UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, o.c 20450
EPA-SAB-EEC-S9-035
September'20, 1989 QFF.^O*
fWt ADV>llMI$THA TOO
Honorable William K. Reilly
Administrator.
U.S. Environmental Protection Agency
401 M Street, S.W. 20460
Subject? Science Advisory Board's Review of the Office of Water's
Proposed Sludge Use And Disposal Regulation (40 CFR Parts 257 and
503)
Dear Mr. Reilly:
This report presents the EPA Science Advisory Board's (SAB)
review of the Office of Water's (OW) proposed Municipal Sewage
Sludge Incineration regulations, dated February 6, If8$. This SAB
review was requested by the OW to fulfill partially the EPA's
agreement with the court in which the Natural Resources Defense
council (NRDC) challenged the timing of the proposed regulations
(other aspects of the regulation involving land disposal and land
application were reviewed by the W170 Committee of the Department
of Agriculture).
The Subcouiiittee review consisted of a meeting in
Washington, D.C. on April 26-27, 1989, a teleconference on July
28, and extensive evaluation of written and oral briefing
Materials provided by the EPA staff, our examination focused on
five major issues raised by the OW, as well as an evaluation of
the need for long-term research by which some of the generic
issues identified can be addressed to improve the scientific and
technical basis for future regulations.
The SutocQmmitte0 believes that incineration is a viable and
appropriate technology for the treatment of sewage sludge. In
reviewing"*l!ie proposed regulation however, the Subcommittee found
that the OW made a good effort to develop a risk-based regulation
for sewage sludge incinerators, but concluded that deficiencies
in the risk and emissions data bases and, particularly, the
analytical methodologies used did not permit the development of a
sound, technically based regulation. Many safety factors are
built into the various analyses; although each individual safety
factor appears reasonable, the multiplicative use of a series of
such factors (as with the definition of the Host Exposed
Individual) makes the final number overly conservative and
subject to misinterpretation.
-------
Statistical analyses of the limited data base for emissions
are flawed due to the practice of indiscriminate averaging over a
range of different types of eombustors, operating conditions, and
types of air pollution control devices,
We endorse the concept of using a stack gas measure of total
hydrocarbons emissions for monitoring sludge incineration and air
pollution control devices as a general performance indicator.
However, the proposed use of total hydrocarbons as a direct
indicator of risk is not supportable due to the lack of a direct
link between total hydrocarbons and the total spectrum of
organics which might be emitted from sewage sludge incinerators,
as well as the major risk assumptions made in developing the
standard.
The air transport dispersion models used in developing the
proposed regulations are approved EPA models with defined
guidelines for their use. However, the OW has discounted Agency
technical requirements in seeking to offer simplified ways of
using results from selected model calculations.
The enclosed report provides detailed discussion of all the
above issues, as well as recommendations for strengthening future
efforts. We are pleased to have had the opportunity to be of
service to the Agency, and look forward to your response on this
report.
S incerely,
Dr. '(lljiymoftd-CT Loehr, Chairman
Executive Committee
Science Advisory Board
u^u^
2&i
Mr. Richard A. Conway, Chgirman
Environmental Engineering Committee
Dr. Francis McMichael, Chairman
Municipal Sludge Incineration Subcommittee
cc H. Habicht
D.Barnes
-------
United States Office of the Administrator EPA-SAB-EEC-89-035
Environmental Protection Science Advisory Board September, 1989
Agency Washington, DC 20460
Report of the
Municipal Sludge
Incineration
Subcommittee
Review of Proposed
Sewage Sludge Incineration
Rules (40 CFR Parts 257
And 503)
-------
U. 8. ENVIRONMENTAL PROTECTION AGENCY
f
NOTICE
This report has been written as a part of the activities of
the Science Advisory Board, a public advisory group providing
extramural scientific information and advice to the Administrator
and other officials of the Environmental Protection Agency, The
Board is structured to provide balanced, expert assessment of
scientific matters related to problems facing the Agency. This
report has not been reviewed for approval by the Agency and,
hence, the contents of this report do not necessarily represent
the views and policies of the Environmental Protection Agency,
nor of other agencies in the Executive Branch of the Federal
government, nor does mention of trade names or commercial pro-
ducts constitute a recommendation for use.
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
SCIENCE ADVISORY BOARD
ENVIRONMENTAL ENGINEERING COMMITTEE t
MUNICIPAL SLUDGE INCINERATION SUBCOMMITTEE
MEMBERS AND CONSULTANTS
CHAIRMAN
Or. Francis McMichael
The Blenko Professor of Environmental Engineering
Department of Civil Engineering
Carnegie-Mellon University
MEMBERS and CONSULTANTS
Dr. Allen Hershkowits;
Senior Scientist
Natural Resources Defense Council
Dr. Robert Huggett
Professor of Marine Science
Virginia Institute of Marine Science
College of William and Mary
Dr. Kun-chieh Lee
Engineering Scientist
Union Carbide Corporation
Dr. Ishwar Muraka
Program Manager
Electric Power Research Institute
Dr. Adel sarofim
Professor of Chemical Engineering
Massachusetts Institute of Technology
Dr. Walter Shaub
Technical Director
Coalition on Resource Recovery
and The Environment
U.S. Conference of Mayors
Dr. Thomas T. Shen
New York state
Department of Environmental Conservation
Dr. w. Leigh Short*
Woodward-Clyde Consultants
-------
Mr. Charles Velzy
Velzy/Weston
Executive Secretary
Mr. Samuel Rondberg
Science Advisory Board
U.S. Environmental Protection Agency
401 M Street, s.w. - A101F
Washington D.C. 20460
Staff Secretary
Mrs. Marie Miller
Science Advisory Board
U.S. Environmental Protection Agency
401 M street, s.w. - AIOIF
Washington D.c. 20460
Chairman. Environmental Engineering Committee
Mr. Richard A. Conway
Senior Corporate Fellow
Union Carbide corporation
Director. Science Advisory Board
Dr. Donald G. Barnes
U.S. Environmental Protection Agency
401 M Street, s.w. - A101F
Washington D.C. 20460
*Corresponding Member
ill
-------
TABIiB OF CONTENTS
1. EXECUTIVE SUMMARY .................... v
I.I Air Dispersion Model selection and Use ^ ..... v
1.2 MBI Definition and Implications ...*...„.. vi
1.3 Assumptions on Efficiency of Air Pollution Control
Devices ..................... vi
1.4 Total Hydrocarbon Emissions Monitoring ...... vii
1.5 Requirements for Incinerator Operations ...... vii
1.6 Recommendations viii
2. BACKGROUND ....................... 1
2.1 Process ......... ... 1
2.2 Charge to The subcommittee 2
3. DETAILED DISCUSSION ... ........... 2
3.1 ISSUE is Air Dispersion Models .......... 2
3.2 ISSUE 2: Definition of The MEI 5
3.3 ISSUE 3; Assumptions RE Air Pollution Control
Devices ..............«.......£
3.4 ISSUE 4; Total Hydrocarbons As A Monitoring
Surrogate ............. 8
3.5 ISSUE s: Requirements foe Good Incinerator
i operations ........ ... 12
4. CONCLUSIONS ....................... 15
S. REFERENCES ....................... 17
6. GLOSSARY . ............. 18
iv
-------
1. EXECUTIVE SUMMARY
This report presents the EPA Science Advisory (Board's (SAB)
review of the Office of Water's (Off) proposed Municipal Sewage
Sludge Incineration regulations, dated February 6, 1989. This
SAB review was requested by the OW to fulfill partially the EPA's
agreement with the Court in which the Natural Resources Defense
Council (NRDC) challenged the proposed regulations. The SAB
Subcommittee has reviewed five major points pertaining to the
Sludge Incineration part of the regulations. The Subcommittee
also looked at the need for long-term research by which some of
the generic issues can be addressed to improve the scientific and
technical basis for future regulations. Another aspect of the
regulations (land disposal, land application) is being reviewed
by the W170 Committee of the Department of Agriculture.
