Groundwater Monitoring
       Network Optimization
Frontier Hard Chrome Superfund Site,
      Vancouver, Washington

-------
               Solid Waste and       EPA 542-R-07-021
               Emergency Response    December 2007
               (5203P)            www.epa.gov
       Groundwater Monitoring
        Network Optimization
Frontier Hard Chrome Superfund Site,
       Vancouver, Washington

-------
            GROUNDWATER MONITORING NETWORK OPTIMIZATION
                 FRONTIER HARD CHROME SUPERFUND SITE

EXECUTIVE SUM MARY

The following report reviews and provides recommendations for instituting  a long-term
groundwater monitoring network for Frontier Hard Chrome Superfund Site in Vancouver,
Washington (FHC Site). The FHC Site consists of a former chrome  plating facility in the
floodplain  of the Colombia River.  Shallow groundwater in the FHC area has been
impacted by residual hexavalent chromium from  chrome-plating operations conducted
between 1958 and 1983. Affected groundwater migrated downgradient from the source
under the influence of industrial groundwater pumping south of the FHC site.

Extensive  site remediation activities were completed at the FHC Site in 2003. The area
around FHC  is currently undergoing rapid  urban redevelopment to residential and
commercial property use. The primary goal  of developing an  optimized groundwater
monitoring strategy  at  the FHC Site is  to create a  dataset that fully supports site
management decisions while minimizing  time and  expense associated with collecting
and interpreting data. The long-term groundwater monitoring network for the FHC Site
should be designed to  support site management decisions while  accommodating on-
going redevelopment.

In the following  report, the current FHC groundwater monitoring network has been
evaluated  using a formal qualitative approach as well as statistical tools found in the
Monitoring   and    Remediation    Optimization    System    software    (MAROS).
Recommendations are made for groundwater sampling frequency and location based on
current hydrogeologic conditions and long-term monitoring (LTM) goals for  the system.
The following report  evaluates the monitoring system using analytical and hydrogeologic
data collected  after installation of the remedy to the  present,  a time-frame between
October 2003 and June 2007. The following report outlines recommendations based  on
a formal evaluation,  but final determination of sampling  locations and frequencies are to
be decided by the overseeing regulatory agencies.

Current Site Conditions

The broad area of shallow groundwater contamination associated  with chrome plating
operations at FHC  was discovered in  the  1980's and investigated and delineated
through the 1990's.  The Record of Decision (ROD) (USEPA, 2001) for groundwater at
FHC produced  in 2001, detailed an in-situ  chemical  reduction of mobile  hexavalent
chromium  (Cr(VI)) as the final remedy. The regulatory screening  level for total chromium
for  the  Site was determined to be  50  ug/L, based on the State of  Washington
Department of Ecology Model Toxics Control Act (MTCA) Standard A value.

As a result of aggressive remedial treatments  and cessation of industrial pumping, total
chromium  concentrations across the site have dropped below the regulatory screening
level.  It should be noted that for the past 3 years, total  chromium levels in groundwater
at FHC have consistently been measured below the clean-up level of 50 ug/L. The FHC
groundwater plume, that is the extent of groundwater affected above the regulatory

Frontier Hard Chrome Site                      i                Groundwater Monitoring
Vancouver, Washington                                       Network  Optimization

-------
screening level, has largely disappeared.  However, for the  purpose of the following
analysis, the term "plume" is used to describe the historic extent of groundwater affected
by chromium originating from the FHC site. In this document, the term 'plume' describes
all  chromium  concentrations  at any  detectible level  within the  current  FHC Site
groundwater monitoring network. Analytical results for total  chromium were used in the
analysis  of the groundwater  network as  a conservative surrogate for  assessing the
concentration of soluble hexavalent chromium.

Site Groundwater Monitoring Goals and Objectives

Primary monitoring goals for the FHC Site groundwater include defining  the extent and
magnitude of residual contamination and evaluating  the efficacy of the chosen remedy.
The specific  groundwater  monitoring  objective for FHC  is  to "ensure dilution and
dispersion  of  affected  groundwater"  until site groundwater  meets  state cleanup
standards (USEPA,  2001). Shallow groundwater in the  FHC  area is protected  by
institutional controls prohibiting construction of water-supply wells in groundwater that
may be  affected  by industrial contaminants. Monitoring data will provide support  for
institutional controls by delineating the extent of affected groundwater.  Data from the
network will provide evidence of concentration stability and indicate if constituents begin
to remobilize.  Analytical data  collected from  the  network will document  continued
efficacy of the remedy and attenuation of chemical constituents confirming that the
remedy is achieving site clean-up goals.

Project Goals and Objectives

The goal of the  long-term monitoring optimization  (LTMO) process  is to  review the
current groundwater monitoring  program and  provide recommendations for improving
the efficiency  and accuracy of the network in  supporting site monitoring  objectives.
Specifically, the LTMO  process provides information on the site characterization, stability
of constituent  concentrations, sufficiency and  redundancy of  monitoring  locations and
the appropriate frequency of  network sampling. Tasks involved  in the LTMO process
include:

    •   Evaluate well   locations  and  screened intervals  within  the context  of the
       hydrogeologic regime to determine if the site is well characterized;
    •   Evaluate overall 'plume' stability through trend and moment analysis;
    •   Evaluate individual well concentration trends over time for target  constituents of
       concern (COCs);
    •   Develop sampling location recommendations based on an  analysis of  spatial
       uncertainty;
    •   Develop  sampling  frequency  recommendations  based  on  qualitative and
       quantitative statistical analysis results;
    •   Evaluate individual  well  analytical  data  for statistical sufficiency and identify
       locations that have achieved clean-up goals.

The end  product of the LTMO process at the FHC Site is a recommendation for specific
sampling  locations  and frequencies that  best address  site  monitoring  goals and
objectives while providing sufficient flexibility for site redevelopment.
Frontier Hard Chrome Site                       ii                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
Results
Statistical and qualitative evaluations of FHC Site analytical data have been conducted
and the following general conclusions have been drawn based  on the results of these
analyses:

   •   After  a  qualitative  evaluation  of well  locations,  screened  intervals  and
       hydrogeologic characteristics, affected groundwater at the FHC Site is delineated
       to the relevant regulatory standards established for the  site (Washington State
       Department  of  Ecology  MTCA A  Standards,  50  ug/L for  total chromium).
       Groundwater areas  where  concentrations  occasionally  exceed  regulatory
       standards are bounded by wells where results are  below  the standard. No major
       data gaps in site characterization were found.

   •   The historic area of affected  groundwater evaluated shows  overall stable  to
       decreasing concentration trends  for total  chromium. None  of  the well data
       reviewed show increasing concentration trends. Many "no trend'  findings result
       from intermittent  detections,  data outliers or apparently cyclical  variation  in
       concentrations, especially in Zone B wells.

   •   Moment trend analysis indicated that total dissolved mass measured within the
       monitoring network is decreasing  over time. The  center of  mass in Zone B is
       retreating toward the source.

   •   Results from the  spatial redundancy analysis indicate that several  wells could be
       removed from the program, as they do not provide  unique information. Wells
       identified as  redundant are listed in Table 5.

   •   No areas of high concentration  uncertainty  were  found;  therefore  no new
       monitoring locations are recommended.

   •   The sampling frequency analysis recommended a reduced sampling frequency
       for  the  majority  of  wells.  Annual  to  biennial  sampling frequencies  were
       recommended by the MAROS  algorithm based on the rate of change and trend
       of well concentrations.

   •   Many  locations  evaluated  were  statistically  below the  screening level  for
       chromium using both  the  student's  T-test with  a  power  analysis  and the
       sequential T-test. Approximately two-thirds of monitoring locations  have achieved
       the cleanup goals with  80%  or  greater statistical  power, given the current
       dataset.
Frontier Hard Chrome Site                      iii                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
Recommendations

The following general recommendations  are made based  on the findings summarized
above and those described in Section 4 below.

   •   Several areas of spatial redundancy were identified. 10 wells are recommended
       for exclusion from the monitoring program.

   •   No new monitoring locations are recommended.

   •   Reduce the frequency of monitoring to annual sampling.

   •   Monitoring data show fairly high variance. In most cases, variance in the data
       can  be explained by site characteristics and geochemical processes. Continue
       monitoring  concentration   trends  for  both  total  chromium and  hexavalent
       chromium  and  potentiometric  water  levels to determine  how  the  hydraulic
       influence of the Columbia River may be contributing to underlying variance in the
       data.

   •   The majority of the analysis above was completed before several wells in the
       network were  damaged  as a result of site redevelopment. Some wells may need
       to  be replaced or  rehabilitated  in  order  to  achieve stated site  monitoring
       objectives. The  recommendation  that no new monitoring  locations are  needed
       does  not  imply that  monitoring wells  damaged or destroyed during  site
       redevelopment do not need to be  replaced. New wells may be required, but their
       placement near 'old' locations identified as important is recommended.

   •   Continue development  and  updating  of the  comprehensive  site  database.
       Results for both total chromium and hexavalent chromium concentrations should
       be added  to the database. Validated analytical data for all wells in the area
       should be added to database within a reasonable time after sampling. Each well
       should have a complete record of historic sampling events.

   •   Survey location coordinates and elevations for all wells.  Share data  with all
       stakeholders.  A common  set  of coordinates  should  be  used by  planners,
       regulators, and construction and development companies.
Frontier Hard Chrome Site                      i                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
1.0 INTRODUCTION

   The Frontier Hard Chrome  Superfund Site (FCH  Site) is a  National Priorities Listed
   (NPL)   site  administered  under  the  Comprehensive  Environmental  Response,
   Compensation  and  Liability Act  (Superfund).  The  site  is  located  in Vancouver,
   Washington in Clark, County near the Columbia River (see Figure 1). The FHC site is
   currently administered by the Washington Department of Ecology (Ecology) with support
   from the  US Environmental  Protection Agency (EPA)  Region 10. The  original  FHC
   property is a 1/2-acre historic chrome-plating facility, built and operated between 1958
   and 1983. The Site has traditionally been organized into soil and groundwater operable
   units (OU). Only the groundwater OU will be considered in this report.

   Groundwater monitoring plays a critical role  in long-term restoration of the FHC Site. The
   purpose of the following LTMO  evaluation  is to  review  the current  groundwater
   monitoring network and  provide recommendations for improving  the efficiency and
   accuracy of the network for supporting site  management decisions during  and after site
   redevelopment.

   At the FHC Site, monitoring goals define why and how data collected from the site will be
   used. The primary groundwater monitoring goal for the site is  to "ensure dilution and
   dispersion  of affected  groundwater", with  monitoring to continue until "all  remaining
   groundwater meets state  standards  for groundwater  cleanup"  (USEPA,  2001).
   Monitoring data from the site  network are used  to support institutional controls, by
   identifying areas of affected groundwater and to document continued attenuation of site
   constituents.

   In order to recommend an optimized  network that addresses the stated monitoring
   objectives, spatial  and analytical data  from the site were analyzed using a  series of
   quantitative and qualitative tools. Tasks performed during LTMO analyses include:

       •   Evaluate well  locations and  screened  intervals  within  the context of  the
          hydrogeologic regime to  determine if the site is well characterized;
       •   Evaluate overall  'plume stability'  through  concentration  trend   and moment
          analysis;
       •   Evaluate individual well  concentration trends over time for target constituents of
          concern (total chromium);
       •   Develop  sampling location recommendations based  on an  analysis of spatial
          uncertainty;
       •   Develop  sampling frequency recommendations based  on  both qualitative and
          quantitative statistical analysis results;
       •   Evaluate individual well  analytical  data  for statistical  sufficiency and identify
          locations that have achieved clean-up goals.

   A discussion of site background and regulatory context for the FHC Site is provided
   below. Section 2 of the report details the analytical and statistical approach taken during
   the LTMO evaluation. A detailed discussion  of results is provided in Section 3. Summary
   conclusions and recommendations are presented in Section 4.

   Frontier Hard Chrome Site                     1                Groundwater Monitoring
   Vancouver, Washington                                        Network Optimization

-------
1.1 Site Background and Regulatory History

The  FHC Site is  located  in a  former  industrial area in the city  of Vancouver in
southwestern Washington near the Columbia River. The site is located within the greater
Portland, Oregon/Vancouver, Washington metropolitan area. Because of Vancouver's
location along the Columbia River and proximity to the Pacific Ocean, the region has
historically been the home to several shipyards and supporting industrial activity.

As the regional economy has changed in recent years, the Vancouver shipyards have
been redeveloped into residential and commercial  property  to support rapid increases in
population.  The area  to the south of  the FHC Site has  been redeveloped,  and the
industrial water supply  wells that contributed to the  spread of chromium-affected
groundwater  to the southwest  have been  removed from service.  The  FHC  Site is
scheduled for redevelopment into commercial properties in the near future.

The  FHC Site is located in a  floodplain, approximately  one-half  mile  north of the
Columbia River. One-quarter mile north of the site, a steep rise in elevation marks an
area of residential land use. In the mid-1950's, much of the floodplain, including the FHC
Site, was filled with hydraulic dredge material and construction rubble. East of the FHC
Site, a topographic depression exists at the original level of  the floodplain where the City
of Vancouver operates two groundwater well fields to provide public water supply. The
Pioneer Plating Company operated a chrome plating facility  on the one-half acre FHC
site from 1958 through 1970. Chrome plating operations continued under Frontier Hard
Chrome management until 1983.

During much of its operational  history, liquid wastes from chrome-plating operations
were  discharged directly to the  public sanitary sewer system. By  1975, the  City of
Vancouver determined that chromium in wastewater was impacting the operation of its
secondary waste  water treatment systems.  FHC was directed to find  an  alternate
disposal  method  for  liquid wastes. In  1976,  FHC received  a permit to discharge
untreated wastes to a drywell behind the facility. The permit included a schedule for the
installation  of a treatment  system for chromium-affected waste  water;  however, no
treatment systems were installed between 1976 and 1981.

By 1982, Ecology found FHC in violation of state waste disposal regulations. During the
same time period, chromium contamination was discovered in an industrial water supply
well  southwest of the  site,  near the  Columbia  River.  A broad  area of  shallow
groundwater  contamination associated  with  chrome plating operations at FHC  was
discovered. In December 1982,  the  FHC  Site was proposed for inclusion on the  NPL
under the CERCLA. In 1983, FHC closed all operations and the site was officially placed
on the NPL.  Under a  cooperative agreement with EPA, Ecology began  the Remedial
Investigation and Feasibility Study (RI/FS) process. Records of Decision (ROD's) for the
site have been published in 1987 (for the soil OU) and 1988 (for the  groundwater OU)
(USEPA, 1987 and 1988).

The 1987 ROD for soil called for excavation, stabilization  and replacement of affected
soils with concentrations over 550 mg/Kg total chromium.  Subsequently, the proposed
method of soil stabilization as a means of preventing leaching of chromium was found to
Frontier Hard Chrome Site                      2                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
be ineffective. The 1988 ROD for groundwater recommended extraction and treatment
of groundwater from areas where concentrations of total  chromium exceeded 50,000
ug/L. However,  groundwater monitoring indicated that the area of affected groundwater
was shrinking after the downgradient industrial supply wells were removed from service.
The combination of changing site conditions and the development of new cost-effective
technologies motivated the EPA to reevaluate the proposed remedies for FHC.

An amended ROD was completed in 2001 (USEPA, 2001) detailing the final remedial
action  planned  for the  site.  The selected  groundwater remedy included treatment  of
mobile hexavalent chromium (Cr(VI))  through in-situ reduction to relatively insoluble
trivalent chromium  (Cr(lll)).  An  In-situ Redox Manipulation (ISRM)  technology  was
chosen as the  groundwater OU remedy  (see Figure  1  for approximate  location  of
groundwater and  soil ISRM  treatment areas). An area  downgradient  from  the source
was injected with reducing agents, resulting in the reduction of naturally occurring iron in
the subsurface. The area  of reduced iron forms an in-situ permeable reactive barrier,
reducing soluble Cr(VI) in groundwater to Cr(lll). The purpose of the reactive  barrier was
to 1) provide containment and prevent downgradient transport of affected groundwater,
2) reduce  mass  of Cr(VI)  in high  concentration  areas; and 3)  provide long-term
protection against future leaching of Cr(VI) (USEPA, 2001).

An ISRM technology was also chosen for the soil OU. The area of the former chrome-
plating tank and main building of FHC was treated with reducing agents, applied directly
to the  soil. Aggressive treatment of the source area was anticipated to prevent further
Cr(VI) inputs to  site groundwater.

Remedial  activities for soil  and groundwater were completed in September 2003.
Regular monitoring of site groundwater was included in the ROD to "ensure dilution and
dispersion  of affected groundwater", with  monitoring to continue  until  "all  remaining
groundwater meets state  standards for groundwater  cleanup" (USEPA, 2001).  The
groundwater cleanup standard for the FHC site has been established at 50 ug/L.  Site
groundwater has been monitored quarterly between 2003 and 2007.

Analytical  data  for  total  dissolved  chromium have  been collected and used  in the
following report, as this chemical  analysis reflects concentrations of the more toxic and
soluble oxidation state of Cr(VI). Chromium solubility and mobility are strongly influenced
by  redox  reactions,  chemical   speciation,  adsorption/desorption  phenomena,  and
precipitation/dissolution  reactions. The reduced form of chromium (Cr(lll)) is significantly
less soluble in water than Cr(VI). Areas of the FHC site shallow subsurface have been
chemically  treated with reducing agents,  converting  Cr(VI)  to Cr(lll).  Groundwater
samples at certain monitoring well locations are under low reducing conditions due  to
the continued presence of reducing agents.

During the process of groundwater sampling some water samples may appear clear
(indicating Cr in the dissolved phase), and subsequently form a precipitate when
exposed to the atmosphere. When groundwater samples are removed from the
subsurface, Cr (III) compounds can precipitate as amorphous hydroxides. When sample
turbidity exceeds 10 Nephelometric Turbidity Units (NTUs), samples are filtered
removing the Cr(lll) species, but for samples with relatively low turbidity, the samples are
Frontier Hard Chrome Site                     3                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
not filtered even though they may contain suspended Cr(lll). The data that are derived
after adjusting for the interfering precipitation are below clean-up standards for the site.
However, the redox changes introduced during sampling may introduce a higher level of
variance in samples collected in the region of the ISRM remedy.

1.2 Geology and Hydrogeology

The FHC Site is  underlain by several geologic units, with the upper two being of interest
for this  report. The top unit  consists of hydraulic  fill  and construction debris  used to
elevate the adjacent floodplain in the 1940's and1950's. Fill materials are largely silt and
sand  and  heterogeneous,   poorly-compacted   construction   waste.   Fill   extends
approximately 12 to 20 feet below ground surface (ft bgs) across  the site. The fill unit is
generally unsaturated, but localized areas of perched  groundwater may be  present.
(USEPA, 2001)

Underlying the fill is an alluvial  unit, consisting of a clayey silt subunit and a sand-and-
ground unit. Groundwater in the alluvial unit is hydraulically connected to the Columbia
River. The clayey silt  is heterogeneous in character and is 3 to 7 feet thick, thinning to
the north of  the site. The  clayey silt  unit separates the lower sand-and-ground unit from
the fill. The  sand-and-ground unit consists  of poorly sorted sandy gravels, silty sandy
gravels and  sandy silts with scattered large cobbles.  Deposits in  this unit  resulted from
overbank deposition during flooding  of the Columbia River and from channel deposition
that resulted in  more particle sorting than  the overbank deposits. The alluvial unit is
approximately 70 feet  in thickness and is highly heterogeneous and anisotropic.

During initial site characterization, the alluvial unit was considered to have three layers.
Upper and lower permeable zones (Zones A and B) separated by an  aquitard were
described in the RI/FS (issued in 1987). Zone A was described  as a sand and gravel
layer  beginning about 20 ft bgs and extending to  about 35 ft bgs. A confining "lower
aquitard" below Zone A is described  in the 1988 ROD (USEPA,  1988) and was the basis
for separating groundwater in the alluvial unit into  A  and B zones. Currently, this  silt zone
is seen as semi-continuous fine-grained unit of dense sandy silt to silty sand. The layer
is now thought to  be semi-confining and not a  significant hydraulic barrier within the
alluvial aquifer.

Zone  B, or the deeper alluvial unit, is also made up of sands and  gravel, but with higher
permeability than Zone A. The  lower alluvial unit extends from approximately 35 ft bgs
down to 80  to 100 ft bgs. Groundwater velocity in this zone is  about 2.25 ft/d to the
south-southwest. There is no distinct vertical gradient between A  and B Zones.  Wells in
the FHC network are designated as  either A or B Zone wells based on the depth of the
screened  interval.  During the  LTMO analysis,  the zone designations were  used  to
separate the data into two analysis  groups to evaluate groundwater in zones based on
permeability. This is done with the  understanding  that Zones A  and B are most likely
hydraulically connected.

Groundwater flow in the region of the FHC site is generally to the south/southwest as the
potentiometric surface  data  indicate a  shallow  slope  to  the  south.   Historically,
groundwater flow direction has been influenced by pumping at downgradient industrial
Frontier Hard Chrome Site                      4                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
water supply wells, but when these wells were deactivated, groundwater flow returned to
a generally southerly flow direction. The average hydraulic gradient is 0.00015 ft/ft and
groundwater velocity is between 0.5 and 5 ft/d. Recharge to  site groundwater occurs
from local infiltration of precipitation and from  the recharge from another alluvial aquifer
north of the site near the topographic rise.  Downgradient from the Site, groundwater
discharges to  the Columbia River and area potentiometric surfaces are influenced by
Columbia River stage. Groundwater parameters used in the LTMO analysis are listed in
Table 2.
Frontier Hard Chrome Site                      5                 Groundwater Monitoring
Vancouver, Washington                                          Network Optimization

-------
2.0 ANALYTICAL APPROACH

Evaluation  of  the  groundwater monitoring  network in  the  vicinity of the FHC  Site
consisted  of  both  quantitative  and  qualitative  methods.  A  quantitative  statistical
evaluation of the site was conducted using tools in the MAROS software. The qualitative
evaluation reviewed hydrogeologic conditions, well construction and placement. Both
quantitative  statistical and  qualitative evaluations  were combined using a 'lines of
evidence' approach to recommend a final groundwater  monitoring strategy to support
site monitoring objectives.

