
                                       
REVISED TECHNICAL SUPPORT DOCUMENT FOR CAPROLACTAM, GLYOXAL, AND GLYOXYLIC ACID PRODUCTION: SUPPLEMENTAL PROPOSED RULE FOR THE GREENHOUSE GAS REPORTING PROGRAM
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                       
                                                    Office of Air and Radiation
                                           U.S. Environmental Protection Agency
                                                                               
                                                                               
                                                                  April 1, 2023


Contents
1.	Industry Description	1
1.1	Caprolactam	1
1.2	Glyoxal	2
1.3	Glyoxylic Acid	3
2.	Total Emissions	3
2.1	Process Emissions	4
2.1.1	Estimated N2O Emissions	4
2.1.2	Estimated CO2 Emissions	5
2.2	Stationary Combustion	6
3. 	Review of Existing Programs and Methodologies	7
3.1	Review of Existing Programs	7
3.2	Calculation Methodologies for Caprolactam, Glyoxal, and Glyoxylic Acid Production Processes	8
4.	Options for Reporting Threshold	10
5.	Options for Monitoring Methods	11
5.1	Option 1: Emission Calculation / Mass Balance	11
5.2	Option 2: Direct Measurement	12
6.	QA/QC Requirements	13
7.	Options for Estimating Missing Data	13
7.1	Procedures for Option 1: Emission Calculation / Mass Balance	13
7.2	Procedures for Option 2: Direct Measurement	13
8.	References	14




