I
I
I
I
I
I
I
I
I
I
I
xvEPA
United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park, NC 27711
EPA-450/1-90-005
September 1990
Air/Superfund
AIR/SUPERFUND
NATIONAL TECHNICAL
GUIDANCE STUDY SERIES
Contingency Plans At Superfund Sites
Using Air Monitoring
-------
I
I
I
I
I
I
I
I
CONTINGENCY PLANS AT SUPERFUND
SITES USING AIR MONITORING
Prepared by
IPEI Associates, Inc.
South Square Corporate Centre One
3710 University Drive, Suite 201
• Durham, North Carolina 27707
I Contract No. 68-02-4394
Work Assignment No. 39
• PN 3759-39
| Norm Huey, Work Assignment Manager
• U.S. ENVIRONMENTAL PROTECTION AGENCY
AIR PROGRAMS BRANCH, REGION VIII
1999 18TH STREET, SUITE 500
ONE DENVER PLACE
DENVER, COLORADO 80202-2405
I
• September 1990 U.S. Environmental Protection Agency
Region 5, Library (PL-12J)
77 Vv'est Jackson Boulevard, 12th Floor
Chicago, IL 60604-3590
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
DISCLAIMER
This report was prepared for the U.S. Environmental Protection Agency by
PEI Associates, Inc., Cincinnati, Ohio, under Contract No. 68-02-4394, Work
Assignment No. 39. The contents are reproduced herein as received from the
contractor. The mention of product names or trademarks are not intended as
endorsements of the products or their use. The opinions, findings, and
conclusions expressed are those of the authors and not necessarily those of
the U.S. Environmental Protection Agency.
ii
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CONTENTS
Figures iv
Tables V
Acknowledgement vi
1. Introduction 1
2. Past Examples of Contingency Plans Using Air Monitoring. . . 3
Castlewood Site 3
Chesnutis Site 9
Hooker-Hyde Park Site 14
Kane and Lombard Site 21
Sand, Gravel, and Stone Site 24
McKin Site 28
Nyanza Vault Site 32
Quail Run Site 39
VERTAC Site 44
Weatherford Residence 46
3. Development of the Air Monitoring Portion of a Site
Contingency Plan 50
Typical contents of a site contingency plan 50
Determining a need for contingency air monitoring 56
Designing a contingency air monitoring network 64
Case example using reverse risk assessment 71
4. References 91
Appendix A - Characteristics of the HNU Photoionizer and Organic
Vapor Analyzer 94
iii
-------
I
I
I
FIGURES
I Number ' ££fle
— 1 Air monitoring stations at Castlewood Site.......... 4
• 2 Fourteen data point running average, Castlewood Site/
Station C-02........................ 8
I 3 TAT perimeter HNU-PID monitoring stations at Chesnutis. ... 10
4 Air monitoring log photoionization detector ......... 11
I 5 Site perimeter sampling locations at Hyde Park........ 15
• 6 McKin pilot study treatment process ............. 30
• 7 Sampling locations and meteorological tower ......... 35
• 8 Diagram of the Quail Run Site................ 40
9 Results of monitoring versus action level at Quail Run. ... 40
I 10 Ambient air sampler locations at VERTAC Chemical Corporation. 45
11 Soil sample locations at Weatherford residence........ 48
J 12 Site work zones....................... 52
13 Sample standing orders.................... 54
I 14 An example of emergency response operations ......... 57
. 15 Factors influencing the Health Assessment Process ...... 62
* 16 Development of a (Contingency) Air Monitoring Plan...... 65
• 17 Example site configuration.................. 72
I
I
I
I
I
-------
I
I
I
I
I
I
I
TABLES
Number Paae
1 Dates of Monitoring Station Operation 7
2 Compounds of Interest (Using TENAX) at Chesnutis 13
3 Air Grab Sample Target Compounds at Chesnutis 14
4 Monitoring Program Summary for Hooker-Hyde Park 17
5 Monitoring Schedule, Action Level, and Required
Action Summary 18
g 6 Monitoring Levels for Semivolatile Organic Parameters. ... 21
7 Action Levels for Kane and Lombard 23
I
I
I
I
113 Public Protection Levels for the Example Site, in
and ppb '..... 82
I
8 Major Safety Threats From Compounds at the Nyanza
Vault Site ........................ 33
9 Air Sampling Results From Nyanza Vault ........... 37
10 Health Effects Summary Table A: Subchronic and Chronic
Toxicity Via Inhalation .................. 76
11 Health Effects Summary Table B: Carcinogenicity Via
Inhalation ........................ 77
12 Summary of Recommended Exposure Limits ........... 81
14 Allowable Emissions at the Example Site in g/s ....... 85
15 Summary of Average Air Emissions During Remediation ..... 87
16 Comparison of Daily Average Emissions With Allowable
I ID L.uiii(jctr liun ui udM
Emissions in g/s
I
I
I
I
87
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
ACKNOWLEDGMENT
The concept of a report on contingency plans at Superfund sites that use
air monitoring originated with Mr. Norm Huey, Air/Superfund Coordinator in
EPA Region VIII. Mr. Huey served as technical representative for this task.
The author was Mr. Roy Paul, PEI Associates, Inc., Durham, North Carolina.
Mr. Gary Saunders of PEI Associates carried out example calculations of alert
levels at a site. Ms. Alicia Ferdo was the PEI Work Assignment Manager and
Mr. David Dunbar was the PEI Project Director.
Material for this report was contributed by many individuals. Peter
Kahn, EPA Region I, provided on overview of air monitoring at a number of
Superfund sites and provided documentation for the Nyanza Vault Site. Mr.
David Webster, Chief of Maine and Vermont Superfund Section, provided
information on the McKin Site. Mr. Dean Tagliaferro, On-Scene Coordinator,
provided documentation on the Chesnutis Site. Peter Ludzia, Remedial Program
Manager, provided material on both the Sand, Gravel, and Stone Site and the
Kane and Lombard Site in Baltimore.
Mr. Tony Babb, IT Corporation-Knoxville, provided information on air
contingency monitoring at the Quail Run Trailer Park. Ms. Gloria Sosa,
Remedial Program Manager, arranged for documents to be provided for the
Hooker-Hyde Park Site and Ms. Nancy Aungst, Ecology and Environment Inc.,
provided the documents. Mr. Glen Schwartz, IT Corporation-Pittsburgh,
provided information on monitoring at the VERTAC Site. Mr. David Gray, EPA
Region VI, provided documents concerning the Weatherford Residence. Dr.
Michael Allred, Agency for Toxic Substances and Disease Registry, provided
documentation on how health assessments are carried out.
This document would not have been possible without the voluntary
assistance of these professionals.
vi
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
SECTION 1
INTRODUCTION
Air emissions from remedial or removal activities at Superfund sites can
potentially have a significant impact on the health and safety of the
individuals living and working around the site. As a result, potential
offsite impacts should be considered by the Remedial Project Manager (RPM) or
Enforcement Project Manager (EPM). He should decide whether to develop and
implement a contingency plan. The contingency plan may require air
monitoring during site disturbance operations (i.e., exploration, removal,
and remediation). Contingency air monitoring is an extension of the onsite
health and safety plan for the protection of workers. It enables the early
detection of releases such that operations can be modified or controlled and
the public adequately warned in the case of an emergency.
Contingency planning, as defined in this document, encompasses the air
program established to protect offsite populations. Monitors for this
purpose are usually located at the site perimeter or within the community.
Monitors located within the site for the safety and protection of workers are
not included in this definition, unless onsite monitors serve the dual
purpose of protecting both the workers and offsite population.
One reason that offsite contingency planning is sometimes overlooked is
that remediation, when carried out according to plan, should not cause
excessive emissions. Remediation plans, however, are only as good as the
data used to characterize the site, which is usually based on soil and air
sampling. Even the best sampling program can have limitations on the
accuracy of data concerning the locations and concentrations of chemicals.
Even a small amount of error regarding these matters can lead to unexpected
emissions and unexpected concentrations offsite, a situation that is
addressed in a contingency plan.
A contingency plan using air monitoring establishes alert levels in
advance of actually collecting monitoring data. Alert levels address the
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
offsite population exposure concentrations that trigger an emergency response
or a change in remedial activities. These alert levels are in addition to
alert levels for onsite personnel.
One role of a contingency plan using air monitoring at a Superfund site
is to support a United States Environmental protection Agency (EPA) community
relations program. This is especially appropriate for sites which are
perceived by the local community to have potentially unacceptable air
impacts. Air monitoring provides an early warning of actual releases and the
results of air studies provide a factual basis for communicating the
potential for exposure (and nonexposure) to the public. Contingency planning
demonstrates responsiveness to the community's concern on the part of the
responsible party.
David Roe, Senior Attorney with the Environmental Defense Fund, defined
the public's perception of its need for information in the EPA Journal.
"The public emphatically does not need to be deluged with "the data" on
health risks from chemical exposures, general or specific, and told to
make its own mind. This, in effect is too often what happens now by
default, particularly in controversial cases. The public is not
interested in government's abandoning the responsibility for deciding
where chemical...limits lie."
"What the public does want and need is a system that delivers a clear
signal where chemical exposure crosses a boundary from the trivial to
the significant, like the red light above a hockey net that flashes when
the puck entered the goal. The public also needs assurance that the
system is hooked up and operating, so that the light goes on when the
line is crossed, no matter which teams are on the ice. And people need
to know that the line itself is not being curved back into the net, or
even erased, just before the playoffs."
The purpose of this document is to: 1) illustrate contingency air
monitoring with examples from past projects, and 2) describe how a
contingency air monitoring program may be established. This document is
illustrative in nature because the application of this type of monitoring is
not consistently prescribed in rules and regulations, but is based on
professional judgment applied in an analysis of individual sites and
particular circumstances.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
SECTION 2
PAST EXAMPLES OF CONTINGENCY PLANS USING AIR MONITORING
Although contingency air monitoring to protect offsite populations has
not been conducted at every Superfund remediation or removal site, there are
a number of cases where it has been employed. This section documents a
sample of such sites, covering varying terrain and different types and
concentrations of chemicals. These examples exhibit wide variations in the
type of monitoring program to be employed. This variation reflects different
site conditions, different phases in the Superfund program, and different
judgements of professionals who manage the sites.
2.1 CASTLEWOOD SITE
Castlewood is a residential neighborhood on the outskirts of St. Louis,
Missouri. It is located 2.7 miles north of the Merrimac River, a tributary
of the Missouri River. Figure 1 shows the configuration of the roads in this
2
neighborhood and some of the residences.
Soil samples were collected and analyzed from the Castlewood area from
February 1983 to 1987, in an attempt to define the limits of the areas
contaminated with dioxin. As of 1987, the known areas of contamination
covered nearly 450,000 square feet, with approximately 50,000 square feet
having dioxin concentrations in excess of 10 ppb. The highest levels of
contamination were found in the parking area for Mel's Tavern located at the
intersection of Sontag and New Ballwin Roads; these concentations exceeded
500 ppb. In 1985 this area was paved by EPA as an interim mitigation
measure.
Most of the contaminated areas were located adjacent to roadways,
extending from the roadway for a few feet in some areas to as much as 25 feet
"in other areas. Many contaminated road shoulders were located next to
residences while others were near heavily wooded or brush-covered lots. In
addition to soil contamination, dioxin contamination was found inside nine
residences and one business (Mel's Tavern). Some decontamination efforts
were taken within these structures to reduce the potential for exposure.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
C-01
Meteorological Station
C-12
Figure 1. Air monitoring stations at Castlewood Site.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
In 1987 the planned remedial actions involved the excavation,
containerization, and onsite consolidation of contaminated materials in
specially designed storage buildings. Excavation was performed while
maintaining air contamination levels specified by State and Federal health
officials.
An issue of concern to the EPA was whether and to what extent offsite
migration of dioxin in the air would occur while these activities were in
progress, an issue of particular concern at the Castlewood Site due to the
close proximity of areas of contamination to human populations. Because of
this concern, EPA conducted air monitoring around the site while remediation
was in progress. The objectives of the air monitoring operation were: 1) to
evaluate the potential for dioxin exposure to general populations and to
populations at greatest risk, 2) to compare measured dioxin air
concentrations to an established criteria which served as a trigger for
abatement actions, and 3) to assess the adequacy of onsite dust suppression
techniques. Specific abatement actions to be taken were decided by the On
Scene Coordinator (OSC). If soils were dry, they would be sprayed with water
to reduce dust and volatilization. If soils were already wet, operations
were to cease. Investigations would be initiated to determine why the alert
level was exceeded.
Removal operations were planned for five different sections of
Castlewood to be excavated in series. Each excavation section could be
treated as an area source or multiple small sources.
The air monitoring network was comprised of twelve (12) dioxin samplers
and one (1) meteorological station. The samplers and the met station were
located as shown in Figure 1. One of the samplers served the dual purpose of
being both a perimeter and sensitive receptor monitor.
Because potential dioxin emissions were expected only where active
removal operations were in progress, air samples were collected only at
sampling locations in those sections undergoing active remediation.
Consequently, the schedule for sampler operation was dependent on the
excavation schedule. One exception to this rule was that one dual purpose
(perimeter/sensitive receptor) sampler was operated throughout the duration
of the project.
All samplers within each section were operated concurrently on the same
schedule. Sampler startup times coincided with periods of minimal removal
activity, such as early in the morning or late in the afternoon. Once
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
started, samplers were allowed to operate continuously for 72 hours with no
more than +/-10 percent time variation. After samples were collected at the
conclusion of the 72-hour sampling period, the samplers were immediately
restarted unless no removal activities were scheduled for that day, in which
case samplers were restarted the next day for which removal activities were
planned. Table 1 shows the dates of start-up and decommission for each
monitor.
Each air monitoring station consisted of a Model PS-1 PUF sampler
(manufactured by General Metal Works, Inc.) mounted on an elevated platform.
Samples were collected using a dual sample collection media comprised of a
glass fiber filter (6FF) and polyurethane foam (PUF) sorbent. The volume of
air sampled was accurately measured, ranging from approximately 900 to 1300
M3 over the life of the project. Samples were analyzed using
rapid-turnaround GC/MS and GC/MS/MS facilities provided by the Contract
Laboratory Program (CLP).
The air monitoring project's performance was evaluated based upon onsite
quality assurance (QA) audit results and on QA summary statistics. Project
performance documentation consisted of approved QA project plans, written
standard operating procedures, QA system audit reports, quality control (QC)
sample results, and QA audit results.
All sample data generated during this project were subjected to a
rigorous data review/validation process to ensure that reported data met all
criteria for acceptability. Initial data validation was performed by the
field sampling personnel. Collected samples not meeting the sample
collection criteria were voided and not submitted to the CLP for analysis.
Samples that were analyzed had to pass an analytical data validation process
conducted by EPA project QA personnel. Sample data meeting all criteria were
considered valid.
An onsite QA audit was performed at the Castlewood Site during July
1987. The audit results showed that with the exception of minor
deficiencies, the project was conducted in compliance with the specified
procedures.
The resulting air monitoring data showed that average ambient
concentrations of airborne 2,3,7,8-TCDD remained below the 3.0 pg/M action
limit throughout the project. The tabulated data contained in the Analysis
Request Report for this project showed that of the total 392 valid individual
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 1. DATES OF MONITORING STATION OPERATION
Monitoring station
C-01
C-02
C-03
C-04
C-05
C-06
C-07
C-08
C-09
C-10
c-n
C-12
Date of Startup
04/27/87
04/27/87
05/03/87
05/03/87
04/27/87
05/15/87
04/27/87
05/06/87
05/12/87
05/09/87
05/09/87
05/09/87
Date decommissioned
06/20/87
06/25/87
t)6/25/87
07/24/87
10/20/87
10/20/87
10/20/87
10/20/87
10/20/87
07/21/87
10/08/87
07/20/87
air sample measurements performed over the course of the project, only three
samples (Nos. 284, 315, and 327) yielded mesasurements of 2,3,7,8-TCDD at
detectable levels.
For the purpose of evaluating maximum population exposures over the
duration of the project, the data were grouped separately for each air
monitoring station and reduced to 14 running averages that were graphically
plotted over time. One of these graphs is shown in Figure 2. When computing
these running averages, all nondetect data points were treated as though they
were positive measurements. Conservative treatment of nondetect data points
in this manner provided an upper bound result. In addition to showing 14
running average concentrations, the graphs also showed the upper and lower 95
percent confidence limits around the averages. For the most part, the 14
data point running average concentrations remained around the 0.8 pg/M
level.
Based on the data generated from the air monitoring project and
presented in this report, it can be concluded that emissions caused by
removal operations were effectively controlled and that human populations
residing in the vicinity of the Castlewood site were not exposed to average
-------
1
1
1
1
3.50
1
3.00
™ QC
uioe
Io^ 2.50
QU.
00
IODUJ
rCC 2.00
»o
I^m
oS 1.50
I3K
««
O£0
SS 1.00
I-tn
^
- 0.50
| 0.0^0
—
—
—
_
—
<
^
1
1
1
<
o\
<
^
<
|
^
i
<
<
a
i
i
i
/
A£
s\
.UOJL.LfW
|
^
<
|
*>\
1
i
,
1
1"
1 1
i V 0, ^1
^>\ ^>\ Sj\
-
':-
1 1 •
i i
^
o MN-oomiw
x pomiu
'
DATES
1
Figure
1
1
1
2.
Fourteen data point running
Site/Station C-02.
8
-
average, Castlewood
-------
I
I
I
I
I
airborne concentrations in excess of the health based action limit
concentration of 3.0 pg/M3 over the course of the removal project.
2.2 CHESNUTIS SITE
The Chesnutis Site consists of approximately 0.5 acres of land located
o'n Lopus Road in the town of Beacon Falls, Connecticut. The site- is bounded
on the north by an antique auto restoration shop, on the east by Lopus Road,
on the south by an auto body shop, and on the west by a partly paved road
leading to Mr. Posick's home. The Beacon Falls Municipal Waste Water
Treatment Plant is located east of the site, on the other side of Lopus
Road. Figure 3 is a sketch map of the site.
Site assessment reports for this site identified buried drums and soil
contaminated with volatile organic compounds (VOCs). Therefore, a strong
potential existed for VOCs to be emitted into the atmosphere during drum and
soil removal. EPA's OSC, Dean Tagliaferro, was concerned that VOCs could
volatilize into the air and be carried offsite to local receptors during soil
and drum removal activities.
The following three types of air monitoring were carried out:
1. The hot zone was monitored to determine if personnel protection
levels were adequate or could be downgraded. Air monitoring was
done with either an HNU or OVA on a minimum of an hourly basis. An
action level of 5 ppm sustained readings above background required
evacuation of personnel not in protective respirators. From 0 to 5
ppm, benzene draeger tubes were to be used as benzene's threshold
limit value (TLV) is below 5 ppm, while all other identified
compounds TLVs were above 5 ppm.
