542R05007
Sensor Technologies Used During
Site Remediation Activities -
Selected Experiences

-------
                               Solid Waste                           EPA 542-R-05-007
                               and Emergency Response                 September 2005
                               (5102G)                              www.clu-in.org
  Sensor Technologies Used During Site Remediation Activities -
                            Selected Experiences
                           Internet Address (URL) • www.epa.gov/
Recycled/Recyclable • Printed with Vegetable Oil Based Inks on Process Chlorine Free Recycled Paper (minimum 50% Postconsumer)

-------
                             TABLE OF CONTENTS

Section                                                                       Page

ACRONYMS AND ABBREVIATIONS	ii
NOTICE AND DISCLAIMER	iv
1.0    INTRODUCTION	1
      1.1    What are Sensors?	1
      1.2    Purpose of Report	1
      1.3    Methodology	2
2.0    OVERVIEW OF SENSORY SYSTEMS	3
      2.1    Characterization or Monitoring	3
      2.2    Automation	4
3.0    EXAMPLES OF REMEDIATION SITES THAT HAVE USED SENSOR
      TECHNOLOGIES	5
      3.1    Site Characterization	5
      3.2    Monitoring	5
      3.3    Automation	5
4.0    REFERENCES	9
Appendices

1     Sensor Technology Case Study - Use of Membrane Interface Probe Technology for
      Detection of VOCs at the Sol Lynn/Industrial Transformer Superfund Site
2     Sensor Technology Case Study - 2D-Recon and EOL Geophysical Survey Techniques
      for Characterizing Hydrocarbon-Contaminated Soils at the Hotel Pier Site
3     Sensor Technology Case Study - Use of Capacitance Probes to Measure Soil Moisture at
      the Badger Army Ammunitions Plant
4     Sensor Technology Case Study - Use of In-Situ Sensors to Monitor Ground Water
      Velocity at the China Lake Naval Air Weapons System Site
5     Automation Technology Case Study - Programmable Logic Controllers and Ozone
      Analyzers at the Moffett Federal Airfield Site
6     Sensor Technology Case Study - Automated Sampling and Analysis of Trichloroethene
      and Hexavalent Chromium Using the Burge System at the North Indian Bend Wash and
      Nevada Test Sites
7     Automation Technology Case Study - Supervisory Control and Data Acquisition Using
      Programmable Logic Controllers at the Sprague Road Superfund Site
                                LIST OF TABLES

Table                                                                        Page

1     Selected Case Studies on Sensor Technologies	6

-------
                           ACRONYMS AND ABBREVIATIONS

1, 2-DCE       1, 2-dichloroethene
2D-Recon      Two-dimensional gradiometer
ASTM         American Society of Testing and Materials
BAAP         Badger Army Ammunition Plant
Bgs            Below ground surface
BRL           Basic relay logic
BTEX         Benzene, toluene, ethyl benzene, and total xylenes
COC           Contaminant of concern
CPU           Central Processing Unit
CPU           Central Processing Unit
Cr-VI          Hexavalent chromium
DA            Data acquisition
DBG           Deterrent Burning Ground
DELCD        Dry electrolytic conductivity detectors
EC            Electrical conductivity
BCD           Electron capture detector
EOL           Electromagnetic offset log
EPA           U.S. Environmental Protection Agency
FID            Flame ionization detector
ft/day          Feet per day
GC            Gas chromatograph
HMI           Human machine interface
HOA           Hand-off-auto
Hz            Hertz
IrriMAX       Vendor-supplied standard calibration model
LED           Light emitting diode
LPZ           Low permeability zones
MFA           Moffett Federal Airfield
MIP           Membrane interface probe
mL/min        Milliliters per minute
mV            Millvolts
NELP         Navy Environmental Leadership Program
NIBW         North Indian Bend Wash
NPDES        National Pollutant Discharge Elimination System
NTS           Nevada Test Site
°C             Degrees Centigrade
Ogden         Ogden Environmental and Energy Services Co., Inc.
OSHA         Occupational safety and health administration
PC            Anywhere (communications software)
PC            Personal computer
PID            Photoionization detector
PLC           Programmable logic controller
PLC           Programmable logic controllers
ppb            Parts per billion
R&D           Research and development
                                          11

-------
RI
SCADA
SDI
Sol Lynn
Sprague Road
SRI
TCE
THM
USDA
UST
UV
VC
VECTOR
VOC
WBZ
WBZ
Remedial investigation
Supervisory control and data acquisition
Serial data interface
Sol Lynn/Industrial Transformer Superfund
Sprague Road Ground Water Plume Superfund
Supplemental remedial investigation
Trichloroethene
Trihalomethanes
U.S. Department of Agriculture
Underground storage tank
Ultraviolet light
Vinyl chloride
Variably Emitting Controlled Thermal Output Recorder
Volatile organic compounds.
Water bearing zones
Water bearing zones
                                          111

-------
                                NOTICE AND DISCLAIMER

Preparation of this report has been funded by the U.S. Environmental Protection Agency (EPA)
Technology Innovation and Field Services Division under EPA Contract Number 68-W-02-034.
This document represents the views of the authors.  However, this document has  undergone EPA
and external review by experts in the field.

A limited number of printed copies of the report are available free of charge and may be ordered
via the Web site, by mail, or by fax from the following source:

EPA/National Service Center for Environmental Publications
P.O. Box 42419
Cincinnati, OH  45242-2419
Telephone: (513) 489-8190 or (800) 490-9198
Fax: (513)489-8695

This document is not U.S. EPA policy, guidance or regulation. It does not create or impose any
legally binding requirements.  The information is not intended, nor can it be relied upon, to
create any rights enforceable by any party in litigation with the United States or any other party.
The information provided may be revised periodically without public notice. Use or mention of
trade names does not constitute endorsement or recommendation for use. Standards of Ethical
Conduct do not permit EPA to endorse any private sector product or service.

For further information about this report, please contact the EPA's Office of Superfund
Remediation and Technology Innovation:

Ellen Rubin
(703) 603-0141
rubin. ellen@epa.gov
                                          IV

-------
                                   1.0     INTRODUCTION

1.1     WHAT ARE SENSORS?

A sensor is a device that produces a discernable response to external stimulus. Some examples
of sensors are thermometers, photoelectric cells, pressure transducers, and smoke detectors.
Electronic sensors respond to stimulus by producing standardized electrical signals. This enables
them to interface with devices that display a readable output or larger systems providing sensory
input to a decision-making device. For example, sensors may be used inside a storage tank to
supply information on fluid levels to a system controller who would in turn use this information
to make decisions on starting or shutting down pumps that fill or drain the tank.  Sensors can be
used in environmental remediation for the following activities:

       •   Characterization
       •   Monitoring
       •   Automation

When properly applied, sensors can provide long-term benefits for remediation projects by
reducing manpower requirements, reducing analytical costs, and generating information that
facilitates process optimization.

1.2     PURPOSE OF REPORT

Environmental remediation includes many activities that require measurement and monitoring of
parameters such as contaminant concentrations, media characteristics, and systemic parameters.
In recent years, there has been an increase in the number and types of sensor technologies used
during site remediation. These include technologies that are used for performing real-time and
continuous measurements, remote monitoring, remote operation, and system automation.

The U.S. Environmental Protection Agency (EPA) prepared this report to provide an overview of
several types of sensor technologies and a summary of selected experiences with using the
technologies during site remediation activities. The report highlights the applications,
implementation, strengths and limitations, and lessons learned from actual projects that have
used one or more sensor technologies as part of an overall site remediation strategy. Appendices
one through seven provide case studies for specific sites that have used sensor technologies
during site remediation activities.

This report does not provide guidance on the selection of a specific type or vendor of sensor
technology; these technologies are most cost-effective under specific environmental, chemical,
and physical conditions.  Numerous site-specific considerations, such as site geology, soil, and
aquifer characteristics, chemical, physical, and biological parameters of affected media, and
chemicals of concern, among many others, can impact the overall cost-effectiveness of a system.

-------
1.3    METHODOLOGY

In preparing this report, EPA collected available information on sensor technologies for remedial
projects performed at Superfund sites, federal/military sites, and other sites.  EPA attempted to
compile information that was readily available and current for each project as of Summer/Fall
2004, however, in some cases, EPA was not able to confirm the available information. Some
case studies include information provided primarily by the technology vendor, with limited input
from a regulatory authority. In addition, for many of the projects, there were gaps in the types of
information available (e.g., for some sites, performance data were not available, or there was a
limited amount of data that independently evaluated sensor performance). This report is not a
comprehensive review of all available sensor technologies or vendors.

-------
                        2.0    OVERVIEW OF SENSORY SYSTEMS

Sensory systems used for automation, characterization, and monitoring can consist of a number
of different components, including mechanical sensors, electronics, analytical (chemical)
sensors, control systems, telemetry systems, and software. These components may be used alone
or together to form relatively simple or highly complex systems.

Mechanical sensors by definition contain moving parts. For instance, turbine flow meters
contain turbines that rotate as water flows through a pipe. Flow rate is measured by counting the
number of revolutions per minute. When coupled with electronic transmitters, flow meters can
form sensory systems that are able to measure and communicate flow data to a control unit or
display. Other examples of mechanical sensors include floats (used in tank float switches) and
pressure gauges.

Electronic sensors are electrically powered and can measure a variety of parameters such as
pressure, specific gravity,  the presence of liquid (water level meters and interface probes), pH,
temperature, and conductivity.

Analytical sensors are typically used to measure chemical parameters such as contaminant
concentrations.  Some examples of analytical sensors include pH probes, and optical sensors
used for colorimetric measurement

Control systems that work in conjunction with sensors include programmable logic controllers
(PLC) and other electronic microprocessor devices.  Control systems are able to receive sensory
inputs, process information,  and trigger specific actions.

Telemetry systems facilitate system control or data acquisition from remote locations. They can
be radio or telephone based.  Radio-based systems use radiofrequency communication devices to
send and receive information.  Telephone-based systems use modems to send and receive
information through telephone lines.

2.1     CHARACTERIZATION OR MONITORING

Sensors used in characterization are typically used to measure environmental parameters. For
example, a membrane interface probe may be used to detect and locate subsurface
contamination; an electrochemical probe may be used to measure ground water parameters such
as pH; and a thermometer may be used to measure sample temperature.  Sensors in monitoring
are typically used to measure both environmental and systemic parameters. For example, an
anemometer may be used  to  measure wind velocity at a site; a water-level sensor may be used to
measure long term fluctuations in ground water elevation; and a flow meter may be used to
monitor flow through a pipe.

-------
2.2    AUTOMATION

Automation systems use sensory devices to measure parameters necessary for proper system
operation.  Some examples of these parameters are water levels in wells and tanks, temperature,
pressure drop, flow rate, and effluent concentration. These parameters are then used by
microprocessor devices such as PLCs to make operational decisions including starting up or
shutting down components of the remediation system.
                  Additional Sources of Information about Sensor Technologies

  Field Analytic Technologies Encyclopedia (FATE) - an online encyclopedia intended to provide
  information about technologies that can be used in the field to characterize contaminated media,
  monitor the progress of remedial efforts, and in some cases, perform confirmation sampling and
  analysis for site close out. FATE includes information on several types of fiber optic chemical
  sensors,  http://fate.clu-in.org/index.htm

  Measurement and Monitoring Technologies for the 21st Century (21 M2) - through this initiative,
  EPA's Office of Solid Waste and Emergency Response (OSWER) will identify and deploy promising
  measurement and monitoring technologies in response to waste management and site cleanup program
  needs by matching existing and emerging technologies with OSWER program and client needs.
  Current projects include open path monitoring and sampling for contaminated sediments, as well as a
  summary of available literature on measurement and monitoring technologies.
  http://www.cluin.org/programs/21m2/

  Remediation and Characterization Technology Database (EPA REACHIT) - an online database
  with powerful search options for information on treatment and characterization technologies, plus
  updated information from remediation projects undertaken by EPA. The database includes the
  following information for characterization technologies (as of March 2004): 158 technology vendors,
  241 technologies, and 186 vendor source sites, http://www.epareachit.org

  EPA's "A Review of Emerging Sensor Technologies for Facilitating Long-Term Ground Water
  Monitoring of Volatile Organic Compounds" - This report summarizes the status of emerging
  sensor technologies for facilitating long-term ground water monitoring for volatile organic compounds
  (VOCs). It also describes a number of factors, including regulatory acceptance and cost-effectiveness,
  that influence the applicability of these technologies, http://www.clu-in.0rg/s.focus/c/pub/i/1040/

  Superfund Innovative Technology Evaluation (SITE) Program - established by EPA to aid
  engineers,  scientists and other remediation professionals  in the efficient monitoring, characterization
  and remediation of hazardous wastes. In this program, technologies are field-tested to assess
  performance.  Cost and performance data are then presented in technology evaluation reports.
  http://www. epa.gov/ORD/SITE/

-------
        3.0    EXAMPLES OF REMEDIATION SITES THAT HAVE USED SENSOR
                                  TECHNOLOGIES

Table 1 identifies seven case studies on sensor technologies that illustrate their use in site
characterization, monitoring, and process automation. The sites discussed in these case studies
used the following types of technologies:

3.1    SITE CHARACTERIZATION

       •   Membrane Interface Probe - for contaminant concentrations
       •   Geophysical surveys - for evaluation of hydrocarbon contamination

3.2    MONITORING

       •   Capacitance probe - for soil moisture content
       •   VECTOR technology - for ground water flow velocity
       •   Surge System - for sampling and analysis

3.3    AUTOMATION

       •   Ozone analyzers and SCADA with PLC - for ground water pump and treat operation
       •   SCADA with PLC - for ground water pump and treat operation

Five of the seven case studies present characterization and monitoring sensor technologies; the
other two (Moffett Federal Airfield, and Sprague Road Superfund Site) discuss sensor-dependent
automation technologies.  The technologies discussed in this report are commercially available,
and have had at least one full-scale implementation.  Projects for which case studies were
completed were selected based on information in available databases and Internet resources, such
as EPA's Clu-In Web site (www.cluin.org), and discussions with remediation project managers
(RPMs), staff of both EPA Headquarters and Regional Offices, project managers from other
Federal, state, and local government agencies, consultants, and vendors.

Each case study includes site background information, an overview of the sensor technology
used and the goal for using the technology, a brief summary of remedial efforts at  the site,
information about the implementation of the sensor technology, and lessons learned. In addition,
each case study presents cost data for the specific sensor technology.  Where actual cost data are
not available, estimated information is provided. Conclusions in the case studies are not limited
to site-specific details. In most cases, conclusions include site-specific information and general
information about the technology that might benefit potential users. References used in
preparation of each case study are provided at the end of the case study.

-------
TABLE 1. SELECTED CASE STUDIES ON SENSOR TECHNOLOGIES

Site Name
Sol Lynn/
Industrial
Transformer
Superfund Site,
Houston, Texas


Hotel Pier Site,
Pearl Harbor,
Hawaii





Badger Army
Ammunition
Plant, WI (Sub-
Site BAAP-06-
Deterrent
Burning
Ground)



Technology
Employed
Membrane
Interface Probe





Geophysical
survey
techniques - 2D-
Recon and 3D
EOL



Capacitance
probe (for soil
moisture
content)






Time Period
of Use
January -
June 2001





Not provided







2004-
ongoing








Media of
Concern
Ground water






Soil and
ground water






Soil










Contaminants
TCE and its
degradation
products




Hydrocarbon
contamination






Munitions
based
compounds







Goal for Use of
Technology
Delineate ground
water
contamination
and screen
locations
requiring further
characterization.
Characterize
areas of
hydrocarbon
contamination
and assist in
evaluation of
remedial
alternatives.
Measure soil
moisture levels
beneath a cap, to
assess potential
for leaching
contaminants to
GWat 100-1 10 ft
bgs.



Comments
MIP technology was used to
identify highly contaminated
regions in soil and ground water,
as well as delineate the extent of
the contaminant plumes in the
various water-bearing zones.

Electromagnetic surveys
characterized hydrocarbon
contamination based on the
concept that soils contaminated
with hydrocarbons feature higher
resistivity than clean soils.


A nutrient infiltration gallery
encouraged biological
degradation of residual
contamination beneath the cap.
The capacitance probes served as
sentinels against inadvertent
flooding of the remediation zone
that could potentially contaminate
the ground water almost 100 feet
bgs.

-------
TABLE 1. SELECTED CASE STUDIES ON SENSOR TECHNOLOGIES
Site Name
China Lake
Naval Weapons
Station, CA
velocity)
Moffett Federal
Airfield, CA
(West-side
Aquifers)
North Indian
Bend Wash,
AZ, and Nevada
Test Sites, NV
Technology
Employed
VECTOR
technology (for
ground water
flow
Ozone analyzers
and
Programmable
Logic
Controllers
(PLCs) in a
ground water
pump and treat
system
Burge System -
(optical sensor)
Time Period
of Use
1999-
ongoing (data
through
September
2004)
2001-
ongoing (data
through
September
2004)
North Indian
Bend Wash:
Jan 2002 -
July 2003
Nevada Test
Site:
December
2003 and
March 2004
Media of
Concern
Ground water
Ground water
Ground water
Ground water
Contaminants
Not provided
TCE
TCE
Cr-VI
Goal for Use of
Technology
Monitor GW
flow along
southern property
boundary, with
potential for
transport to
nearby municipal
well fields.
Automate pump
and treat system
and monitor
ozone in aqueous
and gaseous
media.
Analyze TCE in
influent and
effluent of
ground water
treatment plant
on a daily basis.
Analyze Cr-VI in
ground water
(pilot test).
Comments
Each velocity sensor interfaces
with an above-ground datalogger
that records sensory data at a
predetermined interval.
Downloaded data is fed into an
accompanying computer program
which translates measured data to
ground water flow speed and
direction.
The ozone monitors work in
conjunction with the PLC to
ensure that (1) the correct dosage
of ozone is applied to the influent
water, (2) the off gas treatment
system is meeting the air
emission standards, and (3) the
ambient air meets occupational
safety and health administration
(OSHA) standards.
The TCE monitoring system was
used to provide automated
monitoring of influent and
effluent from a ground water
treatment system.
The Cr-VI monitoring system was
used for sample acquisition and
analysis of Cr-VI contaminated
water in a pilot test. This system
is currently in use at the Hanford
site near Richland, Washington.

-------
TABLE 1. SELECTED CASE STUDIES ON SENSOR TECHNOLOGIES
Site Name
Sprague Road
Superfund Site,
TX










Technology
Employed
PLC and
SCADA











Time Period
of Use
2003-
ongoing











Media of
Concern
Ground water












Contaminants
Cr-VI












Goal for Use of
Technology
Automation of
pump and treat
system.










Comments
PLCs used to control valves and
pumps. They interface with field
sensors and interpret real-time
sensory data to make system-
control decisions (e.g., turn pump
on or shut valve). The PLCs
communicate through a wireless
network and interface with
desktop computers that serve as
data loggers, continuously
recording system operation data
such as flow rates and totalized
flow.

-------
                                   4.0    REFERENCES
EPA.  2004.  21M2 - Measurement and Monitoring Technologies for the 21st Century.
http://www.cIuin.org/programs/2Jm2/(Includes quarterly literature search).
WPI.  2004.  Sensor Technology Information Exchange, www.sentix.org. (Includes searchable
database of sensor information)
Federal Remediation Technology Roundtable.  2004. Field and Innovative Sampling and
Analysis Technology Matrix, version 1.0.  www.frtr.gov.
EPA.  2004.  Field Analytic Technologies Encyclopedia (FATE), www.fate.cluin.org. (Includes
section on technologies)
EPA.  2004.  REmediation And CHaracterization Innovative Technologies (EPA REACH IT).
www. epareachit. org.
EPA.  2004.  Environmental Technology Verification (ETV) Program.
www. epa.go v/etv/index. h tm.
Vendor Web Sites
   •   Advantech at www.advantech.com
   •   Ametech, Inc. at http://www.drexelbrook.com/
   •   Analytical Measurements, Inc. at http://www.anyliticalmeasurements.com
   •   Bowles Corporation, Inc. at http://www.bowles-corp.com/cet.htm
   •   Burge Environmental at http://www.burgenv.com/index.html
   •   Campbell Scientific at www.campbellsci.com/sensors.html
   •   Clean Earth Technology at http://www.bowles-corp.com/cet.htm
   •   Conor Pacific at http://www.conorpacific.com/
   •   Containment  Solutions at http://www.containmentsolutions.com/
   •   Control Development at http://www.controldevelopment.com/
   •   Controlotron  at http://www.controlotron.com/
   •   Diversified Remediation Controls, Inc. at http://www.drcl.com/prod01.htm
   •   Foxboro at http://foxboro.com
   •   Geophysical Survey Systems, Inc. at http://www.geophysical.com/SIR20.htm
   •   Geo-Sense at http://www.geo-sense.com/
   •   GE Industrial Systems at http://www.geindustrial.com/cwc/gefanuc/

-------
Gundle/SLT Environmental Inc. at http://www.gseworld.com/
Horiba at http://global.horiba.com/analy_e/u-20_series/
Hydrotechnics at http://www.hydrotechnics.com/index_6.htm
I-CORP International, Inc. at http://www.geosynthetic.com/
Instrumentation Northwest, Inc. at http://inwusa.com/xlseries.htm
KPSI at http://www.psih.com/
Leakwise? at http://www.leakwise.com/
North East Environmental Products, Inc. at http://www.neepsystems.com/
Omega at http://www.omega.com/
PERMA-PIPE, Inc. at http://www.permapipe.com/
Physical Sciences, Inc. at http://www.psicorp.com/
Raychem Corporation at http://www.raychem.com/products/chemlex/tracetek.htm
Remediation Service, Inc. at http://www.rsi-save.com
Remote Possibilities at http://www.remotepossibilities.com/
Revere Control at http://www.reverecontrol.com
Rockwell Software at http://www.software.rockwell.com/rsviewstudio/
SAIC at http://www.saiceemg.com/harrisburg/ers-siteboss.htm
Sensaphone, Inc. at http://www.sensaphone.com/
Strison Wireless Systems at http://strison.com
SubSurface Leak Detection, Inc. at
http://www.subsurfaceleak. com/zcorr_loggerjrod. html
Tracer Research Corporation at http://tracertight.com/
Troxler Electronic Laboratories at http://www.troxlerlabs.com/ap200.html
Turner Designs,  Inc. at http://www.turnerdesigns.com/
Tyco Thermal Controls, Inc. at http://tycothermal.com/
Waste Technologies of Australia, Party Limited, at http://www.wastetechnologies.com/
Wondenvare at www.wonderware.com
YSI at http://www.ysi.com
ZISTOC Corporation at http://www.zistos.com/
                                    10

-------
                         Appendix 1

                 Sensor Technology Case Study
Use of Membrane Interface Probe Technology for Detection of VOCs
      at the Sol Lynn/Industrial Transformer Superfund Site

-------

-------
                             Sensor Technology Case Study
    Use of Membrane Interface Probe Technology for Detection of VOCs at the
                   Sol Lynn/Industrial Transformer Superfund Site
Summary Information [1, 5, 6]	

The membrane interface probe (MIP) is a semi-
quantitative field-screening tool for the
detection of volatile and semi-volatile organic
compounds. It facilitates quantitative
measurement, but the results produced at any
given location are relative and not absolute. It is
often used as a precursor to future more accurate
analysis. For instance, site characterization with
MIP could paint a qualitative picture of
subsurface contamination that distinguishes
between areas of low, medium or high
contamination. This could then be used to
develop a plan for media sampling and
laboratory analysis to assess the amount of
contamination.