1.	Industry Description
This prospective source category consists of facilities that produce caprolactam, glyoxal, and/or glyoxylic acid. Most commercial processes for producing caprolactam, glyoxal, and/or glyoxylic acid result in nitrous oxide (N2O) emissions, so the inclusion of this source category would improve accounting of U.S. N2O emissions.  
1.1	Caprolactam
Caprolactam is a crystalline solid organic compound with the chemical formula C6H11NO. Caprolactam has a wide variety of uses, including brush bristles, textile stiffeners, film coatings, synthetic leather, plastics, plasticizers, paint vehicles, cross-linking for polyurethanes, and in the synthesis of lysine. Caprolactam is primarily used in the manufacture of synthetic fibers, especially Nylon 6 (NIH, n.d.-a).
Most commercial processes used for the manufacture of caprolactam use the following sequence of reactions.
 Cyclohexanone is mostly formed from benzene, but toluene can also be used:
 Benzene is hydrogenated to cyclohexane; and
 Cyclohexane is oxidized to produce cyclohexanone, which is used in the last step below to form caprolactam.
 Ammonia is oxidized to nitric oxide (NO) and nitrogen dioxide (NO2);
 Ammonia is reacted with carbon dioxide (CO2) and water (H2O) to form ammonium carbonate;
 Ammonium carbonate is reacted with NO/NO2 to form ammonium nitrite;
 Ammonia is reacted with sulfur dioxide (SO2) and H2O to form ammonium bisulfite;
 Ammonium nitrite and ammonium bisulfite react to form hydroxylamine disulfonate;
 Hydroxylamine disulfonate is hydrolyzed to form hydroxylamine sulfate and ammonium sulfate;
 Hydroxylamine sulfate is reacted with cyclohexanone to form caprolactam through the Beckmann rearrangement (IPCC, 2006).
As of 2019, the United States had two companies that produce caprolactam with a total of two caprolactam production facilities: AdvanSix in Virginia (AdvanSix, 2020) and BASF in Texas (BASF, 2020). Caprolactam production in 2019 was estimated at 515,000 metric tons (ACC 2020).
1.2	Glyoxal
Glyoxal is a solid organic compound with the chemical formula C2H2O2. The preferred name assigned by the International Union of Pure and Applied Chemistry (IUPAC) is ethanedial. Since pure glyoxal is not stable in the atmosphere, it is typically supplied in a 40 percent by weight solution in water. Glyoxal has a wide variety of uses, including as a crosslinking agent in various polymers for paper coatings, textile finishes, adhesives, leather tanning, cosmetics, and oil-drilling fluids; as a sulfur scavenger in natural gas sweetening processes; as a biocide in water treatment; to improve moisture resistance in wood treatment; and as a chemical intermediate in the production of pharmaceuticals, dyestuffs, glyoxylic acid, and other chemicals. It is also used as a less toxic substitute for formaldehyde in some applications (e.g., in wood adhesives and embalming fluids) (Fact.MR, 2019; Ecofys et al., 2009; NIH, n.d.-b).
Commercial glyoxal may be prepared in one of two ways. The more common process is the LaPorte process, the gas-phase catalytic oxidation of ethylene glycol with air in the presence of a silver or copper catalyst (Mattioda and Blanc, 2012). This process produces CO2 as a byproduct but not N2O, due to the feedstock not containing nitrogen. The single largest production site, operated by BASF in Germany, uses the LaPorte process and has a capacity of about 60,000 mt/yr (Ecofys et al., 2009).
The less common production process is the liquid-phase oxidation of acetaldehyde with nitric acid; this process emits both N2O and CO2 and is estimated to account for less than 20 percent of total global production (Teles et al., 2015). A WeylChem Lamotte (previously Clariant) facility in France uses the acetaldehyde and nitric acid production route (Ecofys et al., 2009; Teles et al., 2015); it is not clear if any other facilities in the world use this process. The capacity of the WeylChem Lamotte facility is currently unavailable. Much of the worldwide production currently occurs in China (Fact.MR, 2019).
Total annual worldwide production estimates vary; one source estimated at between 110,000 metric tons (mt) and 170,000 mt (Ecofys et al., 2009), another estimated production at about 250,000 mt (Teles et al., 2015), and one study projected total 2019 consumption at 360,000 mt (Fact.MR, 2019).
It is unclear how much glyoxal currently is produced in the United States. A BASF unit in Geismar, Louisiana, had a capacity of about 20,000 mt (Ecofys et al., 2009), but it was shut down in 2014. Of all the facilities discussed in this section, this BASF facility is the only facility that has ever reported for any subpart in any year under the Greenhouse Gas Reporting Program (GHGRP). Emerald Carolina Chemical LLC (now Dystar Carolina Chemical Corp.) in Charlotte, North Carolina, also produced glyoxal, but the glyoxal process unit (capacity unknown) was decommissioned prior to October 2012 and at the time, the company did not intend to recommission the glyoxal process unit in the future (Cogburn, 2012). 
The most recently available data reported under the Toxic Substances Control Act (TSCA) indicate that several facilities imported glyoxal in 2012 through 2016, but no facility except the BASF facility in Geismar identified themselves as a domestic manufacturer. In 2015, however, four facilities claimed that their production status (i.e., as a domestic manufacturer or as an importer) and their quantities domestically manufactured and/or imported, were confidential business information (CBI) (ChemView, 2021). Thus, it is possible that one or more of these four facilities could be a domestic manufacturer. According to the TSCA data, total nationwide production volume (i.e., the sum of the amount domestically manufactured plus the amount imported) from reporting facilities was 18.3 million pounds (lbs) (8,300 mt) in 2011, and the total production volume in each of the years between 2012 and 2015 was between 10 million lbs and 50 million lbs (about 4,500 mt to 23,000 mt). It is also possible that there are other facilities in the U.S. that do not have to report under TSCA because their total production volume is less than 25,000 lbs/yr or they are exempt from reporting because they are a small manufacturer based on their total company sales revenue. 