2. Air monitoring at the perimeter was also conducted with an HNU or
OVA by the Technical Assistance Team (TAT). This was to determine
if air contamination was migrating offsite and if offsite personnel
had to be evacuated. Once each day, when remediation activities
3. Pollutant-specific air monitoring was carried out with carbon tenax
samplers during the excavation. Results were used to identify and
quantify the concentrations of compounds in the air in the hot zone
and at the perimeter.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TO FOSICK RESIDENCE
TO ABTO
BODY SHOP
ANTIQUE
AUTO
RESTORATION
BUILDING
•FILL
NOT TO SCALE
Figure 3. TAT perimeter HNU-PID monitoring stations at Chesnutis.
10
-------
1
1
• CRESKUTZS SITE
m 450 LOPUS ROAD
BEACON FALLS, CT
1
M
ATE: 7/13)1 o 7 i /
tOMTTORS / ' t^ar-eu'CU fr/OH
STATION/LOCATION
(Reading ppm) __„ — ,____
^^1 •
| S/
| ','/.7«l/
*/ <^£T
II 1 7^l/
//.7«v
I/'.7«f^
1
1
1
1
TIME
'a1&
&bo
\&°
r
I33>£
/— '^
1505
\
\
\
\
\
1
1
1
1
1
WD 01_
(\ljv£ o
^5
«/*,& o
fj/fi>£.
A/JlJ'C.
/
N/tf£
£c
O
O
.01
.01
o
0
0
O
O
0
.oS
_03_
0
t>
£>
^ ^
."
.ol
_04_
o
^^
d
O
D
.01
rtl .°'
O i
* w '
0 0
0 0
_05_
^
^
O
.O/
.0'
0
£>
"o-
^5
*
o •
.01
0)
0
_07— _OB—
_! .
r
-
10
1 1 1
1
—
t>»f-*+'*
0
^*)
0
> **/
(O
O
i
Figure 4. Air monitoring log photoionization detector.
11
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The air monitoring program was conducted on July 11 and 12, 1989, during
onsite soil and drum removal activities. Time-weighted average ambient air
samples were collected on and offsite using TENAX air samples, techniques and
analytical methodologies designed to identify and quantitate the compounds
listed on Table 2. In addition to TENAX samples, grab air samples were
Collected and analyzed onsite using a Photovac portable GC. Grab, air
sampling and analysis techniques identified and semiquantitated the compounds
listed on Table 3.
The air grab sampling results and continuous integrated sampling results
both identified similar compounds above background, on and off the site,
namely: Toluene, Tetrachloroethylene, and Dichlorobenzene isomers. These
compounds were detected at levels between 2 and 100 ppb onsite. Total
hydrocarbon readings taken with an HNU at several locations along the
perimeter of the site, however, showed nothing above 1 ppm. These results
indicated that several VOC targeted compounds were emitted at very low levels
(below 1 ppm) into the atmosphere and transported offsite, as a direct result
of excavation and drum removal activities. The HNU total hydrocarbon
analyzer used to routinely monitor the air around the perimeter of the site
during site work was judged to provide adequate air monitoring to determine
if VOCs were migrating offsite.4
12
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 2. COMPOUNDS OF INTEREST (USING TENAX) AT CHESNUTIS
1,1,1-Trichloroethane
Carbon Tetrachloride
' Benzene
1,2-Dichloroethane
Trichloroethylene
1,2-Dichloropropane
Bromodichloromethane
2-Chloroethy1vinyl ether
cis-1,3-dichloropropene
Methyl Isobutyl Ketone
Dibromomethane
Toluene
trans 1,3-Dichloropropene
1,1,2-Trichloroethane
Tetrachloroethylene
1,3-Dichloropropane
Dibromochloromethane
1,2-Dibromoethane
Chlorobenzene
Ethylbenzene
Brorooform
1,2-Dichlorobenzene
NOTE: Compounds that are underlined
samples or in soil samples.
Xylenes (total)
Styrene
Isopropropylbenzene
1,1,2,2-Tetrachloroethane
Bromobenzene
1,2,3-Trichloropropane
n-PropyIbenzene
2-Chlorotoluene
1,3,5-TrimethyIbenzene
4-Chlorotoluene
t-Butylbenzene
1,2,4-TrimethyIbenzene
s-Butylbenzene
p-Isopropyltoluene
1,3-Dichlorobenzene
1,4-Pichlorobenzene
n-ButyIbenzene
1,2-Dibromo-3-chloropropane
1,2,4-Trichlorobenzene
Hexachlorobutadiene
Naphthalene
1,2,3-Trichlorobenzene
Nitrobenzene
have been identified in either soil gas
13
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
TABLE 3. AIR GRAB SAMPLE TARGET COMPOUNDS AT CHESNUTIS
Dichloroethylene isomers
Benzene
Toluene
Tetrachloroethvlene
Chlorobenzene
o-Xvlene
m-Xvlene
Tricloroethvlene
Ethyl Benzene
Note: The compounds that are underlined had been identified in either soil
gas samples or in soil samples previously taken from the site.
2.3 HOOKER-HYDE PARK SITE
The Hyde Park Landfill, approximately 15 acres in area, is an NPL site
located in the northwest corner of the town of Niagara Falls, New York. It
is immediately surrounded by several industrial facilities and property owned
by the power authority for the State of New York (Figure 5). The Niagara
River, an international waterbody, is located 2000 feet to the northwest.
Between 1954 and 1975, Occidental Chemical Corporation (OCC) disposed of
approximately 80,000 tons of chemical wastes at the landfill and 0.6 to 1.6
tons of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) contaminated material.
Between 1975 and 1979, OCC implemented a number of remedial actions. These
actions included capping the site, installing a shallow tile drain, and
initiating a ground water monitoring program. Soil and ground water are
.contaminated with VOCs, organics, toluene, phenol, polychlorinated byphenyls
(PCBs), and dioxin.5
The selected remedy for this site included installation of a prototype
purge well system to extract nonaqueous phase liquids (NAPL) for destruction
by incineration, installation of an overburden tile drain system,
14
-------
91
t/i
n>
•a
a>
-s
en
0>
o
o
a*
O>
a.
a>
-a
o>
2 S*
8. $3
5'
*
I II LJL
' 3 ' ouotoS
3. i JL_—-
I
I
I
I
i
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
Implementation of engineering controls to protect nearby workers,
installation of ground water wells as part of a community monitoring program,
installation of the first stage of a bedrock NAPL plume contaminant system,
installation of purge wells as an aqueous phase liquid plume containment
system, implementation of a lower formation and deep formation study,
implementation of a Niagara gorge seep program, treatment of ground water
with activated carbon, and implementation of a monitoring program.
As a result of negotiations with EPA and the State, an extensive air
monitoring program was developed for OCC by ERT, an engineering firm. This
air monitoring program has five categories:
1. Personal monitoring - occupational health protection
2. Working site monitoring - occupational health protection
3. Downwind of the site - offsite community protection
4. Site perimeter monitoring - offsite community protection
5. Community monitoring - offsite community protection
Table 4 summarizes the monitoring program, including the parameters to
be monitored, the measurement method, and the recommended instrumentation.
Table 5 summarizes the monitoring parameters, frequencies, action levels, and
required actions for each of the five types of monitoring.
Site perimeter monitoring is the type most commonly used to protect the
community. Monitoring station locations for site perimeter monitoring are
depicted in Figure 5 (sites P-l through P-13). For work at Sites A and B,
monitoring locations P-l through P-9 were used. For work at Site C,
monitoring locations P-8 through P-13 were used.
At each of the monitoring locations, two monitors were used, one to
measure total suspended particulates and one to measure semi volatile
organics. Total suspended particulates were measured using a high-volume
particulate sampler (Hi-Vol). The Hi-Vols were operated for 8 hours during
remedial activity. Semivolatile organic compounds were monitored using
sorbent samplers operated for the same period each day as the Hi-Vol
monitors.
Semivolatile organic compounds were collected using the General Metals
Works Model PS-1 sorbent sampler. These units were operated at a flow rate
resulting in a total sample volume of approximately 120 m3. Each PS-1 was
calibrated monthly and the flow rate checked daily.
Semivolatile organic compounds collected on quartz fiber
filter/polyurethane foam (PUF)/XAD-2 sorbent "sandwich" cartridges (PS-1
16
-------
1
1
1
•^•f
1
1
1
I
1
1
1
1
1
1
1
1
1
1
1
TABLE 4.
pArameter(s)
Meteorology
Wind Speed
Wind Direction
teal-Time
Participates
Particulates
TSP
Semi-Volatile
Organlcs
Specific Volatile
Organlcs
Total Volatile
Organlcs
Plammablllty
Semi-Volatile
Organics
H2S
-
"--.
MONITORING
Units
MPH
DEO
>9/"3
vg/m3
iig/m3
vg/m3
•g/m3
PPM
%LBL
ng/m3
PPM
PROGRAM SUMMARY FOR
Measurement Method
Meteorological
Station
Real-time
Aerosol Monitor
NIOSH 0500
Hlvol Samplers
NIOSH P and CAM 343
NIOSH PfcCAM 127
PID
Combustible
Oas analyzer
PS-1. PUF/XAD
Color-Detector Tubes
17
HOOKER-HYDE PARK
Recommended Instrumentation
Cllmatronlcs
BUS
.
MDA PCD-1
OCA HINIRAM
DuPont Alpha-1
OMU1-2000
DuPont Alpha-1
DuPont Alpha-1
HHU-PI-lOl
MSA- 100
OMW-PS-1
Draeger-47s
-------
TABLE 5. MONITORING SCHEDULE, ACTION LEVEL, AND REQUIRED ACTION SUMMARY
Location
Personal
Worksite
oo
Parameter
1) Participates
(TWA)
2) Specific volatile
organic*
(NIOSH P and
CAM 127)
frequencies
Twice weekly Cor first
30 days of Special
Construction activities,
Thereafter, monthly
3) HCB. g
HCCH
1) Planaablllty
2) Total Volatile
Organic* (PID)
1) Continuous
2) Continuous
Action Levels
1) M/A
2) TWA exceeds ACOIH
guidelines
3) HCB > 0.1 ppm
gamma HCCH
> 0.04ppa
1) 25% LEL
2) (a) 5 pp* above
background
2) (b) 10 ppm above
background
3) Specific Volatile
organlcs
(NIOSH P and CAM
127)
4) Real-time
Partlculates
5) Hydrogen Sulflde
(H2S)
3) Twice per week per
working site location
4) Continuous
5) Hourly If odor Is
detected while
Installing extraction
wells
3) TWA exceeds
ACOHI gu Id lines
4) a) 150 ug/m3
b) 150 ug/m3 and
> 2.5 tines
background
5) 10 ppm
Required Action
Review by Safety Officer and EPA/State
on-slte Representative to determine what
action, tf any, shall be taken. Report
data and results of corrective action If any.
to BPA/State within 7 days after receipt of
the data.
1) (1) Suspend construction/notify BPA/State
(11) Proceed per Vapor Emission Response
Plan
2) a) (1) Modify activities to reduce emissions
(11) All working site personnel must upgrade
to full face air purifying respirators
2) b) (1) Proceed per Vapor Emissions Response Plan
(11) Conduct Specific Volatile Organic
Analysis, but not more frequently than
once every two weeks. Complete analysis
within one working day. Proceed per
Action Level, Required Action shown In
(3)
3) Review by Safety officer and EPA/State on-slte
representative to determine what corrective
action If any shall be taken. Report data and
results of corrective action. If any, to
BPA/State within 7 days after receipt of the
data.
4) a) Initiate hourly upwind monitoring, modify
activities to reduce emissions
i i
b) Suspend Const. Activity, notify BPA/State
5) a) Notify Site Safety Officer and KPA/State,
modify activities, and upgrade respiratory
protect Ion
b) Proceed per Vapor Emlslon Response Plan
(Continued)
-------
TABLE 5 (Continued)
location
Parameter
Downwind of the VOC (PID)
Working Site
frequencies
1) Every 2 hours or
as required
2) integrated (real-
tin*) continuous
parttculates
2) Continuous
(Integrated)
Site Perimeter 1) TSP
2) HCB
3) HCCH
Dally during all Special
Construction Activities
4) Perchloro-
pen t acyclodecane
5) TCP
Action Levels
1) a) 2.5 ppM (above
background
b) 5.0 ppM (above
background Cor
2 consecutive
readings)
c) 5.0 ppM (above
background Cor
3 consecutive
readings)
2) a) Significantly >
background Cor
any IS Minute
Interval
b) Significantly >
background Cor
2 consecutive
hourly averages
or any 3 hourly
averages during
one work day
1) Significantly >
background
2) 20 Mg/M3
3) 20 ng/M3
4) 20 ng/M3
5) 20 ng/M3
Inquired Action
1) a) Increase Monitoring frequency to hourly
b) Hodlfy activities to reduce emissions
c) Suspend activity until readings are less
than 2.5 ppM above background, notify
HPJk/Stata
2) a) 1 uspet - Modify const.- activity to reduce
eMlsslons. 2 upsets In 1 hour » Increase data
collection frequency to every 1/2 hour
b) Suspend const, activity, notify EPA/State
1) Analyze collocated PS-1 sample for seml-
volatlles
2) Compare with background levels (Section a.2.3)
and notify EPA/State, proceed per RUT Section
12.10.4
3) CoMpare with background levels (Section 8.2.3)
and notify BPA/8tate, proceed per RRT Section
12.10.4
4) Coapara with background levels (Sections 8.2.3)
and notify BPA/State, proceed per RRT Section
12.10.4
5) Compare with background levels (Sections 8.2.3)
and notify EPA/State, proceed per RRT Section
12.10.4
(Continued)
-------
TABLE 5 (Continued)
Location Parameter
Beyond the site 1) TSP
2) HCB
3) HCCH
frequencies
4) Perchloro-
pentacyclodecane
5) TCP
Action Levels
Dally during «11 Special 1) Significantly >
Construction Activities background at two
•jacent sites
2) 20 ngAri
3) 20 ng/m3
4) 20 ng/«3
5) 20 ng/s£
Action
1) Analyze collocated PS-1 samples for se*l-
vola tiles
2) Notify EPA/State, proceed per off-site
Contlgency Plan
3) Notify EPA/State, proceed per off-site
Contlgency Plan
4) Notify EPA/State, proceed per off-site
Contlgency Plan
5) Notify EPA/State, proceed per off-site
Contlgency Plan
ro
o
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 6. MONITORING LEVELS FOR SEMIVOLATILE ORGANIC PARAMETERS
Semi volatile
organic parameter Monitoring Level
Hexachlorobenzene (HCB) 20 ng/m3
2,3,4-Trichlorophenol (TCP) 20 ng/m3
Perchloropentacyclodecane 20 ng/m3
(C10ci12)
Hexachlorocyclohexane 20 ng//m3
(HCCH) [for each alpha, beta,
gamma, and delta isomer]
samples). Samples were analyzed using gas chromatography/mass spectrometry
(GC/MS) or gas chromatography/electron capture detection (GC/ECD) using a
modification of EPA Method T-04. Either GC/ECD or GC/MS was used for
identification and quantitation of the designated target compounds i.e., HCB,
2,4,5-trichlorophenol (TCP), Perchloropentacyclodecane (Cjoch2^' anc* HCCH
isomers.
During construction activity, perimeter TSP levels were checked to
determine if they exceeded a predetermined upper level action level (ULAL).
If a ULAL was exceeded, the PS-1 sample from the same site location was sent
to the laboratory to be analyzed. If any of the four target chemicals
exceeded a specified monitoring level, then that chemical was considered to
have migrated beyond the site perimeter. If this occurred 3 times in 30
days, then construction had to be stopped. Table 6 lists the monitoring
levels that were established for semivolatile organic compounds for this
site, based on studies of background ambient air monitoring performed
earlier.
2.4 KANE AND LOMBARD SITE
The Kane and Lombard Site is an 8.4 acre parcel of undeveloped land in
Baltimore, Maryland.- Dumping and burning of construction debris, and
disposal of domestic trash and drums occurred at the site from 1962 until
1967 when the city passed an ordinance prohibiting the open burning of
refuse. Illegal dumping continued from 1967 until approximately 1984, during
21
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
which time many citations were issued for illegal burning on the property.
In 1980 Maryland state inspectors observed between 400 and 500 drums, the
majority of which were rusted, damaged, and punctured. Following an onsite
property assessment, EPA authorized the immediate removal of 1,163 drums in
1984. Of those, 822 drums were classified as empty and 341 drums, contained
contaminants which included benzene, toluene, xylene, polyaromatir
hydrocarbons (PAHs), PCBs, and heavy metals. Approximately six inches of
soil below the drums were removed and disposed offsite. The site was
stabilized by regrading, capping, and revegetation. Currently, soil and
ground water are contaminated with prior drum contaminants.
The selected remedial action for this site includes removal of drums,
hot spots, and contaminated soil (approximately 67,000 cubic yards), site
cleaning and removal of vegetation to facilitate the construction of
subsurface containment and diversion structures, construction of a multilayer
soil cap, construction of a drainage system, clearing of the drainage ditch
along the east site of the site, development of necessary surface water
runoff management facilities, and ground water monitoring.
An air monitoring program was established to 1) determine appropriate
safety and personnel protective measures to be implemented during cleanup, 2)
document onsite employee exposures, and 3) assess offsite migration of
contaminants released during remedial activities so that appropriate control
measures and/or contingency plans could be implemented. Two principal
approaches were used to identify and quantify airborne contaminants:
o Real-time air monitoring by use of direct-reading instruments
o Time-weighted averages by use of sampling techniques that capture
samples over periods of time for later identification and
quantification of specific contaminants
Real-time air monitoring was conducted for VOC's, particulates,
explosive atmospheres and oxygen levels. Total organic vapors, given in
parts per million (ppm), were detected with photoionization detector (PID),
manufactured by H-NU, Inc. Real-time readings for combustible gas levels
tgiven in percent of the LEL) and oxygen levels (given in percent 02) were '
taken with an Industrial Scientific Combustible Gas/Oxygen Monitor, Model MX
241. Particulate concentrations in milligrams per cubic meters (mg/m3) were
determined by use of a direct-reading dust monitor, a Miniram PDM-3 model,
22
-------
I
I
I
I
f
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 7. ACTION LEVELS FOR KANE AND LOMBARD
Instrument
Location
Concentration
Action to be taken
H-Nu monitor
H-NU PID
MX 241
oxygen meter
MX 241
LEL monitor
PDM-3 dust
monitor
PDM-3 dust
monitor
PDM-3 dust
monitor
Active work area 4 ppm above back-
ground
Perimeter
Active work area
10 ppm above back-
ground
Two readings
greater than 5 min
apart 4 ppm or
more above back-
ground, or one
reading of 10 ppm
above background
Below 20.9% for 2
readings 5 min
apart within
perimeter
Active work area 10% of LEL
20% of LEL
«•
Active work area Up to 0.5 mg/nT
Active work area >0.5 mg/m
Active work area 15 mg/m
23
Upgrade to Level C
protection
Alert CO of situation
Alert CO of situation
Stop activities until
levels at perimeter
drop below 4 ppm
Stop all work until
source of oxygen
deficiency is found
and corrected
Stop all potential
spark-producing
activities
Evacuate work areas,
isolate problem area
Level C protection
Upgrade to Level B
protection
Notify the CO
Evacuate all work
areas
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
1
I
I
made by MIE, Inc. Action levels for workers as well as perimeter monitoring
are listed in Table 7.