2.1 MAROS Method

The  MAROS 2.2 software was used to evaluate the LTM network at the FHC Site.
MAROS is a collection of tools in one software package  that is used in an explanatory,
non-linear but  linked fashion to statistically evaluate groundwater monitoring programs.
The tool includes models, statistics, heuristic rules, and empirical relationships to assist
in  optimizing a groundwater monitoring  network  system. Results generated from the
software tool can be used to develop lines of evidence, which, in combination  with
professional judgment,  can be  used  to inform  regulatory  decisions  for  safe  and
economical  long-term monitoring of affected groundwater.  A summary  description of
each tool used in the analysis is provided in Appendix A of this report. For a detailed
description of the structure of the software and further utilities, refer to  the MAROS 2.2
Manual (AFCEE, 2003; http://www.gsi-net.com/software/MAROS  V2 2Manual.pdf) and
Azizetal., 2003.

In  MAROS 2.2, two levels of analysis are used for optimizing long-term monitoring plans:
1)  an overview statistical evaluation based on temporal trend analysis resulting in 'plume
stability' information; and 2) a more detailed statistical optimization based on spatial and
temporal redundancy reduction methods (see Appendix A or the MAROS  Users Manual
(AFCEE, 2003)).

2.1.1 COC Choice

MAROS includes a  short module that provides recommendations on prioritizing  COCs
for the entire  network based on toxicity,  prevalence, and mobility of  the compounds
dissolved in groundwater.  However,  the  priority  constituent at  the FHC site is total
dissolved chromium, analyzed as a surrogate for Cr(VI). Volatile organic compounds
(VOCs) are present in small amounts in site groundwater  from off-site sources, but these
compounds  are not  risk-drivers for site management. The COC choice  module was not
used for the  FHC site.

2.1.2 Plume Stability

Within MAROS,  time-series  concentration data are analyzed to  develop  a  conclusion
about 'plume stability'. For the MAROS analysis, a plume is defined as the extent of
groundwater within the  monitoring network affected by any concentration of the target
contaminant over time.  Practically, the 'plume' area is defined as the maximum extent of
affected groundwater over  the time-frame of the investigation. The definition of 'plume'
Frontier Hard Chrome Site                      6                Groundwater Monitoring
Vancouver, Washington                                       Network Optimization

-------
used in this document is distinct from the regulatory definition in that concentrations do
not need to exceed the regulatory screening limit in order to be considered part of the
'plume'.

For the purpose  of this  analysis,  a groundwater plume is  said  to  be stable when
constituent concentrations  at  individual  monitoring  locations  as well as  moments
estimated from  the  entire network are not changing  rapidly. If a plume is  found to be
stable, in many cases, the number of locations and  monitoring frequency  can be
reduced without loss of information.

Individual  well  concentrations are  evaluated  using  both Mann-Kendall  and  Linear
Regression trend tools. The Mann-Kendall  nonparametric evaluation is considered one
of the best methods to evaluate concentration trend as it does not assume the data fit a
particular  distribution   (Gilbert,  1987).  Individual  well  concentration trends  were
calculated for chromium for the time period 2003 to 2007.  Individual well Mann-Kendall
trends were also used in the sampling frequency analysis,  where trends determined for
the 2006 to 2007 interval were  compared with trends calculated using the entire dataset
for each well. During the final 'lines of evidence' evaluation, individual well concentration
trends are considered  along with summary statistics such as percent detection and
historic  maximum  concentration to  make  recommendations for  the final  sampling
network.

Moment  analysis  algorithms  in  MAROS  are  simple  approximations  of  complex
calculations and are meant to estimate the total dissolved mass (zeroth  moment), center
of mass  (first moment) and spread of mass  (second moment) within the monitoring
network and the trend for each  of these estimates over time. Trends for the first moment
indicate the relative amount of mass upgradient vs. downgradient and the change in the
distance of the center of mass from the source over time. Trends in the second moment
indicate relative dispersivity by  evaluating the spread  of mass  about the center of mass
over time.

2.1.3 Well Redundancy and Sufficiency

Spatial analysis modules in MAROS recommend elimination of sampling locations that
have little impact on the historical characterization of contaminant concentrations while
identifying areas within the monitoring network where additional data  are needed. For
details on the redundancy and sufficiency analyses,  see  Appendix A or the MAROS
Users Manual (AFCEE, 2003).

Sample locations are evaluated in MAROS for their importance in providing information
to define concentrations within the  area of affected groundwater. Wells identified as
providing information redundant with surrounding wells are recommended for elimination
from the program.  (Note:  'elimination' from the program  does not necessarily mean
plugging and abandoning the well. See Section 2.3 below.)

Well sufficiency is evaluated in MAROS using the same spatial analysis method as that
for  redundancy. Areas identified as having unacceptably  high  levels  of concentration
uncertainty are recommended for additional monitoring locations.
Frontier Hard Chrome Site                      ^                Groundwater  Monitoring
Vancouver, Washington                                        Network Optimization

-------
The  well redundancy and  sufficiency analyses use  the  Delaunay method and  are
designed to select the minimum number of sampling locations based on the relative
importance of information supplied at each sampling location in the monitoring network.
The importance of each sampling location is assessed by calculating a slope factor (SF)
and concentration and area ratios (CR and AR respectively). Sampling  locations with a
high  SF provide unique information and are retained in the network. Locations with low
SF are considered for removal. Areas  ringed  by wells with high SF's may be candidates
for new well locations. SF's were calculated for all wells at the  FHC Site and the results
were  used to determine the importance of each well  in the  network for defining
chromium concentrations.

The results from the Delaunay method and  the method  for determining new sampling
locations are derived solely from the spatial configuration of the monitoring network and
the spatial pattern of the contaminant plume based on a two-dimensional assumption.
No parameters such as the hydrogeologic conditions are considered  in the analysis.
Therefore, professional judgment and regulatory considerations must be used to confirm
final decisions.

2.1.4 Sampling Frequency

MAROS uses a Modified Cost Effective Sampling (MCES) method to optimize sampling
frequency for each  location based on the magnitude, direction, and uncertainty of its
concentration trends. The MCES method is based on the Cost Effective Sampling (CES)
method developed by Ridley et al. (1995). The MCES method estimates a conservative
lowest-frequency sampling schedule for a given groundwater monitoring  location that still
provides needed information for regulatory and remedial decision-making.

The MAROS algorithm recommended  a preliminary location sampling frequency (PLSF)
for each monitoring  location at the FHC Site based on a combination of  recent (2006-
2007) and long-term (2003-2007) trends and the magnitude and rate of  concentration
change. The PLSF  has been  reviewed  qualitatively  and a  final  optimal sampling
frequency has been recommended consistent with monitoring objectives and regulatory
requirements.

2.1.5 Data Sufficiency

The MAROS Data Sufficiency module employs simple statistical methods to evaluate
whether analytical  data are adequate both in quantity  and in quality  for revealing
changes  in constituent concentrations. Statistical tests for the MAROS  module were
taken from the  USEPA  Methods for Evaluating the Attainment of Cleanup  Standards
Volume 2: Groundwater statistical guidance document (USEPA, 1992).

Two types of statistical analyses have been performed on analytical samples from each
individual well. First, hypothesis testing using a Sequential T-test has been performed to
determine if groundwater concentration is statistically below the screening level for total
chromium (screening levels were set to Ecology MTCA Standard A of 50 ug/L). The
Sequential T-test indicates if the well has a sufficient number of samples at low enough
Frontier Hard Chrome Site                     8                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
concentrations  to  be  categorized  as having "attained" groundwater cleanup  goals
(confidently below the screening level  concentration). The statistical standard set by the
Sequential T-test  is  quite high,  and if  the well  data  indicate  the  groundwater
concentration has 'attained' cleanup, then there is  high confidence that the groundwater
is statistically below the regulatory limit. If measured concentrations are high or there are
an insufficient number of data points, then the well  is recommended for further sampling.

A Student's T-test followed by statistical power analysis was also performed in the Data
Sufficiency module to  assess the reliability  of the hypothesis test and to suggest the
number of additional  samples that may be required to reach statistical significance. The
power analysis uses the number of  samples (n),  the variance of  the  samples,  the
minimum detectible difference and the significance (a) of the test to determine if the well
is below the screening level with very high  confidence. The power analysis provides a
higher level of certainty that the well is not  affected above risk-based levels. Locations
that pass the  power test are considered "statistically clean".

The Data  Sufficiency module is designed to evaluate 6 years of sampling data.  While
quarterly sampling for  the past 3 years has provided a sufficient number of events to
evaluate the data using most techniques, 3  more years of sampling is necessary before
wells at the FHC Site can be confidently evaluated using this module. The analysis was
conducted with the current dataset  and results are  reported, but the results should  be
considered preliminary, at this point.

At the  FHC Site, locations that monitor groundwater areas "statistically below screening
levels" or "statistically  clean" may  be considered for reduced  sampling  frequency or
elimination from the  program. Statistically 'clean'  ring locations  can be retained  in the
program to  help bound  the areas of  affected  groundwater,  set institutional control
boundaries or function as surrogate point of compliance locations.

2.2 Data Input, Consolidation and  Site Assumptions

Groundwater analytical data from the  FHC  Site were supplied by Region 10 EPA and
from the Frontier Hard Chrome  Event  11 Long-Term Monitoring Report (Weston, 2007).
Site data  were supplemented  with  information from historic site reports  including the
RODs.  Groundwater monitoring locations included in the evaluation are listed in  Table
1, with additional aquifer and site details provided in Table 2.

Chemical  analytical  data collected  between October 2003  and June 2007 and well
information data were  organized in a database,  from  which summary statistics were
calculated. In all, 33 sample locations were  considered in the network evaluation for the
FHC Site.  Wells are described in Table 1, and well  locations are illustrated on Figure 1.
Groundwater monitoring data collected prior to 2003 are available for a subset of FHC
wells;  however, the installation of the remedy changed the nature and distribution of
dissolved  constituents  as well as groundwater geochemistry. Therefore, data collected
before 2003 are not comparable with those collected after installation of the remedy. In
order to provide reasonable consistency in statistical comparisons, analyses have been
limited to the 2003 - 2007 time-frame. Individual well trend evaluations were performed

Frontier Hard Chrome Site                      9                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
for data collected between 2003 and 2007. The data represent a roughly 3 year record
for many wells, and provide an indication of long-term, post-remediation trends in site
constituent concentrations.  Spatial analyses and  recent sampling frequency analyses
were conducted for data collected 2006-2007. Duplicate  samples in the dataset were
averaged to develop one analytical result for each quarter. No other data consolidation
was performed.

It should be noted that only total chromium concentrations in groundwater were used for
this evaluation. Analytical data for total chromium concentrations were collected as a
conservative surrogate for Cr(VI), which is the soluble form of the metal. Groundwater
samples at certain monitoring well locations were under low reducing conditions in the
subsurface due to the injection  of the reductant.  During the process of groundwater
sampling,  the clear water samples  would form  a precipitate  when  exposed  to the
atmosphere and filtering of the sample was necessary because turbidity was greater the
10 NTUs.  Use of total chromium results should  be considered conservative as the
method will  over-predict  soluble  chromium  concentrations  by  including  residual
suspended Cr(lll) in the result. Using total chromium analysis at all sites should improve
consistency in evaluating groundwater under a variety of subsurface redox conditions.

2.3 Qualitative Evaluation

Multiple factors should be considered in developing recommendations for monitoring at
sites undergoing long-term groundwater restoration. The LTMO process for the FHC Site
includes developing a 'lines of evidence' approach, combining statistical analyses with
qualitative  review to recommend an improved monitoring  network.  Results from the
statistical analyses  in combination with a qualitative review were used to determine
continuation or cessation of monitoring at  each  well location along with  a proposed
frequency of monitoring for those locations retained in the network.

The  primary consideration  in developing any monitoring  network  is to  ensure that
information collected efficiently supports site management  decisions. Site  information
needs  are  reflected in the  monitoring objectives for the network. For this  reason, any
proposed changes to the network are reviewed to be consistent with and supportive of
the stated monitoring  objectives. The qualitative review process begins with evaluating
each monitoring location for the role it  plays supporting site monitoring objectives. For
example,  a  location   may provide  vertical  or  horizontal  delineation  of  affected
groundwater or may provide information on decay  rates in the source area.  Each well in
the FHC Site  network was evaluated for its contribution  to site  monitoring objectives.
Qualitatively, redundant  locations are  those  where  multiple wells address the same
monitoring objective in approximately the same location.

A recommendation to eliminate chemical analytical monitoring at a particular location
based  on the  data reviewed does  not  necessarily constitute a recommendation  to
physically abandon (plug) the well. A  change in site conditions might warrant resumption
of monitoring at some time  in the future at wells that are not currently recommended for
continued  sampling.  In  some cases,  stakeholders  may  pursue  a comprehensive
monitoring  event  for all historic wells every five to ten years to provide a broad view of
plume  changes over time.  In general, continuation of water level measurements in all
Frontier Hard Chrome Site                      10               Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
site wells is recommended. Data on hydraulic gradients and potentiometric surfaces are
often  relatively inexpensive to collect and can  be used to support model development
and resource planning.  However, when site redevelopment is an issue, optimization of
the network can be used to identify redundant locations that can be plugged without loss
of information.
Qualitative evaluation for sampling frequency recommendations includes consideration
of factors such as the rate of change of concentrations, the groundwater flow velocity,
and the type and frequency of decisions that must be made about the site.  Additionally,
consideration  is  given  to  the concentration at a  particular location  relative  to the
regulatory screening level, the length of the monitoring history and the location relative to
potential receptors.

A  summary of  the lines  of  evidence used to develop  a final  monitoring network
recommendation is presented below.
Key Point: Several lines of evidence were used to develop recommendations for the monitoring
network.
Lines of Evidence
• Individual well trend
• Plume-Wide Trends
• Well Redundancy and Sufficiency
• Sampling Frequency
• Data Sufficiency
• Qualitative Evaluation
Method
• Mann-Kendall (Linear regression)
• Moment Analysis: Total dissolved mass,
center of mass and distribution of mass
trends.
• Delaunay triangulation and slope factor
calculation, along with area ratios and
concentration ratios.
• Modified Cost Effective Sampling
• Sequential T-Test, Student's T-Test and
Power Analysis
• Hydrogeologic factors, monitoring
objectives, stakeholder concerns and all
statistical results to develop final
recommendation.
Frontier Hard Chrome Site
Vancouver, Washington
                                        11
Groundwater Monitoring
Network Optimization

-------
3.0 SITE RESULTS

Data from 33 monitoring wells at depths corresponding to Zones A and B were included
in  the quantitative network analysis for the FHC Site.  Summary statistics for the wells
(including percent detections and maximum concentrations)  are shown  on Table  3.
Qualitative considerations are discussed alongside statistical interpretations below.

3.1 Plume Stability

3.1.1 Concentration Trends

Individual well  chromium concentration trends using the Mann-Kendall method are
summarized in the table below. Trends were evaluated for data collected between 2003
and 2007. Detailed results of the trend evaluations  performed are  summarized on Table
3.  Results of the individual well Mann-Kendall trends are also illustrated on Figure 2 and
Figures 7 and 8. Detailed Mann-Kendall reports for each well in the network are located
in Appendix B.
Alluvial
Aquifer Zone
Zone A
ZoneB
All Wells
Total
Wells
16
17
33
Number and Percentage of Wells for Each Trend Category
Non Detect
0
0
0
PD, D
5(31%)
7(41%)
12 (36%)
S
7 (44%)
2(12%)
9 (27%)
I, PI
0
0
0
No Trend
4 (25%)
8 (47%)
12(36%)
Note: Number and percentage of total wells in each category shown. Decreasing trend (D), Probably Decreasing trend
(PD), Stable (S), Probably Increasing trend (PI), and Increasing trend (I).

All wells had sufficient analytical data to evaluate trends. Because chromium is present
naturally at low levels in the aquifer, all wells groundwater analyzed showed detectable
quantities.   None of the sampling locations showed increasing or probably increasing
trends for total  chromium. The site  cleanup standard for chromium is 50 ug/L.  Overall,
two-thirds of well datasets showed stable to decreasing concentration trends.

Several wells with historic high concentrations near the ISRM zone indicate no trend or
high variance  in the data. Overall  most of the  measured  concentrations at these
locations are quite low,  but occasional spikes in concentration are seen (see MW-12A
and MW-15A and B in Appendix B). Data variability may be a result of artifacts due to
changing redox environment during sampling and subsequent filtering of samples (see
discussion above). Many of these wells monitor former "hot spots"  or areas with possibly
high residual chromium concentrations. Greater variability in the analytical data was
found in data from Zone B wells, as  indicated by the relatively large number of No Trend
(NT) results.

3.1.1.2 Moments

Moment analysis was used to estimate the total dissolved mass (Zeroth Moment), center
of mass (First Moment)  and distribution of mass (Second Moment) for total chromium in
Frontier Hard Chrome Site
Vancouver, Washington
                                        12
Groundwater Monitoring
Network Optimization

-------
the Zones A and  B. The values were determined using the current well configuration.
The Mann-Kendall trends of the moments were determined for data between 2004 and
2007  (4th quarter  2003 data did not include all sampling locations). Estimates of the
zeroth and first moments for the FHC Site are shown in Table 4. Moment trends are
summarized in the table below,  and first moments over time are illustrated on Figure 2.

Total  mass values are rough estimates  of mass in the dissolved phase,  assuming a
constant porosity  and uniform saturated thickness across the site. The mass estimates
are best interpreted as metrics for determining the trend of dissolved mass within the
network. For  both A and B zone groundwater,  mass estimates decreased strongly
between 2004 and  2007. Mass estimates  are greater for Zone  B as the saturated
thickness is greater.

First moments, indicating the  trend in  center of mass, show No Trend  in Zone A.
Concentrations measured  in  Zone  A  wells are  low,  and minor  fluctuations  in
concentration  are seen  in concentration vs.  time graphs  (Appendix B).  The alluvial
aquifer is influenced by stages of the Columbia River, and  the fluctuations in both first
moments and concentrations in Zone A may result from hydraulic influence of the river.
First moments for Zone  B indicate that  the  center of mass is regressing toward the
source, indicating  decreasing concentrations in the tail area relative to the source.
Moment
Type
Zeroth
First
Second
Moment Analysis
Zone A
Decreasing'. Total mass of chromium
showed a strongly decreasing trend
2004-2007
No Trend. The movement of center of
mass in Zone A shows no trend over
time.
No Trend in both X and Y directions
Comment
Zone B
Decreasing'. Total mass of chromium
showed a strongly decreasing trend 2004-
2007.
Decreasing'. The center of mass in Zone B
is moving closer to the source, supporting
the conclusion of a shrinking plume.
Probably Increasing in the X direction
(direction of groundwater flow).
Increasing in the Y direction (orthogonal to
groundwater flow).
Second moments indicate the pattern of dilution and dispersion of mass as it moves
from the center of the network to the edges.  No clear trend in second  moments was
found for Zone A. Zone B second moments indicate relatively more mass is moving to
the edges relative to the center. Increasing second moments support the conclusion that
total  chromium in Zone B  is dispersing in both the X (direction of groundwater flow) and
Y (orthogonal to groundwater flow) directions.

3.2 Well Redundancy and Sufficiency

The spatial redundancy analysis was performed using data collected between 2006 and
2007.  Spatial redundancy results  include slope  factor (SF)  and  area  (AR)  and
concentration ratio (CR) calculations to rank the importance  of the well in the network.
Summary results  for  the redundancy  analysis as  well as a  summary of the data
Frontier Hard Chrome Site
Vancouver, Washington
                                       13
Groundwater Monitoring
Network Optimization

-------
sufficiency  results  (identifying  wells where groundwater is statistically  below the
screening level) are presented on Table 5.

Of the 16 wells screened in the Zone A depth range,  eight were identified as possibly
redundant based on SF and AR and CR results. Interior well locations with SF below 0.3
were considered for removal, while hull wells were retained if SF was above 0.1. The
average SF for each well is shown on Table 5. Wells identified by the MAROS algorithm
as redundant include B85-4, MW-14A, MW-15A, MW-16A, W85-6A, W85-7A, W-92-16A
and W98-20A.

Seventeen  Zone B locations were evaluated and six were identified as redundant based
on the criteria described above. Wells B87-8, MW-11B,  MW-12B and C, MW-13C and
W85-6B were identified as not providing unique information to characterize the affected
or potentially  affected groundwater.

The decision to  remove a  well from routine monitoring  is based on a  combination  of
spatial statistical analyses and qualitative review of the function of the well in supporting
site monitoring objectives. The spatial statistics for Zone A and B wells were considered
along with  other lines of  evidence including whether the well  monitors groundwater
below the screening level,  trend results,  detection frequencies  and overall  monitoring
objectives before a final recommendation was made.

In the case of FHC, location of wells should be compatible with site redevelopment while
still meeting the objectives of the program. Proposed plans for site redevelopment were
received  from stakeholders, and well locations were reviewed to try to  accommodate
proposed development (see Figures  7 and 8).  In the  case  of  nested  well locations
(locations where multiple wells monitor several depths),  if  one well was very important
for monitoring one depth  profile the  other well  is recommended  for  retention in the
program  as well. Final recommendations for wells to retain in the monitoring program are
summarized below and shown on Table 5.

Wells
Retained
Wells
Excluded
Final Network Recommendation
Zone A
685-4, MW-16A, W85-6A, W92-16A
MW-12A, MW-17A, W97-18A,
MW-15A, W97-19A, W98-21A,
W99-5A,
MW-11A, MW-13A, MW-14A, W85-7A,
W98-20A
Zone B
685-3, 687-8, MW-12B, MW-12C, MW-15B
MW-16B, W85-6B, W92-16B, W97-18B,
W97-19B, W98-21B, W99-R5B
MW-11B, MW-13B, MW-13C, MW-14B,
W85-7B
The graphical well sufficiency analyses for Zones A and B are illustrated in Figures 3
and 4, respectively.  MAROS uses the Delaunay triangulation and SF calculations  to
identify areas with high concentration uncertainties. Graphical results illustrate polygons
created by the triangulation method and indicate areas of high uncertainty with an "L"  or
an "E" in the center of the triangle.  For both Zones A and  B,  no areas  of  high
Frontier Hard Chrome Site
Vancouver, Washington
                                       14
Groundwater Monitoring
Network Optimization

-------
concentration uncertainty were found; all areas show an "S" (for small uncertainty) or an
"M" (medium). Overall,  the networks have very low spatial uncertainty. Some areas  of
moderate spatial uncertainty were identified near the source "hot spots", but these areas
do not require additional well locations. No new well locations are recommended for the
monitoring networks.