1.3	Glyoxylic Acid
Glyoxylic acid is a solid organic compound with the chemical formula C2H2O3, and the preferred name assigned by IUPAC is Acetic Acid, 2-oxo. Glyoxylic acid is exclusively produced by the oxidation of glyoxal with nitric acid. It is used mainly in the synthesis of vanillin, allantoin, and several antibiotics like amoxicillin, ampicillin, and the fungicide azoxystrobin (Teles et al., 2015). The production process is expected to emit both N2O and CO2  (see Section 3.2 for estimation methodologies). Production capacity of glyoxylic acid worldwide is estimated to be around 80,000 mt/yr (Teles et al., 2015) or 122,000 mt/yr (for 2015) (Fior Markets, 2018). According to one market researcher, China accounted for more than 88 percent of worldwide production in 2015 (Fior Markets, 2018). In 2015, four facilities reported glyoxylic acid data under TSCA. Each of these facilities reported no domestically manufactured glyoxylic acid. Total glyoxylic acid production volume (presumably all imported) was in the range of 1 million lbs to 20 million lbs (450 mt to 9,000 mt) (ChemView, 2021). As was noted above for glyoxal, there also may be some small glyoxylic acid manufacturers that were not required to report under TSCA. 
2.	Total Emissions
The ammonia oxidation step of caprolactam production results in emissions of N2O, and the ammonium carbonate step results in emissions of CO2 (IPCC, 2006); however, CO2 emissions "... from the conventional process are unlikely to be significant in well-managed plants" (IPCC, 2006). Therefore, only N2O emissions are estimated in this document. Note that abatement of N2O emissions may occur at caprolactam facilities, but any potential abatement has not been considered in emissions estimates. 
The liquid-phase oxidation of acetaldehyde with nitric acid to produce glyoxal emits both N2O and CO2, but as discussed in Section 4 of this document, available methods for estimating emissions address only the N2O (Ecofys et al., 2009; IPCC, 2006). The LaPorte process for producing glyoxal generates CO2 emissions (Ecofys et al., 2009), but the source of such emissions is not described and no methods for estimating such emissions are presented in the available literature. The available methods for estimating N2O emissions that are discussed in Section 3 of this document appear to focus on process vent emissions; it appears that any emissions from equipment leaks are assumed to be small and are not calculated.
The basic reaction equation for the production of glyoxal from acetaldehyde and nitric acid is:
                    2C2H4O + 2HNO3 à 2C2H2O2 + N2O + 3H2O
The stoichiometric relationship indicates that the ratio of N2O produced to glyoxal produced will be 0.38 mt N2O/mt glyoxal; however, under commercial conditions the ratio has been determined to be approximately 0.52 mt N2O/mt glyoxal (IPCC, 2006).
Production of glyoxylic acid also produces N2O emissions (Ecofys et al., 2009; IPCC, 2006); available literature makes no mention of CO2 emissions, but it appears there might be small amounts from control of process vent streams. Glyoxylic acid is produced by the oxidation of glyoxal with nitric acid. A considerable amount of the glyoxal is overoxidized to oxalic acid, and N2O is created through this secondary reaction as follows (Teles et al., 2015; Ecofys et al., 2009):
                    2C2H2O2 + 2HNO3 à 2(COOH)2 + N2O + H2O
Available literature suggests N2O is also produced through the reduction of nitric acid in the primary reaction to produce glyoxylic acid, but this is not stated explicitly, and no chemical equation is presented.
The IPCC indicates the default uncontrolled N2O emission factor for glyoxylic acid production is 0.10 mt N2O/mt glyoxylic acid (IPCC, 2006). Note that another reference estimates the emissions are 0.2 mt N2O/mt glyoxylic acid (Teles et al., 2015).
Section 2.1 presents nationwide estimates of process-related/byproduct N2O and CO2 emissions from caprolactam, glyoxal, and glyoxylic acid production and Section 2.2 discusses estimates of carbon dioxide equivalent (CO2e) emissions from stationary combustion units at caprolactam, glyoxal, and glyoxylic acid production facilities. A summary of annual process and stationary combustion emissions for each chemical is presented in Table 1.
Table 1. Total GHG Emissions by Compound[a]
Compound
Number of Facilities
Annual Process Emissions
(million mt CO2e/yr)
Annual Stationary Combustion Emissions
(million mt CO2e/yr)
Total Emissions
(million mt CO2e/yr)
Caprolactam
2[b]
1.2[c]
1.2[d]
2.5
Glyoxal
Unknown
0.06 (estimated)
0.06 (estimated)
0.13
Glyoxylic Acid
Unknown
0 (estimated)
0 (estimated)
0 (estimated)
[a] Emissions calculated using AR5 global warming potential (GWP) values of 28 for CH4 and a GWP of 265 for N2O. 
[b] AdvanSix, 2020; and BASF, 2020
[c] 2019 value from Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2019 of 1.4 million mt CO2e using AR4 GWP value of 298 for N2O converted to 1.2 million mt CO2e, using AR5 GWP value of 265 for N2O.  
[d] Reported by facilities to GHGRP for RY2019.