MIP technology works by advancing the MIP
through the strata to be explored. The MIP
heats the matrix in contact with it and volatilizes
contaminants.  Volatilized contaminants enter
the probe through a membrane covered window
and are transported to the surface by a
continuous draft of inert carrier gas. The
contaminated gas stream is conveyed  to a
measurement device which then produces a
quantitative result. The result is fed to a
computer which displays it on the screen as a
real-time graph of detected contamination
versus depth of probe penetration.  The
computer also logs this data for future use.

The Sol Lynn/Industrial Transformer Superfund
(Sol Lynn) site is located about 6 miles
southeast of downtown Houston, Texas.
Historical operations at the site resulted in
contamination of several water bearing zones
with trichloroethene (TCE). The 1988 record of
decision selected a pump and treat remedy for
this site which operated for several  years before
ongoing groundwater monitoring revealed that it
was ineffective. A supplemental remedial
investigation (SRI) was initiated in 2001 to
better understand the nature and extent of
contamination.  During this SRI, MIP
technology was used to detect subsurface
volatile organic compounds (VOC).

MIP exploration was conducted at 99 locations
between January 17, 2001, and June 5, 2001. A
truck mounted Geoprobe® rig was used for
probe advancement. Equipment used in
combination with the probe included an MIP
controller, a field computer (to display and log
real-time measurements), and a gas
chromatograph (GC) for gas-phase detection.

Real time results made it possible to employ a
dynamic method of site investigation that
steered locations of subsequent  investigations.
MIP showed limited use as a  quantitative tool.

Technology Description [1, 5,  6, 8]	

MIP is a semi-quantitative screening tool for
detection of volatile and semi-volatile  organic
compounds. Strictly speaking, the probe in
itself has no sensory capability;  it merely
transfers vaporized samples of subsurface
contaminants to gas-phase detectors at the
surface. However, MIPs are used in tandem
with electrical conductivity (EC) sensors. The
two have been integrated into a  single probe that
is still called an MIP. Present day MIPs both
serve as collection devices for subsurface
contaminants, as well as measure soil EC. The
primary use of the EC sensor  in this probe is to
map stratigraphy.

The MIP  is manufactured by Geoprobe®
Systems.  It is a pen shaped device with
stainless steel construction consisting of an EC
sensor, a heater block, and a semi-permeable
                                             1-1

-------
                                          Sol Lynn/Industrial Transformer Superfund Site, Houston, Texas
membrane. Figure 1 displays an MIP. The
membrane is a replaceable steel-impregnated
thin-film polymer approximately 6.35 mm in
diameter, and sits on the probe's heater block.
The block is raised at the leading edge to protect
the membrane. Teflon tubing conveys carrier
gas to and from the probe. Power wires  supply
electrical energy to the heater block, and control
wires facilitate sensory feedback from the EC
unit to the field computer. A model MP3500
surface unit -  called the MIP controller -
monitors and  controls carrier gas supply. The
MIP controller feeds clean nitrogen gas to the
probe, and conveys contaminant laden gas to the
gas detector.  A model FC4000 field computer
interfaces with the EC sensor and the gas
detector to acquire and log all sensory data.
Figure 2 displays the field computer, the MIP
controller, and the gas detector. The field
computer has a screen which provides a
graphical display of real-time measurements as
the probe advances through the soil.  This tells
the operator the location of the contaminant, the
relative concentration, and the type of soil in
which the contaminant is located.

This technology  exploits the phenomenon of gas
transfer across membranes to detect subsurface
contamination. The membrane used in an MIP
is semi-permeable, which means that the
passage of substances through it is selective.
The membrane retains liquids, but allows
compounds in their gaseous state to pass
through.  Cross-membrane transfer of gaseous-
phase contaminants occurs through the process
of diffusion motivated by a concentration
gradient across the membrane.  The heater block
speeds this transfer by elevating the temperature
of the surrounding matrix. The block is heated
to approximately 120 degrees Centigrade (°C).
Heat from the block vaporizes contaminants in
the surrounding matrix  causing higher vapor
pressures, and consequently higher
concentration gradients. Once past the
membrane, contaminants are transported to a
detection unit by a continuously flowing stream
of carrier gas.  Carrier gas sweeps behind the
membrane at a constant flow rate of 35-45
milliliters per minute (mL/min). Travel time
from the membrane to the detector is
approximately 30-60 sec (depending on the
length of trunk line and flow rate).

The ability to detect a contaminant depends on
the type of gas detectors being used.  Any
laboratory grade gaseous phase detector with an
analog output of 1 -5 Volts may be used. Most
commonly used detectors include the
photoionization detector (PID), electron capture
detector (BCD), and the flame ionization
detector (FID).  Each of these detectors is best
suited to a group or type of contaminant.  The
BCD is usually used for the  detection of
chlorinated contaminant (such as TCE, PCE,
etc.); the PID is best suited for the detection of
aromatic hydrocarbons (BTEX compounds);
and the FID is best used for straight chained
hydrocarbons (such as methane, butane, etc.).
These detectors may be used in series with the
least destructive detector being first and the
most destructive detector coming last.  The MIP
field computer (FC4000) system can process up
to four detector signals simultaneously. Figure
2 shows a GC -housing a PID and an FID- being
used as a gas-phase detector.

The essential components of an MIP
characterization effort are:  (1) a direct push
mechanism (such as Geoprobe® or Cone
Penetrometer [CPT]); (2) an MIP; (3) an MIP
controller; (4) gaseous phase detectors; and (5) a
data display and logging system.

MIP exploration typically requires a three-man
crew  including a geologist.  Though not an
unchangeable standard, MIP advancement is
usually accomplished one foot at a time.  The
waiting periods between subsequent
advancements typically last a minute.  This is
necessary for sample collection and transfer to
the surface detector.
                                              1-2

-------
                                 Sol Lynn/Industrial Transformer Superfund Site, Houston, Texas
                                  Figure 1
                         Membrane Interface Probe
                Connector Plug for Heater Power
Thermocouple
        Wre
                                 Teflon Gas Tubing
Removable Membrane
                                                             Membrane Heater
                                                             Blot*
                             Soil Conductivity Dipole
     Source:  Geoprobe" Systems
                                     1-3

-------
                                         Sol Lynn/Industrial Transformer Superfund Site, Houston, Texas
                                          Figure 2
                                  MIP Detection Equipment
                                                                                       FC4000
                                                                                       Field
                                                                                       Computer
                Gas Chromatograph
        MP3500 MIP Controller
            Source: Geoprobe ' Systems
Site Information [1,4, 6]
The Sol Lynn site is located about 6 miles
southeast of downtown Houston, Texas.
Historical activities at the site lead to
contamination of the groundwater with TCE.
Following a remedial investigation (Rl) and
feasibility study, a pump-and-treat system was
designed to address the groundwater problem at
the site. The system was installed in 1990.
After several years of monitoring the
contaminant plume, it became evident that the
pump-and-treat system had failed to mitigate the
problem or achieve plume containment. As a
result, the pump-and-treat system was shut
down in 1999.
An SRI was initiated in late 1999 to gain a
better understanding of the hydrogeology and
contaminant distribution at the Sol Lynn site.

The site hydrogeology at Sol Lynn was found to
be extremely complex. While the first RI had
concluded that there were only 3 water bearing
zones (WBZ) at the site, the SRI found that
there were in fact 9 WBZs in the first 200 feet
below ground surface (bgs). Shallow
groundwater at the site occurred within the more
permeable units of the Beaumont formation.
The surficial hydrogeologic units were a part of
the upper Chicot aquifer.  Each of the WBZs
was separated by low permeability zones (LPZ)
which acted like aquitards between the
individual WBZs.
                                             1-4

-------
                                          Sol Lynn/Industrial Transformer Superfund Site, Houston, Texas
TCE was the primary contaminant of concern
(COC) in groundwater.  TCE by-products,
primarily 1, 2-dichloroethene (1, 2-DCE) and
vinyl chloride (VC), were also present in
groundwater.

MIP was used to delineate groundwater
contamination at the site and serve as a field
screening tool to identify locations requiring
additional characterization.

Ninety-nine MIP advancements were made at
the site between January 17, 2001, and June 5,
2001. Figure 3 displays MIP advancement in
progress at Sol Lynn.

A truck mounted Geoprobe® 5410 unit was used
for probe advancement. A Model 14A
Shimadzu GC was used for gas-phase detection.
The GC housed an BCD and PID detector. The
BCD was particularly sensitive to chlorinated
compounds.  The FC4000 Geoprobe® field
computer was used for data display and logging.
The system displayed membrane temperature,
probe advancement rate, and plotted EC  and
contaminant response versus depth.  The system
produced data that was readily transferable to
spreadsheet programs such as Microsoft  Excel.

The steps in a typical advancement included
setup over the desired location, anchoring
equipment in place, performing pre-
advancement response tests on the MIP,  MIP
advancement, and response monitoring.  The
locations for MIP exploration were not entirely
predetermined.  In many instances the location
of the next MIP advancement was based on the
results of the previous.

Since one objective of the investigative effort
was to delineate the extent of the contaminant
plume, the general methodology  used in
determining exploratory locations was to move
outwards in the direction of decreasing
contamination. The original plan anticipated  a
total of 82 MIP advancements, but the dynamic
nature of the exploration effort lead to a total of
99 advancements.

The advancement rate at Sol Lynn was
approximately one foot per minute.  This
coupled with the waiting period between
advancements resulted in an approximate 2.5 to
3-hour duration for complete penetration at any
given location. As a result, 3 to 4 locations
could be explored per day.  The depth of MIP
penetration was limited by its physical
durability.  Soil resistance made direct push
advancement infeasible beyond a certain depth.
At some locations, direct push was not possible
beyond 20 ft bgs. The actuating force for
advancement beyond this depth involved impact
loading using a hydraulic hammer. The MIP
had a limited tolerance for this type of loading,
and that is why its durability restricted the depth
of exploration.  The average depth of MIP/EC
exploration at Sol Lynn was 42.86 ft bgs,  and
the maximum depth was 64.10 ft bgs.

Gaseous phase detectors  quantified MIP carrier
gas contaminant concentrations in millvolts
(mV). The measurements represented a family
of contaminants rather than one specific
contaminant.  Analytical results from
groundwater samples collected in the vicinity of
specific MIP advancements later revealed that
the minimum detection limit of MIP exploration
at that site was greater than 100 parts per billion
(ppb) of total volatiles.

A response test was performed at every new
location prior to MIP advancement. This  was
done to evaluate the condition of the membrane.
The test involved immersing the probe in  a
standard solution and observing the response on
the detectors. The response was compared to
that of previous tests. A  decline in response
indicated the need for membrane replacement.
Figure 4 presents an example of the same test
being performed at a Geoprobe® facility.
                                              1-5

-------
                        Sol Lynn/Industrial Transformer Superfund Site, Houston, Texas
                         Figure 3
MIP Probe Advancement Using a Geoprobe® Direct Push Rig
 Source:  Tetra Tech EM Inc.
 Source: Tetra Tech EM Inc.
                            1-6

-------
                                             Sol Lynn/Industrial Transformer Superfund Site, Houston, Texas
                                               Figure 4
                                         MIP Response Test
Gas
Chromatograph
                                                  MIP Immersed
                                                  in Standard
                                                  Solution
  Laptop
  Computer
  (running
  same
  software as
  the FC4000
  field
  computer)

MP3500
MIP
Controller
                               ®,
               Source: Geoprobe  Systems
                                                  1-7

-------
                                          Sol Lynn/Industrial Transformer Superfund Site, Houston, Texas
Cost [2, 4]
The total cost of the investigative field effort at
Sol Lynn was approximately $150,000.
However, part of this cost did not pertain to MIP
exploration. Costs quoted by contractors
offering MIP field services varied from roughly
$3,000 to $3,700 per day. On an average, 250
to 300 feet of strata could be explored in one
day.  Costs could be influenced by factors such
as the depth of penetration, and the media to be
penetrated. There could be a surcharge for
requiring penetration beyond a certain depth.
Similarly there could be a surcharge to
compensate for equipment wear if impact
loading were used for MIP advancement. A
typical probe withstands approximately 1,000 ft
of cumulative advancement through clays and
sands.  However, probe advancement through
gravel would likely result in shorter life spans
and consequently, additional charge for
exploration. Difficulty in accessing locations
for MIP exploration could be another factor
increasing cost. Equipment mobilization and
demobilization charges varied from $2 to $ 15
per mile.

Conclusions [1, 2,3, 4]	

In general MIP technology worked well as a
screening tool at Sol Lynn.  Although
repeatability of measurement was observed in a
few cases, MIP was not found to be
significantly useful as a quantitative tool. There
were a few instances when highly contaminated
media were encountered. In such cases
contaminant concentrations were found to lie
outside the detectors' range of measurement.
The results of MIP exploration at Sol Lynn
played a large role in developing the monitoring
well plan for the site.

The remainder of this section discusses
important issues about MIP.  These are not
necessarily site-specific issues, and include
general opinions gleaned from contractors that
offer MIP field services. Two such contractors
are Plains Environmental and Applied Research.
Plains Environmental was the contractor at Sol
Lynn, and Applied Research was contacted to
serve as an alternate source for this case study.

Although theoretically suitable for any VOC,
there seemed to be agreement that MIP worked
especially well for chlorinated VOCs.
Chlorinated VOC in-situ detection limits were
stated to range from 200 to 500 ppb. The use of
MIP as a field  screening tool would not be
advisable at in-situ concentrations below the
aforementioned.  Variation in detection limits
were attributed to probe peculiarities,
contaminant chemistries, and specifics of the
strata being penetrated. At the other end of the
spectrum, extremely high in-situ concentrations
could also be troublesome with MIP
exploration. Penetrating a zone of free-phase
contamination could lead to saturation of the
MIP carrier gas trunk line.  Once saturated,
further exploration would not be possible until
the line were completely purged. Purging could
take from a few to several minutes. Purging is
not a special procedure; it is implemented by
simply letting the system operate without
advancing the probe. While the probe remains
in place, cross-membrane contaminant transfer
continues until the contaminants' gaseous phase
in the vicinity of the membrane is depleted.
Once the ingress of contaminants has stopped,
the continuing flow of nitrogen carrier gas
renders the trunk line purged in the duration it
takes to flush an entire tube volume. An MIP
carrier gas trunk line would be considered
purged when the detectors read zero. In
extreme cases, purging could be impracticable.
In such cases, the trunk line would have to be
replaced to facilitate further exploration.

There seemed  to be some difference of opinion
in the preferred method of MIP advancement.
One contractor stated that they preferred to use
CPT instead of Geoprobe® rigs for MIP
                                              1-8

-------
                                          Sol Lynn/Industrial Transformer Superfund Site, Houston, Texas
exploration. This was because CPT rigs were
said to provide considerably greater pushing
force, which consequently afforded penetration
to greater depths. CPT rigs also made it
possible to use a wider variety of probes in
combination with MIP/EC probes to collect
additional geological data.  Another contractor
stated that they preferred to use Geoprobe® rigs
for MIP advancement because those rigs
facilitated impact loading which permitted
deeper penetration than possible with direct
push.

As implied by the sensitivity of various
detectors to different contaminant groups,
knowledge  of existing contaminant types in the
media to be explored was considered
advantegeous. Prior knowledge of
contamination not only facilitates selection of
the  most suitable detectors, but also lets one
evaluate the applicability of MIP as an
exploration tool. In general, dry electrolytic
conductivity detectors (DELCD) and ECDs are
considered best suited for detection of
halogenated VOCs.  PIDs though able to detect
low molecular weight VOCs including
halogenated VOCs,  cannot distinguish between
halogenated and non-halogenated VOCs. FIDs
are  characterized by low sensitivity, but are
useful for detecting straight chained
hydrocarbons. The likelihood of detection
decreased with increasing molecular weight.
Another detector that has been used with MIP is
the  mass spectrometer. This showed promise
when it was first used in the mid nineties, but -
according to a contractor- has been sparingly
used since.

According to the lead agency's contractor there
seems to be a deficiency of quantitative
guidance to predict the effectiveness of MIP in
detecting any given  compound in a given strata.
Since the working principle is based on
vaporization of contaminants, it is likely that the
contaminant's boiling point, and Henry's Law
constant could help predict its response to MIP.
Consequently, it might be safe to doubt the
effectiveness of MIP for contaminants with
boiling points above the temperature (120 °C)
attained by its heating block. One contractor
stated that MIP  could be expected to lose
effectiveness with compounds at least as heavy
as xylenes (which have a boiling point of
140 °C).  The same contractor also verified that
MIP was unable to detect dichlorobenzene
which has a lower molecular weight than
xylenes, but  a higher boiling point (172 to
174 °C).  However, this does not imply that MIP
will work for any compound with a boiling
point less than 120 °C.  Heavier compounds that
volatilize and pass through the membrane could
still condense in the trunk line and drop out of
the carrier gas before they reach the detector.
Given the current incomprehensive knowledge
on the potential for this technology, one might
best be served by discussing their investigative
needs with the MIP vendor before assuming the
applicability of MIP to their site.

Contact Information	

Lead Agency:
EPA Region 6
Remedial Project Manager
Mr. Ernest R. Franke, PE
Phone: 214-665-8521
Email: franke.ernest@epa.gov

State Agency:
Texas Commission on Environmental Quality
Project Manager
Ms. Carol Dye, P.G.
Phone: (512)239-1504
Email: cdye@tceq.state.tx.us

Lead Agency's Contractor:
Tetra Tech EM, Inc.
Project Manager
Mr. Timothy Startz
Phone: (214)740-2064
Email: tim.startz(a).ttemi.com
                                              1-9

-------
                                        Sol Lynn/Industrial Transformer Superfund Site, Houston, Texas
MIP Vendor:
Geoprobe® Systems
Mr. John Terpening
Phone: (800)436-7762
Email:  terpeningj@geoprobe. com

MIP Contractor (at Sol Lynn):
Plains Environmental
Mr. Lynn Newcomer
Phone: (800)542-0445
Email:  lynn@plains.kscoxmail.com

MIP Contractor (Alternate Source):
Applied Research
Mr. Ray Reed
Phone: (281)290-6493
Email:
       rree
\ara.com
References
The following references were used in the
preparation of this report:

1.  Telephone Conversation. Bill Gagnon, SKA
   Consultants, with Chitranjan Christian, Tetra
   Tech EM Inc., Response to Questions on use
   of MIP Technology at Sol Lynn. May 26,
   2004.
2.  Telephone Conversation. Ray Reed,
   Applied Research, with Chitranjan
   Christian, Tetra Tech EM Inc., Response to
   Request for Information on MIP
   Technology.  April 26, September 15, and
    16, 2004.
3.  Telephone Conversation. Neal Van Wyck,
   Applied Research, with Chitranjan
   Christian, Tetra Tech EM Inc., Response to
   Questions on MIP Effectiveness. September
    15,2004.
4.  Telephone Conversation.  Lynn Newcomer,
   Plains Environmental, with Chitranjan
   Christian, Tetra Tech EM Inc., Response to
   Questions on MIP Technology. September
    15,2004.
5.  E-mail from John Terpining, Geoprobe®
   Systems, to Chitranjan Christian, Tetra Tech
   EM Inc., Response to Request for
   Information on MIP Technology. July 28,
   2004.
6.  Tetra Tech EM, Inc., 2003. Supplemental
   Remedial Investigation Report.  Sol
   Lynn/Industrial Transformer Site. Houston,
   Texas. December 2002.
7.  Worldwide web resource.  Chemical
   Database.  ChemFinder.
   http://chemfmder. cambridgesoft. com
8.  Worldwide web resource.  Geoprobe }
   Systems.
   http://www.geoprobesystems.eom/whatjs/w
   hat_is.htm

Acknowledgements

This report was prepared for the U.S.
Environmental Protection Agency's Office of
Solid Waste and Emergency Response, Office
of Superfund Remediation and Technology
Innovation. Assistance was provided by Tetra
Tech EM Inc. under EPA Contract No.
68-W-02-034.
                                            1-10

-------
                           Appendix 2

                   Sensor Technology Case Study
          2D-Recon and EOL Geophysical Survey Techniques
for Characterizing Hydrocarbon-Contaminated Soils at the Hotel Pier Site

-------

-------
                            Sensor Technology Case Study
              2D-Recon andEOL Geophysical Survey Techniques for
                  Characterizing Hydrocarbon-Contaminated Soils
                                  at the Hotel Pier Site
Summary Information
Surface and subsurface geophysical surveys
were completed at the Hotel Pier site on Pearl
Harbor Naval Base in Hawaii to characterize
hydrocarbon contamination and help evaluate
remedial alternatives. Geophysical resistivity
surveys were useful for characterizing
hydrocarbon contamination sites  because soil
with hydrocarbons within the soil pores
(contaminated areas) have a relatively higher
resistivity to electromagnetic current than
similar soil without hydrocarbons within the soil
pores (uncontaminated areas).  A surface two-
dimensional gradiometer (2D-Recon) survey
was used initially to measure relative resistivity
and determine the probable plume boundaries
across a 7-acre site. The subsurface 3D
electromagnetic offset log (EOL) survey was
then completed to define and map resistivity
using a 10-foot grid spacing over 1.4 acres of
the Hotel Pier site.

Surface and subsurface geophysical surveys
took place in January 1999. The  2D-Recon
survey was completed in 2 days while the EOL
survey required an additional 7 days for data
acquisition.  Final data processing and reporting
required approximately two months to complete.
The geophysical surveys at the Hotel Pier site
provided additional surface and subsurface
information on the extent of hydrocarbon
contamination in a rapid manner. Geophysical
surveys provide the greatest characterization
value for sites with large and complex
hydrocarbon contamination.
Technology Description [10,11]	

While 2D-Recon is used to determine the
horizontal extent of subsurface contamination,
3D EOL is used to determine vertical variations
in contaminant distribution. 2D EM surveys
and EOL surveys have been used for over 15
years to identify and model hydrocarbon
contamination associated with leaking
underground storage tank (UST) sites.

2D-Recon

The 2D-Recon survey (a surface
electromagnetic [EM] gradiometer technique)
evolved from geophysical techniques developed
for petroleum and mining exploration. Such
techniques were used to detect subsurface zones
with unusually high contrasts in electrical
resistivity.

2D-Recon measurements are made by moving a
surface EM gradiometer along closely spaced
transects around and over suspected areas of
contamination. The data acquired is processed
to filter noise and produce a horizontal outline
of subsurface contaminant plumes.

The surface EM gradiometer data acquisition
system uses pairs of EM sensor coils normally
aligned vertically; one sensor near the ground
surface and the other sensor five feet above it.
Figure 1 shows sensor coils mounted on a golf
cart for faster data collection. Each pair of coils
measures the difference in signal voltage (or the
voltage gradient) between the coils.  Anomalies
in soil resistivity are indicated by large changes
in the voltage gradient, and imply the presence
of subsurface contamination.
                                             2-1

-------
                                                                        Hotel Pier Site, Oahu, Hawaii
                                            Figure 1
                               2D-Recon Survey Data Acquisition
                        Source: Pritchard Geophysics
EOL

The EOL survey is an established
electromagnetic induction technique which
measures resistivity variations in the subsurface.
For example, all free hydrocarbons are highly
resistive while subsurface waters are much
lower in resistivity.  By measuring resistivity
variations of the subsurface, one can predict the
presence of hydrocarbon plumes.  The
resistivities are plotted to provide a three
dimensional information on subsurface
contamination.

