TRIAL BURN AND RISK ASSESSMENT RESULTS FOR THREE BIF UNITS
Gerald J. Drake and Melissa L. Douglas
Compliance Strategies & Solutions, Inc.
John R. Jones, P.E.
Reilly Industries, Inc.
ABSTRACT
The United States
Environmental Protection Agency Region V (hereafter, EPA) required the Reilly
Industries, Inc. (hereafter, Reilly) facility located in Indianapolis, Indiana
to modify its container and tank storage RCRA Part B Permit to incorporate
three on-site boilers burning hazardous waste.
The EPA required Reilly to conduct a Trial Burn on each of the boilers
in order to establish permit limits and to collect data for determining the
incremental risk presented by the facility.
Emissions testing was performed on Boilers 70K and 30K and included
dioxin/furans, volatile organics, semivolatile organics, particulate matter,
HCl/Cl2, PCBs, total organics, total hydrocarbons, and hexavalent
chromium.
The Trial Burn testing
generated the necessary data for proposing RCRA Part B Permit limits and for
conducting a direct and indirect human health risk assessment (hereafter, human
health risk assessment) and a screening level ecological risk assessment
(hereafter, ecological risk assessment).
Destruction and removal efficiency (DRE) results of greater than 99.999%
were demonstrated. A hexavalent chromium
conversion ratio of 24.5% was also demonstrated. The risk assessments demonstrated acceptable
risks and hazards for all three boilers operating simultaneously, 365 days a
year, at their maximum hazardous waste feed rates.
INTRODUCTION
The Reilly facility is a
manufacturer of pyridine and pyridine-derived organic chemicals. The production processes at the facility
generate various waste streams that are either burned onsite in Boilers 70K,
30K, and 28K, or are shipped offsite for treatment. The waste streams burned onsite in the
boilers are classified as hazardous waste due to the characteristic of
ignitability (D001), the characteristic of toxicity (D018 and D038), or are listed
from non-specific sources (F003 and F005).
None of the boilers have air pollution control devices.
The EPA has primacy over the
BIF program in the State of Indiana, and requested a RCRA Part B Permit
Application for Reilly’s boilers in a letter received on February 11,
1998. Pursuant to this letter, Reilly
was required to modify their existing RCRA Part B Permit to add the boilers,
perform a Trial Burn and to determine permit conditions based on the results of
the human health and ecological risk assessments. The EPA approved a Demonstration of
Similarity for Boiler 28K and, therefore, emission testing was not required for
this boiler.[1]
This paper briefly presents
the test objectives and approach for gathering the information for permitting
Boilers 70K, 30K, and 28K. In addition,
the significant test results and proposed permit limits associated with the
Trial Burn and risk assessments are presented.
TEST OBJECTIVES
The objectives of the testing
were to generate the data necessary to conduct human health and ecological risk
assessments and to propose RCRA Part B Permit limits for the three
boilers. The test objectives are
summarized as follows:
·
Demonstrate a DRE
for the designated principal organic hazardous constituent (POHC) of at least
99.99%;
·
Demonstrate that
carbon monoxide (CO) emissions were controlled to less than 100 parts per
million dry volume (ppmdv) corrected to 7% oxygen (O2);
·
Demonstrate that
particulate matter (PM) emissions were controlled to less than 180 milligrams
per dry standard cubic meter (mg/dscm) of stack gas corrected to 7% O2
(0.08 grains per dry standard cubic foot (gr/dscf) of stack gas corrected to 7%
O2);
·
Develop the
emissions data necessary to perform the human health and ecological risk
assessments including polychlorinated dibenzo(p)dioxins and polychlorinated
dibenzofurans (dioxins/furans), total hydrocarbons (THC), hydrogen chloride and
chlorine gas (HCl/Cl2), metals, and other organic products of
incomplete combustion (PICs);
·
Determine the
hexavalent chromium conversion ratio for the combustion process; and
·
Establish
limitations on the hazardous waste feed rate and other operating conditions
that ensure compliance with the performance standards and the risk-based
emission limits.
