Reducing the Burdens of Compliance Testing under Increasing Regulatory Requirements[1]
Jerry Drake and Lisa Douglas
Compliance Strategies & Solutions, Inc.
Houston, TX 77058
John Jones, P.E.
Reilly Industries, Inc.
Indianapolis, IN 46241
ABSTRACT
Hazardous waste combustion facilities across the United States are faced with increasing regulatory requirements for compliance testing that result in significant costs and effort for demonstrating compliance and continued operations. Three separate methodologies are assessed for reducing the burdens of compliance testing while satisfying these regulatory requirements. Each of these methodologies offers opportunities to achieve significant cost savings for a facility as well as reduced recordkeeping requirements and permit parameters.
The first methodology assessed is the combination of a recertification of compliance (RCOC) with a Trial Burn. Numerous boiler and industrial furnace operators are currently faced with significant compliance testing in order to meet both the interim status RCOC requirements and the Trial Burn requirements for their RCRA Part B Permit. An approach is discussed to gain Agency approval for combining the testing including any necessary certification of compliance extension requests, the simplified compliance test notification, and the actual test plan, which equates to realized cost savings for the facility operator.
The second methodology assessed is the Trial Burn/Risk Burn test condition minimization. A traditional Trial Burn typically has testing at two separate test conditions. These test conditions include a high temperature test for metals and a low temperature test for destruction and removal efficiency (DRE). A third test condition is typically required by the Agency in order to gather emissions data under “normal” operations for performing a Risk Assessment. These multiple test conditions are costly to implement and minimize the operational flexibility of the facility by establishing multiple operational constraints into a permit (i.e., annual limits in addition to hourly rolling average limits). An approach is presented to combine the Trial Burn and Risk Burn parameters, thus eliminating one test condition in the Trial Burn as well as the annual operating limits in the permit.
The last methodology considered is the Demonstration of Similarity for combustion units. Numerous facilities across the United States have more than one on-site combustion unit. An approach is taken to prepare a Demonstration of Similarity for multiple combustion units, which may or may not be identical, thereby, eliminating any testing for the similar unit.
INTRODUCTION
EPA Region V (hereafter, EPA) requested a RCRA Part B Permit modification for three hazardous waste burning boilers located at the Reilly Industries, Inc. facility in Indianapolis, Indiana in February 1998. The three boilers associated with the permit modification are referred to as Boiler 28K, Boiler 30K, and Boiler 70K. The EPA required the facility to establish limitations on operations pursuant to the requirements of 40 CFR Part 266 (Boiler and Industrial Furnace Regulations) and Part 270 (Part B Permit Regulations), and to demonstrate acceptable risk using their omnibus authority.
The addition of a site-specific Human Health and Ecological Risk Assessment (hereafter, Risk Assessment) to a traditional Trial Burn has many implications for a facility and the Permit Writer. The Risk Assessment greatly increases the amount of emissions testing required to be performed and the amount of data to be evaluated by the Permit Writer. The data used in the Risk Assessment has typically been gathered during a separate test condition under “normal” operating conditions. In addition to the Trial Burn and Risk Assessment testing, interim status boiler and industrial furnace (BIF) facilities must continue to comply with the recertification of compliance (RCOC) testing every three years until a permit is issued.
The emissions testing associated with the Trial Burn, Risk Assessment, and RCOC is the largest and most expensive component of the permitting process. The following sections present several opportunities for cost savings and how they can be implemented. These opportunities include combining a RCOC and Trial Burn, minimizing the number of test conditions associated with a Trial Burn and Risk Assessment, and developing a Demonstration of Similarity for two non-identical combustion units.
COMBINED RCOC AND TRIAL BURN
Interim status BIFs are required to perform periodic recertifications pursuant to the provisions of 40 CFR 266.103(d), Periodic Recertifications. These periodic recertifications are required to be performed within three years of submitting the previous certification or recertification. Conducting the RCOC testing is periodic and is typically planned well in advance.
The issuance of the RCRA Part B Permit application call-in letters or requests for a permit modification generally cannot be predicted with any degree of accuracy. Agency schedules are constantly being changed and the information available today may not be current tomorrow. Also, the receipt of an application request cannot be used to reliably define the schedule of a Trial Burn due to the sometime lengthy notice of deficiency (NOD) resolution process involved with obtaining approval of the Trial Burn Plan, Quality Assurance Project Plan (QAPP), Air Dispersion Modeling (ADM), and Risk Assessment Protocol (RAP). Once these plans have been approved or are near approval, scheduling of the testing can proceed with increased confidence.
