STAFF MEMORANDUM
GROUNDWATER PROTECTION STANDARDS FOR HAZARDOUS WASTE LAND DISPOSAL
FACILITIES: WILL THEY PREVENT MORE SUPERFUND SITES?
Industry, Technology, and Employment Program
Office of Technology Assessment
United States Congress
April .6, 1984
Staff Memoranda are neither reviewed nor approved by the
Technology Assessment Board.
-------
GROUNDWATER PROTECTION STANDARDS FOR HAZARDOUS WASTE LAND DISPOSAL
-FACILITIES: WILL THEY PREVENT MORE SUPERFUND SITES?*
Table of Contents
SUMMARY 2
INTRODUCTION AND BACKGROUND 13
INDUSTRIAL SOURCES OF CERCLA SITES 24
INTERIM STATUS 26
LIMITATIONS OF COVERAGE 32
GROUNDWATER MONITORING WELLS 34
CONTAMINANT TOLERANCE LEVELS 41
MONITORING IN THE VADOSE ZONE 53
DELAYS IN STARTING CORRECTIVE ACTION 63
STATISTICAL ANALYSIS 67
COMPLIANCE MONITORING 72
CORRECTIVE ACTION 75
FINANCIAL RESPONSIBILITY ' 81
REFERENCES 82
*
This staff memorandum is part of an ongoing assessment, Cleanup of
Uncontrolled Hazardous Waste Sites Under Superfund. that is being conducted by
OTA's Industry, Technology and Employment Program. The full assessment, which
will focus on technical problems and issues of the Superfund program to clean
up uncontrolled hazardous waste sites will be delivered to the Congress later
this year. Another assessment, Technologies to Measure, Monitor and Mitigate
Groundwater Contamination, is being conducted by OTA's Ocean and Environment
Program. That assessment, to be released shortly, will provide a
comprehensive technical framework to assist Congress in understanding the
major groundwater contamination issues facing the Nation.
-------
GROUNDWATER PROTECTION STANDARDS FOR HAZARDOUS WASTE LAND DISPOSAL
FACILITIES: WILL THEY PREVENT MORE SUPERFUND SITES?
One of the principal reasons for the passage of the Resource Conservation
and Recovery Act (RCRA) in 1976 was for the regulation of future disposal of
hazardous waste. It. then became evident that additional legislation was
needed to deal with the burgeoning number of uncontrolled sites which resulted
from past practices. In 1980, therefore, Congress passed the Comprehensive
Environmental Response, Compensation and Liability Act (CERCLA), also known as
Superfund. There has been a general impression and hope that these two laws
would eventually provide effective protection of public health and the
environment from hazardous wastes: CERCLA by cleaning up past problems and
RCRA by preventing future ones.
This analysis concludes that, where groundwater is at risk, RCRA
groundwater protection standards are not likely to prevent land disposal sites
from becoming uncontrolled sites that will require cleanup under Superfund.
The problems with the RCRA groundwater protection standards are so numerous
and serious that the standards cannot compensate for what has been found to be
ineffective and unproven* land disposal technology. Although OTA has not
focused on the details of the RCRA statute in this analysis, there does not
appear to be a major statutory problem.
The limitations of the RCRA groundwater protection standards, coupled
with those of land disposal technology are likely to cause serious problems
for future generations. Concern for the future indicates that land disposal
be limited to inert low hazard wastes, such as the stabilized residues from
waste treatment operations, and to facilities where groundwater is not
threatened. Otherwise, Superfund is likely to face a continuing stream of
-1-
-------
substantial burdens in the decades ahead from land disposal facilities : ;
sanctioned by the regulatory structure, but whose operators may not bear
cleanup costs. There remains, moreover, a threat from the billions of tons of
hazardous waste which have been disposed for many decades at what are now the
interim status facilities under RCRA.
SUMMARY
General conclusions
o RCRA groundwater monitoring and protection standards issued by EPA were
not designed to prevent RCRA regulated sites from becoming CERCLA sites
and they are not capable of doing so.
o Many of the RCRA regulations may seem reasonable on their surface;
however, detailed technical analysis reveals serious inadequacies,
especially associated with providing for effective, long-term
management of hazardous wastes.
o Many important decisions in the RCRA regulations were apparently made
without consideration of alternative approaches and without
cost/benefit analysis or risk analysis of alternative approaches. Had
•
such analysis been performed, alternative, approaches for groundwater
protection which cost less over the long term and present fewer risks
to public health and the environment probably would have been
identified. '
o There appears to be almost no consideration 'given, in the RCRA
regulations, to the huge cost of cleaning up groundwater contamination.
Regulatory decisions have had the effect of keeping down the short
range costs of the regulated community.
o The regulations take an optimistic view of the availability of
-2-
-------
technologies to detect and clean up contamination but a pessimistic
view of technologies which can prevent contamination, even when they
are the same technology.
Summary of Specific Conclusions
t
o Interim Status Facilities; Groundwater protection standards for these
facilities are less stringent than for new facilities, and most of them
already are, or are likely to become leaking sites; however, there are no
corrective action requirements.
o Fixing Leaks; With confirmed groundwater contamination there are no
requirements that a facility be closed until the leak is found and corrected,
nor to even find or stop the leak.
o RCRA Coverage Stops at the Fenceline; Although contamination may
spread beyond the legal boundary of a facility and have to be cleaned up under
CERCLA, under RCRA the operator only has to • clean up within the facility
limits.
o RCRA .Coverage Limited to 30 Years: New facilities must be designed
not to leak for 30 years after closure during which time the operator must
•maintain the facility, but later when leaks are more likely CERCLA becomes
responsible.
o Financial Responsibility; There are no RCRA requirements for
financial assurance for corrective action on a leaking site.
o Contaminants Which Are Regulated; Because CERCLA regulates more
substances than RCRA, and detection levels for other substances are set lower
-------
by CERCLA than by RCRA standards, a permitted but leaking RCRA facility can
become an uncontrolled site under CERCLA.
o Tolerance Levels of Contaminants; Acceptable levels of groundwater
contaminants are not based on health effects, and using detection limits of
•analytical techniques may not be protective of human health.
o Geological Standards; There are difficulties in predicting
groundwater movement or the rapid movement of contamination in some geological
environments which makes early detection and correction uncertain at some
sites. However, RCRA has no facility siting standards to restrict hazardous
waste sites to geologically suitable locations.
o Groundwater Monitoring; Technical complexity and site specificity
make it difficult for government rules to set the conditions for effective
groundwater monitoring.
o Monitoring in the Vadose Zone; Although the technology exists, RCRA
standards do not require monitoring in the land between the facility and
underground water; hence, an opportunity to gain an early warning of leaks is
lost. .
•
o Test for Statistical Significance! Tests required by RCRA keep the
probability of falsely detecting contamination low at the expense of high
probability that contamination might go undetected.
o Corrective Action Delays Complex RCRA procedures can lead to delays
of several years, increase cleanup costs, and increase the chances of CERCLA
financing of' cleanup.
o Compliance Monitoring and Corrective Action; Technology does not
-4-
-------
necessarily exist to meet the RCRA standards for taking corrective action, nor;
in all cases for compliance monitoring, required after contamination is found.
RCRA in Relation to CERCLA
There have already been cases of hazardous waste from clean-up at CERCLA
sites going to RCRA regulated sites which were later found to be leaking.
Moreover, although RCRA and CERCLA are managed by the same agency, research
for this analysis has found that the two program offices do not coordinate
- r
closely and that one office appears to be unaware, at times, of what the other
is doing.
Many people view • RCRA as the program which will prevent present and
future hazardous waste sites from becoming CERCLA sites. However, in - the
80,000 word preamble to the final RCRA land disposal regulations standards,
written two years after the passage of CERCLA, there is no reference to the
concept that the standards are to serve the purpose of preventing regulated
sites from becoming uncontrolled sites. Indeed, the only two references to
CERCLA in the preamble a*e in the context of what CERCLA can do for RCRA, not
*
what RCRA can do for CERCLA. Consequently, it appears that RCRA groundwater
monitoring and protection standards were not designed to prevent RCRA
regulated sites from becoming CERCLA sites and they are not capable of doing
so.
Interim Status Facilities
Although they are "grandfathered" by the RCRA legislation, interim status
-5-
-------
facilities do not have EPA-issued permits for operation. In contrast to the
regulations for new facilities, existing interim status facilities are not
designed or operated to EPA's specifications for adequate groundwater
protection. However, these facilities are the ones most likely to have
received wastes which are most inappropriate for land disposal (e.g.,
uncontainerized, highly toxic liquids). No matter what may be done to limit
land disposal in the future, the interim status facilities have already
received billions of tons of hazardous wastes over several decades; they
continue to receive wastes. Moreover, available data and historical
experience-indicates that many of them already are, or are likely to become,
leaking sites which will require corrective or remedial action. It will be
many years before EPA can closely examine interim status sites - even ones
given priority - to determine whether or not, and how, they should be
permitted. But every day that goes by without detecting existing
contamination or correcting contamination once it is found, adds to the cost
of correction and makes it more likely that CERCLA will be involved.
Nevertheless, the groundwater monitoring requirements for interim Status sites
are far less stringent than for new facilities designed to EPA specifications
and there are no corrective action requirements. Alternatives to current
r
regulations which could reduce high future cleanup costs include: requiring
financial assurance for corrective action; improved monitoring and sampling;
requiring prompt corrective action upon discovery of contamination; and
promptly closing down obviously badly designed and badly located facilities.
No requirement to fix leaking land disposal facilities
Although EPA regulations require new hazardous waste disposal facilities
to be designed so that they do not leak for at least 30 years after closure,
-6-
-------
if Che facility does leak there is no requirement in the RCRA regulations that.
the facility be closed until the leak is found and corrected, indeed there is
not even a requirement to find or stop the leak. Cleaning up the consequence
of a leak, such as a plume of pollution, but not the leak itself is only a
temporary expedient. Since the cost of cleaning ground water is generally
^
proportional to the amount of time a site is allowed to leak, inattention to
leaks raises the cost of remedial action to the point where facility owners
may not be able to afford facility modification and cleanup and the result is
an abandoned site. This research found no cost/benefit analysis or risk
assessment"to justify this policy, which runs a considerable risk of creating
more sites and high cleanup costs for CERCLA. The result may be that short-
term benefits will accrue to facility operators and users, and the longer-term
costs likely to be borne by the site operator, the government, and the public
will mount.
RCRA coverage stops at the fenceline
RCRA regulations do not 'require corrective action for groundwater
contamination which goes beyond the fenceline of the regulated facility which
created the problem. The reason given by EPA is that it may not be possible
for the owner to gain access to the neighboring property in order to conduct
•corrective action. EPA assumes that the problem of plumes migrating off the
property boundary would be addressed under CERCLA. However the same agency
administers CERCLA and the same problem of gaining access to the neighboring
property would have to be faced under CERCLA. It is unclear why this problem
can be addressed under CERCLA, but not earlier and less expensively under RCRA
which does not legislatively limit actions to within facility boundaries.
-7-
-------
RCRA coverage limited to 30 years
Even though many toxic wastes will remain dangerous for many decades, if
not forever, RCRA regulations require that a new hazardous waste disposal site
be designed so that it will not leak for 30 years after closure. The
.regulations also require the site owner to be responsible for routine
maintainance of the site for 30 years after closure. However after 30 years,
when the site may be more likely to leak, or for a leak to be detected through
adverse effects, the maintenance cost is turned over to CERCLA.
Financial.responsibility
A major cause for the abandonment of hazardous waste disposal facilities,
and subsequently their becoming CERCLA sites, is the inability of site owners
to finance the high cost of corrective action. This was recognized by
Congress in its explicit requirement that the RCRA regulations provide
assurances of financial responsibility consistent with the risk. Nevertheless
the regulations have no financial assurance requirement for corrective
action. A prudent, precautionary approach in establishing the level of
financial responsibility, considering the historically proven limits of the
technology, would be to assume that a leak will occur, will not be detected
«
very early, and that groundwater contamination will be significant.
Contaminants regulated under RCRA and CERCLA
The universe of toxic groundwater contaminants of concern to CERCLA is
greater than those of concern to RCRA. CERCLA regulates all contaminants
defined by RCRA but not vice versa. Therefore, a RCRA regulated facility in
compliance with all RCRA standards can still become a CERCLA site.
Additionally for many contaminants of concern to both RCRA and CERCLA, the
-------
levels of detection are set higher under RCRA procedures than under CERCLA
procedures.
Tolerance levels of contaminants
Under RCRA, EPA does not appear to have set tolerance levels of
groundwater contaminants based on their danger to human health, yet under
CERCLA EPA is concerned with any contamination which threatens human health.
Under RCRA, tolerance levels appear to be whatever detection limits result
from the chemical analysis techniques used, and the choice of technique
appears to' be based on cost and ease of analysis rather than health factors.
This is borne out by the fact that for many chemicals the tolerance level
(allowable concentration) appears too high to adequately protect human health,
and for many more chemicals, including EDB, dioxin, and DBCP, test protocols
were established without knowledge of their detection levels. There is no
cost/benefit analysis to evaluate whether costlie.r analytical techniques
should be used to lower the detection limits, and no risk analysis to evaluate
whether land disposal of some chemicals should be banned until the detection
limits that are determined to be adequate are set by EPA.
»
Geological standards
There are some geological formations in which groundwater movement cannot
be predicted; hence, groundwater cannot be monitored effectively at these
sites. There are others in'which groundwater contamination moves so rapidly
that it cannot be detected and corrected before it has spread dangerously.
Many states (e.g. California and Illinois) and other government agencies (e.g.
Nuclear Regulatory Commission), therefore, have set standards which preclude
locating land disposal facilities in certain geological formations. The RCRA
-9-
-------
standards, however, do not recognize this problem in permitting hazardous:
waste land disposal facilities. EPA has indicated that corrective action
technology to effectively deal with groundwater pollution will become
available in the future.
Groundwacer monitoring
Groundwater monitoring must be "custom tailored" for each site. There
are numerous complex hurdles to be overcome in order to do the job right, the
failure of any one of which can lead to incorrect results. If the geology of
the site -is suitable (which it frequently is not) and if enough time, money
and expertise are spent in designing and operating a groundwater detection
system, then there is a reasonable chance of detecting pollution. However,
groundwater monitoring has not yet proven its effectiveness as a regulatory
tool. Technical complexity and site specificity make it difficult - for
government rules to set the conditions for effective groundwater monitoring.
As a result, many facilities are inadequately monitored and significant
improvement in the future is unlikely. A possible alternative would be to
have the government (but not necessarily a regulatory agency) conduct
monitoring. *
i
Monitoring in the Vadose Zone
Detecting contamination before it reaches groundwater (i.e. in the vadose
zone underneath the facility) might save millions of dollars in corrective
action costs and might make the difference in keeping a site from becoming a
CERCLA responsibility. Vadose zone monitoring has been used for some years at
hazardous waste facilities in several states. The techniques have been
studied by EPA's research laboratories and many of them are available today.
-10-
-------
Nevertheless, both EPA's RCRA interim status standards and the 1982 standards
for permitted facilities dismiss the use of this technology without analyzing
its effectiveness in reducing groundwater cleanup costs.