TEST APPROACH
The approach taken to design
the test was based on satisfying the test objectives and the need to maximize
the permit limits to achieve the greatest flexibility for facility
operations. The varied aspects of the
test approach provided a number of cost saving opportunities that included a
demonstration of similarity, minimizing test conditions, and minimizing
sampling and analysis, which are discussed in the following sections. Also, a summary of the testing that was
actually performed is included.
Demonstration of Similarity
A demonstration of similarity
for Boilers 30K and 28K was presented to the EPA. The demonstration of similarity compared the
feed streams, design, operation, and maintenance of the two non-identical
boilers.[1] The EPA approved the
similarity for the two non-identical boilers with the stipulation that the
permit limits for Boiler 28K be 3% lower than the limits demonstrated for
Boiler 30K due to sizing differences.
Based on the demonstration of similarity, testing of Boiler 28K was not
required.
Minimizing Test Conditions
The test was designed to
obtain the data required to establish permit limits, maximize these limits, and
confirm that the incremental risks posed by the operations was acceptable to
the EPA and the public. Trial Burns for
boilers are traditionally designed with two test conditions. One test condition is conducted at high
temperatures and maximum feed rates, emissions, and operating conditions. The second test condition is conducted at minimum
temperatures and maximum emission concentrations. Data for risk assessment purposes is often
collected during a third test condition known as a Risk Burn. Risk Burns are usually conducted at “normal”
operating conditions, and are intended to produce lower emission rates that are
representative of long-term operations rather than the higher rates of
emissions measured at the extreme operating conditions of a Trial Burn.
Reilly chose to manage the
incremental risk presented by its operations in such a manner that a separate
Risk Burn test condition was not performed.
Instead, Reilly agreed to demonstrate acceptable risks and hazards by
testing its boilers at the extreme of their operations. Worst case emissions were demonstrated at the
maximum feed rate/maximum temperature test condition and the minimum
temperature DRE test condition of the Trial Burn. The highest dioxin/furan emission rate
determined during these two test conditions was used in the risk
assessments. The number of test
conditions was minimized by eliminating a separate Risk Burn test condition
and, thus, only two test conditions were performed on Boiler 70K and Boiler
30K.[1]
Minimizing Sampling and Analysis
A goal for all testing is to
obtain only the data necessary to satisfy the test objectives. A method for minimizing the amount of samples
and analyses is to use the Adjusted Tier I feed rate screening limits for
metals.[2] Feed stream analyses are
required to demonstrate conformance with the Adjusted Tier I feed rate
screening limits. This approach
eliminates stack gas sampling for metals and provides significant cost savings
for facilities with low metal feed rates.
At Reilly, the Adjusted Tier
I feed rate screening limit for chromium was modified by a hexavalent chromium
conversion ratio determined through stack gas and feed stream sampling and analysis.[3] To accomplish this modification, feed stream
sampling for total chromium and stack gas sampling for hexavalent chromium was
required. The demonstrated conversion
ratio was then applied to the Adjusted Tier I feed rate screening limit to
specify a total chromium feed rate limit that resulted in acceptable risks and
hazards associated with hexavalent chromium emissions. The hexavalent chromium emissions testing is
an additional sampling train that is not normally required for a Trial Burn or
Risk Burn. However, this testing was
performed in order to provide the facility with adequate feed rate limits
having the least impact on normal operations.
Testing to support the risk
assessment typically requires the stack gas sampling and analysis of metals. Rather than testing metal emissions, the
Adjusted Tier I or modified Adjusted Tier I feed rate screening limit was used
as inputs to the risk assessment modeling.
Because it was anticipated that these Adjusted Tier I values would
result in unacceptable risk results (the tiered limits were based on direct
inhalation only and did not include the indirect pathways), multiple iterations
of the risk assessment calculations were conducted until they yielded
acceptable incremental risks and hazards.
In other words, the Adjusted Tier I feed rate screening limits for
metals were modified until acceptable risks and hazards were achieved. The modified values were then translated into
the feed rate limits for the metals.
This method did not require sampling and analysis of the stack gases for
the metals. Instead, it allowed the
metal feed rates to be maximized based on the successful results of the risk
assessments. It also allowed the
development of different modes of operation without performing additional testing. The additional modes of operation serve to
grant Reilly a wider range of acceptable operations and they are discussed in
greater detail later in this paper.