The Subcommittee review consisted of a meeting in
Washington, D.c, on April 26-27, 1989 and extensive evaluation of
written and oral briefing materials provided by the EPA staff. ,
The Subcommittee also relied heavily on the recent draft Products
of Incomplete Combustion (PIC) Subcommittee report f Proposed
Controls for Hazardous Waste. Incinerators; Products of Incomplete
reference 5), along with the SAB's recent resolution
on the use of mathematical models (reference 6) , and an earlier
(1985) SAB report reviewing incineration (reference 9) .
The Subcommittee believes that incineration is a viable and
appropriate technology for the treatment of sewage sludge. In
reviewing the proposed regulation however, the Subcommittee finds
that the OW made a good effort to develop risk-based regulation
for sewage sludge incinerators, but concludes that the
deficiencies in the risk and emissions data bases and
methodologies used do not permit the development of a technically
based regulation. Major findings are summarized below.
1.1 Mr Dispersion Model selection and Use
The OW selected three EPA approved air dispersion/ transport
models which are well documented in Agency publications. The EPA
manuals for these models identify the scientific problems
associated with their misuse. Yet the Agency, in the proposed
regulations, ignores its own technical requirements for the use
of these models in order to simplify their application to the
risk-based approach to incineration regulation. The Subcommittee
found no scientific/technical evidence that dealing with
emissions from sewage sludge incinerators requires the use of
specifically designed models, and concludes that no new
simplified model is justified for these specific regulations.
-------
1,2 MEI Definition and implications
The Agency defines the MEI (Most Exposed lndi\^.dual—a
hypothetical construct) as one who lives at the point of maximum
ground level concentration, is exposed 24.hours per day, breathes
20 cubic meters of air per day, weighs 70 kg and is exposed for
a 70 year lifetime. The subcommittee finds that this definition
is not developed through scientific reasoning or facts. It is
highly unlikely that any such scenario is representative for any
sludge incineration site. The only justification for this
definition can be, traced to seeking simplicity for the purpose of
calculating risks. As indicated in the body .of this SAB review,
the Mil assumptions are extremely conservative and consequently
unrealistic. Not withstanding this observation, the definition
of the MEI, the simplified method of deriving air dispersion
factors, the assumption of q* values (the cancer potency value,
expressed as the slope of the dose-response curve in the low-dose
range—see reference 3), and the assumption of the detection .
level concentrations for chemicals never measured, when combined,
yield large uncertainties and bias the Agency risk calculations.
The uncertainties and multiplicative factors are so large that it
makes interpretation of the calculated risks unreasonable and
misleading. Nevertheless, the Agency uses the MEI exclusively
for estimation of cancer risks without addressing any other type
of adverse health effects.
1*3 Assumptions on Efficiency of Mr Pollution Control Devices
V
Most of the data regarding facilities used in the
development of the proposed regulations are based on old tests
employing analytical methods that were called into question by
the Subcommittee. More recent comprehensive EPA studies are also
of limited value because they represent old incinerators with old
air pollution control devices not representative of the current
state-of-the-art systems. The key issue is whether or not the
Agency's generalizations of control efficiencies made from the
available data base are suitable for the entire universe of
combustion systems and air pollution control devices. The
Subcommittee finds that flawed statistical analyses of these data
vitiate many of the conclusions reached by the Agency»
The subcommittee concludes that the average control
efficiencies as specified in the proposed regulations are not
appropriate due to indiscriminate averaging over a range of
different types of combustion devices, different operating
conditions, and different types of air pollution control devices.
In particular, the metals emissions will be dictated by the
design and operating conditions of the combustion system and air
vi
-------
pollution control device.
1.4 Total Hydrocarbon Emissions Monitoring
*
The Subcommittee reviewed the issue of using total
hydrocarbons (THC) monitoring as an appropriate surrogate for
total organics emissions and whether or not there is a scientific
basis to estimate risk from exposure to organic emissions using
measured THC values. For the following technical reasons, the
Subcommittee concludes that the proposed THC monitorii^ device,
the proposed limits on THC emissions level, and the basis chosen
for estimating the weighted potency of the organic emissions are
flawed and cannot be supported scientifically. The Subcommittee
further finds that the use of a staclc gas total hydrocarbon
monitor is a sound approach to measuring organics destruction
efficiency (this same conclusion was reached by the Pic
Subcommittee), but* it is necessary to differentiate between the
monitoring of unconditioned flue gas samples with a hot THC
analyzer equipped with a flame ionization detector (FID) and the
monitoring of pre-conditioned flue gas samples with a cold THC
analyzer. The hot FID device has not yet been demonstrated to be
successful or reliable for monitoring in a continuous operation
mode.