The EOL survey utilizes a very large surface
transmitter coil at low frequency to induce a
magnetic field in the subsurface.  A receiver coil
is placed in a nearby well and measures the
signal. The strength of the measured signal is
proportional to the resistivity of the soil it
passed through. This measurement is
transmitted to a data collection unit on the
surface. The measurements are taken and
recorded at 0.1 foot intervals. Figure 2 shows a
typical layout for an EOL system. Once
measurements are made at a location, the
transmitter coil is moved to a new data point on
the surface for another set of measurements.

Once all the necessary data has been collected,
it is edited to eliminate extraneous noise. The
data is then normalized to eliminate differences
caused by data acquired from separate receiver
wells.
                                               2-2

-------
                                                                        Hotel Pier Site, Oahu, Hawaii
                                            Figure 2
                                     3D EOL Typical Setup
        Well
       EOL
       Collector
                           EOL
                           Transmitter
                           Coil
                         Contaminant
                         Plume
          Source: Gehm Environmental
Once this is done, apparent resistivity and
second order resistivity logs are generated
versus depth. These logs are then interpreted to
identify contaminated zones.

Site Information [1, 2,3, 4, 5, 6]	

The Hotel Pier site is located at Fleet Industrial
Supply Center, Pearl Harbor Naval Base, Oahu,
Hawaii.  A non-time-critical removal action was
proposed for the Hotel Pier site and EOL was
completed as part of the remedial site evaluation
for the site. Previous investigations at the site
included an underground storage  tank (UST)
investigation, soil gas survey, a site
reconnaissance, and a remedial investigation
(RI). The RI for the site concluded that a plume
of free product existed and that local
groundwater was impacted with petroleum
constituents. Preliminary fate and transport
calculations conducted under the  RI suggested
that free product might be releasing to the
surface waters of Pearl Harbor.
In addition, several fuel spills have occurred at
the Hotel Pier site. The most recent fuel spill
occurred in July 1997 when an estimated 1,500
to 3,500 gallons of diesel fuel leaked from a
damaged fuel line. As a result of historic fuel
leaks, elevated levels of benzene, toluene, ethyl
benzene, and total xylenes (BTEX) were
detected in soils and groundwater samples
collected from soil borings located across the
site.  Based on the RI and other site
investigations, the subsurface free product
contamination at the Hotel Pier site may be
associated with the historic spills, leaking
subsurface product lines,  or the preferential
pathways for plume migration along subsurface
electrical and sewer utility lines.

Hotel Pier site has a gentle relief with elevations
from 2 to 10 feet above mean sea level. Pearl
Harbor is essentially a series of drowned river
valleys that formed between the Waianae and
Koolau Volcanoes through a complicated
history of rising and falling sea levels,
subsequent erosion and deposition of alluvial
                                              2-3

-------
                                                                       Hotel Pier Site, Oahu, Hawaii
material, and additional deposition of
pyroclastic ash from eruptions at Salt Lake and
Makalapa. Pearl Harbor is underlain primarily
by soils of the lualualei-fill land-ewa
associations  which are well-drained and fine- to
moderately-fine textured soils.  The primary
surface soil across Hotel Pier site is composed
of fill material with brown silt with angular tuff
and coral fragments, and occasional basalt
fragments.

2D-Recon at the Hotel Pier Site

The 2D-Recon survey used present EM noise
fields generated by existing overhead power
lines as the source field. The EM fields may
penetrate into the ground to depths of 300 feet.
In cases and sites where electrical noise is low
or weak, a temporary EM power source can be
supplied for the survey by hanging a temporary
line between buildings or other aboveground
objects. The 2D-Recon technology components
included two sensor coils (upper and lower), a
laptop computer with data acquisition software,
and a measuring wheel to record distance and
location. The lower sensor measures the
resistive change in the earth and the upper
sensor is used to account for and correct errors
in the power readings  of the lower sensor. The
2D-Recon survey coils and components were
mounted on a rented electric golf cart for fast
and effective data collection (Figure 1). The
2D-Recon survey measured resistivity  changes
in the soil and geology below it as the sensors
were moved over the surface. The data was
processed to develop a 2-D aerial picture of the
subsurface hydrocarbon contamination.

3D EOL at the Hotel Pier Site

The 3D EOL geophysical survey used  a surface
source coil (transmitter) with an area of
approximately four square meters.  The surface
source coil was constructed with more than 30
loops of low resistance wire and connected to a
transmitter power unit consisting of a 1,600 watt
60 hertz gas-powered generator and a power
amplifier capable of up to 12 ampere output.
The amplifier was set at a 5.00 amperes signal
and the transmitter coil and receiver were tuned
to around 270 Hertz (Hz) with a narrow
bandwidth. This tuning procedure, along with
choosing EM receiver wells in low noise level
areas, can filter out most of the excess and
unwanted electrical noise and allows the EOL
technology to be used in and around most
manmade structures and other sources of
subsurface electrical noise.  A grid pattern with
approximately 10-foot spacing was used to
survey a central 100 feet by 600 feet area of the
Hotel Pier site (1.4 acres).  Figure 3 shows the
EOL transmitter and receiver locations for the
Hotel Pier site. The transmitter coil was a
portable, self-contained unit that was placed at
each grid point one at a time, and the electrical
signal transmitted.

The EM receiver was mounted on a 2-inch
diameter, 4-foot long probe and connected to  a
wire-line winch and EOL data acquisition
system.  To receive the induced signal from the
transmitter, the receiver was mechanically
pulled up through the receiver well hole
measuring the primary and secondary EM fields
produced at the transmitter coil location.  A
large, long wavelength response was created
representing the primary EM field.
Superimposed on this response were responses
related to the secondary EM fields caused by
eddy currents moving around the boundaries  of
resistivity contrasts in the earth.

The EM receiver was mounted on a 2-inch
diameter, 4-foot long probe and connected to  a
wire-line winch and EOL data acquisition
system. To receive the induced signal  from the
transmitter, the receiver was mechanically
pulled up through the receiver well hole
measuring the primary and secondary EM fields
produced at the transmitter coil location.  A
large, long wavelength response was created
representing the primary EM field.
                                               2-4

-------
                                                                        Hotel Pier Site, Oahu, Hawaii
                                            Figure 3
                        EOL Survey Transmitter and Receiver Locations



BUILDING
439



O EOL-1







FENCE


LEG
MWQ
BUILDING gp x
404
EOLO

"


tEMP;
Monitor Wei)
Strata Probe

EOL Recover Well

EOL Survey Point




I?



5

                                    SAN JUAN AVENUE
                                                                 SAN JUAN AVENUE
                 !
                      VS-3 '
                  MW-il  A I
                  O  -^ 1
                   o _,^.6.1,
                  PEARL HARBOR
                 0 40  80 120 16!
                    SCALE (IMI)
-106 X

•

t?~ ~ ' "" ~ "~ ."T < f~
r . 	 , 	 , .(|p
" , WORK " . I 	 T = |
"r=,- BUILDINO I 1 I
. . 4M ; • t::
i. i 17J8i::.L
\ {ENCE- ••• \ •p-"Jl I 	 J ••O
H:^::::::-:-fe:";:;::H::::::-:
O EOL-2

BUILDING
• 405 *
f— v-17 SUMP?


                                                            PEARL HARBOR
                                                            SOURCE Gehm Environmental -1999
                 Source: Gehm Environmental
Superimposed on this response were responses
related to the secondary EM fields caused by
eddy currents moving around the boundaries of
resistivity contrasts in the earth.

The raw data were computer processed to
remove the primary field and calculate and
verify the secondary fields. The secondary
fields were converted to apparent resistivity
measured in ohm-meters to compare directly to
the physical properties in the earth. The
computer processed data were presented in both
three dimensional figures and as depth-specific
slices and cross sectional views across the Hotel
Pier site.

System Operation

Calibration and maintenance checks on the EOL
transmitter and receiver tools were typically
performed at the beginning of each day of
survey and at changes between receiver well
locations. The portable generator required
typical oil and fuel checks.  Both 2D-Recon and
EOL data acquisitions were collected in real
time, therefore it was obvious when the
equipment was not properly functioning. Each
completed resistivity logging data set was
uniquely named and saved to a laptop computer.
The primary operation and maintenance tasks
involved making routine checks of the wiring
and wiring connects and the transmitter's output
and signal level.

EOL measurements were not affected by
asphalt, reinforced concrete features, or buried
steel utility lines and pipelines. However, the
Hotel Pier site had several underground product
fuel lines and sewer lines with cathodic
protection that could have potentially interfered
with the EOL survey.  To avoid interference, the
cathodic protection was turned off at the
beginning of each day and turned back on at the
end of each day, allowing an 8 to 10 hour time
interval for the EOL survey to be conducted
without interference.
                                              2-5

-------
                                                                        Hotel Pier Site, Oahu, Hawaii
Cost [ 7, 9]
A complete 2D-Recon survey costs about
$2,500 per day with coverage of between two to
five acres per day, depending on data point
density. Mobilization and demobilization are
additional costs. The 2D-Recon produces
qualitative data and the data are available within
two days of the final data collection.

A complete EOL geophysical survey costs about
$10,000 per day including all materials,
supplies, equipment rentals, data acquisition,
two field staff, one geophysicist, data
processing, and report preparation.  The daily
rate of data acquisition is dependent on the
contamination depth, groundwater depth, and
overall logging time at each EOL location.
Based on observations at three sites (North
Island, San Diego; Makalapa Field, Hawaii; and
Hotel Pier site, Hawaii), between 50 to 100
individual EOL survey locations (0.1 to 0.25
acres per day) can be completed during an eight
hour day.  Mobilization and demobilization and
completion of any additional monitoring wells
are added costs. Final data processing and
reporting, performed using Silicon Graphics'
ShowCase and word processing software,
required approximately two months to complete.

At Hotel Pier, 2D-Recon and EOL surveys cost
approximately $10,000 and $90,000
respectively.

Conclusions  [8]

Data from the 2D-Recon and EOL survey were
used to characterize the lateral extent of the
hydrocarbon plume and  define the general
boundaries of larger subsurface anomalies.
Qualitative results were  presented in graphical
format using five-color resistivity maps
correlating to low, average, above average, high,
and very-high resistivity values (Figure 4). The
project geophysicist stated that the high and
very high resistivity features identified in plan
view depth intervals and in cross-section may be
associated with the soils with hydrocarbon filled
pores.  Many of the high resistivity anomalies
were found to be positioned around and along
utility and pipeline corridors which are known
to be preferential conduits for hydrocarbon
movement.

Post EOL survey soil sampling was completed
but locations were limited to areas outside of the
buildings and depths only reached fifteen feet
below ground surface. Laboratory analytical
data identified hydrocarbon contaminated soil
which agreed with the resistivity model.
Multiple anomalies were identified from the 2D-
Recon and EOL surveys that likely would not
have been identified through traditional
subsurface soil sampling. The remedial
alternative selection criteria and feasibility and
effectiveness of a subsurface barrier to control
hydrocarbon seepage into the harbor were
supported through this survey.

Technology Performance and Factors

Factors that affected the performance of this
technology at the Hotel Pier site were (1) the
experience of the data acquisition team (two or
three field staff), (2) experience and interpretive
skills of the  senior geophysicist, (3) availability
of the Senior Geophysicist to be on site during
the  field data acquisition and to provide quality
control, (4) access to all areas of the site  (inside
and outside of buildings), (5) timing and
logistics to complete additional monitoring
wells at specific locations within a few days,
and (6) other site-specific factors such as depth
to groundwater, inherent resistivity of the
contaminated and uncontaminated areas, nature
and amount of overhead and buried utilities, and
general size of the contaminated area. The
overall qualifications and capability of the
geophysicist to generate good data and process
the  data into a usable and accurate model were
paramount.
                                              2-6

-------
                                                                        Hotel Pier Site, Oahu, Hawaii
                                            Figure 4
                        Apparent Resistivity Map from 2D-Recon Survey
.108 Jf^
ZA
V
-
BUILDING 489
1 o EOL-1
SpF/S



, ll ,-»






BUILDING
406


                                                                               RESISTIVITY:

                                                                                  VERY HIGH

                                                                                 '•''• HIGH
                                                                              \ .'
                                                                                   AVERAGE
                                                                                      USABLE
                                                                                    1 TRANSECTS
       50   100   150   200

         SCALE (feet)
           Source:  Gehm Environmental (Best available copy)
A weakness of EOL surveys is the need for site
characterization data to distinguish and calibrate
the soil resistivity measurements.  A significant
contrast between the resistivity of contaminated
and the non-contaminated material is needed in
order to have a high degree of confidence in the
data.

Some subsurface site characterization data is
necessary to discern the initial resistivity
measurements. For example, subsurface
geology information from monitoring well
boring logs, previous subsurface sampling, and
groundwater monitoring are all potentially
valuable characterization data helpful for an
EOL survey.
Suitable existing EM receiver wells for the EOL
receiver will not be available at all hydrocarbon
contaminated sites; therefore time and costs may
be needed to properly install new EM receiver
wells.  The most suitable EOL receiver wells are
clean monitoring wells located just outside the
central edge of the plume.  The radial coverage
from a receiver well is dependent on site
geology and subsurface features but is
approximately 200 to 300 feet.

One improvement that has occurred during the
last 15 years is the phased approach in which
the geophysical survey begins with a more cost
effective 2-D surface gradiometer survey. Then,
if contamination and site logistics  concur, the
                                               2-7

-------
                                                                    Hotel Pier Site, Oahu, Hawaii
survey proceeds to the more labor-intensive and
costly 3D EOL survey. Smaller, more portable
laptop computers and the use of commonly
available database software packages have
helped with the data acquisition and storage of
large data sets.

Lessons Learned

A site should have a potentially large and
complex area of hydrocarbon contamination to
afford the use of surface and subsurface
geophysics. The 2D-Recon and EOL
geophysical surveys can quickly provide
additional subsurface information on the extent
of hydrocarbon contamination. Data are
collected in real time allowing for on-site
qualitative assessments of the plume boundaries,
appropriate locations for new monitoring wells,
and subsurface sampling efforts which may all
help to expedite the entire remediation effort.

Though the geophysical survey techniques
provided information on hydrocarbon entrained
in the soil, additional analysis was needed to
determine the fraction of it that was mobile and
recoverable.

The EOL technology performance was
evaluated at North Island Navy Base under the
Navy Environmental Leadership Program
(NELP) in 1997.

Contact Information

Owner's Contractor
Tetra Tech EM Inc.  .
Mr. J. Edward Surbrugg, Ph.D.
Phone (406) 442-5588
Email: edward.surbrugg@ttemi.com

Technology Vendor
Gehm Environmental
Mr. Dave Gehm, President
Phone:  (660)882-3485
Email: dgehm@gehm.com
Technology Vendor
Pritchard Geophysics
Mr. James Pritchard, Ph.D.
Phone: (972)851-3433
Email:  jip.gp-1 @worldnet. att. net

Navy Contact
Pacific Division, Pearl Harbor
Naval Facilities Engineering Command
Ms. Michelle Yoshioka, Remedial Project
Manager
Phone: (808)472-1413
Email:  michelle.yoshioka@navy.mil

References

The following references were used in the
preparation of this report:

1. Gehm Corporation. 1995. Technical
   Details and Case Histories, Electromagnetic
   Offset Log (EOL), by: James I. Pritchard,
   Technical Director. December.
2. Gehm Environmental. 1999. 3-D
   Resistivity Survey, Electromagnetic Offset
   Log (EOL), Hotel Pier Area, Pearl Harbor
   Naval Base, Oahu, Hawaii.  March.
3. Macdonald, G.A., A.T. Abbott, and F.L.
   Peterson.  1983. "Volcanoes in the Sea."
   University of Hawaii Press.  Honolulu. 2nd
    Edition.
4.  Ogden Environmental and Energy Services
    Co., Inc. (Ogden). 1994. "Remedial
    Investigation/Feasibility Study Draft Report
    for Subsurface Fuel Investigation, Naval
    Base Pearl Harbor, Hawaii."
5.  Tetra Tech EM Inc. 1996. Electromagnetic
    Offset Logging at Naval Aviation Deport
    Buildings 379 and 397.  Demonstration
    Evaluation Report. Navy Environmental
    Leadership Program. November 25.
                                             2-8

-------
                                                                    Hotel Pier Site, Oahu, Hawaii
6.  U.S. Department of Agriculture (USDA).
   1972. "Soil Survey of Island of Kuai, Oahu,
   Maui, Molokai, and Lanai, State of Hawaii."
   U.S. Department of Agriculture, Soil
   Conservation Service in Cooperation with
   University of Hawaii. August.
7.  E-mail from Edward  Surbrugg, Tetra Tech
   EM Inc., to Chitranjan Christian, Tetra Tech
   EM Inc., Response to Questions on Cost.
   January 13, 2005.
8.  E-mail from Michelle Yoshionka, Naval
   Facilities Engineering Command, to Ellen
   Rubin, EPA, Office of Superfund
   Remediation and Technology Innovation,
   Response to Request for Comments.
   November 3, 2004.
9.  E-mail from James Pritchard, Pritchard
   Geophysics, to Ellen  Rubin, EPA, Office of
   Superfund Remediation and Technology
   Innovation, Response to Request for
   comments. September 10, 2004.
10. Worldwide web resource.  Gehm
   Environmental.
   http://www.gehm.com/eol.htm.
   Electromagnetic Offset Log.
11. Worldwide web resource.  Gehm - Prichard
   Environmental Surveys.
   http://www.gpesurveys.com/2-d-recon-
   surveys.html.  2D Recon Surveys.
Acknowledgements
This report was prepared for the U.S.
Environmental Protection Agency's Office of
Solid Waste and Emergency Response, Office
of Superfund Remediation and Technology
Innovation.  Assistance was provided by Tetra
Tech EM Inc. under EPA Contract No.
68-W-02-034.
                                            2-9

-------

-------
                 Appendix 3

         Sensor Technology Case Study
Use of Capacitance Probes to Measure Soil Moisture
     at the Badger Army Ammunitions Plant

-------

-------
                             Sensor Technology Case Study
                 Use of Capacitance Probes to Measure Soil Moisture
                         at the Badger Army Ammunitions Plant
 Summary Information [1, 2, 3, 4, 5, 6,10,12]

 There are several devices that can be used to
 measure soil moisture content in the vadose
 zone. Some of these devices include dielectric
 permittivity probes, radio-frequency probes,
 lysimeters, tensiometers, and capacitance
 probes. This case study focuses on capacitance
 probes.

 Capacitance probes have long been used to
 measure soil moisture in the agricultural
 industry for irrigation scheduling.  They have
 also been used in environmental applications as
 part of containment type remedies. They
 measure soil moisture content by measuring the
 capacitance (a measure of charge storing
 capacity) of the soil around them.

 Capacitance probes are being used to monitor
 the moisture content in soil at the Badger Army
 Ammunition Plant (BAAP). The BAAP is a
 7,354 acre site located in Sauk County,
 approximately 7 miles south of Baraboo,
 Wisconsin. Intermittent war-time plant
 operations over a 33 year period resulted in
 contamination of the site with munitions based
 compounds. The remedial approach at Badger
 was to address sub-sites individually. This case
 study presents use of soil moisture sensors used
 at sub-site BAAP-06 also called the Deterrent
 Burning Ground (DBG).

 The DBG is located in the northeastern portion
 of BAAP. The DBG was a former two-acre
 borrow pit that was used as a landfill for
 demolition debris, and for the open burning of
 deterrents, structural timbers,  asphalt shingles,
 cardboard, papers, and office waste. The
primary contaminants are volatile organic
compounds and munitions based compounds.
The most contaminated soil in the top 15 feet of
 the waste pit was removed.  The remedy for the
 site included the construction of a 7-acre low-
 permeability cap above the contaminated zone.
 This was to prevent leaching of contaminants
 into the groundwater during large rainfall
 events. Another component of the remedy
 involved in-situ bioremediation of residual
 contamination in the vadose zone.

 The bioremediation system was installed in the
 summer of 2003. The system supplies water
 and nutrients to native microbes in the vadose
 zone through an infiltration gallery.
 Capacitance probes were used to measure soil
 moisture in the vadose zone to evaluate overall
 system performance. In addition, capacitance
 probes provided the sensory means to observe
 the potential for excessive infiltration, and thus
 prevent inadvertent contamination of the
 underlying groundwater.

 Four capacitance probes were installed at
 BAAP. The probes continuously measure soil
 moisture content, and four separate dataloggers
 periodically retrieve measured data from their
 respective probes.  The data is downloaded from
 the datalogger and  examined to understand soil
 moisture response during infiltration.
 Inferences from soil moisture data are used to
 determine the need for modification in
 infiltration gallery operation.

 To date, the system has not been able to detect
 infiltration beyond  the first two feet beneath the
 infiltration gallery.  This is possibly due to the
 lack of a detectable moisture front during
 infiltration.

 Technology Description [1, 2,3, 4, 7, 8]	

Capacitance probes use the principal of
electrical capacitance to measure soil moisture.
                                             3-1

-------
                                                   Badger Army Ammunitions Plant, Baraboo, Wisconsin
Capacitance is a measure of a device's ability to
store electrical charge.  A device that can store
charge is called a capacitor and consists of two
metal plates separated by a dielectric substance
(or insulator).  The device's capacitance
depends on the dielectric substance used. When
an alternating current is applied to a capacitor, it
charges and discharges sequentially causing an
apparent  change in the frequency of applied
current. This change in frequency is a function
of the device's capacitance, and by measuring
the altered frequency, one can determine the
device's capacitance.

The soil-sensor system  is in essence a capacitor.
The sensor's electrodes serve as the plates, and
the soil around the sensors serves as the
dielectric substance.  The capacitance of the
sensor-soil system is a function of the moisture
in the soil.  By measuring the capacitance of the
system, the probe in essence measures the
amount of moisture in the soil.

Capacitance probes have been used to measure
soil moisture in the agricultural industry  for
several years.  The technology has been
commercial  for at least 15 years.

The probe consists of vertically spaced sensor
assemblies enclosed in  a snug-fitting PVC
access tube (Figure 1).  The number of sensors
used in a probe, and the sensor spacing along
the access tube can be varied as desired.
Different manufacturers have different
constraints on the maximum number of sensors
allowed in a probe. Probes in the agricultural
industry are typically 3 feet long and rarely
exceed 6 feet in length.  The access tubes are
installed  in direct contact with the lithology
using a suitable augur method.

Once the access tube is in place,  the sensor
assembly can be inserted into it.  However, the
sensors typically need to be normalized before
this. Alternately, one can purchase normalized
sensors for an additional charge.  Normalization
is a process through which sensor outputs are
standardized to lie on a common scale.
Normalization is necessary because there is a
slight variation in outputs produced from sensor
to sensor. EnviroSMART™ probes are
normalized using vendor-supplied software to
set the sensor's maximum output to 1, and its
minimum output to zero. The maximum
corresponds to complete saturation, and the
minimum corresponds to zero moisture content.
Therefore normalized probes produce outputs
on a unitless zero to 1 scale.

Probe installation involves drilling  a hole to the
required depth. After the probe casing is
lowered into the hole, a bung is inserted to seal
off the bottom. The sensors and controlling
electronics are then inserted into the casing.
Finally the cap is  installed to seal-in the sensors
and associated electronic devices. The cap is
weather proof, but cannot withstand continuous
submergence.  The cap contains a desiccant to
prevent moisture from entering the casing. The
probe head is typically not flush mounted with
the ground. It usually sits about 5 inches above
ground level.