Site data excluding the wells  recommended for elimination were re-run in the MAROS
data sufficiency module to  determine if eliminating wells from the  program would
increase concentration  uncertainty.  Figures  5  and  6  illustrate  the concentration
uncertainty found after elimination of redundant locations in Zones A and B.  No increase
in statistical uncertainty was found when wells listed above were eliminated, supporting
the redundancy of locations recommended for exclusion from the program.

3.3 Sampling Frequency

Table 6 summarizes  the  results of the  MAROS  preliminary  sampling  frequency
recommendation. The MCES method evaluates overall and recent temporal trends, and
recommends  an optimized sampling  frequency  based on the  rate  of concentration
change. As with the redundancy analysis, a qualitative review of all data is conducted
before recommending a final sampling frequency.

The rate of change of chromium concentrations for FHC wells is very low. The majority
of well locations have decreasing to stable concentration trends for the period analyzed.
For the recent  data,  many  wells  show  stable trends,  indicating that  the rate  of
concentration reduction at most locations has slowed. Many wells show some fluctuation
in the data that may be consistent with hydraulic  influence from the Columbia River  or
redox conditions during sampling.

The majority  of wells  in both Zones A and  B  have preliminary recommendations for
annual to biennial (every two  years) sampling.  The current  sampling frequency  is
quarterly. Quarterly  monitoring has  already provided a statistically significant dataset
(sufficient number of sample  points  to perform statistical analyses). After a qualitative
review,  annual  sampling frequency is recommended for all  wells remaining in the
network during  long-term groundwater monitoring. Annual  sampling is consistent with
the very low rate of change seen over the past 3  years, and relatively low groundwater
flow velocities and limited number of site management decisions to be made.

One well, MW-12A, had a PLSF recommendation for quarterly sampling, based on the
'no trend' concentration trend result and the presence of one  outlying sample result.
With the exception of one possible data outlier, the well shows a  fairly low overall rate of
change, so the MW-12A is also recommended for annual sampling.

The table below summarizes the current monitoring frequency  and the recommended
sampling frequency after the lines of evidence evaluation.
Frontier Hard Chrome Site                      15               Groundwater Monitoring
Vancouver, Washington                                       Network Optimization

-------
Monitoring Wells

Total Samples (average
per year)
Total Wells
Well Sampling frequency Analysis Source OU
Sampling
Frequency
Quarterly
Semi-annual
Annual
Biennial


Current Sampling
Frequency
33
0
0
0
99
33
Sampling Frequency
Recommendation
0
0
23
0
23
23
3.4 Data Sufficiency

The Data Sufficiency module was used to identify wells monitoring groundwater that has
statistically achieved site cleanup goals with >80% statistical power and those that have
attained  cleanup using the Sequential T-test method (even more stringent). Statistical
power increases with the number of samples taken and with reduction in  both the
concentration and detection limits for the dataset. For the FHC data set, the data were
assumed to be log-normally  distributed  and the statistics were performed using this
assumption. The groundwater cleanup goal for the FHC Site is 50 ug/L and the majority
of detection limits are 0.5 ug/L for most samples.

The Data Sufficiency tools are normally  run on  datasets with  greater than 6 years of
data,  but quarterly data for  the past 3 years  provides  enough data to perform a
preliminary analysis. Preliminary results for all  sampling locations are reported in Table
5. Achievement of "clean" status was considered along with other lines of evidence when
considering elimination of wells from the program and for reduction in sample frequency.
Data sufficiency should be revisited when  3 more years of data have been collected.

Results of the data sufficiency indicate that the majority of wells in the network are  at or
approaching cleanup goals and  have a  sufficient number of sample events to provide
confidence in the statistical outcome (although the number of sample years since  source
remediation is insufficient).

The data support the conclusion that the ISRM  groundwater treatment in  combination
with removal of the industrial supply wells has reduced site-wide concentrations. The
groundwater  network indicates groundwater is approaching and  may have achieved
cleanup goals and that a reduction in  monitoring effort may be  appropriate at this time.
The table below summarizes the results of the Data Sufficiency analysis. Identification of
specific wells that have achieved cleanup  can be found on Table 5.
Frontier Hard Chrome Site
Vancouver, Washington
                                       16
Groundwater Monitoring
Network Optimization

-------
Groundwater
Zone
A
B
Total
Total Wells
16
17
33
Data Sufficiency Results
Wells Statistically
Below MTCA with >80%
Power
15(94%)
12(71%)
27 (82%)
Wells Statistically "Attained"
Clean-up Goals
4 (25%)
1 (5%)
5(15%)
   MTCA = Washington State Model Toxics Control Act Standard A = 50|jg/L
Frontier Hard Chrome Site
Vancouver, Washington
                                                    17
Groundwater Monitoring
Network Optimization

-------
4.0 CONCLUSIONS AND RECOMMENDATIONS

4.1 General Conclusions

The primary goal  of developing an optimized monitoring strategy at the FHC Site is to
create a dataset that fully supports site management decisions while  minimizing time
and expense associated with collecting and interpreting the data. A summary of the final
recommended monitoring network is presented in Table 7. The recommended network
reduces  monitoring effort  and  cost by  reducing both the  frequency  of groundwater
sampling and the number of locations sampled.

Tasks identified in the Section 1 were performed for each of the groundwater zones. A
summary of general results and recommendations resulting from each task is presented
below:

•    Evaluate  well  locations  and screened  intervals  within  the   context  of  the
     hydrogeologic regime to determine if the site is well characterized.

     Result: Part of the network optimization process is to identify possible  gaps in site
     characterization that may require additional sampling locations or site investigation.
     Based on well locations, screened intervals  and hydrogeologic characteristics,
     affected groundwater at  the  FHC  Site is delineated to the  specified screening
     levels  (MTCAs Standard A,  50ug/L).  Groundwater areas where concentrations
     historically exceed screening levels (hot spots) are bounded by wells where results
     are below MTCAs. Monitoring locations in the tail of the network have average
     concentrations below the screening levels for both Zone A and  Zone B. A "hot
     spot" was identified in Zone A near location MW-12A, while the "hot spot" in Zone
     B is shifted to the south  near wells MW-15B and B87-8. All wells in the network
     have a sufficiently large data set to perform statistical calculations. No major data
     gaps were identified during the qualitative evaluation.

     Recommendation:  LTMO is  appropriate for the site at this time. No additional
     fundamental site investigation is recommended at this time. In order to comply with
     stated  monitoring  objectives,  future groundwater monitoring should include historic
     "hot spot" wells as well as regulatory compliance points.

•    Evaluate overall plume stability through trend and moment analysis.

     Result: Total chromium concentrations evaluated are largely decreasing to stable,
     even  though some concentration  trends  (for both individual  wells  and plume
     moments) show no trend. Many 'no trend' findings result from occasional outliers in
     the dataset (see MW-14B) or from wells where the concentration fluctuates at very
     low to non-detect concentrations (see W97-18B, W97-19B).   Another source of
     data variance includes possible influence of Columbia River stages on the aquifer
     and conservative sampling artifacts resulting from monitoring total chromium from
     a highly  reduced geochemical regime. Overall,  total  dissolved mass estimates
     (zeroth moment) within the monitoring networks are strongly decreasing.  Center of

Frontier Hard Chrome Site                      1                 Groundwater Monitoring
Vancouver, Washington                                         Network Optimization

-------
     mass estimates show some variation, but are consistent with shrinking extent of
     affected groundwater. The distribution of mass within the Zone B network indicates
     that dilution and dispersion of dissolved chromium is ongoing.

     Recommendation: Reduced monitoring effort is appropriate for stable or shrinking
     plumes. Monitoring frequency can be reduced where groundwater concentrations
     are not changing rapidly. After an initial steep drop in concentrations (2003-2004),
     groundwater  concentrations  are  not changing  rapidly at the  FHC  Site  and
     concentrations within the network appear to have stabilized at a low level, largely
     below  the 50ug/L screening  level. This  finding is  consistent with  reduced
     monitoring effort.

•    Evaluate individual well concentration trends over time for target constituents of
     concern (total chromium);

     Result:  For 33  wells evaluated  at the FHC  Site,  approximately two-thirds of
     locations showed stable to decreasing concentration trends (63%). No increasing
     or probably increasing trends were calculated. No statistically significant trend was
     found at roughly one-third of locations.

     Recommendation:  Individual well trend  evaluations  at  the FHC Site provide
     support for the conclusion  that total chromium  concentrations are largely stable.
     Monitoring frequency  can be reduced for locations where concentrations are not
     changing  rapidly or  are decreasing below screening  levels.  Some variation in
     concentrations  is seen  at "hot  spot"  locations, where  occasional  spikes in
     concentration  have been  recorded. "Hot  spot" locations  should be  monitored
     periodically to develop a longer-term dataset (>6 years).

     In the future,  both dissolved and total chromium analytical data should be collected
     at  the  appropriate  locations.   The  dissolved  chromium  concentrations  in
     groundwater  should  be  used instead  of  or along side total chromium for the
     evaluations in order to reduce variance in  the data introduced through sampling
     artifacts  and variable  redox conditions. The appropriate  locations to  use the
     dissolved data in the evaluations include monitoring well  locations where the
     groundwater samples have greater then 10 NTUs turbidity readings.

     Wells in  the  tail  area of the  network (south  and west of W85-6A/B) are largely
     stable  to decreasing  with  very low concentrations;  these  locations should  be
     monitored in the future as delineation or compliance points to confirm the absence
     of affected groundwater in this area.

•    Develop  sampling location recommendations based on  an analysis of spatial
     uncertainty;

     Result: The  spatial redundancy  analysis  indicated that several  wells  could  be
     removed from  the routine monitoring program, as they do not  provide unique
     information.

Frontier Hard Chrome Site                      2                 Groundwater Monitoring
Vancouver, Washington                                         Network Optimization

-------
     The spatial analysis did not identify any areas of high spatial uncertainty.

     Recommendation:    10  wells  are  recommended  for  exclusion  from  routine
     monitoring. The wells include locations near the remedy and some downgradient
     locations.  A spatial  analysis was  conducted for the reduced  network and  no
     increase in spatial concentration uncertainty was found for data between 2006 and
     2007. The resulting network of 23 locations should provide adequate information to
     monitor "dilution and dispersion" of dissolved chromium until  all areas achieve
     cleanup goals with statistical confidence.

     No new monitoring  locations are recommended.

   Develop sampling frequency recommendations  based  on  both  qualitative and
   quantitative statistical analysis results;

   Result: The  sampling  frequency analysis  recommended a  dramatically  reduced
   sampling frequency for the majority of wells. Annual to biennial sampling frequencies
   were recommended by the algorithm based on the rate of change and trend of well
   concentrations.

   Recommendation:   Reduce the frequency  of monitoring.  An  annual  sampling
   frequency was recommended  for future  monitoring.  While quarterly  sampling has
   been  effective to characterize the success  of the remedy, long-term data over a
   period of years are required to  achieve the stated monitoring objectives. These long-
   term objectives are not achieved by frequent sampling events, but rather by sampling
   a  consistent set of  wells  at  a frequency  comparable to the  rate  of change of
   concentrations. The recommendation is to collect annual data for approximately  six
   more years,  and re-evaluate  the plume for statistical attainment of site cleanup
   objectives.

   Evaluate individual well analytical data for statistical sufficiency and identify locations
   that have achieved clean-up goals.

   Result: 82% of wells are statistically below cleanup standards with greater than 80%
   power. 15% of locations have  achieved cleanup using the  Sequential T-test, a very
   rigorous statistical test.

   Recommendation: Data sufficiency should be revisited when  3 more years of data
   have   been   collected.  Preliminary results   indicate  that  remedial   actions and
   management decisions at the FHC site have resulted in a reduction in groundwater
   concentrations  with  groundwater  concentrations achieving  or close  to  cleanup
   objectives. The  high  number of sampling  locations currently  achieving  cleanup
   objectives is consistent with a reduced monitoring effort. All locations recommended
   for removal  from routine monitoring have achieved the cleanup goal  based on the
   Student's T-test and power analysis.
Frontier Hard Chrome Site                      3                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
Additional Recommendations:

•  The majority of the analysis  above was completed before several  wells in the
   network were damaged as a result of site redevelopment. Some wells  may need to
   be  replaced or rehabilitated in order to  achieve stated site monitoring objectives.
   The general recommendations for the network are to: 1) monitor "hot spots" in Zones
   A and B, and 2) monitor sufficient delineation points down and cross-gradient to
   confirm contaminant containment.

   The recommendation that no new monitoring locations are needed does not imply
   that monitoring wells damaged or destroyed during site redevelopment do not need
   to be replaced. New wells may be required, but their placement near 'old' locations
   identified as important is recommended.

•  Monitoring data at the FHC Site show  some variance relative to  concentrations
   (resulting in no trend).  In most cases, variance in the data  can be explained by site
   characteristics and geochemical processes.  Continue monitoring for concentration
   trends and potentiometric water levels to determine how the hydraulic influence of
   the  Columbia  River  may be  contributing  to underlying  variance  in  the data.
   Additionally, area  redevelopment may cause  changes in  recharge patterns (new
   paved areas, installation of  permeable paving), which may  be  reflected in aquifer
   characteristics and concentration trends.

•  Collect analytical data on total chromium as well as  dissolved  (Cr(VI)) chromium.
   Monitor turbidity in groundwater samples to ensure that only dissolved  chromium is
   being measured in the sample. Flag samples that have been filtered.

•  Continue  development and updating of the comprehensive site  database including
   both total and dissolved chromium analytical  results. Validated analytical data for all
   wells in  the  area should be  added to  database within a  reasonable  time  after
   sampling. Each well should have a complete record of historic sampling  events.

•  Survey location coordinates  and elevations for all wells. Make data available to all
   stakeholders.
Frontier Hard Chrome Site                      4                Groundwater Monitoring
Vancouver, Washington                                        Network Optimization

-------
5.0 CITED REFERENCES

AFCEE.  (2003).  Monitoring  and  Remediation  Optimization System (MAROS)  2.2
       Software  Users  Guide.  Air Force  Center  for  Environmental  Excellence.
       http://www.gsi-net.com/software/MAROS V2  1Manual.pdf

AFCEE. (1997). Air Force Center for Environmental  Excellence,  AFCEE Long-Term
       Monitoring Optimization Guide, http://www.afcee.brooks.af.mil.

Aziz, J. A., C. J. Newell, M. Ling, H. S. Rifai and J. R. Gonzales (2003).  "MAROS: A
       Decision Support System for Optimizing Monitoring Plans." Ground Water 41(3):
       355-367.

Gilbert, R. O. (1987). Statistical Methods for Environmental Pollution  Monitoring. New
       York. Van Norstrand Reinhold.

Ridley, M.N., Johnson, V. M and Tuckfield, R. C. (1995). Cost-Effective Sampling of
       Ground Water Monitoring Wells. HAZMACON. San Jose, California.

USEPA. (2001) Frontier Hard Chrome Superfund Site Amended Record of Decision.
       USEPA Region X. August, 2001.

USEPA (1988) Frontier Hard Chrome Superfund Site Record of Decision: Groundwater
       Operable Unit. USEPA Region X. July 1988.

USEPA (1987) Frontier Hard Chrome Superfund Site Record of Decision: Soil Operable
       Unit. USEPA Region X. December, 1987.

USEPA (1992). Methods for Evaluating the Attainment of Cleanup Standards:  Volume 2
       Ground Water.  Washington, D.C.,  United  States Environmental  Protection
       Agency Office of Policy Planning and Evaluation.

Weston (2007). Frontier Hard Chrome Event 11  Long-Term Monitoring Report (June
       2007 Results). Weston Solutions, Inc. August, 2007.
Frontier Hard Chrome Site                     5                Groundwater Monitoring
Vancouver, Washington                                       Network Optimization

-------
            GROUNDWATER MONITORING NETWORK OPTIMIZATION
                         FRONTIER HARD CHROME
                           Vancouver, Washington
TABLES

Table 1   Monitoring Well Network Summary

Table 2   Aquifer Input Parameters

Table 3   Well Trend Summary Results: 2003-2007

Table 4   Moment Estimates and Trends

Table 5   Well Redundancy and Cleanup Status Summary Results

Table 6   MCES Sampling Frequency Analysis Results

Table 7   Final Recommended Groundwater Monitoring Network Frontier Hard Chrome

-------
Issued: 21-DEC-2007
Page 1 of 2
              TABLE 1
MONITORING WELL NETWORK SUMMARY

  LONG-TERM MONITORING OPTIMIZATION
FRONTIER HARD CHROME SUPERFUND SITE
       VANCOUVER, WASHINGTON
Well Name
Hydrologic
Zone
Screened
Interval [ft
bgs]
Source or
Tail (for
MAROS)
Minimum
Sample Date
Maximum
Sample Date
Number of
Samples
(2003-2007)
Current
Sampling
Frequency
Well Description
Zone A
B85-4
RA-MW-11A
RA-MW-12A
RA-MW-13A
RA-MW-14A
RA-MW-15A
RA-MW-16A
RA-MW-17A
W85-6A
W85-7A
W92-16A
W97-18A
W97-19A
W98-20A
W98-21A
W99-R5A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
21.5-26.5
22.9-27.6
23.2-27.9
22.5-27.1
20.3-25.1
22.1-26.6
22.2-26.7
21.7-26.2
17-27
16.5-26.5
24-34
22.5-27.5
20-25
22-27
21-26
22-32
S
S
S
T
T
S
T
T
T
T
T
T
T
T
T
T
2/5/2004
10/16/2003
10/17/2003
10/15/2003
10/15/2003
10/15/2003
10/14/2003
10/14/2003
2/9/2004
2/6/2004
2/5/2004
2/5/2004
2/6/2004
2/7/2004
2/9/2004
2/7/2004
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
11
12
12
12
12
12
12
12
9
11
11
11
11
11
11
11
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Historic high concentrations, monitors central area downgradient
of plume.
Monitors northern edge of plume, near area of historic high, area
of remedy barrier, paired with Zone B well.
Monitors area of historic highest concentrations and permeable
reactive barrier, nested with Zone B wells.
Very low concentration area near permeable reactive barrier,
nested with Zone B wells.
Monitors area of historic high concentrations and eastern
permeable reactive barrier, nested with Zone B well.
Monitors center plume between remedy and B85-4, low Cr in
Zone A, paired with high Cr well in Zone B.
Monitors center plume between remedy and B85-4, low Cr in
Zone A, paired with high Cr well in Zone B.
Northeastern edge in remedy zone, not paired with Zone B well.
Downgradient, center of plume, paired with Zone B well.
Downgradient, center-west of plume, paired with Zone B well.
Delineates western edge of plume near-downgradient of remedy,
low detection frequency, paired with Zone B well.
Delineates far eastern edge of plume, paired with Zone B well.
Delineates far western downgradient edge of plume, paired wit
Zone B well.
Monitors downgradient, center of plume, not paired with Zone B
well.
Monitors downgradient, eastern edge of plume, paired with Zone
Swell.
Monitors farthest downgradient tail, historic edge of plume, paired
with Zone B well.

-------
Issued: 21-DEC-2007
Page 2 of 2
                 TABLE 1
MONITORING WELL NETWORK SUMMARY

  LONG-TERM MONITORING OPTIMIZATION
FRONTIER HARD CHROME SUPERFUND SITE
        VANCOUVER, WASHINGTON
Well Name
Hydrologic
Zone
Screened
Interval [ft
bgs]
Source or
Tail (for
MAROS)
Minimum
Sample Date
Maximum
Sample Date
Number of
Samples
(2003-2007)
Current
Sampling
Frequency
Well Description
Zone B
B85-3
B87-8
RA-MW-11B
RA-MW-12B
RA-MW-12C
RA-MW-13B
RA-MW-13C
RA-MW-14B
RA-MW-15B
RA-MW-16B
W85-6B
W85-7B
W92-16B
W97-18B
W97-19B
W98-21 B
W99-R5B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
B
24-29
24.5-29.5
28.3-32.9
28.3-32.8
34.5-39
27.3-31.9
34.6-39.5
25.5-30.1
27.7-32.5
27.9-32.5
44-49
44-49
35-45
39.5-44.5
40-45
39-44
44-49
T
S
S
T
T
T
T
T
S
S
T
T
S
T
T
T
T
2/5/2004
2/4/2004
10/16/2003
10/17/2003
10/17/2003
10/16/2003
2/3/2004
10/15/2003
10/15/2003
10/14/2003
2/9/2004
2/6/2004
10/14/2003
2/6/2004
2/6/2004
2/9/2004
2/7/2004
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
6/5/2007
11
11
12
12
12
12
11
12
12
12
9
11
12
11
11
11
11
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Monitors upgradient northwest corner, near former FHC building,
west of reducing zone.
Monitors high concentration area immediately downgradieent of
remedy, not paired with Zone A well.
Monitors northern edge of plume, near area of historic high
groundwater concentrations, paired wit A Zone well.
Monitors northern edge of upper Zone B, near area of historic
high groundwater concentrations, paired with Zone A well.
Monitors northern edge of lower Zone B, near area of historic high
groundwater concentrations, paired with Zone A well.
Monitors upper Zone B near reactive barrier, area of low
concentrations.
Monitors lower Zone B near reactive barrier, area of low
concentrations.
Monitors eastern edge of Zone B near remedy, paired with Zone A
well.
Monitors immediately downgradient of permeable barrier, 'hot
spot' in Zone B.
Monitors center plume between remedy and downgradient area,
low Cr in Zone A, paired with high Cr in Zone B.
Downgradient, center of plume, paired with Zone A well.
Downgradient center of plume, paired with Zone A well.
Delineates western edge of plume near-downgradient of remedy,
variable concentrations, paired with Zone A well.
Delineates far eastern edge of plume, paired with Zone A well.
Delineates far western downgradient edge of plume, paired wit
Zone A well.
Monitors downgradient, eastern edge of plume, paired with Zone
A well.
Downgradient tail edge of plume, paired with Zone A well.
     Notes:
     1. Wells listed are in current monitoring program. Data from USEPA Region 10, Sept. 2007. Well locations illustrated on Figi
     2.  Groundwater zones are based on the depth of the well screened interval. Zone A is in the upper alluvial aquifer; Zone B
     3. Number of samples is the number of quarters the well has been sampled 2003-2007.
                                  lure 1.
                                  is in the more transmissive lower depth of the alluvial aquifer.

