MEMORANDUM   

DATE:		08/03/2020

TO:	National Emission Standards for Hazardous Air Pollutants: Coke Ovens: Pushing, Quenching, and Battery Stacks Residual Risk and Technology Review; Docket ID: EPA-HQ-OAR-2002-0085

FROM:	Donna Lee Jones and Ted Palma, U.S. EPA, Office of Air Quality Planning and Standards (OAQPS), Research Triangle Park, NC; Gabrielle Raymond, RTI International, Research Triangle Park, NC 

SUBJECT:	Conference call with Coke Industry Representatives   
--------------------------------------------------------------------------------


Summary

      A telephone conference was held on July 29, 2020, between the EPA, the Coke Oven Environmental Task Force (COETF), and coke company representatives as a follow-up to the information collection request (ICR) sent to selected facilities in the industry in 2016 and the quality assurance (QA) review performed by the companies of their data, as part of the Pushing, Quenching, and Battery Stacks (PQBS) Residual Risk and Technology Review (RTR) for 40 CFR part 63, subpart CCCCC. 
      
      Industry representatives discussed their ideas on estimating emissions for the facilities who did not test for the ICR, which included using the emission rates from facilities that tested for the ICR along with calculated coal charged/coke produced data to use with emission factors, where emission factors would be used to develop emissions estimates for the nonICR facilities. Industry representatives stated that the emission factors should be developed per type of coke produced (e.g., foundry vs. blast furnace coke). Industry representatives wanted the EPA to confirm that for runs below the detection limit (BDL) the EPA would use half of the detection level to calculate the emissions. Industry representatives also noted that the emission factors could be used with permit limits of coal production to calculate allowable emissions. 
      
      Industry representatives stated that combustion sources and other sources that pass through a control device should be modeled as cumulative HAP risk and not as coke oven emissions (COE). They also stated that the quench tower testing at the ArcelorMittal Monessen, PA, facility over estimated emissions and that the toluene soluble organics testing at the U.S. Steel Clairton, PA, facility had testing issues and, thus, also overstated emissions. 
      
      Industry representatives described a potential methodology for estimating emissions for pushing fugitives for metals and semi-volatiles that included using 1998 uncaptured pushing baghouse inlet test data from ABC Coke and ArcelorMittal Burns Harbor, and capture efficiencies; they offered to provide this information to the EPA. Industry representatives also suggested that the ICR pushing test data could be used with estimated capture efficiencies to estimate the pushing fugitive emissions.
      
      Industry representatives inquired as to how the EPA was planning to estimate emissions for nonPQBS sources such as charging, door leaks, topside leaks, flares, Coke Oven Gas flares, and coke by-product recovery plants. They wanted to know if National Emissions Inventory (NEI) data would be used and if so which year. They also wanted to confirm with the EPA that they would have an opportunity to review the emissions estimates.
      Industry representatives presented a methodology for estimating actual emissions from allowable emissions using a ratio of actual-to-allowable coal charge. Some coke facilities have permits that have limits of allowable coal charge. They found that the ratios ranged from 1:1 to 1:3.
      
      Industry representatives wanted to know how certain sources should be characterized in terms of modeling, i.e., as a point source or volume source. Industry representatives also noted the difficulties of modeling coke oven batteries and with using buoyant line plume (BLP) modeling. Industry representatives suggested a few alternatives for modeling coke battery emissions, such as using facility-specific buoyancy parameters and plume heights, and calculating an BLP buoyancy parameter, "F", for other facilities. 
      
      Industry representatives mentioned that a modeling acute toxicity factor of 10, commonly used by the EPA, was not appropriate for this industry. They described why they believe a factor of two should be used. They conducted an analysis on hourly pushing values from five coke plants. The ratio of the 99[th] percentile of the number of pushes per clock hour to annual mean number of pushes per hour were calculated. The ratios ranged from 1.26 to 2.06, thus the recommendation to use an acute factor of two.
      
      Industry representatives suggested that the EPA examine and adjust several receptor locations at coke facilities that may be onsite or not representative of populations. Industry offered to send the EPA the property boundary information for each facility as well as their proposed receptor location changes. be removed or relocated.
      
      Industry representatives inquired about the rationale behind the recent litigation and court decision regarding the 2005 RTR for the Subpart L facilities. 

EPA Response

      The EPA responded that the ICR data would be used preferentially, supplemented, as needed, by the most recent NEI data (2017) and 2005 Coke Ovens RTR for subpart L. The BLP model is now a part of AERMOD; therefore, it will be used for the subpart CCCCC RTR for buoyant line sources. In addition, the EPA confirmed that emissions from control devices would not be considered COE because the control device alters the nature of the COE stream and post-control emissions have different toxicity profiles than COE. These after-control emissions would be modeled as the individual HAP. The EPA confirmed that using half the detection limit was the typical EPA protocol in risk modeling for BDL data. 

      The EPA stated that the industry will get the opportunity to review the average emissions estimates for the nonICR facilities and the draft modeling file in further QA rounds. Dr. Jones thanked the industry for their help in reviewing their ICR emissions data in the prior two rounds of QA.

Follow-up Action Items

 Industry will provide 1998 data on pushing inlet emissions to potentially use as pushing fugitive emissions profile.
 Industry will provide data used to develop acute toxicity factor of two. 
 Industry will provide information regarding adjusting receptor locations, including property boundaries for each facility.
 Industry will confirm the type of coke produced at the Bluestone Birmingham, AL, facility. 
 EPA will provide the Erie Coke ICR data for review by foundry coke facilities. and then use the Erie data for these facilities rather than the ICR averages based on blast furnace coke facilities.


ATTENDEES
      
Donna Lee Jones, EPA/OAQPS
Ted Palma, EPA/OAQPS
Chris Owen, EPA/OAQPS
Gabrielle Raymond, RTI International

David Ailor, COETF
Mary Kaplan, AECOM
Allen Dittenhoefer, Montrose
Jeffrey Knight, Pillsbury Winthrop Shaw Pittman LLP

AK Steel
Katie Kistler
Jason Aagenes
Michael Long
Pat Gallo
Chris Potts

ABC Coke 
Jay Cornelius 
Abigail Anderson 
Bill Osborn 

ArcelorMittal
Rich Zavoda
Julianne Kurdila
Simonne Benoit
Marian Gammon
Rob Maciel
Danielle Skolnekovich

DTE
Brenna Harden
Robert Sanch

US Steel
Christopher Hardin
Jonelle Scheetz
David Hacker
Mike Dzurinko

SunCoke 
Katie Batten
Aron Schnur
Nick Tang

Bluestone Coke 
Charles Jones
