iry
(Jang and Aral 2007) at the Georgia Institute of Technology. Atlant:1. Georgia (hup:/lmes/.ce.gatechedu)
➔Results from Chapter F report (Faye In press 2007b)
1Results from Chapter G report (.Jang and Aral In press 2007)
Chapter A: Summary of Findings A93
A94 Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-------------------------------Appendix AJ. Questions and Answers
Appendix A3. Ouestions and Answers
Two of the three drinking-water systems that served family housing at
U.S. Marine Corps Base Camp Lejeune were contaminated. One system,
the Tarawa Terrace drinking-water system, was mostly contaminated
with tetrachloroethylcne (or perchloroethylene, PCE) from off-base
dry-cleaning operations. The other system. the Hadnot Point drinking-
water system, was contaminated mostly with trichloroethylene (TCE)
from on-base industrial operations. The contaminated wells were continu-
ously used until 1985 and sporadically used until early 1987. ATSDR's
health study will try to determine if there was a link between in utero and
infant (up to I year of age) exposures to drinking-water contaminants and
specific birth defects and childhood cancers. The study includes births
occuITing during 1968-1985 to mothers who lived in base family housing
during their pregnancy. The birth defects and childhood cancers that will
he studied are:
• neural tube defects (spina bi Iida and anencephaly).
cleft lip and cleft palate. and
• leukemia and non-Hodgkin's lymphoma.
Only a few studies have looked at the risk of birth defects and childhood
cancers among children born to women exposed during pregnancy to
volatile organic compounds (YOCs) such as TCE and PCE in drinking
water. This study is unique because it will estimate monthly levels of
drinking-water contaminants to determine exposures.
Chapter A provides a summary of detailed technical findings (found in
Chapters B-K) for Tarawa Terrace and vicinity. The findings focus on
modeling techniques used to reconstruct historical and present-day
conditions of groundwater flow. contaminant fate and transport, and
distribution of drinking water. Information from the water-modeling
analyses will be given to researchers conducting the health study.
(Future analyses and reports will present information and data about
the Hadnot Point drinking-water system.)
Chapter A: Summary of Findings
What is the purpose of
the ATSDR health study?
Why is ATSDR studying
exposure to voe-
contaminated drinking
water since other studies
have already done this?
What is in the ATSDR
reports about the
Tarawa Terrace
drinking-water system'?
A95
Appendix A3. Questions and Answers-------------------------------
Why is ATSDR using
water modeling to
estimate exposure
rather than real data?
What is a water model?
What information did
ATSDR use to develop
the water models and
what were the sources
of the information'!
A96
Data on the levels of VOC contaminants in drinking water are not available
before 1982. ·n, determine levels before 1982, ATSDR is using a process
called "historical reconstruction." This process uses data on the amount of
the chemicals dumped on the ground. It also uses the properties of the soil.
the groundwater. and the water-distribution system. These data are then
used in computer models. The models estimate when contaminants lirst
reached drinking-water wells. The models also estimate monthly levels
of contaminants in drinking water at family housing units. This information
is important for the health study. It can also be used by those who lived in
base family housing to estimate their exposures.
A water model is a general term that describes a computer program used
to solve a set of mathematical equations that describe the:
• flow of groundwater in aquifers.
movement of a contaminant mixed with groundwater,
mixing of water from contaminated and uncontaminated
water-supply wells at a water treatment plant. or
flow of water and contaminants from reservoirs. wells. and
storage tanks through a network of pipelines.
The historical n::construction process required information and data describ-
ing physical characteristics of the groundwater-flow system. conservation
principles that describe the flow system, the specific data on the contami-
nant (PCE) and its degradation by-products, and the water-distribution
system. The following specific data needs were required:
• aquifer characteristics: gcohydrologic, hydraulic. water production.
fate, transformation, and transport;
chemical properties characteristics: physical. fate. transformation.
and transport; and
water-distribution system characteristics: pipeline characteristics,
storage-tank geometry. pumps, water-production data, and water-
quality parameters.
Information and data used to conduct the historical reconstruction analysis
were obtained from a variety of sources. These sources included ATS DR.
U.S. Environmental Protection Agency, Environmental Management Division
of U.S. Marine Corps Base Camp Lejeune, U.S. Geological Survey, private
consulting organizations, published scientific literature. and community
groups representing former marines and their families. Chapters A and K
of the Tarawa Terrace report provide searchable electronic databases-on
DVD format-of information and data sources used to conduct the histori-
cal reconstruction analysis.