Summary of the Testing
Table I presents a summary of
the sampling and analyses performed during the testing of Boilers 70K and 30K
and the test condition(s) during which it was performed.
Table I
Summary of Sampling and Analyses of
Boilers 70K and 30K
|
Parameter |
Methods |
Test
Condition 1 |
Test
Condition 2 |
|
Hazardous
Waste and Spiking Stream Samples |
|||
|
Metals |
3050B/Various |
X |
|
|
Total Chlorine/Chloride |
5050/9056 |
X |
|
|
Ash Content |
ASTM D482 |
X |
|
|
Volatile Organics |
5030B/8260B |
X |
|
|
Semivolatile Organics |
5030B/8270C |
X |
|
|
Semivolatile POHC |
5030B/8270C |
|
X |
|
PCBs |
HRGC/HRMS |
X |
|
|
Heat Content |
ASTM D240 |
X |
|
|
Stack Gas
Samples |
|||
|
Particulate Matter |
Method 5 |
X |
|
|
Total Chlorine/Chloride |
0050/9057 |
X |
|
|
Hexavalent Chromium |
0061/7199 |
X |
|
|
Volatile Organics |
0030/5041A/8260B |
X |
|
|
Semivolatile Organics |
0010/3542/8270C |
X |
|
|
Semivolatile POHC |
0010/3542/8270C |
|
X |
|
Dioxins/Furans |
0023A/8290 |
X |
X |
|
PCBs |
0023A/HRGC/HRMS |
X |
|
|
Total Organics – VOC |
0040/GC/FID |
X |
|
|
Total Organics – SVOC |
0010/GC/FID |
X |
|
|
Total Organics – GRAV |
0010/Gravimetry |
X |
|
|
Carbon Monoxide |
Method 10 |
X |
X |
|
Oxygen |
Method 3A |
X |
X |
|
Total Hydrocarbons |
Method 25A |
X |
X |
Test Condition 1 – High
temperature and maximum operating conditions.
Test Condition 2 – Minimum
temperature.
Several items should be noted
about the information contained in Table I.
Stack gas sampling for dioxin/furans was conducted during both test
conditions. The highest emission rate
(on a toxicity equivalence basis) was used in the risk assessments. This was an agreement previously made with
the EPA to eliminate long-term permit limits on feed rates. Total hydrocarbon measurements were obtained
during both test conditions to verify that similar emissions occurred at the
different operating extremes. The
hazardous waste feed stream was analyzed for the volatile organics and
semivolatile organics so that compounds not detected in the feed stream, the
stack gases, and in historical analyses of the feed stream could be eliminated
from inclusion in the risk assessments.[4]
SIGNIFICANT TEST RESULTS
The significant results of
the testing on Boilers 70K and 30K are presented in this section. The results of the testing were used in the
performance of the human health and ecological risk assessments, which
demonstrated acceptable risks and hazards.
The results presented include the emissions of PM, dioxin/furans, total
organics and THC, and organic PICs, the DRE determination, and the hexavalent
chromium conversion ratio determination.
Each of these is discussed in the following sections.
PM Emissions
The PM emissions were
determined during Test Condition 1 while injecting ash spiking material into
the waste feed stream. Table II presents
the PM emission results for each run and the sootblow corrected result for the
test condition.[5] The sootblow was
performed during Run 3. The total ash
feed rate is also presented in Table II along with a corresponding partitioning
factor. The partitioning factor is
defined as the ratio of the PM emissions to the total ash feed rate.