2.1	Process Emissions
2.1.1	Estimated N2O Emissions 
The Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2019 (EPA, 2021) (hereafter referred to as the U.S. GHG Inventory) estimates emissions from caprolactam processes at two U.S. facilities in 2019 to be 1.4 million mt CO2e using the GWP values from IPCC's Fourth Assessment Report (AR4), or 1.2 million mt CO2e using the GWP values from the IPCC's Fifth Assessment Report (AR5). Caprolactam, glyoxal, and glyoxylic acid production are all considered part of the same industrial source category in the U.S. GHG Inventory, but the U.S. GHG Inventory does not estimate emissions from the latter two industries due to the lack of available production data.
As noted in Section 1, reporting under TSCA indicates that the total amount of glyoxal imported and domestically manufactured in 2011 was approximately 8,300 mt and was between 4,500 mt and 23,000 mt in each of the subsequent four years, and the liquid-phase oxidation of acetaldehyde with nitric acid process is estimated to account for less than 20 percent of total production (Teles et al., 2015). As noted in Section 3.2, the IPCC estimates the default N2O control efficiency as 80 percent for both glyoxal and glyoxylic acid controls (IPCC, 2007). A global warming potential of 265 was used for N2O. Assuming half of the maximum annual production was domestically manufactured (and the other half was imported), the total annual N2O emissions from domestic production could be estimated as follows:
(23,000 mt glyoxal/yr)(0.5)(0.2)(0.52 mt N2O/mt glyoxal)(1-0.8 eff)(265)
= 63,388 mt CO2e/yr
Based on the annual TSCA reporting for 2012 through 2016, it appears that there may be no glyoxylic acid being produced in the U.S. Thus, the estimated annual N2O emissions from glyoxylic acid production are 0 mt CO2e.
2.1.2	Estimated CO2 Emissions 
Assuming that all hydrocarbon feedstock in the glyoxal process that is not converted to glyoxal is either converted to CO2 in the process or vented and converted to CO2 in the control device provides an estimate of the maximum expected quantity of CO2 emissions. 
The yield of glyoxal in the acetaldehyde-based process is 70 percent (Teles et al., 2015). Thus, 1.43 moles of acetaldehyde would be required to produce 1 mole of glyoxal. Converting to mass using the molecular weights of glyoxal and acetaldehyde (58 and 44, respectively) means 108 mt of acetaldehyde would be required to produce 100 mt of glyoxal (1.43*44/58=1.08). If all of the carbon in the 30 percent of the acetaldehyde that is not converted to glyoxal is ultimately converted to CO2 in flares or combustion units, then 65 mt of CO2 are generated per 100 mt of glyoxal produced (108*0.3*24/44*44/12=65). Applying this factor to the nationwide production of glyoxal using the acetaldehyde process as described above results in the following nationwide estimate of CO2 emissions:
(23,000 mt glyoxal/yr)(0.5)(0.2)(0.65 mt CO2/mt glyoxal)
= 1,500 mt CO2/yr
The yield of glyoxal in the LaPorte process is estimated between 70 and 80 percent (Mattioda and Blanc, 2012). Thus, assuming the typical yield is 75 percent, 1.33 moles of ethylene glycol would be required to produce 1 mole of glyoxal (the reaction equation indicates that 1 mole of glyoxal would be produced from 1 mole of ethylene glycol if complete conversion occurred). Converting to mass using the molecular weights of glyoxal and ethylene glycol (58 and 62, respectively) means 142 mt of ethylene glycol would be required to produce 100 mt of glyoxal (1.33*62/58=1.42). If all of the carbon in the 25 percent of the ethylene glycol that is not converted to glyoxal is ultimately converted to CO2 in flares or combustion units, then 50.4 mt of CO2 are generated per 100 mt of glyoxal produced (142*0.25*24/62*44/12=50.4). Applying this factor to the estimated nationwide production of glyoxal using the LaPorte process results in the following nationwide estimate of CO2 emissions:
(23,000 mt glyoxal/yr)(0.5)(0.8)(0.504 mt CO2/mt glyoxal)
= 4,600 mt CO2/yr
It is not clear how much of the byproducts from both glyoxal processes may be recovered or removed from the processes as liquid wastes. If the amount of carbon in such streams is significant, then the estimated CO2 emissions above would be too high.
As noted above in the discussion of nationwide N2O emissions, there is not enough evidence to suggest that glyoxylic acid is being produced in the U.S. Thus, the estimated annual nationwide CO2 emissions from glyoxylic acid production are estimated to be 0 mt CO2.
2.2	Stationary Combustion
Stationary combustion emissions occur when fossil fuels are combusted to provide energy for the manufacturing equipment and process. For the two facilities that are known to manufacture caprolactam, AdvanSix and BASF, combustion emissions are already reported to the GHGRP under subpart C (General Stationary Fuel Combustion Sources). In Reporting Year 2019, the total combustion emissions for AdvanSix are 547,000 metric tons of CO2 and minimal emissions of methane and nitrous oxide; the total combustion emissions for BASF are 667,000 metric tons of CO2 and minimal emissions of methane and nitrous oxide. For these two facilities, the total combustion emissions of 1.2 million metric tons CO2e are also on the same order of magnitude as the estimated total process emissions of 1.2 million metric tons CO2e.
To understand the general magnitude of the potential stationary combustion emissions for glyoxal production, emissions from facilities that are reporting to a similar chemical production source category, subpart V (Nitric Acid Production), were reviewed. The overwhelming majority of facilities reporting under subpart V between 2015 and 2019 also reported combustion emissions under subpart C (General Stationary Fuel Combustion Sources). In 2019, 75 percent of the facilities that reported process emissions under subpart V also reported stationary combustion emissions that were greater than the associated subpart V emissions (both in terms of mt CO2e). Combustion emissions for glyoxal production could potentially be on the same order of magnitude as the 63,388 mt CO2e emissions from the production process.
For more information on combustion-related GHG emissions (CO2, CH4, and N2O), including calculation methodologies and reporting options, refer to the General Stationary Fuel Combustion Technical Support Document (EPA, 2009a).
3. 	Review of Existing Programs and Methodologies
3.1	Review of Existing Programs
In developing GHG monitoring and reporting options for caprolactam, glyoxal, and/or glyoxylic acid manufacturing processes, a number of existing programs and guideline methodologies were reviewed. Specifically, the following programs were examined:

 2006 Intergovernmental Panel on Climate Change (IPCC) Guidelines for National Greenhouse Gas Inventories. Volume 3, Chapter 3.

 European Union (EU). Commission Implementing Regulation (EU) 2018/2066 of 19 December 2018 on the Monitoring and Reporting of Greenhouse Gas Emissions Pursuant to Directive 2003/87/EC of the European Parliament and of the Council and Amending Commission Regulation (EU) No. 601/2012. January 1, 2021. Available at:  
      https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:02018R2066-20210101&from=EN. 
       
 CARB (California Air Resource Board). Unofficial Electronic Version of the Regulation for the Mandatory Reporting of Greenhouse Gas Emissions. April 2019. https://ww2.arb.ca.gov/mrr-regulation.

 Environment and Climate Change Canada (ECCC). Canada's Greenhouse Gas Quantification Requirements. Version 4.0. December 2020. http://publications.gc.ca/collections/collection_2021/eccc/En81-28-2020-eng.pdf.
      
 United States Environmental Protection Agency (USEPA). Greenhouse Gas Reporting Program (GHGRP).

Each of these programs were reviewed to determine their applicability to the caprolactam, glyoxal, and/or glyoxylic acid production source category and whether they contained information on the types of emissions to be reported, the calculation methodologies, and the monitoring methodologies recommended. The 2006 IPCC Guidelines specify three emissions calculations methods (IPCC, 2006). The EU requires both direct monitoring of abated N2O process emissions from all three processes and mass balance calculations of unabated N2O emissions when the abatement technology fails, or if it is technically infeasible to use abatement technology (EU, 2021). The Canadian and California programs do not require emissions calculation and reporting for caprolactam, glyoxal, or glyoxylic acid processes. While the GHGRP does not require emissions calculation and reporting for caprolactam, glyoxal, or glyoxylic acid processes, it was reviewed for applicable facilities that already report combustion emissions.
It was determined that methods for calculating GHG emissions from the 2006 IPCC Guidelines were appropriate for this source category. These methods are discussed in Section 3.2.
3.2	Calculation Methodologies for Caprolactam, Glyoxal, and Glyoxylic Acid Production Processes
Emissions estimations for the caprolactam, glyoxal, and glyoxylic acid sector are addressed in the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC, 2006). The IPCC describes three methods for estimating national N2O emissions from caprolactam, glyoxal, and glyoxylic acid production, as summarized below:
Tier 1: Multiply total nationwide production quantity by the default uncontrolled N2O generation factor. Assume no control.
Tier 2: Measure plant-specific production quantities and multiply by the default uncontrolled N2O generation factor. If applicable, this quantity should also be adjusted to account for emissions control (i.e., multiply the uncontrolled emissions by the factor [1-control eff * control utilization factor]). The IPCC estimates the default control efficiency as 80 percent for both glyoxal and glyoxylic acid controls. Best practice for estimating emissions from caprolactam production dictates the use of plant-specific measurement data; therefore, no default destruction efficiency was recommended in the IPCC Guidelines for caprolactam. Plant level emissions should then be summed to calculate the nationwide emissions.
The Tier 2 monitoring method requires raw material input or output to be known in addition to a standard emission factor. In addition, Tier 2 requires site-specific knowledge of the use of N2O control technologies. Table 2 gives the default factors for Tier 2. 
The equation for calculating N2O emissions is:
EN2O=i,jEFixPix1-DFjxCDUFj
Where:
EN2O 	= 	Process emissions of N2O, kilograms (kg).
EFi 	=	N2O generation factor for product i (caprolactam, glyoxal, or glyoxylic acid), kg N2O / mt of product produced.
Pi	=	Annual production of product i, (caprolactam, glyoxal, or glyoxylic acid), mt.
DFj	=	Destruction factor for control technology type j, fraction. Default is 0.8 for glyoxal and glyoxylic acid if the facility has a control device and 0 if it does not. Default is 0 for caprolactam.
CDUFj	=	Control device utilization factor for control technology type j, fraction. Use 1 if the control device has no downtime during the year and adjust downward to account for any downtime of the control device (i.e., time the device is not operating).
Table 2. Default Factors
Product
N2O Generation Factor (kg/mt product)
EFi
N2O Destruction Factor (fraction)
DFj
Control Device Utilization Factor (fraction)
CDUFi
N2O Emission Factor (kg/mt product) 
EFi * (1-DFi * CDUFi)

Caprolactam
9.0
0
1.0
9.0
Glyoxal
5,200
0.8
1.0
1,040
Glyoxylic acid
1,000
0.8
1.0
200
Source: IPCC, 2006