-------
Issued 21-DEC-2007
Page 1 of 1
                                       TABLE 2
                           AQUIFER INPUT PARAMETERS

                       LONG-TERM MONITORING OPTIMIZATION
                     FRONTIER HARD CHROME SUPERFUND SITE
                              VANCOUVER, WASHINGTON
Parameter
Current Plume5 Length
Maximum Plume5 Length
Plume5 Width
SeepageVelocity (ft/yr) 3
Zone A
Zone B
Distance to Receptors (Columbia River)
GWFIuctuations

SourceTreatment
Contaminant Type
NAPLPresent
Priority Constituent
Chromium (total)
Parameter
Groundwater flow direction
Porosity
Source Location near Well
Source X-Coordinate
Source Y-Coordinate
Coordinate System
Saturated Thickness
Zone A
Zone B
Value
1000
2500
1000

182.5
821.25
3000
Yes
Permeable reactive
barrier/chemical reductant
Metals
No
Cleanup Goals
50
Value
s/sw
0.3
North of RA-MW-1 1
1091615.515
112599.082
NAD 83 SP Washington South

15
50
Units
ft
ft
ft

ft/yr
ft/yr
ft
~

—
-
~

ug/L

225 degrees
-
~
ft
ft


ft
ft
Notes:
1. Aquifer data from ROD and ROD Amendment (USEAPA, 1988; USEPA, 2001).
2. Source coordinates estimated to center of historic FHC building.
3. * = a wide range of transmissivites are present in the aquifer, and groundwater velocity
  calculations result in a range, with values shown being the best estimate.
4. Cleanup objectives are Model Toxics Control Act Method A promulgated by the
   Washington State Department of Ecology for property with unrestricted use.
5. 'Plume' as used in this report descripes the extent of groundwater affected by
  source-associated chromium at any concentration; rather than groundwater above
  the regulatory screening limit.

-------
Issued 21-DEC-2007
Page 1 of 1
                                                                     TABLE 3
                                                 WELL TREND SUMMARY RESULTS: 2003-2007

                                                      LONG-TERM MONITORING OPTIMIZATION
                                                    FRONTIER HARD CHROME SUPERFUND SITE
                                                            VANCOUVER, WASHINGTON
WellName
Number of
Samples
Number of
Detects
Percent
Detection
Maximum
Result [ug/L]
Max Result
Above
Standard?
Average
Result [ug/L]
Average
Result Above
Standard?
Mann-
Kendall
Trend
Linear
Regression
Trend
Overall
Trend Result
Zone A Wells
B85-4
RA-MW-1 1A
RA-MW-12A
RA-MW-1 3A
RA-MW-1 4A
RA-MW-1 5A
RA-MW-1 6A
RA-MW-1 7A
W85-6A
W85-7A
W92-16A
W97-18A
W97-19A
W98-20A
W98-21A
W99-R5A
Zone B Wells
B85-3
B87-8
RA-MW-1 1B
RA-MW-1 2B
RA-MW-1 2C
RA-MW-1 3B
RA-MW-1 3C
RA-MW-1 4B
RA-MW-1 SB
RA-MW-1 6B
W85-6B
W85-7B
W92-16B
W97-18B
W97-19B
W98-21B
W99-R5B
11
12
12
12
12
12
12
12
9
11
11
11
11
11
11
11

11
11
12
12
12
12
11
12
12
12
9
11
12
11
11
11
11
10
10
12
10
9
11
11
11
8
9
7
6
10
10
10
4

8
11
11
10
12
6
10
9
12
12
8
4
12
8
9
10
10
91%
83%
100%
83%
75%
92%
92%
92%
89%
82%
64%
55%
91%
91%
91%
36%

73%
100%
92%
83%
100%
50%
91%
75%
100%
100%
89%
36%
100%
73%
82%
91%
91%
37.7
50.1
5260
4.4
5.4
37
9.2
10.2
14.3
3.6
6.3
0.6
7.9
5.1
7.1
4.1

6.3
241
69.2
26
12.2
7.1
7.3
7
192
225
13
18
225
1.3
12.5
6.6
9.9
No
Yes
Yes
No
No
No
No
No
No
No
No
No
No
No
No
No

No
Yes
Yes
No
No
No
No
No
Yes
Yes
No
No
Yes
No
No
No
No
8.3
9.7
682.0
1.4
1.8
6.2
3.3
5.1
4.6
1.8
1.6
0.5
2.7
2.1
2.6
0.8

2.84
61.8
9.9
5.3
4.0
1.4
2.6
1.5
78.1
46.4
4.5
3.7
46
0.881
3.41
2.77
3.99
No
No
Yes
No
No
No
No
No
No
No
No
No
No
No
No
No

No
Yes
No
No
No
No
No
No
Yes
No
No
No
No
No
No
No
No
D
D
NT
S
S
NT
D
S
S
S
NT
S
PD
S
PD
NT

NT
NT
D
D
S
NT
S
NT
NT
NT
D
D
NT
NT
D
D
D
D
D
NT
S
NT
NT
D
S
S
S
PD
D
D
D
D
NT

NT
NT
D
NT
S
NT
S
NT
NT
NT
D
D
NT
NT
D
D
D
D
D
NT
S
S
NT
D
S
S
S
S
PD
D
PD
D
NT

NT
NT
D
S
S
NT
S
NT
NT
NT
D
D
NT
NT
D
D
D
A/ores
1 . Trends were evaluated for data collected between 2003 and 2007.
2. Number of Samples is the number of quarterly samples for the compound at this location.
4. Screening level Standard from Washington Department of Ecology =  50ug/L. Values above the Standard indicated irgold •
5. D = Decreasing; PD = Probably Decreasing; S = Stable; PI = Probably Increasing; I = Increasing; N/A = Insufficient Data to determine trend;
^TManW-ReWfilMBraWiaiT^^e nRfirfr!SSsS'
-------
Issued 21-DEC-2007
Page 1 of 1
                                           TABLE 4
                              MOMENT ESTIMATES AND TRENDS

                           LONG-TERM MONITORING OPTIMIZATION
                         FRONTIER HARD CHROME SUPERFUND SITE
                                 VANCOUVER, WASHINGTON


Zone

Effective Sample
Event Date
Number of
wells in
network

Dissolved Cr Mass
Estimate [Kg]







Zone A












Zone B






2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006

6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
16
16
16
16
15
15

16
16
16
16
16
Trend
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006

6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
17
17
17
17
16
16

17
17
17
17
17
Trend
0.32
0.63
0.59
0.26
0.37
0.25

0.18
0.26
0.08
0.25
0.25
D
5.66
3.34
2.70
2.76
1.91
1.41

1.16
1.74
0.58
1.68
1.44
D

Distance of Center of
Mass from Source [ft]

830
1192
957
1038
753
931

962
912
1058
984
1010
NT
901
1092
1026
817
709
878

831
792
673
737
881
D
Notes:
1.  Input parameters for the moment analysis are listed in Table 2.
2.  Moments are based on all wells sampled during the quarter including the effective date indicated.
3.  Number of wells is the total number of locations sampled for the plume during the year indicated.
4.  Estimated mass is the total dissolved mass of total chromium within the network indicated.
5.  Trends are Mann Kendall trends on the moments, S=Stable,  D = Decreasing, NT = No Trend.
6.  First moments are illustrated on Figure 2.

-------
Issued: 21-DEC-2007
Page 1 of 1
                                                                   TABLE 5
                                   WELL REDUNDANCY AND CLEAN-UP STATUS SUMMARY RESULTS
                                                   LONG-TERM MONITORING OPTIMIZATION
                                                  FRONTIER HARD CHROME SUPERFUND SITE
                                                         VANCOUVER, WASHINGTON
WellName
Mann-Kendall
Trend
2003-2007
Maximum
Concentration
Cr(ug/L)
Cr Average Slope
Factor
Statistically Below
Screening Level
>80% Power
Sequential T-Test
Result
MAROS
Statistically
Redundant
Recommendation After
Qualitative Review
Zone A
B85-4
RA-MW-11A
RA-MW-12A
RA-MW-13A
RA-MW-14A
RA-MW-15A
RA-MW-16A
RA-MW-17A
W85-6A
W85-7A
W92-16A
W97-18A
W97-19A
W98-20A
W98-21A
W99-R5A
Zone B
B85-3
B87-8
RA-MW-11B
RA-MW-12B
RA-MW-12C
RA-MW-13B
RA-MW-13C
RA-MW-14B
RA-MW-15B
RA-MW-16B
W85-6B
W85-7B
W92-16B
W97-18B
W97-19B
W98-21B
W99-R5B
D
D
NT
S
S
NT
D
S
S
S
NT
S
PD
S
PD
NT

NT
NT
D
D
S
NT
S
NT
NT
NT
D
D
NT
NT
D
D
D
37.70
50.10
5260.00
4.40
5.40
37.00
9.20
10.20
14.30
3.60
6.30
0.60
7.90
5.10
7.10
4.10

6.30
241 .00
69.20
26.00
12.20
7.10
7.30
6.50
192.00
225.00
12.90
17.70
225.00
1.30
12.50
6.60
9.90
0.12
0.43
0.58
0.40
0.09
0.12
0.08
0.15
0.20
0.07
0.27
0.41
0.10
0.09
0.04
0.32

0.26
0.17
0.17
0.17
0.22
0.42
0.23
0.46
0.39
0.35
0.21
0.52
0.33
0.45
0.23
0.08
0.05
YES
YES
WO
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES
YES

YES
WO
YES
YES
YES
YES
YES
YES
WO
WO
-
YES
WO
YES
YES
YES
YES



Attained





Attained

Attained



Attained














Attained



Yes
No
No
No
Yes
Yes
Yes
No
Yes
Yes
Yes
No
No
Yes
No
No

No
Yes
Yes
Yes
Yes
No
Yes
No
No
No
Yes
No
No
No
No
No
No
Retain
Exclude
Retain
Exclude
Exclude
Retain
Retain
Retain
Retain
Exclude
Retain
Retain
Retain
Exclude
Retain
Retain

Retain
Retain
Exclude
Retain
Retain
Exclude
Exclude
Exclude
Retain
Retain
Retain
Exclude
Retain
Retain
Retain
Retain
Retain
Wotes:
1.
  Slope factors close to 1 show the concentrations cannot be estimated from the nearest neighbors, and the well is important in the network.
2. Slope factors were calculated using data between January 2006 and June 2007.
3. Locations with slope factors below 0.3 and area ratios below 0.8 were considered for elimination.
4. Wells statistically below the cleanup level (50 ug/L) by Student's-T Test and >80% Power indicated. "Attained" indicates wells statistically

                                                                             and the final recommendation reflects both
statistical findings and regulatory and site specific factors.

-------
Issued: 21-DEC-2007
Page 1 of 1
                                                                             TABLE 6
                                                     MCES SAMPLING FREQUENCY ANALYSIS RESULTS

                                                            LONG-TERM MONITORING OPTIMIZATION
                                                           FRONTIER HARD CHROME SUPERFUND SITE
                                                                   VANCOUVER, WASHINGTON



Well Name
Recent
Concentration
Rate of Change
[mg/yr]

Recent MK
Trend (2006
2007)
Frequency
Based on
Recent Data
(2006-2007)
Overall
Concentration
Rate of Change
[mg/yr]

Overall MK
Trend
(2003 - 2007)
Frequency
Based on
Overall Data
(2003 - 2007)

MAROS
Recommended
Frequency

Current
Sampling
Frequency

Final
Recommended
Frequency
Zone A Wells
B85-4
RA-MW-11A
RA-MW-12A
RA-MW-13A
RA-MW-14A
RA-MW-15A
RA-MW-16A
RA-MW-17A
W85-6A
W85-7A
W92-16A
W97-18A
W97-19A
W98-20A
W98-21A
W99-R5A
-4.00E-06
1.31E-06
-1.27E-03
-1.42E-06
-1.84E-07
-3.69E-06
-2.06E-06
-7.68E-06
1.71E-06
5.67E-07
-2.46E-07
-1 .49E-07
1 .84E-06
1 .66E-06
-1 .20E-07
-3.32E-07
S
S
NT
S
S
S
S
S
S
S
NT
PD
NT
NT
S
S
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
-1.48E-05
-1.75E-05
5.99E-04
-7.25E-07
-3.91 E-07
1.17E-06
-2.56E-06
-1.45E-06
-4.82E-06
-2.96E-07
-2.35E-06
-4.89E-09
-3.38E-06
-2.23E-06
-2.70E-06
-1.12E-06
D
D
NT
S
S
NT
D
S
S
S
NT
S
PD
S
PD
NT
Annual
Annual
Quarterly
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Biennial
Biennial
Quarterly
Biennial
Biennial
Biennial
Biennial
Biennial
Biennial
Biennial
Biennial
Annual
Biennial
Biennial
Biennial
Biennial
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Zone B Wells
B85-3
B87-8
RA-MW-11B
RA-MW-12B
RA-MW-12C
RA-MW-13B
RA-MW-13C
RA-MW-14B
RA-MW-15B
RA-MW-16B
W85-6B
W85-7B
W92-16B
W97-18B
W97-19B
W98-21B
W99-R5B
-7.10E-06
9.07E-05
-1 .35E-05
-9.87E-06
-8.58E-07
-5.49E-07
-8.67E-07
-2.35E-07
-2.64E-04
3.04E-05
-7.12E-06
O.OOE+00
2.55E-05
3.45E-07
1.74E-06
-2.01 E-06
-2.71E-06
S
S
NT
S
NT
S
S
S
S
NT
S
S
NT
NT
NT
S
S
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
5.25E-07
-4.39E-05
-2.08E-05
-6.41 E-06
-5.10E-07
-7.81 E-07
-5.17E-07
-8.40E-07
4.74E-06
1.01E-05
-5.15E-06
-1.13E-05
8.99E-06
2.61 E-07
-5.83E-06
-3.20E-06
-5.94E-06
NT
NT
D
D
S
NT
S
NT
NT
NT
D
D
NT
NT
D
D
D
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Biennial
Annual
Biennial
Biennial
Biennial
Biennial
Biennial
Biennial
Annual
Annual
Biennial
Biennial
Annual
Biennial
Biennial
Biennial
Biennial
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Quarterly
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Notes:
1. Concentration rate of change is from linear regression calculations. 'Recent' concentration rate of change and MK trends are calculated from data collected 2006 - 2007.
2. MK trend = Mann  Kendall trend. D = Decreasing, PD = Probably Decreasing, S = Stable, NT = No Trend.
3. Recent data frequency is the estimated sample frequency based on the recent trend.
4. Overall rate of change and MK trend are for the full data set (2003-2007) for each well. The overall result is the estimated sample frequncy based on the full data record.
6. MAROS Recommended Frequency is the final frequency from the MAROS calculations based on both recent and overall trends.
7. Current frequency is the approximate sampling frequency currently implemented.
8. The final recommended sampling frequency is based on a combination of qualitative and statistical evaluations.

-------
Issued: 21-DEC-2007
Page 1 of 1
                                                TABLE 7
              FINAL RECOMMENDED MONITORING NETWORK FRONTIER HARD CHROME

                                 LONG-TERM MONITORING OPTIMIZATION
                               FRONTIER HARD CHROME SUPERFUND SITE
                                       VANCOUVER, WASHINGTON



WellName
Zone A Wells
B85-4
RA-MW-11A
RA-MW-12A
RA-MW-13A
RA-MW-14A
RA-MW-15A
RA-MW-16A
RA-MW-17A
W85-6A
W85-7A
W92-16A
W97-18A
W97-19A
W98-20A
W98-21A
W99-R5A
Zone B Wells
B85-3
B87-8
RA-MW-11B
RA-MW-12B
RA-MW-12C
RA-MW-13B
RA-MW-13C
RA-MW-14B
RA-MW-15B
RA-MW-16B
W85-6B
W85-7B
W92-16B
W97-18B
W97-19B
W98-21B
W99-R5B
Total Chromium

Percent
Detection
Mann
Kendall
Trend
Statistically
Below
Standard?
MAROS
Redundancy
Determination
Recommendation
After Qualitative
Review
Final
Recommended
Frequency

91%
83%
100%
83%
75%
92%
92%
92%
89%
82%
64%
55%
91%
91%
91%
36%

73%
100%
92%
83%
100%
50%
91%
75%
100%
100%
89%
36%
100%
73%
82%
91%
91%
D
D
NT
S
S
NT
D
S
S
S
NT
S
PD
S
PD
NT

NT
NT
D
D
S
NT
S
NT
NT
NT
D
D
NT
NT
D
D
D
V
V

V
V
V
V
V
V
V
V
V
V
V
V
V

V

V
V
V
V
V
V


V
V

V
V
V
V
V



V
V
V

V
V
V


V




V
V
V
V

V



V






Retain
Exclude
Retain
Exclude
Exclude
Retain
Retain
Retain
Retain
Exclude
Retain
Retain
Retain
Exclude
Retain
Retain
Annual
Exclude
Annual
Exclude
Exclude
Annual
Annual
Annual
Annual
Exclude
Annual
Annual
Annual
Exclude
Annual
Annual

Retain
Retain
Exclude
Retain
Retain
Exclude
Exclude
Exclude
Retain
Retain
Retain
Exclude
Retain
Retain
Retain
Retain
Retain
Annual
Annual
Exclude
Annual
Annual
Exclude
Exclude
Exclude
Annual
Annual
Annual
Exclude
Annual
Annual
Annual
Annual
Annual
Notes:
1.  Mann Kendall Trends: D = Decreasing; PD = Probably Decreasing; S = Stable; PI = Probably Increasing; I = Increasing;
   NT = No Trend; ND = well has all non-detect.
2.  Mann-Kendall trends 2003 - 2007 are shown.
3.  Statistically below standard based Student's T-Test with >80% statistical power for data between 2003-2007.
   Cleanup standard is Washington Ecology MTCA A = 50ug/L Total Chromium.
4.  MAROS redundancy indicates well has low SF and high AR and CR.
5.  Final Recommendation based on statistical as well as qualitative evaluation.

-------
FIGURES
            GROUNDWATER MONITORING NETWORK OPTIMIZATION
                          FRONTIER HARD CHROME
                            Vancouver, Washington
Figure 1      Frontier Hard Chrome Groundwater Monitoring Network

Figure 2      Frontier Hard Chrome Concentration Trend and First Moment Results

Figure 3      Zone A Chromium Concentration Uncertainty

Figure 4      Zone B Chromium Concentration Uncertainty

Figure 5      Zone A Optimized Network Concentration Uncertainty

Figure 6      Zone B Optimized Network Concentration Uncertainty

Figure 7      Frontier Hard Chrome Zone A Source Area Summary

Figure 8      Frontier Hard Chrome Zone B Source Area Summary

-------
Average [Cr] [mg/L]

    A    ND-0.001

    A    0.001 - 0.01

    A    0.01 - 0.05

    A    0.05-0.5

    A    >0.5
                      Legend
- Roads

  Cassidy Mfg.

 Soil Treatment
 Area (Approximate)

 ISRM Groundwater
 Treatment Zone
 (Approximate)
Notes:

1. Aerial map from 1994 shows historic site features
  and road locations. FHC buildings and industrial
  area to the south have been demolished.
2. The Soil Remedy and ISRM Groundwater Remedy
   are approximate areas based on USEPAmaps.
3. Groundwater monitoring locations for each Zone
  are indicated. Average Cr concentrations 2003 - 2007
  at each location are indicated by color.
4. Total Cr MTCA Standard A cleanup level = 0.05 mg/L.
Large Map Scale (ft)

0     150     300
FRONTIER HARD CHROME
     GROUNDWATE
 MONITORING NETWORK
      Vancouver, Washington
                             Coord. Sys.
                                   NAD 83 SP Wash. S. FT.
                             Drawn By:
                                      MV
                             Chk'd By:
                                      MV
                                                                                                                                                                                   AppVd By:
                                                                                                                                                                                            MV
                                                     Issued:
                                                          21-DEC-2007
                                                     Revised:
                Map ID:
                                                                                                                                                                         FIGURE 1

-------
Zone A Well  Locations
 Legend
  Mann Kendall Trend Cr

   ^     Decreasing

   •    Probably Decreasing

   O    Stable

   •    Probably Increasing

   •    Increasing
Non Detect (2003-2007)
No Trend

Insufficient Data

First Moment
(Effective Date Shown)
Soil Remedy Area

ISRMTreatment

Cassidy Mfg.

