Document LDmo09knDaOLBKXVK1RpBGx7

G E N E R A L A E L E C T S 1C OCNCIIAL ELECTRIC COMPANY, 1 RIVER ROAD, SCHENECTADY. NEW YORK 13948 PfMMM (618) 374-3211 Subject: Effluent Limitations for Items oa Toxic Substances List: Economic Impact of s Baa on PCB PRODUCT QUALITY STAFF November 21, 1973 Dr* Martha Sigir Chairman, Effluent Stsadaxds and Water Quality Information Advisory Cocol ttee Environmental Protection Ageney. Boos 821, Crystal Hall, Bldg* #2 Washington, 0 C. 20460 Dear Dr. Sager: With reference Co your forthcoming meeting (38FS30577) and pursuant to our understanding that the Committee has expressed an interest in knowing the economic lmpaet of a possible ban of polychlorinated biphenyls (or a restriction tantamount to a ban)., we have the^following comments to offer' The value of the annual production of PCB transfom e r a in the U.S. is approximately $45 million end that of capacitors $140 million. Neither of these figures forms an appropriate base for assessing the socio-economic effect of regulations having the force of a ban on further production. Should sueh a ban occur in the near future (i.e., within five-years) the impact of PCB transformer unavailability would be felt primarily in terms of disruption of building construction, electric distri bution, and railroad equipment projects; and interruption of the supply of PCB capacitors would impact most heavily on the electrical supply situation, exacerbating the "energy crisis." Hose PCB transformers are installed in industrial, commercial, and residential substation applications A considerable number are Installed on electric railway passenger commuter cars serving major metropolitan areas. In all applications, these tranformen distribute electrie power reliably without risk of fire and explosion, and In compliance with specific laws, codes, and Insurance restrictions mandating non-flammable liquid content. There is no readily available alternative for many applications where construction *ls under way oi plans ere complete. As much at $1 billion of construction projects could bo seriously delayed. Involving schools, factories, office buildings, apartment houses, and connuter trains. 774043 IDS Q023u 6NflAl0 UICTIIC Dr. Martha S|ir 2 Tha PCS capacitor* produced la tba U.S. annually hava a major affect la raduelng the electric currant which would otherwise hava to ha generated to carve Induetrial and public needs. The saving may be approximated by sunolag up the incremental coat of oversize generators and that portion of transmission and distribution system equipment which la thermally limited* Equipment in these classes, corresponding to the 44 million kilovars of new capacitors Installed annually, has a value of approximately $ 1 billion* A forced conversion to a non-PCB capacitor would be hampered by shortages of basle capaci tor materials and factory production capability, thus placing la Jeopardy about half of this annual saving* Additionally, thsra would be serious dislocations in the lighting end air-conditioning Industries due to shortages and mismatches of eapaeltora raquirad la thoaa systems* \ the enclosed paper, ``The Role of Polychlorinated Biphenyls In Electrical Equipment", February 4, 1972, was prepared originally for tha Offlea of Science and Technology and la lneludad hera for furthar information on the aconoolc valua of these devices. In addition, we have updated the information about capacitors and addressed ourselves specifically to the consequences of unavailability of PCB capacitors in tha enclosed new study entitled, `Tha Impact of a 'Ban' on tha Use of PCS in Capacitors, ?'November 19, 1973* Important as are economic considerations, vs feel that tha ease for continued use of PCB's in capacitors and transformers rests less on these factors than upon the lack of substantiated need for any tighter constraints on ehe handling of these materials than those already in use or assumed as goals by the electrical industry. These aspects are discussed in ehe enclosed letter from Sr. Simona of General Electric to Dr. Thompson of E.P.A., cementing upon toxic pollutant affluent standards (reference, 38FR21342). Vs hope this discussion will be of use to your coianlttee. Very truly yours. JSN:ev Enclosures ce: Dr. C. Hugh Thompson Chairman-Hazardous and Toxic Substance Regulation Task Force Office of Water Program Operations Environmental Protection Agency. Washington, D. C. 20460 774044 IDS 002 TBS IMPACT OP A "BAN" ON TBS PSE OF PCS IN CAPACITORS SDMMAET Tha FCB-impregnated capacitors produced la tha U.S. annually bava a major effect la reducing tha alaetrle currant which Bust ba generated