Document vVbQv98EXkeq1z3RqN6bKzQQE

HONS 003*6* o Detection end destruction Amen| the molt urgent research re quirement* of die utility industry is a quick, inexpensive way of identifying PCB-contamlnated transformers, which, like PCB equipment, are subfect to dis posal regulations. Hlngorani suggests that perhaps 5% (1.9 million) of distribu tion transformers and about 20% (1 mil lion) of power transformers filled with mineral oil may be contaminated with PCBs in the 90-500 ppm range. But EPA considers all mineral oil transformers to be PCB-contaminated <90-900 ppm) un less tested or otherwise proved to be nonPCB equipment Commercial gas chromatography can identify PCB levels In transformer oil, but the process is inconvenient time-con suming, and expensive. Samples of trans former oil must be drawn in the Acid, then sent to a laboratory for examination by skilled technicians. It costs $50-9100 to test each sample, and more than one ample may have to be tested to make a positive Identification. To test every transformer suspected of PCB contami nation Bright coet the Industry as much as 92-19 billion, estimates Hlngorani. A portabla inspection device that could quickly and inexpensively Identify PCBs in transformer oil might reduce the cost of detection to as low as $900 million. EFRl's ES Division is now developing such a device. Project Manager Vaau Tahillani explains that under EPRI con tract, Central Electric Co. scouted other Industries for applicable devices and found that tha oil Industry uses an X-ray fluorescence Instrument bom Horiba Instruments, Inc., to meamre undesirable sulfur content in off. CMorins atoms In PCB tetpond ihntlxrty to X rays, ao on Central Electric's Toccmnwndittona and specifications, Horiba modified the de vice to measure chlorine to the required sensitivity. General Electric is nowusing the device to Hot tfioussnds of samples of PCBcontaminated oil, tvahisting how well Its chlorine analysis correlates with PCB analysis by standard gas chromatog raphy. The device, which consist* of a portable analyzer and a portable data processor, will soon be taken to the Salt River Project's service area for field test ing and evaluation. Because the new device measures chlorine rather than PCBs directly, some transformer oil samples may require re testing in laboratories by conventional gas chromatography. EPRI ultimately7 plans to develop, a more advanced device that measures PCBs explicitly; the E5 Division Is now exploring possible re search in that area. Once utilities have Identified PCBcontaminated transformers, the contami nated oil could certainly be drained and destroyed and the transformers refilled with fresh insulating fluid. But trans former-quality naphthenic oil may not be available in sufficiently large quantities to replace all the contaminated oil at once, according to Project Manager Gilbert Addis. And because stringent requirements apply to the incineration or tha landfill disposal of PCB-eontarui nated oil, large-scale incineration or landfill disposal program# may ba im practical Instead, utilities might reduce the level of PCBs in transformer oil to below the regulated limit. This could be accomplished by either destroying the PCBs in the oil in situ or extracting them from' the oil for later destruction. The purified oil might then be reused in transformers, assuming no oil degrada tionor further contamination takes place. Commercial methodsof reducingPCBs to acceptable concentration* are not readily available, however. PCB com pounds are both remarkably stable and similar In physical properties to the trans former oil with which they may be mixed. As a result, they resist both destruction andextraction. Only recentlyhave several processes emerged that an potentially suitable for destruction or extraction. The E5 Division has singled out several of the moat promising technique* and b assessing them to see if any art appro priate for the utility industry. Addis notes that the largest EPRI contract for PCB destruction and extraction is with Genera Electric, which researching two pro cesses for PCB destruction and two tor PCB extraction. Based on transportability operability, downtime requirements, am other considerations. General Elcctrii will select the best of the four process* for scale-up to a demonstration project a a later date. The first PCB destruction process em ploys electron-beam irradiation to dr compose PCBs by breaking off chforisu atoms from the compound. The secom process uses a sodium reagent for dr chlorination. One extraction process la liquid-liquid extraction; the other proem is a supercritical fluid-liquid extraction Yet a third extraction process is being to searched by Franklin Research Center in this process, the PCB in the extractor solvent would be destroyed