Evaluation of the Listing of Catalytic Reforming Operations at Petroleum Refineries as a Source Activity of Dioxins and Furans Reporting to the National Pollutant Release Inventory
NPRI Multi-stakeholder Work Group on Substances
D-F and HCB Sub-group (SG) made up of the following individuals -
Alan Penn (Grand Council of the Crees)
Allan Mumby (CWWA representative)
Anne Legault (Environment Canada - NPRI)
Francois Lavallee (Environment Canada - NPRI)
Jim White (ENGO representative)
John Jackson (ENGO representative)
Kal Virk / Adolfo Silva (CPPI representative)
Richard Martin (Health Canada)
Wilfrid Jan – (SG Chair - Environment Canada - NPRI)
Evaluation for the Listing of Petroleum Refining as a Source Activity of Dioxins/Furans Reporting for the National Pollutant Release Inventory
Based on the technical study literature and the U.S. Toxics Release Inventory, petroleum refining is considered a potential source activity of Dioxins/Furans (D-F) releases to the environment. Specifically, the D-F releases would arise from the regeneration of spent catalyst from the reforming process at petroleum refineries.
The reforming process occurs at high temperature and pressure and requires the use of a platinum or platinum/rhenium catalyst. During the reforming process, a complex mixture of aromatic compounds, known as coke, is formed and deposited onto the catalyst resulting into a decrease of its activity. The high cost of the catalyst necessitates its regeneration. Catalyst regeneration is achieved by removing the coke deposits via burning at temperatures of 399 to 454 °C and then reactivating the catalyst at elevated temperatures (454 to 538 °C) using chlorine or chlorine compounds. Burning of the coke produces flue gases that can contain D-F along with other combustion products. In addition, internal effluents might become contaminated with D-F since the flue gases could be scrubbed with caustic or water if not vented directly to the atmosphere.
2. Review of Existing Information
An U.S. EPA study report, An Inventory of Sources and Environmental Releases of Dioxin-like Compounds in the United States for the Years 1987, 1995, and 20001, has provided a summary of relevant studies performed to date, and the status of the U.S. EPA investigations concerning petroleum refining catalyst regeneration and the occurrences of D-F. This information can be found in Section 5.4 of the Report, and in which some pertinent Canadian information has been cited:
- Thompson et al. (1990)2 reported total dioxins and furans concentrations of 8.9 ng/m3 and 210 ng/m3, respectively, in stack gas samples from a Canadian petroleum refinery's reforming operation. They also observed D-F in the picogram- to nanogram-per-liter range in the internal washwater from a scrubber of a periodic/cyclic regenerator.
The Report also presents information for the D-F emission factors from the petroleum catalytic reforming units, on a category basis as well as on a congener basis. However, the emission factors information is being assigned a low confidence rating.
From the U.S. Toxics Release Inventory database, it is found that 63 petroleum refineries have reported the D-F data. Unlike the reporting requirements for NPRI, the U.S. facilities would submit the D-F data only if they have met the annual M, P, O threshold of 0.1 gram prescribed for the substance category. According to the 2004 data retrievals, the petroleum refining sector would account for 2% of the total contributions of the air releases, which would put it among the top sectors responsible for the largest air releases of D-F (see Appendix A). Based on the available U.S. reported data, as well as the estimated ratio between the number of refineries in the U.S. and those in Canada, one could extrapolate that some 2.5 grams of D-F might possibly be released to air from the Canadian petroleum refineries. While this "projected" information is not directly comparable to the Canadian data which are expressed in g-ITEQ, it is nevertheless interested to observe that the petroleum refining sector would be ranked in the 8 th spot (after cement manufacturing) as one of the top contributors of the air releases of D-F in Canada (see Appendix B).
From what is described in the Dioxins and Furans Canada-Wide Standards (CWS) Status Report (October, 2004)3, it seems that the Canadian Council of the Ministers of the Environment (CCME) does not consider the D-F formation and release at petroleum refineries, compared to criteria air contaminants (CACs) and benzene, to be a high priority of concerns for assessment. There is no indication that the D-F from the petroleum refining processes would be thoroughly examined within the CWS framework.