Test for statistical significance
Before a facility 'is required to report the presence of contamination in
a detection monitoring well, a test for "statistical significance" is
performed. EPA has chosen a test procedure which keeps the probability of
falsely detecting contamination low, but this has happened at the expense of
increasingrthe probability that groundwater contamination might go undetected
until it becomes obvious through environmental impacts, when cleanup costs may
soar. Indeed, EPA apparently has not calculated the probability of detecting
contamination with their procedures. Under some circumstances (e.g. interim
status facilities following minimum RCRA requirements) the probability, of
detecting contamination may be such that the plume of contamination goes by
the detection system for many years.
Delays in onset of corrective action
The RCRA regulations contain many complex procedural steps which can
cause delays of several years in implementing corrective action, increase the
•costs of cleanup, and increase the chances of the need for CERCLA.
Compliance monitoring and corrective action
For most cases, the technology does not exist to meet the standards for
taking corrective action required by the RCRA regulations, nor in all cases
for compliance monitoring, required after contamination is found. The
regulations rely on the availability of the technology some time in the
-11-
-------
future. The option of banning land disposal for untreated hazardous wastes
until the technology to clean up groundwater (to background levels as required
by the regulations) is available does not appear to have been evaluated. How
such sites will be treated is unclear. If EPA insists on their meeting an
unachievable monitoring or cleanup standard then the sites may be forced into
«• -~
bankruptcy and into the CERCLA program. If such sites are allowed to operate,
Chen groundwater pollution would likely worsen.
-12-
-------
INTRODUCTION AND BACKGROUND :
This analysis examines how EPA has implemented the RCRA statute, and how
this implementation affects the use of CERCLA. Although OTA has not focused
on the details of the RCRA statute in this analysis, 'there does not appear to
be a major statutory problem. On the other hand, it is possible to conceive
of statutory changes which could 'remedy the problems found in this analysis.
Indeed, in the current RCRA reauthorization process some changes have been
proposed which would direct EPA to remedy some of the problems discussed in
this Memorandum.
The Scope of Superfund. CERCLA provides authority to EPA to arrange for
removal and provide remedial actions whenever any "hazardous substance" is
released or there is substantial threat of such a release. In addition,
whenever there is a release or substantial threat of a release of a "pollutant
or contaminant" which may present an imminent and substantial danger to the
public health or welfare, EPA may also initiate removal and remedial action
(CERCLA §104(a))« The term "hazardous substance" means not just' "hazardous
waste" as defined under RCRA, but includes any material designated as
hazardous or toxic under, the Clean Water Act or the Clean Air Act (CERCLA
§101(14)). "Pollutant or contaminant" is defined even more broadly to cover
any .substance that can cause death or serious health effects (CERCLA
§104(a)(2)).
Thus, CERCLA goes far beyond the original interest in the adverse health
and environmental impact of hazardous waste disposal. It includes the impacts
from such sources as mining operations, leaking pipelines, runoff from raw
materials, piles and spills from loading operations.
-13-
-------
At the time CERCLA was passed, there was no systematic attempt to
ascertain whether any kind of prevention programs were in place. While in
many areas, such as air, surface water, ground water, hazardous waste, and
surface mining, there are Federal laws in effect, for many others there are
none. This places the Federal government in the position of assuming the
responsibility for the failure of operations over which it has no original
regulatory control.
The extent of government control over the several causes of environmental
problems covered by CERCLA merits considerable study. This paper, however, is
limited to the study of land disposal of hazardous waste as regulated by
RCRA. While all modes of pollution are covered by CERCLA, this paper will
only look at groundwater contamination. This is the most significant mode of
contamination accounting for the majority of the sites on the National
Priorities List (1). Moreover, cleanup of groundwater contamination poses
substantial technical complexity as well as very high costs.
RCRA and Land Disposal. Several aspects of the RCRA regulations have
already received considerable analysis. For example, OTA completed a major
study of hazardous waste Control in March, 1983 (2). Another major study was
done by the National Academy of Sciences (7). A large part of these and many
other studies dealt with the technology of hazardous waste land disposal and
its alternatives. Therefore this paper will not focus on EPA's regulations
under RCRA for the design and construction of hazardous waste treatment and
disposal facilities. This analysis concentrates on ' EPA's groundwater
protection standards for land disposal facilities.
There are, however, several conclusions from these earlier works which
will help to better understand the context of this paper. The first is that
-14-
-------
even with the best available land disposal technology, hazardous wastes placed ;
in land disposal facilities will likely migrate into the broader environment
sooner or later. The second is that there are commercially available waste
reduction and waste treatment alternatives to the land disposal of many
hazardous wastes. And the third is that RCRA regulations present technical
(•
and economic disincentives to industry to utilize more fully these alternative
technologies.
Many more resources continue to be allocated to the regulation of
fundamentally flawed land disposal technology than to the development and
demonstration of alternatives to land disposal. EPA has frequently been
criticized for not encouraging alternative technological approaches to the
land disposal of hazardous waste. EPA's response has been (a) that the
technology for recycling and alternative treatment to land disposal may not
exist for all or most wastes, (b) that the technologies are not "off-the-
shelf" but are in some stage of development, and (c) that to the extent to
which technology does exist, the necessary plant capacity may not be in
place. However, it will be seen from this study that EPA did not apply these
same conditions to the writing of the land disposal groundwater protection
standards, as they suffer from all of the same defect-s.
t
To sum up, RCRA regulations cannot overcome the fundamental inadequacies
of land disposal technology, and experience has shown that regulatory
enforcement efforts do not assure compliance with regulations. Just 'as
troubling, the following analysis reveals that even if there was compliance
with RCRA groundwater protection standards, land disposal would still pose
serious risks to health and environment. Moreover, attempts to limit the
future use of land disposal do not address the problem of billions of tons of
hazardous waste already land disposed.
-15-
-------
Data to Illustrate the Scope of the Problem. About 2000 hazardous waste
land disposal facilities required to conduct groundwater monitoring filed for
interim status. EPA has released data for 1981 which provide some indication
of the number of hazardous waste management facilities which operated that
year and which might threaten groundwater. (Note that injection wells are
*• -
regulated under another statute and not by the RCRA groundwater protection
standards even though they are used for hazardous waste disposal.)
surface impoundments 770
landfills 200
injection wells 90
land treatment • 70
waste piles 170
storage and treatment tanks 2040
OTA has analyzed the data from EPA's study of waste management in 1981 to
examine the extent to which land disposal facilities receive hazardous wastes
which are toxic. Such information has not been available previously. Toxic
wastes present long-term chronic health risks and are to be contrasted with
waste which are hazardous only on the basis bf characteristics such as
.reactivity, ignitability, and corrosivlty. These results are given in Table
1, but it should be recognized that the data have poor statistical reliability
and there likely have been changes in hazardous wastes and waste management
practices since 1981. Nevertheless, the data indicate 'that a significant
fraction—perhaps a majority—of the wastes being placed in land disposal
facilities nationwide are toxic chemicals which pose long-term health problems
if released into the environment. For surface impoundments and landfills
almost all the wastes may be toxic, while for injection wells about one-third
-16-
-------
Table 1
NATIONAL ESTIMATES OF HAZARDOUS WASTE TYPE BY PROCESS OF DISPOSAL
FROM EPA 1981 SURVEY
(in millions of metric tons)
Well Injected
Surface Impounded
Landfilled
Land Treated
Other
Total Land
Disposed
Reported
as Toxic1
8
14
- 3
0.2
2
28
Reported as Reported as
Non-Toxic Waste Only
14 45
0.7
0.3 0.1
0.1
0.1
14 4
4
Totals
26.1
15.1
3.3
0.3
2.4
47.27
(Columns and row totals may not check because of rounding.)
Us defined in 40 CFR 261.24, 261.30 -261.33.
As defined in note 1; wastes that are only ignitable, corrosive, and/or
reactive.
Respondants did not specify wastes by appropriate RCRA hazardous waste
numbers.
4
Private communication from EPA to OTA.
5 •
OTA analysis of data in "The CMA Hazardous Waste .Survey for 1981 and 1982"
indicates about 60 million metric tons of hazardous wastewaters were injected
into wells for the entire chemical industry. These wastes do not appear to be
included in the EPA data. Nor is it clear what type wastes these are.
'May include above categories, ocean dumping, etc.
7
The CMA report also indicates that, excluding wastewaters, hazardous wastes
regulated by the states but not Federally can be as much as the amount which
EPA regulates.
Source: OTA
-17-
-------
of the wastes may be toxic.
A recent report by EPA's Superfund Task Force discusses the future of the
Superfund program. (Memo to Alvin L. Aim and Lee M. Thomas, December 8, 1983)
EPA projects a total inventory of 22,000 uncontrolled'sites. As of December,
1983, nearly 900 sites had been evaluated; and 546 of those sites have been
placed on the National Priority List (NPL). Contamination of groundwater is
the number one problem with currently assessed uncontrolled sites. For
example, for the 881 sites scored for the NPL, 526 sites had observed releases
of hazardous substances into groundwater. Over eight million Americans are
potentially exposed to the groundwater from these sites, and in about 350 of
these sites the contaminated groundwater is the only source of drinking water
for the affected population. Another 6.5 million people are potentially
exposed to contaminated surface water at 450 sites. Most of the commonly
encountered of the 444 toxic pollutants found at these 881 sites are
acknowledged by EPA to exhibit chronic toxicity and pose health threats at
extremely low levels of human exposure.
Furthermore, most of the cleanups being conducted under Superfund involve
either leaving the waste^ in the ground and attempting to contain them, or
removing wastes and contaminated materials and placing them in land disposal
«
sites. Of the 546 sites on the NPL, 40 percent were landfills originally and
30 percent were surface impoundments. We are beginning to see cases of land
disposal sites leaking after they have received wastes from Superfund cleanups
(e.g., the BKK facility in California). This is to be expected, as EPA
research, as early as 1975, indicated that more than 90 percent of operating
land disposal facilities were leaking. Therefore, not only is the RCRA
regulatory program contributing to future Superfund burdens, but the Superfund
-18-
-------
program is adding to the uncontrolled site problem through its own cleanup
efforts. While attempts to spread limited Superfund resources among many
sites may seem necessary and reasonable, the longer term risks (often to
different communities) and costs support a different approach.
EPA's Dependence .on Current Groundwater Protection Standards. Current
Federal regulatory control of hazardous waste land disposal facilities is
critically dependent on EPA's groundwater protection standards. Because of
the admitted deficiencies and uncertainties of land disposal technology, such
as the inability of synthetic liners to fully contain liquids and the unproven
long-term effectiveness of leachate collection systems, protection of human
health and the environment rests ultimately on the protection afforded by the
•
groundwater monitoring requirements.
For example, EPA's director of its Office of Solid Waste has said:
While no method of hazardous waste management is failproof, our rules
should protect human health and the environment. Even if a containment
system fails, groundwater monitoring will identify leakage and the
pollutant plume will have to be cleaned up. (Letter from John H. Skinner
to Keith H. Gordon, August 12, 1983.)
However, no mention is made of dealing with the leak itself, nor of
stopping the disposal of hazardous materials in the leaking site. Cleaning up
the pollutant plume is of limited effectiveness when the leaking is allowed to
continue.
And the director for air and waste management in EPA's Region VIII has
said:
In the Agency's view, the cornerstone of our land disposal program rests
on the groundwater protection standards. They were devised to provide
essential environmental and health controls. (Letter from Robert L.
Duprey to Leo Younger, August 10, 1983.)
More recently, EPA has been formulating a national groundwater protection
strategy in response to a growing awareness that this national resource needs
-19-
-------
more effective protection. EPA- recognizes that "In most circumstances it is
prudent to protect the resource from contamination in the first place, rather
than rely on cleanup after the fact." However, because of OTA's conclusions
concerning the inadequaces of the RCRA groundwater protection standards, it is
imperative to note that EPA's new national groundwater protection strategy
*.
guidelines "...will not alter the existing technology and monitoring
requirements for hazardous waste facilities incorporated in existing
regulations." ("Draft A Ground-Water Protection Strategy for the
Environmental Protection Agency," January, 1984.) Thus, OTA concludes that
the goal of protecting the resource rather than cleaning it up after the fact
is in serious jeopardy.
The Economics of Prevention. The national problem of uncontrolled
hazardous waste sites has received much attention not merely as a result of
the threats to human health and the environment, but also because of the high
costs of cleanup. What was once perceived to be a problem that might be
handled with a five year $1.6 billion program, is now generally recognized to
require a long-term commitment - perhaps many decades - with costs which are
still difficult to forecast.
EPA has estimated that 1400 to 2200 uncontrolled sites will require
-Federal action as National Priority List (NPL) sites for a cleanup cost of
$8.4 billion to $16 billion (in 1983 dollars). The EPA estimate does not
include costs for decontaminating polluted aquifers. In March, 1983, OTA
estimated future cleanup costs at $10 billion to $40 billion. However, an
unreleased survey of the States conducted for EPA indicated that State
officials believe that well over 7,000 sites will require cleanup under
Superfund; if true this would bring cleanup costs to the high end of the OTA
-20-
-------
estimated Moreover, such estimates have not included studies to indicate the
extent to which present RCRA facilities, both for hazardous and nonhazardous
solid wastes (municipal and sanitary landfills), may become future
uncontrolled sites. These may total in the thousands. Nor do these estimates
include the costs for cleaning up Federal uncontrolled sites, which now number
about 500 in EPA's inventory and are expected to increase.
A major economic issue is the extent to which it pays to prevent more
uncontrolled sites from being created. The primary consideration is the
widespread use of land disposal rather than alternatives to it. Even if such
alternatives were substantially costlier than land disposal in the short-term,
they would still be cheaper than the ultimate cleanup costs for uncontrolled
sites resulting from land disposal. When such cleanup costs are related to
the amount of hazardous waste originally disposed they are generally 10 to 100
times greater than the costs of currently expensive waste treatment options.
Cleanup costs for uncontrolled sites vary greatly and depend not only on
the nature of the site's problem(s), but also on the extent 'of cleanup
chosen. If permanent rather than "band-aid" cleanups are used, then costs
escalate sharply. For »xample, cleanups which leave wastes in the land or
move them to another land disposal facility are far cheaper than the use of
bnsite or offsite destruction or detoxification of wastes. But such lower
short-term costs for containment and land disposal ignore probable future
costs of cleanup actions at such sites in just the same way that land disposal
.of newly generated wastes does.
Moreover, although there is much groundwater contamination at
uncontrolled sites, there have been very few attempts to actually
decontaminate the water rather than to simply contain the plume of pollution
-21-
-------
by, for example, a slurry wall, or to take no attion. Decontamination of
groundwater is a very costly and time-consuming process which can take tens of:
millions of dollars and many years for an aquifer. However, such cleanup
costs can be minimized by minimizing the extent of groundwater
contamination. Simply put, the greater the volume of contaminated
groundwater, the greater the cleanup costs and time. In addition to
preventing leaking land disposal facilities and correcting leaks themselves,
future groundwater cleanup costs, therefore, can be reduced by early detection
of groundwater contamination and prompt cleanup. There is now some evidence
(albeit of-a statistical nature) that suggests that EPA's strategy may be not
to spend CERCLA funds to decontaminate groundwater. A recent analysis of
EPA's use of the Hazard Ranking System (HRS) and its allocation of CERCLA
funds found the following:
...the HRS ground water scores bear a statistically significant negative
relationship to obligations. This means that when the HRS total score
increases due to a higher ground water score, the increase in obligations
is smaller than if the increase in the total score is attributable to
another component measure of hazard. ...Given the relatively high cost of
cleanup when ground water contamination is present, EPA may have concluded
that the damage associated with other cleanups foregone is too great to
justify cleanup of a particular site's ground water. ...If EPA places
greater weight on short-term dangers, they would be less likely to fund
remedial action in relationship to ground water contamination. ...this
aspect of EPA's Supe^rfund allocation decision making shifts the social
cost of hazardous waste forward to future users of contaminated ground
water or to future tax payers. (Harold C. Barnett, "The Allocation of
Superfund, 1980-1983," Dept. of Economics, Univ. of Rhode Island.)