or remove* by a second solvent. Once PCBs are destroyed or cxtractev from transformer oil, there is no guaran tee that the treated oil will be suitabl* for reuse in a transformer. The sodhw destruction process, for example, mrgh leave behind by-products that could tern to transformer failure. Similarly, extras tom processes might leave behind harm hri traces of extractant or remove chemt cals essential to rite oil's dielectrl qualities. General Electric wilt study th. long-term effects of these processes to bi certain that PCB destruction or extractiot does not leave the oil unsuitable fo retom to a traniformer. The four processes But General Elec trie Is studying are aU performed whili transformers are shut down. But down time is inconvenient for utilities. RTl Corp. has developed a process that can be used while a transformer is in open tion. The process reportedly require? several months to reduce the PCBs in . transformer by a significant amount Through an ES project, RTE researcher are checking to see if this procedure cat be accelerated. The project will also do terorine the rale at which PCB* concen trated in the transformer core will lead* out to recontanrinate the cleaned oil. 22 IP* JOURNAL October 1991 HONS 003466 MONS 003467 A ttmiiur to r*vw detection and de struction techniques for PCBs in utility equipment is btin| planned by CPR1 lot this December; Addis is seminar coordinator. Capacitors are another area of the ES Division's PCB research- Most capacitors manufactured before 1077 are known to contain PCBs over 900 ppm. Researchers are trying to find a way to anticipate failure in PCB-filled capacitors because when these capacitors do fail, they some times rupture and spill PCBs over the im mediate ares. Cleanup costs for a capac itor spill are high and may result in an annual cost to the industry exceeding several million dollars. Yet because capac itor manufacturing capability is limited, at It FCB-capadtor storage and disposal, utilities cannot replace all PCB capacitors overnight with non-PCB capacitors. As a possible alternative, EPR1 and Westinghouse Electric Corp. are investi gating the feasibility of several devices that Identify capacitors with fendpient faults. Field and factory data indicate that weeks to months may go by before a par tial electrical discharge remits in an in ternal failure Urge enough to rupture a capacitor. If a utility had a device that could alert It to partial discharges, units with Incipient fault* could be replaced prior to rapture, and the number of capacitor spills could be minimised. One device being considered is an acoustic detector that senate partial dis charges. A partial electrical discharge In a capacitor produces gas that in turn can lead to capacitor rupture. Since par tial discharges typically emit ultrasonic sound, acoustic datoclors far partial dis charges could alert utilttas to capacitors that arc about to fail. Another device is an Infrared scanner equipped with an optical display. Because a capacitor approaching failure would probably be at a higher temperature than Its neighboring capacitors, the infrared scanner could pick it out according to Project Manager Robert Tackaberry. Once Westinghouse has completed its review of these and other fault-detection devices, (he most successful of them will be installed in van for field-testing on more than 1000 capacitors at a cooperat ing utility. The van and contents are expected to be ready early next year. SpM cleanup and disposal PCB spill cleanup and disposal arc other areas where utilities would be greatly aided by additional research. Although EPA regulations generally permit con tained PCB or PCB-contaminated equip ment to remain in service, sooner or later utilities with such equipment must handle those PCBs. Assuming no spill occurs during operation or maintenance, the equipment eventually wears out and has to be retired and disposed of. PCB spill cleanup and PCB disposal can be complicated and costly for utilities, according to Program Manager lohn Maulbetsch. EPA regulations require, for example, that PCBs spilled on earth, con crete, or other surfaces must be reduced to concentrations of leas than SO ppm. PCB liquids and capacitor wastes of over 900 ppm must be burned in EPA-approved high-temperature incinerators. Barsuae the complete sweep of PCB regulation can be both confuting and costly, EPRI's CCS Division (with sup port from the EAE Division) has devel oped a series of handbooks to provide Information about cleanup and disposal options. The first handbook, Ditpotol of ftbdkrimhf Sipfofjrb sad KB*Co*t*miaatsd Materials (prepared by Steams, Conrad, and Schmidt Consulting Engi neers, Inc.), gives utility engineers gantral information on PCB production and use, details on present and proposed