Management and Interpretation of Data
Under a Triad Approach - Technology Bulletin
May 2007
INTRODUCTION
Since its inception in 1995, the U.S. Environmental
Protection Agency's (EPA) Brownfields Initiative and
other revitalization efforts have grown into major
national programs that have changed the way
contaminated property is perceived, addressed, and
managed in the United States. In addition, over time,
there has been a shift within EPA and state regulatory
agencies in the way that hazardous waste site
cleanups are managed.
Project managers, regulators, technology providers,
and other stakeholders are increasingly recognizing
the value of implementing a more dynamic approach
to site cleanup that is flexible and focuses on real-
time decision-making in the field to reduce costs,
improve decision certainty, and expedite site
closeout. As shown in Figure 1, the Triad approach
uses (1) systematic project planning, (2) dynamic
work strategies (DWS), and (3) real-time
measurement technologies to reduce decision
uncertainty and increase project efficiency (Source:
EPA 2003).
Real-Time Measurement
Technologies
Figure 1: The Triad Approach.
The approaches used in Triad projects require
specific procedures and tools for data interpretation
and management. For example, technologies such
as open-path air monitoring systems and subsurface
geophysical detection tools can generate thousands
of individual data points that must be assimilated and
manipulated by computer to provide the full benefit of
their real-time imaging capabilities. Fortunately, data
management and decision support tools (DST) have
become more available in recent years, and
experienced Triad practitioners are already exploiting
them (ITRC 2003).
About the Brownfields and Land Revitalization
Technology Support Center (BTSC)
EPA established BTSC (www.brownfieldstsc.om} to
ensure that brownfields and other land revitalization
decision-makers are aware of the full range of
technologies available for conducting site
assessments and cleanups and can make informed
decisions about their sites. The center can help
federal, state, local, and tribal officials evaluate
strategies to streamline the site assessment and
cleanup process at specific sites; identify, review and
communicate information about complex technology
options; evaluate contractor capabilities and
recommendations; and plan technology
demonstrations.
Localities can submit requests for assistance through
their EPA Regional Brownfields Coordinators
(http://www.epa.ciov/swerosps/bf/corcntct.htm} or by
calling 1 -877-838-7220 toll free. For more
information about BTSC, contact Carlos Pachon at
(703) 603-9904 or pachon.carlos@epa.ciov.
The Brownfields and Land Revitalization Technology
Support Center (BTSC) created this bulletin to focus
on implementing a data management program for a
Triad project, and includes:
1. A brief introduction to the Triad approach
2. Answers to frequently asked questions about
data management on Triad projects, such as the
following:
How do Triad practitioners plan for data
management and interpretation?
Who prepares the dynamic work strategy
(DWS) and data management plan, and
what are the essential elements of the data
management plan?
How are data collected and used in a Triad
investigation?
Office of Solid Waste and Emergency
Response (5203P)
EPA 542-F-07-001
May 2007
www.brownfieldstsc.org
-------
Management and Interpretation of Data Under a Triad Approach
What types of tools are used to manage,
interpret, and communicate data?
How do Triad practitioners balance the need
for data review with the need for rapid data
turnaround?
3. Three examples of data management are
addressed, with state agencies as the primary
regulatory body:
Milltown Redevelopment Site, Milltown, New
Jersey
Metal Etching Site, Freeport, New York
Underground Storage Tank (UST) Sites,
South Dakota
4. Sources of additional information for project
teams and stakeholders who develop or provide
input on a data management program in support
of a Triad project
THE TRIAD APPROACH
Triad is a three-pronged approach for improving
decision-making and streamlining environmental
cleanup projects. The Triad approach draws on
advancing science, technology, and practitioner
experience to develop strategies for making site work
more defensible, resource-effective, and responsive
to stakeholder concerns.
Data management is an essential cross-cutting
component to the three elements of the Triad
approach: systematic project planning, dynamic work
strategies, and real-time measurement technologies.
Systematic project planning provides the
overarching framework for project planning,
contracting, and stakeholder communication. The
systematic planning process compels the stakeholder
group (which consists of regulators, landowners, civic
leaders, public participants, and their representatives)
to reach consensus on critical issues, such as
schedule, milestones, data management,
communication, and most importantly exit
strategy as early as possible in the project's life cycle.
Although stakeholder participation is necessary for all
hazardous waste site remediation and closure efforts,
it plays a particularly important role in the Triad
approach. Stakeholder participation is particularly
important because of the Triad's reliance on non-
standard analysis to support real-time decision-
making and its use of dynamic work strategies that
often defer significant sampling program decisions to
the field (Triad Resource Center 2006). Developing
data flow plans, and gaining an understanding of the
data user's needs are essential tasks during
systematic planning.
Dynamic work strategies (DWS) are embodied in
dynamic (flexible) planning documents. The key
aspect of a dynamic work strategy is a defined
decision logic that members of the field team can
follow as they collect and evaluate data. The decision
logic is a set of guidelines or decision rules that
provide a logical framework the project team will use
to make decisions on how, when, where, and why
sampling and analysis will be conducted. The DWS
informs field team members and project stakeholders
on the amount and type of data to be collected initially
and as the investigation progresses. The DWS, with
the decision logic at its core, is incorporated into
project planning documents and may be summarized
in a flow chart or diagram (Figure 2).
The Triad approach differs from conventional
investigations in its emphasis on collecting data using
real-time measurement technologies. These
technologies allow data to be collected or analyzed at
a much greater density or rate than is typically
obtained through a conventional remedial
investigation (Rl) or other site characterization
project. The Triad approach uses real-time
measurement technologies to increase the amount of
data available to support decision-making and to
provide these data to the project team in a matter of
minutes or hours. Thus, decision-making for many
projects can occur in real time while equipment is
mobilized to the site and samples are being collected,
rather than after months of data review, report
preparation, and approval. It is no surprise then, that
management and communication of data is of utmost
importance to the success of a dynamically managed
project.