Roads
Notes:

1. Aerial map from 1994 shows historic site features
  and road locations. FHC buildings and industrial
  area to the south have been demolished.
2. Mann-Kendall trends we re determined for data
  collected between 2003-2007.
3. First moments were determined using quarterly data.
  An effective date of the quarterly sampling event is
  indicated.
4. Total CrMTCA Standard A cleanup level = 0.05 mg/L.
    Scale (ft)
    ^•=
0     150    300
                                    FRONTIER HARD CHROME

                                     CONCENTRATION TREND AND
                                        FIRST MOMENT RESULTS

                                          Vancouver, Washington
                            Coord. Sys.
                                  NAD 83 SP Wash. S. FT.
Drawn By:
         MV
                            Chk'd By:
                                     MV
                                                                                                                                                AppVd By:
                                                    Issued:
                             21-DEC-2007
                        Revised:
                        Map ID:
                                                                                                                                                                              FIGURE 2

-------
NORT
1 1 2600.0 -
1 1 2400.0 -
1 1 2200.0 -
1 1 2000.0 -

111800.0-
111600.0-


111400.0-



111200.0-
111000.0-
1 1 0800.0 -
1089
H
Figures RA-MW-HA RA-MW-12A RA_MW_17A
Zone A Chromium \,. \ / RA-MW-ISA
W92-16A ^mk_Jli/
Concentration Uncertainty J^S^*T£-— RA-MW-MA
x' / | / B85-4 //
/S I S / /
/' / i / ;//
/ / s \ i / /I
/ * / i / / / /
/ \\/ / /
/ _— ( inmr n 	 Jr vny°5 BA * /
/ ^- ••— WSS-^A— —— — —— ^ WWO-1 DA y I
xx ^^*^^ / xx ^ / /
W97-1 9A X, ^- ^ 1 xX l^/
/-- S ' X^ 3 / /
/ -----..^ 1 xxX I / /
. ' — Jr* V /
/ X W98-20A ^^ /
7 xX ^^^^ /
/ ^X ^^^ /
/ S ,' ^
/ ^x s ^s ^^_____ Very low spatial uncertainty across the plume
/ /' ^'^-— — -" 	 "
/ xx ^^ 	 "
' x ^
/ /'',/'''
^'
New Location
Analysis for
CHROMIUM, TOTAL
Existing
Locations
Potential areas for
new locations are
indicated by triangles
with a high SF level.
Estimated SF Level:
S - Small
M - Moderate
L - Large
E - Extremely large
High SF-> high
estimation error ->
possible need for
new locations
Low SF -> low
estimation error ->
no need for new
locations

Back to
Access
V J

1 ' EAST
500.0 1090000.0 1090500.0 1091000.0 1091500.0 1092000.0 1092500.0

-------
1 1 9ftnn n New Location
i i zouu.u ~i~


1 1 2600.0 -


1 1 2400.0 -


1 1 2200.0 -


1 1 2000.0 -



111800.0-



111600.0-


111400.0-



111200.0-


111000.0-

1 1 0800.0 -
Figure 4
Zone B Chromium
B85-3
Concentration Uncertainty RAMWHR
' RA-MW-11B . RA-MW-12B/C
2006-2007 v. X/*K /
><< ' \ V\/ RA-MW-13B/C
W92-16B x ^^LA OW. /
~^^>Cy/ •Alsl&jMtr--- " — RA-MW-14B
' ' ^^^^%T^^^Bt~---__
XX / M / | X^^NS^"S^:^ RA-MW-16B
/ / / II B87-8 ^S. ~~/f W97-18B
X / / || //'\v
/ / * II / ' ^\.
/ .' ^ || / / RA-MW-15B
S ^ / / M M Si
/ / ' m \ \ / ,
/ "'/ /
/ '/ J
* _/ 1^ *

X -,-"'*" W85-7B ^ ' /
xx --" ^ s \
™»-^^- s \ > /

/ — — 	 	 x I /
, — — 	 	 V I /
/ "~"~—3^ W98-21B
/ ^-^
/ ^^^
' S ^"'
f ^
/ ^''
/ ^'
/ ^'*'
/ **
/ ^''
/ ^^'"
/ ^
/ ^
/ ^^

11111
Analysis for
CHROMIUM, TOTAL



Existing
•
Locations

Potential areas for
new locations are
indicated by triangles
with a high SF level.
Estimated SF Level:
S - Small
M - Moderate
L - Large
E - Extremely large

High SF-> high
estimation error ->
possible need for
new locations
Low SF -> low
estimation error ->
no need for new
locations



Back to
Access
V J
EAST
1089500.0 1090000.0 1090500.0 1091000.0 1091500.0 1092000.0 1092500.0

-------
! ! ofiNoo o1^ 	 New Location
1 1 ZDUU.U -j~~~^^


1 1 2400.0 -


1 1 2200.0 -



1 1 2000.0 -


111800.0-

111600.0-




111400.0-



111200.0-

111000.0-


1 1 0800.0 -
Figure 5 RA MW I?A
3 KA MVV ^A RA-MW-17A
Zone A
Optimized Network W92-16^ ^^fs RA_MW_16A
Concentration Uncertainty x' j V'TSTV^'NS
2006-2007 ' I/ S^1 s^^N
RA,BW-15A < Tl 	 -^
x , . ~~~^P W97-18A
x 1 / B85-4 .7
1 S / "
/ 1 / / /
X 1 / S / /
' / / /
XX / ' '
.' 1 / / /
xX I / / /
7 S ' ' /X /
X i / / /
/ I'/ s/
X — -"f W85-6A /
X _ — — """""" I /
^c\a929A S 1 /
' "~~"~~~"~ — — I/
/ ~""~—jJ( W98-21A
/ ^^
' ^
/ ^^
/ ^^^
/ S s'
1 ^'
/ ^s'
t s'
/ ^'
, S
f ^
/ /''''
1 ^
1^'


11111
Analysis for
CHROMIUM, TOTAL


Existing
•
Locations


Potential areas for
new locations are
indicated by triangles
with a high SF level.
Estimated SF Level:
S - Small
M - Moderate
L - Large
E- Extremely large

High SF-> high
estimation error ->
possible need for
new locations

Low SF-> low
estimation error ->
no need for new
locations

s -\
Back to
Access

V J
EAST
1089500.0 1090000.0 1090500.0 1091000.0 1091500.0 1092000.0 1092500.0

-------
H9Rnnn™ New Location
I I ZOUU.U -j~~~^^

1 1 2600.0 -


1 1 2400.0 -



1 1 2200.0 -


1 1 2000.0 -


111800.0-



111600.0-


111400.0-




111200.0-

111000.0-


1 1 0800.0 -
Figure 6
Zone B
Optimized Network B85 3
Concentration Uncertainty ^L RA MW 123/0
2006-2007 W92-16B / 1 \^V RA-MW-15B
•x* ^^3T — —— ff^i RA MW 16C

' xx | ^— 	 _?^^^
X XX | M I B87-8 ~~~ — — *p W97-18B
xX xX I | /I
/ / I'M //
xX xxX 1 ' / 1
/ X 1 i X /
x'V s .' x^ /
X x 'I X /
X XX ll XX /
X x 1 / M'
Xx ____^W85-6B /
^y 	 	 — — \ '
W97-19B X'' _____ — — """" . /
^~ — 	 S \ ,
/ ~~~~~~~~~~~^ \ '

/ ~~~-yil W98-21B
/ ^'"'
/ S ^"^

' ^-^
/ ^-^^
/ ^-"
/ ^
/ ^"'
1 ^'
/ ^
• ^
/ ^'"
/ ^, -^
iTW99-R5B

11111
Analysis for
CHROMIUM, TOTAL


Existing
Locations


Potential areas for
new locations are
indicated by triangles
with a high SF level.
Estimated SF Level:
S - Small
M - Moderate
L - Large
E- Extremely large

High SF-> high
estimation error ->
possible need for
new locations
Low SF-> low
estimation error ->
no need for new
locations



Back to
Access
V J

EAST
1089500.0 1090000.0 1090500.0 1091000.0 1091500.0 1092000.0 1092500.0

-------
                                                                                                                               0      If
Average Cr Concentration 2003-2007
Legend
Average Cr Concentration [mg/L]

    A    ND-0.001

    A    0.001 - 0.01

    A    0.01 - 0.05

    A    0.05-0.5

    A    >0.5

    Cr Screening Level = 0.05 mg/L
         Mann Kendall Trend Cr

Decreasing             •     Non Detect (2003-2007)

Probably Decreasing     •

Stable                  •

Probably Increasing
No Trend

Insufficient Data
Increasing
Soil Remedy Area

Proposed Development
Recommended Sampling Frequency

      D    Annual Sampling

            Eliminate from
            routine monitoring
Notes:
1. Average Total Cr concentrations calculated
  for data 2003-2007.
2. Mann Kendall trends were determined
  for Total Cr 2003-2007.
3. All built structures are proposed
  for the property redevelopment.
  Drawingsfordevelopment received
  from developer October, 2007.
FRONTIER HARD CHROME
    ZONE A SOURCE AREA
          SUMMARY

      Vancouver, Washington
                                                                                                                Coord. Sys.
                                                                                                                                                   NAD 83 SP Wash. S. FT.
                                                                                                                                             Drawn By:
                                                                                                MV
                                                                                                                                             Chk'd By:
                                                                                                                                                      MV
                                                                                                                                                                                         AppVd By:
                                                                                                                                                                                                  MV
                                                                                                                                        Issued:
                                                                                                                                                                          21-DEC-2007
                                                                                                                                                                     Revised:
                                                                                                                                        Map ID:
                                                                                                                                                                           FIGURE 7

-------
                                                 T"*	L-L
^ i-r^—j_i4_|   ra -i ™-» - ,•   cj_

   •——IlJ       ^	I—IfM-MMMeB	.
         -  .,!_ I           -        T   X|
   RA-MW-15B p=-—
 Average Cr Concentration 2003-2007
Legend
Average Cr Concentration [mg/L]

    A    ND-0.001

    A    0.001 - 0.01

    A    0.01 - 0.05

    A    0.05-0.5

    A    >0.5

    Cr Screening Level = 0.05 mg/L
                                           Mann Kendall Trend Cr

                                  Decreasing             •    Non Detect (2003-2007)

                                  Probably Decreasing     •

                                  Stable                 •

                                  Probably Increasing
No Trend

Insufficient Data
                                  Increasing
Soil Remedy Area

Proposed Development
Recommended Sampling Frequency

      D    Annual Sampling

            Eliminate from
            routine monitoring
Notes:
1. Average Total Cr concentrations calculated
  for data 2003-2007
2. Mann Kendall trends were determined
  for Total Cr 2003-2007.
3. All built structures are proposed
  for the property redevelopment.
  Drawingsfordevelopment received
  from developer October, 2007.
FRONTIER HARD CHROM
    ZONE B SOURCE ARE
          SUMMARY

      Vancouver, Washington
                                                                                                                Coord. Sys.
                                                                                                                                                                                     NAD 83 SP Wash. S. FT.
                                                                                                                                                                               Drawn By:
                                                                                                MV
                                                                                                                                                                               Chk'd By:
                                                                                                                                                                                        MV
                                                                                                                                                                                          AppVd By:
                                                                                                                                        Issued:
                                                                                                                                                                                                            21-DEC-2007
                                                                                                                                                                                                       Revised:
                                                                                                                                        Map ID:
                                                                                                                                                                                                                       FIGURE 8

-------
          GROUNDWATER MONITORING NETWORK OPTIMIZATION
                       FRONTIER HARD CHROME
                        Vancouver, Washington
APPENDIX A:	

MAROS 2.2 Methodology

-------
                            APPENDIX A
                  MAROS 2.2 METHODOLOGY


                               Contents

1.0 MAROS Conceptual Model	1

2.0 Data Management	2

3.0 Site Details	2

4.0 Constituent Selection	3

5.0 Data Consolidation	3

6.0 Overview Statistics: Plume Trend Analysis	3
      6.1 Mann-Kendall Analysis	4
      6.2 Linear Regression Analysis	4
      6.3 Overall Plume Analysis	5
      6.4 Moment Analysis	6

7.0 Detailed Statistics: Optimization Analysis	8
      7.1 Well Redundancy Analysis- Delaunay Method	8
      7.2 Well Sufficiency Analysis - Delaunay Method	9
      7.3 Sampling Frequency - Modified CES Method	10
      7.4 Data Sufficiency- Power Analysis	11

Cited References

Tables

      Table 1 Mann-Kendall Analysis Decision Matrix
      Table 2 Linear Regression Analysis Decision Matrix
Figures
      Figure 1 MAROS Decision Support Tool Flow Chart
      Figure 2 MAROS Overview Statistics Trend Analysis Methodology
      Figure 3 Decision Matrix for Determining Provisional Frequency

-------
MAROS METHODOLOGY

MAROS is a collection of tools in one software package that is used in an explanatory,
non-linear but linked fashion. The tool includes models, statistics,  heuristic rules, and
empirical relationships to assist the user in optimizing a groundwater monitoring network
system.  The  final optimized network maintains  adequate delineation while providing
information on plume dynamics over time. Results generated from the software tool can
be used to develop lines of evidence, which, in combination with  expert opinion, can  be
used to inform regulatory decisions for  safe  and economical long-term  monitoring of
groundwater plumes.  For a detailed  description  of the structure of the  software and
further utilities,  refer to  the  MAROS  2.2   Manual   (AFCEE,  2003; http://www.gsi-
net.com/software/MAROS V2 1Manual.pdf) and Aziz et al., 2003.

1.0 MAROS Conceptual  Model

In MAROS 2.2, two levels of analysis are used for optimizing long-term monitoring plans:
1) an overview statistical  evaluation with interpretive trend analysis based on temporal
trend  analysis and  plume  stability  information;  and 2) a  more detailed  statistical
optimization based on spatial and temporal redundancy reduction methods (see Figures
A.1 and A.2 for further details). In general, the MAROS method applies to 2-D aquifers
that have relatively simple site hydrogeology.  However, for a multi-aquifer (3-D) system,
the user has the option to apply the statistical analysis layer-by-layer.

The overview  statistics or interpretive trend analysis  assesses the general  monitoring
system  category  by considering  individual well concentration  trends, overall  plume
stability, hydrogeologic factors  (e.g., seepage velocity, and current plume length), and
the location of potential receptors (e.g., property boundaries or drinking water wells). The
method relies on temporal trend analysis to assess plume stability, which is then used to
determine the general monitoring system category.   Since the  monitoring system
category is evaluated for both  source and tail regions of the plume, the  site wells are
divided into two different zones: the source zone and the tail zone.

Source  zone  monitoring  wells could  include  areas  with non-aqueous  phase liquids
(NAPLs), contaminated vadose zone soils, and areas where  aqueous-phase releases
have been introduced into ground water. The  source zone generally contains locations
with historical high ground water concentrations of the COCs. The tail zone is usually the
area downgradient of the contaminant source zone.  Although  this classification is a
simplification of the plume conceptual model,  this broadness makes the user aware  on
an  individual   well basis  that  the concentration  trend results  can  have a different
interpretation depending on the well location in and around the  plume.  The location and
type of the individual wells allows further interpretation  of the trend results,  depending  on
what type of well  is being analyzed (e.g., remediation  well, leading plume edge well, or
monitoring well).  General recommendations for the monitoring network frequency and
density are suggested based on heuristic rules  applied  to the source and tail trend
results.

The detailed   statistics level of  analysis  or  sampling optimization  consists of well
redundancy and  well sufficiency  analyses using the Delaunay method, a sampling
frequency analysis using  the Modified Cost Effective  Sampling (MCES) method  and a

Appendix A                               7                       MAROS 2.2 Methodology

-------
data sufficiency analysis  including statistical  power  analysis.  The  well  redundancy
analysis is designed to minimize monitoring locations and the Modified CES method is
designed to minimize the  frequency of sampling.  The data sufficiency  analysis uses
simple statistical methods to  assess the  sampling  record to determine if groundwater
concentrations are statistically below target levels and  if the current monitoring network
and record is sufficient in terms of evaluating concentrations at downgradient locations.

2.0 Data Management

In MAROS, ground water monitoring data can be imported from simple database-format
Microsoft® Excel spreadsheets, Microsoft Access  tables,  previously created  MAROS
database archive files, or entered manually. Monitoring data interpretation in MAROS is
based  on historical analytical data from  a  consistent set of wells  over  a  series of
sampling events. The analytical  data is composed of the well name,  coordinate location,
constituent, result, detection limit and associated data qualifiers. Statistical validity of the
concentration trend  analysis requires constraints on the minimum data input of at least
four  wells  (ASTM  1998)  in  which COCs have been detected.  Individual  sampling
locations need to include data from at least six  most-recent sampling events. To ensure
a meaningful comparison  of COC concentrations over time and space, both data quality
and  data  quantity  need to be considered.  Prior to  statistical analysis, the user can
consolidate  irregularly sampled data  or smooth data  that might result from seasonal
fluctuations or a change in site conditions.  Because MAROS is a terminal analytical tool
designed for long-term planning,  impacts of seasonal variation  in the water unit are
treated on a broad scale, as they relate to  multi-year trends.

Imported ground water monitoring data and the site-specific information entered in Site
Details can be archived and exported as  MAROS archive files. These archive  files can
be appended as new monitoring data  becomes available, resulting in a dynamic long-
term  monitoring  database that  reflects  the  changing  conditions  at  the  site  (i.e.
biodegradation, compliance attainment, completion of remediation  phase,  etc.).   For
wells with a limited monitoring history,  addition of information as it becomes available
can change the frequency or identity of wells in the network.

3.0 Site Details

Information needed for the MAROS analysis includes  site-specific parameters such as
seepage velocity and current plume  length  and width. Information on the location of
potential receptors relative to  the source and tail regions of the plume is entered  at this
point.  Part of the trend analysis methodology applied in MAROS  focuses on where the
monitoring well  is located, therefore the user needs  to divide site wells into two different
zones: the source zone or the tail zone. Although this classification  is a simplification of
the well function, this broadness makes the user aware on an individual well basis that
the concentration trend results can have a different  interpretation depending on the well
location in and around the plume. It is up to the  user to make further interpretation of the
trend results, depending on what type  of well is being  analyzed (e.g., remediation well,
leading plume  edge well, or monitoring  well).  The  Site Details  section  of  MAROS
contains a preliminary map of  well locations to confirm well coordinates.
Appendix A                               2                       MAROS 2.2 Methodology

-------
4.0 Constituent Selection

A database with multiple COCs can be entered into the MAROS software.  MAROS
allows the analysis of up to 5 COCs concurrently and users can pick COCs from a list of
compounds existing in  the monitoring  data.  MAROS runs separate optimizations for
each compound.  For sites with a single source, the suggested  strategy is to choose one
to  three priority COCs for the optimization.  If, for example, the site contains multiple
chlorinated volatile organic compounds (VOCs), the standard sample chemical analysis
will evaluate all VOCs, so the sample locations and frequency should based  on the
concentration trends  of the most  prevalent, toxic or mobile  compounds.  If different
chemical classes  are present, such  as metals  and chlorinated  VOCs,  choose and
evaluate the priority constituent in each chemical class.

MAROS includes  a short module that provides recommendations on prioritizing COCs
based on  toxicity,  prevalence, and  mobility of the compound.  The toxicity ranking is
determined by examining a  representative concentration for each compound  for the
entire site. The representative concentration is then compared to the screening  level
(PRG or  MCL)  for  that compound  and  the  COCs  are  ranked  according  to the
representative  concentrations  percent exceedence of  the  screening   level.   The
evaluation of prevalence is performed by determining a representative concentration for
each well location  and evaluating the total exceedences (values above screening levels)
compared to the total number of wells.  Compounds found over screening levels are
ranked for mobility based on Kd  (sorption partition coefficient).  The MAROS  COC
assessment provides  the relative ranking of each COC, but the user must choose which
COCs are included in the analysis.

5.0 Data Consolidation

Typically, raw data from long-term monitoring have been measured irregularly  in  time or
contain many non-detects, trace level results, and duplicates. Therefore, before the data
can be further analyzed, raw data are filtered, consolidated, transformed,  and possibly
smoothed to allow for a consistent dataset meeting the minimum data requirements for
statistical analysis mentioned previously.

MAROS allows users to specify the period of interest in which data will be  consolidated
(i.e., monthly,  bi-monthly,  quarterly,  semi-annual,  yearly,  or  a biennial  basis).  In
computing the representative value when consolidating, one  of four statistics can be
used: median,  geometric mean, mean,  and maximum. Non-detects can be transformed
to  one half the reporting or method detection limit (DL), the DL, or a  fraction of the DL.
Trace level results can be represented by their actual values, one half of the DL, the DL,
or  a fraction of their actual values.  Duplicates are reduced in  MAROS by  one of three
ways: assigning the average, maximum, or first value. The reduced data for each  COC
and each well can be viewed as a time series in a graphical form on a linear or semi-log
plot generated  by the  software.

6.0 Overview Statistics: Plume Trend Analysis

Within the MAROS software there are historical data analyses  that support a conclusion
about plume stability  (e.g., increasing plume, etc.) through statistical trend analysis of

Appendix A                              3                      MAROS 2.2 Methodology

-------
historical  monitoring  data.   Plume stability results  are assessed from  time-series
concentration data with the application of three statistical  tools:  Mann-Kendall Trend
analysis, linear regression trend analysis and moment analysis. The two trend methods
are used to estimate the concentration trend for each well and each COC based  on a
statistical  trend  analysis of concentrations versus time at each  well.  These trend
analyses are then consolidated to give the user a general plume stability estimate and
general monitoring frequency and density recommendations (see Figures A.1  through
A.3 for further step-by-step details).  Both qualitative and quantitative plume information
can be  gained by these evaluations of monitoring network historical data trends  both
spatially and temporally.  The  MAROS Overview Statistics are the foundation the  user
needs to make informed optimization decisions at the site. The Overview Statistics are
designed  to  allow site personnel  to  develop  a better  understanding of the plume
behavior over time  and  understand how  the individual well  concentration trends are
spatially distributed within the plume. This step allows the user to gain  information that
will support a more informed decision to be made in the next  level  or detailed statistics
optimization analysis.

6.1 Mann-Kendall Analysis

The Mann-Kendall test is a statistical procedure that is well suited for analyzing trends in
data over time.  The Mann-Kendall test can be viewed as a non-parametric test for zero
slope of the first-order regression of time-ordered concentration data versus time.  One
advantage of the Mann-Kendall test  is that it does not require any assumptions as to the
statistical distribution of the data (e.g. normal, lognormal, etc.) and can be used with data
sets which include irregular sampling intervals and missing data. The Mann-Kendall test
is designed for analyzing a single  groundwater constituent,  multiple constituents are
analyzed separately.   The Mann-Kendall  S statistic measures the trend in the data:
positive values  indicate an increase in concentrations over time  and negative values
indicate a decrease in concentrations over time. The strength of the trend is proportional
to the  magnitude of the Mann-Kendall statistic (i.e., a large  value indicates a strong
trend). The confidence in the trend  is determined by consulting the S statistic and the
sample  size, n, in a Kendall probability  table such as the one reported in Hollander and
Wolfe (1973).

The concentration trend is determined for each well and each  COC based on results of
the S statistic, the confidence  in the trend,  and the Coefficient of Variation (COV). The
decision matrix for this evaluation is shown in Table 3. A Mann-Kendall statistic that is
greater  than 0 combined with  a confidence of greater than 95%  is categorized as  an
Increasing  trend while a Mann-Kendall statistic of less than 0 with a confidence between
90% and 95% is defined as a probably Increasing trend, and so on.

Depending  on  statistical indicators,  the  concentration trend is  classified  into six
categories:

    •   Decreasing (D),
    •   Probably Decreasing (PD),
    •   Stable (S),
    •   No Trend (NT),
    •   Probably Increasing (PI)
    •   Increasing (I).
Appendix A                               4                      MAROS 2.2 Methodology

-------
These trend estimates are then analyzed to identify the source and tail region overall
stability category (see Figure 2 for further details).

6.2 Linear Regression Analysis

Linear  Regression  is  a  parametric  statistical  procedure that is  typically  used  for
analyzing trends  in data over time.   Using this type of analysis, a higher degree of
scatter simply corresponds to a wider confidence interval about the average log-slope.
Assuming the sign (i.e., positive or negative) of the estimated log-slope is correct, a level
of confidence that the slope is not zero can be  easily determined.  Thus, despite a poor
goodness of fit, the overall trend in the data may still be ascertained,  where low levels of
confidence correspond to "Stable" or "No Trend" conditions (depending on the degree of
scatter) and higher levels of confidence  indicate the stronger likelihood of a trend. The
linear regression analysis is  based  on the  first-order linear  regression  of  the log-
transformed concentration data  versus time.   The  slope  obtained from  this log-
transformed regression, the confidence level  for this  log-slope, and the COV of the
untransformed data are used to determine the  concentration trend.  The decision matrix
for this evaluation is shown in Table 4.

To estimate the  confidence in the log-slope, the  standard  error of the log-slope is
calculated. The coefficient of variation, defined as the standard  deviation divided by the
average, is used  as a secondary measure  of scatter to distinguish between "Stable" or
"No Trend" conditions for negative slopes.  The Linear Regression Analysis is designed
for analyzing  a  single  groundwater  constituent;  multiple constituents are  analyzed
separately, (up to five COCs  simultaneously).  For this evaluation, a decision matrix
developed by Groundwater Services, Inc. is also used to determine the "Concentration
Trend" category (plume stability) for each well.

Depending on statistical   indicators,  the  concentration trend  is  classified  into  six
categories:

       Decreasing (D),
       Probably Decreasing (PD),
       Stable (S),
       No Trend (NT),
       Probably Increasing (PI)
       Increasing (I).

The resulting confidence in the trend, together with the log-slope and the COV  of the
untransformed  data,  are  used  in the  linear  regression  analysis  decision matrix to
determine the concentration trend. For example,  a positive log-slope with a confidence
of less than 90% is categorized as having No Trend whereas  a negative log-slope is
considered Stable if the COV is less than 1 and categorized as  No Trend if the COV is
greater than 1.

6.3 Overall Plume Analysis

General  recommendations  for  the  monitoring  network  frequency and  density  are
suggested  based on heuristic  rules applied to  the  source  and  tail trend results.
Appendix A                               5                       MAROS 2.2 Methodology

-------
 Individual well trend results are consolidated and weighted by the MAROS according to
 user input, and  the direction and strength of contaminant concentration trends  in the
 source zone and tail zone for each COC are determined.  Based on
    i)     the consolidated trend analysis,
    ii)    hydrogeologic factors (e.g., seepage velocity), and
    iii)    location of  potential receptors  (e.g., wells,  discharge points, or property
          boundaries),
the software suggests a general optimization  plan for the current monitoring system in
order to efficiently but effectively monitor groundwater in the future.  A flow chart utilizing
the trend analysis results and other site-specific parameters to form a general sampling
frequency and well density recommendation  is outlined in Figure 2.  For example,  a
generic plan  for  a  shrinking  petroleum hydrocarbon plume  (BTEX) in  a  slow
hydrogeologic  environment (silt) with no nearby  receptors would entail minimal,  low
frequency sampling of just a few indicators.  On the other hand, the  generic plan for a
chlorinated solvent plume in a fast hydrogeologic environment that is expanding but has
very erratic concentrations over time would  entail  more extensive, higher  frequency
sampling.  The generic plan is based on a heuristically derived algorithm for assessing
future  sampling  duration,  location and density that  takes  into consideration  plume
stability.  For  a detailed description of the heuristic rules used in the  MAROS software,
refer to the MAROS 2.2Manual (AFCEE, 2003).

 6.4 Moment Analysis

 An analysis of moments can help resolve plume trends, where the zeroth moment shows
 change in dissolved mass vs. time, the first moment shows the center of mass location
 vs. time,  and the second moment shows the spread of the plume  vs. time. Moment
 calculations can predict how the  plume will  change  in the future if further statistical
 analysis is applied to the moments to identify a trend (in this  case, Mann Kendall  Trend
 Analysis is applied). The trend analysis of moments can be summarized as:

    •  Zeroth Moment: An estimate of the total mass of the constituent for each sample
       event
    •  First Moment: An estimate of the center of mass for each sample event
    •  Second Moment: An estimate of the spread of the plume around the center of
       mass

 The role  of moment analysis in  MAROS  is to provide a  relative estimate of  plume
 stability and  condition within the context of results from other MAROS modules. The
 Moment  analysis algorithms  in   MAROS  are  simple approximations  of complex
 calculations and are meant to  estimate changes in total mass,  center of  mass and
 spread of mass for complex well networks.  The Moment Analysis module is sensitive to
 the number  and arrangement of wells in each  sampling event, so,  changes  in the
 number and  identity of wells during monitoring events, and the parameters chosen for
 data consolidation can cause changes in the estimated moments.

 Plume stability may vary by constituent, therefore the MAROS Moment analysis can be
 used to evaluate multiple  COCs simultaneously which can be  used  to provide a quick
 way of comparing individual plume parameters to determine  the size and movement of
 constituents relative to  one another. Moment analysis in the  MAROS software can also

 Appendix A                              6                      MAROS 2.2 Methodology

-------
be used to  assist the  user  in evaluating the impact on plume delineation  in future
sampling events by removing identified "redundant" wells from a long-term monitoring
program (this analysis was not performed as part of this study, for more details on this
application of moment analysis refer to the MAROS Users Manual (AFCEE, 2003)).

The zeroth moment is  the sum of concentrations for all monitoring wells and is a mass
estimate. The zeroth moment calculation can show high variability over time, largely due
to the  fluctuating concentrations  at the most contaminated  wells as well as varying
monitoring  well  network.  Plume analysis  and   delineation  based  exclusively  on
concentration can exhibit fluctuating temporal  and spatial values. The mass estimate is
also  sensitive to the extent of the site monitoring well network over time.  The zeroth
moment trend over time is determined by using the Mann-Kendall  Trend Methodology.
The zeroth Moment trend test allows the user to understand  how the  plume mass has
changed over time. Results for the trend include:  Increasing, probably  Increasing, no
trend, stable, probably decreasing, decreasing or not applicable (N/A) (Insufficient Data).
When considering the results of the zeroth moment trend, the following factors should be
considered which could effect the calculation and interpretation of the plume mass over
time: 1) Change in the  spatial distribution of the wells sampled historically 2) Different
wells sampled within the well network over time (addition and subtraction of well within
the network). 3) Adequate versus inadequate delineation of the plume over time

The first moment estimates the center of mass, coordinates  (Xc and Yc)  for each
sample event and COC. The changing center of mass locations indicate the movement
of the center of mass over time. Whereas, the distance from the original source location
to the center of mass locations indicate the movement of the center of mass over time
relative to the original source.  Calculation of the first moment normalizes the spread by
the concentration indicating the center of mass. The first moment trend of the distance to
the center of mass over time shows  movement of  the plume in relation to the original
source location  over time.  Analysis of the  movement of mass should be viewed as it
relates to 1)  the original source location of contamination 2) the direction of groundwater
flow and/or 3) source removal or remediation. Spatial and temporal trends in the center
of mass can indicate spreading or shrinking or transient movement based on season
variation in rainfall or other hydraulic considerations.  No appreciable movement or a
neutral trend in the center of mass would indicate plume stability. However, changes in
the first moment over time do not necessarily completely characterize the  changes in the
concentration distribution  (and  the  mass)  over  time.  Therefore, in  order  to  fully
characterize  the plume the  First Moment  trend  should  be compared to the  zeroth
moment trend (mass change over time).

The second moment indicates the spread of the contaminant about the center of mass
(Sxx and Syy), or the distance of  contamination from the center of mass  for a particular
COC and sample event. The Second Moment represents the spread of the plume over
time  in both  the x and y directions. The Second Moment trend indicates the spread  of
the plume about the center of mass. Analysis of the spread of the plume should be
viewed as it relates to  the direction  of groundwater flow.  An Increasing  trend in the
second moment indicates an  expanding plume, whereas a declining trend in the second
moment indicates a shrinking plume. No appreciable movement or a neutral trend in the
center  of mass would indicate plume stability.  The second moment provides a measure
of the  spread of the concentration  distribution about the  plume's  center of mass.

Appendix A                              7                       MAROS 2.2 Methodology

-------
However,  changes in the second moment over time  do not necessarily completely
characterize the changes in the  concentration  distribution (and the mass) over time.
Therefore, in order to fully characterize the plume the Second Moment trend should be
compared to the zeroth moment trend (mass change over time).

7.0 Detailed Statistics: Optimization Analysis

Although  the  overall  plume  analysis  shows  a  general  recommendation regarding
sampling frequency reduction and a general sampling density, a more detailed analysis
is also available with the MAROS 2.2 software in order to allow for further reductions on
a well-by-well  basis  for  frequency,  well  redundancy,  well  sufficiency  and sampling
sufficiency. The MAROS Detailed Statistics allows for a quantitative analysis for spatial
and temporal optimization of the well network on a well-by-well basis.  The results from
the Overview  Statistics  should  be considered along with  the MAROS optimization
recommendations gained from the  Detailed  Statistical  Analysis described previously.
The MAROS Detailed Statistics results  should be reassessed in view of site knowledge
and regulatory requirements  as  well  as  in  consideration of the Overview Statistics
(Figure 2).

The Detailed Statistics or Sampling Optimization MAROS  modules can be used  to
determine the  minimal number  of sampling locations  and  the  lowest frequency  of
sampling that can still meet the requirements of sampling spatially and temporally for an
existing  monitoring program.  It also provides an analysis of  the sufficiency of data for
the monitoring  program.

Sampling optimization in MAROS  consists of four parts:

   •  Well redundancy analysis using the Delaunay method
   •  Well sufficiency analysis using the Delaunay method
   •  Sampling frequency determination using the Modified CES method
   •  Data sufficiency analysis using statistical power analysis.

The well  redundancy analysis using the  Delaunay method  identifies and eliminates
redundant locations from the monitoring network. The well sufficiency analysis can
determine the areas where new sampling locations might be needed. The Modified CES
method  determines the optimal sampling frequency for a sampling location based on the
direction,  magnitude,  and uncertainty in its concentration trend.  The data sufficiency
analysis examines the risk-based site cleanup status  and power and  expected sample
size associated with the cleanup status evaluation.

7.1 Well Redundancy Analysis - Delaunay Method

The well redundancy analysis using the Delaunay method  is designed to select the
minimum  number of sampling locations based  on the  spatial analysis  of the relative
importance of each sampling location in the monitoring  network. The approach allows
elimination of sampling locations that have little impact on the historical characterization
of a contaminant plume.  An extended method  or wells sufficiency analysis, based on
the Delaunay  method, can  also  be  used for recommending new sampling locations.
Appendix A                               g                      MAROS 2.2 Methodology

-------
Details about the Delaunay method can be found in Appendix A.2 of the MAROS Manual
(AFCEE, 2003).

Sampling Location determination uses the Delaunay triangulation  method to determine
the significance  of the  current sampling locations  relative to  the  overall monitoring
network. The Delaunay method calculates the network Area and Average concentration
of the plume using data from multiple monitoring wells. A slope factor (SF) is calculated
for each well to indicate  the significance of this well in the system (i.e. how removing a
well changes the average concentration.)

The Sampling  Location optimization process  is performed in a stepwise fashion.  Step
one involves assessing the significance of the well in the system, if a well has a small SF
(little significance to the network), the well may be  removed from the monitoring network.
Step two involves evaluating the information loss of removing a well from the network. If
one well has a small  SF, it may or may not be eliminated depending on whether the
information loss is significant.  If the information loss is not significant, the well can be
eliminated from the monitoring network and the process  of optimization  continues with
fewer wells.  However if the well information loss is significant then the  optimization
terminates. This  sampling optimization process allows the user to assess "redundant"
wells that will not incur significant information loss on a constituent-by-constituent basis
for individual sampling events.

7.2 Well Sufficiency Analysis - Delaunay Method

The well sufficiency analysis, using the Delaunay method, is designed to  recommend
new sampling locations in areas within the existing monitoring network where there is a
high level of uncertainty in contaminant concentration. Details about the well sufficiency
analysis can be found  in Appendix A.2 of the MAROS Manual (AFCEE, 2003).

In many cases, new sampling locations  need to be added to the existing network to
enhance the spatial plume characterization.   If the MAROS algorithm calculates a high
level of uncertainty in predicting the constituent concentration for a particular area, a new
sampling location is recommended.  The Slope Factor (SF) values obtained  from the
redundancy evaluation  described  above  are used to  calculate  the concentration
estimation  error for each triangle area  formed  in the  Delaunay triangulation.   The
estimated SF value for each  area is  then classified into four levels:  Small, Moderate,
Large,  or Extremely large (S, M, L, E) because the larger the estimated SF value, the
higher  the estimation error  at this  area.    Therefore,  the triangular areas  with  the
estimated SF value at the Extremely large or Large level can be candidate regions for
new sampling locations.

The results from the Delaunay method and the method for determining new  sampling
locations are derived solely from the spatial configuration  of the monitoring network and
the spatial pattern of the contaminant plume.  No parameters such as the hydrogeologic
conditions  are  considered  in the  analysis.    Therefore, professional judgment and
regulatory considerations must be used to make final decisions.

7.3 Sampling Frequency Determination - Modified CES Method
Appendix A                               g                      MAROS 2.2 Methodology

-------
The  Modified CES method optimizes  sampling  frequency for each sampling location
based on the magnitude, direction, and uncertainty of its concentration trend derived
from its  recent and historical monitoring records. The Modified Cost Effective Sampling
(MCES)  estimates a  conservative lowest-frequency  sampling  schedule for a given
groundwater monitoring location that still provides needed information for regulatory and
remedial decision-making.  The MCES method was developed on the basis of the Cost
Effective Sampling (CES) method developed by  Ridley et al  (1995).  Details about the
MCES method can be found in Appendix A.9 of the MAROS Manual (AFCEE, 2003).

In order to estimate the least frequent  sampling schedule for a monitoring location that
still provides enough information for regulatory and remedial decision-making, MCES
employs three steps to determine the sampling frequency.   The first step involves
analyzing frequency based on recent trends.  A preliminary location sampling frequency
(PLSF) is developed based on the rate of change of well concentrations calculated by
linear regression along with the  Mann-Kendall trend analysis of  the  most recent
monitoring data (see Figure 3).  The variability within the sequential sampling data is
accounted for by the Mann-Kendall analysis.  The rate of change vs. trend result matrix
categorizes wells as requiring annual, semi-annual or quarterly sampling.  The PLSF is
then reevaluated  and  adjusted based on overall trends.  If the  long-term  history  of
change  is significantly greater than the recent trend, the frequency may be reduced by
one level.

The  final step  in the analysis involves reducing  frequency based  on risk, site-specific
conditions, regulatory requirements or other external issues.  Since not all compounds in
the target being assessed  are equally harmful,  frequency  is reduced  by one level if
recent maximum  concentration for a  compound of high risk is less than  1/2 of the
Maximum Concentration Limit (MCL).  The result of applying this method is a suggested
sampling frequency based on  recent sampling data trends and overall sampling data
trends and expert judgment.

The  final sampling frequency determined from the MCES method  can be Quarterly,
Semiannual, Annual, or Biennial.  Users can further reduce the sampling frequency to,
for example, once every three years, if the trend estimated from Biennial data (i.e., data
drawn once every two years from the original data) is  the same as that estimated from
the original data.

7.4 Data Sufficiency Analysis - Power Analysis

The  MAROS Data Sufficiency module employs  simple statistical methods to evaluate
whether the collected  data  are adequate both in quantity and in  quality  for revealing
changes in  constituent  concentrations.   The first  section  of the  module  evaluates
individual well  concentrations to determine if  they  are statistically below a  target
screening level.   The second section includes a simple calculation for  estimating
projected groundwater concentrations at a specified point downgradient of the plume.  A
statistical Power analysis is then  applied to the projected concentrations to determine if
the downgradient concentrations are statistically below the  cleanup standard.  If the
number  of projected concentrations is below the  level to provide statistical significance,
then the  number of sample events required to statistically confirm concentrations below
standards is estimated from the Power analysis.

Appendix A                               70                     MAROS 2.2 Methodology

-------
Before testing the  cleanup status  for individual wells, the stability or trend of the
contaminant plume should be evaluated. Only after the plume has reached stability or is
reliably diminishing can we conduct a test to  examine the cleanup status of wells.
Applying  the analysis to wells in an  expanding plume may cause incorrect conclusions
and is less meaningful.

Statistical power analysis  is a technique for interpreting the results of statistical tests.
The Power of a statistical  test is a measure of the ability of the test to detect an effect
given that the effect actually exists. The method provides additional information about a
statistical  test:  1)  the  power  of the statistical  test,  i.e.,  the probability of finding  a
difference in the variable of interest when a difference truly exists; and 2) the expected
sample size of a future sampling plan given the minimum detectable  difference  it  is
supposed to detect.  For example, if the mean concentration is lower than the cleanup
goal  but a statistical test cannot prove this,  the power and expected sample size can tell
the reason and how many more samples are needed  to result in a significant test.  The
additional samples can be obtained by a longer period of sampling or an increased
sampling  frequency.   Details about the  data  sufficiency  analysis  can  be found  in
Appendix A.6 of the MAROS Manual (AFCEE, 2003).

When applying the MAROS power analysis method, a  hypothetical statistical compliance
boundary (HSCB) is  assigned to be  a  line perpendicular to  the  groundwater  flow
direction  (see  figure below).  Monitoring  well concentrations are projected onto the
HSCB using the distance from each well to the compliance boundary along with a decay
coefficient. The projected concentrations from each well and each sampling event are
then used in the risk-based power analysis. Since there may be more than one sampling
event selected by the user, the risk-based power analysis results  are given on an event-
by-event  basis. This power analysis can then indicate if target are statistically achieved
at the HSCB.  For instance, at a site where the historical monitoring record is short with
few wells, the  HSCB  would be  distant; whereas, at a  site with longer duration of
sampling with many wells,  the HSCB would be close. Ultimately, at a site the  goal would
be to have  the HSCB coincide  with  or  be within  the actual  compliance boundary
(typically  the site property line).
Appendix A                                77                      MAROS 2.2 Methodology

-------
                                                          Concentrations
                                                          projected to this
                                                       •""* line
                                                             The nearest
                                                             downgradient
                                                             receptor
                Groundwater flow direction
In order to perform a risk-based cleanup status evaluation for the whole site, a strategy
was developed as follows.

    •   Estimate  concentration versus distance decay coefficient from plume centerline
       wells.
    •   Extrapolate  concentration  versus  distance  for  each well  using  this  decay
       coefficient.
    •   Comparing the extrapolated concentrations with  the compliance  concentration
       using power analysis.

Results from  this analysis  can be  Attained  or  Not Attained, providing a  statistical
interpretation of whether the cleanup goal has  been met on the site-scale from the risk-
based  point of view.  The results as a function of time can be used to evaluate if the
monitoring system has enough power at each step  in the sampling record to indicate
certainty of compliance by the plume location  and condition relative to the compliance
boundary.  For example, if results are Not Attained at early sampling events but are
Attained in recent sampling events, it indicates that the recent sampling record provides
a powerful enough result to indicate compliance of the plume relative to the location of
the receptor or compliance boundary.
Appendix A
                                       12
MAROS 2.2 Methodology

-------
CITED REFERENCES

AFCEE 2003. Monitoring and Remediation Optimization System (MAROS) 2.1 Software
Users  Guide.  Air  Force  Center for  Environmental  Excellence,  http://www.gsi-
net.com/software/MAROS  V2 1Manual.pdf

AFCEE. 1997. Air Force Center for  Environmental Excellence, AFCEE Long-Term
Monitoring Optimization Guide, http://www.afcee.brooks.af.mil.

Aziz, J. A., C. J.  Newell, M.  Ling, H. S. Rifai and J. R.  Gonzales (2003).  "MAROS: A
Decision Support System  for Optimizing Monitoring Plans."  Ground Water 41(3): 355-
367.

Gilbert, R. O., 1987,  Statistical Methods for Environmental Pollution Monitoring, Van
Nostrand Reinhold, New York, NY, ISBN 0-442-23050-8.

Hollander,  M. and Wolfe, D.  A. (1973).  Nonparametric Statistical Methods, Wiley, New
York, NY.

Ridley, M.N.  et al., 1995. Cost-Effective Sampling of Groundwater Monitoring Wells, the
Regents of UC/LLNL, Lawrence Livermore National Laboratory.

U.S. Environmental Protection Agency, 1992. Methods for Evaluating the Attainment of
Cleanup Standards Volume 2: Ground Water.

