SCIENTIFIC REVIEW COMMITTEE MEETING -
July 24, 2003

MEETING HIGHLIGHTS

 Marty Kay welcomed the SRC members and the audience to the meeting.  Marty Kay announced that Todd Wong has replaced Grant Chin as the CARB representative on the committee.  The topics listed below were discussed during the meeting.

¾    Minutes of May 22nd Meeting

¾    Responses to Comments from May 22nd Meeting

¾    New and Updated BACT - Part B Listings

¾    Other Business

Minutes of the May 22nd Meeting

The minutes of the May 22^nd meeting were approved as written.

Responses to Comments from the May 22nd Meeting

AQMD stated that changes to the new Part B listings and minutes from the prior meeting that had been agreed upon at the May 22nd meeting had been made, and the updated documents could be viewed on AQMD’s web site.

With regard to the UC San Diego gas turbine/SCONOx listing, two changes had been made to the listing based on the discussion at the May 22nd meeting.  These changes were: (1) more had been added about the SoLoNOx combustion technology used to maintain low NOx emissions from the gas turbine and (2) the listing had been revised to state that the SCONOx sorbent deactivation mechanism was physical (masking) rather than chemical.

With regard to the Jacuzzi Whirlpool Bath fiberglass/polyester resin operation listing, AQMD had considered a number of comments that were made at the meeting.  One comment was received by e-mail from a committee member following the meeting.  The committee member suggested that BACT guidelines of this type, which require a minimum control efficiency, should include an alternative control requirement based on a maximum outlet concentration.  This allows compliance to be maintained if the process operates for some period of time at such a low VOC level that the minimum control efficiency requirement would be unreasonable.  Although there was no basis for incorporating an alternative concentration limit in the Jacuzzi Whirlpool Bath listing (the BACT listing must adhere to the BACT determination in each case), the BACT Team felt that this was a good suggestion and passed it along to the relevant permitting teams in the form of an e-mail memo.

With regard to the same listing, another committee member had commented that the facility may not be meeting the permanent total enclosure criteria and also may experience difficulty with fouling of the zeolite concentrator due to exposure to styrene in hot air.  AQMD had discussed these potential problems with the permitting team, and the team agreed to address these issues in the source test.  First, the test protocol will include U.S. EPA Method 204 to verify that the facility is meeting the permanent total enclosure criteria.  Second, it was verified in discussions with the facility’s consultant that there are no plans to install fresh zeolite prior to testing, and the permitting team will require that the source test include documentation of the age of the zeolite, which should be approximately 16 months, at the time of the source test.

The same committee member had asked what methods would be used to fulfill the permit requirement that the zeolite and oxidation catalyst be activity-tested semi-annually.  AQMD had obtained the test protocol that the facility planned to use for activity-testing the zeolite and it seemed acceptable.  The protocol seemed to be also appropriate for testing the catalyst, however the facility had not yet made that decision.  Another committee member had asked whether urea was being used in this process, and AQMD had determined that it was not.  An audience member had requested cost information on the control system.  AQMD had requested the information, but the facility had declined to share this information. (Howard Lange, AQMD)

New BACT Part B, Section I Listings

I.C. Engine, Emergency, Compression-Ignition, Claremont Manor (A/N 387480)

This 550 hp engine was installed in 2001, but the facility did not apply for a permit until 2002.  The engine is operated every two weeks for 20 to 45 minutes at approximately 20% load.  Due to the engine being located adjacent to a school, the facility agreed to install a diesel particulate filter, which was installed in May of 2002.  Since the engine’s normal operating load is insufficient to bring the filter to a sufficiently high temperature to burn off accumulated carbonaceous material, the filter must be returned to the manufacturer for kilning when its pressure drop becomes too high.  The engine was source tested in July 2002 using AQMD Method 5.2.  The total PM emission was .046 g/hp-hr, and the filterable PM was .011 g/hp-hr.  The U.S.EPA-certified emissions data for the engine family to which this engine belongs indicated that the total PM emissions, using the test method specified by EPA, are approximately .06 g/hp-hr at 20% load.  AQMD’s information was that the filterable portion of PM measured using Method 5.2 is comparable to the total PM measured using the EPA method.  This comparison suggested that the filter is removing a significant amount of the PM emitted from the engine.  Based on the operating history and test results, a diesel particulate filter is now achieved-in-practice technology for emergency diesel engines. (Howard Lange, AQMD)

