




DATE:
May 10, 2012
FROM:
Lisa Grogan-McCulloch, Greenhouse Gas Reporting Branch
SUBJECT:
Potential Future Subpart L Options
      
I. 	Introduction 
Following publication of proposed confidentiality determinations for the Fluorinated Gas Production source category (subpart L), EPA received public comment that the release of process-specific emissions of individual fluorinated greenhouse gases (GHGs) from some fluorinated gas production and transformation processes may reveal trade secrets.  In response to these comments, EPA is undertaking an in-depth analysis on whether the subpart L reporting requirements could be changed to mitigate these concerns without compromising the adequacy of the data set available to the Agency to inform development of future climate policy.  As part of this effort, we are presenting in this Memorandum potential options for changes to the subpart L reporting requirements should EPA determine that a change is appropriate.

There are several possible options for reporting process level emissions from fluorinated gas production.  In this memo, we have attempted to outline these options in order to be as comprehensive as possible.  EPA is currently in the process of evaluating whether these options would generate an adequate data set upon which to inform future policy.  Should EPA determine that a permanent change to the reporting requirements is appropriate, any comments received on this memorandum will facilitate the narrowing of options and development of a future proposal.  
II. 	Overview of Current Subpart L Reporting Requirements

Under the current subpart L, fluorinated gas producers are required to report GHG emissions by process and by chemical.  Specifically, facilities are currently required to report the following: 
   * The total mass of each fluorinated GHG emitted from 
         o each fluorinated gas production process (98.126(a)(2)(i)) , 
         o each fluorinated gas transformation process that is not part of a fluorinated gas production process (98.126(a)(2)(ii)), 
         o each fluorinated gas destruction process that is not part of a fluorinated gas production or transformation process (98.126(a)(2)(iii)), and 
         o venting of residual fluorinated GHGs (heels) from containers returned from the field (98.126(a)(2)(iv)).  
   * Where producers use the emission factor or emission calculation factor methods under section 98.123(c), they are also required to report chemical-specific emissions for each process vent and for equipment leaks, which reflects the fact that emissions are calculated separately for vents and leaks (98.126(c) and (d)).  
   * Where producers use the mass-balance method under 98.123(b), they are also required to report emissions by the role the chemical plays in the production process (i.e., as reactant, product, or by-product), again reflecting the method by which emissions are calculated (98.126(b)(3)  -  (5)).  
   * In addition, fluorinated gas producers are required to report, by process, the chemical identities of the contents of the streams (including process, emissions, and destroyed streams) analyzed under the initial scoping speciation at 98.124(a) as well as the location and function of those streams (98.126(a)(3) and (4)).  
   * The first reporting under subpart L is to occur in 2012, for reporting year 2011 (RY2011).

III. Options for Reporting Emissions from Fluorinated Gas Production and Transformation Processes 
EPA is considering several options to address the concerns raised by the commenters on the proposed confidentiality determinations for subpart L.  Section A below describes an approach under which facilities would be required to meet certain criteria as a condition for reporting their fluorinated GHG emissions in a less detailed manner than currently required by the rule.  Section B describes a range of options for reporting at various levels of detail.  The options described in Section B could be applied together with the approach described in Section A or separately.

        A. Option of establishing criteria to determine the level of reporting 

As discussed above, given the comments on the confidentiality rulemaking (77 FR 1434), EPA is considering whether EPA needs the current level of data reported from all facilities within subpart L, or whether EPA would have an adequate data set with which to inform future policy by collecting this level of information from most facilities and less detailed information from facilities with specialized circumstances.  To this end, EPA is considering an approach under which facilities that meet certain criteria could report their fluorinated GHG emissions in a less detailed manner according to one of the options outlined in Section B below. The criteria could include considerations such as whether a facility has a unique process or product, whether the facility's process information is already in the public domain, and/or the quantity of the product produced.  Under this option, all facilities that do not meet the criteria for reporting in a less detailed manner would be required to report using the current subpart L reporting requirements.  