After the probe has been installed in place  and
all the necessary electrical connections have
been made, the sensors may be field-calibrated.
Field calibration in essence involves correlating
the sensor's normalized output to actual  soil
moisture. However, according to the vendor,
field calibration is a rare practice. Most  of the
time customers use the vendor-supplied
standard calibration model (IrriMAX) which
converts retrieved output data to approximated
volumetric soil moisture contents.  Sometimes
normalized results produced by the sensors meet
the customer's requirements, in which case not
even the standard calibration model is used.
However, field-calibration is recommended for
applications that need accurate measurement of
volumetric soil moisture content. Field
calibration takes into account all the
peculiarities of the soil surrounding a given
                                               3-2

-------
                                        Badger Army Ammunitions Plant, Baraboo, Wisconsin
                                 Figure 1
                           Soil Moisture Probe
                                              Top Cap —
                                              Access tube —
                                              Sensor  '
                                              Electrodes
                                                        \
— Sensor
                                              Sensor  /
                                              Electrodes
                                                       \
— Sensor-
                                                                      Sensor
Source: Campbell Scientific Inc.
                                    3-3

-------
                                                  Badger Army Ammunitions Plant, Baraboo, Wisconsin
sensor thereby minimizing measurement error.
It involves developing a mathematical
relationship between the sensor's normalized
output and actual soil moisture content. Data
required for this correlation is obtained by
sensory measurement followed immediately by
collection of a soil sample for laboratory
analysis.  The EnviroSMART™ probe's user
manual provides instructions on  deriving this
mathematic relationship. Once derived, a
spreadsheet program such as Microsoft Excel
can be used to convert sensor outputs to
volumetric soil moisture content.

The table below provides specifications of the
probes used at the Badger site. These probes
were manufactured in Australia by Sentec Inc.
and marketed in the United States by Campbell
Scientific Inc.
Site Information [2, 5,10,12]
EnviroSMART 1M Probe Specification
Parameter
Maximum Cable
Length
Maximum Sensors
per Probe
Accuracy (field
calibration)
Precision
Measurement Range
Temperature Effects
Operating
Temperature Range
Time to Read One
Sensor
Sphere of Influence
Sensor Diameter
Access Tube
Diameter
Value
200ft
16
R2=0.99
(+/-) 0.03% by volume
Oven-dry to saturation
(+/-) 3% (at
temperatures between
41° and 95°F)
32° to 158°F
1.1 seconds
99% of measurement
within 3.94 inch radius
from the outside of the
access tube
2 inches
2.22 inches
BAAP is a 7,354 acre site located in Sauk
County, approximately 7 miles south of
Baraboo, Wisconsin. The BAAP site contains
several sub-sites.  The DBG, where the probes
were installed, is one of the sub-sites at BAAP.

The lithology at the site comprises layers of
sand and gravel with interspersed layers of silt.
The first water bearing zone is encountered
approximately 100 to 110 feet below ground
surface (bgs).

A low permeability cover was part of the
remedy for the DBG. The low permeability
cover reduced surface water infiltration through
the contaminated vadose zone, thereby
preventing the leaching of contaminants into the
groundwater. The other component of the
remedy involved bioremediation of
contaminants in the vadose zone.
Bioremediation of the vadose zone involved the
supply of water and nutrients to sustain
contaminant-degrading microbial communities.
Water and nutrients were supplied through
infiltration galleries constructed approximately
four feet beneath the surface of the low
permeability cap.

Capacitance probes were installed in June 2003
(as part of the bioremediation system) to
monitor infiltration. In addition to serving as
indicators for conditions favoring
bioremediation, the capacitance probes were
expected to provide the sensory means to
recognize excessive infiltration into the vadose
zone. The probes spanned the entire depth of
the vadose zone, extending from the ground
surface to a few feet above the water table.
Figure 2 presents  a section of the installed
probes at Badger. Each installed probe was
approximately 100 feet deep, but the deepest
sensor in each probe laid only 50 feet bgs.
                                              3-4

-------
                                           Badger Army Ammunitions Plant, Baraboo, Wisconsin
                                    Figure 2
                          Soil Moisture Probe Assembly
Note: Modified from as-built drawing (Shaw Environmental, Inc. 2004)
                                       3-5

-------
                                                   Badger Army Ammunitions Plant, Baraboo, Wisconsin
Two infiltration galleries were constructed in
the remediation area. One measured 100 feet by
80 feet, and the other located southeast of the
first measured 320 feet by 100 feet. Of the four
Sentek EnviroSMART™ probes used at the site,
one was installed beneath the smaller gallery,
two were installed beneath the larger gallery,
and one was installed in an area adjacent to both
galleries to serve as  a control.

Sonic drilling was used to drill 4-inch-diameter
holes for each of the probes.  Segments of the
access tube were then spliced together and
lowered into the borehole. Each segment was
approximately 6.8 feet long and had a 2.22-inch
outside diameter. The deepest segment had a
PVC cap at the bottom to provide a moisture-
tight seal. The segments were spliced together
using glued PVC slip joints.  PVC centralizers
were installed at each slip joint and at the
bottom of the assembled access tube.  After the
access tube was  in place,  a kaolinite clay-
cement grout was used to fill the annular space
between the borehole and the access tube. The
grout was placed along the entire length of
borehole from the bottom of the borehole to the
ground surface.  After the grout had set, the
sensors and related electronics were lowered
into each tube. Each probe contained 14
sensors.  Sensor spacing varied along the length
of the probe. In general, the spacing increased
with increasing depth.

The data retrieval system included a serial data
interface (SDI) operating at a baud rate of 1,200
to receive and record data from each soil
moisture sensor. A Campbell Scientific Model
CR510-2M datalogger was used to query each
probe's SDI and store soil moisture data for
future retrieval.  The datalogger could be
programmed to record soil moisture
measurements in intervals that ranged from as
small as five minutes to greater than one week.

The Intelligent Probe Utility Software provided
by Campbell Scientific, Inc. was used to
normalize each sensor through the probe's SDI.
The probes were not field-calibrated.

System Operation [1, 2,11]

The infiltration gallery operated only one day
per quarter to simulate percolation caused by
precipitation. The moisture probes, however,
operated continuously to measure soil moisture
response during, and after infiltration gallery
operation.  The only human effort in operation
of the moisture probes involved data retrieval
from the dataloggers.  Soil moisture
measurements were automatically collected
from the probes by the dataloggers at 30-minute
intervals for a period of up to 4 weeks.

Soil moisture data downloaded from the
dataloggers was  closely examined to gain an
understanding of the changes in soil moisture
content during infiltration gallery operation. To
date, the only observed spikes in soil moisture
content were observed in the upper two feet of
soil beneath the infiltration gallery.

Over the months, there was a decreasing trend
in soil moisture measured by some of the deeper
sensors. This supported the claim that the low
permeability cap was reducing the amount  of
infiltration through the contaminated zone.

System maintenance included periodic checks
for probe cap seal, and changing the desiccant in
the cap as necessary. The probe head
enclosures were replaced with watertight
enclosures to prevent recurrence of sensor
malfunction previously caused by the entrance
of moisture into the probes.

Cost [4, 6, 9]	

Costs for installation of the probes at the Badger
site were not available.  However, a cost
estimate for a similar system is presented in
Table 1.
                                               3-6

-------
                                                  Badger Army Ammunitions Plant, Baraboo, Wisconsin
                                          TABLE 1
       ESTIMATE FOR A TYPICAL SYSTEM USING ONE SOIL MOISTURE PROBE
EnviroSMART™
Item
EnviroSmart Probe ' (50 feet long)
Sensors 2
SDI-12 Interface
Interface Cable 3 ($20 + $0.23/ft)
Yellow Cutting Edge
Expandable Bung
Gel Bags
Ferrite Beads
Normalization Container
Programming Cable
Probe Utility Software
Datalogger (Campbell Scientific CR510-2M)
Qty
1
16
1
50








Unit
ea
ea
ea
ft
ea
ea
set
pack
ea
ea
ea
ea
Unit Price ($)
2,837
200.00
375.00
0.23
10.25
19.00
34.00
12.60
157.00
120.00
63.00
895.00
TOTAL 4
Subtotal ($)
2,837
3,200
375
12
10
19
34
13
157
120
63
895
6,840
    Note:
    i
        Cost varies with length
        Number of sensors in a probe can vary from 1 to a maximum of 16
        Cable runs from the datalogger to the SDI-12 interface
        Cost does not include installation
    Source: Campbell Scientific Inc. 2004 Price List
There are several variables that affect the cost for
installing such a system. However, the two major
variables are depth of installation and number of
sensors used.  The depth of installation (or the
length of probe) directly impacts total cost of the
system. Shorter probes are less expensive than
longer ones. The depth of installation also
influences the cost of drilling.  For shallow probes
not exceeding 7 feet in length, the hole  may be
drilled using the installation kit purchased (or
rented) from the supplier.  On the other hand,
deep installations such as the one at Badger
require drilling rigs,  and more  involved probe
installation methods. The cost of the system also
depends on the number of sensors used per probe.
The number of sensors can vary from 1 to a
maximum of 16.  As in other systems, the number
of probes purchased influences the selling price
per probe. Optimizing datalogger use by using
the maximum number of probes per datalogger
can reduce unit cost per probe.

The estimate presented in Table 1 assumes that
the probe contains 16 sensors that interface with
a Campbell Scientific Model CR510-2M
datalogger, the deepest sensor being 50 feet
deep. In practice the datalogger to be used
depends on requirements of the application, and
not the probe.  The simplest datalogger that can
be used with this probe is Campbell Scientific's
Model CR200. The estimate does not include
drilling or probe installation.
                                              3-7

-------
                                                   Badger Army Ammunitions Plant, Baraboo, Wisconsin
Conclusions [1, 2, 4,10]
According to the owner's contractor, Sentek's
capacitance probes were used at Badger because
they were the only probes amenable to deep
installation.

The cap on the probe's access tube is
weatherproof but does not provide a watertight
seal when submerged. This should be taken into
account while designing systems that  use this
probe. Moisture entry into a probe caused it to
malfunction at Badger. This resulted  in
replacement of the then-existing probe-head
enclosures with watertight enclosures. The
vendor recommends replacement of the desiccant
inside the access tube's cap as needed to prevent
moisture build-up.

Sensors can be removed and re-used as desired.
The access tube might have to be abandoned in
place (depending on the depth of installation), the
sensors, according to the Vendor, can always be
removed and used at another location.
Modification of installed probes by adding or
removing sensors is also possible. Addition of
sensors to an existing probe would be constrained
by the space available inside the access tube, and
the maximum number of sensors (16) that can be
used per probe.

According to the Vendor, most customers are
satisfied with normalized outputs or the vendor-
supplied standard calibration model.
Nevertheless, in applications where accurate soil
moisture measurement is required, field-
calibration is essential.  Field-calibration is a
tedious and time consuming process,  and should
not be neglected in budgetary estimates. It is
important to note that deep installations (such as
those at Badger) can make field-calibration
impractical. This is because collecting a deep soil
sample within the sensor's zone of measurement,
immediately following sensory measurement, can
be  difficult if not impossible. For this reason, one
must discuss the requirements of the application
with the vendor before incorporating the probe
into the design for a remediation system.

The access tubes supplied by the vendor
required splicing using glue and slip joints.
According to the owner's contractor, this
method of splicing presented problems during
installation. Since glue was used, care had to be
taken not to strain the joint until the glue had
dried.  This increased the time and effort
required to install the access tubes. In this
regard, the owner's contractor was of the
opinion that threaded PVC pipe would have
made for a better  access tube. At one location,
an obstruction was encountered at a slip joint in
the access tube during sensor installation. The
inner walls of the access tube had to be reamed
to smooth the slip seams before the sensors
could be re-inserted. In another instance, grout
entered the access tube presumably through a
separated slip joint. The access tube was
abandoned in place, and a new access tube had
to be installed a few feet away from the first
one.

To date there have been no observed spikes in
measured soil moisture beyond the first two feet
beneath the infiltration gallery.  Although this
seems to imply that there is no infiltration
beyond this point, this is not necessarily the
case. Soil vapor vents used to monitor
biological activity have detected gases in
concentrations indicative of enhanced biological
activity.  This suggests successful nutrient
delivery, and consequently infiltration of water
to the remediation zone.  It can be theorized that
infiltration may not have been detected due to
the lack of a saturated front, and the resulting
irregular migration of moisture through the soil
pores. Whatever the migratory path, it appears
that the moisture  did not pass sufficiently close
(3.94 inches from the outside of the access tube)
to the probes to be detected.
                                               3-8

-------
                                                 Badger Army Ammunitions Plant, Baraboo, Wisconsin
Contact Information
Federal Oversight Agency
EPA Region 5
Office of RCRA
Corrective Action Project Manager
Mr. Bob Egan
Phone: (312)886-6212
Email:  egan.roberl@epa.gov

State Oversight Agency
Wisconsin Department of Natural Resources
Site Contact
Mr. Steve Ales
Phone: (608)275-3310
Email:  Stephen.Ales@dnr.state.wi.us

Owner's Contractor
Shaw Group
Project Manager
Mr. Doug Rubingh
Phone: (303)741-7665
Email:  doug.rubingh@shawgrp.com

Probe Vendor
Campbell Scientific, Inc.
Applications Engineer
Mr. David Meek
Phone: (435)750-9555
Email:  davemeek@campbellsci. com

References

The following references were used in the
preparation of this report:

1.  Telephone Conversation. David Meek,
   Campbell Scientific Inc., with Chitranjan
   Christian, Tetra Tech EM Inc., Response to
   Questions on Soil Moisture Probe. June 22,
   July 1, and September 13, 2004.
2.  Telephone Conversation. Doug Rubingh,
   Shaw Group Inc., with Chitranjan Christian,
   Tetra Tech EM Inc., Response to Questions
   on Sensor Technology use at the Badger
   Army Ammunitions Plant.  June 22 and July
   1,2004.
3.  Email from Corey White, Sentek Sensor
   Technologies, to Chitranjan Christian, Tetra
   Tech EM Inc., Response to Questions on
   EnviroSMART™ operation. July 5, 2004
4.  Email from David Meek, Campbell
   Scientific Inc., to Chitranjan Christian, Tetra
   Tech EM Inc., Response to Questions on
   EnviroSMART™ Cost. July 26, and August
   6, 2004.
5.  Email from Doug Rubingh, Shaw
   Environmental, Inc., to Chitranjan Christian,
   Tetra Tech EM Inc., Response to Questions
   on Probe Operation at Badger Army
   Ammunitions Plant. August 11, 2004, and
   July 25, 2005.
6.  Email from Jason Ritter, Campbell
   Scientific Inc., to Chitranjan Christian, Tetra
   Tech EM Inc., Response to Questions on
   EnviroSMART™ Cost. July 16, 2004.
7.  Campbell Scientific Inc. 2004.
   EnviroSMART™ Soil Water Content
   Probes Instruction Manual. July.
8.  Campbell Scientific Inc. 2004.  Product
   Literature. Soil  Volumetric Water Content
   Probes Models EasyAG® and
   EnviroSMART™.  July.
9.  Campbell Scientific Inc. 2004.  2004 U.S.
   Price List. July.
10. Shaw Environmental Inc.  2004. Draft
   Corrective Measures Implementation
   Report, Enhanced Biodegradation and
   RCRA Cap/Cover System, Deterrent
   Burning Ground, Badger Army Ammunition
   Plant, Baraboo, Wisconsin. Revision 1.
   January.
                                            3-9

-------
                                                Badger Army Ammunitions Plant, Baraboo, Wisconsin
11. Shaw Environmental Inc. 2004. Draft
   Operation and Maintenance Manual, Final
   Remedy Enhanced Biodegradation System,
   Deterrent Burning Ground, Badger Army
   Ammunition Plant, Baraboo, Wisconsin.
   Revision 1. April.
12. Worldwide web resource. Badger Army
   Ammunitions Plant.  2003.  Propellant
   Burning Ground. April.
   http://www. badgeraap. org/sites/pbg/pbg. htm
Acknowledgements
This report was prepared for the U.S.
Environmental Protection Agency's Office of
Solid Waste and Emergency Response, Office
of Superfund Remediation and Technology
Innovation. Assistance was provided by Tetra
Tech EM Inc. under EPA Contract No.
68-W-02-034.
                                            3-10

-------
                    Appendix 4

           Sensor Technology Case Study
Use of In-Situ Sensors to Monitor Groundwater Velocity
   at the China Lake Naval Air Weapons System Site

-------

-------
                            Sensor Technology Case Study
              Use ofln-Situ Sensors to Monitor Groundwater Velocity
                  at the China Lake Naval Air Weapons System Site,
Summary Information [1, 2, 3, 4, 5, 6, 7, 8, 9,
Variably Emitting Controlled Thermal Output
Recorder (VECTOR) technology is used to
measure groundwater flow speed and direction
in three dimensions. It works on the principle of
heat perturbation, whereby measured
displacements in the heat-flow field around the
probe are used to calculate groundwater
velocity. The measurement range for velocity
magnitude is 0.01 feet per day (ft/day) to 2.00
ft/day. Three-dimensional directional
measurement has an accuracy of (+/-) 5 degrees.
These probes are installed in situ, much like a
monitoring well, using a suitable method such
as hollow stem auger drilling. Following
installation, the system is connected to 120 volt
power supply. Once all of the necessary
electrical connections have been made, the
probe is turned on, calibrated and allowed to
collect flow data.

VECTOR technology is being used at the China
Lake Naval Weapons Station to monitor
groundwater flow in various water bearing
zones (WBZ) along the southern boundary of
the property.

VECTOR probes were installed in 1999.  Data
from the probes were used to monitor local
groundwater flow as well as  seasonal and
background impacts to local  flow fields (such as
agricultural or municipal pumping) located
several thousands of feet away from the probes.

Quarterly site visits to download stored data
from automatic data collection devices
constituted the only human effort in system
operation. Although the option for remote
access via telephone or cellular phone was
available, this was not used at China Lake.
Once downloaded, the data was analyzed using
HTFlow, the analysis software provided with
the VECTOR system.

Although a few probes have stopped operating
since their installation in 1999, they have
generally lasted longer than their predicted one-
year lifespan. Operation to date has resulted
only in minor data corruption, attributed to
system wear and electrical anomalies.

The probe requires a minimum five-foot
saturated thickness above it to function
properly. This has been a limitation at China
Lake, where saturated thicknesses are usually
less than 10 feet.

Technology Description [1, 4, 5, 6, 7, 8, 9]

VECTOR technology is used to measure
groundwater flow velocity (Darcy velocity). It
is a full-scale technology and VECTOR systems
are available for purchase from Hydrotechnics
Inc.  Probes are generally made to order with
lead times-on delivery ranging from one to five
weeks.

Since velocity is a vector characterized by
magnitude and direction (three dimensional in
this case), this document uses the term
"velocity" to imply both magnitude and
direction.

Following in-situ installation, the probe is field-
calibrated. Field-calibration is usually
performed just once in the lifetime of a probe
provided there is no drastic change in the
probe's environment. For instance, a 50 degree
change in groundwater temperature caused by a
local thermal influence, or a significant change
in saturated zone composition through intrusion
of another liquid may warrant re-calibration.
                                             4-1

-------
                                                 China Lake Naval Air Weapons System Site, California
Calibration includes heating the probe to a
steady temperature approximately 25 degrees
Centigrade (°C) above the ambient groundwater
temperature. As groundwater flows past the
heated probe an equilibrium temperature
distribution forms on the probe's surface. The
analysis software supplied with the VECTOR
system then correlates measured temperature
distribution to groundwater velocity. Just as the
coldest side of a wet finger held up in the air
indicates which way the wind is blowing, the
coldest side of the submerged probe indicates
the direction from which the groundwater is
flowing.

The VECTOR system is made up of several
different components: a probe, a power source,
wiring to transmit data, a data collector often
called a datalogger, and data analysis software
(HTFlow). Some of these components come
with differing capabilities depending on the
requirements of the application. The datalogger
can be connected to a maximum of 12 probes
and the power supply can be tailored to
accommodate any number of probes. Figure 1
depicts a typical VECTOR system.

The VECTOR probe is a cylinder measuring
2.375 inches in  diameter and 36 inches in length
connected to industry-standard flush-mounted
Schedule 40 PVC well pipe.  Alternating current
(110 Volt) runs a power supply which is used to
constantly heat  the probe with about 60 watts,
maintaining it at a temperature between 20 and
30 °C above ambient groundwater temperature.
An integrated array of 30 carefully calibrated
temperature sensors or thermistors forms the
outermost layer of the probe. This layer
measures temperature variations along the
surface of the probe caused by groundwater
flow. These thermistors are extremely sensitive
and can accurately measure differential
temperature to within (+/-) 0.01 °C.
The VECTOR probe has the following
measurement capabilities:

   Minimum Darcy Velocity = 0.01 feet per
   day (ft/day)
   Maximum Darcy Velocity = 1.00 ft/day
   Resolution = 0.001 ft/day
   Minimum Flow Rate Measured = 0.01 ft/day
   Max. Flow Rate Measured = approximately
   2 ft/day
   Thermistor Accuracy = (+/-) 0.01 °C
   Directional Accuracy = (+/-) 5 degrees

The probes are installed in direct contact with
saturated soil using a method of drilling that
allows uniform collapse of the lithology around
the probe. According to the manufacturer,
hollow-stem auger drilling has so far yielded the
most consistent results. Although this method
of drilling disturbs the lithology, there is the
certainty that a saturated sand lithology will
uniformly collapse around the VECTOR probe
when the auger flights are removed. Other
drilling techniques have successfully been used
for installation in deeper and more difficult
formations.  Good results have also been
obtained using mud-rotary drilling.  According
to the manufacturer, resonant sonic drilling has
the potential to replace hollow-stem auger
drilling as the method  of choice.

Groundwater flow direction measurements
made by the probe are relative to the probe's
reference direction which is carefully recorded
during installation of the probe.  Conversion of
the probe's directional measurements to
azimuths is performed by HTFlow.

The datalogger currently supplied with the
VECTOR probe is a CR10X datalogger
manufactured by Campbell Scientific.  Although
there are other similar dataloggers, the
manufacturer recommends use of the CR10X
because of the variety of remote downloading
capabilities it offers.  The CR10X datalogger
                                              4-2

-------
                                         China Lake Naval Air Weapons System Site, California
                                  Figure 1
                           Vector System Layout

Source: Hydrotechnics Inc.
                                    4-3

-------
                                                  China Lake Naval Air Weapons System Site, California
requires a simple wall-mounted 12 Volt direct
current power source for operation. To make
communication possible, an interface box
(Campbell Scientific Model SC32B) must be
used.  The data logger comes with industry-
standard software allowing a user to program
data collection schedules as well as
simultaneously manage data collection from
several probes. The datalogger can be fitted
with a modem that communicates through
landlines or cellular phones to facilitate remote
data acquisition.

Site Information [3, 8, 9]	

The VECTOR system was used at the Naval Air
Weapons Station in  China Lake, California.

The concern at China Lake was the proximity of
the site to the municipal well fields in the City
of Ridgecrest.  The closest well field was
located less than a mile south of Inyokern Road
which marks the boundary between China Lake
and the City of Ridgecrest.  Although no
contamination was found along the boundary,
the VECTOR probes were installed to monitor
the potential for off-site migration of
contaminants to safeguard against the possibility
of their detection at  a later date.

The Navy's monitoring well network included
64 wells (5 equipped with pressure transducers)
screened in the shallow WBZ, 36 wells (3
equipped with pressure transducers) screened in
the intermediate WBZ, 23 wells screened in the
deep WBZ, and 9 VECTOR probes installed at
four strategic locations along the facility's
property boundary in the shallow and
intermediate WBZs.

The probes were installed in 1999 using hollow
stem auger drilling.  They were installed 50 feet
and 180 feet bgs in the permeable (sand and
gravel) portions of the shallow and intermediate
WBZs respectively. Data collected from the
nine VECTOR probes and pressure transducers,
coupled with the water level measurements
collected quarterly from the fence line
monitoring well network facilitated the
following:

(1) Determination of the direction and rate of
   groundwater movement from the China
   Lake Complex to the City of Ridgecrest

(2) Monitoring changes in horizontal and
   vertical groundwater gradients and velocities

(3) Evaluating the interconnectivity between the
   three WBZs (shallow, intermediate, and
   deep)

According to the Owner's contractor, VECTOR
technology in concert with other monitoring
programs in  place at China Lake, provided a
cost-effective approach to monitoring horizontal
and vertical  groundwater movement (within the
different WBZs) from the China Lake Complex
to the municipal well field.

The VECTOR probes, coupled with the pressure
transducers in adjacent monitoring wells,
collected hourly measurements of groundwater
velocity and groundwater level data to provide
temporal continuity to the network of 123 wells
that were being used to monitor groundwater
flow within  an 84-square mile area.

The entire field effort for installation of the
VECTOR systems, including well drilling,
probe installation, and electrical connections
took place over a two-week period. The
VECTOR manufacturer provided installation
services  during this period. No problems were
encountered during probe installation.

The electronics were housed in weatherproof
NEMA4 enclosures next to electrical utility
poles that served as power sources. The locked
enclosure allowed limited access for
downloading data and adjusting probe power
levels. All power and control wiring to the
probes were run below ground to allow for
unimpeded site access.
                                              4-4

-------
                                                  China Lake Naval Air Weapons System Site, California
System Operation [3, 8]
Cost [5, 7]
Following initial calibration, probe operation
required no human supervision.  The datalogger
continuously logged probe data and recorded
hourly averages. Stored data was manually
downloaded for analysis every three months.
Data retrieval involved connecting the
datalogger to a laptop computer through an
interface box. The supplied software was then
used to download the data onto the laptop
computer's hard drive.  Downloaded data was
analyzed in the office using HTFlow.

According to the Owner's contractor,  the probes
operated as expected. Minor flaws in recorded
data were distinguishable and could be excluded
from analysis.  The HTFlow software was able
to calculate a "fit error," for calculated
velocities. This error was presented as an upper
and lower confidence interval bounding the
velocity curve.  It was a useful feature for
determining the quality of retrieved velocity
data.  In addition, error analysis allowed users to
separate the  few data flaws caused by electrical
anomalies from real measurements.

Being a relatively new and untested technology
(first installed in 1996), the expected lifespan
for a constantly operating submerged probe was
not known.  Although the probes were said to
have life expectancies ranging from one to two
years, nearly all probes installed at this site
operated without malfunction for at least five
years.  Recently, however, several of the probes
ceased to function.  The Owner's contractor
attributed their failure to possible electrical
problems resulting from earth moving
operations on site rather than probe malfunction.
Over the operating life of the probes, there were
only minor problems involving the data logging
and retrieval system, including some data loss
caused by power surges.
Incurred costs for installation of the nine-probe
VECTOR system at China Lake were not
available.  However, the estimated cost
(excluding well drilling) in 2004 dollars is
approximately $50,000. Table 1 presents a cost
estimate for a similar three-probe VECTOR
system. Unit prices are in 2004 dollars. This
estimate is based on the assumption that the
purchaser would have total ownership of the
equipment and all data processing. This type of
procurement is beneficial when equipment
rental costs for long-term groundwater
monitoring exceed that of outright ownership.

Based on Table 1, the  unit price for installation
was $8,145 per probe. This included material
and installation costs for the VECTOR probes
and datalogger; assuming that the well had
already been drilled. Consequently, drilling
costs were not part of this estimate.

Material costs account for the bulk (82 percent)
of the total capital cost of the technology. In
comparison, installation costs are small at only
18 percent of the total  capital cost. Although
costs for operation were not available, the main
cost components are expected to include energy
and quarterly site mobilization for data
download.

Some factors that influence capital costs are:

(1) Number of probes.  The unit cost of
   implementing this  technology is inversely
   proportional to the number of probes
   purchased.
(2) Production volume and demand for the
   VECTOR system.  High production and low
   demand would decrease costs, while low
   production and high demand would increase
   costs.
                                              4-5

-------
                                                 China Lake Naval Air Weapons System Site, California
                                          TABLE 1
                                     COST ESTIMATE
Typical System Using Three Vector Probes
Description
VECTOR In Situ Permeable Flow Sensor includes down-
hole multiplexer, lab-calibration with associated files, and
introductory technical support.
Campbell Scientific CR10X Data Logger (with 2 (two) meg
memory option, RS232 optically isolated interface, manual,
12V power supply and all com port connection cables)
60V/1 .5 A Protek DC adjustable digital readout power supply
for single-probe operation
Probe/cable packaging and handling, basic insured ground-
shipping w/2-3 week advance notice
VECTOR System assembly, down-hole installation with
driller assistance, establish reference direction, assistance
with cable burial and well head completion, all cable/power
connections, enclosure installation, and initial probe
powering for field calibration
NEMA-4 weather-proof hinged enclosure for power supply
and/or datalogger. Construction with cooling fans and back-
plate mounted instruments
HTFLOW software license (Free to US Government)
Qty
3
1
3
3
5
1
1
Unit
Each
Each
Each
Each
Days
Each
Each
Cost ($)
3,080
2,650
400
65
900
350
300
Total ($)
9,240
2,650
1,200
195
4,500
350
300

TOTAL
UNIT COST (per Probe)
24,435
8,145
  Source:  Hydrotechnics Inc.  Estimate # 1802 from Richard Fagioli
(3) Peripheral costs. Some components of the
   VECTOR system (such as the data logger)
   are manufactured by a different entity.
   Capital costs of the VECTOR system would
   therefore be influenced by price changes in
   peripherals.

According to the manufacturer, apparent, large
up-front capital costs are an obstacle to probe
use.  Though the technology might have an
initially high capital cost, the minimal operation
and maintenance costs imply increasing cost-
benefit over time.  For example, if a one-time-
only snapshot measurement of groundwater
flow were required at a site, VECTOR
technology might prove more expensive than a
three-point piezometric analysis. On the other
hand, if continuous measurements were required
over a period of time and over differing flow
scenarios, cost analysis would reveal the
benefits of using VECTOR technology over
conventional methods.

In general, the level of effort and associated
costs for implementing VECTOR technology
are less than those for setting up and performing
a week-long constant rate discharge test
representing a location of equal size.
                                             4-6

-------
                                                   China Lake Naval Air Weapons System Site, California
Conclusions [3, 5, 8,10]
Information presented in this section was
derived from conversations with the Navy's
contractor at China Lake, and the technology
vendor.

This section summarizes technology strengths,
weaknesses, and useful facts learned through
implementation at China Lake. This section
also includes useful knowledge gained from
implementation at other sites, as provided by the
vendor.

The accuracy in measurement of groundwater
flow direction is a function of the magnitude of
velocity and the geometry of flow past the
probe. HTFlow can graphically illustrate  the
potential margin of error in measured flow
direction at a given point  in time. Another
factor influencing directional accuracy is the
diligence exercised by the field-installer in
recording the compass orientation of the probe's
reference meridian.  This  is thought to be  the
greatest contributor to error in directional
measurement.  Discounting human  error, these
probes usually have (+/-)  5 degree azimuthal
accuracy at low groundwater flow velocities.
This is considered more than sufficient for
typical hydrogeologic applications.  Another
important factor in directional measurement is
magnetic declination. After the orientation of
the probe's meridian has been determined using
a compass, a correction must be applied. This is
because compass measurements are in reference
to the earth's magnetic poles, while true
azimuths are in reference  to its geographic
poles.

Continued accuracy of groundwater flow
measurement is contingent on the assumption
that groundwater temperature will remain
relatively constant.  Strictly speaking,
measurement is only valid at the temperature at
which the probe was calibrated.  As a result, one
must be cognizant of heat sources in the vicinity
that could alter groundwater temperature after
calibration.  Early tests on shallow probes
installed in very thin aquifers revealed
detrimental effects related to diurnal heating and
cooling of overlying soils. To mitigate these
effects, the probes must be installed in a
lithology that provides at least 4 feet of
saturated overburden.

The probes have never been installed in aquifers
deep enough to experience more than about 30
pounds per square inch (or 70 feet of water
column) of pressure. Therefore, it is not known
whether depth is a limiting factor in probe
function.  Though there are no theoretical limits,
the depth of installation is currently restricted by
the length of cable  connecting the probe to the
datalogger.  The present configuration allows
for a maximum cable length of about 500 feet.

Improper installation of the probe can produce
errors in measurement. Improper installation
could include a situation where the probe is in
contact with a different lithology than
presumed. This can be avoided by maintaining
accurate lithologic  logs while drilling the well.
Improper installation may also constitute a
probe encompassed by disturbed soil with
hydraulic properties differing from the
surrounding lithology. Minimizing surface
disturbances during probe installation, and
careful removal of the drilling augur can prevent
this from happening.

Factors affecting the life span of the probe are
not known.  However, the manufacturer believes
that the longevity of the thermistors used in the
probe possibly has  the greatest influence on the
probe's life span.  The manufacturer estimates a
one to two year lifespan.  However, the probes
at China Lake lasted at least five years.

VECTOR systems function best in  a wide
variety of sands.  This includes any material
with grain sizes from 0.050 mm to  1.5 mm or
U.S. Standard Sieve Series No. 270 to No. 12.
                                              4-7

-------
                                                 China Lake Naval Air Weapons System Site, California
However, usable results have also been obtained
in other lithologies.

In general, the VECTOR system has widespread
potential application. Its potential uses include
monitoring ambient groundwater flow,
monitoring the capture zone of a groundwater
extraction system, and monitoring bulk
hydraulic conductivities of aquifer media during
remedial processes that require such
measurement.

The cost-benefit of VECTOR technology is a
function of the user's needs.  According to the
manufacturer, the technology is generally
underutilized in that customers typically use it
to measure fewer parameters than it is capable
of. For example, the probe in combination with
pressure transducers in nearby wells can be used
to measure and monitor changes in bulk
hydraulic conductivity. However, if a user does
not need to measure this parameter, the probes
ability to measure it will not be considered a
benefit.

Limitations

VECTOR technology allows users to manually
regulate the power supplied to the probe; users
are expected to record any change made, in
HTFlow. Erroneous results could be produced
if a user did not record the new power level.

Any lithology that does not produce a laminar
groundwater flow past the probe's surface is
inappropriate for the VECTOR. Fractured or
solid bedrock, unsaturated sediments,
excessively high groundwater flow velocities or
grossly heterogeneous lithologies primarily
consisting of cobbles and coarse glacial deposits
will  yield poor results.  Conversely, super fine
sediments such as silts and clays will also yield
poor results because local convection will play a
greater role in heat transfer than groundwater
advection. However, thin lenses of these kinds
of sediments within a suitable lithology can be
accounted for when performing data analysis by
simply isolating those thermistors in direct
contact with the lenses.

This technology cannot be used in environments
such as open boreholes, open streams, rivers,
and open channels.

This technology is not suitable in  any thermally
heterogeneous environment such as within a
narrow permeable reactive barrier, or near a
formation with thermal properties differing from
the surrounding lithology.

Improvements

Although the VECTOR was originally designed
for emplacement in direct contact with a
saturated granular lithology, efforts are
underway to produce a tool that will measure
three dimensional flow within any packed and
backfilled open-borehole configuration.

The VECTOR manufacturer is also beta-testing
their own down-hole datalogger (VeComm)
system for use with the probes.

Contact Information	

Owner
Department of the Navy
Southwest Division
Naval  Facilities Engineering Command
Remedial Project Manager
Mr. Mike Cornell
Phone: (619)532-4208
Email:  michael.j.cornell@navy.mil

State Oversight Agency
Department of Toxic Substances Control
Project Manager
Ms.  Laurie Racca
Phone: (916)255-3668
Email: LRacca@dtsc.ca.gov
                                              4-8

-------
                                                China Lake Naval Air Weapons System Site, California
Owner's Contractor:
Tetra Tech EM Inc.
Project Manager
Mrs. Kathy Monks
Phone: (505) 881-3188 ext 101
Email: kathy.monks@ttemi.com

Technology Vendor:
HydroTechnics, Inc.
P.O. Box 92828
Albuquerque, NM 87199
Mr. Richard Fagioli
Phone: (505)797-2421
Email: richf@hydrotechnics.com

References	

The following references were used in the
preparation of this report:

1.  Campbell Scientific, Inc. CR1 OX
   Measurement and Control Module -
   Operators Manual. Revision 9/01.  2001.
2.  Tetra Tech EM Inc.  1999. Proposed
   Revised Approach for Hydrologic
   Investigation Along Boundary Fence Line,
   Naval Air Weapons Station, China Lake,
   California.  March.
3.  Telephone Conversation. Kathy Monks,
   Tetra Tech EM Inc., with Chitranjan
   Christian, Tetra Tech EM Inc., China Lake
   Response to Questions on VECTOR
   Technology. April 28, 2004.
4.  Telephone Conversation. Craig Knox,
   Campbell Scientific, Inc., with Chitranjan
   Christian, Tetra Tech EM, Inc., Response to
   Questions on CR10X datalogger. May 20,
   2004.
5.  E-mail from Richard Fagioli, Hydrotechnics
   Inc., to Chitranjan Christian, Tetra Tech EM
   Inc., Response to Questions on VECTOR
   technology. May 21, 2004.
6.  E-mail from Richard Fagioli, Hydrotechnics
   Inc., to Chitranjan Christian, Tetra Tech EM
   Inc., Response to Additional Questions on
   VECTOR technology. May 26, 2004.
7.  E-mail from Richard Fagioli, Hydrotechnics
   Inc., to Chitranjan Christian, Tetra Tech EM
   Inc., Cost Estimate for VECTOR
   technology.  June  14, 2004.
8.  E-mail from Kathy Monks, Tetra Tech EM
   Inc., to Chitranjan Christian, Tetra Tech EM
   Inc., Response to Questions on China Lake.
   June 11,2004.
9.  Worldwide web resource. Hydrotechnics
   Inc.  http://www.hydrotechnics.com!
10. Monks, K. and M.Godwin.  2001.  "A Cost-
   Effective Approach to Multi-Parameter
   Hydrologic Monitoring to Characterize
   Groundwater Flow Conditions." 2001
   International Containment & Remediation
   Technology Conference and Exhibition.
   DOE/EM-0620. http://www.em.doe.gov.

Acknowledgements	

This report was prepared for the U.S.
Environmental Protection Agency's Office of
Solid Waste and Emergency Response, Office
of Superfund Remediation and Technology
Innovation. Assistance was provided by Tetra
Tech EM Inc. under EPA Contract No.
68-W-02-034.
                                            4-9

-------

-------
                   Appendix 5

        Automation Technology Case Study
Programmable Logic Controllers and Ozone Analyzers
         at the Moffett Federal Airfield Site

-------

-------
                         Automation Technology Case Study
               Programmable Logic Controllers and Ozone Analyzers
                           at the Moffett Federal Airfield Site
Summary Information [3]
Ozone analyzers and a programmable logic
controller (PLC) are used in an ozone/hydrogen
peroxide groundwater treatment system to
monitor system performance. This case study
presents the use of real-time ozone analyzers in
conjunction with a PLC in a pump and treat
system. The ozone analyzers work in
conjunction with the PLC to ensure that (1) the
correct dosage of ozone is applied to the influent
water, (2) the offgas treatment system is
meeting the air emission standards, and (3) that
the ambient air meets occupational safety and
health administration  (OSHA) standards.

In addition to the ozone analyzers, this case
study presents the use of a PLC as well as a
supervisory control and date acquisition
interface. The supervisory control portion of the
system consists of a PLC, which receives inputs
from various sensors and instruments and
operates the treatment system based on a set of
programmed instructions. The data acquisition
portion of the system  consists of remote
computers, the human machine interface (HMI),
which stores all system operational data to assist
the system operators during system
troubleshooting, data logging for National
Pollutant Discharge Elimination System
(NPDES) and other reports, and system
optimization.

This case study is based on the treatment system
as initially designed and operated.  Subsequent
process modifications occurred, however they
are not relevant to the PLC or ozone analyzers
discussed in the case study.

Air strippers have commonly been used to treat
TCE contaminated groundwater in pump and
treat remedies. Air strippers remove
contamination from the liquid phase and transfer
 it to the gas phase. Regulations do not always
 require treatment of this gas phase prior to
 discharge. Concerns about TCE emissions from
 phase-transfer processes have made destructive
 treatment options more prevalent. The ex situ
 ozone advanced oxidation process discussed
 here is one such treatment option to reduce
 atmospheric TCE  emissions.

 Additionally, ozone has been increasingly
 applied via sparging to oxidize groundwater and
 soil contaminants in situ. Monitoring of ozone
 emissions from these systems is important both
 for system optimization and worker safety.

 Based on information received from the vendor,
 the ozone analyzer costs range from $2,500 to
 $15,000 and are generally reliable and require
 little maintenance.  Typical problems result
 from nuisance shutdowns due to high levels of
 ambient ground level ozone.

 Technology Description [4]

 PLCs enable facility automation.  However,
 they depend on field devices for sensory input.
 The ozone analyzer is one such device used at
 the Moffett Federal Airfield (MFA).

 PLC

 A PLC is a computer system that monitors
 inputs, makes decisions based on its program,
 and controls outputs to automate a process or
machine. PLCs consist of input modules or
points, a Central Processing Unit (CPU), and
output modules or points.  Input modules accept
a variety of digital or analog signals from
various field devices (such as level sensors or
ozone analyzers) and converts them into a logic
signal that can be used by the CPU. The CPU
makes decisions and executes control
instructions based on program instructions in
                                             5-1

-------
                                              Moffett Federal Airfield Site, Santa Clara County, California
memory. Output modules convert control
instructions from the CPU into a digital or
analog signal that can be used to control various
field devices (such as valves or switches). PLCs
can be networked with other computer systems
for data acquisition or display.

Ozone Analyzer

Ozone analyzers are devices that can be used to
measure ozone concentration in gaseous media.
The media sampling system consists of inlet
needle valves for sample and zero gas
collection, a flow meter, a solenoid valve, and a
sample chamber. The zero gas is collected from
atmospheric air that passes through a scrubber
containing an ozone destruction catalyst to
remove any background ozone concentration.
The unit measures ozone concentration by
comparing the absorption of the sample with the
zero gas. Depending on the position of the
solenoid valve, zero or sample gas is forced
through the solenoid valve, the sample chamber,
and flow meter. The intensity of the ultraviolet
light (UV) traversing the sample chamber is
attenuated as prescribed using the Beer-Lambert
Law, that is, the presence of ozone reduces the
intensity of the UV light, the amount of UV
light reduction corresponds to the ozone
concentration. The ratio of the intensities is
determined and the results are processed by the
microcomputer to determine the ozone
concentration. Since the concentration
determined by the photometer is based solely on
the ratio of light intensities, the actual intensity
of the light is not important. Slow changes in
light intensity due to lamp aging or dirt buildup
in the optics will not affect the concentration
reading.

Site Information [1, 2, 3, 5, 6]

MFA is located in Santa Clara County,
California, at the southern end of San Francisco
Bay.  From the 1930s until its closure in 1994,
the base was operated by the Navy as Naval Air
Station Moffett Field.  In 1994, the base was
transferred to the National Aeronautics and
Space Administration (NASA) and continues to
operate as a federal airfield. MFA was placed
on the national priorities list (NPL) in 1987.

In 1991, the EPA and other agencies divided
MFA into six operable units (OUs).  OU4
consisted of the west-side aquifers. These
aquifers were contaminated with a mixture of
gasoline and diesel from several leaking USTs,
as well as tetrachloroethylene (PCE) and its
breakdown products from an on-site dry cleaner.
However, in 1992, the EPA determined that the
west-side aquifers were affected by a regional
plume (primarily TCE and its daughter
products) that emanated from the adjacent
Middlefield Ellis Whisman (MEW) NPL site.
As a result, EPA determined that the west-side
aquifers were subject to the 1989 record of
decision (ROD) already written for the MEW
site, directing remediation of the west-side
aquifers.  Consequently, OU4 was deleted and
has since been referred to as the west-side
aquifers.

The west-side aquifers consisted of two highly
heterogeneous interconnected aquifer zones of
alluvial channel deposits composed of sand and
gravel incised in and interbedded with clayey
floodplain deposits.  The aquifers extended from
5 to 65 feet below ground surface.

The MEW site ROD designated pump-and-treat
with air stripping as the remedy. A subsequent
explanation of significant differences in 1996
allowed the use of liquid-phase granular
activated carbon (GAC). Design of the west-
side aquifers treatment system (WATS) was
completed in 1997 and construction was
completed in 1998. Table 1 lists the
contaminants, concentrations, and treatment
requirements for extracted water at the  WATS.
                                              5-2

-------
                                            Moffett Federal Airfield Site, Santa Clara County, California
                                          Table 1
                            Influent Contaminant Concentrations
                            West-Side Aquifers Treatment System
Contaminant
Trichloroethene (TCE)
Tetrachloroethene (PCE)
Vinyl Chloride
1 , 1 -Dichloroethene
(1,1-DCE)
Cis- 1 ,2-Dichloroethene
(1,2-DCE)
Trans- 1 ,2-Dichloroethene
(1,2-DCE)
1 , 1 -Dichloroethane
(1,1 -DC A)
1,1,1 -Trichloroethane
(1,1,1-TCA)
Freon 113
TPH-purgeable
TPH-extractable
Average Influent
Concentration
(HS/L)1
1,850
24
31
26
270
10
6
1
21
180
11
Federal MCL
(HS/L)
5
5
2
7
70
100
N/A
200
N/A
N/A
N/A
California MCL
(Hg/L)
5
5
0.5
6
6
10
5
200
1,200
N/A
N/A
NPDES
Treatment
Requirement
(HS/L)
5
5
0.5
5
5
5
5
5
5
50
50
Notes:
MCL
N/A
NPDES
TPH
Based on operational sampling data from December 1998 through March 2000
Micrograms per liter
Maximum contaminant level
Not applicable—no MCL for this parameter
National Pollutant Discharge Elimination System
Total petroleum hydrocarbons
Source:  Tetra Tech EM Inc.
                                             5-3

-------
                                             Moffett Federal Airfield Site, Santa Clara County, California
Remediation System

The WATS consists of eight extraction wells
piped to a treatment system located north of
Building 45. All system equipment is controlled
by a PLC. The extraction wells maintained a
constant groundwater drawdown to maximize
extraction rates from each well. Contaminated
water collected in two  on-site sumps near
Hangar 1 was also treated in the WATS.
Contaminated water was pumped from the
extraction wells and treated to remove
groundwater contaminants to levels specified in
the MFA NPDES permit before being
discharged to the MFA storm sewer.

Extracted groundwater was first passed through
bag filters, which remove  sediment from the
influent.  Then, ozone  and hydrogen peroxide
were mixed with the water in three 1,400-gallon
reaction tanks (ozone and  hydrogen peroxide
combine to form hydroxyl radicals that are
stronger oxidants than  ozone and hydrogen
peroxide alone.  The hydroxyl radicals destroy
about 99 percent of the influent contaminants).
The remaining contaminants were removed by
the air stripper and  GAC units and the treated
water was discharged to the MFA storm sewer
under an NPDES permit.  Offgas from the
oxidation tank, consisting  of a mixture of
unreacted ozone and oxygen, was heated and
treated with a proprietary ozone destruction
catalyst. Figure 1 presents a simplified process
flow diagram of the WATS.