Table II
Summary of PM Emissions and Ash Feed
Rates
|
Parameter |
Run 1 |
Run 2 |
Run 3 |
Sootblow Corrected |
|
Boiler 70K |
||||
|
PM Emissions (mg/dscm @ 7%
O2) |
86.5 |
64.5 |
127.9 |
85.9 |
|
PM Emissions (gr/dscf @ 7%
O2) |
0.0378 |
0.0282 |
0.0559 |
0.0375 |
|
PM Emissions (g/hr @ 7% O2) |
2,033 |
1,539 |
3,050 |
2,037 |
|
Ash Feed Rate (g/hr) |
5,001 |
4,902 |
4,873 |
4,925 |
|
Partitioning Factor (%) |
40.6 |
31.4 |
62.6 |
41.4 |
|
Boiler 30K |
||||
|
PM Emissions (mg/dscm @ 7%
O2) |
113.0 |
114.6 |
164.7 |
127.9 |
|
PM Emissions (gr/dscf @ 7%
O2) |
0.0494 |
0.0501 |
0.0720 |
0.0559 |
|
PM Emissions (g/hr @ 7% O2) |
1,442 |
1,502 |
2,121 |
1,652 |
|
Ash Feed Rate (g/hr) |
4,039 |
2,561 |
2,534 |
3,045 |
|
Partitioning Factor (%) |
35.7 |
58.6 |
83.7 |
54.2 |
As indicated in Table II, ash
partitioning to the stack gas ranged from approximately 31% to 59% during
non-sootblow runs. The sootblow runs
showed partitioning of approximately 63% to 84%. Correcting for the sootblow provided
partitioning factors of approximately 41% to 54%.
Dioxin/Furan Emissions
Dioxin/furan emissions were
measured during both Test Condition 1 and Test Condition 2 for Boilers 70K and
30K. Table III summarizes the
dioxin/furan emissions converted to a TEQ basis and corrected to 7% O2. The contribution of risk associated with
these emissions is approximately 1.0E-08, which is three orders of magnitude
less than the typical allowable incremental risks.[4]
Table III
Summary of Dioxin/Furan Emissions
|
Run Number |
Boiler 70K Total TEQ
Emission Rate |
Boiler 30K Total TEQ
Emission Rate |
||
|
ng/hr |
ng/dscm
@7% O2 |
ng/hr |
ng/dscm
@7% O2 |
|
|
|
Test
Condition 1 – Maximum Waste Feed Rate |
|||
|
Run 1 |
< 184 |
< 0.00642 |
< 290 |
< 0.0189 |
|
Run 2 |
< 185 |
< 0.00654 |
< 265 |
< 0.0177 |
|
Run 3 |
< 171 |
< 0.00579 |
< 256 |
< 0.0177 |
|
|
Test
Condition 2 – Minimum Temperature |
|||
|
Run 1 |
< 68.3 |
< 0.00855 |
< 19.9 |
< 0.00507 |
|
Run 2 |
< 49.8 |
< 0.00535 |
< 21.5 |
< 0.00338 |
|
Run 3 |
< 70.5 |
< 0.00785 |
< 23.6 |
< 0.00384 |
On a mass emission rate
basis, the highest dioxin/furan emissions occurred during the high feed rate
test condition. The emissions used in
the risk assessments are those based on the mass emission rate. Therefore, the highest dioxin/furan emissions
occurred during the high feed rate test condition as demonstrated during the
testing of Boilers 70K and 30K.
Total Organic and Total Hydrocarbon Emissions
Total organic emissions were
determined during Test Condition 1 and included volatile (VOC), semivolatile
(SVOC), and gravimetric (GRAV) fractions.
Total hydrocarbon emissions were measured during Test Condition 1 and
Test Condition 2. Table IV summarizes
the total organic emissions by each fraction and as a total, and compares these
values to the total hydrocarbon emissions that occurred during Test Condition
1.
Table IV
Total Organic and Total Hydrocarbon
Emissions
|
Constituent |
Boiler 70K
(g/hr) |
Boiler 30K
(g/hr) |
||||
|
Run 1 |
Run 2 |
Run 3 |
Run 1 |
Run 2 |
Run 3 |
|
|
Total Organics - VOC |
< 64.8 |
< 29.5 |
< 24.5 |
< 12.9 |
< 13.3 |
< 12.6 |
|
Total Organics – SVOC |
56.2 |
40.7 |
50.0 |
63.0 |
92.9 |
60.8 |
|
Total Organics – GRAV |
27.7 |
48.6 |
33.5 |
54.4 |
26.6 |
77.8 |
|
Total Organics |
< 148.7 |
< 118.8 |
< 108.0 |
< 130.3 |
< 132.8 |
< 151.2 |
|
Total Hydrocarbons |
< 4.5 |
< 4.5 |
13.6 |
9.1 |
9.1 |
< 4.5 |
As indicated in Table IV, the
mass emission rates obtained from the total organic trains and the Method 25A
THC analyzer do not compare favorably.