        B. Options for Reporting at Alternative Levels of Aggregation
For fluorinated gas production and transformation processes, EPA is considering a range of reporting options that would provide varying levels of detail regarding the source of the emissions (ranging from reporting by process and by emission type to reporting at the facility level) and the chemicals emitted (ranging from reporting by speciated fluorinated GHG to reporting in CO2e).  Table 1 below illustrates where the possible alternative reporting options (as well as the current reporting approach in the rule) fall along these two axes.  To address the possibility that reporting chemical-specific emissions at the process level could disclose business-sensitive information, many of the options (represented as cells in Table 1) involve reporting at different levels of aggregation for the source and the chemicals.  For example, under one option, the source would be reported at the highest level of aggregation (by facility) while the chemicals would be reported at the lowest level (by fluorinated GHG species).  Under other options, the source would be reported at the lowest level of aggregation (by process and emission type) while the chemicals would be reported at a higher level of aggregation (e.g., by gas group or in terms of CO2e).  In addition, different options could be combined.   
Based on Table 1, there are twelve or more individual options that could be based on different combinations of these levels.  Rather than reviewing these individually, Section 1 focuses on the varying levels of aggregation for the source, while Section 2 focuses on the varying levels of aggregation for the chemicals.  Section 3 then discusses four individual options in more detail.  

      1. Levels of Aggregation for the Emissions Source
      a.  Reporting by process and emissions type (current requirements in subpart L)
Currently, subpart L requires facilities to report their emissions by process.  In addition, they are required to report the identity and mass of the product produced by each production or transformation process. Where the emission estimation method permits, facilities are also required to report their emissions by emission type.
      
The definition of "process" in subpart L reads in part, "Process means all equipment that collectively functions to produce a fluorinated gas product, including an isolated intermediate (which is also a fluorinated gas product), or to transform a fluorinated gas product. A process 


Table 1.  Overview of GHG Reporting Alternative Options for Subpart L

                                       
                            Increasing Aggregation
Emission Source                                       
                                       
                                      Gas
                             Increasing Aggregation
                                       
By process, by emission type
   *  Vents
   *  Leaks
Product of process reported
By process, by emission type
Product of process not reported; instead process identified generically,  e.g.,
   *  Production reaction A
   *  Production reaction B
   *  Transformation distillation process C
By Process only
At Facility-Level
   *   All production, transformation processes combined
     
   
 
By chemical
                                 Current rule
                                       
                                       
                                       

Gas groups (based on chemical structure OR atmospheric lifetime)
                                       
                                       
                                       
                                       

CO2e
                                       
                                       
                                       
                                       



may consist of one or more unit operations. For the purposes of this subpart, process includes any, all, or a combination of reaction, recovery, separation, purification, or other activity, operation, manufacture, or treatment which are used to produce a fluorinated gas product."
Emission type refers to emissions from process vents and equipment leaks.  The emission factor and emission calculation factor methods provide separate estimates for these emissions types while the mass-balance method does not.  
Reporting the identity and mass of the product produced by the process provides a natural means of identifying the process and allows EPA to compare processes that make similar products at different facilities.  Reporting by emission type allows EPA to identify and evaluate potential reduction opportunities.

      b. Reporting by generically identified process 
Given that some commenters have asserted that the identity of the product of the process could disclose sensitive business information (if the product is an intermediate), EPA is considering an option where facilities would identify processes generically.  The process would first be identified as a production process, a transformation process that transformed fluorinated GHGs produced on site, or a transformation process that transformed fluorinated GHGs produced at other facilities.  Within these categories, the process would be further identified as a reaction, separation/recovery, or packaging process, or as a combination of these.  Finally, the process would be tagged with an identifier chosen by the facility (e.g., a letter or number) that would remain constant from year to year to allow year-to-year comparisons.  One area where reporting of the identity of the product of the process may not involve reporting of sensitive business information is where that product is a final product, whose identity would already be known to competitors.   Compared to Option a, this option would not allow EPA to identify and compare processes making the same product across different facilities  
      c. Reporting by  process, with emission types reported in aggregate (across processes) 
Under this option, the emissions reported for each process would not be broken down by emission type.  Emission type could be reported in aggregate across all processes.  Under this approach, potential reduction opportunities for each emission type within each process would be more difficult to evaluate than under Options a and b.  Because the relative magnitudes of emissions from vents and leaks are not likely to be business-sensitive, and similar sets of chemicals are expected to be emitted from vents and leaks for any given process, distinguishing between emissions from vents and leaks would not add to the risk of revealing business-sensitive information (compared to reporting by process).