Finally, there is the issue of whether or not it makes a difference if
cleanup of groundwater at RCRA sites is accomplished under the CERCLA program
(which this analysis concludes is likely to be the case) rather than through
the RCRA program. Aside from the equity of the situation, there is a
difference ', if it is advantageous to have the operators and users of RCRA
land disposal facilities bear the actual or anticipated cleanup costs so that
the market price of land disposal reflects its true long-term costs. Cleanup
-22-
-------
may require CERCLA funding*" without collection of moneys from responsible
parties. Enforcement action under CERCLA may not be effective for the same
reasons that RCRA enforcement actions may not be effective (e.g., due to
bankruptcy of the facility operator). Consequently, through the financing
mechanisms of CERCLA, cleanup costs are borne by industry broadly and the
*
general public rather than directly by the most responsible parties.
Moreover, by shifting cleanup to CERCLA there is likely to be more procedural
delays which contribute to additional cleanup costs as leakage continues and
groundwater pollution spreads.
After closure, responsible parties may not bear full costs. This is true for
a facility which is closed and, after five years, when there is no detection
of leaking, it becomes covered by CERCLA's Post-Closure Liability Fund.
Although the fund is supported by a tax on land disposed wastes, there is no
distinction among facilities on the basis of their design, location, or
operation; hence, there is no incentive for active facilities to reduce taxes
by achieving maximum protection. Nor is there is any assurance that the fund
will be able to fund extensive actions to fix leaks and cleanup groundwater
contamination.
-23-
-------
INDUSTRIAL SOURCES OF CERCLA SITES
Although RCRA regulates many industrial sites, it does not have
jurisdiction over non-waste related activities which may cause a site to be
addressed under CERCLA.
*
A review of the National Priorities List shows many manufacturing sites
where non-waste materials have been spilled or discharged resulting in
polluted groundwater. Some of the mechanisms are:
o spills in loading areas
o leaking tanks
o runoff from storage piles
o spills from floods, hurricanes and fires
o leaking underground pipelines, and
o leaking manufacturing equipment.
Few measures* at the Federal level have been taken to prevent such non-
waste sources of CERCLA sites. There are industrial sites regulated and
inspected by EPA under RCRA which have considerable groundwater pollution from
non-waste sources, but these are largely ignored by EPA. For example, a
manufacturing plant might have a waste pile and' a storage pile of raw
materials on the same site. Both may be capable of polluting groundwater from
runoff. The legal position of EPA is that there is an advantage in having a
groundwater monitoring network which does not detect pollution from the'
material pile, because doing so would confuse any enforcement action the
Agency could take against the site owner under RCRA. However, pollution from
One of the few measures which has improved industrial operations is the
CERCLA reporting and liability requirements for leaks and spills.
-24-
-------
the material pile would trigger action under CER~CLA. Therefore, in the absence;.;
of Federal measures to control pollution from such non-waste sources, it is
reasonable to expect increasing pollution problems to come under the purview
of CERCLA.
There are also several waste-related sources of groundwater pollution
that have been addressed by acts of Congress but for one reason or another are
not required to comply with the most stringent groundwater protection
regulations. Often there is a presumption of effective waste containment
technology (Type A), that wastes do not contain toxic materials (Type B), or
that toxic wastes will not enter the ground (Type C). These are not
necessarily correct. These facilities could, therefore, become uncontrolled
CERCLA sites. Examples of these, which are not the subject of the following
analysis, include:
Type A
o double lined waste disposal sites with leachate collection and leak
detection systems
o injection wells
o closed hazardous w^ste disposal sites not yet leaking
Type B
o facilities for RCRA exempt wastes, including state regulated hazardous
wastes
o impoundments and sanitary landfills (RCRA - Subtitle D) for solid
wastes
o disposal sites for petroleum drilling wastes
Type C
o waste recycling and recovery sites
-25-
-------
INTERIM STATUS ;
When Congress passed RCRA in 1976, it provided a "grandfather" clause for
existing facilities so that they could continue to operate as if they had a
permit until EPA issued them a permit (RCRA §3005(a)). This "interim status"
was to allow for a smooth transition to a condition of federally permitted
hazardous waste treatment, storage and disposal facilities. It was not
envisioned, at that time, that this process would take almost two decades. As
of December 1983, there were about eight thousand interim status sites. Two
thousand of these are required to monitor groundwater because they conduct
waste management activities capable of polluting groundwater, such as
landfilling and placement in surface impoundments (3). Although seven years
have elapsed since the passage of RCRA, none of these two thousand facilities
has yet been issued a permit by EPA (3);* thus all continue to operate under
interim status. While the permitting process has begun, EPA estimates (6)
that it will not complete the permitting of the 2,000 facilities for ten more
years. In the following discussions the use of the terms "new" or ."permitted"
facilities refers to either newly built facilities or interim status ones
which have become permitted.
*
r
EPA's Implementation. Although Congress allowed interim status
facilities to operate without a permit, it did not excuse them from complying
with all the standards necessary for the protection .of human health and the
environment. However, in May of 1980, EPA issued "interim status standards."
(40 CFR 265) as the "minimum requirements" for interim status facilities.
These were, by EPA's admission, considerably less than what would, be necessary
To date EPA has permitted only three disposal facilities under RCRA; all of
these are new facilities (3).
-26-
-------
to meet the legislative requirement for standards adequate to project human
health and the environment. These interim status standards (or Part 265
standards) are "in lieu of" (40 CFR 264.3) the more stringent part 264
standards which only go into effect after the facility is permitted by EPA.
.This action cut off any means of bringing an interim status facility into
compliance with standards "adequate to protect human health and the
.environment" short of issuing (or denying) a permit.
EPA's estimate of the time to permit all interim status facilities is now
ten years,, after having been revised upward several times. Many facilities
could be in interim status for ten years and some for even longer. EPA states
that these facilities will be permitted on a priority basis with the highest
priority going to facilities which show the greatest environmental problems.
Even where problems are identified, it takes over a year to process a pe-rmit
and there is a backlog of over 1500 disposal facilities waiting for their
permits to begin to be processed.
As previously mentioned, the interim status (Part 265) regulations do not
require interim status facilities to comply with the more stringent Part 264
groundwater protection and facility design standards. The technical details
of the groundwater protection standards will be discussed later, but the
importance of stringent groundwater protection can be .seen by the fact that
there are already over fifty RCRA interim status facilities regulated by EPA
on the CERCLA National Priorities List (9). And several interim status sites
*There are provisions in both RCRA and CERCLA for EPA to seek an injunction to
require action if it can be demonstrated that there may be an imminent and
substantial endangerment to health or the environment. These provisions may
have been used in a few cases to require corrective action for groundwater
pollution at an active interim status site. Their use at an active RCRA
regulated site would indicate that there are no pertinent regulations with
which the agency can require compliance.
-27-
-------
in which wastes from CERCLA remedial action clean up activities have been .
disposed have been found to be leaking and could themselves become CERCLA
sites.
Although the interim status groundwater monitoring requirements have only
..recently gone into effect, about 145 facilities are currently "in assessment"
because their groundwater monitoring systems indicate that they are polluting
groundwater (10). This figure takes on more significance when considered with
a 1983 study by the General Accounting Office (6) of several states with above
average regulatory programs. The study found that only 22% of the regulated
facilities were complying with the interim status groundwater monitoring
requirements.
EPA is reported in the press to have estimated that 50% to 60% of the
interim status land disposal facilities are leaking and will require
corrective action (60). There is evidence that the figure is closer to 90% to
100%. A study conducted by EPA in 1975 (12) investigated 50 facilities
randomly selected from these 2,000 hazardous waste disposal facilities and
found that over 90% of them were leaking into groundwater. Therefore, even
.before the passage of RCRA, the poor state of these interim status facilities
was well known.
EPA could have written regulations for financial assurance for corrective
action; regulations to monitor and gather necessary environmental data as well
as regulations to bring them promptly in compliance or close them down.
However, the interim status standards abrogate most of EPA's authority to
regulate interim status sites until they are issued a permit by EPA. These
facilities may continue to operate for a decade or more, perhaps leaking all
the while, increasing the ultimate cleanup cost and increasing the chances of
-28-
-------
their ultimately becoming uncontrolled sites. ;
Indicator Parameters. To illustrate just one aspect of the interim
status standards, consider the parameters required to be monitored in
groundwater at interim status sites. EPA has identified four indicator
parameters to determine whether an interim status hazardous waste facility is
leaking enough to cause "gross contamination." The four indicator parameters
are: specific conductance, pH, total organic carbon, and total organic
halogen. In its interim status permitting standards, EPA limited the
groundwater monitoring requirements for purposes of leak detection to these
four parameters (40 CFR 265.92(b)). EPA gave the following reason for
choosing these four parameters (45 FR 33194):
Increases in specific conductance indicate the presence of
inorganic substances in the groundwater. Likewise, increases
or decreases in pH suggest the presence of inorgar.ic
contamination. Total organic carbon (TOG) and total organic
halogen (TOX) concentrations in groundwater tend to increase
as a result of organic contributions from a hazardous waste
facility. The methodology to sample and analyze for these
indicators is presently available. EPA believes that
monitoring these indicators will be sufficient to make the
threshold assessment of whether a facility is leaking.
However, the more stringent Part 264 standards for EPA permitted sites
;
(40 CFR 264.98) give the EPA permit writer the option of requiring monitoring
of the actual waste constituents or their reaction product rather than the
four indicator parameters. EPA's guidance to. the permit writers (13) says
this about the four indicator parameters:
In some cases, these parameters may not be the most
appropriate, and this use should be carefully reviewed before
they are included as indicator parameters in a detection
monitoring program. For example, TOC and TOX will be of
little value at a facility where no organic wastes are
present, and even at facilities handling organic wastes,
background levels may reduce the utility of these
-29-
-------
parameters. The use of pH and specific conductance may also
not always be appropriate. There are so many geochemical
controls on pH, such as natural buffering capacity, that it is
difficult to predict what changes in pH might occur in a
leachate migrating through the unsaturated and saturated
zones. In addition, unless extremely acidic or basic, the
addition of large amounts of leachate will likely be required
to significantly alter pH. Consequently, pH may be suitable
only as an indicator of gross contamination. Detectable
changes in specific conductance will similarly require a
relatively large increase in ion concentrations.
; Consequently, it may also be useful as an indicator of gross
pollution, and then only at facilities where constituents
migrating to groundwater are primarily inorganic ions.
Further criticism of the ability of the indicator parameters to detect
toxic contaminants at critical concentrations was made at a recent groundwater
symposium (14):
....there can be highly selective migration of contaminants
that are hazardous to human health in drinking waters at
concentrations far less than those that would be detected
using the "indicator" parameters. For example, the analytical
detection limit for TOX is 5 ug Cl/1. The toxic
concentrations of many organohalogens are less than 1
ug/l....for some organohalogens the critical concentrations
are on the order of picograms/1. For TOG, the analytical
detection limit is 1 mg/1. There is a large number of
chemical contaminants that occur in aquatic systems that have
critical concentrations for human health at orders of
magnitude below this detection limit.
•
Number of Monitoring Wells. Another feature of the interim status
standards is that they require only three wells for detecting groundwater
contamination. This is true regardless of the size of the facility, the size
of the aquifer, the extent of pollution, or the' potential for damage to human
health and the environment. In many cases, three wells are far too few to
give a reasonable probability of early detection of pollution. In the
processing of RCRA permits the number of required detection wells is generally
in the range of four to twenty for interim status sites currently operating
with three wells. On the state level, one interim status site in Illinois was
-30-
-------
required by the state to install 40 wells and another over 50 (66), and three
sites in New Jersey are required to have over one hundred wells (62) while
Federal standards require only three wells for the same sites.
In summary, the facilities which are most likely to leak, the two
thousand existing interim status facilities, have a much less stringent
groundwater monitoring standard then the three presumably far better designed
new facilities. EPA's own characterization of these standards is that they
are "minimal and are specifically designed not to be burdensome" (11). There
are no corrective action requirements or requirements to stop dumping should
groundwater contamination be detected. Sites found to be polluting will be
put on a "fast track" for • issuing a permit so that corrective action may be
required, but as of this date no Federal permits have been issued to interim
status facilities requiring groundwater monitoring.
-31-
-------
LIMITATIONS ON COVERAGE
The viewpoint of EPA, as evidenced in the groundwater protection
provisions of Part 264 of RCRA, is to determine when groundwater is getting
polluted enough to cause concern for public health and then to require the
*groundwater to be cleaned up. There tool for this is groundwater monitoring!
Groundwater monitoring is not a feasible substitute for techniques such as
leak detection systems used as a tool to analyze the engineering soundness of
the waste management facility, e.g., to locate a ruptured liner in a landfill
or a leaking storage tank. Permitted facilities are required to be designed
and built to exacting EPA engineering standards whose goal is to "minimize the
formation and migration of leachate to the adjacent subsurface soil or
groundwater" (47 FR 32312). However, when leachate does appear in groundwater
there is no requirement to find out what went wrong, "a landfill -liner which
has been designed not to leak does not violate the design standards if the
liner fails at some future time" (47 FR 32330). There is no requirement under
RCRA regulations for fully permitted facilities that the leak be fixed or that
the waste disposal activities be halted. When pollution may be coming from
one of several sources, there is no requirement to determine which of them it
is. In short, it is not a violation of any RCRA.standard to pollute. There
is only the requirement that the pollution which has reached groundwater be
cleaned up and this, as will be discussed later, is a very limited
requirement.
If the RCRA standards were designed less for the detection of pollution
and more for assurance of the engineering integrity of the facility, they
would have been more protective of human health. If EPA had the viewpoint
that the detection of any pollutant at any level was indictive of the failure
-32-
-------
of che facility to meet the design specifications, then EPA might require that:
waste disposal be halted while the failure is found and corrected, or the
waste removed. Rather than doing this, however, lengthy evaluations of the
extent of groundwater contamination are conducted. However, there is no
evaluation of the implications of a leak for the continued operation of a
facility.
A further measure which tends to suggest that many RCRA sites will become
CERCLA sites is the fact that RCRA groundwater clean up requirements end at
the boundary line of the facility (40 CFR 264.91 (a)(3)). Any pollutant that
runs off the property of a RCRA regulated site becomes a CERCLA problem. The
regulations explain that a site owner cannot be expected to get permission for
cleanup outside of the property under his control. The regulations go on to
state that "plumes migrating beyond the property boundary could, however, be
addressed under other authorities such as CERCLA" (47 FR 32311). The
regulations do not explain why EPA could handle this- problem under CERCLA—
perhaps years later—when EPA cannot handle it under RCRA.
A similar EPA limitation on its RCRA jurisdiction is to limit the site
owner's responsibility fo? site maintenance to thirty years after site closure
?f
(40 CFR 264.117 and 265.117). Since EPA (as well as many others) has
\
concluded that it is "inevitable" that landfills and disposal lagoons will
leak (46 FR 11126-28), it is therefore inevitable that many of these
facilities will eventually fall under CERCLA. Moreover, for a number of
.reasons (e.g. firms going out of business) clean-up costs would then shift
from facility owners and users to the government.
-33-
-------
GROUNDWATER MONITORING WELLS . [
The hydrogeology of the site is important in the design of a groundwater
detection monitoring system for interim status and permitted facilities. A
good knowledge of the hydrology and geology in the immediate area of a waste
disposal site is necessary in order to know where, how many, and how deep to
locate detection monitoring wells. In addition, for compliance monitoring, it
may also be necessary to be able to create a mathematical model of the
groundwater flow in order to be able to predict the speed and direction of
contamination movement.
OTA will shortly be coming out with a study of groundwater pollution
which will go into some detail on the science of hydrogeology so it is not
necessary to repeat that here. This discussion will therefore be limited to
this issue: how realistic and reliable are the RCRA (Part 264 and Part '265)
standards for establishing groundwater monitoring networks?
Hydrogeological structures are very complex. In determining the
location, depth, number, and type of monitoring wells a great many assumptions
have to be made about the underground geological structure at the site, the
•
adjacent area, and the location, depth, quantity, direction and speed of
underground water. Furthermore, the proper location of monitoring wells
depends on a knowledge of how all the above parameters may vary with season,
rainfall, tidal water, and groundwater usage. 'These latter factors can cause
groundwater flow to greatly increase, decrease, or even change direction over
time.
The physically hidden characteristics of hydrogeological structures mean
that they cannot be viewed but must be inferred from limited data. Such data
-34-
-------
are obtained from sources such as core samples, well drilling logs, and
historical rainfall data. The difficulty of doing this was summarized
picturesquely in a recent review by the Princeton University Water Resources
Program (27).
Effective monitoring of a hazardous waste disposal site is an extremely
difficult data collection problem. To understand its complexity, consider
air pollution. Often we can see whether the pollution controls on a
chimney are working: the smoke may be darkened and the odor (downwind)
noxious. As the wind carries the pollution smoke, we can see and follow
in direction.
Now imagine there are thousands of little chimneys around a factor site.
By looking at the smoke, we may be able to tell which air pollution
control devices are working and which are not. Again, we can see the
trail of polluted smoke as it is carried away.
Imagine that we cannot see the sky, we cannot tell the direction or
velocity of the wind, and we ask: Is the factory (with its thousands of
little chimneys) polluting the air? That is our groundwater monitoring
problem—at its easiest. It is made more difficult because the geological
properties of the soil vary with depth and direction, and this variation
is unknown or uncertain. When we look up in the sky, we observe the
spatial variation of the pollutants. If we could look up only through a
small tube or telescope, then the information we gathered from the one
sighting might not be representative of what we would see if we looked
everywhere. The small tube into the sky is like our groundwater
monitoring well: the data we gather may not tell us too much about what
is occurring in other nearby locations.
One of the few studies .of operational land disposal sites was an
investigation of 50 typical hazardous waste disposal sites conducted in 1976-
77 for EPA by the firm of Geraghty & Miller (12). One of the major
conclusions of this study was:
At sites presently monitored the use of wells as an aid in
evaluating groundwater conditions is generally poor, due to
inadequacies with respect to one or more of the following
parameters:
—number of wells
—distance of wells from potential contamination source
—positioning of wells in relation to groundwater flow
—selection of screened intervals
-35-
-------
—use of proper well construction materials
—sealing against surface water contamination, or inter-
aquifer water exchange
—completion methods, such as development, maintenance,
and protection against vandalism
Of the 50 sites evaluated by Geraghty & Miller, 32 of them had existing
.grbundwater monitoring systems which were usually installed to meet 'the
requirements of state law. Of the 32, Geraghty & Miller found 7 monitoring
systems (or 22%) so inadequate that they had to install new wells in order to
conduct the relatively basic monitoring required by the contract.
RCRA.was passed in 1976 during the Geraghty & Miller study. Six years
later, in 1982-83 EPA conducted another study of 148 interim status facilities
which had implemented groundwater detection monitoring programs in response to
RCRA interim status regulations (31). They found that 64 facilities (or 43%)
had "deficiencies related to the number, depths, and/or locations' of
monitoring wells." Among the problems encountered were:
o background wells not in the uppermost aquifer,
o background wells affected by the facility,
o downgradient wells not located in the direction of expected
contamination movement, and
t
o downgradient wells not located at depths which would intercept
contaminants.
These studies show that the percentage of unsatisfactory monitoring
systems was 22% in the 1977 study and 43% in the 1983 study. Since these two
studies are not comparable, it is perhaps too simplistic to conclude that the
practice of groundwater monitoring had deteriorated in those six years, but
there is no basis for believing, in spite of improvements in technology, that
-36-
-------
the practice had gotten any better. There are several possible explanations!
(not mutually exclusive) for this state of affairs. There was a workshop of:
experts on groundwater resources and contamination in the United States
sponsored by the National Science Foundation in March of 1983 (30). One
expert, Keros Cartwright, head of the Hydrogeology and Geophysics Section of
the Illinois State Geological Sur\rey, offered the failure of our institutions
as a major problem. He stated that: "From my experience, very few monitoring
systems today on existing disposal sites are adequately monitoring the
site."(33) And that "the most common reason we have (for monitoring) is
simply a cosmetic procedure to reassure the public. ... too many of our
monitoring systems are cosmetic, not real." (34)
Another expert, Professor John Cherry, pointed to limitations in the
state of the art as a second explanation. He observed, for example, that
"contamination migration in fractured rock is complex and generally
unpredictable" and that "prediction of contaminant travel paths through
fracture networks is generally beyond the state of the art" (35). Not only
fractured rock but fractured clay and fractured silt make for very difficult
monitoring conditions. The best media for predicting pollutant movement and
the one for which there is the most knowledge is sand and gravel. Ironically,
this media is the worst media for land disposal because of the rapidity of
pollutant movement in these very porous soils. The only soils which have good
containment properties and are hydrogeologically predictable are unfractured
silt and clay. However, these soils are found in only about 10 to 20% of the
United States (36).
-37-
-------
There are many other hydrogeological conditions which make the design of :
groundwater monitoring systems very difficult if not impossible:
o There can be connections between different aquifers which are difficult
to detect. (39)
o Groundwater flow can change direction due to: intrusion of tidal
water, seasonal recharge patterns,, nearby production wells, etc. (38)
o Leachate does not always flow straight down to an aquifer, but under
some geological conditions would flow at an angle and enter an aquifer
downstream of the monitoring wells. (24)
o Liquid contaminants in an aquifer do not always flow in the same
direction as the groundwater. (37)
A third possible explanation for the poor state of groundwater detection
monitoring involves a combination of institutional problems and current
technology limitations. Frequently, the establishment of a proper groundwater
monitoring system takes a great deal of money, time and expertise, all of
which are normally in short supply. In order to meet governmental regulatory
requirements without costing too much, reliance is placed on "engineering
judgment" rather than hard data. This warning appears in the EPA RCRA permit
writers guide (5):
Experience with the installation of monitoring systems for
compliance with the ?nterim Status Regulations has indicated that
most owners/operators who have hired a ground-water consultant to
install the groundwater monitoring system have not envisioned
spending the time or money to conduct as thorough an investigation
as is suggested in this chapter. To retrieve all of the information
necessary to design the system in accordance with considerations in
this document, test-boring and piezometer installation programs will
be necessary. Though some local geologic reports usually exist in
the region of most facilities, site specific considerations almost
invariably require extensive test borings. Because of the lack of
time and funds, in most cases parameters such as the direction of
ground-water flow and the nature of subsurface materials have been
determined through evaluation of local topography and, to the.extent
possible, evaluation of existing building foundation borings.
Monitor wells are usually located on the basis of this information
and completed to just below the water table. Variations in ground-
water flow direction and geologic variability have usually not been
considered because of lack of information. The primary factors for
-38-
-------
minimizing the pre-monitor well installation field investigation
have been time and cost.
A similar point about cost was made by David Miller at a
Congressional hearing in 1982 on EPA's Part 264 groundwater protection
s.tandards (41):
There are, of course, certain geologic environments in which
monitoring becomes extremely expensive and may not be cost-
effectively employed. In order to obtain credible information,
dozens of wells and hundreds of groundwater samples may be required
to develop an adequate analysis of the hydrogeologic system.
Although there are probably a large number of existing land diposal
sites located in such areas, it is my recommendation that no new
land "disposal facilities be allowed under these conditions
regardless of engineering design.
What is required for a facility operator to detect groundwater
pollution? The hazardous waste disposal facility operator must want to detect
groundwater pollution, and must determine if the geology of the site is
suitable for groundwater monitoring. The operator must be willing to hire the
experts, spend the time, and spend the money (probably far in excess of EPA's
minimum requirements). Finally, sampling and analysis procedures must be
designed which optimize the ability to detect contamination, even if they are
more stringent than EPA's procedures (see e.g. section on statistical
procedures). There are many facilities operating this way, although they are
not required to do so. However, they are not required to report to EPA the
results of anything over the minimum requirements.
At the other extreme is the facility operator who monitors his
groundwater because he is required to, fulfilling only the ' minimum
requirements of the law. He may hire experts more as the representative of
the interests of the facility operator in dealing with the regulatory agency
than in optimizing the efficiency of the groundwater detection system. Past
-39-
-------
experience has shown that groundwater detection systems designed and operated
under these circumstances have a low probability of detecting groundwater
contamination. Many of the sites on the National Priorities List (1) had such
groundwater monitoring systems.
The latest EPA Part 264 regulations of July 26, 1982, while" an
improvement over the Part 265 standards, do not take account of past
experience on the failure of regulatory groundwater monitoring systems, nor of
expert advice on the unsuitability of many geological formations. It
continues to rely on regulatory groundwater monitoring in any terrain to
detect leaks. But the minimum requirements of the regulations are inadequate
to assure a high probability of detection. As a result, many more sites,
including sites permitted under RCRA. will probably be added to the National
Priorities List.
One additional point should be made. Several experts have pointed out
that a knowledgeable but unscrupulous person could set up a groundwater
monitoring system which met all the legal requirements of Part 264 but which
would not be likely to detect a contaminant plume. This is mentioned to
illustrate the vulnerability of the current regulations.
-40-
-------
CONTAMINANT TOLERANCE LEVELS \
The RCRA regulations for EPA permitted land disposal facilities (40 CFR
264), unlike those for interim status facilities (40 CFR 265), provide for
detection monitoring .of the specific contaminants being disposed as an
alternative to the use of the four indicator parameters (at the discretion of
the EPA permit writer). This would appear to overcome one of the problems
mentioned in the section on Interim Status. Upon close examination, however,
this process raises many other equally troublesome issues having to do with
the tolerance levels of these contaminants.
In regulatory parlance the "tolerance level" of a chemical is the
concentration which is acceptable to the regulatory agency. The Part 264 RCRA
regulations do not have an explicit tolerance level for groundwater
contaminants except for the sixteen chemicals in the EPA primary drinking
water standard. However for the hundreds of toxic constituents listed in the
RCRA regulations (40 CFR 261 appendices VII and VIII) there is an implicit
tolerance level. The regulations specify that the EPA publication "Test
Methods for Evaluating Solid Waste, Physical/Chemical Methods" (17) shall be
used to determine whether* a sample contains a given toxic constituent (40 CFR
261 appendix III).
For most substances, reference 17 lists more than one analytical
method. Some methods are more sensitive than others. In issuing permits, EPA
plans to use relatively low cost scanning techniques, which are the least
sensitive methods, explaining (59):
The Agency feels that a special hiearchical approach is appropriate for
this purpose. These approaches will first use scanning techniques
designed to detect broad classes of compounds. If the presence of a
particular class of compound is detected, more specific analysis to
determine which constituents are actually present can then be initiated.
-41-
-------
Although some sensitivity may be sacrificed by such an approach, the range
of detection of certain scanning methods are clearly adequate....
Therefore, the detection limit of the scanning methods which are least
sensitive of the required test methods, constitutes a de facto tolerance
level, since no action- will be taken for contaminants which appear below that
level. Furthermore, there are more sensitive test methods than those chosen,
and EPA has demonstrated in the case of dioxin that more sensitive methods can
be developed when required. The RCRA regulations give no explanation of why
certain test procedures were chosen and why others were not. Finally,
tolerance levels are only implicit in these procedures for most cases, and
have not actually been determined, and this is discussed below.
Table 2 illustrates the fact that these implicit tolerance levels have
been set without adequate consideration of health effects. The first column
shows the minimum concentrations at which twelve selected chemicals cari be
detected using the RCRA procedures (17). For each of these chemicals, the
second column shows EPA's estimate of the concentration which EPA projects
will cause one cancer per one hundred thousand people drinking two liters a
day of the water over a lifetime (45 FR 79325-41). The concentrations
associated with cancer arT based on animal studies, and projections from high
doses to low doses, and projections from carcinogenic activity in animals to
estimated effects in humans. There are substantial disagreements about the
accuracy of such projections, and the values listed in table 2 are not
universally accepted. They are, however, EPA's own published projections and
they continue to be used by EPA. Since it is EPA's criteria which determine
whether a site should be included in CERCLA, these projections are relevant to
this study despite uncertainties about their derivation.
-42-
-------
Table 2
EPA DETECTION LIMITS FOR SOME CARCINOGENS
Highest permitted
EPA detection limit
Concentration projected**
to cause one cancer per
100,000 people T
Projected**
cancers per
100,000 people
.Chemical (nanograms/liter) (17)*
aldrin
.dieldrin
1,1,2, 2-tetrachloroethane
3,3 '-dichlorobenzidine
heptachlor
PCBs
benzo(a)pyrene
benzidine
chlordane
DDT
1,900
2,500
6,900
16,500
1,900
36,000
2,500
44,000
14
4,700
' nanograras/liter)
0.74
0.71
1700
103
2.78
0.79
28
1.2
4.6
0.24
-
2,600
3,500
4
160
680
46,000
90
37,000
3
20,000
* A nanogram is a billionth of a gram.
one part per trillion.
One nanogram per liter is approximately
* Projections based on the consumption of two liters (a little over two quarts) a
day of the contaminated drinking water over a lifetime. Projections are also
based on animal studies that include assumptions on the transfer of results
from animals to humans, and extrapolation from high doses to low doses.
Despite the uncertainties introduced by these assumptions, these are the
projections EPA uses. Column 3 has been calculated by OTA by dividing Column 1
by Column 2. This calculation converts back towards high doses. Uncertainties
introduced into Column 2 by high-to-low dose extrapolation are thus partially
corrected for in deriving Column 3. Column 3 contains no correction for
uncertainties introduced by applying animal results to humans.
tReference: 45 FR 79325-79341
-43-
-------
By dividing the entry in the first column by the entry in the second
column, the projected number of cancers per one hundred thousand is estimated
in column three. Thus, for example, table 2 shows that a hazardous waste
disposal site operator, permitted by EPA, may, without violating his permit,
pollute groundwater with up to 2,500 nanograms per liter of dieldrin. This is
*• , —
a concentration which EPA data projects may cause 3,500 cancers per hundred
thousand people who drink such water over their lifetime.
To put this in its proper context, EPA is currently seeking to ban the
use of pesticides on the basis that the cancer risk is as low as one in one
hundred thousand (8). Therefore, it is likely that a facility which is
polluting groundwater at a level which is projected to cause 3,500 cancers per
hundred thousand would come to the attention of CERCLA.
The next point to be made concerns the explicit tolerance level
associated with the sixteen contaminants for which there is an EPA drinking
water standard. EPA allows (20) that for pollutants for which there is an
existing EPA primary drinking water standard, RCRA permitted facilities may
contaminate up to the standard. The primary groundwater pollution standards
are shown in table 3. Just as in table 2 (and with the same caveats), this
t
table also projects the cancers per hundred thousand for those substances for
"which data are available from the EPA published source. In addition, the
fourth column indicates the substances known or believed to be carcinogens.
For some of these pollutants, there may be no "zero effects" level and
any amount of the substance is considered a risk to human health. For
example, cadmium is carcinogenic (23) and is not considered without risk at
any level (15). Arsenic, lindane and toxaphene are alleged carcinogens and,
-44-
-------
Table 3
DATA ON RCRA POLLUTANTS WITH PRIMARY DRINKING WATER STANDARDS
Pollutants
arsenic
barium
cadmium
chromium
lead
mercury
nitrate (as N)
selenium
silver
fluoride
endrin
lindane
methoxychlor
toxaphene
2,4-D
2,4,5-T, Silvex
EPA Primary
Drinking Water
Standard
(ug/1)*
50
1000
10
50
50
2
10000
10
50
1400-2400
0.2
4
100
5
100
10
Concentration
projected** to
cause one cancer
per 100,000 people'T
(ug/1)
0.022
Projected**
cancers
per 100,000
people
2300
Comments
a
b
a
0.186
0.0071
22
700
a - known human carcinogen (23)
b - probable human carcinogen Ifssed on animal studies (23)
'.
* ug/1: microgram per liter, or millionth of a gram per liter.
approximately one part per billion.
1 ug/1 is
** Projections based on the consumption of two liters (a little over two quarts) a day
of the contaminated drinking water over a lifetime.. Except for arsenic, projections
are also based on animal studies that include assumptions on the transfer of results
from animals to humans, and extrapolations from high doses to low doses. For
arsenic, projections are extrapolated from the effects of high doses in humans.
Despite the uncertainties introduced by these assumptions, these are the projections
EPA uses. Column 3 has been calculated by OTA by dividing Column 1 by Column 2.
This calculation converts back to high doses. Uncertainties introduced into Column 2
by high-to-low dose extrapolations are thus partially corrected for in deriving
Column 2. Except for the arsenic number, which is based on human data, Column 3
retains the uncertainties introduced by applying animal results to humans.
tReference: 45 FR 79325-79341
-45-
-------
as shown in table 3, are associated with significant cancer risks at the EPA
tolerance level.
The next point in regard to tolerance levels is that not all toxic
pollutants which can cause a site to be regulated under CERCLA are monitored
under RCRA. A most conspicuous example is dioxin contaminated soils which "are
being sent to RCRA regulated landfills although under regulations EPA cannot
currently require the monitoring of some dioxins, although they are proposing
to do so (48 FR 14514). Table 4 is a list of some other hazardous substances
regulated under CERCLA which are not regulated or monitored under RCRA.
Table 4 was drawn up by reviewing the rules proposed under CERCLA on
May 25, 1983 (48 FR 23552). These rules propose "reportable quantities" for a
long list of hazardous substances. A reportable quantity (RQ) is that
quantity of a hazardous substance which if spilled must be reported to the
National Response Center (CERCLA §103) so that, among other things, a
determination can be made if any response under CERCLA is necessary. RQ's are
based on six criteria, i.e., aquatic toxlcity, mammalian toxicity,
ignitability, reactivity, acute toxicity, and carcinogenicity. They are in
five reporting levels: *, 10, 100, 1000, and 5000 pounds. The lower the RQ
the more hazardous the substance is supposed to be.
Table 4 lists those hazardous substances which have proposed RQ's in the
two most hazardous categories of 1 and 10 pounds and which are not regulated
under RCRA. The proposed rules do not indicate the basis of the rating for
each substance; therefore, it is possible that it is inappropriate to regulate
some of these hazardous substances under RCRA, but no discussion of this issue
has been found. Table 4 also shows the oral mammalian toxicity (in LD50)
where this information is available in the NIOSH registry (3).
-46-
-------
Table 4
SOME POLLUTANTS REGULATED UNDER CERCLA
(REPORTABLE QUANTITIES) BUT NOT UNDER RCRA
Pollutant
carbofuran
chlorpyrifos
diazinon
dichlone
alpha - endosulfan
beta - endosulfan
endosulfan sulfate
endrin aldehyde
guthion
mercaptodiraethur
nievinphos
naled
Proposed
Reportable
Quantity
(pounds) t
• 10
1
10
1
1
1
1
1
1
10
10
10
Oral (mammal) LD50*
(mg/kg) (23)
11
97
76
13
34
3.7
250
t 48 FR 23552-23595
LD-Q - Lethal Dose Fifty - a calculated dose of a substance which is
expected to cause the death of 50% of an entire defined experimental animal
population. It is measured in milligrams of substance ingested per kilogram
of animal body weight. For comparison purposes note that the oral toxicity
of iodine is 14,000 mg/kg, arsenic acid is 48 mg/kg, and potassium cyanide
is 10 mg/kg.
-47-
-------
The significance of table 4 is that these substances could be leaking
into groundwater from a RCRA permitted facility without violating the permit,
yet would be candidates for regulations under CERCLA. Even more to the point
is the fact that if these substances are spilled in transportation or
manufacturing operations in excess of their RQ, they must, under CERCLA, be
*•
cleaned up and disposed in a RCRA regulated facility where RCRA regulations
would not require their monitoring.
Table 5 addresses those contaminants of concern to CERCLA that are also
regulated under RCRA. In many cases, the groundwater detection levels are
higher under RCRA, as much as 1000 times higher. This is another example of
the puzzle that often occurs in comparing RCRA regulations with CERCLA. The
cure is more protective of public health than the prevention. Thus an EPA
RCRA regulated site may legally pollute groundwater to a level tolerated by
RCRA but come to the attention of CERCLA for the same pollution.
The last, and perhaps most important point in regard to tolerance levels
is that for many, perhaps even for most of the several hundred hazardous
constituents for which EPA has published test procedures for groundwater
monitoring samples (17), the level at which these contaminants can be detected
has not been published in reference 17 and has "not yet been determined by
-EPA. Although research is underway to determine detection levels, this
further confirms that considerations of human health did not play a major role
in determining the test protocols to use. Some of the hazardous constituents
for which EPA does not yet know the detection limits are listed in table 6.
The substances shown on this table were selected because they are alleged
carcinogens to which preliminary EPA research has given high hazard ratings.
Nevertheless, RCRA rules permit groundwater contamination by these substances
to an, undetermined level.
-48-
-------
Table 5
SOME EXAMPLES OF GROUNDWATER DETECTION LEVELS OF HAZARDOUS
CHEMICALS WHICH ARE HIGHER UNDER RCRA THAN UNDER CERCLA
CERCLA Detection RCRA Detection
Pollutant Levels (ng/1)(21,22) Levels (ng/1)
dieldrin .,5 2,500 (17)
DDT 10 . 4,700 (17)
DDE 5 5,600 (17)
ODD 10 2,800 (17)
heptachlor 5 1,900 (17)
heptachlor epoxide 5 2,200 (17)
aldrin 5 1,900 (17)
antimony 20,000 32,000 (63)
arsenic 10,000 53,000 (63)
cadmium 1,000 4,000.(63)
lead 5,000 42,000 (63)
selenium 2,000 75,000 (63)
thallium 10,000 40,000 (63)
•49-
-------
Table 6
SOME CARCINOGENIC CHEMICALS FOR WHICH EPA HAS NOT YET DETERMINED
THE LEVELS AT WHICH THEY CAN BE DETECTED IN GROUNDWATER
BY THE METHODS OF REFERENCE 17
aflotoxin •
4-aminobiphenyl
aziridine (ethyleneiraine)
bis-(chloromethy!)ether
chloromethyl methyl ether
1,2-dibromo-3-chloropropane (DBCP)
diethylnitrosamine (n-nitrosodiethylamine)
diethylstilbesterol*
dimethylaminoazobenzine
7,12-dimethylbenz(a)anthracene
dimethylcarbamoyl chloride
1,2-dimethylhydrazine
ethyl methanesulfonate
hydrazine
methylnitrosourea
nitrosomethylurethane (n-nitroso-n-methylurea)
n-nitosopiperidine
n-nitrosopyrrolidine
streptozotocin*
2,3,7,8-tetra$hlrodibenzo-p-dioxin (TCDD)
ethylene dibromide (EDB)
*Test methods not yet published by EPA as of January 19, 1984.
-50-
-------
In addition, the RCRA test procedures manual indicates that when several
chemicals are mixed together, as is usually the case in groundwater
monitoring, the ability to detect a specific chemical by a given test
procedure is reduced. These so called analytical interferences raise the
detection limits by an undetermined amount (17). It is clear that not being
able to detect carcinogens, which can be of concern at very low levels of
.contamination, as well as other hazardous materials, is not only dangerous to
human health, but increases the likelihood of CERCLA involvement.
The effects of this can be best illustrated with the example of ethylene
dibromide (EDB). EPA has recently cancelled the use of EDB as a fungicide
because of its carcinogenicity. In recent Congressional testimony, EPA's
pesticide program director, Edwin Johnson said (58):
. . . .we believe that the risks posed by EDB in drinking water at
levels in the low parts per billion are roughly comparable to the
risks posed by grain fumigation. In both cases we consider these
estimated risk levels to be unacceptable for a lifetime of
exposure. . . .According to our information, the State of Florida
has acted to provide alternative drinking water for approximately
500 wells found to contain EDB at or above 0.1 p.p.b. This appears
to be a responsible and effective way of dealing with these
potential risks. In short, the risks of EDB being reported in
Florida ground water (typically 1 to 20 p.p.b.) are probably similar
to risks posed by grain products. ...
EPA's Office of Solid Waste has indicated that the appropriate test
method for EDB in groundwater is the "GC/MS method for volatile organics"
which is test method number 8240 in reference 17. While this reference does
not list a detection level for EDB, it does list detection levels for 21 other
volatile organics. These range from 1.6 parts per billion to 7.2 parts per
billion. Furthermore, the text states that the table "lists detection limits
that can be obtained in waste waters in the absence of interferences.
Detection limits for a typical waste sample would be significantly higher."
-51-
-------
Therefore, the RCRA tolerance level for EDB could be from one to possibly
three orders of magnitude higher than the 0.1 parts per billion indicated as
"responsible" in the EPA testimony quoted above.
In summary, CERCLA is required to address releases of any "hazardous
.substance" which is defined as any substance designated under CERCLA and four
other acts administered by EPA. EPA has chosen to have RCRA regulate a much
narrower universe of substances and many of those are not regulated with the
same stringency as in other EPA programs. Therefore, compliance with a RCRA
permit will not necessarily be sufficient to prevent a site from becoming a
CERCLA site.
-52-
-------
MONITORING IN THE VADOSE ZONE •
EPA regulations for permitted facilities require that groundwater
detection monitoring wells be placed in the uppermost aquifer at the edge of
the waste disposal area (40 CFR 264.98(b)). Any contaminant detected by the
well may have first traveled anywhere from a few feet to several hundred feet
under the waste disposal area before it reaches the aquifer. Then the
contamination may have traveled anywhere from a few feet to several thousand
feet in the aquifer before it reached the well. Furthermore, if the leading
point of the plume of contamination is between two monitoring wells, it could
travel some distance past the wells before it is detected. Therefore, even if
a detection monitoring system works exactly as planned, there could still be
considerable environmental damage before the contamination may be detected in
a monitoring well.
The vadose zone is the ground above the uppermost aquifer. In humid
areas of the United States it is rarely over one hundred feet deep and is
usually much less. In arid western areas, however, the vadose zone can be
several hundred feet deep. Water and associated contaminants from a land
disposal facility will travel through the vadose zone to an aquifer at a rate
*
determined by the soil -characteristics, the depth of the vadose zone, the
" amount of fluids in the waste, and the amount of water. This can take
anywhere from a few months to many decades. P»F. Pratt, Chairman of the Soil
and Environmental Sciences Department at the University of California at
Riverside points out (44):
In irrigated agriculture we have estimates of water movement
and time required for water to move through the vadose zone.
For sandy soils having low water retention properties and
fairly large drainage volumes the time required to move
through 100 feet of the unsaturated zone is 10 to 20 years.
For clayey soils of higher water retention and lower drainage
-53-
-------
volumes the time is 40 to 60 years for 100 ft of the
unsaturated zone. The transit time is proportional to the
water retention properties of the soil material in the vadose
zone and inversely proportional to the amount of water that
leaves the surface zone (root zone in cropland or the storage
facility in case of a waste disposal facility). In irrigated
agriculture the drainage volume usually ranges from about 6 to
20 surface inches per year. If the leakage from a waste
disposal facility is of the same order of magnitude as in
irrigated agriculture the transit Limes will be of the same
order of magnitude. If the leakage is smaller the transit
time will be longer.
The significance of this fact is that by the time contamination is
discovered in a groundwater monitoring well, the vadose zone could have stored
significant amounts of contamination. Such toxic materials could continue to
pollute the groundwater for many decades even if disposal is halted and the
groundwater is initially cleaned up. Furthermore, the trend in regulatory
actions is to require land disposal facilities to be located in areas wich low
porosity clay soils preferably at great depth to groundwater. Such locations
postpone the time it will take the contamination to reach groundwater, but
also increase the amount of contamination stored in the vadose zone.
Not all contamination which reaches the aquifer is carried away by the
groundwater. Some contaminants may be adsorbed on solid surfaces or otherwise
•
contained in the aquifer and only gradually released or desorbed in small
amounts to pollute the groundwater. Professor John Cherry cites one example
for such materials as paint thinners, pesticides and PCB's (45):
These dense halogenated immiscible hydrocarbons currently pose
many intractable problems pertaining to subsurface contaminant
evaluation and prediction. They are so dense that in some
situations, irrespective of the directions of groundwater flow
or water table configuration, they can move downward or
laterally along paths of least resistance offered by granular
beds or fractures. While this movement occurs and after it
occurs, the immiscible liquid yields toxic dissolved
contaminants to the groundwater. The dissolved contaminants
are then transported by the groundwater in directions and at
-54-
-------
rates that may have no relation to the flow of the immiscible
liquid.
At some waste disposal sites, it is suspected that dense
halogenated hydrocarbon liquids have moved downward and have
settled as pools on top of impermeable beds in dead-end
fractures. The pools would then act as a long-term source
providing dissolved hydrocarbons to the flowing groundwater.
Scenarios can be envisioned whereby isolated zones of
immiscible liquids exist at considerable depth below the waste
disposal site and locally contribute hazardous concentrations
of dissolved contaminants to the groundwater. Because of low
solubility these contaminant sources could persist for
hundreds of thousands of years. They would be difficult or
impossible to detect using normal monitoring networks. They
could produce unpreditable small-scale contaminant plumes. In
some circumstances, numerous little pools or zones of
immiscible liquids from numerous leaky drums in a landfill
"could result in a rather chaotic pattern of input of
halogenated hydrocarbons to the groundwater flow system.
Thus, by the time contamination is detected in groundwater (if it is
detected), there may have been significant contamination of the vadose zone
and the aquifer which can continue to slowly re-enter the groundwater even
after it is initially cleaned up.
It is seen from the previous discussion how useful it would be to detect
leachate contamination in the vadose zone beneath a hazardous waste disposal
site before it reaches groundwater. Groundwater cleanup costs and alternative
water supply costs might be avoided and human health and the environment
better protected. EPA does require vadose zone monitoring for land treatment
of hazardous wastes* in the standards for EPA permitted facilities of July 26,
1982. The preamble to the regulations states that "EPA believes that adequate
technology and expertise is available to develop effective and reliable
systems." (47 FR 32329) Yet in the same regulations vadose zone monitoring is
*This method is used for less than one percent of wastes land disposed; also
known as land spreading or land farming of wastes.
-55-
-------
not required for landfills, surface impoundments and waste piles where the
need and the benefits would appear to be far greater.
The technology for which there is the most experience in waste disposal
monitoring in the vadose zone is the suction lysimeter, a porous ceramic cup
^placed in the vadose zone to collect a sample of the fluids there. In-the
interim status standards for existing land disposal facilities, EPA rejected
the use of lysimeters with this explanation in the preamble of May 19, 1980
(45 FR 33191):
,Available leachate monitoring technology generally involves
the placement of probes (lysimeters) beneath the disposal
facility. Since each probe is not generally capable of
monitoring a large area, many of them would have to be placed
under a facility in order to detect a localized flaw in the
landfill design. It may not be possible to place such devices
below an existing landfill or surface impoundment without
completely removing the waste and redesigning the facility.
Moreover, once such a system is in place, the probes tend to
fail over time due to deterioration or plugging. It is
difficult to determine when such a failure occurs and, if
discovered, the damage is generally irreparable. Under these
circumstances EPA does not believe that leachate monitoring
should be a general requirement for landfills and surface
impoundments during interim status.
Other commentors have pointed out that lysimeters do not work well in sub-
freezing or conditions of low soil moisture (50-) or very hot and dry
•
conditions (49).
Upon close examination, many of these points do not stand up. The first
point, that the "probe is not generally capable of monitoring a large area" is
contradicted by field experience. At a recent conference on vadose zone
monitoring a paper was presented which indicated that a suction lysimeter
located 10 feet below an impoundment could measure a distance of 10 to 30 feet
laterally (61). Secondly, placing suction lysimeters under existing land
-56-
-------
disposal sites can and has been done by the simple technique of drilling at a
slant. Thirdly, the plugging problem can be largely overcome by packing the
sampler with silica flour (68), a standard technique which even appears in EPA
manuals (69). Fourthly, the statement that the "damage is generally
irreparable" is unclear since what has been placed ought to be replaceable.
As for the other comments, it is largely irrelevant that lysimeters do
not work well in conditions of freezing or low soil moisture since these are
not conditions in which there would generally be leachate. And as for hot and
dry conditions, as pointed out later, vadose zone monitoring is currently
being conducted in Beatty, Nevada. In any event, it is not necessary that
lysimeters work perfectly (no technology does) or that they be convenient to
use. The important point is whether they are cost-effective in reducing
groundwater cleanup costs.
Lysimeters have been used for many years for monitoring land disposal
sites. At least one state, Texas, uses them for regulatory monitoring (51).
Wisconsin has been requiring vadose zone monitoring since the mid 70*s and
there are currently 19 hazardous waste sites in that state with either suction
lysimeters or collection *Lysimeters (64). California has proposed regulations
',
which would require vadose zone monitoring in new installations.
The United States Geological Survey has installed suction lysimeters
(albeit, not without .difficulty) at two existing low level nuclear waste
landfills. This research projected was started by USGS in.1981 (67).
A two year study of three sanitary landfills by Thomas M. Johnson of the
Illinois State Geological Survey (52) placed lysimeters under the existing
landfills; he found that all three had contamination in the vadose zone which
-57-
-------
had not been detected by groundwater monitoring wells. In one site the
lysiraeters showed that a clay liner had been ruptured and in another site the
lysimeter monitoring showed that contamination detected by a monitoring well
was coming from a different site. The Illinois researchers did not experience
the difficulties reported by EPA.
There is also field experience with geophysical vadose zone monitoring
techniques. A commercial hazardous waste disposal facility in Oregon uses a
vadose zone monitoring system which "integrates lysimeters, dual purpose
tensiometers/lysimeter units, and geophysical arrays to provide an early
warning leak detection and sampling system." (61) A firm in Las Vegas has
installed three resistivity grids since 1980 at hazardous waste lagoons, and
they are all reported to be working well (65).
Keros Cartwright of the Illinois State Geological Survey points out that
(48):
Numerous techniques have been developed for monitoring the
movement and quality of water in the unsaturated 'zone,
including tensiometers, soil moisture blocks, and neutron
logging techniques to monitor soil water content and
pressure. Water quality is generally monitored by soil
sampling using »porous ceramic cups similar to those used in
tensiometers. Whereas soil sampling requires repeated
drilling for extended analyses, soil water sampling using
suction or pressure vacuum lysimeters allows repeated sample
collection.
The usefulness of soil water samplers for monitoring soil
water quality in the vicinity of waste disposal sites has been
demonstrated by several workers in Pennsylvania (Apgar and
Largmuil, 1975; Parizek et al., 1975; Parizek and Lane, 1970;
and Johnson and Cartwright, 1980). More recent applications,
incorporating additional refinements, include the use of
pressure vacuum lysimeters to monitor soil water quality at
depths to 33 meters (108 ft) beneath artificial recharge sites
in Texas (Wood, 1970) and, in a study contemporaneous with
this one, three landfills in Wisconsin were instrumented
(Gerhardt, 1977) to generate data on the attenuation of
leachate in the unsaturated zone.
-58-
-------
As Cartwright points out, there is a fair amount of literature evaluating
the many techniques available for monitoring in the vadose zone for both new
and existing land disposal facilities. In particular, in 1980 L.G. Wilson of
the University of Arizona Water Resources Research Center reviewed a number of
*•
techniques for vadose zone monitoring below waste disposal sites for EPA
.(42). See table 7. Many of these are commercially available and are in
common use. Another survey of state-of-the-art techniques and techniques
under research or development which are capable of localizing a liner leaks
was made for the EPA Cincinnati laboratory (46). Table 8 lists the
technologies evaluated in this study.
Vadose zone monitoring techniques are not generally easy to use nor are
they inexpensive. No one technique is universally applicable and to get a
reasonable assurance of detecting leachate, several of them may have to be
used at any given site. However, as discussed previously, the techniques for
groundwater monitoring are also difficult, fallible and expensive. The cost
of cleaning groundwater which is often in the tens of millions of dollars, is
proportional to the amount of contamination. Thus, even if the technology for
vadose zone monitoring is more difficult and less reliable than groundwater
monitoring there are substantial benefits from early detection of pollution in
the vadose zone.
EPA, in rejecting the use of vadose zone monitoring in 1982 (47),
referred to the work of Wilson but only discussed one of the 26 techniques he
evaluated, the suction lysimeter. This technique was rejected largely because
of cost, although no analysis was made of the trade-off of avoiding the cost
of cleaning the contaminated groundwater. There was apparently no review made
of the many applications of vadose zone monitoring.
-59-
-------
Table 7
TECHNIQUES REVIEWED BY WILSON (42) FOR MONITORING IN THE VADOSE ZONE
Techniques for observing storage changes in the vadose zone at a waste
disposal site;
o Monitoring the spatial distribution of water levels in wells to
delineate the areal thickness of the vadose zone.
«.
o The gamma ray attenuation method to characterize bulk density and water
content of vadose zone sediments
o The neutron moderation method for defining the water content
distribution in the vadose zone
o Tensiometers for estimating water content at discrete points in the
vadose zone
o Electrical resistance blocks for estimating water content at discrete
points
Methods for monitoring water movement (flux) and associated parameters in the
vadose zone;
o Estimating infiltration rates by infilotrometers and test ponds
o Characterizing the quantity of water moving beneath the soil zone using
the water balance approach
o Determining the direction of unsaturated water movement and associated
hydraulic gradients using tensiometers, psychrometers, and che neutron
moderation method
o Measuring the unsaturated flux of water by adapting laboratory
techniques to the field, using water content profiles, estimating from
suction cup response, and using direct techniques such as flow meters
r
o Determining saturated flow in perched groundwater zones using
piezometers and observation wells
o Outlining techniques for determining the saturated hydraulic
conductivity in the soil zone and deeper vadose zone
Indirect methods for monitoring movement in the vadose zone;
o The four-electrode method for soil salinity
o The EC probe for monitoring soil salinity
o The four-electrode conductivity cell for observing soil salinity
o The earth resistivity approach for delineating pollution plumes
-60-
-------
Table 7 (continued)
Techniques for solids sampling in the vadose zone for determination of
associated pollutants are reviewed.
Direct techniques for water sampling during unsaturated flow;
o Ceramic-type samplers (suction lysimeters and filter candles)
o Cellulose-acetate hollow-fiber filters
o Membrance filter samplers
Methods for sampling from shallow perched groundwater zones;
o Sampling tile drain outflow
o Collection pans and manifolds
o Wells
o Piezometers
o Multilevel samplers
o Groundwater profile samplers
Sampling from deeper perched groundwater;
o Collecting cascading water
o Installing special wells
-61-
-------
Table 8
SUMMARY OF CANDIDATE METHODOLOGIES FOR DETECTING
AND LOCALIZING LEAKS IN LANDFILL LINERS
FROM WALLER AND DAVIS (46)
Technique
Electric
Resistivity
SP
Electromagnetic
Low Frequency Electromagnetic
High Frequency Electromagnetic
Acoustic
Seismic
Acoustic Emission
For planned sites
TDR Grid
DC Grid
What is Measured in the Ground?
Resistance over a length vs. horizontal &
vertical position
Voltage generated by electrochemical actions
Conductivity vs. horizontal and vertical
position
Dielectric properties vs. horizontal and
vertical position
Elastic properties vs. horizontal and
vertical properties
Sounds emitted from fluid flow in soils
Dielectric properties vs. position on
transmission line
Change of resistance of a wire due to
corrosion caused by leak
-62-
-------
DELAYS IN STARTING CORRECTIVE ACTION
Under the Part 264 EPA standards for EPA permitted facilities in a
detection monitoring mode (40 CFR Part 264, Subpart F), if hazardous
constituents are detected by the groundwater monitoring system a "compliance
"monitoring" program must be instituted. This program consists of two parts.
First, the EPA permit writer will establish a "groundwater protection
standard" for the unit, which will be specified in the permit for the
facility. Second, a new groundwater monitoring program will be instituted to
determine whether the unit is in compliance with its groundwater protection
standard. This new program will consist of monitoring at the compliance
point, i.e. the edge of the disposal area, to detect any statistically
significant increase in the concentration levels of hazardous constituents
specified in the groundwater protection standards.
The groundwater protection standard includes the hazardous constituents
to be monitored or removed if necessary, the concentration limits for each
hazardous constituent that trigger corrective action, the "point of
compliance" for measuring concentration limits, and the compliance period.
The regulations require that the concentration limits be set at: the
background level of the constituent in the groundwater; or for any of the 16
hazardous constituents covered by the National Interim Primary Drinking Water
Regulations (see table 3), the maximum concentration limits for drinking water
established in these regulations, if the background level-of the constituents
is below this. The facility owner may ask for a variance to establish an
alternate concentration limit if he can demonstrate that the constituent will
not pose a substantial present or potential hazard to human health or the
environment as long as the alternate concentration limit is not exceeded.
-63-
-------
If the groundwater protection standard is exceeded, then still another
step, the "corrective action program" is instituted. The objective of of this
program is to bring the facility into compliance with the groundwater
protection standard by removing the hazardous waste constituents from the
groundwater or treating them in the aquifer. The regulations require that
t- f
corrective action measures be taken to clean up the plume of contamination
that has migrated beyond the compliance point but not beyond the property
boundary.
Earlier it was shown that even in a well designed and properly
functioning groundwater detection monitoring system, a long time, even
decades, could elapse before contamination from a leak from a hazardous waste
disposal site reached a detection monitoring well. However, because of the
structure of the EPA regulations, a long time could also elapse between the
time the contamination reaches a monitoring well and the time anything is done
about it. Table 9 shows a scenario where this elapsed time is over two
years. This example does not present a "worst case" scenario,. but simply
illustrates times required to work through the many steps prescribed by the
regulations.
Furthermore, it should be pointed out once again that the action required
- to be taken is that the plume of groundwater contamination be cleaned up from
the edge of the disposal area to the property line. There, is no requirement
to clean up the contamination beyond the property line; there - is no
requirement to find the source of the leak and to repair it; and there is no
requirement to cease disposal operations.
-64-
-------
Table 9
SCENARIO FOR INSTITUTING CORRECTIVE ACTION AT A RCRA
PERMITTED SITE IN DETECTION MONITORING
January 1, 1984
April 1, 1984
May 1, 1984
August 1, 1984
November 1, 1984
March 1, 1985
April 15, 1985
May 15, 1985
August 15, 1985
September 1, 1985
Contamination reaches groundwater detection monitoring
well.
Sample is drawn from monitoring well. Well must be
sampled semi-annually (40 C.F.R. 264.98(a)). Assume
average time to detect contamination is three months.
Determination is made that there is a statistically
significant increase over background. This
determination must be made "within a reasonable time
period" (264.98(g)(2)). Assume one month, however,
discussion in next section will show this is optimistic.
Submit request to EPA for permit modification to
establish compliance monitoring program. This must be
done within 90 days (264.98(h)(4). Include notice of
intent to seek a variance for alternate concentration
limits under part 264.98(b) (264.98(h)(4)(iv)).
Submit data to justify variance under part 264.94(b) for
every hazardous constituent identified under part
264.98(h)(2). This must be done within 180 days of the
time that a determination is made that there is a
statistically significant increase over background
(264.98(h)(5)(ii)(B)).
EPA rejects request for variance and issues draft
revised permit for compliance monitoring. No time limit
specified in the regulations. Assume it takes four
months for EPA to review the data and prepare a draft
permit. Notice is given for public comment.
•»
End public comment period. Regulations require 45 days
EPA issues revised permit. No time limit specified in
regulations. Assume it takes EPA one month to review
public comments and revise permit accordingly.
Compliance monitoring begins.
Submit request to EPA for permit modification to
establish corrective action program. This must be done
within 90 days (264.99(i)(2) and 270.14(c)).
Submit engineering feasibility plan for corrective
action program. This must be done within 180 days of
the time that the request for variance is rejected,
i.e., March 1, 1985. (264.98(h)(5)(ii)).
-65-
-------
December 1, 1985
January 15, 1986
February 15, 1986
EPA issues draft revised permit for corrective action.
No time limit specified in the regulations. Assume it
takes four months for EPA to review the data and prepare
a draft permit. Notice is given for public comment.
End public comment period.
Cl24.10(b)).
Regulations require 45 days
EPA issues revised permit. No time specified in the
regulations.' Assume it takes EPA one month to review
public comments and revise the permit. Corrective
action begins.
Total elapsed time: two years one and one half months not including delays
from statistical analysis.
-66-
-------
STATISTICAL ANALYSIS
In the previous discussion it was assumed that when contamination had
been found in a well, a finding of a statistically significant increase over
background levels would be made within one month. In fact this is very
"unlikely.
In sampling groundwater, there is considerable variability due -to factors
other than the introduction of waste related contamination. These include
such things as seasonal fluctuations, geochemical processes, perturbations
introduced by the monitoring well, contamination or other changes introduced
by the sampling technique, natural and non-waste contamination, variability in
chemical analysis, and a great many others. It is necessary to distinguish
changes in groundwater due to contamination from those due to random or
periodic effects. The EPA regulations for both Part 264 and Part 265 state
that when a sample of the groundwater is taken from a monitoring well and
analyzed for the required contaminants, that the results be compared with the
previously determined background levels to see if there is any "statistically
significant" increase in contamination (40 CFR 264.97(h) and 265.93(b)).