Perimeter monitoring took place at four locations, designated according
to wind direction. Initially, real time monitoring was performed practically
continuously during the first hour or two of each workday during-active
remediation operations, followed by periodic monitoring (at least every two
hours) for the remainder of the workday.
Air sampling was conducted throughout complete shifts to determine
time-weighted average (TWA) concentrations of selected chemical agents.
These data were used in interpreting real-time monitoring results on a
day-to-day basis, documenting employee exposures, and for determining whether
or not significant contamination extended beyond the site. Selection of air
contaminants for TWA monitoring was based on previous site characterization
and sampling data and included the following:
Organic Vapors: Toluene, xylene, isophorone
Nuisance (inert) dusts: Total dust with subsequent analysis
for heavy metals: arsenic,
chromium, and lead
An action level was established for perimeter monitoring, based on the
H-Nu PID. If one reading of 10 ppm above background was taken or if two
readings greater than 5 ppm above background were taken at least 5 minutes
apart, then all activities had to stop until perimeter levels dropped below 4
ppm.
If monitoring results at the perimeter exceeded action levels,
subsequent readings were taken 100 feet and 200 feet downwind of the
perimeter on a perpendicular traverse approximately 200 to 250 feet in
length. If readings at these locations exceeded action levels in the
direction of nearby schools, the Site Safety and Health Officer was required
to inform the administration of the local school.
2.5 SAND, GRAVEL, AND STONE SITE
The Sand, Gravel, and Stone Site covers approximately 200 acres, and is
located in Elkton (Cecil County), Maryland. The site was previously operated
as a sand and gravel quarry under the name Maryland Sand and Gravel stone
Company. Currently, the site is occupied by the Sand, Gravel, and Stone
24
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
company. Between 1969 and 1974, three acres were used to dispose of waste
processing water, sludge, still bottoms, and approximately 90 drums of solid
and semisolid waste. On July 16, 1974, 1,300 gallons of flammable products
in drums were reportedly received and dumped. On August 5, 1974, 5,000
gallons of nonflammable materials were received at the site. Onsjte pits
were used as surface impoundments, where approximately 700,000 gallons of
o
waste were dumped.
Remedial measures at the site will be implemented in two phases.
Selected remedial actions approved at this time include excavation and
offsite disposal of buried materials (drums and trucks) at an approved RCRA
facility, installation of shallow ground water interceptors downgradient from
waste sources, collection and treatment of contaminated ground water,
recirculating the treated effluent to ponds, and discharging treated waste to
Mill Creek. A decision on remedial measures for contaminated soils, lower
sand and bedrock aquifers, final site closure requirements, and post closure
operations and maintenance has been deferred.
Remediation for this site had not begun at the time of this report, but
the air monitoring plan includes contingencies for the protection of the
public. As in most remediation plans, the site has been subdivided into an
exclusion zone (EZ) with potentially high air concentrations, a clean zone
(where the EZ may be entered and where decontamination takes place), and a
support area.
Air monitoring will be conducted during excavation in the exclusion
zone. If work levels measured with the HNU (intrinsically safe IS101)
monitor (or levels of benzene or chloroform measured with detector tubes) are
in a range from 10 to 50 ppm and remain constant for a period of 10 minutes,
the site will be put on alert. Air concentrations are expected to fluctuate
and increase while waste is uncovered, however, the levels will also be
reduced due to dispersion. The Health and Safety Officer will keep the
Project Manager appraised of the levels in the exclusion zone. The Project
Manager will designate an individual in the clean zone to check the levels in
-this zone as well as obtain perimeter readings. This information will be
"reported back via radio to the Project Manager.
Work will halt in the exclusion zone if levels reach 100 ppm and are
sustained for 10 minutes. Evacuation will take place and evacuations will
depend on wind direction. All clean zone employees will exit through the
25
-------
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
primary gate. The work party will exit a gate determined by wind direction.
They will, however, all exit the site at the same location. During
evacuation, the exclusion zone evacuating personnel and clean zone evacuating
personnel will remain in constant communication via radios. A head count
will be obtained to ensure complete evacuation.
A second situation, involving excessive levels in the clean zone, would
also warrant evacuation. If levels in this zone exceed 10 ppm, the Project
Managers will notify the employee's in the exclusion zone and work will halt.
The site will be evacuated in an orderly manner (abbreviated decon and use of
full facepiece respirator). Evacuation is necessary at this point due to
respirator cartridge limitations. It is possible, but not probable, that
concentrations in this zone could reach 10 ppm without the exclusion zone
reaching the 100 ppm evacuation point. Again, a head count will be obtained.
Reentry to the site will be made by a two-employee investigation team
(Health and Safety Officer and a member of the project management team).
Detector tubes, an HNU monitor, and a combustible/02 monitor will be used
during the investigation. Information on hot spots, suspect containers, and
air levels will be relayed via intrinsically safe two way radio
communication. At this time, the downwind perimeter monitor will be checked
and levels reported. The Project Manager will notify the EPA and the Local
Emergency Planning Commission (LEPC). A decision to evacuate the surrounding
areas will depend on the situation, airborne levels reported, and guidelines
contained in the Clean Air Act.
Any time that PID or OVA levels exceed 20 ppm in the work area, an
individual with an HNU meter will repeatedly walk 100 meters downwind of the
work area. He will repeatedly walk a 100 meter traverse with a center point
100 meters downwind of the work area. Wind direction will be continually
monitored during traversing, and the locations of the traverse endpoints
adjusted as necessary. If the fence line is nearer than 100 meters to the
work area, traverses will be done along the fence line.
During each traverse, the highest 10-minute average HNU reading will be
recorded. If fence line readings within the plume reach 10 ppm volatile
organic chemicals, as measured by the HNU meter, and are sustained at that
level for 30 minutes despite onsite control measures, monitoring will move to
the nearest downwind residence. At this location, 10-minute average HNU
readings will be taken continually traversing the property.
26
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
If offsite monitoring detects an average concentration of 2 ppm volatile
organic chemicals above background for 10 minutes, then the remediation
contractor, Clean Sites Inc. (CSI) will notify the Local Emergency Assistance
Agency as well as the Cecil County Fire Dispatch so that preparation for
evacuation may be initiated. The Director of Emergency Management will make
the decision to actually evacuate residents and will coordinate the
evacuation with the Cecil County Fire Department. CSI will also notify the
EPA RPM.
If at any time, specific circumstances indicate that there is an
imminent threat to the public health and safety, CSI will override the above
procedures and notify via telephone the Local Emergency Assistance Agency as
well as the Cecil County Fire Dispatch so that preparation for potential
evacuation may be initiated.
An interesting issue arose during negotiations between EPA and the
principal responsible parties (PRP's) regarding the types of instruments that
should be used and the community alert levels that should be established.9
In EPA's original proposal, evacuation would be triggered by residential HNU
readings of 2 ppm above background. The PRPs objected to this proposed
criteria, arguing that no adverse effects would result from exposure to much
higher concentrations of some of the compounds found at this site. To take
this factor into account, however, the monitoring system used during
excavation would have to distinguish among the various compounds.
EPA and the PRPs tried to develop a mutually acceptable monitoring and
evacuation strategy. At one point, the PRPs suggested using Draeger tubes to
determine concentrations of individual compounds, but EPA rejected this
method due to the interference that might be caused by the other compounds
found at this site. The PRPs then agreed to use a portable gas chromatograph
(GC) to measure concentrations of individual compounds, but did not agree
with the concentration limits proposed by EPA for those chemicals, which they
considered too conservative. They also became concerned about the ability of
the portable GC to measure concentrations for the wide variety of compounds
found at this site at the alert levels identified by EPA. In light of these
limitations and the inability to reach a consensus, the PRPs agreed to accept
EPA's original evacuation criteria (i.e., 2 ppm above background as measured
with an HNU meter).9
27
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
2.6 MCKIN SITE
Gray, Maine, is the location of a former waste collection, transfer, and
disposal facility operated by the McKin Company from 1965 to 1978. Onsite
waste handling procedures included discharge to the ground, storage in tanks,
incineration, and onsite burial. The site is approximately seven acres in
size. Neighboring lands include residential areas, wooded areas, and rural
farmland with the nearest home located approximately 200 ft from the site.
By 1983, all surface drums and tanks had been removed from the site in a
series of removal actions. Afterward two major contamination problems were
associated with the site. The first was onsite soil contamination which
served as a source for offsite ground water contamination. The second was
ground water contamination of the surficial and bedrock aquifers affected by
the site. Primary contaminants of concern in soils and ground water were
VOC's, particularly trichloroethylene (TCE), and 1,1,1-trichloroethane.
In July 1985, the selected remedial action was onsite aeration of soils
to remove volatile contaminants from soils, extraction and treatment of
ground water from offsite contaminated areas, and certain site removal and
closure activities. EPA established soil performance standards to protect
human health and the environment. For VOC contaminants, TCE was selected as
the indicator compound based on its prevalence, mobility, and toxicity. The
TCE performance standard established by EPA to evaluate soil treatment at the
McKin site was a maximum of 0.1 ppm averaged over a treatment volume of soil.
A soil aeration pilot study was conducted with continuous air monitoring
to evaluate methods of aerating soils for removal of TCE while controlling
air emissions to maintain acceptable air quality. Two private companies that
had potential liabilities, Fairchild Camera and Instrument Corporation and
Sanders Associates, agreed to perform a soil aeration pilot study for the
removal of TCE. The objectives of the pilot study were to determine the
effectiveness of a full scale soil aeration process, to determine optimum
operating conditions, and to assess the impacts of the process on ambient air
quality. The pilot study involved a series of conventional construction and
pollution control technologies used together with an innovative approach: to
aerate soils in an enclosed, heated environment and to capture the organics
vaporized from the soil. Several key pieces of equipment used in this
28
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
project were components of a portable asphalt batch plant. The major
components of the process used to excavate, transport, aerate, solidify, and
redeposit soils, and to treat contaminated air are presented in Figure 6.
A comprehensive air monitoring system was designed for the pilot
study, including the following components:
o Continuous monitoring of excavation, soil transfer, and aeration
for organic vapors using portable flame ionization detectors.
o Continuous monitoring for organic vapors at five permanent site
perimeter stations using five flame ionization detectors with
real-time data acquisition at 15-second intervals.
o Daily monitoring at ten local residences for organic vapors using a
portable flame ionization detector.
o Regular collection and analysis of air pollutants by 8-h charcoal
and Tenax tube adsorption and laboratory extraction. Samples were
taken at upwind and downwind site perimeter locations.
o Daily 24-h sampling for total suspended particulates at three
permanent site perimeter locations, using hi-vol samplers.
o Continuous monitoring for particulates at two permanent site
perimeter stations using real-time particulate analyzers and data
storage in an onsite computer system.
o Continuous monitoring and data storage of wind speed, wind
director, temperature, barometric pressure, humidity and solar
radiation during working hours as measured on an onsite
meteorological tower.
29
-------
I
1
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Three passes
through
^circulating
conveyor
system
Excavation
by caisson
digging bucket
within steel
caissons
_L
p
Materials
dryer
300°F
1
\
^ Cerr
i
•\
tent
er
Redeposition
in excavation
caissons
Exhausted
air
Baghouse
fines
Heated
screw
conveyer
Bag house
1
r
Scrubber
!
r
Vapor phase
carbon
adsorption
bed
Exhaust '
r
Figure 6. McKin pilot study treatment process.
30
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Contingency plans for corrective measures, volatilization abatement, and
public protective responses were based on site-specific guidance from the
Center for Disease Control [now called Agency for Toxic Substances and
Disease Registry (ATSDR)]. Among the recommendations were 1) continuous
monitoring for organic vapors near site activities and 2) public notification
if continuous downwind organic vapors at the site perimeter were more than 2
ppm above background. For the purposes of this monitoring and contingency
plan, the background level was assumed to be the reading of the most upwind
of the five perimeter flame ionization detectors. Remediation and monitoring
were performed by Canonie Environmental Services Corporation of Porter,
Indiana, with oversight provided by the EPA.
Results of air monitoring for organic vapors during the pilot study
indicated that onsite activities had negligible effects on air quality at the
perimeter of the 7-acre site. As monitored with portable onsite flame
ionization detectors calibrated to methane, excavation activities created the
most significant source of airborne VOC's. Total organic vapor
concentrations within 2 ft of a full caisson bucket or front-end loader were
as high as 1000 ppm. At a distance of 20 ft downwind of excavation
activities, however, 5-minute time-weighted average readings did not exceed 5
ppm above background during the pilot study.
Continuous monitoring for organic vapors at the site perimeter
demonstrated little evidence of onsite emissions of volatile organic soil
contaminants. Organic vapor levels of 2 ppm above background did not occur
at the site perimeter. Area background levels as measured upwind of the site
and at surrounding residences with flame ionization detectors varied from
about 1 ppm to 5 ppm during the study. Continuous background levels above
3.5 ppm occurred only during the early portion of the pilot study. During
the spring, background total organic vapor levels typically were 1 to 2 ppm
as measured on portable flame ionization detectors calibrated to methane.
Air monitoring results from 8-h sorbent tube sampling at the site
perimeter indicated that TCE concentrations in the ambient air ranged from
l«ss than 0.002 ppm to 0.01 ppm. Trichlorofluoromethane (Freon 11) was
'measured at slightly higher concentrations ranging from less than 0.010 ppm
to 0.018 ppm. Other compounds including 1,2-dichloroethylene, toluene,
ethyl benzene, and xylene were detected at levels of 0.01 ppm or less on
isolated occasions.
31
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Elevated ambient levels of suspended participates during a brief portion
of the pilot study represented the most significant air quality impact. On
several days during the latter portion of the pilot study, total suspended
particulate levels exceeded 110 /jg/m3 as measured during 24-hour sampling
periods at high volume samplers at the site perimeter. After dust control
measures were implemented visual dust emissions and high volume particulate
concentrations noticeably decreased, with the maximum 24-hour high volume
concentration less than 50
2.7 NYANZA VAULT SITE
The Nyanza site is located in Ashland, Massachusetts, approximately 25
miles west of Boston. The site is a privately owned active industrial
complex comprising approximately 35 acres. Between 1917 and 1978, numerous
companies that manufactured textile dyes and intermediates occupied the land.
The last of these dye manufacturing companies was Nyanza, Inc., which
operated from 1965 to 1978. Industrial wastes generated by these companies
were partially treated and the resulting chemical sludges were disposed
onsite in unlined lagoons and in an underground vault. In 1978, Nyanza, Inc.
12
ceased its operation at the Ashland facility.
In 1986, a State of Massachusetts DEQE site investigation team
discovered that an underground vault or settling basin, which is located in
the lower industrial area of the site, contained high levels of chlorinated
organic compounds and other constituents.
In 1987, a sampling team from the Massachusetts DEQE and an OSC from
EPA's Oil and Hazardous Materials Section took soil samples from the area
immediately downgradient of the vault. High concentrations of nitrobenzene
(1200 to 9100 ppm), chlorobenzene, aniline, dichlorobenzene and other
chemicals were found just below the ground surface (Table 8). Groundwater
was also heavily contaminated. Chemical Brook is located 60 to 70 feet
downgradient of the vault and runs directly by a residential area
approximately 250 to 300 yards from the site. Chemical Brook runs into the
Sudbury River, which empties into the Framingham Reservoir which is currently
used for recreational purposes, however, it may be considered for use as a
supplemental source of drinking water for the City of Boston.
32
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 8. MAJOR SAFETY THREATS FROM COMPOUNDS AT THE NYANZA VAULT SITE
Nitrobenzene Low permissible exposure limit (PEL),
toxic combustion products, inhalation
hazards
Chlorobenzene High flammability, reaction wiih caustic
agents, toxic combustion produ'cts
Aniline Low PEL, toxic combustion products,
inhalation hazards
Trichloroethylene Reaction with strong caustic agents
1,2-dichloroethylene Reaction with strong caustic agents
1,2,4-trichlorobenzene Reaction with strong oxidizers
Mercury
On April 23, 1987 a removal action was authorized. This site is now on
the NPL.
At the time of removal operations, MCL Development Corporation leased
part of the original property to Nyacol Products Incorporated. Nyacol's
plant directly abutted the property, which was the subject of the removal
action.12
In a 1987 memorandum to the OSC, a representative of the ATSDR concluded
that the VOCs detected in the soil downgradient from the underground vault
presented a potential threat to public health. Recommendations included air
monitoring at the site, especially when the soil in the area began to dry
out.13
Removal actions took place under an air supported dome. The purpose of
this building was to control or eliminate the release of volatile chemicals.
The dome was equipped with a carbon adsorption system that allowed for air
exchange within the building without contaminating the ambient air.
Excavated chemical sludges and soils were shredded within the building
and were transported .on a conveyor to a rotary kiln thermal destruction unit
located just outside the building. Excavation and shredding of materials
were conducted in such a manner as to reduce volatilization of chemicals
33
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
within the building structure. The conveyor to the outside was enclosed to
contain the volatilization of chemicals.
A small surface area was excavated and then covered. The material was
delivered to a shredder hopper which was kept covered. After shredding, the
materials were covered for final delivery to the thermal destruction
operation. Vapor control foam was kept on hand inside the building for use
on exposed chemical sludge areas to reduce chemical volatilization.
The first phase of EPA's operations covered excavation and destruction
of the materials within the vault. Then the building was moved to an area of
contaminated soil directly downgradient from the vault. Throughout both
phases, air monitoring was conducted within the building and at the site
14
perimeter.
The purpose of the air sampling program was to collect sufficient data
on volatile emissions from removal activities taking place inside the air
structure and from the incineration process conducted outside in the ambient
air. Samplers were located at 5 sites, both inside and outside the air
structure, as shown in Figure 7. Sampling and analytical methodologies were
designed to identify and quantify the release of six target compounds:
nitrobenzene, chlorobenzene, aniline, trichloroethylene, 1,2-dichlorobenzene,
and 1,2,4-trichlorobenzene. Data were interpreted and compared to worker
exposure limits [8-hour permissible exposure limits (PEL)] for determining
the impact that onsite activities were having on both the workers and the
14
extent to which emissions were migrating offsite and impacting residents.
Calibrated Dupont Alpha 1 and P125A personal constant flow air sampling
pumps were used to collect eight-hour samples on conditioned 1.5 grams TENAX
GC adsorbant, packed in 12.7 mm OD x 100 mm stainless steel tubes, which were
inserted into stainless steel sampling cartridges. Two eight-hour sampling
events were conducted during the operational phase of the removal and
incineration process. For each event, three samples were collected inside
the air structure collocated with real time instrumentation: one sample
(approximately 5 liters) was collected over an eight hour period; the other
-two samples (approximately 1 liter each) were run consecutively for four
"hours. All samples were collected in the breathing zone, three to five feet
above ground level.
34
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
xSite 5
D
Sitel*
fHlMllt
»CllWHt
• IHI
1, ^i ^•••SI • a
\_ cooos 3 At I I
>» HW»U ___ S >»
coritnotiioc.