       d.  Reporting at the facility level
The least detailed option under this section would be to require reporting emissions at the facility level for all production and transformation processes combined.  EPA would anticipate using this requirement to supplement a requirement to report emissions at the process level.  For example, a requirement to report emissions by chemical group at the process level could be supplemented by a requirement to report emissions by chemical at the facility level.  Such an approach would retain some of the benefits of chemical-specific reporting (e.g., comparison of emissions estimates through atmospheric measurements) without linking particular chemicals to particular processes. 
  
      1. Levels of Aggregation for Chemicals
      a.  Reporting by individual (speciated) fluorinated GHG (current subpart L requirement)
Subpart L currently requires fluorinated gas producers to report their emissions by chemical.  

For subpart L, information on the identities and characteristics of GHGs is particularly important.  First, the range of GWPs and atmospheric lifetimes spanned by the fluorinated GHGs is quite large.  Lifetimes range from a few days (e.g., for several unsaturated fluorocarbons) to thousands of years (e.g., for saturated perfluorocarbons), while GWPs range from less than one (e.g., for several unsaturated fluorocarbons) to above 20,000 (for SF6).  Often, the same fluorinated gas production facility may emit fluorinated GHGs at both ends of the GWP and lifetime ranges.  Knowledge of the GWPs of the chemicals emitted is necessary to distinguish between processes emitting many tons of a low-GWP chemical and processes emitting a few tons (or kilograms) of a high-GWP chemical.  While the CO2-equivalent emissions of both processes may be the same, appropriate emission reduction strategies may differ.  Knowledge of the lifetimes of the chemicals is important to understanding how emissions from different processes would fit into policies that focus particularly on short-lived or long-lived GHGs.

Second, fluorinated gas producers are a significant emission source for many fluorinated GHGs, and for some fluorinated GHGs, they are the only source.  This makes them especially important in efforts to reconcile atmospheric measurements of fluorinated GHGs with estimates of emissions of fluorinated GHGs from all emitting industries.  (Most fluorinated GHGs lack significant natural sources.)    

Chemical-specific reporting is also consistent with GHG Inventory reporting under the United Nations Convention on Climate Change (UNFCCC), which encourages chemical-specific reporting.   Under the UNFCCC, other countries report chemical-specific emissions from comparable fluorinated gas production facilities. For example, in 2011 and previous years, Belgium's GHG inventory reported emissions from "an electrochemical synthesis (electro-fluorination) plant, which emits, or has emitted SF6, CF4, C2F6, C3F8, C4F10, C5F12 and C6F14 as well as fluorinated greenhouse gases not covered by the Kyoto Protocol (among which CF3SF5, C7F16, C8F18 and C8F16O)."   From this plant, Belgium reported 2009 emissions of CF4, C4F8, C5F12, and C6F14 in tons of each gas.   France and Italy have also reported chemical-specific emissions from their fluorinated gas production facilities.

      b.  Reporting by fluorinated GHG group based on chemical structure 
One set of options that would retain information on chemical characteristics but not on precise chemical identities would be to require reporting of emissions by fluorinated GHG group.  Such fluorinated GHG groups could be based on chemical structure, GWP, atmospheric lifetime, or some combination of these.   Whatever the organizing principle, to be useful for policy analysis, the groups should meet the following criteria:
      1. They should be distinct (i.e., not overlapping);
      2. Considered together, they should be comprehensive (i.e., cover all fluorinated GHGs);
      3. They should track atmospheric lifetime and GWP.
One set of groups that EPA is considering is based on the structure of the chemicals.  The groups would include:
   * Fully fluorinated GHGs.  These would include fluorinated GHGs that contain only single bonds and in which all available valence locations are filled by fluorine atoms.  This group would include but is not limited to saturated perfluorocarbons, SF6, NF3, SF5CF3, fully fluorinated linear, branched and cyclic alkanes, fully fluorinated ethers, fully fluorinated tertiary amines, fully fluorinated aminoethers, and perfluoropolyethers. Fully fluorinated GHGs generally have lifetimes of over 500 to several thousand years and GWPs of 6,500 to 23,900.
      