Tier 3: Use the same calculation equation as for Tier 2, but use direct measurement to determine plant-specific uncontrolled N2O generation factors and control efficiencies. The measurements may be either flow and concentration data based on periodic sampling or based on the use of continuous emission monitoring. Also measure production at each plant. In some cases, direct measurement may instead include measurements of the GHG concentration in the stack gas and the flow rate of the stack gas using a continuous emission monitoring system (CEMS); however, there is no CEMS for N2O. 
Available literature does not describe mechanisms for the generation of CO2 emissions from caprolactam production, glyoxal production, or glyoxylic acid production. One reference indicates that CO2 is the main byproduct of the LaPorte process (Teles et al., 2015); other CO2 emissions are likely from oxidation of hydrocarbons in control devices. No methods for estimating such CO2 emissions have been identified in the literature as part of this review. Based on the emission estimates in Section 2.1.2, the CO2 process emissions are small compared to the N2O process emissions or the CO2 combustion emissions. While a CEMS option could be considered for CO2 emissions, it appears unlikely that facilities currently monitor their CO2 emissions using CEMS. Therefore, CEMS are not discussed further in this document, although they would result in the most accurate emissions measurements. Either these CO2 process emissions can be neglected, or a carbon balance can be used to estimate the CO2 emissions (i.e., assume all carbon not converted to product is emitted as CO2).
4.	Options for Reporting Threshold
The requirements provided in subpart A of the GHGRP (40 CFR 98.2) specify the reporting thresholds for each direct emitter and supplier source category included under the GHGRP. These thresholds determine whether a facility's annual emissions must be reported under their respective subpart(s). For direct emitters of GHGs, reporting, monitoring, recordkeeping requirements apply to the owners and operators of any facility emitting over 25,000 mt CO2e from all direct emitter source categories, unless the facility is part of an "all-in" source category provided in Table A-3 of subpart A. If the facility is part of an "all-in" source category, they must initially report to the GHGRP under all applicable subparts regardless of their GHG emissions quantities. As part of this source category analysis, the number of facilities expected to report at different reporting thresholds was evaluated. Specifically, we evaluated the number and percent of emissions that would be reported using a 25,000 mt CO2e threshold (combined from all direct emitter source categories) versus no threshold (i.e., "all-in" subpart).  
As discussed in Sections 1 and 2 of this document, there are two known facilities in the United States producing caprolactam, up to four potential facilities in the United States producing glyoxal, and no known facilities understood to produce glyoxylic acid. The two known facilities producing caprolactam already report combustion emissions to the GHGRP well above the potential 25,000 mt CO2e reporting threshold. The estimated process emissions are also well above the potential 25,000 mt CO2e reporting threshold. It is estimated that 20 percent of the glyoxal is produced using the acetaldehyde and nitric acid process, so we assumed that one of the two potential domestic producers of glyoxal uses that process. The remaining two potential glyoxal facilities are assumed to be importers with zero emissions from the production of glyoxal. The process emissions from the facility using the acetaldehyde and nitric acid process are estimated to be approximately 65,000 mt CO2e, which is greater than the reporting threshold. The process emissions from the one potential domestic producer of glyoxal assumed to be using the LaPorte process is approximately 4,600 mt CO2e per year.
Because none of the potential glyoxal manufacturers currently report to the GHGRP, it is difficult to determine which, if any, additional subparts may cause the facilities using the LaPorte process to exceed a potential 25,000 mt CO2e reporting threshold. It is likely that stationary combustion sources would be present at these facilities. However, for this analysis, we assumed all glyoxal producers using the LaPorte process were below the 25,000 mt CO2e reporting threshold because the additional LaPorte process CO2 emissions are small and unlikely to cause a facility not currently reporting to now exceed that threshold. 
Table 3 summarizes the results of this threshold analysis. Adding caprolactam, glyoxal, and glyoxylic acid production as an all-in source category (i.e., regardless of their emissions profile) is a conservative approach to gather information from all applicable facilities and accounts for the uncertainty in the data and assumptions used in this emissions analysis. However, adding caprolactam, glyoxal, and glyoxylic acid production as a 25,000 mt CO2e reporting threshold category is expected to require 99.6 percent of the emissions from the source category to be reported. 