It is necessary to address and resolve the following discussion points in formulating the recommendations for the issue:
- What is the intent to list the "regeneration of spent catalyst at petroleum refineries" as a source activity of D-F reporting for NPRI? Is the reporting need justified?
- Is there sufficient information available to allow facilities to estimate the D-Freleases from the "regeneration of spent catalyst at petroleum refineries" on a substance category basis? On a specific congener basis?
- Is there justification to add the same source activity for the Hexachlorobenzene (HCB) reporting, considering that noHCB reports are identified in the U.S. Toxics Release Inventory? If the addition is warranted, would there be enough information available to support the capability of the data estimation of HCB releases from this source activity?
In order to be certain that the emission scenarios behind the D-F reporting for the U.S. petroleum refineries would be no different in Canada, the SG has decided to conduct a mini-survey to gather relevant information regarding the catalytic reformer process operations at Canadian refineries. The SG will be relying on the survey results to facilitate the determination on whether to include the source activity of "regeneration of spent catalyst in petroleum refineries" to the list for D-F reporting to the NPRI.
The top sectors for the D-F data (2004) in the U.S., in terms of the quantities reported for the total air releases:
|Primary SIC Code and Description||# of Facilities||On-site Fugitive Air (grams)||On-site Point Source Air (grams)||Total Air Releases (grams)||% of Total Air Releases|
|4939||Combination Utilities, NEC||6||0||157.54||157.54||12.88%|
|2812||Alkalies and Chlorine||14||55.28||27.90||83.17||6.80%|
|3312||Steel Works, Blast Furnaces||16||0.15||59.30||59.45||4.86%|
|3341||Secondary Smelting and Refining of Nonferrous Metals||53||2.37||46.10||48.47||3.96%|
|2821||Plastics Material and Synthetic Resins||20||9.22||19.71||28.93||2.37%|
D-F data (2004) in Canada, in terms of the quantities reported for the total air releases by sectors in descending order:
|NAICS Code||NAICS Code Description||# of Facilities Reported||Air Releases (g-ITEQ)||Percentage Contribution|
|562210||Waste Treatment & Disposal||43||46.573||56.3%|
|331110||Iron & Steel Mills & Ferro-Alloy Mfg.||12||8.623||10.4%|
|221112||Fossil-Fuel Electric Power Generation||27||7.381||8.9%|
|322112||Chemical Pulp Mills||31||3.123||3.8%|
|331410||Non-Ferrous (except Al) Smelting & Refining||9||2.956||3.6%|
|212232||Nickel-Copper Ore Mining||5||2.817||3.4%|
|331317||Alum. Rolling, Drawing, Extruding & Alloying||5||2.371||2.9%|
|322121||Paper (except Newsprint) Mills||7||1.944||2.3%|
|321212||Softwood Veneer & Plywood Mills||6||0.458||0.6%|
|337213||Wood Office Furniture Mfg.||1||0.38||0.5%|
|332999||All Other Misc. Fabricated Metal Product Mfg.||2||0.335||0.4%|
|331529||Non-Ferrous Foundries (except Die-Casting)||4||0.309||0.4%|
|336310||Motor Vehicle Gasoline Engine & Parts Mfg.||1||0.306||0.4%|
|336390||Other Motor Vehicle Parts Mfg.||1||0.185||0.2%|
|212233||Copper-Zinc Ore Mining||2||0.14||0.2%|
|321911||Wood Window & Door Mfg.||1||0.128||0.2%|
|321111||Sawmills (except Shingle & Shake Mills)||53||0.107||0.1%|
|212231||Lead-Zinc Ore Mining||1||0.099||0.1%|
|221320||Sewage Treatment Facilities||6||0.086||0.1%|
|331490||Non-Ferrous (except Cu & Al) Secondary Proc.||2||0.04||0.0%|
|113210||Forest Nurseries & Gathering Forest Products||3||0.025||0.0%|
|321211||Hardwood Veneer & Plywood Mills||2||0.022||0.0%|
|325210||Resin & Synthetic Rubber Mfg.||2||0.011||0.0%|
|321999||All Other Misc. Wood Product Mfg.||2||0.01||0.0%|
|114113||Salt Water Fishing||1||0.003||0.0%|
|335990||All Other Electrical Equip. & Component Mfg.||1||0.003||0.0%|