The total organic trains produced mass emission rates approximately one
order of magnitude greater than the THC analyzer.
Table V presents a summary of
the total hydrocarbon emissions measured during Test Condition 1 and Test
Condition 2.
Table V
Summary of Total Hydrocarbon Emissions
|
Run Number |
Boiler 70K
THC Emissions (g/hr) |
Boiler 30K
THC Emissions (g/hr) |
||
|
Test
Condition 1 |
Test
Condition 2 |
Test
Condition 1 |
Test
Condition 2 |
|
|
Run 1 |
< 4.5 |
< 4.5 |
9.1 |
< 4.5 |
|
Run 2 |
< 4.5 |
4.5 |
9.1 |
18.1 |
|
Run 3 |
13.6 |
9.1 |
< 4.5 |
18.1 |
In general, the mass
emissions measured by the THC analyzer during the maximum waste feed rate and
minimum temperature test conditions were equivalent.
In the uncertainty section of
the risk assessments, a comparison of the total speciated organic emissions and
the total unspeciated organic emissions was performed. The total speciated organic emissions were
obtained from the volatile organic, semivolatile organic, dioxin/furan, and PCB
sample trains. The total unspeciated
organic emissions were obtained from the total organic VOC, SVOC, and GRAV
fractions discussed above. Table VI
presents the speciated and unspeciated organic emissions that occurred during
the testing.
Table VI
Speciated Versus Unspeciated Organic
Emissions
|
Parameter |
Boiler 70K |
Boiler 30K |
||||
|
Run 1 |
Run 2 |
Run 3 |
Run 1 |
Run 2 |
Run 3 |
|
|
Speciated Organic Emissions
(g/s) |
0.0017 |
0.0042 |
0.0018 |
0.0025 |
0.0029 |
0.0029 |
|
Unspeciated Organic
Emissions (g/s) |
0.0398 |
0.0309 |
0.0274 |
0.0348 |
0.0356 |
0.0406 |
|
Percent Speciated (%) |
4.27 |
13.53 |
6.42 |
7.16 |
8.26 |
7.22 |
As shown in Table VI, the
percentage of speciated organic compounds detected during the testing was
generally less than 10% of the unspeciated organics detected.
Organic PIC Emissions
Volatile and semivolatile
organic and PCB emissions were measured during Test Condition 1 for inclusion
in the risk assessments. There were 38
volatile organic compounds and 23 semivolatile organic compounds measured
pursuant to the target compound list developed for the testing.[6] In addition, the coplanar and homologue group
PCB emissions were also measured. A
summary of the results is not included due to the volume of information and its
relatively low impact on the risk assessment results. However, these data were incorporated into
the risk assessment modeling, which yielded acceptable risks and hazards.
DRE Determination
The POHC designated for the
Trial Burn was 1,2-Dichlorobenzene (1,2-DCB). Table VII presents the POHC feed
rate and emission rate data and the calculated DRE for each run of the testing
on Boilers 70K and 30K. The DRE for each
run was demonstrated to be greater than 99.999%.
Table VII
Summary of DRE
Determination
|
Parameter |
Run 1 |
Run 2 |
Run 3 |
|
|
Boiler 70K |
||||
|
POHC Feed Rate (g/s) |
5.6701 |
5.6663 |
5.6663 |
|
|
POHC Emission Rate (g/s) |
< 5.06E-05 |
< 5.12E-05 |
< 5.00E-05 |
|
|
DRE (%) |
> 99.99911 |
> 99.99910 |
> 99.99912 |
|
|
Boiler 30K |
||||
|
POHC Feed Rate (g/s) |
2.6453 |
2.6377 |
2.6440 |
|
|
POHC Emission Rate (g/s) |
< 1.51E-05 |
< 1.56E-05 |
< 1.52E-05 |
|
|
DRE (%) |
> 99.99943 |
> 99.99941 |
> 99.99942 |
|
Hexavalent Chromium Conversion Ratio Determination
Testing was performed to
demonstrate a hexavalent chromium conversion ratio for the boilers. The hexavalent chromium conversion ratio is
defined as the ratio of hexavalent chromium emissions to the total chromium feed
rate. The purpose of this demonstration
was to measure the amount of hexavalent chromium that is formed (or left
unchanged if hexavalent chromium is present in the feed) in the combustion
process based on the feed rate of total chromium into the boiler.