The Agency has proposed that total hydrocarbons can be
related to risk for exposure to all organics. The Agency
performed a statistical correlation analysis to relate total
hydrocarbon measurements to measurements of selected volatile and
semi-volatile organics. The data indicated that 90% of the
organics were unaccounted for; other work indicated that THC
measurements cannot be reliably correlated with specific or total
organics emissions (see the PIC Subcommittee draft report noted
earlier). The Subcommittee concludes that the current data and
correlation analyses do not support the use of THC as a direct
measure of risk at sludge incinerators.
1.5 Requirements for Incinerator Operations
The Subcommittee considered many guestions in evaluating the
technical basis for the regulations proposed for controlling
operations of the combustion and air pollution control device
that together constitute a sewage sludge incineration system.
The Agency made, in the proposed incineration regulations,
an effort to reduce costs by proposing to establish minimally
acceptable operating conditions, that if satisfied, will prevent
the need for performance compliance testing of sewage sludge
incineration facilities. The Subcommittee finds that there is
vii
-------
neither an adequate database nor adequate understanding of the
relationship between operating parameters and emissions to
support prescribing minimally acceptable conditions in lieu of
performance testing*
•
1.6 Recommendations
As a result of the findings described above, the
Subcommittee developed the following two groups of
recommendations for future work!
Group I—specific recommendations for incinerator systems:
1. Develop air dispersion factors with the defined
technical bounds of existing air models.
2. Garry out "multiple1* simulations, reflecting
"real world" incinerator sites, to develop a data
base sufficiently large to yield valid statistical
summaries of air dispersion factors.
3. Develop a probabilistic approach to replace
"constant value" definitions of the MEI.
4. Approach metals control removals issues
separately for combustion and pollution control
devices»
5. Collect the needed key data on emissions and
control device performance and apply the
appropriate statistical analysis techniques
necessary to develop sound, scientifically
defensible bases for this regulation.
6. Emphasize the need for better training of PQTW
operators,
7. Emphasize the need to develop performance-based
regulations which provide incentives for improving
incineration technology and pollution control
equipment.
viii
-------
Group II—recommendations for long-term research and
development:
1. Conduct additional model validation research,
in support of both the ow regulatory effort, and
the many other Agency programs utilising such
data.
2. Support epidemiological research to determine
the incidence of adverse health effects in
populations residing near existing incineration
facilities.
3* Emphasize the need for source reduction of
toxic metals in the municipal waste stream to
prevent such agents from reaching the incineration
feedstock.
2 . BACKGROUND
2.1
The !PA Office of Water, on February 6, 1989 , proposed a new
regulation on sewage sludge use and disposal (40 CFR Parts 257
•and 503) . Shortly thereafter, the Natural Resources Defense
Council (NRDC) challenged the proposed regulation and filed a
suit in court raising a variety of issues, in some preliminary
discussions with the NRDC and the Judge hearing the case, the
Office of Water agreed to provide to the Court external peer
reviews of a number of technical issues raised in the litigation.
In late February, Water Program Officials asked the SAB to
provide an a_d hoc, review of five major issues pertaining to the
sludge incineration portion of the sewage sludge regulation
(Another aspect of the proposed regulation, land application/land
disposal, is being reviewed by the W170 Committee of the
Department of Agriculture) .
The Subcommittee met in Washington, D.C, on April 26-27,
1989. Members had been supplied with extensive documentation
prior to the meeting, and received detailed briefings from EPA
staff at the meeting. In addition, the subcommittee relied
heavily on the draft PIC Subcommittee report, Proposed_Control s
for Hazardous Waste Incinerators; Products of IncjomSlete
Combustion, the SAB's recent resolution on mathematical models
(EFA-SAB-EEC 89-012, January 1989) , and an earlier SAB review of
incineration (Report on the Incineration of Liguid Hazardous
Wastes, April 1985) .
-------
2.2 Charge to The Subcommittee
*
In a memorandum dated February 28, 1989, the' Office of water
identified five specific issues for SAB review:
1* Are appropriate and correct air dispersion models
used to relate stacK emissions to ground level
concentrations for the most exposed individual (MEI) ?
2, Is the MEI defined properly ?
3, Are the assumptions concerning the efficiency of air
pollution control devices correct and appropriate ?
4. Is it appropriate to use total hydrocarbon emissions
as a monitoring surrogate for total organics emitted ?
5. Are the requirements for good incinerator operations
properly stated ?
3. DETAILED DISCUSSION
3.1 ISSUE 1: Mr Dispersion Models
*
Use of air dispersion models is a long-standing practice.
Numerous simple-to-complex models are used routinely by the
Agency and the regulated industries to assess compliance with air
emissions and air quality standards. Accuracy and precision of
predictions from these models depend heavily on data concerning
meteorological conditions, source composition, emission rates,
velocity at the emission point, and physical configuration
inputs.
Dispersion values and ground level concentrations calculated
by these air dispersion models change as the inputs for the
meteorological conditions, emission rates, emission velocities,
and staclc heights are changed. Topography surrounding a source
plays an important role in estimating expected concentrations at
ground level,
The EPA chose three approved air dispersion models. The
ISCLT model is constructed to determine air concentrations for
urban and rural situations where terrain elevations clo not exceed
stack height. The LONG2 and COMPLEX-1 models are designed for
use in complex terrain situations. These models were recently
-------
evaluated by the ,PIC Subcommittee of the SAB arul were found to be
scientifically sound to calculate ground level concentrations
within the half an order-of-magnitude precision range. The
models are scientifically supportable, but require reliable input
data, which are not easy to obtain. Many potential users may not
be familiar with the assumptions and uncertainties involved with
the models, and indiscriminate use of the models and annual
average values can be misleading.
This Subcommittee chose not to conduct another in-depth
review of these models; because several members of the
Subcommittee also served on the PIC Subcommittee and were
familiar with its conclusions, the majority of the Incineration
Subcommittee accepted the findings of the PIC Subcommittee on air
models as reported in the PIC draft report referenced earlier
(see page v, and also reference 5}.