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
-------------------------------Appendix AJ. Questions and Answers
A water model requires information on the specific properties or '·parameters ..
of the soil. groundwater. and water system at the base. Often assumptions
arc needed because complete and accurate data are not available for all the
parameters that must be modeled. In particular. historical data are often lack-
ing. To be sure that water-modeling results arc accurate and represent historical
"real-world'' conditions, a model needs to be calibrated. A calibration process
compares model results with available "real-world" data to see if the model's
results accurately reflect "real-world .. conditions. This is done in the follow-
ing way. Models arc constructed using different combinations of values for the
parameters. Each model makes a prediction about the groundwater-flow rate.
the amount of water produced by each well. and the contamination level in the
drinking-waler system at a particular point in time. These predictions are then
compared lo "real-world" data. When the combination of parameter values that
best predicts the actual "real-world'" conditions are selected, the model is --cali-
brated.'' The model is now ready 10 make predictions about historical conditions.
At lirst. ATSDR developed a model that simulated the fate and transport
(migration) of PCE that was completely mixed in groundwater in the satu-
rated zone (zone below the water table). The model code used is known as
MT3DMS. ATS DR developed a second model because of suggestions from a
panel of experts and requests from former marines and their technical advisers.
The second model is capable of simulating the fate and transport of PCE and
its degradation by-products of TCE. trans-1,2-dichlorocthylcne ( 1,2-tDCE),
and vinyl chloride (VC) in the unsaturated zone (area ahovc the water table)
and the saturated zone. This model, known as TechFlowM P. is based on
significantly more complex mathematical equations and formulations. This
highly complex model also can simulate PCE and its degradation by-products
in both the vapor and water phases. Values of simulated PCE concentrations
in the saturated zone obtained using the two different models (MT3DMS and
TechFlowMP) arc very close.
ATSDR did in-depth reviews of historical data, including water-supply well
and WTP operational data when available. ATSDR concluded that the Tarawa
Terrace waler-distribution system-including the WTP-was not intercon-
nected with other water-distribution systems at Camp Lejeune for any time
longer than 2 weeks. All water arriving at the WTP was obtained solely from
Tarawa Terrace water-supply wells. Also it was assumed to be completely and
uniformly mixed prior to delivery to residents of Tarawa Terrace. On a monthly
basis. the concentration of PCE delivered to specific family housing units at
Tarawa Terrace was assumed to be the same as the simulated concentration of
PCE in finished water at the WTP.
No. The available data are not specific enough to accurately estimate daily
levels of PCE in the Tarawa Terrace water system. The modeling approach
used by ATSDR provides a high level of detail and accuracy to estimate
monthly PCE exposure concentrations in finished water at the Tarawa
Terrace WTP. II is assumed that simulated monthly concentrations of PCE
represent a typical day during a month.
Chapter A: Summary of Findings
How can ATSDR be sure
that water-modeling results
represent historical "real-
world" conditions?
Why did ATSDR develop
and calibrate two models
for simulating the
migration of PCE from
ABC One-Hour Cleaners
to Tarawa Terrace
water-supply wells?
Why isATSDR
providing simulated PCE
concentrations in finished
water at the Tarawa
Terrace water treatment
plant (WTP) rather than
at locations of specific
family housing units?
Can ATSDR water
modeling results be
used to determine the
concentration of PCE
that my family and I were
exposed to on a daily basis?
A97
Appendix AJ. Questions and Answers-------------------------------
Were my family and I more
exposed to contaminated
drinking water than other
families hecause we lived near
one of the contaminated Tarawa
Terrace water-supply wells?
Were my family and I exposed
to other contaminants hesides
PCE in finished drinking
water while living in family
housing at Tarawa Terrace?
How can I get a list of the
monthly PCE (and PCE
degradation hy-product)
concentrations in finished water
that my family and I were
exposed to at Tarawa Terrace?
ATSDR's historical
reconstruction analysis
documents that Tarawa
Terrace drinking water was
contaminated with PCE that
exceeded the current maximum
contaminant level (MCL) of
5 micrograms per liter (µg/L)
during 1957 and reached a
maximum value of 183 µg/L.
What does this mean in terms
of my family's health'?
A98
No. Water fro111 all Tarawa Terrace water-supply wells (unconta111inatcd
and conta111inatcd) was mixed at the WTP prior to being distributed
through a network of pipelines to storage tanks and family housing areas.
On a monthly basis. the concentration of PCE delivered to specific fa111ily
housing units at Tarawa Terrace has been shown to he the same as the
concentration or PCE in finished water at the WTP.
Yes. A small amount of PCE degrades in the groundwater to other VOCs.