to serve industrial and public needs for power. The annual saving Bay be viewed as equivalent to 44 large aleetric generators, each of the 1000 aegavatt size, wbleh do not need to be installed* aa aggregation comparable to the nation's annual growth requirement. The following consequences will result from a FCB effluent standard so strict as to have tha effeet of a bans If tha industry is forced to convert to a non-PCB lspregnant, lew dielactrie-conateat liquids, such as nineral oil. offer the best initial fallback position. - Mineral oil would roughly double the pounds of capacitor grade dielectric paper and aluminum foil needed to produce present eapacitor requirements. Current world-wide shortages of eopaeltor paper and foil would not support this. - Materials.availability would support about 50T of .the current capacitor, requirements using mineral oil. - - If adequate materials supplies were somehow attained, current capaeitor manufacturing facilities could produce at best only 50X of cspaeltor unit requirements. - Forced conversion to a non-PCB would result in capacitor shortages and major production cutbacks in the lighting, air conditioner, and power supply Industries, and in their customer operations. - Doubling the sise of capacitors would present serious replacement problems where PCB capacitors had been used. - Ballast and air conditioner manufacturers would have to redesign certain products to accept the larger capacitors.. - Power system stability would be adversely affected by a shortage of cepecitors, and blaek-outs could be expeeted. - New non-PCB*s, such es G.E.'s Eenol, could subtantially reduce the magnitude of the materials problem. However, Econol still ' represents an unknown risk in reliability and safety in the major consumer application areas. A mass conversion is not Justified at this tlma. 774045 IDS 0 U E N P 005394 2 I. INTRODUCTION This Is a forecast of eh eonsaquancas of PCB fflune standard which cannot bo mat by eapaeltor manuiscturar* , and consequently become* an affective ban. It la submitted by the General Eleetric Company to assist In the determination of appropriate effluent standard* relative to eapaelcor grade PCB* 12* BACKGROUND Xn the tf.S*v an Impregnated eapaeltor ean be very simply described as: J Aluminum foil (around 0*22 mil thick) which serves as the electrodes. - Sheets of dielectric consisting of capacitor grade .paper or combinations of this paper with plastic film* The dielectric is positioned between the alisalnum electrodes as it is wound into rolls* An appropriate liquid Impregnant. ' - *A sealed container, for the capacitor roll and Impregnant* The choice of dleleetrle materials, the number of sheets, and their thicknesses are all the result of many years of research and testing for reliability and conauser use safety. About 90-95X of liquid impregnated capacitors contain PCB, and are critical components to the fluorescent and mercury vapor lighting, the electric motor and^ air conditioning industries, and power transmission by alcetrie utilities* The dielectric paper and plastle films (primarily polypropylene), aluminum foil, and PCB have all been specially developed and controlled for eapaeltor use and are produced as "specialty" materials, Capaeltor grade PCB la A r o d o r lOlfi, manufactured only by Monsanto, It is primarily triehloroblphenyl. It differs from all previously used formulations in that practically all of the higher boiling homologs-- which have given rise to reports of PCB in wild-life residues-- have been removed. Introduced in 1971, it la now the only PCB used by the U.S. capacitor Industry* Compared to ocher formulations, - it has a higher degree of biodegradability; - lc has a lover degree of persistence in the environment. tiao*6 IDS G023< U h N P 00539 -A XXI. COHSEOPEHCES Of A "BAH" TO THE CAPACITOR INDUSTRY This svmluselon addresses itself to tht situation wherein ths capacitor industry is affectively banned from using PCS'a. A. Alternative lmpregnant Alternative Iapregnants are unproved in the U.S. capacitor industry. Mineral oil vaa Che major Impregnant before PCX's, and It appears to represent the best Isnediste fallback position. Even though mineral oil technology would put the industry back 30*40 years in progress, it still represents the best combination of known technology, materials availability, and reliability. For these reasons, mineral oil lmpregnant has bean selected at this time as the basis of comparison with PCB's. There have been a number of very active programs to develop non-?C3 iapregnants, especially after the environmental problem was recognised. These