The Agency chose to use the three models to calculate
initially the maximum ground-level concentrations for ten
existing sewage sludge incinerators. EPA then conducted a
sensitivity analysis of the ISCLT model to determine the
feasibility of developing a simplified model for obtaining
dispersion factors. The analyses were designed to evaluate the
sensitivity of the following parameters: incinerator stack
design, including stack height and emission temperatures;
meteorological conditions; and location of the HEX.
After these limited sensitivity analyses were completed, the
Agency developed the regression coefficient for the relationship
between dispersion factors and stack height for the reasonable
worst case facility. The meteorology and building dimensions
were held constant while stack height was varied.
The Agency thus used one model (i.e., ISCLT) with only one
meteorological input datum and then developed a simplified model
which used stack height as the only variable to derive dispersion
factors. This new simplified model was subsequently used in an
additional conservative mode to arrive at the dispersion values
for the incineration regulations. The Subcommittee observes that
emission rates , emission velocities, and transformation
processes (wet and dry deposition and chemical transformation),
all important variables, were not considered in obtaining air
dispersion factors. The subcommittee has the following
additional comments and suggestions concerning the use of these
models in the proposed regulations;
1, The three models noted above are all EPA approved
models. Each was formulated for use under its own
specific set of topographic conditions. The Agency
publications on these models specifically recognize the
inaccuracies and imprecisions that result when a model
-------
designed for flat terrain is used to predict ground
level concentrations for hilly terrain. Although the
EPA manuals for model use identify these scientific
problems associated with misuse, the Agency, in
developing the proposed regulations, ignores its own
technical requirements and chooses to use* "maximum'1
predicted values from any of the three"models,
regardless of their applicability. The Subcommittee
recommends that the Agency use these models to arrive
at "dispersion factors" within the bounds defined in
their technical reports. There is insufficient
scientific evidence that dealing with emissions from
sewage sludge incinerators requires specifically
designed models, and the Subcommittee concludes that no
new simplified air transport model is needed for these
specific regulations,
2. Initially, the Agency selected three approved models
to estimate air dispersion factors? but in the final
regulation development, the Agency used "stack height"
as the only variable to calculate these factors. This
approach is misleading in that it gives the impression
that EPA approved air dispersion models were used
appropriately, when in fact, "dispersion factors11 were
calculated for only one meteorological and site
condition with only the stack height varied for the
purpose of dispersion calculations. No matter how
scientifically sound the construct of a mathematical
model may be, the output or the prediction will be no
more accurate than the input information. This aspect
has not been given enough attention by the Agency, We
also comment on the unscientific and poor practice of
reporting calculated values to many more significant
figures than can be technically justified.
Since these models provide a crucial numerical input
for further regulatory calculations, the Subcommittee
recommends that "multiple" simulations be done where
data from several different sites are used as input to
represent the many locations of incinerators in the
country. These simulations would represent a
scientifically sound, "realistically large" number of
outcomes that can be statistically summarized to give
estimated "most probable,11 "x-th percentile level,11
"mean," and "median" values for the dispersion factors.
3, The Subcommittee observes that validation of a given
model's ability to predict ground level concentrations
of organic compounds, or the back-calculated
permissible stack emission levels that would be
protective of the MEI has not yet been achieved by the
Agency. In a recent SAB resolution on modeling
-------
(referenced above), it was pointed that the accuracy
and precision of many models remains largely unknown.
The Subcommittee believes that, in the long run, the
Agency would be well served to support research on
"model validation11 (a generic issue) in qcder to reduce
the prediction uncertainties which may -be unnecessarily
costing our society large sums of money.
3.2 ISSUE 2: Definition of The MET
The Agency defines the MEI as a hypothetical construct who
lives at the point of maximum ground level concentration, is
exposed 24 hours per day, breathes 20 cubic meters of air per
day, weighs 70 kg, and is exposed for a 70 year lifetime. These
assumptions are best described as a policy choice rather than one
that is developed by scientific reasoning or facts*
%
For example, we know of no evidence to show that such an Mil
exists at any current or future sewage incineration site or come
close to living outdoors continuously 24 hours a day for 70 years
at the point of maximum ground level concentration. If one looks
at the census information, it can be easily seen that at least 10
hours per day are spent elsewhere by any individual living in any
geographical area, similarly if we use human life cycle data, it
is obvious that human body weight is a function- of time with
childhood mass increasing from 3 kg at birth to 35 kg in the
first 10 years of growth. Additionally, because of shifts in
meteorological conditions, the maximum ground level concentration
can not coincide with the MEI for 24-hours per day for 70 years,
The only justification for choosing a "constant value" for all
the factors used to define the MEI could be traced to seeking
simplicity in calculating risks,
The Agency has taken an overly conservative approach in the
definition of the MEI, The selected scenario is not
representative of sludge incinerator sites. The previously
mentioned SAB Subcommittee which reviewed PICs also addressed the
MEI issue. Again, because some members of our Subcommittee
participated in that review, we did not initiate any new
investigations to cover the same work* With the exception of one
member, the Subcommittee supports the comments and conclusions
presented in the PIC Subcommittee report.
Finally, the 1985 report noted above (reference 9)
summarized the issue of determining exposure to incinerator
emissions with a statement which, unfortunately, is still largely
applicable? MThe efforts of the Agency to assess such exposures
have been inadequate because they resulted from either individual
judgments or computer models without adequate laboratory or field
validation."
-------
The definition of the MEI, the simplified method of deriving
air transport dispersion factors, the assumptions of q* values,
and the assumption of detection level concentrations for
chemicals never measured, when combined, yield largfe
uncertainties and bias in the calculations of cancer rislcs. In
order to achieve a "reasonable worst case" for these calculations
the subcommittee suggests that it is scientifically more
appropriate to use a "probability distribution approach'1 to
represent variations of ground level concentrations as a function
of time and the definition of the MEI with appropriate functions
of time, space, and exposure durations. We recognize that
developing more scientifically defensible rislc calculations will
require more support of long-term, sustained -research within the
Agency,
3,3 ISSUE 3; Assumptions RE Air Pollution Control Devices
\
Most of the data used in the development of the proposed
regulations regarding incinerator facilities are based on old
tests employing analytical methods with questionable quality
control? these were called into question by the Subcommittee.
More recent comprehensive EPA sampling studies are also limited
in value because they represent old incinerators with old air
pollution control devices—not representative of the state-of-
the-art. Also, the statistical methods employed did not separate
the effect of differences in emissions from the incinerator from
the effects of differences in the efficiency of the control
device. Thus, the quality of the data base hampers the Agency's
ability to establish regulations.
To the extent that EPA's good efforts were able to obtain
any data at all, which were necessarily from technologically
substandard facilities, the Agency focused most heavily on metals
control. While metals emission control for certain volatile
elements is most efficient at lower temperatures (1200°-1600 F
to avoid volatilization), the destruction of organics can require
higher temperatures (>1SOO° F) , In principle, it is possible to
maintain a low solids temperature to minimize metals vaporization
and a high gas temperature to ensure high burnout efficiency of
the volatile^. To maximize simultaneously control efficiencies
through a single temperature limit for both metals and organics
is difficult.
The key issue is whether or not the Agency's generalizations
on control efficiencies made from the available data base is
suitable to represent the entire universe of combustion systems
and air pollution control devices. The Subcommittee concludes
that the generalization is flawed due to the practice of
indiscriminate averaging over a range of different types of
-------
combustion devices, different operating conditions, and different
types of air pollution control devices. In particular, the
metals emissions will be dictated by the design and operating
parameters of the combustion system and air pollution control
device, and it is not appropriate to average contrdl efficiencies
across the entire system without consideration of the individual
design and operating practices of the incinerators,
The mechanisms by which metals escape from the incinerator
are correctly recognized in the proposed regulation development,
However, the ultimate emissions of metals from the incinerator
system, including the air pollution control device, will be
dictated by a wide variety of other parameters that were not
correctly evaluated and considered in the proposed regulations.
Specifically, the ultimate size of metal particulates and
the partitioning of the metals as a function of size will be
dictated not only by the sewage sludge characteristics, but also
by the subsequent quench rate that occurs downstream of the
incinerator device. Volatile metals will condense and form
different types of particles depending on the metal species
formed in combustion and on the rate of quench. The resulting
size distribution will determine the capture of metals that will
occur within the particulate removal device. The temperature
within the air pollution control device will dictate the species
formed of the highly volatile metalsr for example, mercury and
lead can have forms that are highly volatile even at air
pollution control device temperatures. The temperature of the
air pollution control device will normally result in the
condensation of these metals on particulates and the subsequent
ability to remove material from the flue gas.
At a minimum, the subcommittee recommends that the
combustion device be considered separately from the air pollution
control device. The averaging of metals control efficiencies
over a wide variety of parameters as was done is clearly
misleading and can lead to misleading conclusions*
The development of control efficiencies for different metals
is a critical need and the date base must be broadened if such
generalizations to the universe of combustion devices and air
pollution control devices is to be accomplished, in addition,
the regulations must be accompanied by appropriate guidance to
permit writers to allow them to implement the regulations.
The parameters specified in the regulations fall into
several areas. One set of parameters are dictated when testing
will not be required for incinerators? a second set is dictated
when an incinerator is tested for metals emissions. The second
set is generally more flexible and based on operating conditions
during the test burn.
-------
If the average control efficiencies as specified in the
proposed regulations are not appropriate due to flaws in the
averaging process, then it becomes necessary for permit writers
to specify the appropriate operating conditions. The
establishment, within the regulations, of operating conditions
may not be the most appropriate means to construct \he
regulations. Having requirements that dictate the design and
operating conditions of incinerators removes the incentive for
the development of new combustion devices and can tie the hands
of permit writers who are dealing with special problems that were
not foreseen in the regulations.
The design of an air pollution control device is important
for emissions control, A venturi scrubber can remove some
submicron-sized particles (depending on the pressure drop), but
a spray impinger can not. If one does not discriminate removal
efficiency data with respect to equipment typef one can encounter
the types of misinformation in the regulations that relates
control efficiency^to excess air levels.
A more generalized regulation which specifies an emission
standard and suggests guidelines for the development of
appropriate permit conditions would provide the flexibility both.
for innovative process and permit development. Again, the
permit writer must be provided with very good guidance about the
impacts of design and operating conditions on the partitioning of
metals throughout the system.
With regard to the performance of air pollution control
devices, the proposed excess air limitation (12% oxygen
limitation for multiple hearth incinerators and 7% oxygen
limitation for fluidized bed incinerators) is unjustified. The
important factor is the air pollution control device design and
operation. For a venturi scrubber, a reduction in excess air may
result in lower pressure drop and lower scrubbing efficiency.
Total air flow through the air pollution control device is more
important than excess oxygen level in terms of air pollution
control device performance.
3.4 ISSUE 4s Total Hydrocarbons A3 A Monitoring Surrogate
The Subcommittee reviewed the issue of using total
hydrocarbons (TEC) monitoring as an appropriate surrogate for
total organics emissions and whether or not there is a scientific
basis to estimate risk from exposure to organic emissions using
measure THC values.
For a number of technical reasons, the Subcommittee
concludes that the proposed THC monitoring device, the proposed
limits on THC emissions levels, and the basis chosen for
3
-------
estimating the weighted potency of the organic emissions are
flawed and can not be supported scientifically. We support the
Agency's efforts to find a direct or indirect surrogate for
organics concentration in stack flue gas as a means of regulating
emissions, but find that additional work will be needed before
this goal can be put into practice.
The Subcommittee offers these specific observations about
using the THC level as a surrogate for organic emissions, the
monitoring device, and the role that they may play in sewage
incinerator emission controls
1. THC monitoring plays an important role in the
proposed regulations, as a means of. judging
the adequacy of the operation of a sewage sludge
incinerator. The draft PIC Committee report
recommends the use of THC monitoring "as a
measure of good combustion practice." The Agency
proposes -to use a FID (flame ionization detector)
to measure, on a continuous basis, THC as a
surrogate for assessing total organic emissions.
The Subcommittee endorses this concept for
monitoring; but because the accuracy and
reliability of such instruments have not been
ascertained under actual continuous operating
conditions, we conclude that the THC as a
surrogate is not technologically supportable at
the present time.
The key issue of the use of the total hydrocarbon
monitors for incineration systems is the long-term
operability. There is indeed instrumentation
available, based on the flame ionization detector
system, for continuously monitoring total hydrocarbons.
However, it has not been demonstrated that systems
using unconditioned hot stack gas samples can operate
continuously in the sewage sludge incinerator stack gas
environment. The Agency Office of Solid Waste has
recently recommended to permit writers that they
consider the use of pre-conditioned stack gas sampling
with a cold THC analyzer equipped with an FID for Tier
II hazardous waste incinerator stack gas (EPA
Memorandum, April 7, 1989, Guidance on PIC Controls for
Hazardous Waste Incinerators). A cold sampling train
FID eliminates many of the operating problems
associated with using hot systems for continuous
monitoring (reference 2). Given the state-of the-art
of continuous THC hot stack gas monitoring, we believe
that it is not appropriate to propose regulations prior
to a demonstration of feasibility.
2. The underlying assumption for the development of
-------
the proposed regulation for THC is that total
hydrocarbons, as measured, can be related with total
organic carbon, and that total organic carbon can be
related to risk. The approach to develop this
relationship is based on the available da%a on total
hydrocarbon emissions and measurement of target
organics by sampling and gas chroMatograph/mass
spectrometer (GC/MS) analysis.
The Agency performed a statistical correlation
analysis to relate FID total hydrocarbon measurements
with samples collected for GC/MS analysis of selected
volatile and semi-volatile organics. The correlation
coefficient determined was only for those species
selected for GC/MS measurement and does not include
all organics. In fact, the data indicated that 90% of
the organics were unaccounted for by this analysis.
Other work in the office of Air Quality Planning and
Standards, as well as work reported by Environment
Canada, have indicated that, at this time, total
hydrocarbon measurements by FID can not be reliably
correlated with specific or total organics emissions,
and additionally, that correlations between total
hydrocarbons and polychlorinated dioxins and furans,
compounds of concern, are poor (reference 1).
3. The Subcommittee concluded that the use of THC
measurements as indicators of incineration system
performance extends current technical practice. It is
clearly poor combustion practice to have high levels
of THC emissions. The state-of-the-art of waste
combustion should allow total hydrocarbon emissions to
be minimized. The data base from alternative types of
waste combustion systems generally reflects a total
hydrocarbon emissions level of less than 10 ppm
(corrected to 7% oxygen).
Most waste combustion systems have total hydrocarbon
emissions that are near the detection limit of
traditional FID systems. Nonetheless, the regulation,
as proposed, would allow hydrocarbon emissions well in
excess of the state-of-the-art of other waste
combustion devices. Allowable THC emissions would be
from 20, to well over 100, ppm for sewage sludge
incinerators, depending on stack height and other
operating variables. The data base currently available
for sewage sludge incinerators indicates that the
operating range is typically between zero and 30 ppm.
The standards proposed for THC would not additionally
restrict the performance of sewage sludge incinerators
and they would not force an improvement of technology,
10
-------
but would, in fact, allow a relaxation of the
appropriate operating conditions-
4. The THC measurements at best may indicate the
combined performance of combustion and air quality
control devices. How these measured concentrations at
the stack relate to environmental concentrations of
carcinogens remains unknown at this time. Because of a
few very large values of risk calculated by the Agency
for incinerator emissions, the Subcommittee looked
carefully at the several steps and assumptions in these
calculations and concluded that the accuracy and
precision of the Agency analysis is questionable.
The Agency used "detection limit" concentrations for
carcinogens to calculate the weighted q values (the
cancer potency value, expressed as the slope of the
dose-response curve in the low-dose range—see
reference 3)* There is no scientific basis for
choosing "these values. In fact, a weighted q* value of
8,7 x 10'4 [mg/kf/day]"1 and 3.5 X 10"4 [mg/kg/day]"1 for
diethylstilbestrol and 2-nitropropane respectively are
introduced by this arbitrary choice. The total
potency-sum of these two non-detected compounds is
nearly 10% of the calculated total potency for
carcinogens* The very limited data base from sludge
incineration also does not indicate the presence of
these compounds.
On the other hand, the §5% emission concentrations for
benzene, benzo(a)pyrene, chloroform, and formaldehyde
are found to be in "measurable" amounts giving rise to
weighted q* values of 1,6 X 1Q"4, 3.6 X 10 , 3.1 X
10~4, and 6.0 X 10"4 [Big/kg/day]"1, respectively. The
total potency, that is the sum of the q* values, from
these four compounds, is 4.67 X 10"3 [mg/kg/day]"1,
which is about the same as from the assumption of the
"detection limit" values for the two undetected
compounds cited above. It is our recommendation that
the scientific foundations for the measurements of
individual chemical carcinogens be better defined.
The apparently flawed risk estimation methodology used
by the Agency to develop these regulations leads to
forecasts of very high cancer risks associated with
sewage sludge incinerator technology. As indicated
earlier in this SAB review, the MEI assumptions, ^
simplified air dispersion factors, assumptions on q
values, and the assumption of detection level
concentrations for chemicals never measured, inherent
in these regulationsr are extremely conservative and
consequently unrealistic. The risk estimates are
11
-------
suspect. The Agency should either re-evaluate and
improve the basis for producing estimates of risk or
alternatively, if shown to be feasible, initiate
epidemiologieal studies to determine the incidence of
carcinomas or other adverse health effect^ in persons
residing around existing incinerator facilities. This
would be a direct test of the validity of the risk
estimates and hence the appropriateness of the data and
assumptions used in the calculations. This information
will not be obtainable in the short run.
5. Monitoring of emissions at the stack will at best.
provide data on how well the incineration and air
pollution control devices are working. To provide
surveillance of how the Mil is protected/ periodic
monitoring should be done at ground level where maximum
concentrations must not be exceeded. Contributions by
other- sources should also be considered.
3.5 ISSUE 5: Requirements for Good Incinerator Operations
The Subcommittee considered many guestions in evaluating the
regulations proposed for controlling operations of the combustion
device and the air pollution control device that together
constitute the incinerator system. The Agency's approach seeks
to establish minimally acceptable operation conditions, that if
satisfied, will prevent the need to perform compliance testing.
This approach potentially could reduce the cost of compliance for
municipalities and operators of sewage sludge incinerators.
However, at the present time, we do not find that there is an
adequate data base nor an adequate understanding of the
relationship between operating parameters and emissions to
support such a non-testing rationale.
An important factor in the proposed regulations is the means
by which they will be implemented. The establishment of
operating parameters and specific values for operating parameters
in the regulation could be a road block to the successful
implementation of these regulations. The Agency correctly
recognized the general relationship between metals volatility and
the operating temperature of the combustion device. However, it
was not recognized that the one temperature to be specified is
the temperature that the metals actually experience within the
burning bed of sludge. That particular temperature is difficult
to measure and is not appropriate to be considered in a
regulatory strategy. A possible alternative use'of temperature
as a control parameter would be one in which it is characteristic
of the operating incinerator system. If a compliance test is run
and metals1 limits are established to be acceptable, then the
12
-------
operating temperature can be monitored as a single temperature
for the entire combustion system device. It is not appropriate
to use a temperature that is generalizable to all combustion
systems without specifying exactly how it would be determined.
*
To achieve the maximum environmental benefits, it is
important to have regulations that impel society to take actions
that result in a reduction in any harmful emissions. The
Subcommittee suggests three major areas for emphasis: training
and education of incineration plant operators; at-source
reduction of toxic metals that presently enter the municipal
wastewater system? and performance-based regulations that provide
incentives for better pollution control equipment with lower life
cycle costs.
The most important operational parameter is the training
level of the operator of the sewage sludge incinerator system, A
high level of operator training should be required in order to
ensure that the system is continuously operated in a safe manner.
The regulations could require a minimum level of training as well
as a training certificate. Economic costs and benefits
associated with the use of highly trained personnel should be
considered.
It is important in the proposed regulations to establish
those parameters that influence the emissions of metals and
organics. However, because of the inter-dependence of the many
operating parameters, it is difficult to define a single set of
parameters that will ensure that all systems of different designs
and operating conditions are functioning in a safe manner. The
Implementation of the regulations by permit writers must allow
the establishment, on a case-by-case basis, the emissions and
relationships to operating conditions. A compliance test could
be conducted and the operating conditions established at the same
values for which compliance was assured.
The Subcommittee finds that the Agency has correctly
identified metals emissions from sewage sludge incinerators as
requiring high priority for regulation. However, due to faulty
statistical analysis of the incinerator emissions data base, the
proposed control of incinerator operating conditions is incorrect
and unsupported. Our review of the test data indicates that the
control efficiency for metals removal is mainly a function of the
air pollution control equipment, for the operating range
typically seen in sewage sludge incineration.
We also recognize the advance in regulatory strategy that
the Agency shows in proposing to limit the THC in the stack flue
gas and thus make it unnecessary to regulate other parameters
such as excess air and minimum operating temperatures which
influence organics destruction. This is an improvement over
current regulatory practices for hazardous waste incinerators
13
-------
which reach an impasse when a multitude of dependent operating
parameters are limited as though they are independent and can be
regulated separately.
Problems with the level of emissions of th®4Plfe from sewage
sludge incinerators can be exacerbated if operating temperatures
of 1200°-1600° F exist, as below 1500° F, reaction rates start to
decrease* At higher temperatures, the reaction rates for organics
destruction are fast and mixing or mass transfer rates need to be
considered. As most sewage sludge incinerators operate in the
range where time and temperature history have an impact on
destruction and removal efficiencies (ORE) for organics, low
levels of THC should be a good indicator of operations desirable
for the protection of human health and the environment.
A maximum operating temperature limitation appears rational,
as data indicate that for combustion of organic compounds* little
improvement in destruction efficiency is realized if the
temperature is higher than about 1600° F and the incinerator is
properly designed. Hazardous waste incinerators are operated at
1800° F to provide some safety margin. Sewage sludge
incinerators frequently process waste streams characterized by
elevated metals content. Consequently, higher operating
temperatures can volatilize certain toxic metal salts and produce
submicron particles which may require special and costly air
pollution control devices to capture. In view of these
observations, and concerns regarding data quality, the proposed
limitation of 1650 F is not justified. More research is needed
to support setting a specific maximum temperature limit „ In
addition, a maximum temperature limit which excludes sludge
vitrification and other high temperature technologies is
unwarranted as alternatively designed systems, e.g., for
vitrification, may accommodate higher operating temperatures.
The Subcommittee recommends that additional attention be
given to the composition of the sludge and that appropriate
issues of source reduction and pre-treatment merit further study,
It is generally agreed that a superior option for control is to
limit the amount of metals entering the incinerator.
For good operating management of an incinerator, it is
important for the operator to have continuous monitoring of
oxygen, temperature, and sludge feed rate. We think that such
monitoring should be required. However, except to relate these
operating variables to certain control options, and to relate
them to site specific conditions of successful compliance
testing, we find it unwise to set universally applied operating
numbers.
14
-------
S. REFERENCES
1. Lee, Kun-chieh, "Research Areas for Improved Incineration
Performance,11 J. Air Pollution Control Aa5ocia.fc ion f Vol. 38,
No. 12, December, 1988, pp 1542-1550,
2. U.S.E.P.A., memorandum, July 11, 1989, Office of solid Waste
and Emergency Response, "Review comments on Draft SAB Report
on PIC Controls for Hazardous Waste Incinerators," Bob
Holloway. Cites Midwestern Research Institute (MRI) report
"THC Monitor Survey,1' Project f 9102-L(57120), June 20,
1989.
3. U.S.E.P.A;, undated, "incineration of sewage Sludge,w
Technical Support Document, office of Water Regulations and
standards, Washington P.C.
4. U.S.E.P.A., memorandum, Office of Solid Waste, Waste
Treatment Branch, "Guidance on PIC Controls for Hazardous
Waste Incinerators," Shiva Garg, April 7,1S89,
5. U.S.E.P.A,, Science Advisory Board, Environmental
Engineering committee. July, 1989. "Proposed Controls for
Hazardous Waste Incinerators: Products of Incomplete
Combustion," Draft Report, (EPA-SAB-EEC-89-number to be
assigned). Washington D.C.
6. U.S.E.P.A. Science Advisory Board, Environmental Engineering
Committee, January, 1989. "Resolution on Use of
Mathematical Models by EPA for Regulatory Assessment And
Decision Making," (EPA-SAB"EEC-89-12). Washington D.C.
7. U.S.E.P.A., Science Advisory Board, Environmental Effects,
Transport, and Fate Committee. April, 1987, "Review of The
RisK Assessment Methodology for Municipal Incinerator
Emissions," (EPA-SAB-EIFT-87-Q27). Washington D.C.
8. U.S.E.P.A., Science Advisory Board, Environmental
Engineering committee. January, 1987. "Review of Technical
Documents Supporting Proposed Revisions to The EPA
Regulations for The Disposal/Reuse of Sewage Sludge under
Section 405(d) Of the Clean Water Act," (SAB-EEC-87-015),
Washington D.C.
9. U.S.E.P.A. Science Advisory Board, Environmental Effects,
Transport, and Fate Committee, April, 1985. "Report on the
Incineration of Liquid Hazardous Wastes." Washington D.C.
10. "Technical standards for The Use And Disposal of Sewage
Sludge, Federal Register. 54 H 5746-5902, February 6, 1989.
17
-------
6. GLOSSARY
ADF...........Air dispersion factor, major parameter
generated by air models and used to determine
ground-level concentrations of contamnants from
source mass emissions; stated in ug/m (g/sec)"1
CFR. code of Federal Regulations
COMPLEX-l.....USEPA air model for use in rural areas with
complex terrain
EEC. ,.Environmental Engineering Committee, SAB
FID. . Flame lonization Detector; so called "hot11 FIDs
utilize un-conditioned, high temperature stack gas,
while "cold" FIDs work with pre-conditioned (via
traps, filters, etc) lower temperature stack gas.
*
GC/MS.........Gas Chromatograph/Mass Spectrometer
ISCLT.........USEPA long-term air model for industrial source
complexes
LONGZ USEPA air model for use in urban areas with
complex terrain
MEI...........Most Exposed Individual; a hypothetical
person living at the point of maximum ground-level
concentration of the pollutant of concern, who is
exposed 24 hours/day, breathes 20 m3 of air/day,
weighs 70 kg, and is exposed for a 70 year span
NRDC..........Natural Resources Defense Council
ow....» *.Office of Water, USEPA
pic. .Products of Incomplete Combustion
POTW. Publicly Owned Treatment Works
q*............Cancer potency value, expressed as the upper-bound
estimate of the dose-response curve in the low-
dose range in [mg/kg/day]"1
SAB. Science Advisory Board of the EPA
THC. ,..Total Hydrocarbons
W170...,. A Committee of experts, assembled by the U.S. Dept.
of Agriculture, on the land disposal/land
application aspects of sewage sludge treatment
18
-------
4. CONCLUSIONS
The Office of Water has made a strong effort to use a
limited data base on emissions and the associated hfcalth risk of
toxicants to develop a risk-based regulation for "sewage sludge
incinerators. These regulations focus on the control of metals
and organics emissions froro incinerators. The Subcommittee
believes that incineration is an important and viable technology
for sewage sludge disposal; we conclude however, that there are a
number of technical flaws with the currently proposed regulations
which are likely to preclude the effective regulation of sewage
sludge treatment by incineration.
These risk-based regulations do not have a strong enough
technical basis to allow actual standards to be developed
directly, due to a wide range of uncertainties associated with
the risk-based analysis. Many safety factors are built into the
analysis. Each individual safety factor appears reasonable, but
the multiplicative use of a series of such factors makes the
final number unreasonable. The methodology as presented does not
explicitly assign measures of uncertainty or confidence to the
calculations, A false and misleading impression of confidence is
conveyed by the final expression of risk as a single number.
Even though the Agency tried to simplify the regulatory
process, the proposed regulations will be difficult to implement
because the simplified compliance requirements are based on
faulty statistical analysis. They will put a heavy burden on
permit writers.
Statistical analysis of the limited data base for emissions
are flawed due to the practice of indiscriminate averaging over a
range of different types of operating devices, different
operating conditions, and different types of air pollution
control devices.
We endorse the concept of using a stack gas measure of total
hydrocarbons emissions for monitoring sludge incineration and air
pollution control device performance. However, the use of total
hydrocarbons as a direct indicator of risk is not possible due to
the uncertainties associated with the field implementation of hot
FID systems and the lack of a direct link between total
hydrocarbonst as measured by the FID, and the total spectrum of
organics which might be emitted from sewage sludge incinerators.
Also, since it has not been demonstrated that the proposed hot
FID systems can operate continuously in the stack gas environment
of sewage sludge incinerators, it is not appropriate to propose
regulations that will demand such operation in order to be in
compliance. Recent information provided by the Office of Water
(reference 2), based on a survey of facilities using cold FID
monitoring systems, suggests that this approach may well be a
IS
-------
feasible alternative method to monitoring THC, Adopting cold FID
monitoring would thus appear to obviate the problems noted above.
Although we can understand the saving of resoiarces that
would follow if there were a technical basis to exempt operators
of sewage sludge incinerators from required compliance testing,
it is insupportable at this time to establish regulations in
which individual incinerators would not have to be tested for
compliance.
The air transport dispersion models used in developing the
proposed regulations are approved EPA models with defined
guidelines for.their use. However, the Agency has ignored its
own technical requirements in seeking to offer simplified ways of
using results from selected model calculations. There is no
scientific evidence that sewage sludge incinerators reguire
specifically designed models to deal with their emissions. No
new simplified air-transport model is needed for these specific
regulations.
This subcommittee did not undertake a full review of the air
transport models because it was aware of the findings of a draft
report in preparation by the SAB PIC Review Subcommittee which
dealt in detail with this issue (several of the incineration
Subcommittee's members participated in that review). The
majority of the Subcommittee accepts the comments and conclusions
of that report.
The Agency has taken an overly conservative approach in the
selection of the most exposed individual. It is highly unlikely
that any such exposure and physiology will actually exist.
Again, since this topic was also addressed in the previously
mentioned PIC report, we support the conclusions of that report.
16
------- |