regula tion, projected requirements for disposal capodly, and an overview of available incineration and landfill technology. The second and third handbooks pro vide guideline! on how to develop spillpcevention techniques and countermea sure control plans to ensure that risks associated with PCB activities are mini mal Induded are model operation plans that address assembly and servicing of PCB-filled components, proper use and maintenance of these components, and containment procedures for preventing ind confining accidental releases. AH handbook, an available from Unearth Report* Center; according to Profit, Manager Dean Golden. Bnidn developing guidn to pdl cleanup and disposal, the CCS Diviaion has bran working on cleanup and dfcpoeal technique!. One kay protect ia the development of an bnmirnem capable of measuring Mil PCB level, jn the Held. At thi, time, according to Protect Mar-- ager Ralph KomaL no Md-usabic internmenu are available that can do Ihto job. When a spill ocean. Mil in the immedi ate area to removed. Commonly, tamples of the remaining Mil are taken to a laboratory for analysis to ice if the rpdl baa been cleaned to legal limit,. Typically much more Mil la removed then would be nccewary if a BeM-ueaMe detection method were available. PCB detection coat, vary, depending on spill ,iae and ament, but repment a significant annual expenditure to the industry. A portable instrument that could perform inexpensive. on-the-spet analyte, would greatly simplify rpdl cleanup. This type of instrument mute be entetive enough to detect the many reg ulated PCB compound,, yet not ro sentetivc that other wbteancea in the toil can obscure the analytes. EFRI to developing luch a device with Oak Ridge National Laboratory. One approach ia to extract PCBe from toil sample, and analyte thorn on location. The other approach to to vaporise oil from sod sample, in a port able chamber, than iontee the vapors with an uHravtolat lamp; the level of ioniaMioat indicates the amount of PCB prewnr The ORNL protect to expected to be com pleted later dale year, and the multant prototype wtH be made available to interetted uMUtiee for fiold-letetotg. Other ImportantCCS Division woeh hr rpill technology todevelopment ofa treat ment that wHI quickly decompos, or re move PCB ipill istedoaaon *uch surface, as soil or cemant. Such a treatment would eliminate the necessity of having to ex- 24 EPmjourtNAi. oeioo--teat MQNS 003frhB UTILITY ROLI IN PC* MANAQEMKNT From 1929 to 1977. 1.4 billion lb (635,026 Mg) ol PCBs were manufactured in me United States and routinely used for appli cations that included electrical equip ment, plastics, adhesives, lubricants, carbonless copying paper, and inks. A small percentage of that amount was eventually destroyed; close to hall of it vms buried In landfills, escaped into the environment, or exported and is not ad dressed by ERA regulations. But over half of it--some 750 million lb (340,193 Mg)-- is still in service, practically all in sealed electrical equipment. Electric utilities have a large part to play in the manage ment of these PCBs. According to preliminary estimates made by Narain Hingorani, chairman of EPRI's interrtivision R3B working group, perhaps 25% of the PCBs in electrical equipment are in small capacitors in fluo rescent lights, motors, and appliances. Somewhat less than 40% are in large ca pacitors, moot of which are in utility and industrial Jurisdiction and can be re trieved; about 40% are in PCS transform ers, most of which are also in utility or industrial jurisdiction and can be re trieved. And about 0.05% is believed to be In PCB-contaminated transformers, which are essentially titled with trans former oil. Ironically; this 0.05% in con taminated transformers is mixed with bHone of pounds of valuable transformer oil: finding and retrieving it may cost utili ties far more than the retrieval of the PCB capacitors and transformers that contain higher concentrations of the compound. / m cavete contaminated soil or debris and later dispose of it at approved landfills. Through newly negotiated contract, Batielle, Pacific Northwest Laboratories is evaluating proprietary chemical treat ments that promote accelerated photodecompotition of PCBs. The CCS Division it planning to explore other methods as well, including application of a poultice* like substance to PCB-conuminated areas at spill sites. The substance would dis solve PCB residues in a volatile solvent, then adsorb them as the solvent vaporiocs. The substance and the entrapped PCBs could then be removed, possibly in the form of a dry powder. Disposal options are another area on the CCS Division's PCB research list. These options are limited at this time. EPRI sponsored a test bum at die Energy Systems Co. (Ensco) incinerator in El