The primary tool that guides a Triad project is the
conceptual site model (CSM). The CSM is a planning
tool that organizes what is known about a site and
helps the project team identify what more must be
known to make defensible project decisions. The
CSM unifies existing data into a concise description of
a site's physical setting, contaminant release,
transport mechanisms, and exposure points that
describe the problem. The CSM is treated as a
working hypothesis of site conditions that will be
refined and improved over the course of the project.
-------
Management and Interpretation of Data Under a Triad Approach
Horizontal Delineation of Deep Groundwater Plume
1 Drive Membrane interface Probe with direct push to 150 ft
2 Collect direct-push groundwater profiling samples
3, ^ajyzesamples using VOC test kits and fixed-taa
1. Step-out 250 ft down-gradient of criteria-
exceeding sample and advance direct-
push boring to 150 ft. Sample bonng and
groundwater as before.
1. Step-out 250 down-gradient of criteria-
exceeding sample and advance direct-
push boring to 150 ft Sample boring and
groundwater as before.
2 Step out perpendicular to GW flow 50 ft
to either side of ban rig in step (I j above
and sample as above.
---"" Total VOC ~^-
Concentrations 30
"-^-^ toSOQppb? ..,.=-"-'
I No
.^--"" Total VOC """--
Concentrations greater
"----....than 300 ppb?^...---
No
--""" Total VOC "---
Concentrations 30 ppb
^-.^ or less?
'res
Horizontal delineation of deep
plume considered complete.
Technital Contingencies
Groundwater*
plume has moved off site.
Discuss ne*t steps with
\. Project Team
Any samples taken
__^ beyond site boundary?
Install multi-chamber, double-cased well using sonic drilling method if
compliance point is iocated. Collect continuous cores.
Figure 2. Example of a Subset of a Decision-logic Diagram. The decision-logic subset depicted in
these diagrams forms part of a dynamic work strategy (DWS) plan, and was created during
systematic planning. Enormous amounts of data from parallel sampling schemes may be
generated once field work begins, and a data management plan is critical to allow quick
information dissemination and presentation so that the project team can interact and make
decisions in real time.
A well-implemented Triad project must use data
management procedures that promote clear
visualization of the CSM and encourage participation
of all stakeholders in its development and refinement.
Furthermore, data management and collection
procedures should allow for review and refinement of
the CSM in nearly real time. This distinction is a
critical difference between the Triad and more
conventional approaches to site investigation and
poses the central data management challenge for
Triad projects.
Triad practitioners recognize that the uncertainty
inherent in site characterizations results largely from
(1) sampling uncertainties, which are minimized by
high-density, sometimes lower-quality sampling of
heterogeneous media using real-time sampling and
analytical methods, and (2) analytical uncertainties,
which are minimized by higher-quality, lower-density
laboratory analysis. The Triad practitioner combines
both types of data into a collaborative data set that
maximizes the strengths of each data type to
minimize overall uncertainty. Collaborative data sets
combine information from multiple data collection
tools, ranging from test kits and hand-held meters to
survey information and analytical laboratory data. It is
essential that the data management system of a Triad
project provides an efficient means for assembling
and using collaborative data sets.
The Triad practitioner also looks for tools and
techniques that allow the core technical team the
environmental professionals who conduct the project
to share and interpret data as rapidly as possible
without compromising overall data quality. Tools
such as relational databases, data reduction or
visualization software (also called "decision support
tools" [DST]), and project Web sites or file sharing
sites improve the ability of the core technical team to
share data and new insights with project
stakeholders. As a result, all interested parties may
meaningfully participate in refining the CSM and
-------
Management and Interpretation of Data Under a Triad Approach
adjusting the sampling plan. Further information
about the Triad approach is available at
www.triadcentral.orci.
FREQUENTLY ASKED QUESTIONS
Five frequently asked questions are answered below
to help potential Triad practitioners understand how
familiar data management issues are handled under
the Triad approach.
How do Triad practitioners plan for data
management and interpretation?
Triad investigations are dynamic. A DWS is
developed before the project team mobilizes to the
field, and the DWS rather than a static grid-based
sampling plan is used to guide the investigation. The
DWS may identify a number of initial sampling
locations, but subsequent (referred to as "adaptive")
sampling locations are based on predetermined
criteria and are driven by collection and interpretation
of real-time data.
The DWS must have a parallel data management
strategy that allows real-time data interpretation and
decision-making by project stakeholders. For this
reason, it is important for project planning documents
to include a data management plan that contains a
detailed discussion of data management procedures,
equipment (software and hardware), lines of
communication, reporting guidelines, and time
frames. Careful planning further allows the project
team to adequately assess costs and resource needs
associated with data management.
Who prepares the DWS and data management
plan, and what are the essential elements of the
data management plan?
Preparation of the DWS and data management plan
is usually the responsibility of the core technical team;
however, these plans should not be prepared until
stakeholder input and concerns are obtained during a
series of systematic planning meetings. As the Triad
methodology is applied to a broader class of
environmental projects and programs, some may
initially view its principles as conflicting with regulatory
standards and approaches. Thus, it is critical to the
success of a Triad project to encourage the
participation of regulators, landowners, and the public
from the outset and to maintain a spirit of
transparency and cooperation throughout the project.
The Triad approach can incur more costs upfront
because critical issues are addressed at the outset of
the project; however, the initial expenditure should be
more than offset by the savings that will result from
fewer mobilizations and phases of work, with less
conflict and more efficient coordination between
stakeholders later in the project.