However predictable the RCOC scheduling and unpredictable the Trial Burn scheduling first appear, there is usually an opportunity to plan the testing to occur simultaneously. The RCOC provisions do not require a facility to wait for the end of the three-year period prior to conducting the next RCOC test. The primary reason for waiting the full three-years is to make the previous testing as cost effective as possible. However, moving the RCOC testing up in time may save money by not incurring duplicate testing costs such as mobilization and demobilization of test equipment, identical emissions sampling and analysis, test planning, report generation, and the efforts of the on-site personnel.
Extending the RCOC testing beyond the 3 year period is also an option the facility should consider. The decision to extend the time period for the RCOC is a regulatory decision requiring an approval by the appropriate Agency. Requests for an extension of time are made pursuant to 40 CFR 266.103(c)(7)(ii), Extensions of Time. The facility should base their extension request on the unpredictable timing of the Trial Burn, the close timing of the RCOC and Trial Burn, and continued operations under the current RCOC will not pose a hazard to human health or the environment. The extension request should be detailed enough so that the Agency can provide prompt approval without requesting additional information, and should provide a thorough discussion on the permitting process and expectations for testing. A good working relationship with the Permit Writer is essential as with any permitting process.
RCOC Test Protocol and QA/QC Plan
Once a decision has been made to combine the Trial Burn and RCOC, the facility must plan and implement the project to satisfy both sets of requirements. Specifically, even though a Trial Burn Plan and QAPP are being developed, the facility must continue to comply with the BIF requirements for a RCOC. Therefore, as required by 40 CFR 266.103(c)(2), Prior Notice of Compliance Testing, the 30-day prior notice of testing must be submitted and must include the testing protocol and a Quality Assurance/Quality Control (QA/QC) plan. However, compliance with these requirements may provide the facility with another cost savings opportunity as follows. The Trial Burn Plan and QAPP must be developed and approved by the Agency prior to conducting the Trial Burn. If agreed to by the Agency, an abbreviated notification can be provided that addresses the RCOC items not addressed by the Trial Burn Plan and includes references to the appropriate sections of the Trial Burn Plan and QAPP for the remaining requirements.
The requirements for the 30-day notification of testing are presented in 40 CFR 266.103(c)(2), Prior Notice of Compliance Testing. The abbreviated notification must ensure that all of the required information is addressed and, for the information included in the Trial Burn Plan or QAPP, proper references to the information must be made. An important item to not overlook is whether the current RCOC limits will be exceeded during the combined RCOC and Trial Burn. If the Trial Burn is designed to exceed the current RCOC limits, a notification requesting 720-hours of operational time to establish the new limits must be submitted pursuant to 40 CFR 266.103(c)(8), Revised Certification of Compliance. If this notification is not made in a timely manner, the facility will technically be out of compliance with the BIF regulations during the combined Trial Burn/RCOC testing. The following items should be addressed in the abbreviated notification:
Cost Savings
Combining the RCOC test with the Trial Burn can result in significant cost savings for facilities and the EPA. Cost savings are realized by eliminating duplicative items including stack sampling and analytical, material spiking services, test plan and QAPP generation, test reporting, on-site personnel efforts, and contractor oversight and document review.
Reilly Industries, Inc. recently combined the RCOC with a Trial Burn that was completed in November 1999. The savings associated with eliminating the duplication of effort are estimated at approximately $250,000 for three boilers. The Agency was also able to maximize the use of their resources in a manner that was beneficial to the welfare of the public. Furthermore, the Agency was able to provide on-site oversight of the RCOC testing, something that typically is not available due to the self-implementing nature of the BIF interim status requirements. Therefore, cost savings can be realized by combining the RCOC with the Trial Burn.
MINIMIZING TRIAL BURN/RISK BURN TEST CONDITIONS
The addition of a Risk Assessment to a traditional Trial Burn greatly increases the amount of emissions testing required for the test. The data used in the Risk Assessment is usually gathered during a separate test condition under “normal” operating conditions. Any minimization in the number of test conditions has the opportunity to result in significant cost savings for a facility. The following sections present a summary of the typical parameters associated with a Trial Burn and Risk Burn, several options for combining test conditions, and potential cost savings.
Trial Burn Test Conditions and Parameters
Trial Burns are usually designed with two test conditions. One test condition is typically conducted at high temperature, maximum feed rates (i.e., waste, ash, chlorine, etc.), maximum emissions (i.e., particulate matter, hydrogen chloride, chlorine gas, and metals), and maximum operating conditions (i.e., gas velocity, production rate, etc.). A second test condition is typically conducted at minimum temperature and maximum emissions (i.e., POHC, CO, etc.).