In order to determine the
hexavalent chromium conversion ratio, the total chromium emissions were not
measured during the testing, only the hexavalent chromium emissions and the
total chromium feed rate were measured.
The hexavalent chromium conversion ratio is important because risk
assessments only address the risks and hazards associated with hexavalent
chromium. Using an emission rate based
on total chromium artificially increases the risks and hazards associated with
the emissions. Additionally, the total
chromium feed rate limit is decreased for this same reason. Therefore, the hexavalent chromium conversion
ratio is used to increase the total chromium feed rate limit to a level that
reflects the actual amounts modeled in the risk assessments. Extrapolation to higher feed rate limits is
allowed by the EPA.[7] In addition,
worst-case conditions were realized in the testing by using a hexavalent
chromium spiking material to increase the total chromium feed rate.
The average total chromium
feed rate that occurred during the testing was 9.40 g/hr, of which
approximately 7.72 g/hr was contributed by the hexavalent chromium spiking
material and approximately 1.68 g/hr was present in the hazardous waste feed
stream (presumed to exist as trivalent chromium). The hexavalent chromium emission rate
measured during the testing was 2.30 g/hr.
Therefore, the hexavalent chromium conversion ratio was demonstrated to
be 24.5%.
PROPOSED PERMIT LIMITS
Permit limits were proposed
to the EPA based on the testing and risk assessment results. The limits represent an incremental risk of
less than 1x10-5 and a hazard index of less than 0.25. The significant proposed permit limits are
discussed in the following sections.
Operating Limits
The operating limits
established during the testing included those parameters specified in Table
VIII. The maximum limits were
established during Test Condition 1, and the minimum limits were established
during Test Condition 2.
Table VIII
Summary of Proposed Operating Limits
|
Operating Parameter |
Proposed
Permit Limits |
||
|
Boiler 70K |
Boiler 30K |
Boiler 28K |
|
|
Maximum Waste Feed Rate |
0.47 Kg/s (3,719 lb/hr) |
0.29 Kg/s (2,272 lb/hr) |
0.28 Kg/s (2,204 lb/hr) |
|
Minimum Combustion
Temperature |
773°C
(1,424°F) |
768°C
(1,414°F) |
745°C (1,372°F) |
|
Maximum Steam Production
Rate |
8.22 Kg/s (65.20 Klb/hr) |
4.06 Kg/s (32.25 Klb/hr) |
3.94 Kg/s (31.28 Klb/hr) |
|
Minimum Steam Production
Rate |
3.38 Kg/s (26.80 Klb/hr) |
1.31 Kg/s (10.39 Klb/hr) |
1.27 Kg/s (10.08 Klb/hr) |
Feed Rate Limits
Feed rate limits were
established for ash, total chlorine/chloride, and the ten BIF metals. The established ash feed rate limits were
4,926 g/hr, 3,045 g/hr, and 2,954 g/hr for Boilers 70K, 30K, and 28K,
respectively, and were based on compliance with the PM emission standard.
The feed rate limit for total
chlorine/chloride was established based on the Adjusted Tier I feed rate
screening limits and, for the final permit, the risk assessments. For Boiler 70K, the chromium feed rate limit
was based on the amount fed during the testing as required by the EPA. For Boilers 30K and 28K, the feed rate limit
for chromium was established based on the Adjusted Tier I feed rate screening
limits, the demonstrated hexavalent chromium conversion ratio from the Trial
Burn and, for the final permit, the risk assessments. The feed rate limits for the remaining metals
were established based on the Adjusted Tier I feed rate screening limits from
the Trial Burn and, for the final permit, the risk assessments. Table IX summarizes and compares the feed
rate limits for total chlorine/chloride and the metals based on the Trial Burn
and the risk assessments. Multiple risk
assessment iterations were conducted in order to maximize the feed rate limit
of chromium at the expense of the remaining metals.