-4060CALHC1.
, M.ON CAU
Figure 7. Sampling locations and meteorological tower.
35
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
One 20 liter sample was collected for eight hours outside the air
structure adjacent to the conveyor system. The sampler was positioned to
capture fugitive volatile emissions being emitted from this system.
One 20 liter sample was collected at a distance of 300 feet upwind of
the site conveyor system. This location provided background data for
comparing data generated from other sampling locations. ;
Two 20 liter samples per event were collected at a distance greater than
300 feet downwind of the site center. These locations provided data for
evaluating the extent that volatile emissions were migrating offsite.
The following instrumentation was used:
o HP 5970 Mass Selective Detector
o HP 5890 Gas Chromatograph equipped with a 60 meter VOCOL capillary
column
o HP 1000 computer using the RTE and Aquarius software
o Tekmar 5000 Thermal Desorber.
Results of the air sampling study are presented in Table 9. The trip
blank and lab blank both had levels below the lower limits of detection for
nitrobenzene (<4ng), chlorobenzene (<2ng), aniline (<30ng), trichloroethylene
(<2ng), 1,2-dichlorobenzene (<2ng), 1,3-dichlorobenzene (<2ng),
1,4-dichlorobenzene (<2ng), 1,2,4-trichlorobenzene (<2ng), and a
trichlorobenzene isomer (<2ng). The results of background sampling (Site 4)
and the downwind sampling (Site 5) showed that very little, if any,
contaminants migrated offsite. Levels were significantly higher, however,
inside the air structure (Site 1) and near the incinerator (Site 2), as
expected. Much lower concentrations were detected just outside the air
structure between the two Nyacol buildings (Site 3), with nitrobenzene being
the highest contaminant measured at 34.3 ppb. Inside the air structure and
next to the incinerator (where levels for certain compounds could not be
quantitative above their upper limits of detection), estimated concentrations
were calculated. For sample 3 (inside the air structure) nitrobenzene was
estimated to be approximately 660 ppb. Sample 2 (inside the air structure)
nitrobenzene, 1,2,4-trichlorobenzene, and trichloroethylene were estimated to
be approximately 1600 ppb, 445 ppb, and 550 ppb, respectively. Sample 4
(next to incinerator) nitrobenzene and 1,2,4-trichlorobenzene were estimated
to be approximately 270 and 70 ppb, respectively. These estimated values are
36
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 9. AIR SAMPLING RESULTS FROM NYANZA VAULT
SITE LOCATION
1
inside air
structure
m
1
inside air
structure
2
3 ft fron
incinerators
feed system
and hopper
3
outside air
structure
between the
two Nyacol
buildings
4
background,
140 feet
North of
Nyacol' s
office and
lab building
5
downwind,
300 feet
VBW of air
structure
along rail-
road tracks
SAMPLE
3
2
4
6
7
8
FLOW RATE
std( ml/fain)
6.6
5.6
19.4
38.2
43.4
37.7
SJWPLE TIME
(hours)
0910 - 1259
1259 - 1649
0910 - 1710
0910 - 1710
0900 - 1700
0910 - 1706
SAMPLE VOL.
(liters)
1.5
1
9
18
21
16
COMPOUND
nitrobenzene
chlorobenzene
aniline
trichloroethylene
1,2 dichlorobenzene
1,3 dichlorobenzene
1,4 dichlorobenzene
1,2,4 trichlorobepzene
trichlorobenzene
nitrobenzene
chlorobenzene
aniline
trichloroethylene
1,2 dichlorobenzene
1,3 dichlorobenzene
1,4 dichlorobenzene
1,2,4 trichlorobenzene
trichlorobenzene
nitrobenzene
chlorobenzene
aniline
trichloroethylene
1,2 dichlorobenzene
1,3 dichlorobenzene
1,4 dichlorobenzene
1,2,4 trichlorobenzene
trichlorobenzene
nitrobenzene
chlorobenzene
aniline
trichloroethylene
1,2 dichlorobenzene
1,3 dichlorobenzene
1,4 dichlorobenzene
1,2,4 trichlorobenzene
trichlorbenzene
nitrobenzene
chlorobenzene
aniline
trichloroethylene
1,2 dichlorobenzene
1,3 dichlorobenzene
1,4 dichlorobenzene
1,2,4 trichlorobenzene
trichlorobenzene
nitrobenzene
chlorobenzene
aniline
trichloroethylene
1,2 dichlorobenzene
1,3 dichlorobenzene
1,4 dichlorobenzene
1,2,4 trichlorobenzene
trichlorobenzene
AMOUNT
(nq)
>3297
26.5
ND <30
397
447 -
19.3'
115
1929
21.7
>3297
113
ND <30
>2009
1188
58
340
>2009
55.4
>3297
60
2225
826
1963
79.2
519
>2009
129
3104
28.69
ND <30
369.55
209.18
9.83
81.97
1801
13.3
4.4
ND <2
ND <30
ND <2
ND <2
ND <2
2.86
2.64
ND <2
7.78
ND <2
ND <30
ND <2
ND <2
ND <2
1.97
6.44
ND <2
CONCENTRATION
(DDb V/v)
>440 «
3.9
ND <5.3 t
49.1
49.6
2.1
12.8
174
2.0
>656 *
24.7
ND <7.9 t
>372 *
198
9.7
56.7
>272 *
7.5
>73 *
1.5
65.0
17.0
36.3
1.4
9.6
>30 *
1.9
34.3
0.3
ND <0.4 t
3.8
1.9
0.1
0.8
13.5
0.1
0.04
ND <0.02 t
ND <0.38 t
ND <0.02 t
ND <0.02 t
ND <0.02 t
0.02
0.02
ND <0.02 t
0.09
ND <0.02 t
ND <0.4 t
ND <0.2 t
ND <0.2 t
ND <0.2 t
0.02
0.05
ND <0.02 t
(Continued)
37
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 9 (Continued)
SITE LOCATION
trip blank
"
lab blank
SAMPLE
9
5
FLOW RATE
std(ml/Wn)
HA
NA
SAMPLE TIME
(hours)
NA
NA
SAMPLE TOL.
(liters)
(A
NA
COMPOUND
nitrobenzene
ohlorobenzene
aniline
trichloroethylene
1,2 dichlorobenzene
1,3 dichlorobenzene
1,4 dichlorobenzene
1,2,4 trichlorobenzene
trichlorobenzene
nitrobenzene
chlorobenzene
aniline
trichloroethylene
1,2 dichlorobenzene
1,3 dichlorobenzene
1,4 dichlorobenzene
1,2,4 trichlorobenzene
trichlorobenzene
AMOUNT
(ng)
NIX4 t
MX2 t
Nixaot
MX2 t
NIX2:t
NX2 t
NIX2 t
NIX2 t
NIX2 t
NIX4 t
NIX2 t
NDOOt
NIX2 t
NIX2 t
NIX2 t
NIX2 t
NIX2 t
WX2 t
OONCFNTRATION
(ppb v/v)
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
* - upper limits of detection
t - lower limits of detection
ND - nondetectable
NA - not Applicable
ng - nanograms
ppb v/v - parts per billion volume per voltme
38
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
above the highest standard used for quantitation; therefore, they should not
be compared to any air quality standard. They are only reported to indicate
the range at which these concentrations might exist. '
2.8 QUAIL RUN SITE
In the early 1970's, hexachlorophene and Agent Orange were produced in a
small industrial facility in southwest Missouri and 2,3,7,8-tetrachlorodi-
benzo-p-dioxin (TCDD) was formed as a byproduct. Waste materials containing
this dioxin byproduct were mixed with used oil and applied to roads and other
surfaces for dust control. Quail Run trailer park near St. Louis, Missouri,
was one of the places where this waste material was applied. The site
covered an area of approximately 11 acres and had an irregular shape as shown
by the site map (Figure 8).
Surface and subsurface soils at Quail Run were tested and were found to
contain 2,3,7,8-TCDD at concentrations above 1.0 ppb at most locations within
the site. A mitigation plan was prepared to control exposure to the
contaminated soil. This plan called for the removal of the contaminated soil
from the surface and for storage of the removed material in a safe location
onsite until detoxification procedures were available. An air monitoring
plan was developed to protect the general public in the immediate vicinity of
the cleanup operation.
A report from the Center for Disease Control (presently called ATSDR)
recommended 5.5 picograms per cubic meter (pg/m3) as an estimated no observed
effect level (NOEL). This level was based on chronic health effects. EPA
and CDC agreed to setting a "warning" or action level at 3.0 pg/m3. The
design of the monitoring network incorporated a measurement detection limit
in the range of 0.1 to 1.0 p/m3 to obtain reliable measurements at the 5 p/m3
level. Initially, 14 samples were collected at each monitoring site. With
the 14 data values at each of the six locations shown in Figure 9, a
definitive conclusion would then be reached if the data were 14 percent
either above or below the action level.
^' In order to obtain sufficient data to assess the effects of onsite
activities on the offsite ambient air under variable wind conditions, the
monitoring network was designed to provide long-term monitoring of the air at
or near the property boundaries of the site. On the basis of the physical
39
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
M - E MONITORS
SAMPLE
LOCATION
— CONTAMINATED
• AREA
Figure 8. Diagram of the Quail Run Site.
700
6.00
gfe 500
'
4.00
feg 30°
ig
53 200
1.00
0.00
t:a OBSERVED EFFECT LEVEL
—
—
—4 » o • $
• i i • i
i
. ACTION LEVEL
i i
O a l (
4 4 •
O NON-
DO WHVIND
Q DOWNWIND
• NON-
SAMPUNC
Figure 9. Results of monitoring versus action level at Quail Run,
40
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
configuration of the site, which is long relative to its width, a minimum of
six fixed monitoring locations were needed to ensure consistency throughout
the study and to have one upwind and one downwind sample for most days. This
would minimize the probability that contaminated air would pass between the
monitors without being detected. The three largest variables affecting this
"probability were the size of the area within the site that was re3easing
dioxin, local variation in the wind direction during a sample collection time
period, and the duration of the study. An onsite meteorological monitoring
station was incorporated into the network design for the purpose of obtaining
adequate wind speed and direction data.
Prior to finalizing the network design, a site visit was made to obtain
detailed information on the topography and to choose the specific sampling
locations relative to anticipated activities at the site. Air sampler
locations were selected so that they would be near, but just inside the
perimeter fence (for security purposes), would be consistent with the
accepted siting guidance for criteria pollutant monitoring, would provide
permanent placement throughout the life of the project (i.e., the samplers
would not have to be relocated during the course of the excavation activities
at the site), and would provide adequate coverage for most wind directions.
Sampling locations that were selected are shown in Figure 8.
Because of the high cost of sampling and analysis of dioxins, a
statistical analysis was carried out to determine the minimum number of
samples that could be collected to have a good probability (95% confidence
level) of showing an exceedance of the action level, if one occurred, and of
showing that no exceedance occurred if one did not occur. The number of
samples needed at each location was calculated as a function of the
difference between the action level (true mean) and the mean of a given
number of measured values.
To collect representative samples, commercially available modified
high-volume air samplers were used that employed both a filter for collecting
particulate matter and a solid adsorbent for collecting vapors. A known
-volume of air (calculated from the flow rate and time of sampling) was drawn
"through a sampling module and exhausted to the air via a 10-ft exhaust duct.
The upper portion of the sampling module holds a 4 in. diameter glass fiber
filter, which collects the particulate matter, and the lower portion consists
of a cylindrical glass cartridge (65 mm x 125 mm) containing a solid
41
-------
adsorbent, which entraps selected vapor-phase compounds. Polyether-type
polyurethane foam (PUF) plugs were use as the solid adsorbent material.
To measure the very low concentrations of dioxin required in this
project, a large sample volume of air was required. Because the air samplers
were only capable of providing a flow rate of approximately 0.280. m3/mi n, it
'was determined that the samples should be collected on a 24-h basris (+15%).
This time and flow gave sampled air volume of 300 to 400 m3 of air.
Samplers were placed on platforms to obtain samples of the ambient air
in the breathing zone. Initially, wind direction could not be predicted for
any 24-h sampling period. Therefore, all samplers were operated each day.
During the first 14 days of sampling, all of the samples were analyzed for
both particulate matter and 2,3,7,8-TCDD to obtain baseline data for the
site. Subsequent to this initial sampling period, only one upwind and one
downwind sample were submitted for 2,3,7,8-TCDD analysis each day. The
selection was based on the prevailing wind direction and the amount of
particulate matter collected on each filter for the sampling period.
Monitoring data were used for a daily risk estimation. The maximum
amount of dioxin a person located just offsite would experience during the
time of the cleanup activities was calculated. Data values from each monitor
were averaged separately because the levels from a single monitor are
representative of exposure for a person living near that monitor. An average
concentration was used because it is more consistent with an action level
based on chronic effects than is a single measured value. Because pollution
abatement actions had to be taken as quickly as possible after data were
available, a 14-day running average (average concentration of the most recent
14 days), was calculated daily. The detection limit of the instrument was
used as a measured value when calculating the running averages for samples
that did not contain a measurable concentration of 2,3,7,8-TCDD. Because
none of the nondownwind samples showed any measurable 2,3,7,8-TCDD, an
average of these numbers was used for those days when data were not
available. A daily data point might be missed for a monitor if that monitor
was neither upwind nor downwind of the site or if no work was occurring at
"the site. All samples were analyzed by an EPA contractor in accordance with
a Region VII standard method entitled Determination of 2,3,7,8-TCDD in Air
Samples Using Gas Chromatography-Mass Spectrometry.
Both system and performance audits were included in the air monitoring
plan to ensure that the established procedures were actually being followed.
42
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The audit process provided a means for continually evaluating the quality of
the data being generated, identifying apparent problems quickly, and making
changes to correct apparent problems.
During the course of the project, tests were conducted to-assess the
quality of data from the monitoring network. On the basis of the. results of
"field and laboratory audits, the procedures described in the monitoring plan
were being followed as written and with the exception of two data points, all
data were of acceptable quality.
During the course of the project, two experiments were conducted to
determine what fraction of the 2,3,7,8-TCDD and 2,3,7,8-TCDF
(2,3,7,8-tetrachloro-p-dibenzofuran) would pass through the filter and what
fraction would remain on the filter under the sampling procedures that were
established.
These experiments indicated that dioxin was very slowly migrating from
the filter to the PUF. Therefore, analyses of only the particulate matter or
only the vapor phase in the sample would give erroneous results. It also
appeared that the furans were more easily transferred from the filter to the
PUF than were the dioxins.
An analysis of monitoring data shown that the volume of air sampled was
just large enough to give adequate data precision (the maximum RSD at the
monitor showing highest concentrations of 2,3,7,8-TCDD was approximately 22%)
in the 3 to 5 pg/m3 concentration range.
Figure 9 is a graph showing the 14-day running averages and the
associated 95 percent confidence levels of those averages for the monitoring
site with the highest concentrations. Data points are representative
averages of measured or estimated concentrations for the most recent 14 days.
A downwind sample (labeled 0 in Figure 9) is one in which the monitor was
downwind of the cleanup activity during the last day of the 14-day averaging
period. A nonsampling value is shown for those days when the concentrations
at site 0 were estimated rather than measured on the last day of the
averaging period. The average of all measured (or estimated) concentrations
-for 14 consecutive days is plotted on the y axis against the last day of the
'14-day averaging period on the x axis for monitor 0. Detection limits were
taken as measured concentrations. If a measurement was not taken for a given
day because the monitor was not downwind or upwind on that day, an estimate
of the concentration was made by averaging all of the nondownwind values to
43
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
date. Because concentrations above the detection limit (usually in the range
0.4 to 0.8 pg/m3) were obtained only when a monitor was downwind of the site,
all estimated values were averages of detection limits. The data are
presented in this manner to illustrate the best estimate of the maximum
exposure of any offsite population that might have resulted from .this cleanup
activity. :
Note that all of the 14-day averages were well below the warning level
and the NOEL at all times. Therefore, concentrations of 2,3,7,8-TCDD that
cause an insignificant risk to the public can be measured in ambient air
using the procedures of this project. During this study the public was not
exposed to a significant concentration (5.5 pg/m3 for a "few months11) of
2,3,7,8-TCDD at any time.17
2.9 VERTAC SITE
The VERTAC Site is a Superfund site located in Jacksonville, Arkansas.
It was a World War II army ammunition plant that was converted to the
production of pesticides in 1948. The plant was converted to the production
of 2,4,5-T and "agent orange" during the 1960s and 1970s by the VERTAC
Chemical Corporation. Many parts of the property and its equipment became
contaminated with 2,3,7,8-TCDD. In 1982 the VERTAC Site was listed on the
NPL.
In 1985, VERTAC Corporation was under an order [based on Resource
Conservation and Recovery Act (RCRA)] to carry out site remediation. French
drains were constructed to capture storm water and tanks were built to settle
solids contaminated by dioxins. Solids were to be stored in above-ground
vaults. Construction activities stirred up dust that was potentially
contaminated with 2,3,7,8-TCDD.
During the period June 27, 1985 to October 25, 1985, an air monitoring
program was conducted by IT Corporation and EMI Consultants to determine
airborne concentrations of 2,3,7,8-TCDD in the vicinity of the remedial
18
activity. ° The sampling network consisted of four sampler locations that
Included three perimeter sites to evaluate the potential for airborne
migration of_TCDD beyond the boundaries of the work site and one internal
site to evaluate potential work exposure (Figure 10). A total of 236 ambient
air samples were collected of which 174 were perimeter samples. All samples
were analyzed for suspended particulate matter (PM). Thirteen of these,
44
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
e
OLD
• OUALIZATION
BASIN
NEA4OH
HILL '
• ••00
• •000
• ••00
• •000
E1BOO
• 1000
C «00
Figure 10. Ambient air sampler locations at VERTAC Chemical Corporation.
45
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
which exhibited the highest potential for measurable levels, were analyzed
for 2,3,7,8-TCDD. Two samples were analyzed for 2,4,5-T and 2,4-D.
Sampling for airborne levels of PM and TCDD was performed using a PUF
equipped high volume air sampler manufactured by General Metal-Works, Inc.
This system used a dual chambered, aluminum sampling module, which contained
'both particulate and vapor phase collection media. The upper chamber
supports a 4" diameter airborne particulate GFF. The lower chamber
encapsulates a glass cartridge, which contains the PUF for vapor entrapment.
A "Method Validation" study was performed to evaluate the effectiveness of
the PUF for collecting TCDD.
Above background PM concentrations were calculated by assuming the
lowest measured PM concentration during a sampling period was background and
subtracting this value from all measured PM concentrations during that time
period. Samples with the highest suspended particulate matter concentrations
were analyzed for TCDD.
Above background PM concentrations in the work area (No. 1) averaged 3.9
to 7.8 times higher than those measured at the perimeter samplers (Nos. 2, 3,
and 4). The average above background PM concentration in the work area was
34.9 A»9/m3 and the range was from 0.0 to 154.4 /*g/m3. The North perimeter
sampler ranged from 0.0 to 78.5 ^g/m3 and the average above background
concentration was 7.5 pg/m3. Above background PM concentrations at the South
perimeter sampler ranged from 0.0 to 6.1 pg/m3 and the average above
background concentration was 8.9 pg/m3. Above background PM concentrations
at the West perimeter sampler ranged from 0.0 to 107.6 ng/ma and the average
above background concentration was 4.5 /*g/m3.
Measurable levels of TCDD onsite ranged from nondetected to 14.9 pg/m3.
No measurable levels of TCDD were found at any of the perimeter sampler
locations.18
2.10 WEATHERFORD RESIDENCE
The Weatherford Residence is located at 25 Cordelia Drive in
"Jacksonville, Arkansas. Mr. Robert Weatherford is an ex-employee of VERTAC
Chemical Corporation at their herbicide production plant in Jacksonville.
Mr. Weatherford purchased three trucks and several clothes washers and
dryers from VERTAC. One of the trucks contained two railroad ties and about
60 gallons of debris, possibly dirt contaminated by still bottoms. He
46
-------
apparently used debris to fill in a 10 foot by 10 foot area adjacent to his
earthen driveway. The two railroad ties were used to support a shed located
in his backyard.19'20
On December 7, 1988, samples were collected from the International lift
truck, the Dodge flatbed truck, and the railroad ties. These samples showed
evidence of 2,3,7,8-TCDD, as follows: :
International lift truck:
scraped material from truck bed = >180 ppb 2,3,7,8-TCDD
exterior wipe - >340 pg/cm2 2,3,7,8-TCDD
interior cab wipe - 20 pg/cm2 2,3,7,8-TCDD
Dodge flatbed truck:
exterior wipe = >290 pg/cm2 2,3,7,8-TCDD
interior cab wipe « 20 pg/cm2 2,3,7,8-TCDD
Railroad tie:
bulk sample = >420 ppb 2,3,7,8-TCDD
During April 10 to 12, 1989, the TAT conducted extensive sampling of
surface and subsurface soil at Mr. Weatherford's residence. Two locations,
an area around the railroad ties and an area around the suspected soil burial
site, were determined to have 2,3,7,8-TCDD above the Agency's action level of
<1 ppb at the 95 percent UCL (Figure 11).
Remedial action involved removal of dioxin contaminated soil from Mr.
Weatherford's residence and transporting it to the VERTAC Superfund site.
Contaminated trucks and the railroad ties were to be transported to the
VERTAC site and decontaminated with steam and solvent. Removal activities
were performed by ERCs.
TAT conducted ambient air monitoring during the removal activities. The
objectives of air monitoring were to: 1) evaluate the potential for airborne
migration of dioxin contamination offsite (general population exposure)
during removal activities, 2) evaluate the potential for exposure of
populations at highest risk during the removal actions, and 3) assess the
adequacy of dust suppression methods being used on site as a function of the
'site-specific meteorological conditions.
Five ambient air monitors (General Metals PS-1 PUF sampler) were
permanently placed along the perimeter of Mr. Weatherford's property so that
the network would contain an upwind and a downwind perimeter monitor for any
given wind direction. A sixth monitor was located at an occupied residence
47
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Figure 11. Soil sample locations at Weatherford residence.
48
-------
judged to be at greatest risk of exposure (sensitive receptor). Air monitor
locations were selected with consideration for wind direction, excavation
areas, obstructions caused by buildings and trees, and probable access. Air
monitors were placed at a distance greater than 30 feet from the excavation
areas and monitor probes were at least twice the height of obstructions.
Monitors were securely positioned atop elevated platforms, placing the
21
sampler intake at a height of 2 to 4 meters above ground level.
49
-------
SECTION 3
DEVELOPMENT OF THE AIR MONITORING PORTION . *
OF A SITE CONTINGENCY PLAN
All Superfund remediation or.removal projects are required to have a
health and safety plan to protect workers. A contingency plan may be an
extension of the health and safety plan. Contingency plans for a site begin
with basic site control procedures. Then the plan covers emergency
situations to protect, as a minimum, onsite workers. In some cases the
offsite population will be addressed in the plan. Whether or not a
contingency air monitoring program is required is determined by analyzing the
site. If required, a monitoring program plan must be developed with the
objective of protecting offsite populations. A special analysis will be
required to determine the alert levels in the contingency plan and the
appropriate actions to take.
Note that this document is intended for use in the Superfund remediation
program. Although examples from the removal program are used to illustrate
the types of air monitoring that have been used in support of Superfund
cleanups, the guidelines are not directly applicable to removal cleanups.
When time permits and sufficient site date are available, however, many of
the procedures contained in this document could be used in developing an air
monitoring plan for a cleanup operation.
3.1 TYPICAL CONTENTS OF A SITE CONTINGENCY PLAN
To understand the context of a contingency air monitoring plan, the
nature of a contingency plan must be understood. Based on existing EPA
documents, the following subsections describe the types of information that
are usually provided.
-3.1.1 Site Control
To maintain a safe environment, activities at a Superfund remediation or
removal site must be controlled. A site control program should be instituted
50
-------
22 23
prior to site characterization and continue throughout site activities. '
Site control serves several purposes: it minimizes the potential for worker
contamination or injury, protects the public from the site's hazards,
facilities work activities, and prevents vandalism. Site control 1s
especially important in emergency situations. Site control procedures may
include: :
o Compiling a site map showing topographic features, prevailing wind
direction, drainage, and the locations of pits, ponds, and tanks.
The map should be updated throughout the course of site operations
to reflect changes in site conditions and activities. The map may
be used to plan site activities; to assign personnel; to identify
access routes, evacuation routes, and problem areas; and to
identify areas of the site that require the use of personal
protective equipment.
o Preparing the site for cleanup activities. Preparation includes
constructing roads, removing physical hazards, installing antiskid
devices, constructing operation pads, constructing docks, and
installing electrical wiring. Preparation can be as hazardous as
cleanup. For this reason, extreme care for worker safety must be
taken.
o Establishing work zones. To reduce the accidental spread of
hazardous substances by workers from the contaminated area of the
site to the clean area, hazardous waste sites should be divided
into as many different zones as necessary to meet operational and
safety objectives. Three frequently used zones are (Figure 12):
The Exclusion Zone, i.e., the contaminated area of a site
The Contamination Reduction Zone (CRZ) where decontamination
takes place
The Support Zone, i.e., the uncontaminated area within which
hazardous conditions should not exist.
o Defining separate zones and tracking entry and exit from these
zones helps ensure that personnel are protected against hazards,
that work activities and contamination are confined to the
appropriate areas, and that personnel can be located and evacuated
in the event of an emergency.
o Using the buddy system when necessary. Most activities in
contaminated or otherwise hazardous areas should be conducted with
a "buddy" who is able to provide assistance, observe for signs of
-"" chemical or heat exposure, periodically check the integrity of his
or her partner's protective clothing, and notify others if
emergency help is needed. The buddy system alone may not be
sufficient to ensure that help will be provided in an emergency.
At all times, workers in the Exclusion Zone should be in
line-of-site contact or communications contact with a person in the
Support Zone.
51
-------
Hotline forming the
boundary between
the CRZ and the
Exclusion Zone.
Estimated boundary
of area with highest
Contamination
Control Line
Support Zone
Command Post
(Upwind of Exclusion Zone)
D
Access Control Points which control the flow of personnel and
equipment into and out of the Exclusion Zone
Decontamination Reduction Corridor where decontamination takes place.
Contamination Reduction Zone (CRZ).
Exclusion Zone. Buffer zone between CRZ and area of highest
contamination.
Figure 12. Site work zones.
52
-------
o Establishing and strictly enforcing decontamination procedures for
both personnel and equipment.
o Establishing site security measures to prevent the exposure of
unauthorized, unprotected people to site hazards; to prevent theft;
to avoid increased hazards from people seeking to abandon other
wastes on the site; and to minimize interference with safe working
procedures. Security measures may include establishing"an ID
system, erecting a fence, posting signs, hiring security guards,
and enlisting public law enforcement agencies.
o Setting up communication networks. An internal communications
network is required to alert workers to emergencies, pass along
safety information, communicate changes in the work to be
accomplished, and maintain site control. An external communication
system between onsite and offsite personnel is necessary to
coordinate emergency response, report to management, and maintain
contact with essential offsite personnel.
o Enforcing safe work practices, including a list of standing orders
stating practices that must always be followed and those that must
never occur in the contaminated areas of the site (Figure 13).
The degree of site control needed depends on site size and
characteristics, and on the proximity of the surrounding community. The site
control program should be established in the planning stages of a project and
should be modified as necessary based on new information and site
22 23
assessments. '
53
-------
FOR PERSONNEL ENTERING THE CONTAMINATION REDUCTION ZONE:
No smoking, eating, drinking, or application or cosmetics in this zone.
No Batches or lighters in this zone.
Check in at the entrance Access Control Point before you enter this
zone.
Check out at the exit Access Control Point before you leave this zone.
FOR PERSONNEL ENTERING THE EXCLUSION ZONE:
No smoking, eating, drinking, or application or cosmetics in this zone.
No matches or lighters in this zone.
Check in at the entrance Access Control Point before you enter this
zone.
Check out at the exit Access Control Point before you leave this zone.
Always have your buddy with you in this zone.
Wear an SCBA in this zone.
If you discover any signs of radioactivity, explosivity, or unusual
conditions such as dead animals at the site, exit immediately and report
this finding to your supervisor.
Figure 13. Sample standing orders.
54
-------
3.1.2 Site Emergencies
Unforeseen circumstances may suddenly create unexpected emergencies at
Superfund remediation or removal sites. An emergency may be as limited as a
worker experiencing heat stress, or as vast as an explosion that spreads
toxic fumes throughout a community. Some common causes of emergencies
include fire or explosion, chemical leaks, chemical reactions, container
collapse, release of toxic vapors, heat stress, personal protection equipment
(PPE) failure, and physical injury.
Site emergencies are potentially complex because uncontrolled toxic
chemicals may be numerous and unidentified, and their effects may be
synergistic. Advance planning, including anticipation of emergency scenarios
and through preparation for contingencies, is therefore essential to protect
worker and community health and safety. Emergency response delays of minutes
can create life-threatening situations; the rapidity of response can mean the
difference between life and death. Therefore, it is essential that personnel
be able to immediately respond or rescue, and that equipment be on hand and
in good working order.
A Contingency Plan that sets forth policies and procedures for
93 p« pC pe
responding to emergencies should be developed for each site. ' ' ' A
Contingency Plan is a written document (generally a separate section within
the Site Safety Plan) that usually incorporates the following:
o All individuals and teams who will participate in emergency
response, and their roles, responsibilities, and lines of
authority.
o A detailed site map showing the locations and types of hazards,
site terrain, evacuation routes, refuges, decontamination stations,
and offsite populations at risk.
o Procedures for communicating onsite (e.g., bullhorns, sirens, hand
signals) and offsite (e.g., key phone numbers, contact names,
two-way radio).
o Equipment necessary to rescue and treat victims, to protect
response personnel, and to mitigate hazardous conditions on the
_. site.
o Medical treatment/first aid techniques.
o Emergency response procedures that encompass all phases of response
operations, from initial notification through preparation of
equipment and personnel for the next emergency.
55
-------
o Procedures for emergency decontamination, including decontaminating
the victim(s), protecting medical personnel, and disposing of
contaminated PPE and decontamination solutions.
Figure 14 outlines a possible flow of operations during an actual
emergency. Operations can be divided into three categories:
o Preparation, which involves assessing the situation, allocating
personnel and equipment for response, and requesting aid from
outside sources.
o Response, which involves rescuing, decontaminating, and treating
victims; evacuating personnel and/or the public as necessary; and
controlling the hazard.
o Followup, which involves replacing equipment, documenting the
incident, and reviewing and revising the Site Safety and
Contingency Plans.
3.2 DETERMINING A NEED FOR CONTINGENCY AIR MONITORING
The need for contingency air monitoring is established by analyzing the
site and the remediation or removal to be carried out. The two methods of
analysis are the air pathway analysis (APA), conducted under the supervision
of EPA, and the health assessment, conducted by the ATSDR. These assessment
methods are briefly described in the following subsections.
3.2.1 Using Air Pathway Analysis
One method for determining the need for contingency air monitoring is
26
the APA. As indicated in the Air/Superfund National Technical Guidance
Study (NTGS) Series, APA is applicable to every activity in the Superfund
process.
56
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Potential offsite impacts are an integral component of the Superfund risk
assessment process and should be considered in remedial
investigations/feasibility studies (RI/FS), remedial designs, and planned
removal actions.
According to the NTGS, Volume 1, an APA conducted for a RI/FS should
involve the following steps: :
Step 1 - Identify and evaluate potential applicable relevant and
appropriate requirements (ARARs) governing the air pathway
for remediation/removal sources.
Step 2 - Perform routine air monitoring during the remedial and removal
operations.
Step 3 - Implement a combination of modeling and monitoring techniques
to characterize nonroutine air releases.
When carrying out an APA for remediation planning, a combination of
monitoring and modeling techniques should be used to characterize unplanned
releases. Dispersion modeling can be used to extrapolate monitoring data
from the source to the downwind receptor locations of interest.
An APA for remediation planning consists of the following five steps
Step 1 - Review Existing Site Information
First, information and data relevant to a site's potential air impacts
are reviewed, and data inputs for a modeling exercise or monitoring program
are developed. This information consists of source and pollutant data,
receptor data, and environmental data.
Source and Pollutant Data
Numbers and types of potential air sources located at the site should be
identified and contaminants listed. The potential for each contaminant to be
released to the atmosphere should also be listed. Characteristics of air
pollutants should be identified (i.e., gaseous or particulate, or a toxic
constituent adsorbed onto dust particles). For a modeling exercise, source
dimensions for area and volume sources and stack parameters for elevated
sources, need to be listed.
Environmental Data -
Environmental data encompasses climatology, topography, land use
classification, and meteorology. Climatological data are usually in the form
58
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
of wind roses that identify the frequency of occurrences of all wind
directions. Local topography, such as shown on a topographic map, can
influence pollutant transport. Topographic features can channel and divert
large-scale regional wind flow such that wind direction onsite-may be much
different from measurements taken offsite. Land-use classificatipn affects
'"whether the area should be modeled as "urban" or "rural," a choice that must
be made in most dispersion models.
Meteorological data are important to dispersion models. A procedure for
developing an onsite meteorological data base is included in Volume IV of the
NTGS.
Emission rates from potential onsite sources can be determined according
to the procedures outlined in Volumes II and III of the NTGS. Short-term
maximum emission rates and long-term averages should be developed. Due to
the complexity of emissions mechanisms for some Superfund sources, the
process of specifying an emission rate may involve a fairly complex protocol,
including field measurements of emissions or monitoring and
"back-calculating" an emission rate based on an assumed concentration
distribution.
When developing information in preparation for monitoring, the critical
concentrations of each pollutant should be identified, i.e., what are the
ARARs, concentrations, and risk levels that will determine whether or not a
concentration is acceptable. Risk assessments for carcinogens and systemic
toxicants for which reference doses have been established are frequently
bases upon long-term averages. Nonetheless, contingency plan action limit
air concentrations will typically be based upon exposures occurring only
during the time period of the cleanup operations. For a few very toxic
chemicals which can have an immediate adverse health effects, the action
limits can be based upon very short term average concentrations (time of
exposure ranging from 15 minutes up to 8 hours) or an instantaneous
concentration.
Receptor Data
A dispersion model can calculate concentrations at any location.
----- Usually, a gridded receptor field is mapped in the model to identify
concentration gradients and maximum concentrations. Population in the
general vicinity of the site and the location of individual residences near
59
-------
the site should be determined. Sensitive receptors (e.g., hospitals,
schools) should also be identified. This information can be used to design a
receptor grid to interpret modeling results in terms of exposure to maximum
concentrations, and to design a monitoring network focusing on areas of
greatest concern.
'Step 2 - Select APA Sophistication Level :
Next, the level of sophistication to be employed in the analysis is
decided. Modeling procedures will usually begin with a screening model.
Steps to refine the analysis are taken only if screening results indicate
unacceptable concentrations. For most Superfund sources, the Industrial
Source Complex (ISC) model, in its short-term (ISCST) or long-term (ISCLT)
version, is directly applicable. This model can be run in a screening mode
for short-term predictions. An updated version of EPA's screening procedures
for point sources contains a computerized SCREEN model that can also be
applied to Superfund sources. The choice of APA model sophistication level
depends on what levels of detection will provide the site manager with
meaningful information, and what lead time is acceptable.
Step 3 - Develop an APA Protocol
A APA employing modeling should be documented in a protocol that
describes how the analysis will be carried out. The protocol should show
what sources will be modeled and how emissions will be calculated. Source
characterization, including sizes and initial dispersion for area sources,
and stack parameters for point sources, should be specified. Other important
topics for the protocol are selection of meteorological data, specification
of a receptor grid, choice of model, a detailed list of model options, and
background concentrations.
Step 4 - Conduct the APA
This step involves carrying out the selected APA through modeling or
monitoring or a combination of the two. Qualified personnel must conduct the
APA to ensure that all QA/QC elements of the monitoring plan are followed and
.that APA results are reported and displayed. Modeling results can be used to
generate isopleths of concentrations around a site. Superimposing isopleths
on a site map is a useful way to display results.
60
-------
Step 5 - Summarize and Evaluate Results
Monitored and modeled concentrations need to be evaluated to determine
1f critical concentrations might be exceeded during remediation. If an
emergency condition could cause offsite populations to be jeopardized or to
experience undesirable exposures, a contingency contingency monitoring
'program is justified. :
3.2.2 Using Health Assessments
A second method for determining a need for contingency monitoring is a
health assessment. The ATSDR was originated to implement the
health-related sections of CERCLA 1980, as amended. The primary vehicle for
meeting this mandate is the health assessment. When a health assessment is
carried out concurrent with an RI/FS, it may help to establish the need for a
contingency plan and air monitoring program to protect offsite populations.
Nature of a Health Assessment--
An ATSDR Health Assessment is an analysis and statement of the public
health implications of the facility or release under consideration. ATSDR
Health Assessments are based on factors such as the nature, concentration,
toxicity, and extent of contamination at a site, the existence of potential
pathways for human exposure, the size and nature of the community likely to
be exposed, and any other information available to ATSDR that is relevant to
a determination of potential risks to public health. This analysis and
attendant health recommendations are based on professional judgments and the
weight of evidence. In this respect, Health Assessments are similar to the
Hazard Identification step of risk assessment.
Basically, every Health Assessment includes the following six steps
(Figure 15):
1. Evaluate information on the site's physical, geographical,
historical and operational setting, and identify health concerns of
the affected community(ies).
2. Determine contaminants of concern associated with the site.
^ 3. Identify environmental pathways.
61
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Evaluation of
Site Infonrwtfon
Determination of
Contaminants of Concam
Identification and
Evaluation of
Environmental Pathways
1
Identification and
Evaluation of
Human Exposure Pathways
CHmate
Sol Types
Hydrogeotogic Information
Surface Cover
Land Use
Water Use
StteAccesslMity
Determination of
Public Health Implications
Recommendations and
Conclusions
Figure 15. Factors influencing the Health Assessment Process.
62
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
4. Identify human exposure pathways.
5. Determine public health implications based on available medical and
toxicological information.
6. Determine conclusions and recommendations concerning "the health
threat posed by the site.
A Health Assessment is written for the "informed community associated
with the site", which would include citizen groups, local leaders, health
professionals, and other government agencies (e.g., EPA, State health
agencies, and environmental agencies). As more complete information is
collected and evaluated, the conclusions and recommendations of the health
assessment may be modified or altered to reflect the public health
implications of additional information. Note that an RPM or OSC may request
a health assessment from ATSDR.
Both Comprehensive Environmental Response, Compensation, and Liability
Act (CERCLA), as amended by Superfund Amendments and Reauthorization Act
(SARA), and the RCRA, as amended by the Hazardous Solid Waste Amendments of
1984, permit concerned parties to petition ATSDR to conduct a Health
Assessment. ATSDR has promulgated regulations on the petitioned health
assessment process (42 CFR Part 90). These regulations were published for
comment in 53 Federal Register 32259-32263, 24 August 1988.
Difference Between Health Assessments and Risk Assessment--
Deliberate differences exist between ATSDR's Health Assessments and
EPA's Risk Assessments. The two agencies have distinct purposes that
necessitate different goals for their assessments. Risk Assessments include
one or more of the following components: hazardous identification,
dose-response assessment, exposure assessment, and risk characterization.
Statistical and biological models are used in quantitative and
compound-oriented risk assessments to calculate numerical estimates of risk
to health using data from human epidemiologic investigations (when available)
and animal toxicology studies. The product of quantitative risk assessment
is a numerical estimate of the public health consequences of exposure to an
_agent. EPA Risk Assessments are used in risk management decisions to
establish cleanup levels; set permit levels for discharge, storage, or
transport or hazardous waste;and determine allowable levels of contamination.
63
-------
Health Assessments conducted by ATSDR use the components of risk
assessment that stress the hazard identification component. Although Health
Assessments may use quantitative data, they are usually qualitative in
nature, focus on medical and public health perspectives associated with a
site, and discuss sensitive populations, toxic mechanisms, possible disease
outcomes, and especially community health concerns. Based on the Health
Assessment findings, health advisories or additional health studies may be
initiated.27
Thus, while a Risk Assessment conducted under EPA's RI/FS process might
lead to the selection of a particular remediation measure at a site, an ATSDR
Health Assessment may be used by local health professionals and residents to
understand the potential health threats posed by a specific waste site and
may lead to further health actions or studies. One example of an action in
response to a health assessment could be a contingency plan and air
monitoring program.
3.3 DESIGNING A CONTINGENCY AIR MONITORING NETWORK
The objective of a contingency air monitoring plan is to document the
pC
design specifications for a site-specific monitoring program. The plan
should be submitted for agency technical review and RPM/EPM approval.
According to NTGS, Volume 4, developing a site monitoring plan involves the
following major elements as illustrated in Figure 16.
o Select the air monitoring constituents
o Specify the time frame for decisions
o Specify the meteorological monitoring constituents
o Design the air monitoring network
o Document the air monitoring plan.
Contingency air monitoring is likely to be one component of a
comprehensive air monitoring plan at a site. Thus many principles of a broad
program also apply to contingency monitoring. This section, however, will
emphasize the part of the plan that is to protect offsite populations.
3.3.1 Select Air Monitoring Constituents
The selection of compounds to be addressed in a monitoring program is a
challenging task because of the extensive number of potential contaminants at
Superfund sites. Technical limitations and budget limitations generally
64
-------
Input Data
EPA Guidance
Select
Air Monitoring
Constituents
i
Specify
Meteorological
Monitoring Parameters
Monitoring
Constituents
Target List -
Other Technical
Guidance
Design Air
Monitoring Network
Document Air
Monitoring Plan
T
CONDUCT MONITORING
Figure 16. Development of a (Contingency) Air Monitoring Plan.
65
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
necessitate the selection of a limited subset of target compounds. Selection
of target contaminants involves the following factors:
o Physical and chemical properties of chemicals (e.g., physical
phase, volatibility, and water solubility)
o Potential health effects of the chemicals (usually based on risk
assessment) :
o Estimated air concentration of a contaminant relative to other
source constituents
o Availability and performance of standard sampling and analysis
methods
o Project objectives
o Resource constraints.
Compounds included in the Hazardous Substances List (HSL) developed by
EPA for the Superfund program are listed in the NTGS, Volume IV. This list
is a composite of the Target Compound List (TCL) for organics and the Target
Analyte List (TAL) for inorganics. Thus the HSL represents a comprehensive
list of compounds from which target air toxics compounds can be selected for
a particular site.
Emission rate measurements, air modeling results, air monitoring data
from the site, and ARARs identified during previous studies should be used to
identify target compounds for air monitoring. In order to rank the relative
importance of compounds, a hazard index (HI) should be calculated. HI is the
ratio of the estimated (expected) concentration divided by the appropriate
health-based action level. HI values should be ranked from highest to lowest
to develop a priority list of candidate target compounds. Compounds selected
for air monitoring should be a function of the estimated HI, and the
technical feasibility of collecting and analyzing the various compounds.
The target compound list should be periodically reevaluated, and revised
if warranted, as monitoring results become available. This is particularly
useful for air monitoring programs conducted during remedial actions.
Periodically (e.g., monthly), a more comprehensive list of compounds may be
•sampled and analyzed to confirm the representativeness of the routine target
compound list.
66
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
3.3.2 Specify Meteorological Constituents
A meteorological monitoring program should be an integral part of a
contingency air monitoring program. A meteorological survey can be used to
design the air monitoring network based on local wind patterns.-
Meteorological data can be used for the interpretation of downwind air
"concentration data and exposure conditions offsite.
The required number and location of meteorological stations depends on
local terrain. One meteorological station is generally sufficient for
flat-terrain sites. For complex-terrain sites, multiple stations may be
necessary to represent major onsite/offsite air flow paths. Generally, one
to three stations will be sufficient. Meteorological stations should be
located away from any nearby obstruction at a distance equal to at least 10
times the height of the obstruction.
Meteorological monitoring parameters can be classified as follows:
o Primary parameters
Wind direction
Wind speed
Sigma theta (horizontal wind direction standard deviation,
which is an indicator of atmospheric stability)
o Secondary parameters
Temperature
Precipitation
Humidity
Atmospheric pressure
Primary parameters are representative of site dispersion conditions and
should be included in all meteorological monitoring programs. Secondary
parameters are representative of emission conditions and are generally only
recommended for refined air monitoring activities.
Recommended meteorological monitoring system accuracies/resolutions and
sensor response characteristics are summarized in the NT6S, Volume IV. A
meteorological survey may be conducted to support air monitoring network
design. It may be necessary, however, to use historical offsite data to
-estimate seasonal effects for planning purposes if the air monitoring program
'is scheduled to last for more than a few months.
67
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
3.3.3 Design the Monitoring Network
Design of a contingency air monitoring network will be affected by
site-specific factors such as source characteristics, receptor locations, and
environmental characteristics, and thus must be decided on a case-by-case
basis. Components of the monitoring network design include: -
,»
o Number and location of monitoring station
o Probe siting criteria
o Program duration and measurement frequency
o Sampling and analysis methods
o Air monitoring equipment
The number and location of monitoring stations for an air monitoring
network depend on the following site characteristics:
o Results of APA air dispersion modeling and monitoring
o Environmental characteristics (e.g., meteorology, topography, soil
characteristics)
o Receptor characteristics (e.g., population centers, sensitive
populations, residence locations, and estimated locations of high
concentrations of air contaminants)
o Source characteristics (e.g., type and extent of contamination,
locations of hot spots)
o Siting constraints
o Duration of the monitoring program.
Contingency air monitoring programs that last for 2 weeks or less
require some judgment about the placement of monitoring stations and their
numbers. Historical meteorological data would generally not provide accurate
information on the meteorological conditions for the few days of sampling and
analysis. A meteorological survey conducted just prior to air monitoring,
however, can help to identify expected wind patterns and downwind sampling
locations, and help to characterize wind direction variability.
Meteorological forecasts can also be used to deploy air sampling equipment.
Air monitoring station numbers and locations are highly site-specific,
tiowever, a single downwind stationary monitor is not adequate to monitor for"
maximum concentrations. Placement of air monitoring and meteorological
stations must conform to a consistent set of criteria and guidance to ensure
data comparability and compatibility. Factors to be considered in the
placement of air quality monitoring stations are:
68
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
o Vertical placement above ground
o Horizontal spacing from obstructions and obstacles
o Unrestricted air flow
o Spacing from roads.
Monitoring duration and frequency depends on the specific "project
objectives and resources. A representative number of air samples-should be
collected during the project to ensure a reasonable data base. The number of
representative samples depends on many factors and guidelines for estimating
the required number are given in NTGS, Volume IV. The recommendations
specified in NTGS are based on the following factors:
o Augmentation of integrated sampling with continuous monitoring for
steps that require more detailed data to enhance the data base
o The resource requirements for laboratory analysis for organic and
inorganic compounds
o QA/QC requirements such as collocated field and trip blank samples
and spike samples.
Selection of air monitoring methods and equipment should be based on a
number of factors, including the following:
o Physical and chemical properties of compounds
o Relative and absolute concentrations of compounds
o Relative importance of various compounds in program objective
o Method performance characteristics
o Potential interferences present at site
o Time resolution requirements
o Cost restraints.
Various classes of contaminants must be monitored by different methods,
depending on the compounds and their physical/chemical properties. One
factor that affects the choice of monitoring technique is whether the
compound is a gas, an aerosol, or is adsorbed to solid particles.
Screening for the presence of highly toxic air constituents involves
techniques that are rapid, portable, and provide real-time monitoring data.
Air contamination screening will generally be used to confirm the presence of
an onsite release. Quantification of individual compounds is not as
.-important during screening, however, the technique must have sufficient
"specificity to differentiate hazardous constituents of concern from potential
interferences, even when the latter are present in higher concentrations.
Detection limits are usually much higher for screening devices than for
69
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
quantitative methods, and unfortunately, may be higher than alert levels
that are desirable to employ. Appendix A contains the characteristics of the
HNU photoionizer and the organic vapor analyzer for future reference. These
two instruments are commonly used to obtain real-time rapid monitoring data.
This material was derived from the EPA Standard Operating Safety .Guides,
'Appendix I, published in July 1988. ;
Laboratory analytical techniques provide identification of components
and accurate measurement of concentrations. Preconcentration and storage of
air samples will usually be required. Therefore, refined monitoring
techniques usually involve a longer analytical time period, more
sophisticated equipment, and more rigorous QA procedures. Turnaround time
for data is a key factor to evaluate when considering offsite analyses.
3.3.4 Document the Air Monitoring Plan
A site-specific air monitoring plan, including contingency monitoring,
should be documented to facilitate implementation. In addition, EPA requires
that any project involving environmental measurement must have a QAPP. The
QAPP, which is distinct from any general project plan, describes the
organization of the project and the assignment of responsibility for those
specific QA/QC activities required to meet the project data quality
objectives (DQOs). The following is a list of subjects addressed in a
typical QAPP:
o Project description
o Project organization and responsibility
o Facilities, services, equipment, and supplies
o DQOs for measurements
o Sample collection
o Sample custody
o Calibration procedures
o Laboratory analysis procedures
o Data management
o Recordkeeping and documentation
o Internal QC checks
o External QA audits
o Preventive maintenance
o Procedures to assess data quality
.^- o Feedback and corrective actions
o Quality assurance reports
o Review and.approval of QAPP
Authority for final approval of the contingency air monitoring plan is with
the RPM/EPM.
70
-------
3.4 CASE EXAMPLE USING REVERSE RISK ASSESSMENT
This case example uses a hypothetical Superfund site that is a
composite of several real sites. This example illustrates a procedure that
may be used to determine whether contingency monitoring is needed, and to
determine action levels for a contingency monitoring plan.
The usual process of risk assessment may be likened to looking down a
gun barrel from the source of emmissions to the receptor. In this example
the barrel was turned around to look down the barrel from the receptor to the
source of emissions. To determine whether or not monitoring would be needed
at a remediation site, the following steps were taken:
1. At the location of the maximum exposed individual (MEI), determine
the maximum concentration of chemicals to which the MEI may safely
be exposed.
2. Conduct dispersion modeling to determine the dilution ratio from
source to receptor.
3. From the dilution ratio, determine the maximum emission rate that
will allow the MEI to be protected.
4. Estimate the average and highest possible emission rates to
determine whether the maximum emission rate could possibly be
exceeded.
If these steps lead to the conclusion that contingency monitoring will
be needed, then alert levels were required as part of the monitoring plan.
These levels were derived from the levels of public protection established in
step 1.
3.4.1 Site Description
The site was a ten acre square (201.2 by 201.2 meters) that contained
o
5000 yd of contaminated soil. The contaminated zone was an area 91.4 meters
wide and 22.9 meters long located in the middle of the site. Five volatile
compounds were present in the soil. These chemicals were benzene,
1-butanol, methylene chloride (MC), methyl ethyl ketone (MEK), and o-xylene.
According to the soil samples taken from the site, these chemicals were
tibmogeneously distributed, with no major hotspots. The site plan is shown in
Figure 17. _ _
The site was bounded on the north by single family residences, bounded
on the east by an apartment complex, and bounded on the west by undeveloped
71
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
201.2 m
Horth
< 91.4 m
Contaminated soil 22.9 m
100.6 m
Haul
road
201.2 m
Figure 17. Example site configuration.
property (a vacant lot and a drainage easement). A paved public road
bounded the property on the south side. Across the road from the site was
a partly developed industrial property which was unoccupied.
This site once contained drums of chemicals. In a previous removal
action, several hundred drums were removed and the site was capped with a
layer of relatively impervious soil. Since that time the cap had become
partly contaminated from below. Contaminated soil was the subject of the
remediation action under consideration. An interesting note from the
previous removal action was that the drums were removed in various stages
of disintegration. Some were empty. While the soil sampling procedure did
not reveal any hot spots, there were incomplete records of the disposition
of chemicals from the drums.
The selected remedial option was excavation and offsite thermal
treatment. The remediation period was expected to last for two months and
would be carried out during the summer. On each day of remediation,
removal activities would take place during an 8-hour period. The sequence
of remediation steps was excavation (and exposure of contaminated soil),
dumping soil onto trucks, and transport down the haul road to exit from the
72
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
site. Excavation would remove a portion of the soil cap and a portion of
the contaminated soil. The haul road was 100.6 meters long. Emissions
from offsite activities were not taken into account in this example because
they did not affect persons next to the site.
3.4.2 Health Protection Levels
:
In an EPA risk assessment procedure , health protection levels are
determined for the MEI. The MEI may be a person more or less sensitive to
toxic pollutants than the general public and may or may not live at the
fenceline of the site. This approach does not address the problem of
meeting ARARs and other standards to be considered (TBCs). The level of
protection is based solely on protecting health.
In this example, the MEI was assumed to be an adult sensitive to the
effects of pollutants who lived at the fenceline directly north of the
contaminated soils and in the path of prevailing winds. Selection of this
worst-case MEI assumed that other members of the public would automatically
be protected to a greater degree than the MEI.
The RPM, or other responsible party, may consult with a toxicologist
to determine the appropriate levels of protection for the MEI. These
levels are based on potential hazards associated with the chemicals known
to be on the site and the activity patterns and pollutant sensitivity of
the MEI. These protection levels may be expressed as maximum allowable
concentrations at the location of the MEI.
In this example, the RPM consulted an EPA toxicologist to determine
health protection levels for the MEI. The toxicologist used data from the
28
Health Effects Assessment Summary Tables (HEAST), as shown in Tables 10
and 11, in his evaluation. These tables have been abbreviated to show data
for the inhalation route of exposure only, because the oral route does not
apply in evaluating air exposures to volatile organic chemicals. The HEAST
tables summarize reference doses (RfD) for toxicity due to subchronic and
chronic inhalation exposure, and provide unit risk factors and unit risk
slope values for carcinogenicity due to lifetime exposure. Chemicals
included in the tables are the subjects of draft documents from various
tjroups within~EPA and may or may not be verified as inputs to the
Integrated Risk Information System (IRIS).
73
-------
TABLE 10. HEALTH EFFECTS SUMMARY TABLE A:
SUBCHRONIC AND CHRONIC TOXICITY VIA INHALATION
s«
Compound
Benzene
1-Butanol
Methyl ene
chloride
chronic
(RfD)
Methyl
ethyl
ketone
subchronic
(RfD.)
s
Chronic
(RfD)
o-Xylene
subchronic
(RfD )
s
Chronic
(RfD)
Exposure Species
Not listed
Not listed
695 mg/m3 Rat
6 h/day,
5 days/wk
for 2 yrs
o
693 mg/nr Rat
7 h/day,
5 days/wk
for 12 wks
693 mg/m3 Rat
7 h/day,
5 days/wk
for 12 wks
150 mg/m3 Rat
continuous
on days
7-14 of
gestation
4750 mg/m3 Rat
8 h/day,
7 days/wk
for 1 yr
Efects
of
concern
NA
CNS
CNS
Feto-
toxicity
Hepato-
megaly
Reference
dose*
mg/m
(mg/kg-day)
3 (NA)
(9E-l)a
3E-0
(9E-1)
3E-1
(9E-2)
3E+0
(1E+0)
7E-1
(2E-1)
Uncer-
- tainty
factor
100
100
1000
100
100
Calculated using a standard body weight of 70 kg, a standard ventilation rate
of 20 m3 /day, and the RfD in M9/m3.
76
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TABLE 11. HEALTH EFFECTS SUMMARY TABLE B: CARCINOGENICITY VIA INHALATION
Compound
Benzene
1-Butanol
Methyl ene
chloride
Methyl ethyl
ketone
o-Xylene
Inhalation
exposure
Occupational
Not listed
2000 or
4000ppm
Not listed
Not listed
Tumor Unit risk,
Species site (/ig/m3)"
Human Leukemia A/8.3E-6
Lung/
Mouse liver B2/4.7E-7
Slope
factor, 1
- (mg.kg-day)"1
- 2.9E-2
*
(1.65E-3)a
Calculated using conversion formula from HEAST and unit risk per /*g/m .
The RfD in HEAST is an estimate of the daily exposure of the human
population that is likely to be without an appreciable risk of deleterious
effect during a lifetime. In the case of subchronic RfD (RfDs), it is the
daily exposure during a portion of a lifetime that is without appreciable risk.
The chronic RfD is appropriate for exposures from seven to seventy years in
duration. The subchonic RfDs is applicable to exposures from two weeks to
seven years. In this example, both RfD and unit risk factors were prorated
to the actual exposure of the MEI over the period of remediation.
The unit risk factors in Table 11 are each preceded by a letter code, A,
B, C, or D. These codes indicate the strength of evidence for
carcinogenicity, as follows:
Group A - Human carcinogen
Group B - Probable human carcinogen
Group C - Possible human carcinogen
Group D - Not classifiable
Group E - Evidence of noncarcinogenicity
HEAST suggests that quantitative risk assessments be conducted only for
chemicals in groups A-and B. The unit risk factor for benzene is coded "A"
77
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
and the unit risk factor for methylene chloride is coded "B". Both these
chemicals are addressed in this reverse risk estimate.
In general, toxicity is the principal concern in short-term exposure and
carcinogenicity is the principal concern in long-term exposure. . In this
example the time period of concern for both toxicity and carcinogenicity was
*Ke two-month period of remediation; therefore, there was no distinction
between long-term and short-term exposures.
Note that benzene is not listed on the toxicity table, but is listed on the
carcinogenicity table. For this reason, the level of protection for benzene
was based on carcinogenicity. MEK and o-xylene are listed on the toxicity
table, but not on the carcinogenicity table; therefore, the health protection
level was based on toxicity. MC is listed on both tables; therefore the health
protection level was based on either systemic toxicity or carcinogenicity,
whichever was the more stringent level of protection. 1-butanol does not
occur on either table; therefore, it was not considered for toxicity or
carcinogenicity based protection levels.
In this example, subchronic reference doses from HEAST were considered
as public protection levels when adjusted to the time of exposure of the MEL
An equation from EPA risk assessment guidance , page 6-44, applies to
residential exposure of airborne organic vapors via inhalation, as follows:
CA = [(IN)(BW)(AT)]/[(IR)(ET)(EF)(ED)] (1)
where CA = concentration in air (mg/m )
IN = intake (mg/kg-day)
BW = body weight (kg)
AT = averaging time (davs)
IR = inhalation rate (m/h)
ET = exposure time (h/day)
EF = exposure frequency (days/yr)
ED = exposure duration (yrs)
For each chemical (x), the exposure equation was solved for the MEI as
fol1ows:
CAX - [(INX mg/kg-day)(70 kg)(60 days)]/
[(0.8 m3/h)(8 h/day)(43 days/yr)(l/6 yr)]
The value of IN for each chemical was taken from the HEAST table for
RfD (Table 10). Other factors were taken from the Exposure Factors
78
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Handbook32' An inhalation rate of 0.8 m3/h was used for an adult with an
average ventilation rate of 20 m3 per day. Exposure time was eight hours
per day; this assumed that the MEI was located outdoors at the same time
that remediation was in progress. The exposure frequency was 4a days per
year, based on a 60-day remediation period comprised of 43 weekdays (when
-remediation takes place) and 17 weekends (no excavation activity).. The
exposure duration was 1/6 years (the two-month period of remediation). The
body weight of an average adult is approximately 70 kg, a standard weight
used in many risk calculations. The averaging time was 60 days, the planned
period of remediation. Using these data, the resulting health protection
levels for subchronic exposures to toxic chemicals were:
AC(MEK) - 82.4 mg/m3 (28 ppm)
AC(o-xylene) = 91.6 mg/m (21 ppm)
Next, protection levels were estimated for carcinogenic chemicals based
on unit risk factors. The National Contingency Plan allows a maximum
individual lifetime risk (MILR) in the range 10"4 to 10"6 for Superfund sites,
The point of departure is 10"6. In this example, a simple and conservative
guideline was chosen: 10 risk for each individual chemical. Because there
were only two carcinogens, the risk for all chemicals combined (assuming
additive effects) was 2xlO"6.
The following formula was used for calculating the air concentration
corresponding to a given upper-bound increased lifetime risk:
CA = [(MILR)(BW)(AT)(CON)]/
[(IR)(ET)(EF)(ED)(SLI)] (2)
where CA = concentration in air Ug/m )
MILR = maximum individual lifetime risk (dimensionless)
BW = body weight (kg)
AT - averaging time (days)
CON = conversion factor i^g/mg)
IR = inhalation rate (m /h)
ET - exposure time (h/day)
EF = exposure frequency (days/yr)
ED = exposure duration (yrs) ,
SLI «= slope factor for inhalation (mg/kg-days)
For each chemical (x), the risk equation was solved for the MEI as
follows:
79
-------
CAX - [(10'6)(70 kg)(25,550 days)(103 /*g/mg)]/
{(0.8 m3)(8 h/day)(43 days/yr)(l/6 yr)[SLIx (mg/kg-days)"1]}
The value for each SLIX was taken from the respective slope factor in
Table 11. The value of 10"6 for MILR was previously selected for this
example. The number of days (25,550) represents the number of days in an
entire lifetime. Other parameter values are the same as used in equation 1.
Using these data, the carcinogens in this example have the following
protection levels:
CA(Benzene) - 1345 /*g/m3 (0.4 ppm)
AC(MC) - 23,600 Mg/m3 (6.7 ppm)
The EPA toxicologist was not entirely satisfied with the calculated
protection levels because they only provided average air concentrations over a
two-month period. The EPA procedures did not address an emission event
that could occur during a single working day. For this reason, the EPA
toxicologist requested a review of this site by ATSDR, with special attention
to be given to the issue of a short-term health protection level. An ATSDR
toxicolgist reviewed the data concerning this site provided to him by EPA.
In addressing the short-term protection level, the ATSDR toxicologist
used data from the NIOSH Pocket Guide to Chemical Hazards. These data are
reproduced in Table 12 for the chemicals present at the example site. From
this table he selected the most conservative recommended standard for each
chemical. Because these values were for a healthy adult male, he then
divided these values by 10 to allow for a sensitive individual. Using this
procedure, the following 8-h average protection levels were selected:
CA(benzene) = 0.01 ppm
CA(MC) - 10 ppm
CA(MEK) = 20 ppm
CA(o-xylene) * 10 ppm
The following health protection levels for less than one-hour average
exposures were taken from Table 12 and adjusted in the same manner:
CA(benzene) = 0.1 ppm, 15 min
~r CA(MC) - 100 ppm, 5 min
CA(MEK) - --None
CA(o-xylene) = 20 ppm, 10 min
80
-------
TABLE 12. SUMMARY OF RECOMMENDED EXPOSURE LIMITS
1
Chemical name
Benzene
Methyl ene
chloride
MEK (2-butanone)
o-Xylene
1-Butanol
PEL,
ppm
10
500
200
200
Not
listed
OSHAa
PEL ceil-
ling, ppm
50 (10 min)
1000, 2000
(5 min/2 h
peak)
NA
NA
Not listed
NIOSHb
REL,
ppm
0.1 (8 h TWA)
Lowest feas-
ible limit
200 (10 h TWA)
100 (10 h TWA)
Not listed
ACGIHC
REL ceil-
ing, ppm
1.0 (15 min)
NA
NA
200 (10 min)
Not listed
TLV,
ppm
10
100
NA
NA
Not
listed
TLV ceil-
ing, ppm
NA
NA
NA
NA
Not
listed
Occupation Health and Safety Administration (OSHA) Permissable Exposure Limits (PEL) published in CFR
1910, Subpart Z. as 8-hour time-weighted averages, unless otherwise noted. Ceiling values are not to be
exceeded at any time.
National Institute of Occupational Health and Safety (NIOSH) recommended exposure limit (REL), listed
when it is less than the corresponding PEL.
cAmerican Conference of Governmental Industrial Hygienists (ACGIH) threshold limit value (TLV), listed
when it is less than the corresponding PEL.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 13 summarizes the public protection levels based on subchronic
toxicity levels, carcinogenicity, and ATSDR recommendations. Note that the
theoretical ATSDR-recommended protection levels appear to be more conservative
than calculated concentrations based on subchronic toxicity or
carcinogenicity. In addition, they used shorter averaging times of less than a
full day. Because the ATSDR levels are less than one-day averages", they could
be used to control day-to-day remedial activities at the site. Note that if
these levels were always met, the longer-term protection levels would always
be met.
3.4.3 Dilution Ratios
Two different models were employed to determine three different dilution
ratios for three different averaging times. The first model, SCREEN, was used
for determining short-term (one-hour and eight-hour) concentrations. This
model estimated one-hour average air concentrations at various distances from
29
the source . These one-hour estimates were used to approximate 10 or 15
minute peak concentrations because (1) there was no standard procedure for
estimating concentrations less than one-hour average concentrations, and (2)
the health protection levels for averaging times of 15 minutes or less were
already set one order of magnitude less than the lowest recommended level.
TABLE 13. PUBLIC PROTECTION LEVELS FOR THE EXAMPLE SITE,
IN ^g/mj AND ppb
Rick assessment,
two-month average
Chemical
Benzene
Methyl ene
chloride
MEK
o-Xylene
Molecular
weight
78
85
72
106
Subchronic
toxicity
None
None
82,400 (28)
91,600 (21)
Carcino-
genicity,
1,300 (0.4)
23,600 (6.7)
None
None
ATSDR
recommendatons
8-hour
30 (0.01)
35,000 (10)
59,000 (20)
43,000 (10)
< 15
minutes
300 (0.1)
350,000 (100)
None
87,000 (20)
82
-------
Another conservative estimate, e.g., estimating a 15-minute average higher
than a one-hour average, did not appear to be in order.
Unlike most dispersion models, the SCREEN model does not use
meteorological data, but gives a series of estimates under different possible
meteorological conditions. The worst meteorological case was used^to show the
-worst impact that could occur at this example site. " -
To determine a dilution ratio for eight-hour and less average protection
levels, a nominal emission rate of 1 gm/sec was used in an extremely simple
emission scenario. In this simplified scenario, all emissions were assumed to
be derived from excavation operations. (A more complicated and more accurate
emissions calculation was carried out in a later step.) Excavations were
planned to be carried out at different locations across the contaminated zone
from day to day, but the worst case excavation would occur along the outer
boundary of the contaminated zone on the same day that worst-case meteorology
took place. The distance between the outer edge of the contaminated area and
the closest fenceline (54.9 m), was used for this worst case estimate.
Accurately estimated emission rates from the site were not needed at this
point, because only a normalized emission scenario (1.0 gm/s) was required to
produce the necessary dilution ratio.
A worst-case meteorological day consisted of 24 hours at stability class
D, including the eight hours when remediation took place. For stablity class
D at the closest distance (54.9 m) from source to MEI, and with a nominal
emission rate of 1 g/s, the SCREEN model gave a predicted one-hour average
concentration of 3031 pq/m . An estimated concentration of 3031 /ig/m divided
by a nominal emission of 1.0 gm/s yielded a dilution ratio of 3031 »g/m per
gm/s of emissions. This dilution ratio was used with public protection levels
of less than one hour.
OQ
According to the SCREEN manual , a ratio of 0.7 may be multiplied times
the one-hour average to approximate the average concentration over a
eight-hour period. Thus, for an estimated 3031 /^g/m one-hour concentration,
3
the corresponding eight-hour concentration is 2122 »g/m A concentration of
2122 A^g/m divided by a nominal emission of 1.0 gm/s yielded a dilution ratio
of 2122 A»g/m per gm/s of emissions. This dilution ratio was used with
eight-hour average public protection levels.
83
-------
The second model, ISCLT, was used to estimate long-term average
concentrations. This model estimated one-year average concentrations based on
one year of hourly meteorological data. In this example, meteorological data
from a nearby meteorological station was used. A nominal 1 g/s "emission rate
was used to simulate the same simplified emissions scenario used i-n the
'eight-hour case. A series of receptors were defined in the model, -
corresponding to the points where radial lines from the center of the site
crossed the site fenceline. These radial lines were 22.5 degrees apart, the
standard radii used in the ISCLT model. The center of the site was considered
to be the location of emissions because it is the average location of
excavation activities over the remediation period.
One year of meteorological values is usually used with ISCLT to obtain
one-year average concentrations at receptors. In this example, remediation
was not planned to take place during the whole year, but only during June and
July (summertime). Thus the objective was to estimate average concentrations
over a particular two-month period. To obtain these estimates, a special
meteorlogical year was constructed to consist of 12 months of summer data
only. Because the input meteorological data were for summer months only, the
one-year averages produced by the model were considered reasonable estimates
of a two-month average concentration for the June and July period of
remediation.
Note that unlike SCREEN, worst case meteorology canot be defined in
ISCLT. The worst case is simply the location of the receptor providing the
highest average concentration In this example, the highest concentration was
145 ng/m and it occurred at the 0 degree radial line (due north) at a
distance of 100.6 meters from the center of the site. An estimated
concentration of 145 /*g/m divided by a nominal emission of 1.0 gm/s yielded a
o
dilution ratio of 145 pg/m per gm/s of emissions. This dilution ratio was
used with public protection levels based on long-term (two-month) carcinogenic
risks or long term chronic exposure levels.
3.4.4 Allowable Emissions
Using the dilution ratios calculated for short-term (15-minute and
eight-hour) and long-term (two-month) protection levels, the maximum
15-minute, eight-hour, and two-month average emission rates were calculated
for each chemical. These calculations were simply the following:
84
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
where
:ha
'ha
emission rate allowed for health concern h and
averaging time a, in g/s
public protection level for health concern h and ..
averaging time a, in
per
D, - dilution ratio for averaging time a, in /*g/m
a 9/s
A summary of allowable emissions, as calculated, is shown in Table 14.
Note that the chemical that is most stringently controlled was benzene. The
most stringent protection level for benzene was a recommendation of National
Institute of Occupational Safety and Health (NIOSH) which was recommended to
the RPM by the ATSDR toxicologist. This stringent protection level then was
translated into a stringent allowable emission rate. This was the key
protection level, which determined what contingency plan was required.
TABLE 14. ALLOWABLE EMISSIONS AT THE EXAMPLE SITE IN g/s
Risk assessment,
two-month period
ATSDR
recommendations
Chemical
Subchronic
toxicity
Carcino-
genicity
8-hour
< 15
minutes
Benzene
Methylene
chloride
MEK
o-Xylene
None
568
568
632
9
163
None
None
0.014
17
28
21
0.010
115
None
29
85
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
3.4.5 Estimated Emissions Versus Allowable Emissions
A more detailed analysis of emissons showed that the example site
consisted of a number of emission sources, not just one. Emissions were
generated from digging with a backhoe, dumping dirt into a truck, and
driving the truck down the haul road. One of the major emissions xame from
the moving truck. The objective was to estimate emissions from each of these
actvities as accurately as possible and to sum these emissions for the site.
Emissions were estimated by a complex procedure that is only outlined
here, but is detailed in a separate report . In these emission calculations,
area sources are represented as squares. If a site is rectangular or
irregular in outline, it is represented in the model as a series of squares
having the same area (in square meters) as the actual site. Thus, for any
one remediation day the area of excavation was represented as a 4x4 meter
square area. The area of dumping dirt into the truck was represented as
another 4x4 meter square area. The truck passing down the haul road was
represented as a series of ten squares, each of which was 22 meters on a
side.
In this example, the rate of emissions of VOCs depended not only on
their concentration in the soil, but also on characteristics of the soil and
characteristics of the chemicals, especially volatility. Because MC is
quite volatile and was present in the highest concentration in the soil, it
was found in highest concentration in air emissions. For the overall
remediation, MC comprised 81.9 percent of VOC emissions, and benzene
comprised 4.9 percent. The other three compounds comprised the remaining
13.2 percent. The relative contribution of the five compounds did not
change appreciably during each step in the remediation.
A summary of the estimated site emissions for each chemical species is
shown in Table 15. These estimates are based on an overall VOC emission of
0.299 g/s for an average remediation day during the two-month remediation
period. Table 16 lists the highest estimate of the average emission for each
chemical over the two-month period (from Table 15) and the most stringent
public protection level over the two-month remediation period (from Table 14).
-Table 16 then-lists the estimated eight-hour average emissions and the
eight-hour allowable emission (based on the recommended eight-hour
86
-------
TABLE 15. SUMMARY OF AVERAGE AIR EMISSIONS DURING REMEDIATION
00
-vj
Soil cap
Chemical ,
Benzene
1-Butanol
i
Methyl ene chloride
Methyl ethyl ketone
o-Xylene
Emission
fraction
0.1737
0.0049
0.5815
0.0218
0.2181
Emission,
g/s
0.052
0.001
0.174
0.007
0.065
Contaminated soil
Emission
fraction
0.0407
0.0052
0.8374
0.0911
0.0256
Emission,
g/s
0.012
0.002
0.250
0.027
0.008
All excavations
Emission
fraction
0.0495
0.0052
0.8191
0.0857
0.0404
Emission,
g/s
0.015
0.002
0.245
0.026
0.012
TABLE 16.
COMPARISON OF DAILY AVERAGE
EMISSIONS WITH ALLOWABLE
EMISSIONS IN
g/s
Chemical
Benzene
Methyl ene chloride
Methyl ethyl ketone
o-Xylene
Two-month
Estimated
0.052
0.250
0.027
0.065
average
Allowable
risk
9
163
None
None
Eight-hour
Estimated
0.156
0.750
0.081
0.195
average
Allowable
risk
0.014
17
28
21
< 15
minutes
A1 1 owabl e
Estimated3 risk
(0.156)
(0.750)
(0.081)
(0.195)
0.35
6.17
0.87 ,
i. '
2,030 '
Estimated 15-minute emissions are the same as 8-hour emissions because of constraints of methodology.
-------
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
protection level). It also lists the estimated 15-minute estimated emissions
and allowable emissions. The 15-minute emissions are the same as eight-hour
emissions because the estimation methodology did not allow for a time
discrimination of less than eight hours. According to the data_in Table 16,
each emission is well below its respective public protection level, except in
;the case of the eight-hour benzene emission. Based on these data, a
contingency plan for this remediation would be needed and control measures
carried out during the remediation process.
While an analysis based on the available data showed that remediation
could proceed (with control measures), the site manager questioned whether
the available information was suffient. Soil sampling showed no hot spots,
but he questioned whether the number of samples was sufficient to discover a
hot spot in a unexpected location. In this example, the site manager opted to
analyze a worst-case scenario in which an unknown hot spot might exist.
Such a hot spot could be caused by someone dumping benzene and other
chemicals into a temporary pit and then covering it so that no evidence of
this pit was visible on the surface of the ground.
Benzene has been found in concentrations above 30,000 ppm in soil.
Using the same emission estimation methodology used for the average
remediation day, emissions due to excavation, dumping, and hauling of soil
with 30,000 ppm benzene were estimated. According to this procedure, this
worst-case scenario would result in an air emission of 0.186 g/s of benzene
averaged over a eight-hour period.
The worst case excavation due to an unknown hot spot would produce
benzene emissions in considerable excess of the allowable emisssion rates for
public protection. The site manager judged that this possible, even if it was
not a high probability, which further justified a contingency plan using air
monitoring to protect offsite populations. Because the surrounding community
was very sensitive to operations at the site, the contingency plan was
intended to provide a margin of safety and reassurance to the community.
The site manager proceeded to prepare a simple plan to address daily
operations and the worst-case contingency.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
3.4.6 Contingency Plan Alert Levels
One principal in setting contingency plan alert levels is that the
averaging time of an alert level must match the averaging time used in
real-time monitoring. In this example, estimated ambient concentrations that
corresponded to a 10 risk of cancer had been adopted as the level of
protection for the public over the two-month period of remediation*. An
eight-hour recommended standard was also adopted to control daily
remediation operations. This meant that when an eight-hour instrument
reading at the fenceline (near the MEI) implied an excursion of the allowable
emission rate, the remedial process was to be stopped or slowed down until
the instrument reading dropped below the alert level. Thus, while there
could be one-hour excursions above the alert level, it is unlikely that a
eight-hour protection level would be exceeded or that a 10 risk level would
be exceeded over a two-month period.