progrsms are continuing in ths industry, and it la expected that an improvement will be found which would be more acceptable than mineral oil. As an example, General Electric is using a phehalats sstsr as the base for its non-PCB Econo1 lmpregnant. Econo1 capacitors are being sold eonsserclally where special situations caquirs a non-PCB. This is ths case where U.S. elactrleal equipment, such as computers, is to be hipped to Japan where PCS'* era. banned. EcSiol has a higher dielectric constant (sea next Section B for discussion of this property) than mineral oil, and consequently will require less additional materials and manufacturing capacity. However, limited availability of the phthalace ester, lack of appropriate manu facturing facilities, need for developing new efficient msnufseturlng processes, and particularly the unknown factors of-reliability and safety in the broad consumer use areaa such as lighting and air conditioners are soma of tha problems that prevent a near-tans substitution. Unless eenslderabla time for tha development and avaluatlon of alternecl liquids is permitted, the most suitable iemediate stratagem appears to be tha use of mineral oil. Tha virtues of PCS lmpregnant and tha disadvantages of mineral oil have bean discussed in detail in a number of recent documents. Among these documents are the report of-the Interdepartmental Task Force on PCB's (COM-72-10419), a paper titled, '`The Role of Polychlorinated Biphenyls in Electrical Equipment" presented by General Electrie on February 4, 1972, and NEMA's Official Standards Proposal CP-P1-1973/TR-P6-1973, dated January 25, 1973. All of these documents heve pointed out that PCS-impregnated capacitors are significantly smaller, sure reliable and safer than mineral oil impregnated capacitors and that use of mineral oil would return IDS p 002366 C 774047 / r r A A i k Tnr Ow 4 capacitor technology to Iti pro-1932 level* The substitution of mineral oil for ?C8 impregnante vould necessitate redesign of oad equipment to utilize ebo largor capacitor, largo lacraos In capacitor manufacturing facilities (both plant and equipment) and would result in increased use of basic matarais which are presently in short supply* The following is an attempt to quantify some of the consequences related to the use of mineral oil as an alternative lmprsgnant* 1 Capacitor Manufacturing Capacity Mineral oil- Impregnated capacitor sizes relative to PCS capacitors are shorn in Table 1. Table 1 - Relative Size Comparison Types* AROCLOR 1016 Mineral Oil Small Industrial 1.0 2-2*5 Power 1*0 2*5-3* 2 Small Industrial Used in lighting, air conditioning, motors and .electronics* Power - Large, capacitors ussd by electric utilities' and by industry for induction-furnaces and power feecor correction* The dielectric conjtant of s capacitor's dielectric system is the ratio of its ability to stors electrostatic energy relative to air. The dielectric constant of PCS lmpregnant is 5.85 while that of mineral oil is 2.25. Use of the lovsr dielectric constant mineral oil in s paper dleleetrle system, assisting no change in the thickness of the paper dielectric, would result in a capacitor which la 60T larger than the PCB impregnated design* Additionally, however, use of mineral oil requires that the thickness of the paper dielectric be increased to reduce the operating stress on the system if equal life and reliability are to be attained. In Che esse of lower voltage industrial type capacitors, such as those used in fluorescent ballast end sir conditioning applications, the dleleetrle thickness must.be increased 10 to 25Z. In power capacitors operating At higher voltages, the dleleetrle thickness must be increased 25 t 40Z. The larger Increase in the high voltage units is related to the face that corona initiation and extinction voltages of mineral oil impregnate dielectric systems are lover than those of PCB impregnated systems. The approximate doubling of physical size results in about half the throughput of capacitors in existing manufacturing equipment. The eapacieo j- industry currently has little if any excess capacity to handle this, even i - adequate capacitor materials supplies are assumed to be available. i 774048 C t. IDS 002367 v 5- CTtSEQUENCE - THE CAPACITY AND OUTPUT OF THE CAPACITOR INDUSTRY WILL RE CUT APPROXIMATELY IN HALF. C Capacitor Material Availability Tba following Tabla 2 sstimates tha capacitor industry requirenanta or papar, foil, and impragnant based on tho currently used AROCLOR 1016 ad for mineral oil* Tabla 2 - 1976 Capacitor Matarlala Raoulregents fibs, in million) Paper A1 Foil lmpregnant AROCLOR 1016 26 