           Table 3.  Evaluation of Alternative Threshold Options[a]
                            Option/Threshold Level
                             Emissions Covered[b]
                             Facilities Covered[b]
                                       
                                 mt CO2e/year
                                    Percent
                                    Number
                                    Percent
                               Option 1:  All-in
                                  1.2 million
                                      100
                                       6
                                      100
                         Option 2:  >25,000 mtCO2e
                                  1.2 million
                                     99.6
                                       3
                                      50
[a] Emissions calculated using AR5 GWPs of 28 for CH4 and a GWP of 265 for N2O. 
[b] Considers process emissions only.

5.	Options for Monitoring Methods
Two separate monitoring methods were considered for this technical support document: an emission calculation for N2O and mass balance for CO2 (Option 1) and direct measurement (Option 2). Option 1 for N2O and Option 2 are based on the IPCC Tier 2 and 3 methodologies, respectively, for caprolactam. The IPCC guidelines provide the methodology for caprolactam in detail, but the introduction also specifically notes that the methodology is "suitable for application to estimation of emissions from glyoxal and glyoxylic acid" (IPCC, 2006). Option 1 for CO2 is based on other subparts in the GHGRP that use a mass balance method (e.g., subpart X). All of these options require annual reporting.
5.1	Option 1: Emission Calculation / Mass Balance
Option 1 for N2O generally follows the IPCC's Tier 2 protocol. The Tier 2 monitoring method requires raw material input or output to be known in addition to a standard emission factor. In addition, Tier 2 requires site-specific knowledge of the use of N2O control technologies. The volume or mass of each product would be measured with a flow meter or weigh scales. 
The mass balance procedures for CO2 would be similar to mass balances required for existing subparts in the GHGRP such as petrochemical production (subpart X). Specifically, site-specific quantities and carbon contents of all carbon-containing feedstocks and products would be determined, and CO2 emissions would be calculated assuming all carbon from the feedstock that does not end up in product is emitted as CO2. Note that liquid wastes that are not combusted should be considered to be a product for the purposes of the mass balance emissions calculation. The volume or mass of each feedstock and product would be measured with a flow meter (other options such as weighing might also be considered). As noted above, the production quantities are expected to be small, and the CO2 process emissions relatively low. Thus, instead of measuring carbon content, the carbon content of each feedstock and product would be determined based on the quantity of each stream and the molecular formula of the applicable pure compounds, adjusted as necessary to account for significant byproducts or diluents in the stream (e.g., water in the glyoxal solution product stream). Separate equations like Equations X-1, X-2, and X-3 in 40 CFR part 98, subpart X, would be used to determine the monthly difference in carbon content between inputs and outputs (i.e., all gaseous feedstocks/products, all liquid feedstocks/products, and all solid feedstocks/products), and then summed over all months in the reporting year. The three resulting annual values would be summed and multiplied by the ratio of the molecular weight of CO2 to the atomic weight of carbon to calculate the annual mass of CO2 emissions. If process vent emissions are routed to a combustion unit that is subject to subpart C reporting, then the quantity of the vent gas should not be included in the total quantity of fuel reported under subpart C. (Note: if the N2O is controlled using something other than thermal or catalytic destruction that would not convert hydrocarbons to CO2, then this option would need to be modified for CO2. Additionally, if the quantity of carbon-containing liquid waste streams is substantial so that the difference in the amount of carbon between the feedstocks and products is small, then measurement inaccuracies may make it difficult to calculate accurate emissions.)
5.2	Option 2: Direct Measurement
For industrial source categories for which the process emissions and/or combustion GHG emissions are contained within a stack or vent, direct measurement may include periodic measurement of the GHG concentration in the stack gas and the flow rate of the stack gas using periodic stack testing on a specified schedule and whenever a plant makes any significant process changes that would affect the N2O or CO2 emission rates. 
For direct measurement using stack testing, sampling equipment would be brought to the site when needed and installed temporarily in the stack to withdraw a sample of the stack gas and measure the flow rate of the stack gas. If there are any emissions controls, the testing should be conducted for both the uncontrolled stream from the process (i.e., the control device inlet) and the control device outlet. The emissions are calculated from the concentration of GHGs in the stack gas and the flow rate of the stack gas. Specifically, the results are used to develop the numerator for a site-specific emission factor (kg N2O) as well as a site-specific destruction factor for the control technology (if applicable). The volume or mass of caprolactam, glyoxal, and glyoxylic acid during the test period would be determined using flow meters (or potentially other devices) and used as the denominator in the site-specific emission factor. The factors are then used along with the site-specific production rates in the same equation as in Option 1. (One exception is that emissions from any separately vented streams to flares would be calculated as currently specified in 40 CFR 98.253 (subpart Y).) Under this Option, fugitive emissions would be assumed to be small and would not be calculated.
A stack test provides a periodic measurement of the emissions rather than continuous measurement of the emissions. A method using periodic, short-term stack testing would be appropriate for those facilities where process inputs and process operating parameters remain relatively consistent over time. In cases where there is the potential for significant variations in the process input characteristics or operating conditions, more frequent testing (or potentially continuous measurements, for CO2) would be needed to accurately record changes in the actual GHG emissions from the sources resulting from any process variations.
As in Option 1, the volume or mass of each product would be measured with a flow meter or weigh scales.
 QA/QC Requirements
In order to ensure the quality of the reported GHG emissions, the following quality assurance/quality control (QA/QC) activities are considered important:
      (1)	All meters or monitors (e.g., feedstock/product flow, gas composition, etc.) that are used to provide data for the GHG emissions calculations should be calibrated prior to the first reporting year, using a suitable method published by a consensus standards organization (e.g., ASTM, ASME, API, AGA, etc.), or as specified by the meter/monitor manufacturer. Said meters or monitors shall be recalibrated either annually or at the minimum frequency specified by the manufacturer. 
      (2)	Calibration documentation should be maintained. The estimated accuracy of measurements made with these devices should also be recorded, and the technical basis for the estimates should be provided. 
