March 4, 1994 Mr. Jack Boley Safety & Environmental Affairs Manager DELHI GAS PIPELINE CORPORATION First City Center 1700 Pacific Avenue Dallas, TX 75201 Re: Control Requirements for Glycol Dehydration Units Dear Mr. Boley: This letter is in response to the meeting held in Austin on February 8, 1994 between industry representatives and several members of the Chemical Section. The purpose of the meeting was to discuss the use of condensers as final abatement control for reboiler vents in glycol dehydration units. During the meeting, you and other industry representatives presented additional testing results on condenser efficiency at controlling benzene, toluene, ethylbenzene, and xylene (BTEX) emitted from reboiler vents. All of the testing results discussed during the meeting were on condensers which use rich glycol coming off the absorber as the cooling medium and have a flash tank upstream. Therefore, our comments below will only concern condensers which meet these two requirements. We will consider other designs on a case-by-case basis. For those individuals who submit PI-8 registrations for Standard Exemption No. 66, we will review your registration and send a letter which explains whether or not we agree with the exemption claim. If a company proposes to use a condenser as a final abatement on a glycol reboiler, we will require no further demonstration as part of the exemption claim, only if the following conditions are met: (1) no more than 80 percent reduction of annual emissions is claimed, (2) the company completes a one-time efficiency testing on the condenser to assure that item (1) is being met, (3) 25 tons per year VOC limitation is not exceeded, and (4) all other requirements of Standard Exemption No. 66 are met. It is our position that if more than 80 percent reduction is needed in order to meet the 25 tpy limitation, much more stringent monitoring and testing (i.e., on a continuous basis) for the condenser would be necessary. In that case, a condenser probably should not be used as the final abatement device. At those levels, it would be more appropriate to route the condenser's vent to a combustion device, such as flare, incinerator, reboiler firebox, or equivalent level of control. I hope this letter addresses the concerns that were raised during our meeting. We would appreciate if you would share this letter with all interested parties and industry trade groups. If you should have any additional comments or questions, please contact me. Sincerely, Tammy Villarreal Manager, Chemical Section New Source Review Program cc: Mr. Allan Dees, Texaco Exploration & Production, Inc. Mr. Vernon H. Schievelbein, Texaco Mr. Curtis O. Rueter, Radian Corporation bcc: Chemical and Combustion Section Engineers Air Program Managers, Regional Offices No. 1 - 15 Texas Natural Resource Conservation Commission Interoffice Memorandum To: Interested Parties Date: May 31, 1996 From: Tammy Villarreal, Chemical Section Chief Subject: Control Requirements for Glycol Dehydration Units Reference: TNRCC Letter, subject as above, dated March 4, 1994, signed by Tammy Villarreal Many glycol dehydration vent condenser systems are being designed with one or more features which ensure the uncondensed vent stream fraction is always combusted before it reaches the atmosphere. Operators with such units now ask exactly what requirements of the referenced letter apply to their claim of standard exemption or standard permit. The simple answer is none. The March 4, 1994 letter applies to dehydrators whose FINAL vent control is a condenser. Systems which direct the uncondensed fraction to the firebox or an afterburner have final control beyond the condenser. We typically recognize a destruction and removal efficiency of 98% for properly designed combustion units such as a dehydrator firebox or flare. The two percent uncombusted VOC amount should be determined based on the estimated vent load. The estimated vent load may be computed using any of the conservative software programs accepted by the TNRCC Office of Air Quality or through a rich-lean analysis conducted according to the protocol prepared by the Gas Research Institute. No condenser efficiency test is required when the dehydrator still vents are combusted in the final control step.