   * Saturated hydrofluorocarbons.  This group would include hydrofluorocarbons that contain only single bonds (i.e., hydrofluoroalkanes such as HFC-134a and HFC-23).  Saturated HFCs generally have lifetimes from 5 to 55 years and GWPs from 650 to 4,000, but there are exceptions at both extremes (see discussion below).

   * Other fluorinated GHGs.  This group would include fluorinated GHGs that are neither fully fluorinated GHGs nor saturated hydrofluorocarbons.  This includes but is not limited to hydrofluoro-olefins, fluoroketones, hydrofluoroethers, and hydrofluoropolyethers .  These GHGs generally have lifetimes of less than 10 years and GWPs of less than 1000, but there are some exceptions with higher GWPs and lifetimes.
      
To ensure that the gas groups are both distinct and comprehensive, the last gas group is designed to capture everything not in either of the first two groups.  With the exception of HFE-125 and HFE-134, which have lifetimes of 136 and 26 years respectively, EPA is not aware of any medium- or long-lived fluorinated GHGs that fall into this group.  
This approach would avoid linking particular chemicals to particular processes, which might reveal business-sensitive information.  At the same time, reporting by chemical group, defined by chemical structure, would provide general information on the chemical and atmospheric properties of the emissions because fluorinated GHGs with fairly similar atmospheric lifetimes and GWPs would be grouped together.  For example, most saturated HFCs whose lifetimes and GWPs have been evaluated by the IPCC have lifetimes ranging between 5 and 55 years and GWPs ranging between 650 and 4,000 (based on the Second Assessment Report).  It would therefore provide useful information for assessing potential emission reduction approaches (e.g., if reported at the process level) and for informing policies that distinguish among chemicals based on their atmospheric lifetimes and GWPs.  It is also similar to the approach of current international conventions that address chemicals with impacts on the global atmosphere (e.g., UNFCCC, Montreal Protocol).

However, the range of lifetimes and GWPs spanned by each chemical group is still relatively large, and it becomes larger when outliers are factored in.  For example, HFCs include HFC-152a (with a lifetime of 1.4 years and a GWP of 140) and HFC-23 (with a lifetime of 270 years and a GWP of 11,700).  This reduces the usefulness of the data compared to chemical-specific reporting or to reporting by other possible chemical groups.  
       
      a.  Reporting by fluorinated GHG group based on atmospheric lifetime or GWP
Another option that EPA is considering is to group gases explicitly by atmospheric lifetime or GWP.    This grouping approach, like that described in Option b, would meet the three criteria for gas groupings described above.  Table 2 below lists one possible set of groups based on atmospheric lifetime.  A similar example could be drawn up based on GWP.  For the fluorinated GHGs whose GWPs are included in Table A-1 of  40 CFR 98, subpart A, EPA could propose to  add lifetimes to Table A-1 based on the most recent IPCC update of GWPs and lifetimes (e.g., the Table 2.14 errata).  For fluorinated GHGs whose GWPs and lifetimes were not on Table A-1, EPA could propose that reporters would either independently assess the lifetime or GWP of the gas or report the gas as unclassified.  

             Table 2.  Fluorinated GHG Atmospheric Lifetime Groups
                   Groupings by atmospheric lifetime, years
                                     <1
                                      1-5
                                     5-20
                                     20-60
                                    60-300
                                    >300
                                       

Under this approach, the ranges of atmospheric lifetimes and GWPs would be considerably lower than under Option b, increasing the usefulness of the data compared to Option b.  However, this option would require facilities to know the lifetime of each chemical to report it.  Although lifetimes have been estimated for the fluorinated GHGs listed on Table A-1 (and could therefore be easily added to this table by EPA), they have not necessarily been estimated for fluorinated GHGs that are not on Table A-1.  (Atmospheric lifetimes are easier to assess than GWPs, however.)
Hybrids of the two grouping options (chemical structure and atmospheric lifetime or GWP) could partially address the drawbacks of each (e.g., by breaking saturated HFCs into two or more groups based on lifetime), but this would increase the complexity of the rule.  
                                        