A successful data management strategy depends on
input not only from data management specialists but
also from those who will be generating and using the
data, including vendors, geoscientists, chemists, and
other technical specialists. The data management
plan must address how data from different sources
will be integrated to support decisions by considering
the following elements:
Origin of data streams: Data may be
generated by a variety of field or laboratory
instruments, and results may be documented
using various manual and electronic formats.
The quality and format of the various data
streams must be understood so that they may be
merged effectively into a common database.
Data pathways and communication: Project
managers must recognize that a number of key
participants in project decision-making may be
remote from the site during significant periods of
time. The data management plan for the project
should recognize this issue as important and
provide for the rapid turnaround and distribution
of results and data analysis to stakeholders. The
data management plan should also assign
responsibility for all aspects of the management
and security of the data throughout the project
life cycle.
Level and timing of quality assurance and
quality control (QA/QC): Triad-based programs
often include QA/QC components that are either
unique or significantly different in scope and
nature from what a conventional program would
require. Fewer QC requirements are typically
pre-defined for real-time measurement systems
that are deployed in the field. This lack of pre-
defined requirements offers the opportunity to
develop QC protocols that are customized to the
site-specific needs and performance goals of a
project. The concept of customized QC
requirements is consistent with EPA's
-------
Management and Interpretation of Data Under a Triad Approach
Performance Based Measurement System
(PBMS) initiative (http://www.clu-
in.org/search/t.focus/id/195/).
Data analyses (visualization and statistical):
Triad practitioners should have a clear idea of the
statistical analyses, geostatistical programs, and
visualization software that will be appropriate to
interpret and evaluate the data at early planning
stages of the project. Planning for data
evaluation up front assures that the data are
formatted appropriately for the analytical tool and
that the analytical suite is appropriate for the
project's objective and the site's characteristics.
Data stream integration: Care is required if
databases are used to store collaborative data
sets. The two separate data sets should not be
indiscriminately mixed together because they
often will not be statistically comparable. Blind
merging of data sets with different QC protocols,
(such as in statistical programs to calculate
means and standard deviations) should be
avoided, and clear, consistent data identification
protocols should be used to minimize time-
consuming errors in identifying and managing
data.
Correspondence between data and the
decisions they support: It is important to
understand how the data will be used to support
different decisions as the project progresses.
The linkage between data and decisions may be
best summarized in a table or matrix format,
providing a direct means to assess whether data
collection and analysis techniques are being
used appropriately.
Data archive and repository: As part of the
systematic planning process, it is important for
project managers to be aware of the records that
must be included in the project archives and
make sure the mechanisms are in place to
capture the appropriate documentation.
How are data collected and used in a Triad
investigation?
The key to using data to drive decision-making in the
field is to ensure that the data are readily available as
the investigation is being conducted. Real-time data
collection and analysis methods may yield immediate
graphical results. For instance, subsurface
investigations increasingly rely on characterization
tools that use direct-push drilling equipment. One
example is the cone penetrometer test (CPT), which
generates a lithologic profile as a sensor is pushed
through soil. Other tools create profiles of organic
contaminants as the sensors are advanced through
the subsurface environment, such as the membrane
interface probe (MIP) or the laser-induced
fluorescence (LIF) sensor.
Vendors of these services can usually plot the profiles
as the sensors are advanced, giving the field team a
hard-copy graphic depiction of subsurface
characteristics of interest, such as a continuous-
profile CPT tip resistance (indicating soil type),
dynamic pore pressure (indicating permeability), or
MIP response (indicating contamination by volatile
organic compounds [VOCs]). Many vendors can
quickly generate additional visualizations, such as
modeled plan view or three-dimensional plume maps.
The team can select the following day's sampling
locations after they have reviewed these work
products and by following the predetermined decision
logic established during systematic project planning.
Similarly, a variety of instruments and tests kits are
available to rapidly analyze samples for metals,
polycyclic aromatic hydrocarbons (PAH),
polychlorinated biphenyls (PCB), and other classes of
chemicals. Test kit results are usually recorded in
logbooks but may be logged electronically using a
personal data assistant (PDA). Other hand-held
instruments such as the photoionization detector
(PID) and X-ray fluorescence (XRF) detector include
data loggers or more comprehensive data systems
with electronic data porting capabilities. Location
data, such as locations of surface sampling or
boreholes, can be recorded electronically using global
positioning system (GPS) equipment.
Although many decisions can be made using
graphical printouts or data recorded in log books, a
Triad investigation is most efficient when data can be
captured electronically and then downloaded in a
seamless manner to a relational database (a
database where relationships between data items are
organized in tables that can be linked through a
common field). As noted previously, many vendors of
real-time analytical technologies can quickly generate
representations of electronic data in the field.
Alternatively, different types of data can be
transferred to a central on-site or off-site location for
incorporation into the relational database and for
-------
Management and Interpretation of Data Under a Triad Approach
further analysis using widely available software.
Numerical detector responses can be tabulated and
correlated with laboratory data within a relational
database.
What types of tools are used to manage, interpret,
and communicate data?
The streamlined manner used to acquire data during
a Triad project is mirrored by the tools and techniques
used for data management and interpretation. DSTs
are interactive software tools decision-makers use to
answer questions, solve problems, and support or
refute conclusions. They may have a single function,
but more often incorporate multiple functions, such as
data acquisition, spatial data management, modeling,
and cost estimation. DSTs are designed to bring data
management and interpretation capabilities as close
(in both time and space) to data collection activities
and decision-making as possible while still
maintaining the integrity of the work products they
develop.
The DST matrix (http://www.frtr.gov/decisionsupport/)
is a Web-based resource EPA created to assist Triad
practitioners in selecting a DST to meet project
objectives. The matrix provides detailed information
for 25 DSTs. Additional information about the DST
matrix is provided at the end of this bulletin.
Examples of data management, interpretation, and
communication tools are presented below. Links to
specific programs are provided in the DST matrix.
Examples of available tools include:
Scribe (http://www.ertsupport.org/scribe_home.htm) is
a data management DST developed by EPA's
Environmental Response Team (ERT) that allows a
greater number of project teams working at small to
medium sites to realize the benefits of maintaining
data in a relational database. Scribe can import
electronic data, including analytical laboratory results
in electronic data deliverable (EDO) format and
sampling location data such as GPS coordinates.
Scribe can print sample labels and chain-of-custody
documents. Scribe is integrated with a software
extension called "Scriblets" to capture and import
sampling and monitoring data collected using hand-
held PDAs during field work. The ability to sort,
query, post, and plot data in the field frees project
teams from the need to load data in an office setting.
On Triad projects, this ability means that the field
team can conduct an initial evaluation of the data
moments after a measurement is made using a real-
time tool.
Spatial Analysis and Decision Assistance (SADA) is a
package of data analysis tools developed by the
University of Tennessee using grants provided by
EPA. SADA incorporates tools for visualization,
geospatial analysis, statistical analysis, human health
risk assessment, ecological risk assessment,
cost/benefit analysis, and design of sampling plans. It
includes a data import module that allows the user to
incorporate electronic data rapidly, thus making it
ideal for use in the field.
The need to distribute real-time data to geographically
remote stakeholders and members of the project
team has increased the use Q\ project-specific Web
pages. Sample results and maps can be posted to
the Internet almost as soon as they are generated.
Electronic access encourages greater participation
among stakeholder groups, which leads to an
enhanced sense of ownership among group
members.
How do Triad practitioners balance the need for
data review with the need for rapid data
turnaround?
Triad practitioners use a variety of techniques to
maintain data quality while expediting processing,
reviewing, and reporting field data. Before the project
team designs a sampling program around a real-time
sampling and analysis technique, it should be tested
during a demonstration of methods applicability
(DMA). Confirmation is often provided by collecting
collaborative data sets.
A DMA is an initial site-specific performance
evaluation for a method, a series of related methods,
or a data collection tool. Investigations that use the
Triad approach depend more heavily on field-based
analytical equipment and emerging technologies. A
DMA usually consists of using the real-time
technology on a limited suite of samples while
sending paired samples to an off-site laboratory for
standard laboratory analysis for the same analytes or
parameters. The DMA sample results are evaluated
using statistical methods such as regression analysis
and analysis of variance to (1) evaluate the adequacy
-------
Management and Interpretation of Data Under a Triad Approach
of the real-time method for achieving the project's
data quality objectives (DQO), (2) establish
appropriate sampling and analysis procedures, and
(3) develop action levels for the real-time methods.
By conducting the DMA before the field sampling
program, many problems and data "bottlenecks" that
may arise during the sampling program can be
anticipated and corrective measures can be
established proactively.
Making full use of collaborative data sets is another
key to balancing rapid generation of data with
thorough review. As noted previously, collaborative
data sets integrate data obtained rapidly using real-
time techniques with analytical data from an off-site
laboratory. A subset of samples that were analyzed
using real-time measurement technologies may be
split, with the other portion sent to an off-site
laboratory to manage uncertainty in real-time
measurements. These samples are selected based
on their representativeness (already established by
the refined CSM) to support specific decisions that
require more analyte-specific information or more
accurate quantitation (ITRC2003).
The real-time data can be used to support most
aspects of the DWS, as long as the adequacy of the
real-time techniques is established ahead of time
during the DMA. If the DMA is not completed before
the field program begins, quick data reporting
turnaround (24 or 48 hours) is required for a portion of
the samples analyzed off site to allow timely
evaluation of the data adequacy.
Similar to conventional investigations, data review
requirements should be documented in the quality
assurance project plan (QAPP). It is important that
the provisions of the QAPP be carefully followed
during the investigation (Triad Resource Center
2006).
EXAMPLES OF DATA MANAGEMENT AND
INTERPRETATION UNDER THE TRIAD
APPROACH
The three examples below illustrate the use of
procedures and tools for data management and
interpretation for Triad projects.
Example 1: Milltown Redevelopment Site,
Milltown, New Jersey
Milltown Redevelopment is a Brownfields site in
Milltown, New Jersey. A large, diverse stakeholder
group was formed to plan, conduct, and oversee the
project. The stakeholder group included
representatives from federal, state, and local
agencies, as well as a potential site developer.
The primary project goals were to (1) develop and
revise a CSM of site geology, hydrology, and
contaminant fate and transport by collecting soil,
sediment, and ground water samples and geologic
logs on a regular grid, and (2) delineate potential
areas of concern (AOC) on a closer grid spacing of
adaptive sampling locations.
Site Facts
The 22-acre Brownfields site is in the heart of
downtown Milltown, New Jersey. More than 50
percent of the site is covered with warehouses
and industrial buildings.
Industrial use of the site began with a rubber
manufacturing plant in the late 1800s, succeeded
by numerous other industries.
The Milltown Redevelopment Authority and
Middlesex County Improvement Agency entered
into an agreement with a developer to improve
the parcel for mixed uses, including more than
300 age-restricted residential units, commercial
space, and open space along Mill Pond, the main
waterway through central Milltown.
Principal contaminants of concern (COC)
included VOCs (particularly chlorobenzene),
PAHs, PCBs, pesticides, and total petroleum
hydrocarbons (TPH).
The stakeholder group identified several concerns
through a series of systematic planning meetings,
including the management of the large volume of data
that would be generated by field analytical methods
over a relatively short, 2-month time frame (Mack
2006a). EPA's Brownfields Technical Support Center
(BTSC) identified Scribe and Scriblets (see "Scribe"
on page 6) as DSTs that would meet the project
team's needs for data management (Figure 3) and
address the stakeholder concerns. Another outcome
of systematic planning was the development of a
DWS for selecting successive rounds of sampling
locations in real time, thus allowing rapid delineation
of potential AOCs.
-------
Management and Interpretation of Data Under a Triad Approach
Milltown Redevelopment Project
Data Storage, Retrieval & Display
GPS Survey Dpta |
\
Updatt to Wtbsite for
StiNt holdtr Rev it w
J
PDA Sampling ^
info
*| Re»uit« to Fiold Plotter |
/ \
j Chains to Lab
Source: Mack
1 Results from Lab |
Figure 3: Generalized Schematic Diagram of
Data Pathways. At Milltown, Scribe was used as
the central data management software, with
Scribblets used on PDAs to feed field data to
Scribe. This allowed the project team to make
real-time decisions during the sampling
mobilization.
Data Management Tools and Procedures
Scriblets allows the user to set up a sampling
template before field work begins; a field technician
uses the template to enter data electronically as they
are generated in the field. A sample numbering
sequence was assigned in Scriblets before sampling
began, providing sample numbers automatically as
the sampling event progressed. Sample data, real-
time analytical results, and sampling location data
(coordinates generated using GPS survey equipment)
were entered into Scriblets in the field and then
imported to the Scribe database by connecting the
PDA to a laptop computer with a USB cable. The
result was an all-electronic data pathway that
minimized the potential for transcription errors.
Scribe facilitated "in the trailer" review of data and
enabled the project team to plan the next day's
sampling locations before team members left the site
for the day. After each round of sampling, the data
were imported to Scribe, reviewed, queried, and
exported to AutoCAD (visualization software) through
an electronic data format the project team created in
Scribe. Most importantly, Scribe made available in
real time many of the features and advantages of a
relational database while the team was still in the
field.
Data Interpretation Tools and Procedures
The DWS anticipated the need for rapid delineation of
AOCs, and so it provided decision logic diagrams in
the work plan. Decision logic diagrams facilitate
delineation of contaminated zones in the field
because they provide general rules, rather than
predetermined sample locations. (An example of a
decision rule would be, "step out 5 feet in each
cardinal direction from each grid location where the
field instrument's action level was exceeded and
collect an additional instrument reading at each new
location.") Scribe provided quick turnaround of
preliminary sample results; as a result, optimal
adaptive sampling locations were identified shortly
after sample results collected previously were
reviewed in Scribe and plotted in AutoCAD. Scribe's
data querying capabilities allowed sample results to
be sorted by analyte or depth and plotted with any
associated data item (such as data qualifier) that was
entered to the database through Scriblets, while still
assuring the relational integrity of the data were
maintained. In this manner, the ground water plume
delineation was expedited, and the CSM was
developed to explain the likely source of the plume
and mechanisms that contributed to its spread.
Data Communication Tools and Procedures
A project-specific Web site was set up on the EPA
ERT's Web server. Passwords were issued to all
stakeholders to provide access to the Web site.
Maps were posted to the Web site daily, along with
progress reports and information about meeting times
and places. The Web site allowed stakeholders,
including any who were not present at the site, to
review tables and maps summarizing the project
team's updated understanding of the distribution of
COCs at the end of each day and then provide input
for the next day's sampling.
Project Results
The field team was able to sample more than 400
locations in approximately 5 weeks. The field team
collected more than 130 ground water samples and
600 soil samples, generating 30,000 analytical results
that were loaded into the database and that
underwent QC review in a short time.
-------
Management and Interpretation of Data Under a Triad Approach
The benefits of streamlined data acquisition and
processing using Scribe were evident during the
dynamic investigation of the chlorobenzene plume.
During a site walk-through immediately preceding the
Triad investigation, a vat was discovered under a
formerly used loading dock that had been obscured
by heavy brush. A sample from the vat verified the
presence of chlorobenzene. Although the vat would
likely have been discovered during a conventional
study, the Triad DWS provided a flexible means to
adapt the sampling strategy immediately after the
discovery was made in the field, without modifying the
written plans. The DWS provided clear direction to
the sampling teams and the means for quick
concurrence on the sampling strategy from project
stakeholders. The plume was delineated in
approximately 4 days after 63 ground water and 28
soil samples were collected from 46 sampling
locations (shown as black dots on Figure 4).
Points of Contact
James Mack
New Jersey Institute of Technology
138 Warren Street
Newark, NJ 07102
Telephone: (973) 596-5857
E-mail: mack@adm.niit.edu
Ms. Denise Nickel
Senior Project Manager
Middlesex County Improvement Authority (MCIA)
101 Interchange Plaza
Cranbury, NJ08512
Telephone: (609) 409-5002
E-mail: DRN@mciauth. com
Example 2: Metal Etching Site, Freeport, New
York
The Metal Etching Site is a New York State
Superfund site in Freeport, Nassau County, New
York, adjacent to Freeport Creek. The New York
State Department of Environmental Conservation
(NYSDEC) conducted an Rl of the site to identify and
delineate contamination in surface and subsurface
soil, sediment, soil gas; source areas; ground water
contamination; and impacts to surface water.
Site Facts
The 2-acre site is currently used for commercial
boat storage, sales, and maintenance.
The site operated from 1966 through 1999, and
products were printed or etched using
anodization, chromate conversion, and chrome
or nickel plating processes.
The primary method for disposal of sanitary and
industrial wastewater at the site was through
sanitary sewer lines. AOCs identified at the site
included the former plating area, chemical and
waste storage areas, wastewater treatment
operations, failed sewer lines and connections,
waste storage areas, and historical spill areas.
Site-related COCs included chromium, cadmium,
nickel, and chlorinated hydrocarbons such as
tetrachloroethene (PCE), and trichloroethene
(TCE).
NYSDEC required that the Triad approach be used
for the Rl. The goal of using the Triad approach was
to complete the Rl during one mobilization effort,
thereby decreasing time and costs for the
investigation, while increasing confidence in project
decisions through a greater sampling density.
Data Acquisition Tools and Procedures
Surface soil, subsurface soil, and ground water
samples were collected on a grid with a borehole
spacing of 70 feet throughout much of the site, with a
more focused grid (20-foot spacing) in areas of
particular interest. The DWS called for adaptive
samples to be collected at two new locations offset 10
feet from areas where contaminants were detected at
concentrations above cleanup goals.
The team evaluated the use of field analytical
methods (XRF for metals and a field gas
chromatograph [GC] for VOCs) but decided against
them because of concerns regarding the high number
of confirmation samples that would be required to
collaborate the XRF with the field GC measurements.
Furthermore, elevated detection limits would restrict
the usefulness of the XRF to sources and other high-
concentration areas. In the case of VOC analysis, the
sample throughput for the field GC would be too low
to support the cost of the instrument and analyst.
Instead, a nearby laboratory was used to analyze
samples using modified EPA SW-846 methods, which
-------
Management and Interpretation of Data Under a Triad Approach
required less QA/QC and could be completed
overnight. Samples were delivered to the laboratory
by noon of each day, and the results were returned to
the field team by 11:00 a.m. the next day, thus
providing data almost as quickly as on-site analytical
methods.
A collaborative data set was collected to support the
objectives of the project; real-time data was combined
with data collected using standard methods and each
type of data had a specific purpose. The advantage
gained by collecting data in real time was in obtaining
the most complete "snapshot" of contaminant
distribution that time and budget would allow.
Confirmation sampling, using standard (SW-846)
laboratory methods, indicated that the real-time data
for soil and ground water could be used to prepare
contaminant distribution plots and plume maps, which
were used to refine the CSM in real time and select
locations for ground water wells. The confirmation
sampling indicated real-time ground water data did
not meet all of the DQOs for making decisions about
the need for remediation; therefore, ground water
monitoring wells were installed and samples collected
for analysis using EPA SW-846 methods. The
locations of the wells, however, were based on
contaminant distributions plotted from the real-time
data.
Data Management Tools and Procedures
A relational database was constructed using
geographical information system (GIS) software. The
database contained chemical data, water levels, and
survey data. Tables presenting the analytical results
as well as maps of contaminant distribution and
ground water flow were all prepared from the
database using the GIS software interface. Because
field personnel obtained electronic data deliverables
(EDDs) from the laboratory daily, updated
contaminant distribution maps were quickly prepared
and the field program was redirected as indicated by
the DWS.
The GIS software was used to automatically integrate
the collaborative data set, which included real-time
and compliance data. Real-time data were used to
delineate the extent the media were contaminated
and to develop the CSM; compliance data were used
to quantify the level of contamination to facilitate
comparison with regulatory standards and guidance
values.
As discussed above, the database queries were used
to prepare graphical representations of contaminant
distributions, which were used in the Rl report to
identify areas that would require remediation. Ground
water flow and the impacts of tidal fluctuations were
evaluated using spreadsheets containing data
exported from the database.
Data Communication Tools and Procedures
Data were shared with the stakeholders through e-
mail messages. Contaminant distribution maps and
tables were prepared in .pdf format and transmitted in
the e-mail messages.
Project Results
The real-time analysis required by the Triad approach
(daily redox potential [Eh] measurements using a
downhole Eh sensor) quickly revealed that a strongly
reducing environment predominated the site
subsurface. The reducing environment had
immobilized metals, and metal contamination was
limited to former waste storage and disposal areas.
Furthermore, it was established that tidal recharge
promoted biodegradation of the VOC plume.
Daily analysis permitted early elimination of
hexavalent chromium as a COC, which resulted in
significant cost savings because ground water
samples did not require analysis for hexavalent
chromium. The real-time data defined a plume of
chlorinated VOCs extending across the site, with the
reducing environment in the shallow aquifer
effectively degrading PCE to TCE to cis-1,2-
dichloroethene and finally to vinyl chloride (Figure 4).
The chlorinated solvents in the deeper portions of the
aquifer, however, did not appear to be degrading
under existing conditions.
Real-time analysis coupled with the GIS software
permitted stakeholder input in redirecting sample
collection to more accurately define the affected
areas. Without real-time analysis and visualization,
the chlorinated VOC plume may have been
incompletely characterized because installation of the
monitoring well network would have been based more
on ground water flow than on chemistry. Finally, the
10
-------
Management and Interpretation of Data Under a Triad Approach
GRAPHIC SCALE
Soil Boring Location
- Concentration of VC
Monitoring Wall Location
- Concentration of VC In groundwater In ug/L
i groundwater In ug/L
VINYL CHLORIDE (VC) CONCENTRATIONS
IN GROUNDWATER (UG/L)
METAL ETCHING SITE, FREEPORT, NY
Figure 4: Vinyl Chloride contaminant distribution map, Metal Etching site, Freeport, NY. For
rapid, economic results, an off-site laboratory returned analytical results by 11 AM the day after
samples were taken. The project team utilized GIS software to visualize data and direct
sampling efforts to accurately define affected areas. The team communicated the contaminant
distribution maps and tables to all stakeholders via e-mail. Source: Shkuda and others 2006
identification of chlorinated VOCs as COCs allowed
the timely delineation of a soil vapor plume and
identified the potential for soil vapor intrusion.
Points of Contact
Kevin Carpenter
New York State Department of Environmental
Conservation
625 Broadway
Albany, NY 12233-7020
Telephone: (518)402-9553
E-mail: kjcarpen@gw.dec.state.ny.us
Gregory K. Shkuda, Ph.D.
Environmental Resources Management, Inc.
520 Broad Hollow Road
Melville, NY 11747
Telephone: (631) 756-8947
E-mail: Greg.Shkuda@erm.com
Example 3: "Legacy" UST Sites, South Dakota
The South Dakota Petroleum Release Compensation
Fund (PRCF) conducted a pilot program in the fall of
2004 to evaluate the effectiveness of the Triad
approach to expedite closure of "legacy" UST sites.
Specifically, the PRCF wanted to evaluate whether
the Triad approach could be used to reduce decision
uncertainties associated with petroleum release sites
on a real-time basis.
Site Facts
Five sites were chosen for the study, including
three active gas stations, one closed gas station,
and a railroad fueling site.
The time since the petroleum releases had been
first discovered on the sites ranged from 1 to 14
years.
All the sites were considered "legacy" sites
because, although the petroleum releases had
been known about for years, none of the sites
was moving toward regulatory closure.
Three of the sites had been previously assessed
at costs ranging from $35,000 to $103,000
(including field work and report preparation), but
the results did little to reduce decision
uncertainty.
11
-------
Management and Interpretation of Data Under a Triad Approach
COCs included benzene-toluene-ethylbenzene-
xylene (BTEX), methyl tert butyl ether (MTBE),
diesel fuel, gasoline, ethylene dibromide (EDB),
naphthalene, terephthalic acid (TPA), and total
petroleum hydrocarbons (TPH).
The Triad pilot program sought to reduce decision
uncertainty by increasing data density through the
use of the MIP as the primary investigation tool. The
Ml P is a direct-sensing tool that enables investigators
to obtain rapid, continuous profiles of subsurface
characteristics and display logs of results in real time.
These capabilities promote a fast, iterative
investigation process that minimizes mobilizations
and ideally can collapse multiple phases of
assessment under a conventional approach into a
single event. These events can generate huge
amounts of data, and require a pre-plan to organize
and utilize the data in real time. The PRCF estimates
that a UST site assessment using a conventional
approach would require 12 to 18 weeks from the time
the samples were sent to an off-site lab to when
stakeholders are notified additional information is
needed (South Dakota PRCF, 2005). A primary goal
of this pilot program was to shorten that time frame to
a few days.
Legacy Data and Systematic Planning
A systematic planning meeting presided over by an
independent facilitator was held before the beginning
of the field program and was essential to the project's
success. Existing assessment data were made
available to all stakeholders at the meeting, along
with three-dimensional models of each site's
subsurface geology and chemistry. This process
provided all stakeholders with a single representation
of the preliminary CSM of existing conditions to
promote a common understanding at the outset of the
problems to be solved. The stakeholders were able
to establish objectives and ground rules (such as the
required participation of all project team members at
the site every day) and to develop a generic protocol
for use at each site to obtain data. All data that were
needed to comply with South Dakota Department of
Environment and Natural Resources (DENR) Tier 1
petroleum release reporting requirements were
acquired within 4 days per site.
Data Acquisition Tools and Procedures
Confirmation samples (including VOC contamination
in both soil and ground water matrices because MIP
sensors are not matrix-specific) were collected from
all sites to establish the site-specific sensitivity of the
MIP sensors to fuel hydrocarbons. The development
of this collaborative data set mitigated regulatory
concerns that decisions were being made based on
data that had not been analyzed at a fixed laboratory.
Logs of sensor response were printed immediately
after the MIP profile was conducted. In many cases,
the log itself can be used to identify contaminated
intervals and target depths for confirmation soil
sample collection and to select locations and depths
for subsequent borings. Direct-sensing methods are
ideal for use in conjunction with a DWS and usually
allow for new sample or boring locations to be
selected in real time as the work progresses. The
MIP borehole location was established on the site
coordinate system within minutes after it was
completed using GPS survey equipment.
Data Interpretation and Communication Tools and
Procedures
The project team uploaded the MIP results and the
GPS survey data to the vendor's proprietary data
processing and contouring program twice daily to
update the existing three-dimensional model of the
site. On-site, three-dimensional renderings of the
geologic structure and the plume extents (Figure 5)
allowed the project team to identify data gaps
immediately and fill them before the team demobilized
from the site.
Each morning, results from the previous day were
evaluated by team members and compared to project
objectives. These results were also uploaded to a
project Web site daily. The Web site allowed project
team members who could not be present in the field
to fully participate in the ongoing evaluation of data.
The quick data turnaround time enabled the project
team to follow through on one of the ground rules
established from the start of the project - that all
parties be satisfied that the objectives were met
before the team demobilized from the site.
12
-------
Management and Interpretation of Data Under a Triad Approach
Figure 5. Example of a rendering of petroleum
plumes and high conductivity values, for the
South Dakota LIST Sites. The Project team fed
newly-acquired data into a 3-D visualization
program in order to make decisions about where
sample next. Source: PRCF 2005
Project Results
Throughout the course of the project, 133 borings
were advanced, generating almost 350,000 MIP data
points. In contrast, only about 1,400 data points
would have been generated if conventional drilling
methods had been used at the same number of
boreholes (assuming one PID reading for every 2.5
feet of core and one discrete sample per borehole).
The greater data density not only reduced the
potential for data gaps, but three previously unknown
USTs were found during the project. All three USTs
were believed to have contributed to the plumes, and
have been removed.
Stakeholders met after the assessment to select the
next course of action. Additional corrective action is
planned for two of the sites, consisting of monitored
natural attenuation, and it is expected that another
site will be eligible for a no further action (NFA)
determination.
The possible NFA determination at one site highlights
the value of increased data density and the benefits
of continuous evolution of the CSM. The ground
water plume was believed to be located within a
source water protection zone and therefore needed to
meet EPA maximum contaminant levels (MCL).
High-density geologic delineation using the soil
conductivity probe revealed that the plume lay within
a geologic unit that was not hydraulically connected to
the aquifer. Thus, MCLs did not apply; depending on
the results of a soil vapor survey, the site may be
closed with an NFA determination.
The costs of the Triad investigations at the two sites
that had not been previously investigated were
$25,000 and $32,000. The costs of the Triad
investigation are difficult to compare with the
conventional investigations at the other three sites
because the Triad investigations built on knowledge
gained from the conventional assessments.
However, the conventional assessments left the
investigators with uncertainty regarding the CSM,
while the Triad assessments tended to remove those
uncertainties. More information about the costs and
results of all five assessments can be obtained from
the Triad profile for this site (Triad Resource Center
2006).
Points of Contact
Dennis Rounds
South Dakota PCRF
445 East Capitol Avenue
Pierre, SD 57501
Telephone: (605) 773-3769
E-mail: dennis.rounds@state.sd.us
John Sohl
Columbia Technologies, LLC
1448 South Rolling Road
Baltimore, MD 21227
Telephone: (410)536-9911
Web: http://www.columbiadata.com
SOURCES OF ADDITIONAL INFORMATION
The Triad approach is gaining ever greater
acceptance by EPA and other federal and state
agencies, as well as professional and industrial
organizations. Communities and project teams
interested in implementing the Triad approach are
encouraged to contact the BTSC for more information
and for successful examples of Triad applications.
More detailed information about data management
and interpretation and on the Triad approach is
available in the Brownfields Technology Primer Series
document titled "Using the Triad Approach to
Streamline Brownfields Site Assessment and
Cleanup," which is available at
http://www.brownfieldstsc.org.
13
-------
Management and Interpretation of Data Under a Triad Approach
Project profiles, case studies, and other information
on applying the Triad approach are available at
http://www.triadcentral.org. As additional bulletins
about other aspects of the Triad approach are
developed, the BTSC will make them made available
through these Web sites.
The DWS is the element of the Triad approach that
provides the mechanism for making real-time
decisions in the field. It consists of stakeholder-
approved decision trees and decision logic that are
tied to the CSM. A guide to developing a DWS is
provided at:
http://www.triadcentral.om/tech/dsp_sub.cfm?id=23
Further information about DSTs is provided at
http://www.frtr.gov/decisionsupport/index.htm. This
Web site provides information on 25 DSTs in a matrix
format. The Web site describes the DSTs, input and
output file requirements, and user comments and
provides a link to a software home page providing a
free copy of the DST.
DMAs can be critical for a better understanding and
management of decision-related uncertainty. DMAs
are discussed in technical bulletins on the Triad
approach available from the Triad Resource Center:
http://www.triadcentral.org/tech/dsp sub.cfm?id=4.
Performance-based Management System (PBMS)
provides the site investigator, regulators, and
stakeholders the leeway to adjust method
specifications (including QC) to address site-specific
needs and issues. A discussion of how PBMS may
be integrated with the Triad systematic planning
process is provided at:
http ://www.cl u -i n .org/search/t .f ocus/id/195/
REFERENCES
Interstate Technology and Regulatory Council (ITRC).
2003. Technical and Regulatory Guidance for
the Triad Approach: A New Paradigm for
Environmental Project Management. Prepared
by the ITRC Sampling, Characterization and
Monitoring Team. December.
Mack, J. 2006a. Milltown Redevelopment Project,
Applying Triad Under a State Regulatory
Program. New Jersey Institute of Technology.
Mack, J. 2006b. Milltown Redevelopment Project
Stage 2 SI/RI Report. New Jersey Institute of
Technology and Najarian Associates. April 2006.
Shkuda, G.K., and others. 2006. "A Simple Twist on
the Triad Approach." 1st International
Conference on Challenges in Site Remediation,
Chicago, Illinois, October 2005.
South Dakota Petroleum Release Compensation
Fund(PRCF). 2005. "A Study of Managing
Decision Uncertainties Using the Triad
Approach." June 15.
Triad Resource Center. 2006. "Triad Resource
Center." Accessed in May. On-line Address:
http://www.triadcentral.org/index.cfm
U.S. Environmental Protection Agency (EPA). 2003.
"Using the Triad Approach to Streamline
Brownfields Site Assessment and Cleanup."
Brownfields Technology Primer Series. EPA
542-B-03-002. June.
EPA. 2005. "Case Study for the Use of a Decision
Support Tool: Using SCRIBE to Manage Data
during a Triad Investigation, Milltown
Redevelopment Site, Milltown, New Jersey."
Case study available through
http://www.frtr.gov/decisionsupport/index.htm.
NOTICE AND DISCLAIMER
This bulletin was prepared by EPA's Office of Solid
Waste and Emergency Response under EPA
Contract No. 68-W-02-034. The information in this
bulletin is not intended to revise or update EPA policy
or guidance on how to investigate or clean up
Brownfields or other revitalization sites. Mention of
trade names or commercial products does not
constitute endorsement or recommendation for use.
This bulletin can be downloaded from EPA's
Brownfields and Land Revitalization Technology
Support Center at www.brownfieldstsc.org. For
further information about this bulletin or about the
Triad approach, please contact Dan Powell of EPA at
(703) 603-7196 or powell.dan@epa.gov.
14
------- |