Table IX
Summary of Proposed Chlorine and Metal
Feed Rate Limits
|
Parameter |
Boiler 70K
(g/hr) |
Boiler 30K
(g/hr) |
Boiler 28K
(g/hr) |
|||
|
Tier IA |
RA |
Tier IA |
RA |
Tier IA |
RA |
|
|
Chlorine |
216,000 |
8,500 |
45,800 |
5,000 |
44,426 |
4,500 |
|
Antimony |
1,620 |
31.86 |
344 |
18.86 |
334 |
18.25 |
|
Arsenic |
12.42 |
6.80 |
2.63 |
4.03 |
2.56 |
3.89 |
|
Barium |
270,037 |
8.50 |
57,245 |
5.00 |
55,577 |
4.86 |
|
Beryllium |
22.68 |
1.02 |
4.81 |
0.60 |
4.67 |
0.58 |
|
Cadmium |
30.24 |
4.25 |
6.41 |
2.51 |
6.23 |
2.43 |
|
Chromium |
9.40 |
9.40 |
3.88 |
7.76 |
3.76 |
3.76 |
|
Lead |
486 |
35.64 |
103 |
21.06 |
100 |
20.41 |
|
Mercury |
1,620 |
0.27 |
344 |
0.16 |
334 |
0.15 |
|
Silver |
16,202 |
14.69 |
3,435 |
8.68 |
3,335 |
8.42 |
|
Thallium |
2,700 |
0.85 |
572 |
0.50 |
556 |
0.49 |
Tier IA – Adjusted Tier I
Feed Rate Screening Limit
RA – Risk Assessments
The risk assessment values
presented in Table IX represent the simultaneous operation of the three
boilers. The risk assessment modeling
was also conducted for other modes of operation, where combinations of the
boilers or individual boilers were modeled.
In all, seven different modes of operation were modeled, and were
identified as the 70K/30K/28K, 70K/30K, 70K/28K, 30K/28K, 70K, 30K, and 28K
modes of operation. The feed rate limits
for the metals were allowed to change under the different modes of
operation. Table X presents the feed
rate limits for the different modes of operation for antimony, arsenic, barium,
beryllium, and cadmium. Table XI
presents the feed rate limits for the different modes of operation for
chromium, lead, mercury, silver, and thallium.
Table X
Proposed Metal Feed Rate Limits for
Different Modes of Operation
|
Operating Mode |
Proposed
Feed Rate Limits (g/hr) |
||||
|
Antimony |
Arsenic |
Barium |
Beryllium |
Cadmium |
|
|
70K/30K/28K |
31.86/18.86 /18.25 |
6.80/4.03 /3.89 |
8.50/5.00 /4.86 |
1.02/0.60 /0.58 |
4.25/2.51 /2.43 |
|
70K/30K |
47.88/28.26 |
11.16/6.48 |
84.96/50.04 |
1.02/0.60 |
6.37/3.78 |
|
70K/28K |
47.88/28.26 |
11.16/6.41 |
84.96/50.04 |
1.02/0.58 |
6.37/3.04 |
|
30K/28K |
33.91/33.91 |
4.32/4.32 |
50.04/50.04 |
0.60/0.58 |
2.51/2.43 |
|
70K |
89.28 |
20.52 |
84.96 |
3.06 |
21.24 |
|
30K |
67.68 |
11.34 |
100.08 |
1.20 |
5.00 |
|
28K |
67.68 |
10.80 |
100.80 |
1.20 |
5.00 |
Table XI
Proposed Metal Feed Rate Limits for
Different Modes of Operation
|
Operating Mode |
Proposed
Feed Rate Limits (g/hr) |
||||
|
Chromium |
Lead |
Mercury |
Silver |
Thallium |
|
|
70K/30K/28K |
9.40/7.76 /3.76 |
35.64/21.06 /20.41 |
0.27/0.16 /0.15 |
14.69/8.68 /8.42 |
0.85/0.50 /0.49 |
|
70K/30K |
10.74/8.82 |
47.52/28.08 |
0.33/0.20 |
22.03/13.03 |
1.27/0.75 |
|
70K/28K |
10.74/8.82 |
47.52/28.08 |
0.33/0.20 |
22.03/13.03 |
1.27/0.75 |
|
30K/28K |
9.63/9.63 |
35.10/35.10 |
0.33/0.33 |
14.98/14.98 |
0.86/0.86 |
|
70K |
16.68 |
91.44 |
0.48 |
38.52 |
2.21 |
|
30K |
11.24 |
70.20 |
0.67 |
29.95 |
1.73 |
|
28K |
11.10 |
70.20 |
0.67 |
29.95 |
1.73 |
Because the chromium feed
rate limits use the hexavalent chromium conversion ratio, a slightly different
approach was used in addition to the risk assessment results to specify a feed
rate limit for chromium. The maximum
feed rate demonstrated during testing was 9.40 g/hr of total chromium, of which