         a. Reporting emissions in terms of CO2e across all fluorinated GHGs
EPA is also considering requiring reporting of fluorinated GHG emissions in terms of CO2e across all fluorinated GHGs.  EPA would anticipate combining this option with another option (e.g., process emissions could be reported in CO2e while facility-level emissions could be reported by chemical).  If CO2e emissions were reported at the process level, this approach would allow ranking of processes in terms of their GWP-weighted emissions.  However, unlike reporting by chemical group, it would not allow EPA to distinguish between processes that emit large quantities of low-GWP chemicals and small quantities of high-GWP chemicals.  Similarly, it would not inform policies that distinguish among chemicals based on their atmospheric lifetimes and GWPs.  Finally, for fluorinated GHGs that are not on Table A-1, this approach would require facilities to know or estimate GWPs or to use default GWPs (see the discussion below) .

      1. Converting fluorinated GHG emissions into CO2e
To convert non-CO2 emissions into CO2e, the EPA generally requires facilities to use Equation A-1 of subpart A, which multiplies the mass of each chemical by the GWP for that chemical in Table A-1.  For fluorinated GHGs that do not have a GWP listed in Table A-1, EPA is considering an option under which facilities would either use a default GWP or use their best estimate of the GWP, based on the information described in 40 CFR 98.123(c)(1)(vi)(A)(3) . The default GWP used would depend on the type of fluorinated GHG. For fully fluorinated GHGs, the default GWP would be 10,000, which is based on the average GWP of the fully fluorinated GHGs on Table A-1. For other fluorinated GHGs, the default GWP would be 2,000, which is based on the average GWP of the other fluorinated GHGs on Table A-1. EPA would propose to distinguish between fully fluorinated GHGs and other fluorinated GHGs because the former have significantly longer lifetimes and higher GWPs than the latter.
      
Facilities using best estimate GWPs would also be required to report those GWPs to EPA along with the data and analysis that were used to develop the GWPs.  This would enable EPA to verify the GWPs and to consider them for inclusion on Table A-1. 

The default GWPs would be an option for facilities that either do not know the GWPs of the fluorinated GHGs that they emit or do not wish to report best-estimate GWPs for one or more chemicals because this would involve disclosure of the identities of the chemicals.  The default of 2000 (for fluorinated GHGs other than fully fluorinated GHGs) would be the same as that used by facilities to calculate whether they must use stack testing to establish an emission factor for a vent that emits one or more fluorinated GHGs that do not have GWPs on Table A-1.  Both the 2000 and 10,000 default GWPs are intended to represent averages across their respective gas groups; they are not conservative.  EPA expects that this would generally result in unbiased estimates of facility-wide CO2e emissions that include fluorinated GHGs without GWPs on Table A-1; however, some facilities might choose to use the default rather than their own best-estimate if that estimate was higher than the default.  In this case, facility-wide CO2e emissions would be underestimated, and EPA might never receive the information necessary to update Table A-1 to correct the underestimate.  Other options include making the defaults more conservative (i.e., higher numerically) to encourage facilities to estimate and report their own GWPs, and requiring facilities to report speciated emissions of fluorinated GHGs (and estimated GWPs, if available) that are not on Table A-1 and are emitted in significant quantities (e.g., 1 ton or more for each fully fluorinated GHG and 5 tons or more for each other fluorinated GHG, which would be expected to equate to approximately 10,000 mtCO2e in either case).

IV.	References
77 FR 1434.  January 10, 2012.  Proposed Confidentiality Determinations for Data Elements under the Mandatory Reporting of Greenhouse Gases Rule.
Muffat, Jeffry C., to EPA Docket No. EPA-HQOAR-2011-028. Public comment letter to EPA's Proposed Confidentiality Determinations for Data Elements under the Mandatory Reporting of Greenhouse Gases Rule.  EPA-HQ-OAR-2011-0028-0018. Dated March 12, 2012.
Mascarenhas, Brendan, to EPA Docket No. EPA-HQOAR-2011-028. Public comment letter to EPA's Proposed Confidentiality Determinations for Data Elements under the Mandatory Reporting of Greenhouse Gases Rule.  EPA-HQ-OAR-2011-0028-0017. Dated March 12, 2012.