16 26 Mineral Oil 52-65 32-36 35-42 X Chance 100-15 O X so-ioox' 35-601 There la a worldvide shortage of capacitor grade naocr and aluminum foil. Tha eapaeltor paper supply is expected to continue at Its present level of availability. Mineral oil Impregnation vlll require at least twice the amount- of materials toprbduce the same capacitance as with PCS'a. Using the most optimistic of the above numbers, the total capacitor industry's requirements of Kraft paper dielectric, all of it using Imported pulp for its production, would increase by 2 to 2.5 times and its requirements for aluminum electrode foil would increase by 1.80 to 2 times. Total industry requirements in 1974 for paper dielectric in PCB Impregnated capacitors is estimated at 26 million pounds. Based on recent experience, this requirement is the limit that suppliers of this type of paper can produee with existing equipment. Adding facilities to more than double tha present domestic output of paper is a task that could not bo completed in less than A to 5 years and the total Investment required by the paper manufacturers is estimated to exceed $200 million. It la questionable that tha paper industry would make this Investment, facin tha probability that future lmpregnant and design developments would negate its ased. Aluminum foil is also in short supply. Some investment.in foil-rolling equipment Is already under way at General Electric to alleviate the shortage but an adequate continuing supply of reroll aluminum scoek from which the th; gauges of aluminum electrode material are made cannot presently be assured The supply of dielectric-grade mineral oil required to replace FC8 Impregnant* is also unassured. Doubling the size of capacitors would /-rv/~v 774049 IQS 0023 t. v < approximately double the voluaa of impregnant required. The eole manufaetui of dielectrle-grade mineral oil hee recently announced thee its manufacture la being discontinued. The feasibility of supplying the capacitor Industry on a timely basis with 35 million pounds annually,' starting from a prodcele rate of aero, is doubtful to say the least. CONSEQUENCE - THE CAPACITOR INDUSTRY WOULD BE ABLE TO SUPPLY ONLY ABOUT 501 OP THE CURRENT CAPACITOR DEMAND BECAUSE OF MATERIAL SHORTAGES, D. Costs U.S. Impregnated capacitor tales for 1974 are shown la Tabla 3. Table 3 - Estimated 1974 Total Impregnated Capacitor Salas $ Millions Small Industrial 100 Power 55 About 90-95Z of these ere FCB Impregnated capacitors. It is estimated that the cost of eepaeitors would increase by at least .1001. in going to mineral oil. It is likely that there would be an added inflationary force with demand far exceeding supply. CONSEQUENCE - COST OF CAPACITORS WOULD AT LEAST DOUBLE. AT CURRENT PCB CAPACITOR SALES LEVELS, THIS WOULD ADD ABOUT $145 MILLION IN CAPACITOR COSTS TO EQUIPMENT MANUFACTURERS. TO THIS WOULD BE ADDED AT LEAST ANOTHER $200 MILLION IN COSTS AT THE END-CONSUMER LEVEL. E. Reliability and Warranty Tha U f a expectancy of capacitors exceeds 10 years for lighting j applications and more than 20 years in electric utility power transmission [ applications. Although capacitors are considered long-lived products, there la an aver present danger of short life and early failure. This can be eaused by a number of factors such as poor process control, materials c quality, mls-application, and new product introduction with inadequate < knowledge and experience of reliability. v The General Electric Company warrants its capacitors for one year, and has extended this warranty where capacitor failures have been v excessive and due to shortcomings in capacitor quality. The major reliability and warranty problems lie in Che small industrial type capacitors 774050 IOS 00239 7 neh as used for lighting, no tor*, *nd power supplies. The capacitor ladottTj produces in eh* order of 100 Billion c*p*citora pr year. The M j o r l t y of eh*** unit* **11 la eh* prlc* rang* of 50c to $4. The *rviee cost for replacing a failed cepeeltor 1* $20 Co $25. The in-warranty failure rate of General Electric capacitor* during the firat year la v*ll wader 0.51. V* bli*v* that fehla reliability la typleal of the capacitor laduatxy in general.. Thi* performance ha* b**a attained after many yeara of developing the PCS capacitor, aad establishing expected norma for failur rate* end service life. The capacitor industry historically is very careful in introducing product aad process changes because small changes in failur* rate, which ere extremely difficult to determine without years of service experience aad statistics, could result la excessive complaint and replacement costa, lor example, Zable 4 shows the replacement cost if the failure rate during the warranty period increased by different levels. Zable 4 * Replacement Costa for Early lallurea I of Annually Installed Capacitors Palling, over Existing Rata Incraasc i n 1 Fercentaee Points Millions Reelaeed Cost at $25/Service Call 1.0 .... . 