Discussion:  A committee member asked why a clean-fuel engine or turbine was not required.  AQMD responded that the added expense of a clean-fuel engine is not justified for emergency engines since they experience very little operation. (Hal Taback, Hal Taback Consulting; Marty Kay, AQMD)

A committee member asked how the engine family emissions are calculated.  AQMD responded that a weighted average of results at various loads is used for comparison to emission standards, but the .06 g/hp-hr figure was interpolated for 20% load using the certification data at 10% load and 25% load.  Another committee member noted that the EPA method is not synonymous with Method 5.2 filterable particulate because it incorporates a dilution tunnel, which is believed to allow some condensation to occur upstream of the filter.  It was agreed that the listing would be corrected in this regard. (Greg Adams, Los Angeles County Sanitation Districts; Howard Lange, AQMD; Gary Rubenstein, Sierra Research)

A committee member noted that the listing stated the maximum heat input to the engine in terms of gallons per hour of fuel.  AQMD responded that this was the only information available in the manufacturer specifications, but also noted that an approximate figure can be calculated based on a typical diesel fuel heating value of 140,000 Btu/gallon. (Phillip Hodgetts, Clean Air NOW; Howard Lange, AQMD; Marty Kay, AQMD)

A committee member asked if the facility was required under Title 22 to operate the engine every two weeks.  AQMD responded that that was the information provided by the facility contact person. (Greg Adams, Los Angeles County Sanitation Districts; Howard Lange, AQMD)

A committee member suggested that the cost of kilning the filter should be included in the listing.  AQMD agreed to do so.  The committee member also commented that the emissions produced by the kiln should be compared to the emissions controlled by the filter.  AQMD responded that the toxic character of diesel particulate should also be considered.  The committee member responded that a BACT requirement should make sense in terms of criteria pollutants.  AQMD agreed that it probably should and noted that this subject would be considered further in conjunction with the Kings County emergency diesel engine listing to be discussed later in the meeting. (Gary Rubenstein, Sierra Research; Marty Kay, AQMD)

Fluidized Catalytic Cracking Unit, ExxonMobil (A/N 344221)

ExxonMobil, at it’s Torrance refinery, has installed a selective catalytic reduction (SCR) reactor downstream of the fluidized catalytic cracking unit (FCCU) to control NOx emissions.  The SCR process selectively reduces NOx using ammonia.  This installation, which was done to reduce NOx emissions for purposes of the RECLAIM program, is the first application of SCR to an FCCU in the U.S.  The SCR reactor was installed downstream of an electrostatic precipitator (ESP), which treats the FCCU flue gas.  This ESP operates at relatively high temperature, and the flue gas leaving the ESP is in a temperature range which is also a suitable for SCR operation.  Ammonia is injected upstream of the ESP for fly ash conditioning, and the flue gas leaving the ESP contains sufficient ammonia for the SCR reactions.  The SCR reactor thus not only reduces NOx but also reduces the amount of residual ammonia in the flue gas.  The SCR system was started up in April of 2000.  Debugging was complete by the end of 2000, and the system has operated well since that time.  The FCCU is equipped with continuous emission monitors both upstream and downstream of the SCR, both of which are subject to semi-annual relative accuracy test audits (RATAs).  Data from the most recent two RATAs show that the SCR system is reducing the NOx emission by more than 90%.  SCR technology is thus achieved in practice technology for NOx control on FCCUs. (Howard Lange, AQMD)

Discussion: A committee member noted that the ammonia emissions measured in the RATA tests were much higher ppm levels then are allowed for SCR systems as applied to power plants.  AQMD responded that proposed Rule 1105.1 will limit emissions of ammonia as well as PM10 from FCCUs.  Another committee member commented that the success of the SCR process when applied to FCCUs may depend to some extent on the crude stock being refined, due to the differing metals contents of various crudes and the possibility that some metals contained in crudes may affect the performance of the SCR catalyst.  AQMD responded that two other refineries in the SCAQMD region are planning to install SCR systems on FCCUs, which indicates applicability to a broader range of crudes.  However, it was decided that staff should investigate the range of crudes refined in the SCAQMD region as compared to the full range of crudes refined in the U.S. (Martin Ledwitz, Southern California Edison Co.; Gary Rubenstein, Sierra Research; Marty Kay, AQMD; Howard Lange, AQMD)

Tunnel Washer, Textek (A/N 392961)