Dorado. Arkansas, to see if this facility could incinerate PCB wastes in accor dance with EPA regulations. The bum, which took place In 1979, included liquid PCBs drained from transformers, ss well as shredded capedtore. (Because most of the PCBs In capacitors are absorbed into the capacitor's paper insulation, capaci tors cannot be easily drained and so are shredded and fed into incinerators.) The test showed incineration at the Ensco facility was highly efficient: samples from the Incinerator's stack, scrubber, and ash were analysed for PCBs. Only in the ash were PCBs confirmed and those were at levels below regulation. The Ensco facil ity subsequently received EPA approval for PCB incineration, makfcg It the sec ond of two facilities in the nation to be approved for liquid PCB incineration and thaonly facility to he approved for solids shredding and birineralton. The coat of deatmetkm of PCB liquids ranges front 9300 to M30 per 99-gal drum. Meanwhile, the CC5 Division is in vestigating alternative ways to dispose of solid PCB materials, such as capacitors, to expand the range of utility options. Rather than shred and incinerate this equipment, chemical processes that ex tract the PCBs for later incineration might be developed, and the division is plan ning to fund research in that area later this year. Hoottti effects Beyond PCB detection, destruction, spill cleanup, and disposal, still another area of PCB research Is basic to the utility industry: health effects, particularly occu pational health effects. The utility per sonnel who service transformers, dean up capacitor spills, and ultimately re tire this equipment from service must be assured that work practices protect them from any potential risks from PCBs. Utilities must also be confident that cur rent practices protect the general public. However, os with most low-level ex posures to contaminants, the PCB risk to humans has not yet been exhaustively Investigated. Explains Walter Weysen, EAE project manager, 'The immediate toxic health effects of undiluted PCBs in animals have been well documented. Similar health affects in humans were observsd following ingestion of rela tively large amounts of PCB-conttminated rice olTlh japan. The results of animal studies indicate that pure PCBs have a low toxicity but some PCB con taminants or reaction products are potent toxicants. "However, the passible long-term health consequences of PCB exposure have been lets well documented. Results of some animal studies appear to suggest that PCBs at high levels of exposure are either carcinogenic or promote the cardnogenicity of other chemicals. The important question for utilities, their em ployees, and the public Is whether lowlevel exposures pose risks to humans. Epidemiologic studies of humans, acci dentally or occupationally exposed, have not demonstrated PCB-usodatad cancer; but many more years of observation are desirable to confirm such findings," con cludes Weysen. Epidemiologic studies require detailed information on well-defined end statisti cally significant population samples over long periods of time. But retrospective studies are difficult to accomplish be cause exposure and health records of qualtty riuy not irt. !f, risk from vmy lw hl,of TCB *pmm exists, it will take records on a very las^ number of potentially exposed personal* conclusively demonstrate the pnasnes m absence of health rieka. EPRI'* EAE Division, with the coopssa- tion of six utilities, is new attempting la set if a study of PCB-exposed utility em ployees is feeefole. If the study rtventa that sufficient information is available; a study of workms exposed in the pnat might be undertaken by EPRI, eccnrdfog to Weysen. Meanwhile, foe EAE Division is plan ning a PCB weskahop for utility health and safety personnel The workshop, ten tatively scheduled for this fall, wouic present the iota* information on the po tential health risks of occupational PCI exposure. Industrial hygiene, work prac tices, and legal and regulatory issues wti. also be disemeed. Workshop attendees con use tttia Information in their owe utility information programs. Weysen is workshop coordbialcir The utility Induatr/ must moke sure that PCB subUtutss see environmental!} acceptable. The EAE Division is planning a two-year peofnet that will provide bank health end environmental information or PCB substitute* used ki utility equip ment This study is expected to begir in 1993. Without a doubt die utility induetr/t continued fitaMpmint of PCBs will re quire time, effort,and money. But through coordinated research,those inevitableex penditures can be significantly reduced The trilateral effort of EPRI's three divi sions on the PCB problem should yiete results that wiB make the industry's ef fort easier. 9 29 EPRI JOURNAL October IBM HONS 003470