These include TCE, 1.2-tDCE, and VC. Degradation by-products or PCE
were round in water samples obtained on January 16, 1985, from Tarawa
Terrace water-supply wells TT-23 and TT-26. Historical reconstruction
analyses conducted by ATSDR and its partners provide simulated monthly
concentrations of PCE and its degradation by-products in finished water
at the Tarawa Terrace WTP.
ATS DR and its partners have developed a Web site where former Camp
Lejeune residents can enter the dates they lived on base and receive i11for-
111atio11 on whether they were exposed to VOCs and to what levels. The Web
site will list the simulated 111onthly concentrations or PCE and its degrada-
tion by-products in finished water at the Tarawa Terrace WTP. The Web site
can he accessed at h11p:/lwww.atsdr.cdc.gov/sites/lejeu11e/i11dex.l11111/.
ATSDR's exposure assessment cannot he used to determine whether you,
or your family, suffered any health effects as a result or past exposure
to PCE-conta111inated drinking water al Ca111p Lejeune. The study will
help determine if there is an association between certain birth defects and
childhood cancers among children whose mothers used this water during
pregnancy. Epidemiological studies such as this help improve scientific
knowledge of the health effects of these chemicals.
The National Toxicology Program of the U.S. Department of Health and
Human Services has stated that PCE "is reasonably anticipated to be a human
carcinogen." However, the lowest level of PCE in drinking water at which
health effects begin to occur is unknown. The MCL for PCE was set at 5 µg/L
(or 5 parts per billion) in 1992 because, given the technology at that time,
5 11g/L was the lowest level that water systems could be required to achieve.
Many factors determine whether people will suffer adverse hcallh effects
because of chemical exposures. These factors include:
dose (how much),
duration (how long the contact period is).
when in the course of life the exposures occurred (for example,
while in utero. during early childhood. or in later years of life).
• genetic traits that might make a person more vulnerable to the
chemical exposure, and
• other factors such as occupational exposures, exposures to other
chemicals in the environment. gender, diet, lifestyle, and overall
slate or health.
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune. North Carolina
-------------------------------Appendix A3. lluestions and Answers
Soil vapor or soil gas is the air found in the open or pore spaces between
soil particles in the soil above the water table (also called the '·unsaturated
zone"). The source of the soil vapor is the contaminated groundwater.
PCE and its degradation by-products are VOCs; therefore. some amounts
of these chemicals volatilize (or vaporize) off the groundwater plume and
enter the soil in the unsaturated zone as gases. The soil vapor plu111e (also
known as the "vapor-phase" plume) is the area where the gases or vapors
have entered the soil in the unsaturated zone above the water table.
Soil at Camp Lejeune is sandy. so the vapors can readily vaporize up
to the surface. The buildings are on concrete slabs, so soil vapor can
enter these buildings through cracks or perforations in slabs or through
openings for pipes or wiring. In addition, because the vapor enters
the building due to pressure differences. the operation of heating or
air-conditioning systems can create a negative pressure in the building
that draws the vapors from the soil into the building. This is similar to
the situation with radon gas.
The results or the PCE and PCE degradation by-product soil vapor mod-
eling will not have a major impact on the current epidemiological study
of specific birth defects (neural tube defects, cleft lip, and cleft palate)
and childhood cancers (leuke111ia and non-Hodgkin's lymphoma-also
known as childhood hematopoietic cancers). The focus of the study is on
drinking-water exposures to the fetus up to the child's first year of life.
The drinking-water exposure is considerably greater than any exposure
that 111ight occur due to soil vapor infiltration into a ho111e. However,
the analysis 111ay incorporate the soil vapor results to determine if these
exposures significantly change the results obtained from the analysis of
drinking-water exposures.
Historical data on the levels or conta111inants in the drinking water is very
li111ited. That is why there is uncertainty and variability concerning when
the MCL of 5 ftg/L was reached at the Tarawa Terrace WTP. Therefore.
ATSDR and its partners conducted exhaustive sets of simulations to
quantify this uncertainty and variability. Based on these analyses. finished
water contaminated with PCE exceeding the MCL of 5 ftg/L could have
been delivered from the Tarawa Ten-ace WTP as early as December 1956
but most likely during November 1957.
Chapter A: Summary of Findings
What is soil vapor'?
Could the soil vapor
enter buildings at
Tarawa Terrace?
Could historical exposure
to soil vapors contami-
nated with PCE and PCE
degradation by-products
affect the current ATSDR
epidemiological study?
How certain is ATSDR
that finished water
exceeding the current
MCL for PCE of 5 µg/L
was delivered from
the Tarawa Terrace
WTP beginning in
November 1957'?
A99
Appendix A3. Questions and Answers-------------------------------
How does ATSDR know
where all of the Tarawa
Terrace water-supply
wells were located if they
have been destroyed?