|339990||All Other Miscellaneous Mfg.||2||0.002||0.0%|
|812990||All Other Personal Services||1||0.002||0.0%|
|622111||General (exc. Paediatric) Hospitals||2||0.002||0.0%|
|913150||Municipal Regulatory Services||1||0.001||0.0%|
|325410||Pharmaceutical & Medicine Mfg.||1||0.001||0.0%|
|325999||All Other Misc. Chemical Product Mfg.||1||0.001||0.0%|
|331523||Non-Ferrous Die-Casting Foundries||1||0.001||0.0%|
|321216||Particle Board & Fibreboard Mills||1||0.001||0.0%|
Dioxin & Furan Releases at Canadian Petroleum Refineries Based on the Industry Survey Report
According to the 2004 U.S. Toxics release inventory data, the petroleum refining sector represents 2% of the air releases of Dioxins ands Furans (D-F) in the U.S. even though the release quantity might be small on an individual refinery basis. (See the related background consultation document "Evaluation for the Listing of Petroleum Refining as a Source Activity of Dioxins/Furans Reporting for the National Pollutant Release Inventory" of March 28, 2007 as shown in Part A). Consequently, the NPRI Multi-stakeholder D-F and HCB Sub-group (SG) is interested to know the magnitude of the potential contribution of petroleum refineries to Canadian inventory of D-F emissions, and to determine if it is worth the merit to add the activity of "regeneration of spent reformer catalyst in petroleum refineries" to the list for the D-F reporting to NPRI. In order to facilitate this determination, it was decided that a simple survey would be conducted to gather information regarding the catalytic reformer process operations at Canadian refineries.
Canadian Petroleum Products Institute (CPPI) collected and compiled the data for the SG in the summer of 2007. Each of the refineries within Canada was asked to provide the following information:
- Type of regeneration process in the catalytic reforming process
- Design (maximum) processing capacity of the Reformer
- Treatment (scrubbing) of the vent gas from the regeneration process
- Material used to scrub the vent gas
- Any treatment of the scrubbing and means of disposal of the scrubbing agent
- Any sampling of D-F conducted in any effluent stream within last 10 years, and knowledge of any detection ofD-F, including the detection limit of D-F, from the samplings
Catalytic reformers are known to produce D-F during the regeneration of the catalyst. Emission factors have been published by the EPA and other sources for different types of reformers, but are of quality of low confidence rating. Limited testing at refineries has shown a wide variation in D-F results, even for the same type of reformer. The following methodology was assembled based on searches of emission factors, discussions within the industry, and engineering judgment. The intent of this estimation methodology is to provide a common basis for estimating D-F releases for this survey, as well as to determine in relative sense the significance of releases of D-F from petroleum refineries. If refineries had specific data estimates, then these were used in the release totals.
For catalytic reformers the emission factors in Table 1 were utilized in estimating emissions to air.
|Type of Catalytic reforming regeneration process**||Emission factors of dioxin/furan (total of the congeners) ng/bbl of reformer throughput|
|Continuous Catalytic Reforming Unit||44.4 (1)|
|Cyclic Reforming Unit||6.2 (2)|
|Semi-Regen Reforming Unit||0.014 (1)|
(1) "An Inventory of Sources and Environmental Releases of Dioxin-Like Compounds in the United States for the Years 1987, 1995, and 2000", EPA/600/P-03/002F, November 2006
(2) From within the industry.
** There are three types of reformers defined by the type of catalyst regeneration process:
Continuous - aged catalyst is continuously removed from the reactor and regenerated in an external regenerator and then returned to the system.