Weight, W. D. and J. L. Sonderegger (2001). Manual of Applied Field Hydrogeology.
New York,  NY, McGraw-Hill.
Appendix A                              73                     MAROS 2.2 Methodology

-------
Mann-Kendall
Mann-Kendall
Statistic
S>0
S>0
S>0
S<0
S<0
S<0
S<0
TABLE 1
Analysis Decision Matrix
Confidence in the
Trend
> 95%
90 - 95%
< 90%
< 90% and COV > 1
< 90% and COV < 1
90 - 95%
> 95%
(Aziz, et. al., 2003)
Concentration Trend
Increasing
Probably Increasing
No Trend
No Trend
Stable
Probably Decreasing
Decreasing
Linear Regression
Confidence in the
Trend
< 90%
90 - 95%
> 95%
TABLE 2
Analysis Decision Matrix (Aziz, et. al., 2003)
Log-slope
Positive Negative
COV < 1 Stable
No Trend
COV > 1 No Trend
Probably Increasing Probably Decreasing
Increasing Decreasing

-------
                                   MAROS: Decision Support Tool
    MAROS is a collection of tools in one software package that is used in an explanatory, non-linear fashion.  The tool
    includes  models, geostatistics,  heuristic  rules,  and empirical relationships  to assist  the  user  in  optimizing  a
    groundwater monitoring network system while maintaining adequate delineation of the plume as well as knowledge
    of the plume state over time. Different users utilize the tool in different ways and interpret the results from a different
    viewpoint.
                                                    T
                                          Overview Statistics
    What it is: Simple,  qualitative and quantitative plume information can be gained through evaluation of monitoring
    network historical data trends both spatially and temporally. The MAROS Overview Statistics are the foundation the
    user needs to make  informed optimization decisions at the site.

    What it does: The Overview Statistics are designed to allow site personnel to develop a better understanding of the
    plume behavior over time and understand how the individual well concentration trends are spatially distributed within
    the plume.  This step allows the user to gain information that will support a more informed decision to be made in the
    next level of optimization analysis.

    What are the tools: Overview Statistics includes two analytical tools:

         1)  Trend Analysis: includes Mann-Kendall and  Linear Regression statistics for individual wells and results in
             general heuristically-derived monitoring categories with  a suggested  sampling  density and monitoring
             frequency.

         2)   Moment Analysis: includes dissolved  mass estimation (0th Moment),  center of mass (1st Moment), and
             plume spread (2nd Moment) over time.  Trends of these moments  show  the  user another  piece of
             information about the plume stability overtime.

    What is the product: A first-cut blueprint  for a future long-term  monitoring program that  is intended to be  a
    foundation for more detailed statistical analysis.
                                                    T
                                           Detailed Statistics
    What it is: The MAROS Detailed Statistics allows for a quantitative analysis for spatial and temporal optimization of
    the well network on a well-by-well basis.

    What it does: The results from the Overview Statistics should be  considered along side the MAROS optimization
    recommendations gained from the Detailed Statistical Analysis.  The MAROS Detailed Statistics results should be
    reassessed in view of site knowledge and regulatory requirements as well as the Overview Statistics.

    What are the tools: Detailed Statistics includes four analytical tools:

         1)   Sampling Frequency Optimization: uses the Modified CES  method to establish a recommended future
             sampling frequency.

         2)   Well  Redundancy Analysis:  uses the Delaunay  Method to  evaluate  if any wells within the  monitoring
             network are redundant and can be eliminated without any significant loss of plume information.

         3)   Well  Sufficiency  Analysis:  uses the  Delaunay Method to  evaluate areas  where new  wells are
             recommended within the monitoring network due to high levels of concentration uncertainty.

         4)  Data  Sufficiency Analysis: uses  Power Analysis to assess  if the historical monitoring data record has
             sufficient power to accurately reflect  the  location of the  plume  relative  to the  nearest receptor or
             compliance point.

    What is the product: List of wells to remove from the  monitoring program,  locations where monitoring wells may
    need to be added, recommended frequency  of sampling for each well,  analysis if the  overall system is statistically
    powerful to monitor the plume.
Figure 1.  MAROS Decision Support Tool Flow Chart

-------
                            Select Representative Wells in "Source" and "Plume" Zone
                                           Source Zone   i  Tail Zone

                            Identify Site Constituents of Concern (COCs).
                            (Assistance provided by software.)

                           Analyze Lines of Evidence (LOEs)
                           for Plume Stability (by well and by COC)
                           Categorize concentrations of COC in each well as:
                                    • Increasing (I)   	
                                    • Probably Increasing (PI)    	
                                    • No Trend (NT)  	
                                    • Stable (S)
                                    • Probably Decreasing  (PD)   	
                            for Each Well Based On All
                            LOE's
                                                                     SOURCE   PLUME
                            "Lump Lines of Evidence"
                            Determine General Trend for Source and
                            Tail Zones
                                  Increasing (I)
                                  Probably Increasing (PI)
                                  No Trend (NT)
                                  Stable (S)
                                  Probably Decreasing (PD)
                                  Decreasing (D)

                             "Lump Wells" in Source and Tail Zone
                           Determine
                           LTMP
                           Monitoring
                           Category
                           for COC By
                           Source / Tail
                           
-------
       Sampling
       Frequency
       Q: Quarterly
       S: SemiAimual
       A: Annual
0>
CE3
T3

I
   PI
        Rate of Change (Linear Regression)
          High  MH Medium  LM  Low
Figure 3. Decision Matrix for Determining Provisional Frequency (Figure A.3.1 of the
        MAROS Manual (AFCEE 2003)

-------
           GROUNDWATER MONITORING NETWORK OPTIMIZATION
                       FRONTIER HARD CHROME
                         Vancouver, Washington
APPENDIX B:	

MAROS Reports

Zone A
     Mann-Kendall Reports
     Moment Reports
Zone B
     Mann-Kendall Reports
     Moment Reports

-------
 MAROS Mann-Kendall Statistics  Summary
 Project:  FHC
 Location:  Vancouver
                User Name: MV
                State:  Washington
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
Source/ Number of
Well Tail Samples
Number of
Detects
Coefficient
of Variation
Mann-Kendall
Statistic
Confidence
in Trend
All
Samples Concentration
"ND" ? Trend
CHROMIUM, TOTAL
RA-MW-11A
RA-MW-12A
RA-MW-15A
B85-4
W85-6A
RA-MW-13A
RA-MW-14A
RA-MW-1 7A
W99-R5A
W85-7A
W92-16A
W97-18A
W97-19A
W98-20A
W98-21A
RA-MW-1 6A
S
S
S
S
T
T
T
T
T
T
T
T
T
T
T
T
12
12
12
11
9
12
12
12
11
11
11
11
11
11
11
12
10
12
11
10
8
10
9
11
4
9
7
6
10
10
10
11
1.23
2.18
1.61
1.48
0.96
0.75
0.78
0.57
1.29
0.51
1.17
0.07
0.81
0.74
0.73
0.65
-43
-18
-2
-25
-6
-9
0
-17
0
-14
-5
-9
-21
-17
-21
-26
99.9%
87.5%
52.7%
97.0%
69.4%
70.4%
47.3%
86.0%
46.9%
84.0%
61 .9%
72.9%
94.0%
89.1%
94.0%
95.7%
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
D
NT
NT
D
S
S
S
S
NT
S
NT
S
PD
S
PD
D
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A)-
 Due to insufficient Data (< 4 sampling events); Source/Tail (S/T)

     The Number of Samples and Number of Detects shown above are post-consolidation values.
MAROS Version 2,.2 2006, AFCEE
Thursday, September 13, 2007
                                                                                           Page 1 of 1

-------
 MAROS  Mann-Kendall Statistics Summary
Well: B85-4
Well Type: s
COC: CHROMIUM, TOTAL
                                                   Time Period: 10/15/2003   to  6/5/2007
                                                   Consolidation Period:  No Time Consolidation
                                                   Consolidation Type: Median
                                                   Duplicate Consolidation: Average
                                                   ND Values:  Specified Detection Limit

                                                   J Flag Values : Actual Value
                                      Date
O)


o
I
o
o

3.5E-02 -
3.0E-02 -
2.5E-02 -
2.0E-02 •
1.5E-02-
1.0E-02-
5.0E-03 -

n np4-nn .



*



•
* * *
* • * +
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   97.0%

                                                                               Coefficient of Variation:
                                                                                         1.48
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
B85-4
B85-4
B85-4
B85-4
B85-4
B85-4
B85-4
B85-4
B85-4
B85-4
B85-4
Well Type
s
s
s
s
s
s
s
s
s
s
s
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
3.8E-02
8.1E-03
3.7E-03
1.1E-03
2.7E-02
5.8E-03
9.0E-04
1.5E-03
5.0E-04 ND
2.8E-03
2.4E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                       9/13/2007
                                                                            Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: RA-MW-11A
Well Type:  s
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Median
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


5nc n9
.UE-U^ •
4.5E-02 -
2- 4.0E-02 -
|" 3.5E-02 -
r 3.0E-02 •
o
s 2.5E-02 •
i 2.0E-02 •
| 1.5E-02-
0 1.0E-02-
5.0E-03 -

Data Table:
Well Well Ty
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
RA-MW-11A S
Date
/ /VVVV* /VVV* /,/
^^^>O^'i
-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-12A
Well Type:  s
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Median
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


6npi.nn
.utruu •
_ 5.0&-00 -
£ 4.0&-00 -
o
s 3.0&00 •

§ 2.0&-00 -
o
0 1.0&-00 -
Oncu-nn
m\ICr\I\I
Data Table:
Well Well Ty
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
RA-MW-12A S
Date
/ /vvvv* //*
^*^*
•




•
•
* * •* * + *


Effective
Pe Date Constituent
10/15/2003 CHROMIUM, TOTAL
2/10/2004 CHROMIUM, TOTAL
4/5/2004 CHROMIUM, TOTAL
8/15/2004 CHROMIUM, TOTAL
5/5/2005 CHROMIUM, TOTAL
12/12/2005 CHROMIUM, TOTAL
3/8/2006 CHROMIUM, TOTAL
6/15/2006 CHROMIUM, TOTAL
9/25/2006 CHROMIUM, TOTAL
12/15/2006 CHROMIUM, TOTAL
3/30/2007 CHROMIUM, TOTAL
6/5/2007 CHROMIUM, TOTAL

/vv\/
<) <$• ^





* + *


Result (mg/L) Flag
6.4E-01
1.8E-01
1.4E-01
8.0E-02
1.1E-01
1.3E+00
8.5E-02
1.3E-01
5.3E+00
8.1E-02
7.9E-02
1.1E-01

Mann Kendall S Statistic:
I ^?8
Confidence in
Trend:
1 87.5%
Coefficient of Variation:
I 2'18

Mann Kendall
Concentration Trend:
(See Note)
I NT

Number of Number of
Samples Detects
2 2
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                     9/13/2007
                        Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-13A
Well Type:  T
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Median
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


5nc ni
.uc-uo •
4.5E-03 -
;;[• 4.0E-03 -
|" 3.5E-03 -
r 3.0E-03 •
o
s 2.5E-03 •
i 2.0E-03 •
| 1.5E-03-
0 1.0E-03-
5.0E-04 -

Data Table:
Well Well Ty
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
RA-MW-13A T
Date
/ /vvvv* /vt
^*^*




• • *
* * *


Effective
Pe Date Constituent
10/15/2003 CHROMIUM, TOTAL
2/10/2004 CHROMIUM, TOTAL
4/5/2004 CHROMIUM, TOTAL
8/15/2004 CHROMIUM, TOTAL
5/5/2005 CHROMIUM, TOTAL
12/12/2005 CHROMIUM, TOTAL
3/8/2006 CHROMIUM, TOTAL
6/15/2006 CHROMIUM, TOTAL
9/25/2006 CHROMIUM, TOTAL
12/15/2006 CHROMIUM, TOTAL
3/30/2007 CHROMIUM, TOTAL
6/5/2007 CHROMIUM, TOTAL

/vv\/
<) <$• ^




•
•
•


Result (mg/L) Flag
1.3E-03
4.4E-03
5.0E-04 ND
1.3E-03
5.6E-04
2.0E-03
1.9E-03
1.5E-03
6.3E-04
5.0E-04 ND
1.4E-03
1.1E-03

Mann Kendall S Statistic:
I ^9
Confidence in
Trend:
1 70.4%
Coefficient of Variation:
I °75
Mann Kendall
Concentration Trend:
(See Note)
I S

Number of Number of
Samples Detects
2 1
1 1
1 0
1 1
1 1
1 1
1 1
1 1
1 1
1 0
1 1
1 1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                     9/13/2007
                        Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: RA-MW-14A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
                                      Date

, — ,
B>
o
1
Concent
6.UE-03 -
5.0E-03 -
4.0E-03 -
3.0E-03 •
2.0E-03 -
1.0E-03-
n np4-nn .
•


•
' • •. '•
•
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   47.3%

                                                                                Coefficient of Variation:
                                                                                         0.78
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
Data Table:

Well
RA-MW-14A
RA-MW-14A
RA-MW-14A
RA-MW-14A
RA-MW-14A
RA-MW-14A
RA-MW-14A
RA-MW-14A
RA-MW-14A
RA-MW-14A
RA-MW-14A
RA-MW-14A

Well Type
T
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
10/15/2003
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007

Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL

Result (mg/L) Flag
5.0E-04 ND
5.4E-03
5.0E-04 ND
2.0E-03
7.3E-04
3.2E-03
1.8E-03
1.8E-03
1.4E-03
5.0E-04 ND
2.2E-03
1.6E-03
I S
Number of Number of
Samples Detects
2 0
1 1
1 0
1 1
1 1
1 1
1 1
1 1
1 1
1 0
1 1
1 1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/13/2007
                          Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-15A
Well Type:  s
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Median
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


A np no
4.UC-U^ •
3.5E-02 -
^ 3.0E-02 •
~ 2.5E-02 -
o
s 2.0E-02 •

| 1.5E-02-
o 1.0E-02-
O
5.0E-03 -

Data Table:
Well Well Ty
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
RA-MW-15A S
Date
/ /vvvv* /vvv* /,/
^^<^^><^'i
-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-16A
Well Type:  T
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Median
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


1 HP no
I .UC-U^ •
9.0E-03 -
^ 8.0E-03 -
|" 7.0E-03 -
r 6.0E-03 •
o
s 5.0E-03 •
i 4.0E-03 •
| 3.0E-03 •
0 2.0E-03 •
1.0E-03-

Data Table:
Well Well Ty
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
RA-MW-16A T
Date
/ /vvvv* /v%
^*^*
•



•
•
*


Effective
Pe Date Constituent
10/15/2003 CHROMIUM, TOTAL
2/10/2004 CHROMIUM, TOTAL
4/5/2004 CHROMIUM, TOTAL
8/15/2004 CHROMIUM, TOTAL
5/5/2005 CHROMIUM, TOTAL
12/12/2005 CHROMIUM, TOTAL
3/8/2006 CHROMIUM, TOTAL
6/15/2006 CHROMIUM, TOTAL
9/25/2006 CHROMIUM, TOTAL
12/15/2006 CHROMIUM, TOTAL
3/30/2007 CHROMIUM, TOTAL
6/5/2007 CHROMIUM, TOTAL

>*vv\/
<) <$• ^





* •


Result (mg/L) Flag
4.8E-03
9.2E-03
2.0E-03
3.5E-03
2.2E-03
4.1E-03
3.7E-03
2.8E-03
1.7E-03
5.0E-04 ND
2.9E-03
2.6E-03

Mann Kendall S Statistic:
_

Confidence in
Trend:
1 95.7%
Coefficient of Variation:
1 0.65
Mann Kendall
Concentration Trend:
(See Note)
I °

Number of Number of
Samples Detects
2 2
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 0
1 1
1 1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                     9/13/2007
                        Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: RA-MW-17A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
                                      Date

B)
o
1
c

o
o

1.2E-U2 -
1.0E-02-
8.0E-03 -
6.0E-03 •

4.0E-03 -
2.0E-03 -
n np4-nn .
•
•
•
•
* * *
•
•
*
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   86.0%

                                                                                Coefficient of Variation:
                                                                                         0.57
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
Data Table:

Well
RA-MW-17A
RA-MW-17A
RA-MW-17A
RA-MW-17A
RA-MW-17A
RA-MW-17A
RA-MW-17A
RA-MW-17A
RA-MW-17A
RA-MW-17A
RA-MW-17A
RA-MW-17A

Well Type
T
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
10/15/2003
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007

Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL

Result (mg/L) Flag
6.3E-03
1.0E-02
2.6E-03
5.0E-03
9.2E-04
7.6E-03
8.6E-03
5.7E-03
4.0E-03
5.0E-04 ND
5.0E-03
4.9E-03
I S
Number of Number of
Samples Detects
2 2
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 0
1 1
1 1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/13/2007
                          Page 1 of 1

-------
 MAROS  Mann-Kendall Statistics  Summary
Well: W85-6A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to  6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit

J Flag Values : Actual Value
                                      Date
    £

_J
B)
o
1
Concent
1.4E-02-
1.2E-02-
1.0E-02-
8.0E-03 •
6.0E-03 -
4.0E-03 •
2.0E-03 -
n np4-nn .
*

*
* * * * *
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   69.4%

                                                                               Coefficient of Variation:
                                                                                        0.96
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
W85-6A
W85-6A
W85-6A
W85-6A
W85-6A
W85-6A
W85-6A
W85-6A
W85-6A
Well Type
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
1.4E-03
1.4E-02
9.1E-03
2.9E-03
2.2E-03
4.1E-03
5.0E-04 ND
3.4E-03
3.2E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                       9/13/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall Statistics  Summary
Well: W85-7A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to  6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit

J Flag Values : Actual Value
                                      Date


_J
O)
c
O
1
1
O
O



3.5E-03 -
3.0E-03 -
2.5E-03 -


2.0E-03 •
1.5E-03-
1.0E-03-

5.0E-04 -
n np4-nn .

»

* *


^
* * •
* * *


» »

                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   84.0%

                                                                               Coefficient of Variation:
                                                                                        0.51
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
W85-7A
W85-7A
W85-7A
W85-7A
W85-7A
W85-7A
W85-7A
W85-7A
W85-7A
W85-7A
W85-7A
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
1.7E-03
5.0E-04 ND
3.6E-03
2.8E-03
1.9E-03
1.7E-03
1.5E-03
1.6E-03
5.0E-04 ND
2.7E-03
1.5E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
0
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                       9/13/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall Statistics  Summary
Well: W92-16A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to  6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values :  Actual Value
                                      Date

_J
B)
o
1
Concer
6.0E-03 •
5.0E-03 •
4.0E-03 -
3.0E-03 -
2.0E-03 -
1.0E-03-
n np4-nn .
•

•

•
* * * * * *
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   61.9%

                                                                               Coefficient of Variation:
                                                                                        1.17
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
                                                                                    [     NT
 Data Table:
Well
W92-16A
W92-16A
W92-16A
W92-16A
W92-16A
W92-16A
W92-16A
W92-16A
W92-16A
W92-16A
W92-16A
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L)
4.2E-03
5.0E-04
6.3E-03
7.0E-04
5.0E-04
5.0E-04
1.1E-03
2.1E-03
5.0E-04
5.6E-04
9.4E-04
Flag

ND


ND
ND


ND


Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
0
1
1
0
0
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                       9/13/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall Statistics  Summary
Well: W97-18A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to  6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit

J Flag Values : Actual Value
                                      Date
                                                                                Mann Kendall S Statistic:

_j
B>
o
1
Concen

6.0E-04 -
5.8E-04 -
5.6E-04 -
5.4E-04 •
5.2E-04 •
5.0E-04 -
4.8E-04 -
4.6E-04 -
A AF-flA .
*

• •

* *
* * * * * *

                                                                                Confidence in
                                                                                Trend:
                                                                                     I   72.9%

                                                                               Coefficient of Variation:
                                                                                        0.07
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
W97-18A
W97-18A
W97-18A
W97-18A
W97-18A
W97-18A
W97-18A
W97-18A
W97-18A
W97-18A
W97-18A
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L)
5.6E-04
5.0E-04
5.0E-04
5.0E-04
5.6E-04
5.3E-04
6.0E-04
5.3E-04
5.0E-04
5.0E-04
5.0E-04
Flag

ND
ND





ND
ND
ND
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
0
0
1
1
1
1
1
0
0
0
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                       9/13/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall Statistics  Summary
Well: W97-19A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to  6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
                                      Date

_j
1
o
1
Concen
8.0E-03 •
7.0E-03 -
6.0E-03 -
5.0E-03 •
4.0E-03 •
3.0E-03 -
2.0E-03 -
1.0E-03-
n np4-nn .
^


*

• * * *
* » »
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   94.0%

                                                                               Coefficient of Variation:
                                                                                        0.81
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
                                                                                    [     PD
 Data Table:
Well
W97-19A
W97-19A
W97-19A
W97-19A
W97-19A
W97-19A
W97-19A
W97-19A
W97-19A
W97-19A
W97-19A
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
2.2E-03
7.9E-03
5.4E-03
3.7E-03
1.4E-03
1.2E-03
1.2E-03
2.1E-03
5.0E-04 ND
2.0E-03
2.2E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                       9/13/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: W98-20A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit

J Flag Values : Actual Value
                                      Date
                      «°  ^

O)
o
1
Concent

5.0E-03 -
4.0E-03 -
3.0E-03 •
2.0E-03 -
1.0E-03-
n np4-nn .
* *


•
•
* * *
• • •
•
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   89.1%

                                                                                Coefficient of Variation:
                                                                                         0.74
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
W98-20A
W98-20A
W98-20A
W98-20A
W98-20A
W98-20A
W98-20A
W98-20A
W98-20A
W98-20A
W98-20A
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
2.0E-03
4.8E-03
5.1E-03
1.7E-03
1.0E-03
1.5E-03
1.0E-03
1.0E-03
5.0E-04 ND
1.7E-03
2.3E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                       9/13/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall Statistics  Summary
Well: W98-21A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to  6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
                                      Date

_j
B>
o
1
Concent

7.0E-03 -
6.0E-03 -
5.0E-03 -
4.0E-03 •
3.0E-03 -
2.0E-03 •
1.0E-03-
n np4-nn .
*

*
* *
* *
*
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   94.0%

                                                                               Coefficient of Variation:
                                                                                        0.73
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
                                                                                    [     PD
 Data Table:
Well
W98-21A
W98-21A
W98-21A
W98-21A
W98-21A
W98-21A
W98-21A
W98-21A
W98-21A
W98-21A
W98-21A
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
1.7E-03
7.1E-03
4.9E-03
2.1E-03
2.8E-03
1.9E-03
1.2E-03
2.5E-03
5.0E-04 ND
1.7E-03
1.9E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                       9/13/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall Statistics Summary
Well: W99-R5A
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to  6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Median
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
                                      Date

_j
1
o
1
Concent
4.0E-03 -
3.5E-03 -
3.0E-03 -
2.5E-03 •
2.0E-03 •
1.5E-03-
1.0E-03-
5.0E-04 -
n np4-nn .
*




^ *** *****
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   46.9%

                                                                               Coefficient of Variation:
                                                                                         1.29
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
                                                                                    [     NT
 Data Table:
Well
W99-R5A
W99-R5A
W99-R5A
W99-R5A
W99-R5A
W99-R5A
W99-R5A
W99-R5A
W99-R5A
W99-R5A
W99-R5A
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L)
4.1E-04
4.1E-03
5.0E-04
5.0E-04
5.0E-04
7.0E-04
5.0E-04
5.5E-04
5.0E-04
5.0E-04
5.0E-04
Flag


ND
ND
ND

ND

ND
ND
ND
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
0
0
0
1
0
1
0
0
0
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                       9/13/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
 Project:  Frontier Hard Chrome
 Location:  Vancouver
                User Name: MV
                State:  Washington
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Geometric Mean
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
Source/ Number of
Well Tail Samples
Number of
Detects
Coefficient
of Variation
Mann-Kendall
Statistic
Confidence
in Trend
All
Samples Concentration
"ND" ? Trend
CHROMIUM, TOTAL
RA-MW-15B
B87-8
RA-MW-11B
W92-16B
RA-MW-16B
RA-MW-12B
RA-MW-12C
RA-MW-13B
B85-3
RA-MW-14B
W99-R5B
W85-7B
W97-18B
W97-19B
W98-21B
RA-MW-13C
S
S
S
S
S
T
T
T
T
T
T
T
T
T
T
T
12
11
12
12
12
12
12
12
11
12
11
11
11
11
11
11
12
11
11
12
12
10
12
6
8
9
10
4
8
9
10
10
0.94
1.10
1.62
1.47
1.45
1.09
0.70
1.32
0.80
1.21
0.74
1.58
0.39
0.99
0.62
0.85
4
2
-48
2
8
-28
-4
-13
2
-6
-33
-34
12
-24
-34
-6
58.0%
53.0%
100.0%
52.7%
68.1%
96.9%
58.0%
79.0%
53.0%
63.1%
99.5%
99.6%
79.9%
96.4%
99.6%
64.8%
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
NT
NT
D
NT
NT
D
S
NT
NT
NT
D
D
NT
D
D
S
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A)-
 Due to insufficient Data (< 4 sampling events); Source/Tail (S/T)

     The Number of Samples and Number of Detects shown above are post-consolidation values.
MAROS Version 2,.2 2006, AFCEE
Thursday, September 06, 2007
                                                                                            Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: B85-3
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Geometric Mean
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
                                      Date


j"
1
o
«
>I
c
S
c
o
o



6.0E-03 •
5.0E-03 •
4.0E-03 -

3.0E-03 -

2.0E-03 -


1.0E-03-
n np4-nn .