Textek provides a shop towel cleaning service to industries such as printers, furniture finishers, etc.  This facility extracts VOC from dirty shop towels using a ten-chamber tunnel washer, which is essentially ten washing machines arranged in series.  Tunnel washers are normally used for washing, but this facility has adapted the tunnel washer to serve as a VOC extraction system.  The dirty towels enter the first chamber on a conveyor belt, which is located in an enclosed loading area, and move through the tunnel washer chamber-to-chamber.  The first six or seven chambers operate mainly on steam, and the last three or four chambers operate on hot water.  The devolatilized towels go through a spin dryer and are then sent elsewhere for final washing.  The air pollution control system consists of a condenser followed by a thermal oxidizer.  The tunnel washer chambers are all vented to oxidizer via the condenser, and the loading area enclosure is vented directly to the oxidizer.  The permit requires 90% overall VOC control by the VOC collection system and oxidizer.  The system was started up in May 2002, and a source test in February 2003 showed 99.3% overall VOC control.  VOC control has therefore been achieved in practice on a tunnel washer used for VOC extraction. (Howard Lange, AQMD)

Discussion:  A committee member was concerned about the potential for VOC to escape the process in the condensate.  AQMD acknowledged the potential for this to occur but noted that the source test had shown the condensate leaving the process to be very low in VOC.  Another committee member asked whether this was a unique process that may not be generally encountered.  AQMD responded that it was the judgment of the permitting team that this would be a useful listing, and another committee member commented that this is probably not a unique process.  Another committee member commented that the oxidizer appears to be oversized and thermally inefficient.  AQMD agreed that this does appear to be the case and noted that the oxidizer had been purchased used. (Katy Wolf, IRTA; Hal Taback, Hal Taback Consulting; Anoosheh Mostafaei, Ship and Shore Environmenta;, Howard Lange, AQMD; Marty Kay, AQMD)

Boiler, Fullerton College (A/N 413617)

Fullerton College has installed two new boilers rated at 10 MMBtu/hr each and has been operating them for several months.  These boilers, which are used for water heating, utilize a new type of ultra low-NOx burner to meet the existing NOx BACT limit of 12 ppmvd@3%O2.  This burner, offered by the Johnston Burner Co., employs a ceramic cylinder extending into the furnace with the air and fuel being injected as high velocity flame jets surrounding the ceramic cylinder.  The jet action surrounding the cylinder induces recirculation of furnace gas from the end of the cylinder to the base of the flames via openings in the cylinder at its base.  This internal flue gas recirculation (FGR) reduces NOx in a manner similar to the more familiar external FGR systems, lowering NOx formation by cooling the flame.  Attractive features of this ultra low-NOx burner are that it operates at normal excess air levels and requires only normal burner maintenance, i.e., no special attention required.  The burner was developed by the Gas Technology Institute with support from the Southern California Gas Company.  Johnston Burner Co. has an exclusive license to market this technology for firetube boilers.  The burner has been functioning well in both boilers, and the boilers were source tested with NOx levels in the 6-9 ppmvd range and CO levels in the sub-25 ppmvd range (both corrected to 3% O2). (Howard Lange, AQMD)

Discussion:  An audience member asked why the CO results in the source test were reported as <22 ppmvd on one boiler and <23 ppmvd on the other boiler (both corrected to 3% O2).  AQMD explained that the CO measurements were done on a 100 ppm instrument range, and AQMD does not recognize values below 20% of instrument range.  The CO values as measured were  less than 20 ppmvd, so AQMD therefore considered them to be reportable as <20 ppmvd on both boilers.  These <20 ppmvd values are as-measured, i.e., at the as-found flue gas O2 levels, and correspond to <22 and <23 ppmvd@3%O2, when corrected to standard dilution. (Damon Erickson, Muira Boiler; Marty Kay, AQMD)

I.C. Engine, Stationary, Tidelands Oil (A/N 405789)

Tidelands Oil will use this engine to drive a water pump at a petroleum production field.  The engine is fueled on well gas, which is a waste gas occurring at the site.  This is simply an example of a stationary engine being required to meet existing BACT guidelines, but is noteworthy in view of the off-spec fuel being used.  The usual engine NOx/CO/VOC control technology, consisting of a rich-burn engine with a three-way catalyst and air/fuel ratio controller, will be used.  The well gas contains nominally 1500 ppm sulfur, and existing BACT guidelines require that the fuel sulfur content be reduced to 40 ppm or less.  This will be achieved using sulfur absorption vessels manufactured by Sulfatreat™ (two in series).  The Sulfatreat system is part of the oil/gas/water separation system, which is a normal component of a petroleum production operation. (Howard Lange, AQMD)