Control System

PLC control and sensors served five main
functions on the system, (1) they allowed the
system to operate automatically, (2) they shut
down the system if water or offgas treatment
was not functioning correctly, (3) they shut
down the system before an unsafe worker safety
condition, such as an ozone leak, existed,
(4) they allowed remote monitoring of system
status, allowing the operator to schedule site
visits, and (5) they logged operational data,
which could then be mined and trended to
troubleshoot problems or optimize system
performance. A pump-and-treat system of the
complexity at the WATS would be extremely
difficult to operate manually or even with a
solid-state control system.

In order to maintain a constant groundwater
drawdown, a capacitance-type level probe  was
used in each extraction well to continuously
monitor the water level and relay this data  to the
PLC. The PLC then sent a signal to a flow
control valve.  If the water level was rising, the
flow control valve opened slightly to increase
the pumping rate, and thus the drawdown.
Conversely, if the water level was decreasing,
the flow control valve closed slightly to
decrease the pumping rate. This automatic
adjustment of well flow rate adjusted for
changes in well yield due water table
fluctuations caused by precipitation events.

The PLC was also used to regulate ozone and
hydrogen peroxide dosage which was critical to
obtaining maximum destruction efficiencies.
This was accomplished using an influent
flowmeter which measured the flowrate into the
treatment system.  The PLC processed this
information and calculated the corresponding
hydrogen peroxide injection rate and sent a
signal to the hydrogen peroxide pump to inject
the correct dosage. The gas flow rate into the
oxidation tanks was constant, but ozone
percentage of gas varied to achieve the proper
ozone/hydrogen peroxide ratio for optimal
treatment. Thus,  the PLC also processed the
water flow rate, calculated the correct ozone
concentration, and sent a signal to the ozone
generator which converted oxygen into the
proper percentage of ozone sparged into the
tanks to maintain the correct dosage.
                                             5-4

-------
                                                       Moffett Federal Airfield Site, Santa Clara County, California
                                                    Figure 1
                                    West-Side Aquifers Treatment System
     Air Stripper
     Ozone
     Analyzers
        To Granular
   Activated Carbon
Filters and Discharge
     Air
     Compressor
                                                                                                Ozone/Peroxide
                                                                                                Reaction Tanks
    Ozone
    Generator
From Bag Filters and
Extraction Wells
                                                                                                  Oxygen
                                                                                                  Generator
  Ozone/Peroxide
  Reaction Tanks
    Hydrogen
    Peroxide
    Tank
    Air Stripper
   PLC Enclosure
                                                                                                 Granular
                                                                                                 Activated
                                                                                                 Carbon Vessels
                                                       5-5

-------
                                              Moffett Federal Airfield Site, Santa Clara County, California
The universal shutdown interlock programmed
into the PLC shut down the entire WATS,
including the eight extraction pump motors,
based on activation of any of the following
system alarms:
       High-high bag filter differential pressure
       (indicates that the bag filters are
       clogged)
       Low hydrogen peroxide flow (indicates
       that the hydrogen peroxide pump has
       failed)
       Low hydrogen peroxide tank level
       (indicates that the hydrogen peroxide
       tank is nearly empty)
       Air compressor general fault
       Air receiver low pressure
       Low ozone generator voltage
       Ozone generator  fault
       Oxygen low flow
       Low oxidation tank inlet ozone
       concentration (ensures that correct
       dosage of ozone is injected into the
       oxidation tanks)
       High-high ozone  generator discharge
       temperature (prevents the ozone
       generator from overheating)
       Low catalytic oxidizer temperature
       (indicates that the ozone destruction
       system is not functioning)
       Low-concentration ozone analyzer
       general failure (indicates that the
       analyzer has malfunctioned)
       High catalytic oxidizer exhaust ozone
       concentration (indicates that the ozone
       destruction system is not working
       properly)
       High air stripper sump level (indicates
       that the air stripper drain is clogged)
       Low air stripper blower pressure
       (indicates that the air  stripper is not
       functioning)
       High-high secondary  containment sump
       level (indicates that a  spill has occurred
       within the treatment pad)
These alarms were selected to activate the
universal shutdown interlock because they
indicate that a spill could occur, that the system
may not be treating the water effectively, or that
the offgas treatment system may not be working
correctly.  Immediately shutting down the
system in response to any of these alarms would
avoid spills, the discharge of untreated water, or
the discharge of untreated offgas.

An ozone analyzer continuously monitored the
ozone concentration in the ambient air. If an
ozone leak occurred, it could immediately
endanger the health of any on-site worker.
Thus, when the ozone analyzer detected an
ambient air ozone concentration of 0.1 ppm, the
entire system (including the ozone generator,
presumably the source of the ozone) would shut
down, well before the OSHA permissible
exposure limit of 0.1 ppm could be exceeded for
an 8-hour time-weighted average (TWA).

Data Acquisition

The WATS PLC was connected via modem to
remote PC-based human-machine interfaces
(HMI) at offsite locations.  This allowed the
system operator to view  system status. Thus,
some troubleshooting could be conducted from
offsite. Additionally, the operator could
remotely view the status of various parameters,
such as bag filter pressure drop, and hydrogen
peroxide volume to remotely schedule bag filter
changeouts and chemical deliveries
respectively.  Finally, the operator could
remotely turn extraction wells on or off as
needed. Figure 2 presents the HMI for a typical
extraction well that shows the current pump
status (auto), pump discharge pressure (17.27
psi), flow rate (5.91 gpm), and water level -1.36
feet mean sea level).

The HMI logged all system parameters, such as
flowrates,  extraction well water levels, and
pressures,  on a remote computer. This
information was useful in the preparation of the
                                              5-6

-------
                                       Moffett Federal Airfield Site, Santa Clara County, California
                                    Figure 2
                          Human Machine Interface
WELL EXWt WATER
 LEVEL CONTROL
                 HCWEXW1
WELLEXW1     TOIAt FI.CW WEIL EWVl - ! S!T2,9*>4 ti it-V.
                                PUMPEXW1
                                                                        EWELEXW1
                          Extraction Well Status Screen
                  4 j 04 rr
                        Process Variable Trending Screen
                                                                                WHfR
                                        5-7

-------
                                             Moffett Federal Airfield Site, Santa Clara County, California
NPDES reports and quarterly progress reports,
where individual well flow rates could be
plotted for each reporting period.  The
extraction well water level data could be trended
over any user-defined period. Figure 2 displays
a typical process variable trending screen on the
HMI.

The datalogging also maintained a record of all
alarms and the causes of all system shutdowns,
which significantly simplified system
troubleshooting.  Otherwise, with  such a
complex system, it could often be difficult to
determine what equipment failure or process
variable initiated the system shutdown
sequence.  Reviewing the data logs allowed the
operator to quickly identify the initial cause or
causes of the problem.

Design of the WATS began in 1996, before use
of the internet was widespread. Thus, the HMI
was originally connected to the PLC via
modem. In 2001, the HMI was upgraded to an
internet-based version which allowed easier
access to the system operating conditions.  The
internet-based version had a firewall and was
password-protected to prevent unauthorized
system modifications.

Ozone Analyzers

The most complex of these sensors were the two
ozone analyzers. There were two ozone
analyzers (Figure 3), a high concentration ozone
analyzer and a low concentration ozone
analyzer. Both worked in the same manner.

The ozone analyzers (PCI Model LC-400 and
HC-400) provided a digital readout of ozone
concentrations. The high concentration ozone
analyzer measured percent ozone  (0 percent to
15 percent) in oxygen from the ozone generator.
The low concentration ozone analyzer measured
ozone in ppm from 0.001 ppm to  10 ppm.  The
high concentration ozone analyzer measured the
actual ozone concentration of the  gas that was
sparged into the oxidation tanks. The PLC
compared this value to the target ozone
concentration based on the influent water flow
rate. If the concentration was not within 10
percent of the target ozone concentration, the
system shut would down before partially-treated
water was released.

The offgas ozone analyzer measured the ozone
concentration in the system offgas. The Bay
Area Air Quality Management District
(BAAQMD) limit for nonpermitted ozone
emissions was  1 pound per day.  The offgas
ozone analyzer monitored the ozone
concentration in the system offgas. When the
offgas ozone concentration exceeded 9 ppmv
(based on the maximum concentration the unit
can detect without damage), the PLC would shut
down the entire system. At 9 ppm, the WATS
ozone emissions would be about 0.026 pound
per day.

System Operation [7]	

The pump and treat system has been in
operation since 1998 and has treated
approximately 190,000,000 gallons of
contaminated water. The PLC control system
has not experienced significant problems and
has not required maintenance.

The ozone analyzers required little maintenance.
The units were factory-calibrated.  For reasons
stated previously (that is, always comparing the
UV absorption by sample to absorption by a
zero gas), the units did not require field
calibration.  However, there were some minor
maintenance items. First, the zero  gas and the
sample gas particulate filters required
replacement every 6 months (this depended on
the site gas quality). Second, the UV bulbs
required annual replacement. An internal check
warned the operator when UV bulbs were
nearing the end of their operational life.
                                              5-S

-------
                                             Moffett Federal Airfield Site, Santa Clara County, California
                                           Figure 3
                                       Ozone Monitors
The gas scrubber that produced zero gas for the
ozone analyzers did not required maintenance.
However, future maintenance could be required
depending on exposure of the zero gas scrubber
to ozone and potential catalyst poisons.

Both ozone analyzers were replaced with newer
models in fall 2004 as replacement parts became
hard to find. The high concentration ozone
analyzer was replaced with a Mini-HiCon
manufactured by IN USA, Inc.  The low
concentration analyzer was replaced with a
Series 930 ozone analyzer manufactured by
Aeroqual Ltd.

Cost [3, 6, 7]	

According to the construction contractor, the
instrumentation and controls system installation
cost was $95,454 (Tetra Tech 2001). However,
this number may not reflect the total control
system installation cost, some of which is likely
lumped in with the $319,060 system startup
cost. Tetra Tech estimates that the actual
installation and startup cost for the
instrumentation and control system was
approximately $200,000.  The ozone analyzers
cost approximately $15,000 at the time of
installation.  However, costs have decreased
considerably since then and now range from
$2,500 to $15,000 depending on sensitivity of
the instrument, the input/output options, and the
type of enclosure (NEMA 3, NEMA 4X, etc.)

Operation and maintenance cost for the PLC
and HMI system are minimal. Once the
software license is purchased and the system is
programmed, installed, and debugged, O&M
costs are similar to the costs of operating several
PCs with internet access.

System changes or upgrades can be more costly,
especially when a programming subcontractor is
unfamiliar with the system and must learn the
system operations as well as familiarize
himself/herself with the program.

Ozone analyzer O&M costs are approximately
$500 per year in parts and approximately 16
hours in labor.

Lessons Learned [1, 2, 3, 5, 6]

The primary troubleshooting problem with the
offgas ozone analyzer occurred during system
startup and shakedown when the unit began
displaying extremely variable ozone
concentrations (from 0.00 ppmv to exceeding
                                              5-9

-------
                                             Moffett Federal Airfield Site, Santa Clara County, California
100 ppmv, beyond the range of the instrument).
It was determined that the zero gas ozone
destruction catalyst had become poisoned (due
to hydrochloric acid in the oxidation tank
off gas). Thus, the zero gas actually contained
ozone. Replacement of the catalyst remedied
the problem.

Another problem with the ozone analyzer
resulted from high ground-level ozone
concentrations due to normal sources, such as
automobile exhaust. The ambient air ozone
analyzer is designed to detect leaks in the ozone
injection system and shut the system when
worker safety is at risk. However, using the
OSHA 8-hour TWA as the system shutdown
criteria would potentially delay system
shutdown during a catastrophic leak.  To
prevent such a delay in system shutdown, the
more conservative National Institute for
Occupational Safety and Health (NIOSH)
instantaneous recommended exposure limit of
0.1 ppm was chosen as the system shutdown
criteria. Using this instantaneous criterion, the
system would respond immediately to a leak in
the ozone piping or failure of the ozone offgas
destruction.

However, during the first few years of system
operation, numerous system shutdowns
occurred because of high ambient ozone
concentrations.  The offgas ozone destruction
system was found to be functioning correctly
and leak testing of all ozone piping (via pressure
testing and soap) did not identify any leaks.
After further investigation, it was determined
that the cause of the shutdowns was elevated
background ozone concentrations, which are
common urban areas. In fact, in  1998, the
maximum instantaneous ozone concentration in
Santa Clara County was 0.15 ppm, and  the
highest 8-hour TWA was 0.11 ppm.

To balance the need for minimizing nuisance
system shutdowns with worker safety, the
ambient ozone system shutdown setpoint was
adjusted from a 0.1 ppm for one instantaneous
reading, to 0.1  for a 1-minute rolling average
(approximately 2 readings). This change
eliminated nuisance shutdowns while
maintaining worker safety.

Conclusions [5, 6, 7]	

PLC system control and instrumentation at the
WATS has been a success to date in that there
have been no exceedences of the BAAQMD
ozone discharge requirements or exceedences of
the NPDES permit due to incomplete oxidation
of the chlorinated solvents, and there have been
no spill events.  Additionally, system
troubleshooting has been greatly simplified by
studying the datalogs of system shutdown.

However,  programming the HMI requires
advanced programming knowledge of the
system software (Wonder Ware). Making
minor changes to the HMI often required a
significant effort and could be expensive as it
required specialized contractors. Newer
versions of HMI software have  incorporated
more user-friendly programming tools that
should allow minor changes to the HMI to be
made more easily.

During system startup, Tetra Tech conducted
proof of performance testing of the ozone offgas
analyzer using a handheld ozone analyzer
(Bionics TG-800).  The handheld monitor uses
an entirely different method for ozone detection
(gas membrane galvanic cell) and its range of
detection was substantially different (0 to 2
ppmv, in 0.05 ppmv increments) than the low
concentration ozone analyzer.  Based on a
comparison of results, Tetra Tech was able to
determine  that the analyzers were accurate to
within 0.05 ppmv.
                                             5-10

-------
                                            Moffett Federal Airfield Site, Santa Clara County, California
Contact Information
State Agency:
San Francisco Bay Regional Water Quality
Control Board
Remedial Project Manager
Ms. Adriana Constantinescu
Phone: (510)622-2353
Email: avc@rb2.swrcb.ca.gov

Owner:
Department of the Navy
Base Realignment and Closure Program
Management Office West
Ms. Andrea Espinoza
Phone: (619)532-0911
Email: andrea. espinoza@navy. mil

Owner's Design Contractor:
Tetra Tech EM Inc.
Project Engineer/Project Manager
Mr. David J. Berestka, P.E.
Phone: (303)312-8856
Email: david.berestka@ttemi.com

Owner's O&M Contractor:
Tetra Tech FW,  Inc. (formerly Foster Wheeler
Environmental)
Technical Lead
Mr. Michael Klosky, P.E.
Phone: (770)825-7144
Email: mklosky@ttfwi.com

References	

The following references were used in the
preparation of this report:

1.  Bay Area Air Quality Management District
   (BAAQMD). 1997.  Letter Regarding
   West-Side Aquifer Treatment System.  From
   Robert E.  Cave, Air Quality Engineer II,
   Permit Services Division. To Stephen Chao,
   BRAC Environmental Coordinator,
   Engineering  Field Activity West,
   Department of the Navy. March 10.
2. California Air Resources Board (CARB).
   1999. Summaries of Ozone Data (1980-
   1998), Santa Clara County. March 26.
   http://www. arb. ca.gov/aqd/OLDozone/
   AlC43.htm
3. PCI Ozone & Control Systems. 1993.
   Ozone Monitor Operating and Instruction
   Manual, Models LC-400, HC-400.
   November.
4. Siemens Energy & Automation, Inc.
   Siemens Technical Education Program
   (STEP) 2000 series.  Basics of PLCs.
5. Tetra Tech EM Inc.  (Tetra Tech). 2000.
   Final Operation and Maintenance Manual
   for the West-side Aquifers Treatment
   System. DS.0310.14581-01.  October 20.
6. Tetra Tech. 2001. Draft Final Interim
   Remedial Action Report, West-side
   Aquifers Treatment System (WATS),
   Moffett Federal Airfield, Moffett Field,
   California. DS.0310.14585-01. January 12.
7. E-mail from David Berestka, Tetra Tech EM
   Inc., to Chitranjan Christian, Tetra Tech EM
   Inc., Draft Case Study. November 9, 2004.

Acknowledgements

This report was prepared for the U.S.
Environmental Protection Agency's Office of
Solid Waste and Emergency Response, Office
of Superfund Remediation and Technology
Innovation.  Assistance was provided by Tetra
Tech EM Inc. under EPA Contract No.
68-W-02-034.
                                            5-11

-------

-------
                         Appendix 6

                 Sensor Technology Case Study
Automated Sampling and Analysis of Trichloroethene and Hexavalent
               Chromium Using the Burge System
       at the North Indian Bend Wash and Nevada Test Sites

-------

-------
                             Sensor Technology Case Study
 Automated Sampling and Analysis ofTrichloroethene and Hexavalent Chromium
   Using the Burge System at the North Indian Bend Wash and Nevada Test Sites
Summary Information [1, 2, 3, 4, 5, 6, 7, 8, 9,
,10]	

The research and development (R&D) of
automated sampling and analysis systems
described in this report began in 1988 and
initially focused on a device called an optrode
developed by Lawrence Livermore National
Laboratory, for the analysis of volatile
chlorinated hydrocarbons.

As the etymology of "optrode" suggests; it is an
instrument that provides a pathway for light.
Although sometimes incorrectly used in a
broader sense, the term optrode was originally
meant to refer to a colorimetric instrument that
measured a rate of change in light intensity as
the compound of interest (analyte) diffused into
the optical pathway through the instrument's
semi-permeable membrane. Following beta
testing, Burge Environmental made the optrode-
based analytical system available to the
environmental industry in March  1998.  The
system was only able to detect trichloroethene
(TCE).  Since then the vendor has developed the
skeletal technology to permit the use of other
sensors to measure various analytes.

There are several sensors today that use
analytical methods suited to specific, or
sometimes an entire range of analytes.
However, they sometimes never make it past the
R&D stage due to the lack of a suitable platform
for field deployment. This report discusses a
modifiable automated sampling and analysis
platform developed by Burge Environmental.
The report focuses on two systems in particular.
One is an optrode based system developed for
the analysis of TCE, and the other is a non-
optrode based colorimetric system developed
for the analysis of hexavalent chromium
(Cr-VI).
Case studies on these systems cover
implementation of this technology at two
different sites: (1) The North Indian Bend Wash
(NIBW) Superfund site; and (2) The Nevada
Test Site (NTS).  The implementation at the
NIBW site involved sampling the influent and
effluent of a groundwater treatment plant and
analyzing  the samples for TCE on a daily basis.
The field effort at the NTS was a pilot test
preceding  deployment of the Cr-VI sampling
and analysis system for groundwater monitoring
at the Hanford site along the Columbia River
near Richland, Washington.

Implementation of this technology at both sites
proved to be satisfactory. Lessons were learned
along the way and improvements have been
made.

This technology is potentially useful in
situations requiring frequent sampling and
generation of large amounts of data.  Since this
technology has not yet been approved by the
U.S. Environmental Protection Agency (EPA)
for  regulatory monitoring, it has limited
application in compliance monitoring.
However,  it has been used in the past to
augment regulatory monitoring for groundwater
and treatment processes.

Technology Description [1, 3, 5, 6]	

The Burge system facilitates automated
sampling and on-site analysis for monitoring of
groundwater or treatment processes.  The
system was specifically designed for
deployment of sensors in the field.

The system essentially consists of two separable
modules; a sampling module and an analytical
module. As their names suggest, the sampling
module is responsible for sample acquisition,
                                            6-1

-------
                                                     North Indian Bend Wash Site and Nevada Test Sites
and the analytical module quantifies the amount
of analyte present in the acquired sample. The
system also has a waste module that manages
wastes produced during analysis and system
purging.  The entire system is powered by a 12
Volt energy source.

Sampling Module

The sampling module is built to suit the
sampling effort.  For instance, the module used
to sample the effluent of a wastewater treatment
system might consist of a peristaltic pump with
associated tubing for sample conveyance. A
groundwater sampling module on the other hand
may be significantly more complex especially if
it involved sampling at multiple depths. The
sampling module would be built to  fit into or
just above the well casing and would include
valves, pneumatic tubing, conveyance tubing,
and electrical wiring.

In general most sampling modules are
pneumatically actuated.  Compressed air serves
as the driving force to convey liquids from one
point to another. High pressure air  (15 pounds
per square inch [psi]) drives sample collection.

Analytical Module

Figure 1 presents an illustration of a well-
sampling and analysis module.  The illustration
shows the sampling module inside the well and
the analytical module to its right in  a subsurface
enclosure.  The enclosure also contains the
battery power source shown connected to the
solar, cell on the surface. The waste module and
compressed air tank are seen to the  left of the
well.

The analytical module is in essence a miniature
laboratory small  enough to fit in a two-foot
cube.  The module can be configured for use
with multiple sensors. In practice, the module
may be located within the monitoring well,
adjacent to the monitoring well, or both.  The
analytical module houses interconnected vials
used to hold samples, store reagents, and
facilitate proper chemical reaction.  Low
pressure compressed air (1.5 psi) drives fluids in
the analytical system from one stage to the next.

Although the specifics of the analytical system
vary with the analyte, the module in general
facilitates reagent component storage, reagent
blend preparation, sample preparation, sample
dilution, reactant mixing, system flushing, and
measurement using a sensor. Reagent storage
containers typically vary in size  from 250 to 500
milliliters.  However the amount of reagent
stored can be increased to reduce the frequency
of reagent replacement.  Assuming one sample
per day, reagent storage is usually designed to
last at least 6 weeks.

According to Burge Environmental, the
analytical platform is flexible enough to even
allow the integration of sensors manufactured
by other entities. Some of the analytical
methods available for use are: (1) molecular
spectroscopy; (2) electrochemical measurement;
(3) colorimetric measurement, and (4) photo-
ionization.

The analytical process consists of three
important stages:  (I) instrument calibration;
(2) sample analysis; and (3) post-test calibration
check.