2.30 g/hr (24.5%) was attributed to hexavalent chromium. Reilly also proposed an incremental feed rate
increase above the 9.40 g/hr limit by assuming that 100% conversion to
hexavalent chromium occurs at rates above the demonstrated 9.40 g/hr. This approach was accepted by the EPA, and it
provided Reilly with greater flexibility and assurance that the final feed rate
limit would not present a hindrance to its operations. The chromium feed rate limits presented in
Table XI reflect this limitation.
Miscellaneous Limits
Permit limits were proposed
for several miscellaneous parameters including combustion gas velocity,
combustion chamber pressure, and waste firing system. Since Reilly does not have a combustion gas
flow rate measurement device, the permit limit for combustion gas velocity was
proposed based on the specifications for the combustion air fans. These specifications included the horsepower,
revolutions per minute (RPMs), and the maximum design flow rate. The combustion chamber is totally sealed as
the means for controlling fugitive emissions, and it is not equipped with
pressure monitoring devices. Therefore,
a permit limit on the combustion chamber pressure was not required. The waste firing system permit limit was
proposed as the atomizing media pressure being greater than the hazardous waste
feed stream pressure as measured at the burner gun. In these instances where miscellaneous permit
constraints were needed, the facility was able to present alternative solutions
to demonstrate compliance that further helped to reduce the financial cost and
burden on facility operations.
CONCLUSIONS
The Trial Burn and Risk Burn
testing was combined into a single test consisting of two test conditions. The results of the testing satisfied the
objective of generating data for obtaining a RCRA Part B Permit for three
boilers and for conducting human health and ecological risk assessments. DRE results of greater than 99.999% were
demonstrated for the 1,2-DCB POHC. A
hexavalent chromium conversion ratio of 24.5% was demonstrated for the
combustion process. Permit limits were
proposed based on the testing and the successful risk assessment results. Feed rate limits for total chlorine/chloride
and metals were proposed for seven different modes of operation.
REFERENCES
1. Drake, G.J., Douglas, M.L., and Jones, J.R.,
“Reducing the Burdens of Compliance Testing Under Increasing Regulatory
Requirements”, 2000 International Conference on Incineration and Thermal
Treatment Technologies, Portland, Oregon, May 8-12, 2000.
2. 40 CFR 266.106(d) and (e).
3. 40 CFR 266.106(g).
4. U.S. EPA, “Human Health Risk Assessment
Protocol for Hazardous Waste Combustion Facilities”, Peer Review Draft,
EPA530-D-98-001A, July 1998.
5. U.S. EPA, “Technical Implementation Document
for EPA’s Boiler and Industrial Furnace Regulations”, EPA530-R-92-011, March
1992, Section 5.2.3.9.
6. Drake, G.J., Douglas, M.L., and Jones, J.R.,
“A Sophisticated Approach for Trial Burn Plan Development to Support a
Post-Trial Burn Risk Assessment”, 2000 International Conference on Incineration
and Thermal Treatment Technologies, Portland, Oregon, May 8-12, 2000.
7. U.S. EPA, “Technical Implementation Document
for EPA’s Boiler and Industrial Furnace Regulations”, EPA530-R-92-011, March
1992, Section 10.5.