Discussion:  A committee member asked what is the fate of the sulfur removed in the Sulfatreat system.  AQMD responded that the Sulfatreat literature indicates that the sulfur is converted to a chemically inert form.  Another committee member suggested that the information regarding the Sulfatreat system be moved from the Comments section of the listing to the technology description section.  AQMD agreed to do so.  There was some discussion of the pressure level of the well gas in the Sulfatreat system. (Hal Taback, Hal Taback Consulting; Gary Rubenstein, Sierra Research; Greg Adams, Los Angeles County Sanitation Districts; Howard Lange, AQMD; Marty Kay, AQMD)

New BACT Part B, Section II Listing

I.C. Engine, Emergency, Compression-Ignition, Kings County Dept. of Public Works (C-1010958)

This 2000 hp diesel engine was originally installed to drive an emergency generator but was later “promoted” to “limited use” status, a permit category that allows a limited amount (614 hours per year, in this case) of operation beyond emergency and testing.  When the engine was promoted, it became subject to a more stringent BACT guideline, which required installation of a particulate filter and use of ultra low-sulfur diesel (ULSD) fuel.  The particulate filter installed in this case is similar to that installed on the Claremont Manor emergency diesel engine that was discussed previously in the meeting, although supplied by a different manufacturer.  The filter manufacturer decided to use this installation to obtain a CARB verification of its filter technology, and the engine was therefore subjected to a number of source tests.  Testing was done at three loads—50, 75 and 100 percent of rated load—before filter installation, after filter installation and again after six months of filter service.  The test data showed removal efficiencies of >75% for PM, >97% for CO and >93% for VOC.

CARB issued a verification letter verifying >85% PM removal on the basis of a weighted-average of the three loads tested.  CARB’s verification letter also stipulated that ULSD must be used and that the verification is restricted to engines that produce <0.1 g/hp-hr PM. (Howard Lange, AQMD)

Discussion: A committee member asked how the filter is cleaned.  AQMD responded that CARB’s verification letter requires that the engine operate at 40% load or above (so that the filter will be heated to a sufficient temperature for self-cleaning to occur) or that the filter be cleaned after every four hours engine operation.  The recommended cleaning method is to run the engine at or above 40% load for at least two hours.  However, since running the engine this much produces a substantial amount of pollution, a preferable method would be to have the filter kilned off site, a service that the filter manufacturer offers. (Hal Taback, Hal Taback Consulting; Howard Lange, AQMD)

A committee member noted that the test data indicate that the filter causes an increase in NOx.  AQMD responded that this had been discussed with CARB, and CARB does not consider the apparent increase in NOx to be proven in view of the limited amount of testing and the long time frame over which the tests were conducted.  Another committee member pointed out that the test results as shown in the listing indicated >100% removals for some PM components.  AQMD agreed to look at the original data and determine whether these entries are correct.  A committee member noted that the manufacturer’s guarantee of 98% VOC removal was not being met at the times of the source tests.  AQMD agreed to inquire about the VOC removal guarantee now being offered.  A committee member noted that the filter manufacturer had guaranteed 85% PM removal and that the filter did not meet this at full load after six months.  AQMD explained that the CARB verification, and presumably the manufacturer’s guarantee, are based on weighted averages of results at three test loads.  The CARB representative on the committee offered to find out what weighting factors are used and report back at the next meeting.  The committee member who had asked the question stated that the fact that the 85% PM removal verification and guarantee is based on a weighted average should be noted in the listing.  AQMD agreed to do so.  AQMD noted that the listing recommends that filter cleaning be done by means other than engine operation in view of the large amount of NOx that would be produced by the engine during filter cleaning relative to the amount of PM being controlled by the filter. (Martin Ledwitz, Southern California Edison Co.; Greg Adams, Los Angeles County Sanitation Districts; Gary Rubenstein, Sierra Research; Todd Wong, CARB; Howard Lange, AQMD; Marty Kay, AQMD)

 Other Business

AQMD stated that the next meeting is scheduled to take place at 1:00 p.m. September 18 in Room CC-2.

There was no further discussion, and the meeting was closed.

 Attachments