Semi-regen - the entire catalytic reformer is taken off-line and regenerated and then returned to service (typically every 6-12 months).
Cyclic - one of the two (or more) reforming reactors is taken off-line for catalyst regeneration and the remaining reactors remain on-line so that reforming operation continues (typical one reactor regenerated every two weeks).
Based on the source information for the emission factors from EPA, it appears that the continuous catalytic unit factors were based on a test of units that vented directly to air and had no treatment of the regeneration vent gas. For unit identified in the survey as continuous reforming and that treated the vent gas (scrubbing), a 70% removal factor was used (engineering judgment). The removed D-F would end up in the facility wastewater treatment system, and based on other studies, would likely adsorb to solids (biological and /or oily sludge) in the system. The semi-regenerative reforming unit's emission factors were based on scrubbing of the vent gas. The cyclic reformer emission factor background data was not available and it was assumed for the purpose of this study that the emission represented a scrubbed stream.
The total D-F emissions were converted to TEQ using a multiplier of 0.1. This was based on conversion of the individual congeners to TEQ and summing to give a factor that ranged in the 7-10% of the total D-F emission factors listed above; and the most conservative approximation of 10% was used.
The results of the survey and calculated emission for the petroleum refining sector are presented in Table 2. The ranking of the petroleum refining among the industry sectors in terms of air emissions based on 2005 NPRI reported D-F emissions are presented in Table 3. A different perspective of ranking by means of 6-digit NAICS code is presented in Table 4, and the petroleum refining would fall into the place after the "copper-zinc ore mining" sector.
The sector's total air emissions of 0.15675 TEQ g would have only represented 0.26% of the total Canadian D-F emission inventory in 2005. The number calculated for the sector is based on maximum design throughput of the reformers and hence represent the maximum emissions. It is likely in the absence of site specific sampling that majority of refineries would utilize the approach taken in the survey, and the numbers would not vary significantly year to year.
|D-F Emissions Estimated based on factors (g/year)||D-F Emissions to Air|
|D-F Emissions to Water/Waste|
|D-F Emissions Estimated based on factors|
|D-F Emissions to Air|
|D-F Emissions to Water/Waste|
|Periodic Cyclic Total||0.14495||0.14495||NA||0.01449||0.01449||NA|
NA Not available
(1) Estimate based on the rated (Design) Reformer capacity and use of EPA factors
* According to CPPI guidance, 70% of D-F are removed when scrubbed in continuous operation
** To convert to 2,3,7,8 TCDD TEQ multiply the total D-F value above by 0.1 per CPPI Guidance document
|Air Emissions only|
|Industry||TEQ grams||# site||% of Total|
|1||Local Gov't services||27.267||29||45.91%|
|2||Other Utility Industry||9.863||29||16.61%|
|3||Primary Steel Industries||8.037||33||13.53%|
|4||Paper & Allied products||4.856||44||8.18%|
|5||Fabricated Metal Products||2.44||4||4.11%|
|6||Non-metallic mineral products||1.976||13||3.33%|
|10||Other manufacturing industries||0.499||3||0.84%|
|11||Furniture & Fixtures||0.423||1||0.71%|
|12||Wholesale - Motor Vehicles, etc.||0.307||1||0.52%|
|13||Other Service Industry||0.188||3||0.32%|
|NAICS Code||NAICS Code Description||# Facility||Air Releases g-ITEQ||Percentage|
|562210||Waste Treatment & Disposal||40||33.247||56.10%|
|221112||Fossil-Fuel Electric Power Generation||31||5.583||9.40%|
|331110||Iron & Steel Mills & Ferro-Alloy Mfg.||12||3.98||6.70%|
|331410||Non-Ferrous (except Al) Smelting & Refining||9||2.564||4.30%|
|322112||Chemical Pulp Mills||27||2.392||4.00%|
|331317||Alum. Rolling, Drawing, Extruding & Alloying||5||2.347||4.00%|
|212232||Nickel-Copper Ore Mining||5||1.291||2.20%|
|322121||Paper (except Newsprint) Mills||9||1.084||1.80%|
|115310||Support Activities for Forestry||7||0.713||1.20%|