Statistical significance is determined by one of several mathematical formulas
approved by EPA.
There are four possible outcomes from such a calculation:
1. The test could indicate that groundwater is contaminated when in fact
it is not (false positive).
2. The test could indicate that groundwater is contaminated when in fact
it is (true positive).
-67-
-------
3. The cest could indicate that groundwater is not contaminated when in
fact it is (false negative).
4. The test could indicate that groundwater is not contaminated when in
fact it is not (true negative).
In designing a test for statistical significance one wishes, of course,
to minimize the false positives and the false negatives. This can be done by
increasing the sample size, i.e. by increasing the number of monitoring wells,
the frequency of sampling and the number of samples taken. But for a given
sample size, any test of statistical significance which reduces the
probability of false negatives also increases the probability of false
positives and vice versa.
There are two ways to design a test for statistical significance. One is
to decide in advance the probability of detecting groundwater contamination
one wishes to achieve (the probability of detection being one minus the
probability of a false negative). In this case the probability of a false
positive .will be a function of the sample size and the variability of the
data. Another way is to determine in advance the probability of a false
positive (called the level of significance) and allow those same factors to
determine the probability of detection. In the former case the probability of
.a false positive will not be known in advance and in the latter case the
probability of detecting contamination will not be known in advance. EPA has
chosen the latter approach.
The cost of a false positive could be several thousand dollars e.g. the
cost of additional sampling and testing to establish that there is actually no
contamination. The cost of a false negative, groundwater contamination which
has gone undetected, could be substantial: in the worst case, millions of
-68-
-------
dollars in additional clean up costs and increased threats to human health and;
the environment. And if the plume of contamination had passed the property
boundary or if the owner cannot afford the necessary corrective action, the
site would become a candidate for CERCLA action. Minimizing the occurrence of
false positives reduces the short range costs of disposal site operators but
this analysis found no mention in any EPA document of why this approach was
chosen over the other.
EPA proposed standards for monitoring interim status sites on December
18, 1978 (43 FR 58982) which proposed a statistical test with a probability of
false positives (the level of significance) of five percent. In the final
regulations for interim status sites of May 19, 1980, EPA decreased the
probability to one percent. But this increased the probability of false
negatives. In the preamble discussion of this change (45 FR 33195) it is
implied that the change was made because of industry concerns over the cost of
a false positive. There is no mention of an attempt to balance this against
the cost of false negatives borne by industry and the public.
In the regulations for EPA permitted sites published July 26, 1982, EPA
raised the probability *of false positives to five percent once again,
*
t
explaining (47 FR 32303):
EPA is fixing the level of significance for the Student's t-test at
0.05 for each parameter at each well. When-the Agency proposed this
significance level for interim status groundwater monitoring, it
received some criticism that this would produce too many
notifications of contamination where none had actually occurred.
EPA recognizes that this could be a problem, particularly when there
are many comparisons being made for different parameters and for
different wells. However, EPA is concerned that a lower
significance level would unduly compromise the ability to detect
contamination when it did, in fact, occur.
-69-
-------
EPA did not, however, raise the probability of false negatives from one to
five percent at the approximately 2000 existing interim status sites which, as
was mentioned before, may be leaking. No explanation was given for not
including interim status facilities in this decision.
Considerable effort has been expended by OTA to find any estimate by-EPA
of the probability of detecting groundwater contamination by this statistical
procedure. While EPA reports and background documents contain many
discussions and calculations of false positives, no estimate of a false
negative can be found. The only related material that has been found is a
study for EPA by JRB Associates (4) which was supposed to "estimate the 'false
positive' and 'false negative' probabilities for various statistical
procedures" (11). However, they estimated the probabilities of false
negatives for only one statistical procedure, and that one is not the one that
EPA uses for detection the reason for this is not given. However, since this
is the only estimate of detection probability found, a sample calculation is
presented in table 10.
Table 10 shows the probability of detection, i.e., the probability of
concluding that there i-» a statistically significant difference, when the
level of contamination of TOX in the detection wells is in fact double the
background level. It can be seen that the probability of detection in one
test is only nine percent and that even after five years of sampling twice a
year, the probability of detecting the contamination is only forty percent.
.The JRB study claims that this statistical procedure gives lower detection
probabilities than EPA's procedure. Attempts to ascertain from JRB and EPA
the significance of these results, in relation to the EPA statistical
procedures, have not been successful.
-70-
-------
Table 10
PROBABILITY OF DETECTING TOTAL ORGANIC
HALOGEN CONTAMINATION
NUMBER OF
TESTS YEAR
1 1
2
3 2
4
5 3
6
7 4
8
9 5
10
PROBABILITY OF DETECTION
9.0%
14.6
20.0
24.6
28.6
32.0
34.2
35.8
37.8
40.0
Assumptions:
background mean equals 20
monitoring well mean equals 40
averaged-replicate test is used at one percent level of significance
other parameters are average of data from 52 sites studied
Source: JRB Associates (4)
-71-
-------
COMPLIANCE MONITORING
The purpose of compliance monitoring at permitted facilities is to
determine the degree and extent of the groundwater contamination or at least
that part of it which is inside the disposal site property boundaries. This
••is especially important in designing and evaluating corrective actions. This
is a very difficult and expensive proposition as EPA has testified in
Congressional hearings held in 1980 (40):
In a typical case. . .determining the extent and severity of a plume
emanating from one single source in a shallow aquifer requires
dozens^of monitoring wells and hundreds of samples. It also takes a
great deal of time and several hundred thousand dollars. If the
geology is more complex or several potential contamination source
exist, the cost will be on the order of $0.5 million. In a case
where the aquifer is deep or surface features cannot help in
determining the hydrogeology, costs could soar to two or three
million dollars.
Here again, as with the placement of the wells for detection monitoring,
the science of hydrogeology enters but with the additional requirement to
model and predict underground contaminant flow. However, groundwater modeling
with an emphasis on the flow of contaminants and not merely water is not a
routinely available technique like well drilling or chemical analysis. Such
modeling is state-of-the-art scientific research generally carried out in
universities and a few companies. Even in those cases where modeling
groundwater flow is possible, predicting contaminant flow may still be very
difficult (see section on vadose zone) if possible at all (45). Groundwater
consultant David Miller pointed this out at Congressional hearings in 1982
(41):
Unlike detection monitoring, compliance monitoring with its
dependence on predictions of contaminant migration through the
subsurface may be beyond the current state-of-the-art of the
groundwater science. It is not presently possible to determine how
-72-
-------
thousands of Individual chemicals will react in the groundwater ::
environment or to confidently predict the aggregate effects of
numerous processes such as attenuation, dispersion, and diffusion.
A vast amount of field data would be required to develop a reliable
basis for such predictions.
It is frequently suggested that modeling could serve as an adequate
predictive tool -for this purpose. However, even detailed
investigations which might cost on the order of $250,000 to $500,000
per site may not provide enough data to develop a model to be used
. . in this capacity. Furthermore, a relatively successful model based
on adequate data can only be expected to yield results within an
order of magnitude of the actual situation. This level of accuracy
may not be acceptable when public health is at risk and critical
concentrations are measured in parts per billion.
The process of obtaining the data for predicting groundwater
. conditions, interpreting the information and making accurate
decisions to implement compliance monitoring is a scientific
endeavor. It can only be carried out in a confident manner by well
trained groundwater technicians. There is presently a severe
shortage of trained groundwater scientists in the public and private
sector, and it is doubtful that there is sufficient talent available
to work on more than a relatively small percentage of the existing
sites that would fall under the compliance monitoring aspects of the
new hazardous waste regulations.
Similar views were put forward by Professor John Cherry at the
aforementioned National Science Foundation Workshop (26):
The ability of hydrogeologists to determine the present position of
zones of migrating contaminants and to develop reliable predictions
of future contaminant migration and of the effects of proposed
remedial measures is 'critical to the task of evaluating the degree
of risk and the cost/benefit ratios of remedial action.
Unfortunately, the processes of contaminant migration are poorly
understood in all except the most simple hydrogeological conditions.
The study of contaminant migration processes in groundwater is in
its infancy. Because knowledge of the processes affecting
contamination migration in groundwater is. slight, the predictive
capabilities of current mathematical models for all except unusually
uniform grandular deposits or clayey, diffusion-controlled deposits
are inadequate or unknown.
EPA shares this opinion of the shortcomings of the science of modeling
and predicting contaminant flow when it comes to using such techniques to
evaluate the geological suitability of a site location. The preamble to the
regulations state (47 FR 32283):
-73-
-------
EPA wants to make sure that the issuance of a RCRA permit for a
facility means that a certain level of protection is provided and
that the public can be assured that the prescribed level of
protection will be achieved. The way to meet this objective is to
avoid regulatory schemes that principally rely on complicated
predictions about the long term fate, transport, and effect of
hazardous constituents in the environment. Such predictions are
often subject to. scientific uncertainties about the behavior of
particular constituents in the hydrogeologic environment and about
' the effects of those constituents on receptor populations.
However, the RCRA. permit writers manual in its instructions for
evaluating the design of a corrective action program takes a somewhat
different--"view of the capability of hydrogeology in predicting contaminant
flow (43):
On the basis of the proposed design, the applicant should also
provide an analysis that identifies the expected hydraulic impact of
the recovery system on groundwater flow at the site. This analysis
should include prediction of flow rates to wells and drains.
Predictions of groundwater flow patterns throughout the contaminated
areas, including the drawdowns and hydraulic gradients, that will be
established by the recovery system should be provided. On the basis
of predicted withdrawal rates, estimates should be provided for the
time required to exchange an amount of groundwater equivalent to
that originally contaminated.
The applicant will need to use either analytical solutions or
numerical (computer) models to provide these predictions of the
response of groundwaier on site to the proposed recovery system.
Where aquifer conditions are simple or can be easily simplified, the
use of analytical solutions will generally be most appropriate. If
the groundwater flow system is complex or irregular boundaries are
involved, the use of numerical models may be more appropriate.
To summarize, the requirement that compliance monitoring predict plume
movement is a regulatory requirement that depends on a technology which does
not really exist. As has been seen before, EPA puts more reliance on state-
of-the-art technology to clean up pollution than it does to prevent pollution.
-74-
-------
CORRECTIVE ACTION .
The RCRA corrective action regulations for permitted facilities call for
contaminated groundwater to be cleaned up to background levels. Since
background contaminant levels can be, and frequently are, at extremely low
•levels, the regulations require a technology which is capable of removing
contaminants to below the level of detection. Even more so than with
compliance monitoring, the corrective action requirements of RCRA are
requirements for a technology which does not really exist. This fact is
acknowledged by EPA in the preamble to the regulations requiring the
technology which states that "the technology of performing corrective action
is new. The Agency's and the regulated community's experience in conducting
remediation activities (beyond the feasibility study stage) is fairly limited
to date" (47 FR 32313). The standards are based on the hope that, technology
will become available in the future as stated in the preamble which says that
"The national experience with groundwater cleanup ... is relatively limited
at this time. EPA expects that over time, the state of knowledge about
groundwater cleanup measures will improve" (47 FR 32286).
The most comprehensive study of attempts to clean up sites where
groundwater had been polluted was made by EPA in 1980 (25). This was a study
of 169 hazardous waste sites requiring remedial action. Groundwater was
polluted at 110 sites. In most of these cases the groundwater supply was
abandoned and replaced by a pipeline to another source. In very few cases,
because of the high costs, was there any attempt made to clean up the
groundwater, and none were cleaned to background levels.
-75-
-------
Although there is little or no experience in restoring polluted
groundwater to zero detection levels, there is experience in attempting to
restore groundwater to some degree. It is difficult, very expensive and the
results have been mixed. Typically, treatment of a plume is considered
adequate when levels of volatile organics are at or below 100 ug/1 (18). It
is possible to have cleanup costs, for a single site of over a million dollars
a year for 20 or 30 years. Groundwater consultant Kenneth Schmidt summed up
his experiences in a recent paper (19).
Substantial efforts are now being made to reclaim polluted
groundwater. In the southwestern U.S., where highly prolific
alluvial aquifers are common, a number of problems can be
encountered when attempting to reclaim polluted groundwater. First,
many of the zones of polluted water are large—often in the range of
thousands or tens of thousands of acre-feet. This results in the
need to pump substantial amounts of water, which must then be
treated and/or disposed. Decades will be required to remove
polluted water in many situations. Second, pumpage of groundwater
for reclamation often has legal constraints. Third, land ownership
often present a formidable problem, because polluted zones
frequently extend beyond property controlled by the responsible
entity. Fourth, relatively deep water levels usually allow
substantial amounts of pollutants to be in the vadose zone, where
pumping is not effective. Fifth, pumping schemes are inherently
inefficient in heterogenous, non-isotropic alluvial aquifers, due to
inflow of unpolluted water during pumping. Because of the many
limitations of reclamation, groundwater quality management should
focus on aquifer protection.,
•
The regulations allow for two basic approaches for corrective action.
" The first is to pump out the contaminated groundwater. This is not always so
simple as pointed out by the American Petroleum Institute (28):
.... in very arid portions of the country, groundwaters are
generally located well below the ground surface. Therefore, it may
be extremely difficult, if not impossible, to pump such underground
waters. In complex geologic environments, contaminants may perch on
clay layers. In such circumstances, even if pumping of surrounding
waters were possible, such pumping would not succeed in bringing
contaminants to the surface. In addition, in these circumstances,
the depth of the contaminant layer may prohibit trenching to reach
the contaminants. . . .Shallow aquifers may not have sufficient
-76-
-------
waters to permit effective pumping. In addition, certain tight clay
formations may prohibit effecting pumping from shallow aquifers. In
these circumstances, if excavation is not possible, it is impossible
to remove all contaminants.
The EPA RCRA permit writers guide recognizes, these difficulties and
points out the technological approaches for handling them (18). Where there
is insufficient groundwater for .efficient pumping, then fresh water must.be
injected into the aquifer by injection wells so as to flush out the plume of
contamination. But the plume itself is the lesser problem.
.... in most cases compliance with the groundwater protection
standard will not be achieved after the removal of only an amount of
groundwater equivalent to that originally contaminated. Rather, the
removal of additional amounts of water, frequently many times in
excess of that originally contaminated, will be required to reduce
contaminant concentrations to acceptable levels. ... Many of the
hazardous constituents present in any plume of contamination
migrating from a hazardous waste management facility will likely be
subject to some amount of adsorption to the geologic materials on
site. ... as contaminated groundwater is removed from the
subsurface and replaced by water of lower contaminant
concentrations, contaminants' will desorb from subsurface solids and
establish new equilibrium concentrations of contaminants in the
groundwater. Thus, the process of restoring groundwater quality
will become a process, in most cases, of not only removing
contaminants originally present in groundwater but also of removing
contaminants adsorbed -to subsurface solids.
This describes a ve5y expensive process of pumping huge amounts of water
for many decades with no guarantee that it will ever achieve the EPA
• standard. The issue of whether EPA will insist on the full measure of
compliance with its standards when faced with such costs, becomes important.