      (3)	All CO2 CEMS and flow rate monitors used for direct measurement of GHG emissions should comply with QA procedures for daily calibration drift checks and quarterly or annual accuracy assessments, such as those provided in Appendix F to Part 60 or similar QA procedures. 
 Options for Estimating Missing Data
Options and considerations for missing data will differ depending on the proposed monitoring method. Each option would require a complete record of all measured parameters as well as parameters determined from company records that are used in the GHG emissions calculations (e.g., production rates).
7.1	Procedures for Option 1: Emission Calculation / Mass Balance
      For process sources that use an emission calculation for N2O emissions, the emission calculation is derived from default emission factors and activity data. No missing data procedures would apply for production rates because businesses closely track those rates and 100 percent data availability would be expected. For the control device utilization factor, assuming that the control device operation is generally constant from year to year, the substitute data value could be the most recent quality-assured value.
      For process sources that use a mass balance for CO2 emissions, determine substitute values for missing feedstock and product quantities using the procedures for missing fuel usage in 98.35(b)(2), and for missing feedstock and product carbon contents and molecular weights (for gaseous feedstocks and products), using the same procedures as for missing carbon contents and molecular weights specified for fuels in 98.35(b)(1).
7.2	Procedures for Option 2: Direct Measurement
      For options involving direct measurement of vent stream flow rate and composition using stack testing, "missing data" is not generally anticipated. Stack testing conducted for the purposes of compliance determination is subject to quality assurance guidelines and data quality objectives established by the U.S. EPA, including the Clean Air Act National Stack Testing Guidance published in 2009 (EPA, 2009b). The 2009 Guidance Document indicates that stack tests should be conducted in accordance with a pre-approved site- specific test plan to ensure that a complete and representative test is conducted. Results of stack tests that do not meet pre-established quality assurance guidelines and data quality objectives would generally not be acceptable for use in emissions reporting, and any such stack test would need to be re-conducted to obtain acceptable data.
      The U.S. EPA anticipates that test plans for stack tests that are expected to be used to obtain data for the purposes of emissions reporting would be made available to EPA prior to the stack test and that the results of the stack test would be reviewed against the test plan prior to the data being deemed acceptable for the purposes of emissions reporting.
      No missing data procedures would apply for production rates because businesses closely track those rates and 100 percent data availability would be expected. For the control device utilization factor, assuming that the control device operation is generally constant from year to year, the substitute data value could be the most recent quality-assured value.
 References
ACC (2020) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA. 
AdvanSix (2020). AdvanSix Hopewell Virginia Information Sheet. Accessed June 21, 2021. Available at: https://www.advan6.com/hopewell.
BASF (2020). BASF: Freeport, Texas Fact Sheet. Accessed June 21, 2021. Available at: https://www.basf.com/global/documents/en/about-us/strategy-and-organization/verbund/BASF_Freeport.pdf.
CARB (California Air Resource Board). Unofficial Electronic Version of the Regulation for the Mandatory Reporting of Greenhouse Gas Emissions. April 2019. https://ww2.arb.ca.gov/mrr-regulation.
ChemView (2021). Compilation of data submitted under TSCA in 2012 and 2016. Accessed April 2021. Available at https://chemview.epa.gov/chemview.
Cogburn, M.O. (2012). United States v. Emerald Carolina Chem., LLC. Consent Decree, Civil Action No. 3:12-cv-00554. United States District Court for the Western District of North Carolina, Charlotte Division. Decided October 25, 2012. Available at https://casetext.com/case/united-states-v-emerald-carolina-chem.
Environment and Climate Change Canada (ECCC). Canada's Greenhouse Gas Quantification Requirements. Version 4.0. December 2020. http://publications.gc.ca/collections/collection_2021/eccc/En81-28-2020-eng.pdf.
Ecofys, et al. (2009). Methodology for the free allocation of emission allowances in the EU ETS post 2012: Sector Report for the Chemical Industry. Prepared by Ecofys, Fraunhofer Institute for Systems and Innovation Research, and Oko-Institut for the European Commission. November 2009. Available at https://ec.europa.eu/clima/system/files/2016-11/bm_study-chemicals_en.pdf.
EPA (U.S. Environmental Protection Agency) (2009a). General Stationary Fuel Combustion Technical Support Document, U.S. EPA Office of Air and Radiation, January 30, 2009. Available at https://www.regulations.gov/document/EPA-HQ-OAR-2008-0508-0004.
EPA (U.S. Environmental Protection Agency) (2009b). Clean Air Act National Stack Testing Guidance, U.S. EPA Office of Enforcement and Compliance Assurance, April 27, 2009. Available at https://www.epa.gov/sites/default/files/2013-09/documents/stacktesting_1.pdf
EPA (U.S. Environmental Protection Agency) (2021). Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2019. EPA 430-R-21-005. April 14, 2021. Available at https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks-1990-2019. 
EU (European Union) (2021). Commission Implementing Regulation (EU) 2018/2066 of 19 December 2018 on the Monitoring and Reporting of Greenhouse Gas Emissions Pursuant to Directive 2003/87/EC of the European Parliament and of the Council and Amending Commission Regulation (EU) No. 601/2012. January 1. Available at:  https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:02018R2066-20210101&from=EN.
Fact.MR (2019). "Glyoxal Sales Surge as a Versatile Chemical Intermediate that Complements Sustainability-driven Efforts of Manufacturers, finds Fact.MR." February 20, 2019. Accessed April 2021. Available at https://www.factmr.com/media-release/882/glyoxal-market.
Fior Markets (2018). Summary of Global Glyoxylic Acid Market by Manufacturers, Regions, Type and Application, Forecast to 2023. July 2018. Available at: https://www.fiormarkets.com/report/global-glyoxylic-acid-market-by-manufacturers-regions-type-268394.html. 
IPCC (The Intergovernmental Panel on Climate Change) (2006). 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the IPCC National Greenhouse Gas Inventories Programme. H.S. Eggleston, L. Buendia, K. Miwa, T Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan. Available at https://www.ipcc-nggip.iges.or.jp/public/2006gl/.
IPCC Fourth Assessment Report (2007). Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, Pachauri, R.K and Reisinger, A. (eds.)]. IPCC, Geneva, Switzerland, 104 pp. Available at: https://www.ipcc.ch/report/ar4/wg1/.
Mattioda, G., and A. Blanc (2012). "Glyoxal." Ullmann's Encyclopedia of Industrial Chemistry. Wiley-VCH Verlag GmbH & Co. KGaA. Weinheim. DOI: 10.1002/14356007.a12_491.pub2. 2012.
NIH (National Institutes of Health) (n.d.-a). "PubChem Compound Summary for CID 7768, Caprolactam." Accessed June 2021. Available at https://pubchem.ncbi.nlm.nih.gov/compound/Caprolactam.
NIH (n.d.-b). "PubChem Compound Summary for CID 7860, Glyoxal." Accessed April 2021. Available at https://pubchem.ncbi.nlm.nih.gov/compound/glyoxal.
Teles, J.H. et al. (2015). "Oxidation." Ullmann's Encyclopedia of Industrial Chemistry. Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim. 10.1002/14356007.a18_261.pub2.