Instrument calibration  is generally performed
before the first analysis, after which calibration
is performed manually (from the remote
computer) only as required.  For instance, re-
calibration may be necessitated by a change in
the ambient temperature recorded by the
instrument's thermometer. Calibration involves
analysis of samples bearing known
concentrations of the analyte. Analytical results
are compared with the known concentrations
and the instrument is adjusted to minimize the
difference between the two.   Samples used for
calibration represent a minimum (zero), a
                                              6-2

-------
                                    North Indian Bend Wash Site and Nevada Test Sites
                           Figure 1
       Burge Groundwater Sampling and Analysis System
^JfpPlJlJIIfc • __ :
i / V, ••!,»„„.-''
V^Wtei, / -.•••-
AIR ASSEMBLY





Well Casing 	 »•




,
Sampling Module /
3.5" X 22" J
v


.*^
Tubing to Collect *g^——-
Multi-Level Samples ^^-^^
h
"'""&











•
' i

i*H
j


__->-
^











VI!
I
In
,J




k
Solar ,/.:
is" I Cell ft
~~~ 	 *"! IM
iff , Battery
!\ L
u r
rS
JIM AnalyUcal Module
v j 7.5" X 26"
• i-fl,
Cabling With
Electric Water
and
Nitrogen Lines

> ApproximateV 2' to 3'



Surface support equipment (solar cells,
battery, compressor, air tank, etc.) may
be housed in (or attached to) a 2' X 2'
X4r field box.


Source:  Burge Environmental
                             6-3

-------
                                                     North Indian Bend Wash Site and Nevada Test Sites
maximum (varies), and a mid analytical range.
The analytical range is designed such that the
required detection range for the analyte falls
between the minimum and the maximum, and is
bisected by the mid. The analytical module
produces the minimum-, maximum-, and mid-
calibration samples from varying blends of
stored standard solution and prepared blanks.
Blanks are analyte-free samples that are made in
the analytical module by passing contaminated
water through the appropriate treatment
medium. The treatment medium in the TCE
analytical module is activated carbon, and in the
Cr-VI analytical module is granular ferric
hydroxide. All process wastes are pumped to a
waste storage tank.

Sample analysis is a two or three stage process
involving the transfer of a metered volume of
prepared sample to the analytical cell followed
by measurement using a sensor.

To verify the validity of the result of analysis,
the module performs a post-test mid-calibration
check. This involves analysis of a prepared
mid-calibration sample soon after analysis of the
acquired media sample.  If the result of post-test
mid-calibration shows more than 20 percent
deviation from the true value, the module marks
the previous result for review.

TCE Analysis

Analysis for TCE involves a colorimetric
measurement technique that uses a device called
an optrode. The optrode consists  of a U-tube, a
light source and a collector. The U-tube is
constructed of a semi-permeable material with a
reflective inner surface.  Being in essence an
optic fiber, the u-tube allows the passage of
light from one end to the other with
undiminished intensity. A green light emitting
diode  (LED) at one end of the u-tube serves as
the light source and a collector at the other end
measures the intensity of exiting light.  The
optrode is located in the head space (volume
above liquid level) of the sample vial.

Prior to every analysis, the optrode is filled with
a freshly prepared solution of a TCE-specific
reagent.  When an aqueous sample is introduced
into the vial containing the optrode, an
equilibrium results between the liquid and vapor
phases of volatile contaminants in that sample.
Vapor in the headspace permeates the wall of
the optrode and enters the reagent. The reagent
reacts with TCE forming a red colored product.
This product absorbs light produced  by the
green LED as it passes through the u-tube.
Consequently, the intensity of light exiting the
u-tube decreases. As the reaction proceeds and
more product is formed, the collector records
continuously diminishing light intensity. At the
end of the three-minute reaction period, the
instrument uses the measured rate of decrease in
light intensity to estimate the concentration of
TCE in the aqueous sample.

The detection limit of the TCE analytical
module is 1 ppb. However, since chloroform
interferes with TCE analysis, the TCE detection
limit is a function of the amount of chloroform
present in the sample.

Cr-VI Analysis

The analysis package used for Cr-VI also uses a
colorimetric technique. However, the technique
does not involve the measurement of a reaction
rate and thus does not use an optrode.

The analytical cell is a cylindrical tube with a
green LED at one end and a collector at the
other. Metered portions of the sample and
reagent (1,2-diphenylcarbazide) are mixed in a
reaction cell.  Cr-VI in the sample reacts with
the reagent to form a red colored product.  The
intensity of the coloration is directly
proportional to the amount of Cr-VI present in
the sample. After a predetermined waiting
period, the red liquid is transferred to the
                                              6-4

-------
                                                      North Indian Bend Wash Site and Nevada Test Sites
analytical cell.  The red liquid absorbs green
light produced by the LED causing a decrease in
intensity of green light measured by the
collector. This decrease in intensity is used to
estimate the concentration of Cr-Vl in the
sample.

The detection limit of the Cr-VI analytical
module was found to be 1  ppb during precision
tests. However, detection limit can be reduced
by increasing the length of the optical path
through the sample.
                                     Control System

                                     The control system for the sampling and
                                     analysis modules comprises level sensors,
                                     electrically actuated valves, and an on-board
                                     logic controller.  The logic  controller interfaces
                                     with a personal computer (PC) through an RS-
                                     232 interface to receive instructions and transfer
                                     data. Figure 2 presents a schematic of an
                                     optrode-based analytical system.
Key Pad
                                            Figure 2
                               Schematic of the Analytical System
                                         Source/Lock-In
                                         Amplifier Driver
                                                                                       Waste
                                                                                       Bottle
        -Reagent Delivery System
        - Electronic System
       j Optrode
   Source: Burge Environmental
                                               6-5

-------
                                                    North Indian Bend Wash Site and Nevada Test Sites
The system has fail-safe provisions. A
watchdog program continuously monitors
system components and shuts down the system
if a malfunction is detected. An error code on
the PC screen identifies the component of the
system that caused shutdown.

Remote operation of the system is facilitated by
PCAnywhere (communications software).  This
commercially available software developed by
Symantec makes it possible for the user's PC to
communicate with the system's PC over a
phone line. The graphical user interface used to
operate the system was developed by Burge
Environmental. The program allows the user to
monitor system operation, perform analytical
system calibration, actuate sampling and
analysis, view results, and store data in a
Microsoft Excel spreadsheet.

Construction

Components of the sampling and analysis
system including all containers and electrical
components are typically housed in an insulated
weather-proof box unless they are being used
indoors.  Figure 3 provides an illustration of
both an indoor and an outdoor system.

Other than the above  common features, each
system usually has to be custom-designed to suit
a given site. This is due to variations in client
requirements, site infrastructure, and the nature
of the system being tied-in to.  For example, at a
site where electricity  is easily available, there
would be no need for an autonomous solar-cell
and battery power supply.  Similarly, where a
phone connection is easily available for an
indoor system, the data transfer system would
not need radio 'telemetry.  In extremely cold
climates, below ground construction and
allowance for a space heating system would
influence system enclosure design.  Reasons
such as these are what force the vendor to be
flexible; and the very same reasons preclude the
manufacture of a standardized off-the-shelf
system.

Site Information (1, 2, 6, 9]	

This report presents two different sites to
illustrate implementation of both the TCE and
Cr-VI systems.

The TCE system had been previously used to
monitor both groundwater and treatment
processes. This report presents a site that used
sampling and analysis modules to test the
influent and effluent of a groundwater treatment
system.

The Cr-VI analytical system became available
for use more recently.  A groundwater sampling
and Cr-VI analytical system was designed for
use at the Hanford site near Richland,
Washington and was deployed in mid July,
2004. However, before this, as a precursor to
full-scale implementation, a pilot-scale test of
the system was performed at the Nevada Test
Site near Las Vegas, Nevada.  This report
presents details of this pilot test.

From this point forward in this report, the
sampling and analysis system will be called the
Burge system.

North Indian Bend Wash Site

Groundwater at the North Indian Bend Wash
(NIBW) Superfund site in Arizona was
contaminated with several volatile organic
compounds (VOC). A groundwater pump and
treat system was being used to extract and treat
contaminated groundwater.  TCE was the
primary contaminant of concern.

The groundwater treatment system used air-
stripping and carbon adsorption to remove
VOCs from the groundwater.  The treatment
plant was operated by the City of Scottsdale and
                                              6-6

-------
                                                     North Indian Bend Wash Site and Nevada Test Sites
                                           Figure 3
                           Sampling and Analysis System Photos
   Indoor Wall-Mounted Sampling and Analysis System for Process Monitoring.  Source:  Burge Environmental
Outdoor Sampling and Analysis System for Groundwater Monitoring - Used at the Nevada Test Site. Source: Burge
                                         Environmental
                                             6-7

-------
                                                    North Indian Bend Wash Site and Nevada Test Sites
processed 10.7 million gallons of water per day.
The Burge system was installed as part of an
agreement with the EPA to allow reduced
regulatory monitoring at the facility.

The Burge system was installed at the plant in
January 2002 to monitor pre- and post-treatment
concentrations of TCE in extracted
groundwater. Influent and effluent monitoring
served as a measure for treatment plant
performance. Effluent monitoring also served
to warn against accidental discharge above the
regulatory threshold of 5 ppb. The influent
water had a TCE concentration averaging 100 to
120 parts per billion (ppb) and the effluent had a
concentration below the laboratory's detection
limit (0.5 ppb). The Burge system tapped
directly into the influent line for influent sample
acquisition.  Since the treatment plant effluent
outlet was much farther away, the facility
provided a continuously flowing open loop
system to make fresh  effluent water
continuously available to the Burge system.
The Burge system tapped into this open loop for
sample acquisition.

The Burge system performed daily analysis of
influent and effluent waters, and performed a
post-test mid-calibration check after each
analysis as a quality control measure. The
calibration process used a blank,  6 and 12 ppb
standards and a 6 ppb post-test check standard.
The calibration curve range was selected to
bracket the regulatory threshold concentration
of 5 ppb. Consequently influent water
underwent dilution prior to analysis.

Burge Environmental remotely operated the
system from their office in Tempe, Arizona.
Operation included setting times  to initiate
sampling and analysis, as well as retrieving and
reviewing testing data.  All other functions
including sampling, calibration and analysis
were automated.
Nevada Test Site

As a precursor to full scale implementation at
the Hanford site, the Cr-VI Burge system was
pilot-tested at the Nevada Test Site near Las
Vegas, Nevada. These tests were conducted in
December 2003 and March 2004.

The sampling module used a peristaltic pump to
collect water samples from storage containers.
The module facilitated sampling from four
different sources.

Communication between the Burge system and
the remote PC was made possible through a
wireless modem link. This was functionally
equivalent to using a serial cable. One modem
was connected to the on-board logic controller
in the field deployment box, and the other was
connected to a PC at a remote facility. The
modems required line of sight, and had a 12-
mile communication range.  The radio modem
in the field deployment box was powered by the
system's 12-Volt power supply.  The PC in the
remote facility hosted a data acquisition and
control software developed by Burge
Environmental.

The test system included a waste module that
was used to remove  and store all analytical
process wastes. There were two waste streams
created by the monitoring system.  One was
water from the monitoring wells used to rinse
the sampling lines. The other waste stream
contained leftover reagents, standards, and
sample after an analytical run.

The test system was not an exact replica of the
system to be used at Hanford because it had the
benefit of operator supervision. The
communication system used at Hanford was
more involved because system monitoring,
control, and data acquisition was to be based in
Arizona while the field unit was located in
Washington. A combination of a wireless
modem link and a phone line was required to
                                              6-8

-------
                                                     North Indian Bend Wash Site and Nevada Test Sites
link the field unit to the monitoring station. The
wireless modems linked the Burge system to the
on-site PC. Telephone modems linked the on-
site PC to the remote PC.  Figure 4 provides an
illustration of the Burge system used at Hanford.

System Operation [1, 2]	

This section discusses operation of the Burge
systems installed at the two sites.

North Indian Bend Wash Site

System operation was automated and did not
demand supervision. However, Burge
Environmental remotely monitored the system
as part of the  service offered with the contract.
Daily monitoring was performed in ten minutes
and was meant to observe system function and
view test results. Results of the post-test mid-
calibration determined the validity of the test. If
the result was within 20 percent of the true
concentration of the standard, the test was
considered valid. On the other hand, deviation
beyond the stipulated 20 percent limit resulted
in either repeating a test, or re-calibrating the
instrument. Table 1 presents monthly average
TCE concentration for treatment plant influent,
treatment plant effluent, and the mid-calibration
standard (6 ppb).

The  system experienced only two mechanical
problems from January 2001 to July 2003.  Both
problems were attributed to malfunctioning
valves.

Maintenance  involved  a site visit once every
three weeks to visually inspect the system and
discuss system performance with the treatment
plant operator. Other maintenance activities
included refilling reagents and standards.
Maintenance  frequency was therefore a function
of storage capacity and sampling frequency.
Temperature was the main factor that affected
performance. Varying temperatures sometimes
made sample analysis invalid because the test
temperature was different than the calibration
temperature. As a result the instrument had to
be re-calibrated several times. Other instances
of failed tests were triggered by factors such as
insufficient reagent, insufficient standard, and
power surges.  Tests were repeated in each case.

Nevada Test Site

The system was tested at the Nevada Test Site
on two different occasions. The first testing
event spanned  December 1, 2 and 3, 2003, and
the second spanned March 16, 17,  18 and 19,
2004. During these events the system
underwent extensive  laboratory and field
testing.  Since the implementation  at the NTS
was a supervised limited-duration operation,
there was no need for system maintenance.

The well-sampling module was not tested at the
NTS. Samples to be  analyzed were prepared
and stored in a container. A peristaltic pump
delivered the sample  from  the storage container
to the analytical module. In order to simulate
site-specific sample chemistry, some of the
samples were prepared using groundwater from
the Hanford site. All samples tested by the
analytical  module were also tested by an
analytical  laboratory. The average
concentration of Cr-VI measured by the Burge
system in  11 separate tests was 1,161 ppb. The
standard deviation of the results was 2.1
percent. The result of laboratory analysis for
the same sample was 1,200 ppb. The sampling
module operated autonomously using its own
power supply unit.  Power supply constituted a
solar cell and a battery. The battery was sized
to facilitate four days of continuous operation
without solar recharging and was found to have
more than sufficient capacity for uninterrupted
operation.
                                              6-9

-------
                                              North Indian Bend Wash Site and Nevada Test Sites
                                    Figure 4
                 Hanford Field Implementation Conceptual Model
                                                   COMMUNICATIONS UNIT
Source: Burge Environmental
                                      6-10

-------
                                                   North Indian Bend Wash Site and Nevada Test Sites
                                          Table 1
                         North Indian Bend Wash Analytical Results
TCE Concentration Monthly Averages (October 2002-April 2003)
Month
October
November
December
January
February
March
April
Treatment Plant Effluent
Burge System
Laboratory 4
Treatment Plant Influent
Burge System *
Laboratory 2' 4
Mid-
Calibration 3
Average Concentration (ppb)
<1
<1
<1
<1
<1
<1
<1
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
<0.5
109
98
106
105
111
110
106
112
118
113
106
132
138
122
6.6
6.3
5.9
5.2
5.5
6.2
6.2
Note:
TCE   Trichloroethene
ppb    Parts per billion
1       Based on an average of at least 28 samples per month
2       Based on an average of 4 to 5 samples per month
3       Mid-calibration standard sample had 6 ppb TCE concentration
4       Laboratory analysis used EPA Method 8260B
Source: Burge Environmental
The system performed well at the NTS.  The
analytical module -as seen above- demonstrated
precision and good agreement with the results of
laboratory analysis.

Cost [1,4]	

Actual costs for implementation at the NIBW
site and the NTS were not available.
Approximate present day costs for various
components of a Burge system are presented in
the following table:
Burge System Cost Estimate
Item
Sample acquisition
system '
Base system (includes
analytical system) 2
Cost Range
$500-$3,000
$10,000 -$12,000
Burge System Cost Estimate
Item
Solar cells & batteries
(sold without markup) 3
Wireless modems 4
TOTAL
Cost Range
$1,500
$3,200
$10,000 -$19,700
Note:

1   Depends on whether sample acquisition is
   from a treatment process or a well
   Depends on the type and number of sensors
   Only for remote system where no power is
   available
   Only for remote units where no phone line is
   available
                                                 Source: David Hoffman, Burge Environmental,
                                                 Inc.
                                            6-11

-------
                                                    North Indian Bend Wash Site and Nevada Test Sites
The bare cost for purchase of a Burge system
could therefore range from $10,000 to $19,700.
Installation costs are sensitive to several factors
and could vary. The main operation and
maintenance cost items would include:
(1) monthly fee for monitoring and data
acquisition services provided by Burge
Environmental; (2) monthly fee for a telephone
line; (3) reagent and standard solution
replacement; and (4) treatment media (activated
carbon or granular ferric hydroxide)
replacement.

If the unit is leased instead of purchased, Burge
Environmental estimates total cost for lease,
operation, and services at  $1,000 per month.

Conclusions [1, 2]

This section discusses the advantages and
disadvantages of this technology, and its
potential uses based on information provided by
the vendor and one of its users.

Although the Burge system uses the same
method for colorimetric analysis of Cr-VI as the
EPA-approved method, it  is not yet recognized
by the EPA as suitable for regulatory
monitoring. Consequently, neither the TCE nor
the Cr-VI analytical system can currently
replace laboratory analysis.  However at the
NIBW site, the combined  effect of the facility's
track record and the use of the Burge system
resulted in reduced regulatory monitoring.

The Burge system is currently more of an
optimization and management tool than a
regulatory tool. Its potential applications
include treatment process  monitoring, long term
groundwater monitoring, and breakthrough
monitoring in permeable reactive barriers.
Though never used in this capacity, the Burge
system, according to Burge Environmental, is
amenable to integration into larger systems
controlling facility  operation.  In such a role, the
Burge system might be used to trigger
independent alarms, or provide input to a
supervisory control system.

Customers have never sought autonomous
control of the Burge system and have always
required that it be operated and maintained by
Burge Environmental. As a result, Burge
Environmental provides remote oversight of
system operation, and keeps system automation
at a minimum. However, further automation
can be incorporated upon request.  For example,
though the system does not currently flag results
that do not meet testing standards it can be
programmed to check for compliance of results
with testing standards. Similarly, the system
can be programmed to trigger an alarm or call
an operator if an analytical result exceeds a
compliance limit.

According to Burge Environmental, system
operation is easy to learn, and new users can be
walked through and taught to operate the control
program in about 15 minutes.  System
troubleshooting however is more complex and
requires substantial knowledge of the Burge
system.

To reduce the amount of waste produced, the
Burge system does not purge well volumes
while sampling monitoring wells.  Analysis
usually requires 500 ml of sample for TCE and
20 ml of sample for Cr-VI. According to Burge
Environmental the sample is acquired at a rate
that is not significant enough to produce
drawdown in the well, or alter chemical
equilibrium in a way that invalidates the sample
as being representative of the groundwater in
the formation.

Mechanical problems experienced by the Burge
system during previous field deployments have
included air compressor malfunction, valve
malfunction, and analytical system clogging due
to excessive solids. The  solids problem was
solved by using an in-line filter.
                                             6-12

-------
                                                     North Indian Bend Wash Site and Nevada Test Sites
The Burge system had minor problems with
hardness-related scale formation at the NIBW
site.  Although the analytical system itself is
protected by inline solids removal, the sample
acquisition line needs to be monitored for scale
when sampling water with high hardness.
Monitoring durations for sample collection has
been an effective method of checking for scale.
According to Burge Environmental, none of
their submerged sample acquisition systems
have had problems with scale or biofouling. So
far the longest duration of submergence for any
of their system components has been 6 months.

Temperature is the most important
environmental variable influencing TCE
analysis. The Cr-Vl analytical system on the
other hand is not as  sensitive to temperature.
Although possible, the Burge system does not
adjust analytical results for temperature.  To
mitigate temperature effects, the Burge system
incorporates sufficient engineering controls and
testing strategies.  Engineering controls include
thermal insulation of outdoor units, space
heating, construction below the frost line in
extremely cold environments, and a
thermometer to record temperatures during
testing. The validity of a test rests on the
proximity of temperature during testing to the
temperature during calibration. Temperature
variations can be minimized by constraining
testing to a certain part  of the day.

Although the use of an optrode for headspace
analysis of TCE negates turbidity-related
problems, it inherits problems associated with
vapor-phase analysis. Headspace analysis
operates on the assumption that TCE vapor in
the headspace is at equilibrium with its
dissolved phase during  sample analysis.  This
complicates analysis, as changes in sample
chemistry can affect equilibration times.  TCE
analysis therefore requires  greater diligence to
detect potentially flawed results.  The validity of
questionable results  has in  the past been
confirmed by test repetition.  Consistency in
results during repetitions is considered a strong
indicator of equilibrium. According to the
NIBW facility's representative, the TCE
analytical system seemed to perform better at
high, than with low concentrations.

Carbon dioxide was identified as a hurdle in
TCE analysis.  Groundwater samples bearing
high carbonate concentrations when brought to
the surface tend to re-equilibrate under the
reduced hydrostatic pressure.  This re-
equilibration sometimes results in the release of
carbon dioxide into the head space. According
to Burge Environmental, carbon dioxide was
found to inhibit the migration of TCE across the
optrode's semi-permeable membrane.

The shelf life of reagents used in the Burge
system was cause for concern during early
stages of development. However, the TCE
system now stores components with indefinite
shelf lives separately and uses them to prepare
fresh reagent prior to sample analysis.  The
reagent used in the Cr-VI system on the other
hand has a finite shelf life.

According to Burge Environmental, none of the
analytical modules have experienced problems
with method contamination. Method
contamination can theoretically result from
leakage of standard solution into the sample, or
residual contamination from a previous analysis.
According to Burge Environmental, the Burge
system has sufficient controls in place -
including long flushing cycles- to prevent this
from occurring.

Although not an issue at the sites in this case
study, it is important to note that the presence of
cosolvents or surfactants in an aqueous sample
can affect the measurement of TCE. Since both
co-solvents and surfactants affect the
partitioning of TCE into the vapor phase, it is
possible that the Burge system would
underestimate TCE concentration when such
compounds are present in the water sample.
                                              6-13

-------
                                                   North Indian Bend Wash Site and Nevada Test Sites
The Burge system has not yet been able to
resolve the problem posed by trihalomethanes
(THM) in the analysis of the TCE. The TCE
optrode is unable to distinguish between THMs
and TCE. However, since THMs in the
environment have usually been related to
chlorination activities (such as disinfection), this
has not been much cause for concern. At the
NIBW site, chlorination of some of the wells to
discourage biological growth resulted in
elevated concentrations of THMs in the
treatment plant influent. The Burge system
detected this as elevated concentrations (1 to 2
ppb)ofTCE.

At the end of their current field deployments the
system will undergo testing for compliance with
American Society of Testing and Materials
(ASTM) requirements for analytical systems.
The standards to be met will be the similar to
those required of EPA Method 8021. This will
be the first in a series of steps towards
recognition by the EPA as a technology capable
of regulatory monitoring.

Contact Information	

North Indian Bend Wash Site
CityofScottsdale
Mr. Mark Seamans
Phone: (480)312-0390
Email: mseamans@ci.scottsdale.az.us

Nevada Test Site
Bechtel Laboratories
AMSI Program
Senior Scientist
Dr. Rick Venedam
Phone: (702)295-5487
Email: venedarj@nv.doe.gov
Federal Oversight Agency - North Indian
Bend Wash Site
EPA Region 9
Remedial Project Manager
Ms. Melissa Pennington
Phone: (415)972-3153
Email:  pennington. melissa@epamail. epa.gov

State Oversight Agency - North Indian Bend
Wash Site
Arizona Department of Environmental Quality
State Project Manager
Mr. William DePaul
Phone: (602)771-4654
Email:  wad@ev.state.az.us

Sampling and Analysis Technology Vendor
Burge Environmental
Mr. David Hoffman
6100 South Maple Avenue, Suite 114
Tempe, AZ 85283
Phone: (480)968-5141
Email:  burgeenv@globalcrossing. net

