<


Correspondence



October 19, 1998

    (AR-18J)


Mike Hopkins, Manager
Air Quality Modeling and Planning
Division of Air Pollution Control
Ohio Environmental Protection Agency
P.O. Box 1049
Columbus, Ohio 43216-1049


Dear Mr. Hopkins:

The purpose of this letter is to provide the Ohio Environmental Protection Agency (OEPA) with the United States Environmental Protection Agency’s (USEPA) unresolved comments regarding British Petroleum Chemicals (BP) draft Permits to Install (PTI) for their proposed Butanediol (BDO) Manufacturing Plant (PTI #03-1250) and Urea Granulation Plant (PTI #03-9943).  This letter is a result of USEPA’s phone conversation with Mr. Alan Lloyd of the OEPA on October 7, 1998.  USEPA comments are as follows:  


PTI #03-1250, BDO Manufacturing Plant

BP did not perform a proper Best Achievable Control Technology (BACT) analysis for Volatile Organic Carbon (VOC) and Carbon Monoxide (CO), both of which exceed threshold emission levels with increases in emissions of 456.06 tons/year and 844.92 tons/year, respectively.  BP did perform a BACT analysis for Nitrogen Oxide (NOx) emissions (134.77 tons/year) on each of the major emission units (the Scrubber Off-gas Incinerator (SOGI) and Hydrogen Reforming Furnace).

BP must perform a proper BACT analysis for VOC and CO as required in 40 CFR 52.21 and outlined in the 1990 USEPA Draft New Source Review Manual.  This BACT analysis includes:  (1) identifying and listing all control technologies including Lowest Achievable Emission Rate technologies; (2) eliminating technically infeasible options by clearly documenting technical difficulties, through physical, chemical, and engineering principles, that would prove the system as being infeasible; (3) ranking the remaining control technologies using: (a) control effectiveness, (b) expected emission rates, (c) expected emission reductions, (d) environmental impacts, and (e) economic impacts;
(4) evaluating and documenting the most effective controls on a
case-by-case basis considering the energy, environmental, and
-2-

economic impacts; and (5) selecting the most effective control option as BACT.

The BACT analysis performed on NOx emissions did not adequately address the top down BACT criteria as described above, as all possible control technologies were not investigated thoroughly.  These possible control technologies may include Flameless Thermal Oxidation (addressed in a October 16, 1998, meeting between BP, OEPA, and USEPA), a Scrubber, Afterburner, Low NOx Burners  and/or Condenser.  

In addition, the NOx BACT analysis was performed on the major emission units of the process rather than the entire process as a whole.  This resulted in an increased cost analysis that omitted a potentially feasible control technology (Selective Catalytic Reduction (SCR)).  USEPA, through the aid of a vendor, determined that had BP performed the BACT analysis on the entire process as a whole, SCR could prove to be cost effective at an estimated rate of approximately $3,000.00 per ton of NOx removed at a 90% removal efficiency.  This estimated cost is considerably less than the $13,421.00 per ton of NOx removed at a 90% removal efficiency as proposed by BP for the single SOGI emission unit.    

PTI #03-9943, Urea Granulation Plant

BP submitted 1994-1995 baseline NOx emissions data and netting credits from 1992 to avoid Prevention of Significant (PSD) review for NOx for the 1998 construction of PTI #03-9943.  

When making a netting analysis, a source must submit the two most recent years of historical emissions data or submit sufficient information to justify otherwise by demonstrating that the two most recent years are not representative of normal source operation (see page A.39 of the 1990 USEPA Draft New Source Review Manual).  

40 CFR 52.21(b)(1)(3) states that an increase or decrease in actual emissions is contemporaneous with the date five years before actual construction of the modification.  As a result, the NOx emission credits that BP received in 1992 are not applicable to the 1998 PTI #03-9943 as the 1992 credits are beyond the five year contemporaneous period of the construction date for
PTI #03-9943.  


PTI #03-1250 and #03-9943

OEPA operates a delegated PSD program which allows a thirty day appeal period following the PTI’s thirty day comment period.  OEPA should add language in the permits that states both
-3-

PTI #03-1250 and #03-9943 can be appealed to the Environmental Appeals Board within the thirty day appeal period under 40 CFR 124.

In conclusion:

1) BP must perform a proper BACT analysis as required in
40 CFR 52.21(j) and outlined in the 1990 USEPA Draft New
Source Review Manual for the following pollutants from
PTI #03-1250: VOC, CO and NOx.  This analysis must be
included in the public record.

2) BP must use the netting analysis procedures as described in
40 CFR 52.21(b)(1)(3) and outlined in the 1990 USEPA Draft New Source Review Manual for PTI #03-9943.  This analysis
includes submitting the two most recent years of historical
emissions data or submitting sufficient information to
justify otherwise.  BP may not use netting credits beyond
the five year contemporaneous period, which includes the credits from 1992, to avoid PSD review.  Without the 1992 NOx credits, this project would have an increase in NOx emissions of 359.86 tons per year.  Therefore, PSD review is required for NOx.

3) OEPA must include the appeals language, as specified in 40
CFR Part 124, for both PTI #03-1250 and #03-9943 to provide for the thirty day appeal period.  

Based on the issues outlined above, it is the position of the USEPA that both PTI #03-1250 and #03-9943, as currently drafted, do not meet the requirements of the Clean Air Act.  We will continue to work with you to develop an acceptable permit.  If we can answer any questions regarding these comments, please contact Steve Gorg, of my staff, at (312) 353-4145.


Sincerely yours,

  /s/

Genevieve Damico, Acting Chief
Permits and Grants Section (IL/IN/OH)