Trial Burn testing is designed at the extremes of the operating envelope of a combustion device. Testing at such extreme operating conditions theoretically validates operations at any point between the extremes.
The permit limits associated with the combustion operations during the Trial Burn are established on a short-term basis, such as instantaneous limits or hourly rolling average limits. Some metal limits can be established on a rolling average basis of 2-hours up to 24-hours.
Table I presents the parameters typically associated with a hazardous waste burning boiler and the Trial Burn test condition during which they are established.
Table I
Establishment of Trial Burn Parameters
|
Parameter |
Maximum Temperature |
Minimum Temperature |
|
POHC Incinerability (DRE Test) |
|
X |
|
Chlorine Feed Rate, Maximum |
X |
|
|
Ash Feed Rate, Maximum |
X |
|
|
Metals Feed Rate, Maximum |
X |
|
|
Combustion Temperature, Minimum |
|
X |
|
Combustion Temperature, Maximum |
X |
|
|
Combustion Gas Velocity, Maximum |
X |
X |
|
Waste Feed Rate, Maximum |
X |
|
|
Production Rate, Maximum |
X |
|
|
Production Rate, Minimum |
|
X |
|
CO Concentration, Maximum |
X |
X |
|
Particulate Matter Emissions |
X |
|
|
HCl/Cl2 Emissions |
X |
|
|
Metal Emissions |
X |
|
Risk Burn Test Conditions and Parameters
Risk Burns are usually designed with one test condition. This test condition is typically conducted at “normal” operations (i.e., temperatures, feed rates, emissions, and operating conditions). The “normal” operations during the Risk Burn are intended to represent the operations over a long period of time. The Risk Assessment addresses the direct and indirect risks associated with the operations of the unit over a 30-year period.
Emissions testing during a Risk Burn can include the following parameters:
The permit limits associated with “normal” combustion operations are established on a long-term basis. Such long-term limits can be established as monthly, quarterly, semi-annual, or annual limits. Typical long-term parameter limits may include maximum waste feed rate, minimum temperature, and maximum combustion gas velocity (1). Given the nature of a long-term limit, a unit might be forced to shutdown if the limit on a maximum value is being approached. However, if the limit is being approached on a minimum value, compliance cannot be achieved by shutting down the unit because, in theory, operations would need to be increased. This presents a complex issue that must be resolved between the facility and the Permit Writer.
Options for Combining Test Conditions
To minimize testing and reduce costs associated with a Trial Burn, the option of combining the Risk Burn testing with the Trial Burn testing at maximum conditions should be examined. Boilers typically operate at or near their maximum operating range since this is the most efficient means of producing steam and the steam is used to maintain production processes. The maximum end of their operating range is also at conditions targeted for the Trial Burn. Combining the Risk Burn with the maximum feed rate test condition is logical because the maximum boiler operations targeted for the Trial Burn are usually within 20 percent of the targeted “normal” operations.
Because the Risk Assessment would utilize the data generated during the Trial Burn testing, spiking of compounds to achieve higher feed rate limits had to be closely scrutinized. For example, if chlorine or metals were spiked during the Risk Assessment testing, an adverse affect on the emissions data for the Risk Assessment could occur. However, for hazardous waste burning BIFs, the tiered limits are allowed for the metals and chlorine. Adjusted Tier I feed rate screening limits could be established for the metals and chlorine without performing emissions testing or spiking.
A problem with the Adjusted Tier I feed rate screening limits is that it is assumed that the entire amount fed to the boiler partitions to the combustion gas. For the Reilly boilers, this equates to direct emissions from the stacks since there are no air pollution control devices on the boilers. Therefore, the Adjusted Tier I feed rate screening limit represents the emissions for chlorine and the metals. It was assumed that if the Adjusted Tier I feed rate screening limits were used as an input to the Risk Assessment, unacceptable risk would occur. The basis for this assumption was that the Adjusted Tier I limits were established using a Risk Assessment model which only accounted for the direct exposure routes. The current Risk Assessments assess direct and indirect exposure routes. Because of this, it was assumed that the Adjusted Tier I limits would cause unacceptable risk.
A solution to this problem was to use the Adjusted Tier I feed rate screening limits as a starting point in the Risk Assessment and perform iterations on reduced feed rate limits for the chlorine and metals until the model achieved an acceptable risk. This approach was proposed to, and accepted by, the Agency and was used to establish the feed rate limits for each of the metals and chlorine. This approach offers several advantages as follows:
A second problem was encountered when the issue of long-term operating limits was raised by the Agency. The Agency indicated that long-term limits would be established for the maximum waste feed rate and the minimum combustion temperature. Establishing the maximum waste feed rate during the maximum operations test condition was not a problem. However, establishing a long-term limit on the minimum combustion chamber temperature during this same (high feed rate) test condition was not acceptable. An option that was considered pertained to eliminating the long-term limits by performing the Risk Assessment using the data obtained at both the maximum and minimum operating test conditions. This option was determined to be acceptable to the Agency. Therefore, dioxin/furan testing occurred during both test conditions because this was the primary Risk Assessment parameter of concern to the Agency. In addition, the Agency required the use of a total hydrocarbon analyzer during both test conditions to verify consistent total organic emissions. Therefore, based on Agency input and the additional testing, the long-term permit operating limits were not necessary and the minimum temperature limit was established during the DRE test condition.
Performing dioxin/furan testing during the minimum temperature test condition simultaneous with the DRE determination can also cause a problem if the DRE testing involves the spiking of a chlorinated POHC. If a chlorinated POHC is spiked into the waste stream so that the DRE can be determined, it has the potential to artificially increase the dioxin/furan emissions, which affects the Risk Assessment. This situation must be considered if dioxin/furan emissions are of a significant concern.
Cost Savings
Combining the maximum operations test condition of the Trial Burn with the Risk Burn test condition can result in significant cost savings for facilities. Cost savings are related to eliminating a test condition and thus reducing the stack sampling and analytical, test plan and QAPP generation, test reporting, and on-site personnel efforts. Additional cost savings result from the avoidance of long-term permit limits. The elimination of any permit limit will result in cost savings for a facility. Long-term limits can be very costly due to the programming of a data acquisition system to obtain and calculate the data to demonstrate compliance, and for the personnel who review and report the results to the Agency.
Reilly Industries, Inc. completed a Trial Burn in November 1999 with the combined test conditions. Costs were incurred due to the additional dioxin/furan and total hydrocarbon testing. The savings associated with eliminating a test condition, eliminating metals and chlorine emissions testing, and eliminating the long-term permit limits are estimated at approximately $150,000 for each boiler. Therefore, real cost savings can be realized by combining the Risk Burn test condition with a Trial Burn test condition.
DEMONSTRATION OF SIMILARITY FOR TWO NON-IDENTICAL UNITS
Reilly Industries, Inc. operates three hazardous waste burning boilers. These boilers are referred to as Boiler 28K, Boiler 30K, and Boiler 70K. These identifications are based on the design steam production capacity of the boilers. Boilers 28K and 30K were both manufactured by the Babcock and Wilcox Company, and were the identical model number and boiler type. Because of the similar sizing of the two boilers and the fact that the hazardous waste fed to each is identical, it was possible to demonstrate the similarity of the boilers and avoid the costs associated with testing both units. The following sections describe the process used to demonstrate the similarity of these non-identical units.
Regulatory Requirements
A review of the regulatory requirements was performed to determine the information required for demonstrating similarity on two units. The following regulations and guidance provided information on demonstrating similarity:
The regulatory review indicated that four primary areas needed to be addressed in order to demonstrate similarity. These four areas included the comparison of the feed streams, design, operating conditions, and maintenance. Each of these areas is addressed in the following sections.
Feed Stream Comparison
A comparison of the feed streams to each boiler was conducted. The first step was to define the streams that are burned in each boiler. Then each stream could be compared to determine their similarity.
In Reilly’s case, each boiler is capable of burning two streams during hazardous waste combustion operations. These two streams are identified as the hazardous waste fuel and the city gas (i.e., natural gas) auxiliary fuel.
The hazardous waste feed stream is generated from production processes at the facility and is stored in four tanks prior to being fed to either boiler. All hazardous waste feed streams for either boiler must originate from these four tanks. The boilers burn the exact same hazardous waste feed streams interchangeably. Therefore, there was no reason to compare the physical and chemical properties of the feed streams between the two boilers because they are one in the same. However, if the feed streams were not exactly identical, a comparison of the following properties would have been performed: viscosity, physical form, heat content, antimony, arsenic, barium, beryllium, cadmium, chromium, lead, mercury, silver, thallium, total chlorine/chloride, ash content, and the 40 CFR Part 261 Appendix VIII hazardous organic constituents.
The city gas feed stream is supplied to the entire facility from the Citizens Gas and Coke Utility. The city gas is conveyed to and throughout the facility by a pipeline, which supplies the boilers. The city gas is obtained from the same source and is in all ways identical for each boiler. Therefore, because the city gas is the same, there was no reason to compare the physical and chemical properties for each boiler. However, if the auxiliary fuel streams were not exactly identical, a comparison of the following properties would have been performed: viscosity, physical form, heat content, antimony, arsenic, barium, beryllium, cadmium, chromium, lead, mercury, silver, thallium, total chlorine/chloride, ash content, and the 40 CFR Part 261 Appendix VIII hazardous organic constituents.
Based on the above information, the hazardous waste feed streams and the auxiliary fuel streams to each boiler were shown to be identical. Therefore, the feed streams to the boilers were demonstrated to be similar.
Design Comparison
The design comparison of the boilers addressed each of the following components: the boilers, the feed systems associated with the boilers, the combustion air system, the Air Pollution Control System (APCS), the operational controls, and the Automatic Waste Feed Cut-Off (AWFCO) system. Appendices were included that contained equipment information related to each boiler, such as design specifications, vendor manuals, etc.
The design comparison of the boilers consisted of comparing the specific design criteria associated with each boiler. Table II summarizes the specific criteria that were compared.
|
Parameter |
Boiler 28K |
Boiler 30K |
|
Manufacturer |
Babcock and Wilcox |
Babcock and Wilcox |
|
Model Number |
Type FM |
Type FM |
|
Boiler Type |
Water Tube |
Water Tube |
|
Manufacture Date |
1959 |
1964 |
|
Design Capacity – Heat Release |
10.79 W (36.8 MM Btu/hr) |
11.52 W (39.3 MM Btu/hr) |
|
Design Capacity – Steam Production |
3.53 Kg/s @ 1724 KPa (28,000 lb/hr @ 250 psig) |
3.78 Kg/s @ 1724 KPa (30,000 lb/hr @ 250 psig) |
|
Linear Dimensions |
5.49 m (18’) |
4.67 m (15’4”) |
|
Cross-Sectional Area |
2.54 m x 1.83 m (8’4” x 6’0”) |
2.57 m x 1.93 m (8’5 ¼” x 6’4”) |
|
Volume |
22.96 m3 (811 CF) |
22.09 m3 (780 CF) |
|
Materials of Construction |
Carbon Steel |
Carbon Steel |
|
Heat Transfer Area |
288.65 m2 (3107 ft2) |
305.74 m2 (3291 ft2) |
The Boiler 30K design capacity is approximately 6.5 percent greater than Boiler 28K with a heat transfer area of approximately 5.6 percent larger. The boilers were built by the same manufacturer and are of the same type, model number, and material of construction. It should be noted that the internal volume of Boiler 28K is slightly larger than that of Boiler 30K. This is of significance later as the residence time is shown to be greater in Boiler 28K than in Boiler 30K. This indicates that the combustion of hazardous waste in Boiler 28K should be to a higher degree of efficiency than that of Boiler 30K. Therefore, by applying similarity to Boiler 28K and using worst case data from Boiler 30K, a margin of safety is implied.
The design comparison of the feed systems consisted of comparing the design information associated with each boiler. During hazardous waste operations there are only two feed streams to the boilers. These feed streams are the hazardous waste fuel and auxiliary fuel. Table III summarizes the design information that was compared on the two feed streams and their associated feed systems for each boiler.
The hazardous waste feed streams are liquids supplied from the same tanks and atomized by the same methodology for each boiler. The auxiliary fuel feed stream is city gas that is supplied from the same source with identical burner designs.
|
Feed Stream |
Parameter |
Boiler 28K |
Boiler 30K |
|
Hazardous Waste |
Matrix |
Liquid |
Liquid |
|
Source |
Tanks 64, 65, 66, and 69 |
Tanks 64, 65, 66, and 69 |
|
|
Burner Design |
Atomized with Sprayer Plate |
Atomized with Sprayer Plate |
|
|
Atomization |
Steam |
Steam |
|
|
Capacity |
0.25 Kg/s (1,960 lb/hr) |
0.26 Kg/s (2,100 lb/hr) |
|
|
Auxiliary Fuel |
Type |
City Gas (Natural Gas) |
City Gas (Natural Gas) |
|
Matrix |
Gaseous |
Gaseous |
|
|
Source |
Citizens Gas and Coke Utility |
Citizens Gas and Coke Utility |
|
|
Burner Design |
Burner Ring |
Burner Ring |
|
|
Atomization |
None |
None |
|
|
Capacity |
21.75 m3/s (46,090 CFM) |
23.31 m3/s (49,400 CFM) |
The design comparison of the combustion air systems consisted of comparing the design information associated with each boiler. Both boilers utilize a combustion air fan that functions as a forced draft fan. Neither boiler has an induced draft fan. Table IV summarizes the design information that was compared on the combustion air systems for each boiler.
|
Parameter |
Boiler 28K |
Boiler 30K |
|
Manufacturer |
Green Fuel Economizer Co., Inc. |
Clarage Fan Company |
|
Model No. |
Arrangement 4 Diffuser Fan |
Uni-Combustion Fan |
|
Type |
Centrifugal |
Centrifugal |
|
Capacity |
3.95 m3/s (8,369 CFM) |
4.23 m3/s (8,970 CFM) |
|
Use |
Forced Draft |
Forced Draft |
The combustion air fans were manufactured by different companies, but are of the same type (centrifugal) and use (forced draft). The combustion air fan capacity of Boiler 30K is approximately 6.7 percent larger than the combustion air fan of Boiler 28K. Based on this information, the combustion air systems for the boilers were deemed to be similar.
Neither boiler has any type of APCS installed. Since neither boiler has an APCS, information on the basic design or critical design specifications is not applicable or provided and thus is, by definition, similar. However, if APCS equipment had been installed, a comparison of the equipment type and associated design and performance data would have been performed to determine their similarity.
The operations of each boiler are continuously monitored using several instruments to monitor temperatures, flow rates, and stack gas concentrations. The following paragraphs discuss the instruments used to monitor the boiler operations as well as the associated computer systems. Each of these instruments are identical to each boiler and are, by definition, similar.
The feed rate of hazardous waste to each boiler is continuously monitored by use of mass flow meters that use the Coriolis effect for measuring the mass flow rate. The mass flow meters are located at each boiler in the feed line prior to the flow control valve and burner gun and downstream of the feed pumps.
The auxiliary fuel flow rate to each boiler is continuously monitored by use of mass flow meters that use the cooling effect of the flow medium (city gas) across thermal elements to determine the rate of flow. The auxiliary fuel mass flow meters are located in the city gas supply line to each boiler upstream of the flow control valve and downstream of other users of city gas.
The combustion chamber temperature of each boiler is continuously monitored by use of Type K thermocouples located on the back wall of the firebox. Two thermocouples are used on each boiler. Each thermocouple operates in conjunction with a temperature transmitter, which forwards the temperature signal to the Distributed Control System (DCS).
The steam production rate for each boiler is measured using vortex flow meters. The vortex flow meters are located in the steam header piping from each boiler prior to entering the main plant steam header system.
Each boiler has a continuous emissions monitoring system (CEMS) installed in their stacks. The CEMS analyze the stack gases on a continuous basis for carbon monoxide (CO) and oxygen (O2) content. The CO analyzer for each boiler is of the same make and model, analysis method, and span. The O2 analyzer for each boiler is of the same make and model, analysis method, and span. The analyzer sampling points for each boiler are located in the stack at least two equivalent stack diameters downstream from a point where changes in pollutant concentration or emission rate occurs and at least one-half an equivalent stack diameter upstream of the stack exhaust.
Table V provides a summary of the process monitoring devices for the two boilers including flow rate, temperature, and emission monitoring devices.
|
Operating Parameter |
Monitoring Device |
Boiler 28K Tag No. |
Boiler 30K Tag No. |
|
Hazardous Waste Feed Rate |
Mass Flow Meter |
28OILNUM |
30OILNUM |
|
Auxiliary Fuel Feed Rate |
Mass Flow Meter |
28CGSNUM |
30CGSNUM |
|
Combustion Chamber Temperature |
Type K Thermocouple |
28TMPNUM |
30TMPNUM |
|
Steam Production Flow Rate |
Vortex Flow Meter |
28STMNUM |
30STMNUM |
|
Stack Gas O2 Concentration |
Paramagnetic O2 Analyzer |
AT2804 |
AT3004 |
|
Stack Gas CO Concentration |
NDIR CO Analyzer |
AI2805 |
AI3005 |
Each boiler is connected to a computer system that controls, monitors, and records the boiler operations. The primary computer system for monitoring and controlling operations for each boiler is the Honeywell TDC3000 DCS. The Honeywell manually and/or automatically controls each boilers operations based on setpoints input by the Boiler Operators and by algorithms programmed into the logic system. The Honeywell also includes the programming for the AWFCO system and provides the documentation of the operations. A single Honeywell DCS unit is used to control, monitor, and record the operations of both boilers.
The AWFCO system for each boiler automatically cuts off the hazardous waste feed prior to an applicable operating condition exceeding its associated limitation. In addition, the AWFCO system does not allow waste feeds to be initiated unless all the applicable limitations are being satisfied. The AWFCO system consists of a DCS and automatic block valves located on the hazardous waste feed lines of each boiler directly upstream of the burner gun and downstream of the mass flow meter for each boiler. These valves operate as a normally closed valve, which means that it must receive a signal from the DCS to allow it to open. Any loss of power to the boiler or to the DCS will cause the valve to fail closed, thereby stopping the feeding of the hazardous waste. Table VI provides a listing of the AWFCO limitations in place at the time of the Demonstration of Similarity. These limits were established based on the Certification of Compliance (COC) testing conducted in June 1996.
|
Parameter |
Boiler 28K |
Boiler 30K |
Percent Difference |
|
Hazardous Waste Feed Rate, Max, HRA |
0.273 Kg/s (2164 lb/hr) |
0.279 Kg/s (2214 lb/hr) |
2.26% |
|
Combustion Chamber Temperature, Max, HRA |
889°C (1633°F) |
939°C (1722°F) |
5.17% |
|
Combustion Chamber Temperature, Min, HRA |
616°C (1140°F) |
638°C (1181°F) |
3.47% |
|
Steam Production Rate, Max, HRA |
3.816 Kg/s (30.29 Klb/hr) |
4.188 Kg/s (33.24 Klb/hr) |
8.88% |
|
Stack Gas CO @ 7% O2, Max, HRA |
75 ppmv |
75 ppmv |
0% |
The design comparison of Boilers 28K and 30K included a review of each of the following systems: boiler system, feed systems, combustion air system, APCS, operational controls, and the AWFCO system. Each of these systems was demonstrated to have a similar design.
Operating Condition Comparison
The operating condition comparison of the boilers consisted of comparing the operating data and the stack emissions data obtained during the last COC test performed on the boilers. This COC report, dated August 21, 1996, was previously submitted to EPA Region V and the Indiana Department of Environmental Management. The residence time values were calculated using the data from the COC report and the volume of each boiler. Table VII summarizes the operating conditions that were compared for each boiler.
|
Operating Parameter |
Boiler 28K |
Boiler 30K |
Percent Difference |
|
Hazardous Waste Feed Rate, Max. |
0.273 Kg/s (2170 lb/hr) |
0.282 Kg/s (2236 lb/hr) |
2.95% |
|
Auxiliary Fuel Feed Rate, Max. |
0.048 Kg/s (383 lb/hr) |
0.044 Kg/s (350 lb/hr) |
- 9.43% |
|
CO @ 7% O2, Max. |
< 75 ppmv |
< 75 ppmv |
0% |
|
Steam Production Rate, Max. |
3.906 Kg/s (31.00 Klb/hr) |
4.247 Kg/s (33.71 Klb/hr) |
8.04% |
|
Maximum Combustion Temperature, Max. |
895°C (1643 ºF) |
954°C (1749 ºF) |
6.06% |
|
Minimum Combustion Temperature, Min. |
612°C (1134 ºF) |
638°C (1181 ºF) |
3.98% |
|
Stack Gas Velocity, Avg. |
7.29 m/s (1436 ft/min) |
6.71 m/s (1321 ft/min) |
- 8.71% |
|
Stack Gas Temperature, Avg. |
243°C (469 ºF) |
211°C (411 ºF) |
- 14.11% |
|
Stack Gas Flow Rate, Avg. |
3.88 dscm/s (8221 dscfm) |
4.01 dscm/s (8498 dscfm) |
3.26% |
|
Stack Gas Flow Rate, Avg. |
8.34 acm/s (17,676 acfm) |
7.75 acm/s (16,426 acfm) |
- 7.61% |
|
Residence Time, Min. |
1.19 seconds |
1.10 seconds |
- 8.18% |
The operating conditions presented in Table VII for Boilers 28K and 30K are very similar. The two parameters with the largest percent differences are the auxiliary fuel feed rate and the stack gas temperature. Both of these parameters are not regulated and would not have associated permit limits. All of the remaining parameters are shown to have differences of less than nine percent. Based on the above information, it has been demonstrated that the operations of Boilers 28K and 30K are similar.
As mentioned earlier, the residence time for Boiler 28K is greater than that of Boiler 30K. This in theory allows for better combustion in Boiler 28K. Therefore, applying emissions data from Boiler 30K to Boiler 28K should provide a level of conservatism to the Risk Assessment. As also demonstrated in Table VII, the stack gas flow rates and exhaust temperatures were greater for Boiler 28K than those for Boiler 30K. This is significant in that Boiler 28K actually has better dispersion characteristics than Boiler 30K. Therefore, using the Boiler 30K emissions data in-lieu of data from Boiler 28K provides more conservatism to the Risk Assessment, helping to further assure the EPA and the public that emissions from Boiler 28K would not present an unacceptable risk to human health and the environment.
Maintenance Comparison
The maintenance comparison of the boilers addressed each of the following: daily inspections, daily CEMS calibrations and audits, quarterly and annual CEMS testing, AWFCO testing, and annual boiler inspections.
At a minimum, the boilers are visually inspected by the Boiler Operators on a daily basis for leaks, spills, fugitive emissions, and signs of tampering. Documentation of inspections is maintained in the facility operating records. Items noted as unacceptable during these inspections can be handled in several ways depending on the severity of the item. For catastrophic problems, waste feeds and boiler operations may be shutdown for repair. Minor items may be repaired by the Boiler Operators or are turned over to Maintenance personnel for repair by use of a work order system.
The CO and O2 analyzers on each boiler are calibrated on a daily basis. In addition, a daily system audit is performed that includes a review of the calibration data and inspections of the recording system, control panel lights, and sample transport and interface systems. Quarterly calibration error testing is performed on each analyzer. Annual performance specification testing is also performed on the analyzers.
The AWFCO system on each boiler is tested on a monthly basis. Monthly testing is conducted in accordance with a demonstration, maintained in the operating record, that weekly testing unduly restricts or upsets boiler operations and that monthly testing is sufficient.
In accordance with State regulations and facility insurance requirements, the boilers are thoroughly inspected annually and issued a certificate for operation. Any significant safety or operational problems noted during this inspection require repair before the boilers can be returned to service.
Both boilers are subjected to identical types of inspections, calibrations, and testing on specified frequencies. Therefore, the maintenance of Boilers 28K and 30K were demonstrated to be similar.
Agency Discussions
Based on the results of comparing the feed streams, design, operation, and maintenance, the boilers were determined to be similar. A document demonstrating this similarity was developed and submitted to the Agency as a portion of the RCRA Part B Permit application.
The Agency performed a review of the Demonstration of Similarity document and agreed with the findings. In order to provide another layer of safety and conservatism to those already presented, the Agency requested that the permit limits for Boiler 28K be three percent lower than the limits for Boiler 30K based on sizing differences. This was agreeable to Reilly Industries, Inc. and will be incorporated in the final permit.
Cost Savings
Demonstrating similarity for two boilers that are not identical can result in significant cost savings. The cost savings are realized by eliminating the Trial Burn and RCOC testing for the similar boiler. The additional costs incurred for demonstrating similarity relate to the Demonstration of Similarity document and discussions with the Agency.
Reilly Industries, Inc. developed a Demonstration of Similarity for Boiler 28K and Boiler 30K that received approval from EPA Region V in October 1999. The costs incurred for developing the Demonstration of Similarity and discussions with the Agency amounted to approximately $10,000. The savings associated with eliminating a Trial Burn and RCOC testing on the similar boiler is estimated at approximately $500,000. Therefore, real cost savings can be realized by demonstrating similarity for two boilers that are similar although they may not be exactly identical. These savings should also perpetuate into the 3 year RCOC testing (if needed) and 10 year RCRA permit renewal requirements.
SUMMARY
This paper presented three opportunities where cost savings can be realized when a RCRA Part B Permit application or permit modification for a hazardous waste combustion device is requested that requires a Risk Assessment. These opportunities include combining a RCOC and Trial Burn, minimizing the number of test conditions associated with a Trial Burn, and developing a Demonstration of Similarity for combustion units. Overall, implementing these opportunities resulted in net savings of over $1,000,000 of the facility’s operating expenses. Savings realized by the EPA and the flexibility provided to the facility occurred, but have not been quantified. Ultimately, the reduction in the burdens typically associated with Trial Burns provided for a more conservative and safe assessment of the risk associated with burning hazardous waste.
REFERENCES
1. U.S. EPA, “Human Health Risk Assessment Protocol for Hazardous Waste Combustion Facilities”, Volume One, Peer Review Draft, EPA530-D-98-001A, July 1998.
2. U.S. EPA, “Technical Implementation Document for EPA’s Boiler and Industrial Furnace Regulations”, EPA530-R-92-011, March 1992.
[1] As presented at the 2000 International Conference on Incineration and Thermal Treatment Technologies, Portland, Oregon, May 8-12, 2000.