In addressing such public concerns, an EPA official recently wrote "It may be
costly and take decades, but it can be done and under the regulations the
owner is required to undertake it." (29) However, EPA's instructions to
their permit writers are much less optimistic (32):
-77-
-------
.... the permit writer should also consider the relative costs of
these measures when determining the adequacy of flushing rates
predicted for proposed recovery systems. Increasing flushing by
increasing pumping rates and the number of wells, well points,
and/or drains will certainly increase the costs associate with the
recovery system. Similarly, requiring the use of injection wells
and/or increasing their number and rates of injection will increase
cost. In some cases, particularly as flushing rates become higher,
the cost of increasing flushing rates by requiring these design
changes will become disproportionally high relative to the
additional flushing achieved and the advantages gained.
Thus, the permit writer will need to balance a number of factors
when reviewing the adequacy of flushing rates expected from a
proposed recovery system.
The EPA permit writers guide also points out many problems which may be
encountered in attempting corrective action and it does not have solutions to
all of them. For example, the problem of cleaning up immiscible fluids,
mentioned earlier by Professor Cherry is poorly understood.
Experience in the recovery of separate layers of immiscibles is
currently limited and pertains almost exclusively to the recovery of
lenses of light hydrocarbons, most notably petroleum products,
floating on the surface of the water table (53). . . .Procedures for
the cleanup of dense, immiscible contaminants are even more poorly
documented and more experimental in nature than those for the
cleanup of light, immiscible, contaminants (54). . . .At the present
time the state of the art in monitoring immiscible fluid content is
imprecise. . . .The Agency is planning to develop further guidance
on this topic as the state of the art is advanced. (55)
•
Once the contaminated water is pumped out of the ground, something must
be done with it. One solution is to filter out the contaminants and return
the cleaned water to the aquifer. This has been tried at some CERCLA sites.
Table 11 shows some examples of the kind of levels of cleanup which can be
practically (albeit at great cost) achieved using the most commonly favored
techniques. Although impressive, these results are far from what would often
be background levels, or even what are generally accepted to be safe levels.
-78-
-------
Table 11
REMOVAL OF SELECTED SPECIFIC ORGANICS FROM GROUNDWATER
FROM ABSALON AND HOCKENBURY (16)
Process Effluent Concentration Range
Organic Compound
t-
phenol
toluene
benzene
ethyl Acetate
formaldehyde
aceton
methyl Ethyl Ketone
aniline
nitroaniline
methanol
isopropanol
isobutanol
methylene Chloride
trichloroethylene
1,1, 1-Trichloroethane
1,1, 2-Trichloroethane
tetrachlorethylene
nitrobenzene
Adsorption Stripping
<10
<100 <10
<50 <10
—
—
—
25,000
<10
50-100
15,000
10,000
40,000
<100 200
<10 5-10
<10 50
<10 ' 50
5-10 5-10
<10
Biological
10-50
10-50
10-100
10-20
50-100
10-20
10-20
10-50
10-50
10-50
10-50
10-50
<50
<10 ;
10-50
10-50
10-200
100-1000
Note: All values in ug/1 or ppb.
-79-
-------
A second technology which the RCRA groundwater protection standards allow-
for corrective action is "in situ" treatment. This is the introduction of
chemical or biological agents into the aquifer to react with and destroy the
hazardous constituents without pumping out the groundwater-. If anything,
there is even less known about these technologies than the previous ones^ as
the permit writers guide points cut (56):
.... to date in situ treatment has been applied in only limited
circumstances, and little experience is available that can be
related directly to the cleanup activities required in Part 264
corrective actions programs. ... In most cases, use of these
techniques will assume the character of a field experiment.
-80-
-------
FINANCIAL RESPONSIBILITY
An additional problem with compliance monitoring and even more so with
corrective action at permitted facilities is the question of assurance that
there will be funds available for the huge expenditures these programs
involve. A great many of the sites being cleaned up under CERCLA simply went
bankrupt when the costs of groundwater cleanup became greater than the
company's assets. EPA regulations are supposed to prevent this from happening
at RCRA regulated sites and to this end EPA regulations do require financial
assurance for closure costs and the costs of post-closure maintenance.
However, there are no financial assurance requirements for the very expensive
requirements of compliance monitoring and the even more expensive corrective
action. Therefore, when companies are faced with these huge costs, some may
chose bankruptcy and the costs will be borne by CERCLA as they have been in
the past.
Because pollution will not be detected in the vadose zone and because
corrective action may not . begin promptly, greater build-up of groundwater
contamination may occur. Since the cost of groundwater clean-up is roughly
proportional to the quanu^ty of groundwater polluted (57), these delays built
into the regulations increase the cost of clean-up and enhance the probability
" that the site owners will not be able to afford the clean-up costs and that
the sites will have to be cleaned up with CERCLA funds.
-81-
-------
REFERENCES
1. UtS. Environmental Protection Agency, Hazardous Waste Site Descriptions;
National Priorities List, Final Rule, and Proposed Update, (Washington,
DC Office of Solid Waste and Emergency Response, EPA, August 1983).
2. U.S. Congress, Office of Technology Assessment, Technologies and
Management Strategies for Hazardous Waste Control. (Washington, DC: U.S.
Government Printing Office, March 1983).
3. U.S. Environmental Protection Agency, "Summary Report on RCRA Activities -
January 1984," (Washington, DC: Office of Solid Waste, EPA, January 1984).
4. JRB Associates, "Evaluation of Statistical Procedures for Groundwater
Monitoring," (paper submitted to U.S. Environmental Protection Agency,
under contract no. 68-03-3113, Dec. 22, 1983).
5. GeoTrans, Inc., "RCRA Permit Writers Manual, Ground-water Protection, 40
CFR Part 264 Subpart F," (submitted to U.S. Environmental Protection
Agency under contract no. 68-01-6464, October 4, 1983), p. 16.
6. U.S. General Accounting Office, Interim Report on Inspection, Enforcement,
and Permitting Activities at Hazardous Waste Facilities, GAO/RCED-83-241,
September 21, 1983.
7. National Materials Advisory Board, "Management of Hazardous Industrial
Wastes: Research and Development Needs," NMAB-398, (Washington, - DC:
National Academy Press, 1983).
8. Pesticide & Toxic Chemical News, 12, 4, (January 11, 1984), p. 15.
9. Private communication, EPA computer printout from the "Hazardous Waste
Data Management System" provided by Jeffrey Tumarkin, June 19,-1983.
10. Private communication with EPA's Lee Daneker, January 16, 1984.
11. U.S. Environmental Protection Agency, Ground-water Monitoring Guidance for
Owners and Operators of Interim Status Facilities. SW-963, (Washington,
DC: Office of Solid Waste and Emergency Response, EPA, March 1983).
12. U.S. Environmental Protection Agency, The Prevalence of Subsurface
Migration of Hazardous Chemical Substances at Selected Industrial Waste
Land Disposal Sites. SW-634, (Washington, DC: Office of Solid Waste, EPA,
1977).
13. Reference 5, page 192.
14. G. Fred Lee and R. Anne Jones, "Water Quality Monitoring at Hazardous
Waste Disposal Sites: Is Public Health Protection Possible Through
Monitoring Programs?" (paper presented at the Third National Symposium on
Aquifer Restoration and Groundwater Monitoring sponsored by the National
Water Well Association, Columbus, Ohio, May 1983).
-82-
-------
15. U.S. Environmental protection Agency, Scientific and Technical Assessment
Report on Cadmium, EPA-600/6-75-003, (Washington, DC: Office of Research
and Development, EPA, March 1975).
16. J.R. Absalon and M.R. Hockenbury, "Treatment Alternatives Evaluation for
Aquifer Restoration," (paper presented at the Third National Symposium on
Aquifer Restoration and Groundwater Monitoring sponsored by the National
Water Well Association, Columbus, Ohio, May 1983).
17. U.S. Environmental Protection Agency, Test Methods for Evaluating Solid
Waste, Physical/Chemical Methods, SW-846, 2nd edition, (Washington, DC:
Office of Solid Waste, EPA, 1982).
18. Reference 5, p. 235.
19. Kenneth D. Schmidt, "Limitations in Implementing Aquifer Reclamation
Schemes," (paper presented at the Third National Symposium on Aquifer
Restoration and Groundwater Monitoring sponsored by the National Water
Well Association, Columbus, Ohio, 1983).
20. Private communication with EPA RCRA/CERCLA hotline (Toney Baney), November
29, 1983.
21. U.S. Environmental Protection Agency, "Statement of Work, Organics
Analysis, Contract Laboratory Program," (Washington, DC: EPA,
September 1983).
22. U.S. Environmental Protection Agency, "Statement of Work, Inorganics
Analysis, Contract Laboratory Program," (Washington, DC: EPA, May 1982).
23. U.S. Department of Health and Human Services, Registry of Toxic Effects of
Chemical Substances, (Washington, DC: Public Health Service, Centers for
Disease Control, National Institute for Occupational Health -and Safety,
February 1982).
24. Private communication with EPA's Burnell Vincent, Oct. 21, 1983.
25. N. Neely,. D. Gillespie, F. Schauf, and J. Walsh, Remedial Actions at
Hazardous Waste Sites; Survey and Case Studies.' EPA 430/9-81-05, SW-910,
(Washington,DClOiland SpecialMaterials Control Division, EPA,
January 1981).
"
26. Reference 35, p. 144.
27. Princeton University Water Resources Program, Groundwater Contamination
from Hazardous Wastes, (Englewood Cliffs, NJ: Prentice-Hall, 1984).
28. "Comments on Interim Final Hazardous Waste Regulations Promulgated by the
United States Environmental Protection Agency Pursuant to Sections 3004
and 3005 of the Resource Conservation and Recovery Act, .Docket 3004,
Permitting Standards for Land Disposal Facilities", (Washington, DC: The
American Petroleum Institute, November 23, 1983).
-83-
-------
29. Lee M. Thomas, Assistant Administrator, U.S. Environmental Protection
Agency, letter to Senator Robert C. Byrd, December 30, 1983.
30. "Workshop on Groundwater Resources and Contamination in the United
States," (Washington, DC: Division of Policy Research and Analysis,
Advanced Technologies and Resources Policy Group, National Science
Foundation, March 14 and 15, 1983).
31. U.S. Environmental Protection Agency, "Resource Conservation and Recovery
. Act, Ground-water Monitoring Interim Status Regulations - §§265.90-94,
Evaluation of the Requirements, Phase II Report to OMB, Implementation of
the Requirements," (Washington, DC: Office of Solid Waste, EPA, March 10,
1983).
32. Reference 5, p. 237.
33. K. Cartwright, "Detecting and Monitoring Contaminated Groundwater,"
(printed in reference 30, p. 208).
34. Reference 30, p. 13.
35. J.A. Cherry, "Contaminant Migration in Groundwater with Emphasis on
Hazardous Waste Disposal," (printed in reference 30, p. 147).
36. J.A. Cherry, private communication, December 7, 1983.
37. Reference 5, p. 19.
38. Reference 5, p. 27.
39. U.S. Environmental Protection Agency, "Permit Writers Training Course on
Groundwat-er Monitoring, RCRA 264, Subpart F", (Washington, DC: Office of
Solid Waste, EPA, July 1983), p. 3-7.
40. Statement of Swep T. Davis, Associate Assistant Administrator for Water
and Waste Management, U.S. Environmental Protection Agency, before the
joint hearing of the Subcommittee on Health, and the Environment and the
Subcommittee on Transportation and Commerce, August 22, 1980.
t
41. Testimony by David W. Miller, Geraghty & Miller, Inc., before the House
Subcommittee on Natural Resources, Agriculture Research and Environment
Hearing, November 30, 1982.
42. L.G. Wilson, Monitoring in the Vadose Zone; . A Review of Technical
Elements and Methods." EPA-600/7-80-134, (Las Vegas, NV: Environmerial
Monitoring Systems Laboratory, U.S. Environmental Protection Agency, June
1980).
43. Reference 5, p. 233.
44. P.F. Pratt, University of California, Riverside, letter to Dwight Baier,
October 20, 1983, submitted to U.S. Environmental Protection Agency docket
for regulations of July 26, 1983 (47 FR 32274), docket No. PLDF II 043.
-84-
-------
45. Reference 35, p. 142. ::
46. M.J. Waller, J.L. Davis "Assessment of Techniques to Detect Landfill Liner
Failings," printed in Land Disposal of Hazardous Waste, EPA-600/9-82-002,
(Cincinnatti, OH: Municipal Environmental Research Laboratory, U.S.
Environmental Protection Agency, March 1982), p. 239.
47. U.S. Environmental Protection Agency, "Summary and Analysis of Comments
(40 CFR Part 264, Subparts F, K, L, M and N)," (Washington, DC: Office of
Solid Waste, EPA, July 9, 1982), p. 72.
48. Reference 33, p. 209.
49. Terry L. Thoem, Conoco Inc., letter to U.S. Environmental Protection
Agency docket for regulations of July 26, 1982 (47 FR 32274), docket No.
PLDF II 090.
50. Law Engineering Testing Co., "Lysimeter Evaluation Study", (Washington,
DC: American Petroleum Institute, 1983).
51. Reference 31, p. 27.
52. Thomas M. Johnson and Keros Cartwright, Monitoring of Leachate Migration
in the Unsaturated Zone in the Vicinity of Sanitary Landfills. Circular
514, (Urbana, IL: State Geological Survey Division, Illinois Institute of
Natural Resources, 1980).
53. Reference 5, p. 238.
54. Reference 5, p. 244.
55. Reference 5, p. 259.
56. Reference 5, p. 250.
57. John Ehrenfeld and Jeffrey Bass, Handbook for Evaluating Remedial Action
Technology Plans, ^EPA-600/2-83-076, (Cincinnati, OH: Municipal
Environmental Research Laboratory, U.S. Environmental Protection Agency,
August 1983), p. 33.
,58. Testimony of Edwin L. Johnson, Director of the Office of Pesticide
Programs, U.S. Environmental Protection Agency, before the Senate
Committee on Agriculture, Nutrition and Forestry, January 23, 1984.
59. Reference 5, p. 115.
.60* Inside EPA. 5, 7 (February 17, 1984), p. 3.
61. Robert D. Morrison, Kenneth A. Lepic, John A. Baker, "Vadose Zone
Monitoring at a Hazardous Waste Disposal Facility," (paper presented at
the conference on Characterization and Monitoring in the Vadose Zone
sponsored by the National Water Well Association, Las Vegas, Nevada,
December 8-10, 1983).
-85-
-------
62. Private communication with William Brown, Supervisor with the New Jersey
Bureau of Groundwater Discharge Permits, March 19, 1984. .:
63. Lee M. Thomas, Assistant Administrator, U.S. Environmental Protection
Agency, Memorandum to The Administrator proposing additional test methods
for reference 17, October 17, 1983.
64. Private communication with Peter Kmet of the Wisconsin Department of
Natural Resources,'March 20, 1984. ^
65. Private communication with Dr. Robert Kaufmann of Converse Consultants, ^ T
Las Vegas, Nevada, March 20, 1984. ' _
66. Private communication with Michael Nechvatal, Illinois Environmental
Protection Agency, March 23, 1984.
67. Private communication with Dr. John B. Robertson, U.S. Geological Survey,
March 23, 1984.
68. Private communication from Dr. L.G. Everett of Kamen Tempo, March 23,
1984.
69. L.G. Everett, L.G. Wilson and E.W. Hoylman, Vadose Zone Monitoring for
Hazardous Waste Sites, performed under contract no. 68-03-3090 for the
U.S. Environmental Protection Agency, (Santa Barbara, California, Kamen
Tempo), p. 5-63.
-86-
------- |