What is the accuracy
of this information'!
What did ATSDR do
to he sure that water-
modeling analyses are
scientifically credible'!
Where and how can
I get a copy of this
ATSDR report and the
information and data
that were used in the
Tarawa Terrace water-
modeling analyses'!
A100
t;:.Q· U.S. GPO: 2007 -660-117 / 62001 Region No. 8
ATS DR relied on a variety of sources to obtain infor111ation on the location
of Tarawa Terrace water-supply wells. These included historical water utility
maps, well construction and location maps, aerial photographs, use of geo-
graphic information system technology. and assistance from Environmental
Manage111ent Division staff at U.S. Marine Corps Base Ca111p Lejeune. The
accuracy of this information is believed to be within± 50 feet of the actual
well location.
Throughout this investigation, ATSDR has sought external expert input
and review. Activities included convening an expert peer review panel and
submitting individual chapter reports to outside national and international
experts for technical reviews. For exa111ple, on March 28-29. 2005, ATS DR
convened an external expert panel to review the approach used in conduct-
ing the historical reconstruction analysis. The panel also provided input and
recommendations on preliminary analyses and modeling. ATSDR used a
nu111ber of recom111endations 111ade by the panel me111bers. ATS DR also used
technical comments from outside expert reviewers when finalizing reports
on l~irawa Terrace water-modeling analyses.
A small number of printed copies of this report and subsequent chapter
reports (A-K) will be available to interested parties and placed in public
repositories. Electronic versions of all chapter reports will be available on
the ATSDR Camp Lejeune Web site at h111,:llwww.,11sdr.cdc.gov!s-ites/
/eje1111e/i11dex.ht111/. Chapters A and K provide a searchable electronic data-
base-on DVD format-of information and data sources used to conduct
the historical reconstruction analysis for Tarawa Terrace and vicinity.
Historical Reconstruction of Drinking-Water Contamination at Tarawa Terrace
and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
1
34°45'
34°42'30"
0 0 0 0 ;ll
34°40'
I
34°37'30"
2477000
77°2.5'
' rl "'~
,, ~c..:,,,.--
J~~~
~~~\
/0-:Jj;
ts; J'/i3~
!
' ; q,~
_}
~r ~11
Q'l
r.r ~ ' ,J/i~
2486000
MO-.i-gan.
il(I J
Jown
Pein.I
Plate 1
Location of Wells, Boreholes, and Water-Supply Areas, U.S. Marine Corps Base Camp Lejeune, North Carolina
Analyses ot Groundwater Flow, Contaminant Fate and Transport and Distribution of Drinking Water at Tarawa Terrace and Vicinity,
U.S. Marine Corp Base Camp Lejeune, North Carolina· Historical Reconstruction and Present-Day Conditions; 2007
Historical water-supply
areas of Camp Lejeune
Military Reservation
hlf&il Montford Point
D Tarawa Terrace
I'.:,] Holcomb Boulevard
lill Hadnot Point
D Other areas of Camp Lejeune
Military Reservation
-25-Topographic contour-
Interval 10 feet
2513000 77°17'30'
ONSLOW COUNTY
EXPLANATION
Water distribution
-Water pipeline-2004
A 52323 Elevated storage tank and number
.._STT-39 Ground level storage tank and number
D Water treatment plant (closed 1987)
Well and identification-(□), first
well used with that assigned number;
(n), second well used with that
assigned number
9 HP-663 Water supply o X24S7 Monitoring
o T-6 Borehole
2522000
; 2.5
o 2.s! s
IOMMS
10 KILOMETEl'IS
Base frOl" tamp l.ejeuM GIS Office, June 2003
Groundwater-flow and
fate and transport
model boundaries
Domain ..... __ Active area
Boundary conditions for
groundwater-flow model
General head
Drain
·'"" No flow
• Specified head
■ ABC One-Hour Cleaners
,:;;, ~
2 KILOMETERS
Base from U.S. Marine Corps and U.S. Geological Survey digital data files
Location of Wells and Boreholes, Groundwater-Flow Model Boundary, and Present-Day (2004) Water-Distribution Systems Serving Tarawa Terrace,
Holcomb Boulevard, and Hadnot Point and Vicinity, U.S. Marine Corps Base Camp Lejeune, North Carolina
By
Morris L. Maslia, Jason B. Sautner, Robert E. Faye, Rene J. Suarez-Soto, Mustafa M. Aral, Walter M. Grayman, Wonyong Jang, Jinjun Wang, Frank J. Bove, Perri Z. Ruckart, Claudia Valenzuela, Joseph W. Green, Jr., and Amy L. Krueger
2007