*
* *

*


*


» ^
* * *
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   53.0%

                                                                                Coefficient of Variation:
                                                                                         0.80
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
                                                                                    [    NT
 Data Table:
Well
B85-3
B85-3
B85-3
B85-3
B85-3
B85-3
B85-3
B85-3
B85-3
B85-3
B85-3
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
5.0E-03
5.0E-04 ND
5.0E-04 ND
1.1E-03
6.3E-03
4.9E-03
5.4E-03
9.0E-04
5.0E-04 ND
2.5E-03
3.6E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
0
0
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: B87-8
Well Type: s
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Geometric Mean
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
                                      Date
                      «°  *s


_J
B)
_£
c
~
k.
C
c
o
o


2.5E-01 -


2.0E-01 -

1.5E-01 •

1.0E-01 -


5.0E-02 -
n np4-nn .


•




•


•
* *
» *****
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   53.0%

                                                                                Coefficient of Variation:
                                                                                         1.10
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
                                                                                    [    NT
 Data Table:
Well
B87-8
B87-8
B87-8
B87-8
B87-8
B87-8
B87-8
B87-8
B87-8
B87-8
B87-8
Well Type
s
s
s
s
s
s
s
s
s
s
s
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
1.8E-02
2.4E-01
7.1E-02
1.9E-02
3.1E-02
5.0E-02
2.2E-02
4.6E-02
3.1E-02
2.0E-02
1.3E-01
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
1
1
1
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-11B
Well Type:  s
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Geometric Mean
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


7nc n9
.UE-U^ •
6.0E-02 •
^j
1 5.0E-02 •
§ 4.0E-02 -
S 3.0E-02 -
c
01
c 2.0E-02 -
o
O
1.0E-02-


Data Table:
Well Well Ty
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
RA-MW-1 1B S
Date
/ /vvvv* /vvv* /,/
^^<^^><^'i
-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-12B
Well Type:  T
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Geometric Mean
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


2ep n9
.OE-U^ •
2- 2.0E-02 -
E
c 1.5E-02-
o
1
•£ 1.0E-02-
c
0 5.0E-03 •
Oncu-nn
m\ICr\I\I
Data Table:
Well Well Ty
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
RA-MW-12B T
Date
/ /vvvv* /v
^* ^*

•


•
•
* * * * *
•


Effective
Pe Date Constituent
10/15/2003 CHROMIUM, TOTAL
2/10/2004 CHROMIUM, TOTAL
4/5/2004 CHROMIUM, TOTAL
8/15/2004 CHROMIUM, TOTAL
5/5/2005 CHROMIUM, TOTAL
12/12/2005 CHROMIUM, TOTAL
3/8/2006 CHROMIUM, TOTAL
6/15/2006 CHROMIUM, TOTAL
9/25/2006 CHROMIUM, TOTAL
12/15/2006 CHROMIUM, TOTAL
3/30/2007 CHROMIUM, TOTAL
6/5/2007 CHROMIUM, TOTAL

vvv\/
% <) <$• ^





* *
*


Result (mg/L) Flag
2.1E-02
7.6E-03
5.0E-04 ND
4.2E-03
4.1E-03
1.1E-02
3.3E-03
2.4E-03
2.4E-03
5.0E-04 ND
3.4E-03
3.0E-03

Mann Kendall S Statistic:
_

Confidence in
Trend:
1 96.9%
Coefficient of Variation:
1 1.09
Mann Kendall
Concentration Trend:
(See Note)
I °

Number of Number of
Samples Detects
2 2
1 1
1 0
1 1
1 1
1 1
1 1
1 1
1 1
1 0
1 1
1 1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                      9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-12C
Well Type:  T
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Geometric Mean
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


1 HP no
I .UC-U^ •
9.0E-03 -
^ 8.0E-03 -
|" 7.0E-03 -
r 6.0E-03 •
o
s 5.0E-03 •
i 4.0E-03 •
| 3.0E-03 •
0 2.0E-03 •
1.0E-03-

Data Table:
Well Well Ty
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
RA-MW-12C T
Date
/ /vvvv* /vi
^*^*
*



•
* *
«


Effective
Pe Date Constituent
10/15/2003 CHROMIUM, TOTAL
2/10/2004 CHROMIUM, TOTAL
4/5/2004 CHROMIUM, TOTAL
8/15/2004 CHROMIUM, TOTAL
5/5/2005 CHROMIUM, TOTAL
12/12/2005 CHROMIUM, TOTAL
3/8/2006 CHROMIUM, TOTAL
6/15/2006 CHROMIUM, TOTAL
9/25/2006 CHROMIUM, TOTAL
12/15/2006 CHROMIUM, TOTAL
3/30/2007 CHROMIUM, TOTAL
6/5/2007 CHROMIUM, TOTAL

e//v\/
' ^



*
•




Result (mg/L) Flag
9.1E-03
2.8E-03
2.7E-03
9.8E-04
4.4E-03
8.7E-03
2.2E-03
6.0E-04
1.5E-03
5.1E-03
5.6E-03
4.2E-03

Mann Kendall S Statistic:
I -4
Confidence in
Trend:
1 58.0%
Coefficient of Variation:
I °7°
Mann Kendall
Concentration Trend:
(See Note)
I S

Number of Number of
Samples Detects
2 2
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                      9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-13B
Well Type:  T
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Geometric Mean
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value

8.0E-03 •
7.0E-03 -
? 6.0E-03 -
~ 5.0E-03 -
| 4.0E-03 •
g 3.0E-03 -
o
o 2.0E-03 •
O
1.0E-03-
Data Table:
Date
d^X/////////.
*

* * *
* * * * » » •



$v Mann Kendall S Statistic:
I -13
Confidence in
Trend:
1 79.0%
Coefficient of Variation:
I 1'32
Mann Kendall
Concentration Trend:
(See Note)
_


Effective Number of Number of
Well Well Type Date Constituent Result (mg/L) Flag Samples Detects
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
RA-MW-13B T
10/15/2003 CHROMIUM, TOTAL 5.0E-04 ND 2 0
2/10/2004 CHROMIUM, TOTAL 2.3E-03 1 1
4/5/2004 CHROMIUM, TOTAL 5.0E-04 ND 1 0
8/15/2004 CHROMIUM, TOTAL 1.3E-03 1 1
5/5/2005 CHROMIUM, TOTAL 7.1E-03 1 1
12/12/2005 CHROMIUM, TOTAL 1.4E-03 1 1
3/8/2006 CHROMIUM, TOTAL 5.0E-04 ND 1 0
6/15/2006 CHROMIUM, TOTAL 7.0E-04 1 1
9/25/2006 CHROMIUM, TOTAL 5.0E-04 ND 1 0
12/15/2006 CHROMIUM, TOTAL 5.0E-04 ND 1 0
3/30/2007 CHROMIUM, TOTAL 1.2E-03 1 1
6/5/2007 CHROMIUM, TOTAL 5.0E-04 ND 1 0
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                      9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: RA-MW-13C
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Geometric Mean
Duplicate Consolidation: Average
ND Values: Specified Detection Limit

J Flag Values : Actual Value
                                      Date


_J
O)

• — '
c
o
1
1
o
o

7.0E-03 -
6.0E-03 -


5.0E-03 -


4.0E-03 •
3.0E-03 -
2.0E-03 •
1.0E-03-
n np4-nn .

*


*



« *

*
* . ' *
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   64.8%

                                                                                Coefficient of Variation:
                                                                                         0.85
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
RA-MW-13C
RA-MW-13C
RA-MW-13C
RA-MW-13C
RA-MW-13C
RA-MW-13C
RA-MW-13C
RA-MW-13C
RA-MW-13C
RA-MW-13C
RA-MW-13C
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
3.7E-03
1.4E-03
6.8E-04
7.3E-03
1.2E-03
1.4E-03
4.1E-03
5.5E-03
5.0E-04 ND
2.2E-03
9.1E-04
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/6/2007
                          Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-14B
Well Type:  T
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Geometric Mean
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value

7.0E-03 -
6.0E-03 •
U
1 5.0E-03 •
§ 4.0E-03 -
S 3.0E-03 -
c
01
c 2.0E-03 -
o
O
1.0E-03-
Data Table:
Date
0oVv
*
*
» * * »**» *



$v Mann Kendall S Statistic:
Confidence in
Trend:
1 63.1%
Coefficient of Variation:
I 1'21
Mann Kendall
Concentration Trend:
(See Note)
_


Effective Number of Number of
Well Well Type Date Constituent Result (mg/L) Flag Samples Detects
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
RA-MW-14B T
10/15/2003 CHROMIUM, TOTAL 5.5E-04 2 1
2/10/2004 CHROMIUM, TOTAL 3.5E-03 1 1
4/5/2004 CHROMIUM, TOTAL 5.0E-04 ND 1 0
8/15/2004 CHROMIUM, TOTAL 8.1E-04 1 1
5/5/2005 CHROMIUM, TOTAL 6.5E-03 1 1
12/12/2005 CHROMIUM, TOTAL 1.5E-03 1 1
3/8/2006 CHROMIUM, TOTAL 5.0E-04 ND 1 0
6/15/2006 CHROMIUM, TOTAL 7.0E-04 1 1
9/25/2006 CHROMIUM, TOTAL 6.4E-04 1 1
12/15/2006 CHROMIUM, TOTAL 5.0E-04 ND 1 0
3/30/2007 CHROMIUM, TOTAL 1.5E-03 1 1
6/5/2007 CHROMIUM, TOTAL 6.6E-04 1 1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                      9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-15B
Well Type:  s
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Geometric Mean
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


2ep ni
.OE-U I •
;;[• 2.0E-01 -
E
c 1.5E-01 •
o
1
•£ 1.0E-01 •
8
c
0 5.0E-02 •
Oncu-nn
.uc^uu
Data Table:
Well Well Ty
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
RA-MW-15B S
Date
o^VVVVVV* «*%
O X T^ i^ N* ^? x* 5 ^


*
•
•
•
•
•


Effective
Pe Date Constituent
10/15/2003 CHROMIUM, TOTAL
2/10/2004 CHROMIUM, TOTAL
4/5/2004 CHROMIUM, TOTAL
8/15/2004 CHROMIUM, TOTAL
5/5/2005 CHROMIUM, TOTAL
12/12/2005 CHROMIUM, TOTAL
3/8/2006 CHROMIUM, TOTAL
6/15/2006 CHROMIUM, TOTAL
9/25/2006 CHROMIUM, TOTAL
12/15/2006 CHROMIUM, TOTAL
3/30/2007 CHROMIUM, TOTAL
6/5/2007 CHROMIUM, TOTAL

>*vv\/
<) <$• ^





. * *


Result (mg/L) Flag
2.0E-02
1.4E-01
5.5E-03
2.2E-03
1.9E-01
1.1E-01
1.9E-01
1.5E-01
3.3E-02
2.1E-02
3.2E-02
4.1E-02

Mann Kendall S Statistic:
I 4
Confidence in
Trend:
1 58.0%
Coefficient of Variation:
1 0.94
Mann Kendall
Concentration Trend:
(See Note)
I NT

Number of Number of
Samples Detects
2 2
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
1 1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                      9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: RA-MW-16B
Well Type:  s
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Geometric Mean
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


2ep ni
.OE-U I •

;;[• 2.0E-01 -
E
c 1.5E-01 •
o
1
•£ 1.0E-01 •
8
c
0 5.0E-02 •

Oncu-nn
.uc^uu
Data Table:

Well Well Ty
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
RA-MW-16B S
Date
/ /vvvv* /vvv* /,/
^^<^^><^'i
-------
 MAROS  Mann-Kendall  Statistics Summary
Well: W98-21B
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Geometric Mean
Duplicate Consolidation: Average
ND Values: Specified Detection Limit

J Flag Values : Actual Value
                                      Date
                                                  c


'a
_§
o
«
>i
§
c
o
o


6.0E-03 •
5.0E-03 •

4.0E-03 -

3.0E-03 -
2.0E-03 -

1.0E-03-
n np4-nn .
^


•

*

* * *

•
*
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   99.6%

                                                                                Coefficient of Variation:
                                                                                         0.62
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
W98-21B
W98-21B
W98-21B
W98-21B
W98-21B
W98-21B
W98-21B
W98-21B
W98-21B
W98-21B
W98-21B
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
3.6E-03
6.6E-03
4.6E-03
2.7E-03
3.2E-03
2.2E-03
1.2E-03
2.2E-03
5.0E-04 ND
1.5E-03
2.2E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: W97-19B
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Geometric Mean
Duplicate Consolidation: Average
ND Values: Specified Detection Limit

J Flag Values : Actual Value
                                      Date


'a
_§
c
o
«
>i
c
S
c
o
O


1.2E-02-
1.0E-02-

8.0E-03 •


6.0E-03 •

4.0E-03 -

2.0E-03 •
n np4-nn .

*








•
* * * * *
• •
                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   96.4%

                                                                                Coefficient of Variation:
                                                                                         0.99
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
W97-19B
W97-19B
W97-19B
W97-19B
W97-19B
W97-19B
W97-19B
W97-19B
W97-19B
W97-19B
W97-19B
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
1.3E-02
5.1E-03
5.1E-03
3.4E-03
5.0E-04 ND
1.8E-03
2.1E-03
2.1E-03
5.0E-04 ND
2.0E-03
2.4E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
0
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: W97-18B
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Geometric Mean
Duplicate Consolidation: Average
ND Values: Specified Detection Limit
J Flag Values : Actual Value
                                      Date
                                                  c


j"
_§
o
«
>I
c
S
c
o
o



1.2E-03-
1.0E-03-

8.0E-04 •

6.0E-04 •

4.0E-04 •


2.0E-04 •
n np4-nn .
• *
*
* 4 ^





» * * *




                                                                                Mann Kendall S Statistic:
                                                                                     I    12
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   79.9%

                                                                                Coefficient of Variation:
                                                                                         0.39
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
                                                                                    [     NT
 Data Table:
Well
W97-18B
W97-18B
W97-18B
W97-18B
W97-18B
W97-18B
W97-18B
W97-18B
W97-18B
W97-18B
W97-18B
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
4.1E-04
5.0E-04 ND
1.1E-03
1.3E-03
1.0E-03
5.0E-04 ND
1.0E-03
1.3E-03
5.0E-04 ND
8.8E-04
1.2E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
0
1
1
1
0
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/6/2007
                         Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics  Summary
Well: W92-16B
Well Type:  s
COC: CHROMIUM, TOTAL
Time Period: 10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type: Geometric Mean
Duplicate Consolidation: Average
ND Values:  Specified Detection Limit
J Flag Values : Actual Value


2ep ni
.OE-U I •

;;[• 2.0E-01 -
E
c 1.5E-01 •
o
1
•£ 1.0E-01 •
8
c
0 5.0E-02 •

Oncu-nn
.uc^uu
Data Table:

Well Well Ty
W92-16B S
W92-16B S
W92-16B S
W92-16B S
W92-16B S
W92-16B S
W92-16B S
W92-16B S
W92-16B S
W92-16B S
W92-16B S
W92-16B S
Date
/ /vvvv* /vvv* /,/
^^<^^><^'i
-------
 MAROS  Mann-Kendall  Statistics Summary
Well: W85-7B
Well Type: T
COC: CHROMIUM, TOTAL
                                                   Time Period: 10/15/2003  to  6/5/2007
                                                   Consolidation Period:  No Time Consolidation
                                                   Consolidation Type: Geometric Mean
                                                   Duplicate Consolidation: Average
                                                   ND Values:  Specified Detection Limit

                                                   J Flag Values : Actual Value
                                      Date
O)

o
1

I
o
o

1.8E-02-
1.6E-02-
1.4E-02-
1.2E-02-
1.0E-02-
8.0E-03 •
6.0E-03 •
4.0E-03 •
2.0E-03 -
n np4-nn .

*



*
•




                                                                                Mann Kendall S Statistic:
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   99.6%

                                                                                Coefficient of Variation:
                                                                                         1.58
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
W85-7B
W85-7B
W85-7B
W85-7B
W85-7B
W85-7B
W85-7B
W85-7B
W85-7B
W85-7B
W85-7B
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L)
1.8E-02
1.1E-02
8.0E-03
8.4E-04
5.0E-04
5.0E-04
5.0E-04
5.0E-04
5.0E-04
5.0E-04
5.0E-04
Flag




ND
ND
ND
ND
ND
ND
ND
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
0
0
0
0
0
0
0
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/6/2007
                                                                            Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: W85-6B
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Geometric Mean
Duplicate Consolidation: Average
ND Values: Specified Detection Limit

J Flag Values : Actual Value
                                      Date
                                                                                Mann Kendall S Statistic:


'a
_§
o
«
>i
§
c
o
o



1.2E-02-
1.0E-02-

8.0E-03 •

6.0E-03 •
4.0E-03 -


2.0E-03 •
n np4-nn .






^
» »

W W
*
•
                                                                                     I    ~23
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   99.1%

                                                                                Coefficient of Variation:
                                                                                         0.79
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
W85-6B
W85-6B
W85-6B
W85-6B
W85-6B
W85-6B
W85-6B
W85-6B
W85-6B
Well Type
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
1.3E-02
4.7E-03
5.6E-03
2.9E-03
4.8E-03
3.8E-03
5.0E-04 ND
2.9E-03
2.0E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/6/2007
                          Page 1 of 1

-------
 MAROS  Mann-Kendall  Statistics Summary
Well: W99-R5B
Well Type: T
COC: CHROMIUM, TOTAL
Time Period:  10/15/2003  to 6/5/2007
Consolidation Period: No Time Consolidation
Consolidation Type:  Geometric Mean
Duplicate Consolidation: Average
ND Values: Specified Detection Limit

J Flag Values : Actual Value
                                      Date
                      «°  *s

B)
o
?
c
o
o

1.0E-02-
8.0E-03 -
6.0E-03 •
4.0E-03 -
2.0E-03 -
n np4-nn .
•
•
•
» •
• . * • *
•
                                                                                Mann Kendall S Statistic:
                                                                                     I    ~33
                                                                                Confidence in
                                                                                Trend:
                                                                                     I   99.5%

                                                                                Coefficient of Variation:
                                                                                         0.74
                                                                                Mann Kendall
                                                                                Concentration Trend:
                                                                                (See Note)
 Data Table:
Well
W99-R5B
W99-R5B
W99-R5B
W99-R5B
W99-R5B
W99-R5B
W99-R5B
W99-R5B
W99-R5B
W99-R5B
W99-R5B
Well Type
T
T
T
T
T
T
T
T
T
T
T
Effective
Date
2/10/2004
4/5/2004
8/15/2004
5/5/2005
12/12/2005
3/8/2006
6/15/2006
9/25/2006
12/15/2006
3/30/2007
6/5/2007
Constituent
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
CHROMIUM, TOTAL
Result (mg/L) Flag
7.5E-03
9.9E-03
4.8E-03
6.7E-03
4.5E-03
1.8E-03
1.4E-03
2.5E-03
5.0E-04 ND
1.9E-03
2.4E-03
Number of
Samples
1
1
1
1
1
1
1
1
1
1
1
Number of
Detects
1
1
1
1
1
1
1
1
0
1
1
 Note: Increasing (I); Probably Increasing (PI); Stable (S); Probably Decreasing (PD); Decreasing (D); No Trend (NT); Not Applicable (N/A) -
 Due to insufficient Data (< 4 sampling events); ND = Non-detect
MAROS Version 2.2, 2006, AFCEE
                                                        9/6/2007
                          Page 1 of 1

-------