1..P.... .... $25 million 2.0 2.0 $50 " 3.0 3.0 $75 In order to avoid eh* consequences .of such warranty costa, the eapacito: Industry would design any new non*?C3 produets on the conservative side. However, ehera would still be a substantial risk associated with a mass changeover to a new Ispragnant, especially in the end-consider product area, with essentially no experience on reliability and safety. CONSEQUENCE - IF THE FAILURE RATE OF THE-100 MILLION CAPACITORS PRODUCES IN ONE YEAR HERE TO INCREASE BY 1 TO 3 PERCENTAGE POINTS, THE REPLACEMENT COSTS AT $25 PER SERVICE CALL WOULD RUN FROM $25 to $75 MILLION. THIS IS THE MAGNITUDE OF THE CAPACITOR INDUSTRY'S EXPOSURE IN GOING TO A NEW, UNTRIED IMPRECNANT ON A HASS BASIS. IV. CONSEQUENCES OF A "BAN" TO CAPACITOR USERS This saetlon estimates the effeet of a PCB ban on customers and users of capacitors. 774051 IDS 002370 i\t\4->C(\(\ CRITICAL 'CCHPCNENT AVAILABILITY Cipirlt o n are a critical cad neceesary component la end-uaer equipment,! particularly in smell Industrial applications. Without the capacitor, the cad produet would not operate ee designed. The capacitor shortage resulting from a "ban" on PC8 capacitors vould force a reduction by at least half from the social production rata la the plants of capaeltor CONSEQUENCE - INCREASED LAYOFF AND UHQiFLOYHEHT A3 MANUFACTURERS ADJUST TO THE SHORTAGE. ELECTRIC POWER SYSTEM SAVINGS The magnitude of the energy levels under discussion eaa be assessed from the following: - The U.S. capacitor industry produces the equivalent of 55 million kllovar per year. - A kllovar Is equivalent to at least 1 kilowatt of generating capacity (some estimates place this ratio of kilowatt/kllovar at up to 1.4). About 20Z of the eapaeltors are for replacement purposes, and the .remaining 80Z are for new service. Thus,- 44 million kllovars are installed on new systems. Without these capacitors, the steam turbines and fuel requirements vould be relatively unaffected, but generators vould be larger to handle more current, and certain components of transmission and distribution systems (e.g., step-down transformers) vould have to be larger to prevent burn-out. The additional generating capacity for 44 million kilovolt-amperes, priced Incrementally, vould amount to about $300 million; the additional transmission and distribution equipment vould amount to about $700 million. CONSEQUENCE - REDUCED AVAILABILITY OF CAPACITORS WOULD ADD' FURTHER SUBSTANTIAL BURDEN TO THE ENERGY SUPPLY SZZUATIQN. Y k T l~~f C \ POWER- SYSTEM STABILITY The preceding diacusalon indicates that electrical energy requirements could theoretically be met by using fever capacitors and building oversize generation, transmission and distribution components (at a severe economic penalty). Such a system vould be operating at low power factor -- or, in other words; the system vould be carrying s large component'of useless reeetlve volt-amperes in addition to the in-phase component that represents < < < 774052 IDS 002371 energ7 delivery. System studies beve shown (see Economics of Kllovar Supply by Campbell, Ros, and ?lee, Cenerai Electric Company, June 20, 1972] that low power factor severely limits the capability of a power pool to Import power from its neighbors during a losa-of-generation emergency. CONSEQUENCE - CAPACITOR SHORTAGES WOULD MARE LARGE POWER SYSTEMS SUBJECT TO CASCADING OUTAGES, OR "BLACK-OUTS." SIZE AND REPLACEMENT As pointed out earlier, the else and eost of mineral oil impregnated capacitors would be at least tvlca that of FC2 capacitors. The lighting Industry has designed their ballasts for the smell size PCB capacitors. Approximately 2OX of ballast capacitors are used in the replacement market. Many room air conditioners have no apaea for a larger capacitor if a replacement la required. The capacitor Industry has continually moved to reduce the size of capacitors. Equipment manufacturers have taken advaneag of these size reductions, and have no room for size increases. In certain eases, size limitations in the replacement market would force the scrapping of good light fixtures and air conditioners. CONSEQUENCE - UNAVAILABILITY OT SMALL SIZE CAPACITORS WOULD FORCE THE EARLY OBSOLESCENCE OF GOOD, FUNCTIONAL EQUIPMENT. . ,. General Electric Company November 19, 1973 7 fV b C fifl IDS 002372 774053