|337213||Wood Office Furniture Mfg.||1||0.423||0.70%|
|331529||Non-Ferrous Foundries (except Die-Casting)||4||0.39||0.70%|
|336310||Motor Vehicle Gasoline Engine & Parts Mfg.||1||0.307||0.50%|
|321212||Softwood Veneer & Plywood Mills||6||0.229||0.40%|
|212233||Copper-Zinc Ore Mining||2||0.167||0.30%|
|321911||Wood Window & Door Mfg.||1||0.12||0.20%|
|212231||Lead-Zinc Ore Mining||1||0.099||0.20%|
|221320||Sewage Treatment Facilities||6||0.092||0.20%|
|321111||Sawmills (except Shingle & Shake Mills)||52||0.077||0.10%|
|114113||Salt Water Fishing||1||0.065||0.10%|
|331490||Non-Ferrous (except Cu & Al) Secondary Proc.||2||0.038||0.10%|
|113210||Forest Nurseries & Gathering Forest Products||3||0.022||0.00%|
|321211||Hardwood Veneer & Plywood Mills||2||0.021||0.00%|
|812990||All Other Personal Services||1||0.012||0.00%|
|321999||All Other Misc. Wood Product Mfg.||2||0.01||0.00%|
|325210||Resin & Synthetic Rubber Mfg.||2||0.009||0.00%|
|339990||All Other Miscellaneous Mfg.||2||0.003||0.00%|
|622111||General (exc. Paediatric) Hospitals||1||0.002||0.00%|
|331523||Non-Ferrous Die-Casting Foundries||2||0.002||0.00%|
|325410||Pharmaceutical & Medicine Mfg.||1||0.001||0.00%|
|322111||Mechanical Pulp Mills||2||0.001||0.00%|
|321216||Particle Board & Fibreboard Mills||1||0.001||0.00%|
The D-F and HCB Sub-group, after reviewing the initial results, had a number of follow-up questions posed to CPPI. These questions and the responses are included below.
Responses to question shown in blue different font
1. Do the numbers shown in column 1 and column 2 of Table 2 represent the sum of each congener calculated for the total D-F values? Or are the numbers simply the sum of total D-F values as available on a category basis?
- The numbers represent total of the D-F congeners. In the source documentation for the emission factors (EPA) each congener has an emission factor and these are summed to provide an emission factor for total D-F.
2. Each congener within the D-F category would have a different TEQ value. How valid it is to solely multiply the total D-F value with the 0.1 factor for the total TEQ? How does the CPPI guidance determine this factor? Is it possible that the CPPI guidance be made available as a reference document outside the CPPI community?
- The individualD-F congener emission factors when multiplied by their respective TEQ factor and then summed represent 7-10% of the total emission factor as calculated in "1" above. Taking 10% represents a conservative basis and was simplified way of obtaining the total TEQ number.
- The document is at this time not available for external use; largely it was developed to meet a need at the time based on best information available. The document did not receive reviews or approval for it to become an official CPPI document. The draft CPPI document was used as starting point for determining the best factors to use.
3. How does the CPPI guidance determine the scrubber efficiency of 70% for the D-F removal?
- It is engineering approximation.
4. If 70% of D-F is removed when scrubbed, then should it be that the D-F releases to water/waste greater than the air emissions? Table 2 indicates a different scenario about the ratio of D-F air releases to water/waste releases. Please verify the numbers.
- Not all the refineries have scrubbers. In the current survey only 7 were identified as having scrubbers. The numbers in the table are correct as calculated.
5. What are the EPA emission factors that CPPI has adopted for the calculations? EC has previously referenced the relevant D-F emission factor information in the draft consultation document on "D-F Reporting for Petroleum Refineries" (2007-03-28 version). This information can be found in Table 5-9 of Page 5-29 of the US EPA published document, which can be downloaded at the following website address:
Is this the same information source that CPPI has identified for the D-F calculations of "air emissions"? If it is, how has CPPI reconciled the use of the emission factors information for the periodic cyclic process of the reforming operations which is not listed in the document?