References

The following references were used in the
preparation of this report:

1.  Telephone Conversation. David Hoffman,
   Burge Environmental, with Chitranjan
   Christian, Tetra Tech EM, Inc., Response to
   Questions on the Burge Sampling and
   Analysis System. June 18 - 22, 29, July 9,
   23, August 6 and 23, 2004.
2.  Telephone Conversation. Mark Seaman,
   City of Scottsdale, with Chitranjan
   Christian, Tetra Tech EM Inc., Response to
   Questions on Burge System Operation at the
   North Indian Bend Wash site. June 29,
   2004.
3.  E-mail from David Hoffman, Burge
   Environmental, to Chitranjan Christian,
   Tetra Tech EM Inc., Response to Questions
   on the Burge System. June 28, 2004.
                                            6-14

-------
                                                   North Indian Bend Wash Site and Nevada Test Sites
4.  E-mail from David Hoffman, Burge
   Environmental, to Chitranjan Christian,
   Tetra Tech EM Inc., Response to Questions
   on Cost. July 12, 2004.
5.  Burge, Scott and Hoffman, David. 2003.
   "Automated Monitoring of Chloroform and
   Trichloroethylene Using a Halocarbon
   Specific Optrode." American Laboratory.
   November.
6.  Burge Environmental.  2004. Cr(VI)
   Detection System. Results of the
   Laboratory and Field Tests at Nevada Test
   Site. April.
7.  Worldwide web resource.  Burge
   Environmental.
   http://www. burgenv. com/about, htm.
   Optrode Background.
8.  Worldwide web resource.  Burge
   Environmental.
   http://www. burgenv. com/case. htm#doe_
   monitoring. DOE Phase I  SBIR Summary.
   Ground water monitoring system for
   multiple sensors.
9. Worldwide web resource. Burge
   Environmental.
   http://www. burgenv. com/casejndianbend_
   wash.htm. North Indian Bend Wash
   Superfund Site Case Study.
10. Worldwide web resource. Burge
   Environmental.
   http ://www. burgenv. com/case_edwards_afb.
   htm. Edwards Air Force Base Case Study.

Acknowledgements	

This report was prepared for the U.S.
Environmental Protection Agency's Office of
Solid Waste and Emergency Response, Office
of Superfund Remediation and Technology
Innovation. Assistance was provided by Tetra
Tech EM Inc. under EPA Contract No.
68-W-02-034.
                                            6-15

-------

-------
                         Appendix 7

               Automation Technology Case Study
Supervisory Control and Data Acquisition Using Programmable Logic
          Controllers at the Sprague Road Superfund Site

-------

-------
                          Automation Technology Case Study
 Supervisory Control and Data Acquisition Using Programmable Logic Controllers
                           at the Sprague Road Superfund Site
Summary Information [1, 2, 3, 4, 5, 6, 8, 9]

Programmable logic controllers (PLC) are used
to automate processes. Before the
commercialization of PLC technology, process
automation was achieved by hard-wired relays
and timers, also called basic relay logic (BRL).
Any change in process operation required
rewiring of the control circuits.  PLC technology
changed the emphasis in automation from
hardware to software. Though skilled
electricians were still required to integrate the
PLC into the control system, the logic that drove
process automation was embedded in computer
programs as opposed to hard-wired circuits.

PLC technology is being used at the Sprague
Road Ground Water Plume Superfund (Sprague
Road) site to operate the remediation system.
The site is located in Ector County outside the
northwest city limits of Odessa, Texas, and
consists of three inactive or abandoned metal
plating facilities located within 1 mile of each
other: Machine & Casting, Inc. (M&C), Leigh
Metals (LM), and National Chromium
Corporation (NC).

Historical operations at those facilities resulted
in contamination of the groundwater with
hexavalent chromium.  The contaminant plumes
beneath each facility are physically distinct, but
due to their proximity the remedial approach
was to address the three facilities as one single
site. The selected remedy for the Sprague Road
site was groundwater extraction, treatment (ion
exchange) and re-injection.

The system was designed for continuous
operation with minimal human intervention,  and
continuous centralized monitoring over the
anticipated ten-year remediation period. This is
facilitated by a supervisory control and data
acquisition (SCADA) system.

PLC technology is used at this site to facilitate
supervisory control over the extraction,
treatment and re-injection system. However, the
control system encompasses much more than
just the PLC. Sensory equipment, actuators, and
the PLC form the bone structure of the control
system.  Sensors are the "eyes," of the control
system, and provide the PLC with information
on the status of the system. The PLC is the
decision-making unit of the control system and
uses its programmed instructions to processes
sensory information ultimately leading to a
response.  The response may be: to do nothing;
to command an actuator to start an associated
device; or to command an actuator to stop an
associated device.

Sensors at this site include liquid level and
pressure sensing devices. Actuators include
switchgears and hand-off-auto (HOA) switches
that start and shut down pumps^ or open and
close solenoid valves.  All pumps at this site
have manual overrides, that is, the "hand,"
mode. An actuator will respond to PLC
commands only when the HOA switch is placed
in the "auto," mode.

The data acquisition (DA) system is comprised
of a network of three personal computers (PC)
that interface with the PLC at each of the three
sites.  A program called Lookout installed on
each of the PCs accomplishes the "data
acquisition," from the PLCs and depicts the
information on user interface screens. Lookout
is essentially a window into the sensory
information available to the PLC.

The remediation and automation system has
been in operation since September 2003, and
                                             7-1

-------
                                                          Sprague Road Superfund Site, Odessa, Texas
there have been no PLC breakdowns as of this
case study (Fall 2004). The DA system
malfunctioned in April 2004. The cause was
software-related, but the exact nature of it was
never determined. The SCADA system was
operational within two weeks of the observed
malfunction.  The time taken to implement the
solution was small compared to that consumed
by administrative tasks and logistical hurdles.
The remediation system itself functioned
normally during this time.

Technology Description [2, 3, 6]

PLCs, also referred to as programmable
controllers, are in the computer family. They
are most commonly used in commercial and
industrial applications. Figure 1 shows the PLC
used at the Sprague Road site.

A PLC monitors inputs, makes decisions based
on its program, and controls outputs to automate
a process or machine. PLCs consist of input
modules or points, a Central Processing Unit
(CPU), and output modules or points.  An input
accepts a variety of digital or analog signals
from various field devices (sensors) and
converts them into a logic signal that can be
used by the CPU. The CPU makes decisions
and executes control instructions based on
program instructions in memory. Output
modules convert control instructions from the
CPU into a digital or analog signal that can be
used to control various field devices (actuators).
A programming device is used to input the
desired instructions. These instructions
determine what the PLC will do for a specific
input. An operator interface device allows new
control parameters to be entered.

DA systems go hand-in-hand with PLC
technology. A DA system refers to DA
software and the computer system hosting the
software.  DA software can be loaded  on any
personal computer (PC) that meets the
software's minimum requirements.  Some
examples of DA system software are:
(1) Lookout developed by National Instruments;
(2) InTouch developed by Wonderwear;
(3) Intellutions developed by GE Fanuc;
(4) Cimplicity developed by GE Fanuc; and
(5) Iconics developed by Iconics Inc.

The DA system interfaces with the PLC and
continuously gathers real-time process
information. The DA system uses this
information to update graphical or numerical
depictions of system status on the user interface
screen. DA systems can be programmed to
maintain electronic records of desired process
data at any desired interval, and generate
reports.  The total amount of data that can be
recorded would be limited only by the host
computer's hard disk storage space.  The DA
system does not control any components of the
remediation system.

Site Information [1, 3, 4, 5, 8,9]	

The site is located in Ector County outside the
northwest city limits of Odessa, Texas, and
consists of three inactive or abandoned metal
plating facilities (M&C, LM and NC) located
within 1 mile of each other.

Past waste management practices at those
facilities led to contamination of the
groundwater with hexavalent chromium. The
plumes associated with each facility were
physically distinct, but the three facilities were
collectively termed the Sprague Road Ground
Water Plume  Superfund (Sprague Road) site.

The selected remedy was a comprehensive
approach for all three facilities. It entailed
extraction of contaminated groundwater through
three localized networks, treatment by an ion
exchange process at LM, followed by
distribution of treated water for re-injection at
the three facilities.
                                              7-2

-------
                                                          Sprague Road Superfund Site, Odessa, Texas
                                           Figure 1
                                Control Cabinet Housing PLC
          PLC with CPU, input,
          output, power, interface
          and communication
          modules
Remediation System

A network of recovery wells (7 at M&C, 27 at
LM, and 23 at NC) formed the ground water
recovery system. Recovery systems at M&C
and NC each pumped contaminated ground
water into local collection tanks. Pumps
transferred water from these collection tanks to
the treatment system. The recovery wells at LM
pumped water directly to the treatment system.

The treatment system was located at LM and
included a surge tank, a pump tank, pumps, bag
filters, and an ion exchange system. The ion
exchange system consisted of two banks, each
consisting of five resin tanks. Resin tanks
within a bank were connected in parallel, and
the two banks were connected in series.  At any
given time one bank acted as the worker (or
primary) resin, and the other acted as the
polisher (or secondary)

The injection system consisted of three separate
networks of injection wells (8 at M&C, 8 at LM,
and 27 at NC), and a vadose zone flushing
system at NC.  Injection pumps in the treatment
building at LM delivered treated water to each
of these networks.
                                             7-3

-------
                                                          Sprague Road Superfund Site, Odessa, Texas
System monitoring data along with groundwater
monitoring, and hydrogeological modeling was
used to periodically modify operation of the
remediation system to address the current state
of the plume in the most efficient manner.

The process flow diagram of the remediation
system is shown in Figure 2.

Control System

System control was shared by three local control
centers, one each at M&C, LM, and NC. The
three control centers communicated through
telemetry making real time data globally
available.  However, field devices associated
with a given facility could only be actuated by a
local control center.  The M&C and NC control
centers had supervisory control over extraction
pumps, injection well shutoff valves, and
transfer pumps at their respective facilities. The
LM control center performed similar functions
at the LM facility, except that it also controlled
operation of the treatment system, and injection
pumps. The control centers were prefabricated
air-conditioned metal buildings that contained a
motor control center, panelboards, a control
cabinet housing the PLC, and a desktop
computer that displayed real time system
information and recorded operating data.

The PLC at each facility was a SIMATIC S7-
300 manufactured by Siemens.  The PLC's
central processing unit was a CPU-315-2-DP-
new.  The PLC had digital input modules - each
capable of receiving 16 inputs- as well as analog
input modules - each capable of receiving 8
inputs. The number of modules used at a given
control center depended on the number of local
field sensors providing input. The PLC also had
digital output modules, each permitting output
to 16 different devices. The number of output
modules depended on the number of field
device actuators to be controlled by the PLC.
The PLC had an interface module that permitted
interface with other devices to receive
programming or transfer system status
information.

Each PLC had a communication module.  The
communication module had its own processor
dedicated to information exchange. This
module facilitated communication between
PLCs, and with the internet through various
available protocols.  At this site the three PLCs
used their communication modules to
communicate with each other through a wireless
industrial Ethernet link. A PLC at M&C was
therefore able to recognize a high level in a tank
at LM and respond by shutting down the
appropriate transfer pump at M&C.

Each PLC contained custom- programmed
instructions on the appropriate response to be
taken based on the input received from sensory
devices. For example if the PLC at M&C
received a low level signal from the level sensor
in an extraction well, it would respond with an
output signal to the corresponding pump
actuator telling it to shut down that pump.

Some of the functions performed by the PLC
included:

   (1)  Start and shut down of groundwater
       extraction pumps based on the water
       levels in the extraction wells.
   (2)  Start and shut down of transfer pumps
       based on the water levels in receiving
       and supplying tanks.
   (3)  Opening and closing of injection well
       valves based on the water levels in the
       injection wells.
   (4)  Start and shut down of transfer pumps
       based on the presence or absence of
       radio communication between facilities.
   (5)  Shut down of ion exchange system feed
       pumps based on the differential pressure
       across the bag filters.
   (6)  Complete system shutdown during alarm
       conditions such as high-high (above
       high), or low-low (below low) levels in
       tanks.
                                              7-4

-------
                                                                              Sprague Road Superfund Site, Odessa, Texas
                                        Figure 2
                                Process Flow Diagram
                            BAG
                           FILTERS
                          (BF-1A2)
TRANSFER
 PUMPS
(P-142)
                                                            si
                                                            2>«
                                                            33
                                                                                      INJECTON PUMPS  CONNECTION
                                                                                                I     TO HOSE
TREATED WATER
STORAGE TANK
   (T-2)
                                                                                        UP-T
                                                                                                               0
                                                                                                               UJ
                                             7-5

-------
                                                          Sprague Road Superfund Site, Odessa, Texas
Data Acquisition System

The DA system at Sprague Road consisted of
three desktop computers - one at each facility -
running the Enterprise version of a DA software
(Lookout) developed by National Instruments.
This version of Lookout was necessary for the
development of applications facilitating process
monitoring through the internet. Lookout was
used to develop the facility-specific applications
that performed the required DA functions at
those facilities.  The computers hosting the DA
software were all Dell PCs with 2.0 gigahertz
Pentium 4 processors.  Each PC was connected
to its local PLC through a communication port.
The DA software retrieved real-time system
status information from the PLC and graphically
depicted this information on the various user-
interface screens.

Some of the functions  performed by the DA
system included:

(1)   Displaying operating status (on or off) of
     extraction, transfer, and injection system
     pumps.
(2)   Displaying water levels in tanks  as a
     percentage of total volume.
(3)   Displaying water level status (high or low)
     in injection and recovery wells.
(4)   Displaying injection well valve status
     (open or closed).
(5)   Computing and displaying flow rates
     (gallons per minute) through  each
     extraction well, injection well, and transfer
     pump.
(6)   Displaying fluid pressure at various nodes.
(7)   Displaying presence or absence of radio
     communication between facilities.
(8)   Recording minimum, average, and
     maximum flow rates.
(9)   Recording totalized flow.
(10) Summarizing data and automatically
     printing daily, weekly, and monthly
     reports.
The DA system is in the process of being
improved to provide internet access to real-time
system-status information. When this update
has been made, authorized users will be able to
access the system's website from anywhere in
the world to view the various user interface
screens showing real-time process information.

System Operation

The system has been in operation since
September 2003. As  of May 16, 2004,
97,160,690 gallons of water were recovered,
treated and re-injected.

The only maintenance associated with the
supervisory control system was the replacement
of PLC input/output fuses. There had been no
PLC malfunctions since system startup. The
DA system malfunctioned in April 2004. The
cause of the malfunction was never determined,
but it apparently resulted in corruption of the
DA application.  The problem was fixed by
updating the Windows operating system,
installing the latest Windows service packs, then
re-installing the DA application. Since the DA
system played no role in supervisory control,  the
remediation system functioned normally even
when the DA system  was inoperable.

The flexibility offered by PLC technology was
evidenced during the initial stages of system
operation. The radio  antennae used for remote
communication between the three PLCs
occasionally lost contact with each other.  When
this occurred there was the potential for transfer
pumps at M&C and NC to flood tanks at LM.
To facilitate safe system operation, the
automation logic had to be modified. PLC
programmers were able to easily change the
program to cause transfer pumps at M&C and
NC to shut down when communication was lost
with LM.
                                              7-6

-------
                                                           Sprague Road Superfund Site, Odessa, Texas
Cost [6, 7, 8]
The total cost for purchase and installation of
the control system at Sprague Road was
$415,000.  This included labor and materials
costs for the motor control centers, panelboards,
control cabinets, sensors, DA system, PLCs and
peripherals. The actual portion of this cost that
was spent on PLC and DA system purchase and
installation was not available. However, the
table below provides the estimated cost of PLC
hardware at one of the facilities (LM) at
Sprague Road. The unit price for each
component is the manufacturer's listed price.
PLC (SIMATIC S7-300) COST ESTIMATE
Item
CPU
Digital input module
Digital output
modules
Analog input
module
Interface module
Communications
processor
Power supply
module
Qty
1
5
3
1
1
1
1
Unit
Price
($)
1,760
198
275
765
230
1,780
365
Subtotal
($)
1,760
990
825
765
230
1,780
365
TOTAL ' 6,715
Note:

1    Cost does not include installation

Unit Price Source:  Siemens Energy and
Automation

Lessons Learned [3, 8]	

According to the lead agency's contractor, the
following were lessons learned with respect to
construction of the SCAD A system at the
Sprague Road site.
The bid should require that all project submittals
be approved by the prime contractor prior to
mobilization by the subcontractors. The
subcontractors must bear the burden of meeting
every requirement of the submittal process,
including accuracy, relevance, detail,
completeness and timely submission. The sheer
volume of submittals associated with such
projects requires the above measures to ensure
strict compliance with design specifications, and
timely approval by the engineer.

Subcontracts for this type of work should be
"performance-based." This implies payment for
services based on the achievement of project
milestones as opposed to payment for time and
effort.  Subcontractors would therefore be
compelled to complete all tasks on schedule in
order to maintain their profit margins.

The inclusion of a professional Electrical
Engineer in the design, procurement and
oversight team proved to be absolutely critical
to the successful installation of the SCAD A
system.

After about a year of operation, it was suggested
that the project could have benefited from
certain SCADA system functions that were not
believed to be essential during system design.
One such function was for the system to contact
the plant operator via telephone or pager if a
problem occurred.  Another was remote access
to the SCADA system via direct dial that would
allow programmers to fix DA software glitches,
or modify the SCADA program from remote
locations.  This was thought to be potentially
more cost effective than having programmers
travel to the site to fix problems.
                                              7-7

-------
                                                         Sprague Road Superfund Site, Odessa, Texas
In general, it is important to remember that the
overall functionality of the SCADA system is
only as good as the information being provided
to it. The selection, installation, and
maintenance of the sensors that communicate
with the PLC are therefore just as important as
the PLC itself.

Conclusions [3, 6]	

Information in this section was derived from
conversations with the project Electrical
Engineer.

The complexity and scale of the pump and treat
system at Sprague Road makes manual
operation impractical. This section uses BRL
technology as a benchmark to discuss system
automation using PLC technology.

The main advantage of PLC technology is its
flexibility. System automation is based in
software as opposed to hard wiring in the older
basic relay logic (BRL) technology.  PLC
technology makes it possible to change the
operating logic of the system merely by
modifying the programmed instructions. On the
other hand changing the operating logic of a
system using BRL would require re-wiring and
the addition of relays and interlocks. This is a
tedious and complex task.

PLC technology requires a lot less space than
BRL technology.  For example, if BRL were
used at Sprague Road, the control cabinet could
be at least twice its current size.

PLC technology is relatively inexpensive
compared to BRL. For example a $1,500
investment in a PLC would purchase a lot more
process control power than BRL technology of
equal cost.

A drawback to using PLC technology is the high
level of skill and sophistication required of
those who implement this technology.  PLCs are
programmed by educated and skilled computer
programmers.  In addition, various PLC
manufacturers market their own PLC
programming software.  Even though
programming in different software
environments may lack conceptual differences,
there often are differences in method. Therefore
an expert programmer in one software
environment may not be as skilled in another.
This somewhat restricts the options available
when looking for programmers after a specific
PLC has been selected. On the other hand, BRL
does not require the same level of skill.  Any
skilled electrician experienced with control
panel wiring would be able to interpret design
drawings to efficiently implement BRL
technology irrespective of who the components
are manufactured by.

Another drawback to PLC technology is
centralized control. If the PLC breaks down, the
entire process shuts down. A damaged relay in
BRL on the other hand only leads to shut down
of the component controlled by it. For example
if a groundwater extraction pump were
controlled by BRL, break down of a relay would
only affect the one pump, and the rest of the
system would function normally.

PLC technology is weaker than BRL in its
weatherability. PLCs require greater
environmental controls than BRL to function
normally.

Generally system reliability can be increased by
blending PLC and BRL technology. However,
this type of blend may not be required for a
system such as that used as Sprague Road. This
is because there is no noticeable consequence to
system shut down for a few days or even a week
at a time. Critical systems on the other hand
would benefit from backup automation circuitry
in the event of PLC failure. One method uses
parallel processing to add reliability to the
control system.  This essentially means two
PLCs are simultaneously used to control the
                                             7-8

-------
                                                        Sprague Road Superfund Site, Odessa, Texas
                                                 1.
                                                 2.
same process. One is the primary and the other        References
serves as a backup. The backup PLC takes over
when the primary fails. The two techniques
used for this type of redundancy are called "hot
backup," and "fault tolerant backup."  Other
than the few inherent software-depended
operational differences, these terms for
redundancy techniques can be used
interchangeably. The system at Sprague Road is
not critical enough to require such measures.

Contact Information	
                                                 3.
Lead Agency:
EPA Region 6
Remedial Project Manager
Mr. Vincent Malott
Phone:  (214)665-8313                             4
Email: malott.vincent@epa.gov
                                                 The following references were used in the
                                                 preparation of this report:
                                                 5.
State Agency:
Texas Commission on Environmental Quality
Project Manager
Mr. Subhash Pal
Phone: (512)239-4513
Email: spal@tceq.state.tx.us

Lead Agency Contractor:
Tetra Tech EM, Inc.
Project Manager
Mr. Keith Westberry
Phone: (214)740-2034
Email: keith. westberry@ttemi. com

Project Electrical Engineer:
Frank J. Dillard and Associates
Mr. Frank Dillard, P.E.
Phone: (713)526-5054
Email: dillardl@netropolis.net
PLC Technology Manufacturer:
Siemens Energy and Automation                    8.
Mrs. Diana Bowman
Phone: (423)262-2510
Email: diana.bowman@siemens.com
EPA, 2000. Superfund Record of Decision.
Sprague Road Ground Water Plume Site,
Ector County, Texas. September 2000.
Siemens Energy & Automation, Inc.
Siemens Technical Education Program
(STEP) 2000 series. Basics of PLCs.
Telephone Conversation. Terry Nauman,
Wunderlich-Malec, with Chitranjan
Christian, Tetra Tech EM Inc., Sprague
Road Response to Questions on Data
Acquisition System. May 12, 2004.
Telephone Conversation. Andrew Long,
Laguna Construction Company, with
Chitranjan Christian, Tetra Tech EM Inc.,
Sprague Road Response to Questions on
Data Acquisition System.  May  12, 2004.
Telephone Conversation. Ronald McClung,
Tetra Tech EM Inc., with Chitranjan
Christian, Tetra Tech EM Inc., Sprague
Road Response to Questions on Supervisory
Control and Data Acquisition  System. May
12, and May 17, 2004.
Telephone Conversation. Frank Dillard,
Frank J. Dillard and Associates, with
Chitranjan Christian, Tetra Tech EM Inc.,
Response to Questions on  Supervisory
Control and Data Acquisition  System. May
14, and May 17, 2004.
Telephone Conversation. Diana Bowman,
Siemens Energy and Automation, with
Chitranjan Christian, Tetra Tech EM Inc.,
Response to Questions on Cost of PLC
Components.  May 17, 2004
Tetra Tech EM, Inc., 2003. Remedial
Action Report. Sprague Road Ground
Water Plume Superfund Site, Odessa, Ector
County, Texas. September 2003.
                                            7-9

-------
                                                        Sprague Road Superfund Site, Odessa, Texas
9.  Tetra Tech EM, Inc., 2004. Ground Water         Acknowledgements
   Recovery, Treatment, and Injection System        ^—-"—~^-——^—^
   Operation and Maintenance Plan. Sprague         „. .                  , ,,   .  T, 0
   D   A r-    A \ir i  ™    o    -e.   j ov          This report was prepared for the U.S.
   Road Ground Water Plume Superfund Site,        r   .      , , £     .   .      ,  „,,,-    ,,
   ~,     c .   „    .  _     rT                  Environmental Protection Agency s Office of
   Odessa, Ector County, Texas. June.              c  ,.,„,      , c         n         ^ff-
                                                 Solid Waste and Emergency Response, Office
                                                 of Superfund Remediation and Technology
                                                 Innovation.  Assistance was provided by Tetra
                                                 Tech EM Inc. under EPA Contract
                                                 No. 68-W-02-034.
                                            7-10

-------