The key public protection level is 30 ug/m (0.01 ppm) benzene, based
on recommendations from NIOSH. This is below the 2.0 ppm level of detection
for a total organics monitor, the instrument frequently used for real-time
detection of ambient VOCs in the field. As a matter of practicality, the alert
level must be set to 2 ppm of VOC at the fenceline. An argument could be
made that a lower alert level would be justified in case the VOC were
comprised solely of benzene, however, the chosen alert level is the lowest
level that can be implemented due to the limits of technology.
One approach is to back-calculate from the emissions rate and dilution
ratios to determine the exact downwind distance from the source that will
register 2 ppm on a monitor when the fenceline concentration is 0.01 ppm.
The monitor can then be placed at this location rather than at the fenceline.
A procedure for doing this has not yet been specified.
Note that the standrd 10.2 eV photoionization detector (PID) lamp on
commonly used instruments will not detect methylene chloride, a compound
that is likely to be emitted from this example site. Either and 11.7 eV PID or
a flame ionization detector (FID) could be used to solve this problem.
Emission control measures at the site consisted of spraying foam onto the
.soil being excavated and using foam and a tight-fitting tarpaulin over the soil
being carried down the haul road. These measures were estimated to reduced
emissions well below the eight-hour allowable emissions.
89
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
The contingency plan to protect the public consisted of the following
rules:
1. A wind direction instrument is to be installed just outside the hot
zone.
2. At the beginning of each day of remediation, a portable "HNU will be
^' • placed at the fence!ine downwind from the area to be excavated that
day. This procedure was an added protective measure, to ascertain
the highest impact regardless of whether it actually affected the
MEI.
3. At two-hour intervals, wind direction will be checked and the
location of the HNU adjusted as necessary to be downwind of
excavation.
4. At hourly intervals, the HNU will be checked for the latest
one-hour average reading.
5. If an exceedance is detected, it will be reported to the site manager
for action under item 6 below.
6. When a reading above 2 ppm occurs, excavation of soil and truck
hauling must stop. Continuous HNU readings will be taken at the
fenceline until three continuous readings at 10-minute intervals
indicate ambient concentrations below the alert level. At this point,
excavation may proceed at a rate of one-half the previous rate of
excavation. If the alert level is not exceeded during a 30-minute
period, the original rate of excavation may be resumed.
90
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
SECTION 4
REFERENCES
1. Roe, David. What Kind of Data Does the Public Need: A Forum. EPA
Journal, Volume 5, Number 3, May/June 1989.
2. U.S. Environmental Protection Agency Region VII. Castlewood
Site-Specific Air Monitoring Plan. March 17, 1987.
3. Hudson, Jody L. Castlewood Site 2,3,7,8-TCDD Air Monitoring Summary
Report. 1987.
4. Kahn, Peter R. Ambient Air Sampling Results During a CERCLA Removal
Action, Chesnutis Superfund Removal Site, Beacon Falls, CT. Memorandum
to Dean Tagliaferro, July 27, 1989.
5. Aungst, Nancy, Ecology and Environment, Inc. Background Information on
Hyde Park Landfill. (FAX) February 22, 1990.
6. Hyde Park, R02-86/038, Abstract of the Record of Decision, November 26,
1985. Listed from the EPA RODs data base, February 16, 1990.
7. ERT, Special Construction Activities - Part II Air Monitoring Plan for
Hyde Park Landfill, Niagara Falls, New York. Prepared for Occidental
Chemical Company, May 1988.
8. Work Plan for Buried Drum Removal, Maryland Sand, Gravel and Stone Site,
Elkton, Maryland. Prepared for Clean Sites Inc. by GSX Services Inc.,
July 7, 1989.
9. Ludzia, Peter J. Maryland Sand, Gravel, and Stone Excavation Criteria,
Memorandum to File, February 6, 1990.
10. McKin Site ROD. Memorandum from Superfund Implementation Group,
Department of Health and Human Services, ATSDR to John Figler, EPA
Region I, August 1, 1985.
11. Webster, David W. Pilot Study of Enclosed Thermal Soil Aeration for
Removal of Volatile Organic Contamination at the McKin Superfund Site,
Journal of the Air Pollution Control Association, Vol. 36, No. 10,
October 1986, pp. 1156-1163.
JL2. U.S. Environmental Protection Agency, Site Safety Plan Nyanza Vault
Site. (Revised) September 28, 1987.
91
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
13. DiSirio, Marilyn R. Agency for Toxic Substances and Disease Registry,
Memorandum to Frank W. Lilly, On-Scene Coordinator, U.S. Environmental
Protection Agency, Region I, Re: Nyanza Hazardous Waste Site Soil Data:
Vault, April 10, 1987.
14. U.S. Environmental Protection Agency, Region I Environmental Services
Division, Ambient and Health and Safety Air Sampling Plan." -Nyanza Vault
Site Ashland, Massachusetts. October 1987.
15. Kahn, Peter R. Results of Ambient and Health and Safety Air Sampling
Study Nyanza Vault Site Ashland, Massachusetts, U.S. Environmental
Protection Agency, Region I, Environmental Services Division.
16. Kahn, Peter R. Results of Ambient and Health and Safety Air Sampling
Study Second Round Nyanza Vault Site Ashland, Massachusetts, U.S.
Environmental Protection Agency, Region I, Environmental Services
Division.
17. Firless, B. J., D. I. Bates, J. Hudson, R. D. Kleopfer, T. T. Holloway,
D. A. Morey, and T. Babb. Procedures used to measure the Amount of
2,3,7,8-TCDD in the Ambient Air Near a Superfund Site Cleanup Operation.
Environmental Science and Technology, Vol. 21, p. 550, June 1987.
18. Ambient Air Monitoring Program During Remedial Action, Vertac Chemical
Corporation Plant Site, Jacksonville, Arkansas. Prepared for IT
Corporation by EMI Consultants, March 19, 1986.
19. Site Assessment Report for Weatherford Residence. Prepared for EPA
Region VI Emergency Response Branch by Ecology and Environment, Inc.,
June 20, 1989.
20. Health Consultation: Mr. Robert Weatherford Residence, Jacksonville,
AR. Memorandum from Senior Public Health Advisor - ATSDR/EPA-6 to Mr.
David Gray, OSC, Emergency Response Branch, EPA Region VI, April 18,
1989.
21. Quality Assurance Project Plan for Air Monitoring at Weatherford
Residence and Jacksonville Crane Site. Prepared for U.S. Environmental
Protection Agency, Region VI, by Ecology and Environment, Inc., April
19, 1989.
22. 29 Code of Federal Regulations Chapter 17, Section 1910.120, Hazardous
Waste Operations and Emergency Response, July 1, 1989.
23. Protecting Health and Safety at Hazardous Waste Sites: An Overview.
EPA 625/9-85-006, U.S. Environmental Protection Agency, September 1985.
24. Field Standard Operating Procedures (FSOP) #9 Site Safety Plan. U.S.
Environmental Protection Agency Office of Emergency and Remedial
Response, Washington, D.C., April 1985.
25. A Compendium of Superfund Field Operations Methods. EPA/540/P-87/001,
Office of Emergency and Remedial Response, U.S. Environmental Protection
Agency, Washington, D.C., December, 1987.
92
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
26. Procedures for Conducting Air Pathway Analyses for Superfund
Applications, Volumes I through IV EPA-450/1-89-001 through
EPA-450/1-89-004, July 1989.
27. Health Assessment GUIdance--Information transmitted by Dr. Mike Allred,
ATSDR, to Roy Paul, PEI Associates, Inc., February 1990.
28. U.S. Environmental Protection Agency, Health Effects Assessment Sumamry
Tables Third Quarter 1989. OERR 9200.6-303(89-3), July 1989;
29. Brode, Roger W. Screening Procedures for Estimating the Air Quality
Impact of Stationary Sources. EPA 450/4-88-010 Technical Support
Division, OAQPS, U.S. Environmental Protection Agency. August 1988.
30. U.S. Environmental Protection Agency, Risk Assessment Guidance for
Superfund Volume I Human Health Evaluation Mnual (Part A).
EPA/540/1-89/002, Office of Emergency and Remedial Response, Washington,
D.C., December 1989.
31. U.S. Department of Human and Health Services. NIOSH Pocket Guide to
Chemical Hazards. National Institute of Occupational Health and Safety.
September 1985.
32. U.S. Environmental Protection Agency. Exposure Factors Handbook.
EPA/600/8-89/043. Office of Health and Environmentla Assessment. July
1989.
33. U.S. Department of Health and Human Services. NIOSH Pocket Guide to
Chemical Hazards. NIOSH Pub. No. 85-114. National Institute for
Occupational Safety and Health. September 1985.
34. PEI Associates, Inc. Development of Example Procedures for Evaluating
the Air Impacts of Soil Excavation Associated with Superfund Remedial
Actions. Prepared under Contract No. 68-02-4394, Work Assignment 38.
July 1990.
93
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
|(*) The use of any trade names does not imply their
endorsement by the U.S. Environmental Protection Agency.
I
APPENDIX A
CHARACTERISTICS OF THE HNU PHOTOIONIZER
AND
ORGANIC VAPOR ANALYZER
I. INTRODUCTION
The HNU Photoionizer* and the Foxboro Organic Vapor Analyzer*
(OVA) are used in the field to detect a variety of compounds
in air. The two instruments differ in their modes of
operation and in the number and types of compounds they detect
(Table 1-1). Both instruments can be used to detect leaks of
volatile substances from drums and tanks, determine the
presence of volatile compounds in soil and water, make ambient
air surveys, and collect continuous air monitoring data. If
personnel are thoroughly trained to operate the instruments
and to interpret the data, these instruments can be valuable
tools for helping to decide the levels of protection to be
worn, assist in determining other safety procedures, and
determine subsequent monitoring or sampling locations.
II. OVA
The OVA operates in two different modes. In the survey
mode, it can determine approximate total'concentration of all
detectable species in air. With the gas chromatograph (GC)
option, individual components can be detected and measured
independently, with some detection limits as low as a few
parts per million (ppm).
In the GC mode, a small sample of ambient air is injected into
a chromatographic column and carried through the column by a
stream of hydrogen gas. Contaminants with different chemical
structures are retained on the column for different lengths
of time (known as retention times) and hence are detected
separately by the flame ionization detector. A strip chart
recorder can be used to record the retention times, which are
^ then compared to the retention times of a standard with known
~C chemical constituents. The sample can either be injected into
the column from the air sampling hose or injected directly
with a gas-tight syringe.
I
94
-------
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
ACTION
Response
x-'" .
Application
Detector
Limitations
Calibration gas .
Ease of
operation
Detection limits
Response time
Maintenance
Useful range
Service life
TABLE 1-1
COMPARISON OF THE OVA AND
OVA
Responds to many organic gases
and vapors.
In survey mode, measures total
concentration of detectable
gases and vapors. In GC mode,
identifies and measures
specific compounds.
Flame ionization detector (FID)
Does not respond to inorganic
gases and vapors. Kit available
for temperature control.
Methane
Requires experience to inter-
pret correctly, especially
in GC mode.
0.1 ppm (methane)
2-3 seconds (survey mode)
for CH4
Periodically clean and inspect
particle filters, valve rings,
and burner chamber. Check
calibration and pumping
system for leaks. Recharge
batteries after each use.
0-1000 ppm
8 hours; 3 hours with strip
chart recorder.
95
HNU
HNU
Responds to many organic
and some inorganic gases
and vapors.
In survey mode, measures
total concentration of
detectable gases and
vapors .
Photoionization detector
(PID)
Does not respond to
methane. Does not detect
a compound if probe has a
lower energy than
compound's ionization
potential.
Isobutylene
Fairly easy to use and
interpret .
0.1 ppm (benzene)
3 seconds for 90% of
total concentration of
benzene .
Clean UV lamp frequently.
Check calibration
regularly. Recharge
batteries after each
use.
0-2000 ppm
10 hours; 5 hours with
strip chart recorder.
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
In the survey mode, the OVA is internally calibrated to
methane by the manufacturer. When the instrument is adjusted
to manufacturer's instructions it indicates the true
concentration of methane in air. In response to all other
detectable compounds, however, the instrument reading may be
higher or lower than the true concentration. * Relative
response ratios for substances other than methane are
available.
To correctly interpret the readout, it is necessary to either
make calibration charts relating the instrument readings to
the true concentration or to adjust the instrument so that it
reads correctly. This is done by turning the ten-turn gas-
select knob, which adjusts the response of the instrument.
The knob is normally set at 3.00 when calibrated to methane.
Calibration to another gas is done by measuring a known
concentration of a gas and adjusting the gas select knob until
the instrument reading equals that concentration.
The OVA has an inherent limitation in that it can detect only
organic molecules. Also, it should not be used at
temperatures lower than about 40 degrees Fahrenheit because
gases condense in the pump and column. It has no column
temperature control, (although temperature control kits are
available) and since retention times vary with ambient
temperatures for a given column, determinations of contam-
inants are difficult. Despite these limitations, the GC mode
can often provide tentative information on the identity of
contaminants in air without relying on costly, time-consuming
laboratory analysis.
III. HNU
The HNU portable photoionizer detects the concentration of
organic gases as well as a few inorganic gases. The basis
for detection is the ionization of gaseous species. Every
molecule has a characteristic ionization potential (I.P.)
which is the energy required to remove an electron from the
molecule, yielding a positively charged ion and the free
electron. The incoming gas molecules are subjected to
ultraviolet (UV) radiation, which is energetic enough to
ionize many gaseous compounds. Each molecule is transformed
into charged ion pairs, creating a current between two
electrodes.
Three probes, each containing a different UV light source,
are available for use with the HNU. Ionizing energies of the
probe are 9.5, 10.2, and 11.7 electron volts (eV). All three
detect many aromatic and large molecule hydrocarbons. The
96
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
10.2 eV and 11.7 eV probes, in addition, detect some smaller
organic molecules and some halogenated hydrocarbons. The 10.2
eV probe is the most useful for environmental response work,
as the lamp's service life is longer than the 11.-7 eV probe
and it detects more compounds than the 9.5 eV probe^
The HNU factory calibration gas is benzene. The span
potentiometer (calibration) knob is turned to 9.8 for benzene
calibration. A knob setting of zero increases the response
to benzene approximately tenfold. As with the OVA, the
instrument's response can be adjusted to give more accurate
readings for specific gases and eliminate the necessity for
calibration charts.
While the primary use of the HNU is as a quantitative
instrument, it can also be used to detect certain
contaminants, or at least to narrow the range of
possibilities. Noting instrument response to a contaminant
source with different probes can eliminate some contaminants
from consideration. For instance, a compound's ionization
potential may be such that the 9.5 eV probe produces no
response, but the 10.2 eV and 11.7 eV probes do elicit a
response. The HNU does not detect methane or inorganic
compounds.
The HNU is easier to use than the OVA. Its lower detection
limit is also in the low ppm range. The response time is
rapid; the meter needle reaches 90% of the indicated
concentration in 3 seconds for benzene. It can be zeroed in
a contaminated atmosphere and does not detect methane.
IV. GENERAL CONSIDERATIONS
Both of these instruments can monitor only certain vapors and
gases in air. Many nonvolatile liquids, toxic solids,
particulates, and other toxic gases and vapors cannot be
detected. Because the types of compounds that the HNU and
OVA can potentially detect are only a fraction of the
chemicals possibly present at an incident, a zero reading on
either instrument does not necessarily signify the absence of
air contaminants.
^ The instruments are non-specific, and their response to
different compounds is relative to the calibration setting.
Instrument rea.dings may be higher or lower than the true
concentration. This can be an especially serious problem when
monitoring for total contaminant concentrations if several
different compounds are being detected at once. In addition,
the response of these instruments is not linear over the
97
-------
I
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
entire detection range. Care must therefore be taken when
interpreting the data. All identifications should be reported
as tentative until they can be confirmed by more precise
analysis. Concentrations should be reported in terms of the
calibration gas and span potentiometer or gas-select-knob
setting.
Since the OVA and HNU are small, portable instruments, they
cannot be expected to yield results as accurate as laboratory
instruments. They were originally designed for specific
industrial applications. They are relatively easy to use and
interpret when detecting total concentrations of individually
known contaminants in air, but interpretation becomes
extremely difficult when trying to quantify the components of
a mixture. Neither instrument can be used as an indicator for
combustible gases or oxygen deficiency.
The OVA (Model 128) is certified by Factory Mutual to be used
in Class I, Division 1, Groups A,B,C, and D environments. The
HNU is certified by Factory Mutual for use in Class I,
Division 2, Groups, A, B, C, and D.
98
-------
TECHNICAL REPORT DATA
(Pleat nod Instructions on the reverse before completing)
1. REPORT NO.
EPA-450/1-90-005
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Contingency Plans at Superfund Sites Using
Air Monitoring
5. REPORT DATE
September 1990
6. PERFORMING ORGANIZATION CODE
7. AUTHOH(S)
• Roy Paul
8. PERFORMING ORGANIZATION REPORT NO
DCN 90-203-080-61-02
10. PROGRAM ELEMENT NO.
61
9. PERFORMING ORGANIZATION NAME AND ADDRESS
PEI Associates, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
11. CONTRACT/GRANT NO.
68-02-4394
12. SPONSORING AGENCY NAME AND ADDRESS
U.S. Environmental Protection Agency
Region VIII
999 18th Street, Suite 500, One Denver Place
Denver, Colorado 80202
13. TYPE OF REPORT AND PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
16. SUPPLEMENTARY NOTES
6. ABSTRACT
Contingency planning, as defined in this document, encompasses the air
program established to protect offsite populations. Monitors for this
purpose are usually located at the site perimeter or within the community.
Monitors located within the site for the safety and protection of workers are
not included in this definition, unless onsite monitors serve the dual purpose
of protecting both the workers and offsite population.
A contingency plan using air monitoring establishes alert levels in
advance of actually collecting monitoring data. Alert levels address the
offsite population exposure concentrations that trigger an emergency response
or a change in remedial activities. These alert levels are in addition to
alert levels for onsite personnel.
The purpose of this document is to: 1) illustrate contingency air
monitoring with examples from past projects, and 2) describe how a contingency
air monitoring program may be established. This document is illustrative in
nature because the application of this type of monitoring is not consistently
prescribed in rules and regulations, but is based on professional judgment
applied in an analysis of individual sites and particular circumstances.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Superfund
Monitoring
Contingency Plans
Contingency Monitoring
8. DISTRIBUTION STATEMENT
IB. SECURITY CLASS (This Report)
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
22. PRICE
I
EPA Far* 2220-1 (R»». 4-77) PREVIOUS COITION is OBSOLETE
-------
^ Protean fcency
u>s tnvucr^ _UJ) r
^lon ?'i?v[son Boulevard, 12W
77 West {fKs6°0n604-3590
Chicago, \L
------- |