- TheEPAfactors for Continuous and Semi-regen are as per theEPAsource document indicated.
- Periodic Cyclic process factors based on a company developed emission factor.
6. The EPA emission factors are meant for "air emissions", and therefore would not be applicable for the calculations of "produced D-F emissions" (see column 1 of Table 2). How does CPPI determine their estimation in the first place?
- For the continuous regen operation there was no scrubbing operation in the facilities tested therefore we applied the 70% removal to those units identified as having a scrubbing operation.
- For semi-regen it is correct that the units had water scrubbing operation thus 70% removal should not be applied. This will add 0.00003g TEQ D-F to the total of 0.154TEQ D-F a change of 0.0.02%.
- For the Cyclic regen process we do not have the source data information as to the basis and if test were with units with or without scrubbers. If we take conservative approach and assume they had scrubbers then 0.0024 TEQD-F will be added to total air emission of 0.154 gTEQ D-F and an increase of 1.5%.
7. Is it a fact that no surveyed facility has used the measurement technique to generate the D-F data? It was mentioned at the SG meeting that one facility did not use the EPA emission factors because of concerns on their reliability. What did the facility do instead to provide the requested data?
- The facility in question had its own data and used that to estimate their emissions.
Draft Date: 2007-09-14
Recommendations of D-F and HCB Sub-group on the Listing of Catalytic Reforming Operations at Petroleum Refineries as a Source Activity of Dioxins and Furans Reporting to the NPRI
After assessing the available information materials relevant to the issue, the dioxins and furans (D-F) and HCB Sub-group (SG) has come to the conclusion that a consensus can not be reached for the recommendations on whether to add the petroleum refining sector to the listing of source activities for the D-F reporting to NPRI.
Some SG members have recommended that the source activity not be added for the reporting due to the following reasons:
- The sector based on survey results is not significant (would rank 14th for 2005). Emissions would have to increase by over 300% for sector to be in top 10
- Results provided in survey represent a picture of the contribution that is not likely to change year over year (survey results were based on maximum throughputs and only methodology available for estimating)
- On individual basis no refinery would rank higher than 100th and thus there would be no drivers to improve
- Sampling of D-Fis costly, and based on the relative contribution of the sector there exists no drivers to undertake it
However, the ENGO stakeholders do not agree with the above position. They recommend that the petroleum refining sector that has the catalytic reformers be added to the NPRI list of facilities that must report D-F. Their reasons for this are as follows:
- The total of D-Freported to NPRI for 2006 was 52.865 g. These numbers, though small, are significant because of the nature of the human and wildlife health effects from D-F. Therefore, it is important to report all releases of D-F, even if they are very small, provided they are above the LOQ. Therefore to apply a factor such as not being in the top 10 of releases, as proposed by the industry members of this sub-group, is completely irrelevant when discussing this set of substances
- The Thompson study referred to on page 2 of this report gives air emissions concentrations in nanograms, dramatically above theLOQ of 32 pg/TEQ given for reporting of D-F. Therefore, it is feasible to accurately report air releases
- As shown by the petroleum industry report, emissions factors for calculating emissions are available
- The number of facilities involved is not large and is easily identified so it will be easy for facilities to determine whether they should be reporting. As a result it will not be a burden on facilities that would not have to report
The SG feels that the analysis of survey results is the best available information with which to make the decision. The SG also feels that any further work would not be productive, and consequently is referring the issue back to the WG.
Draft Date: 2007-09-19
1US EPA. An Inventory of Sources and Environmental Releases of Dioxin-Like Compounds in the United States for the Years 1987, 1995, and 2000 (EPA/600/P-03/002f, Final Report, November 2006). U.S. Environmental Protection Agency, Washington, DC, EPA/600/P-03/002F.
Available online at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=159286
- Date Modified: