Document pp6vZ63X70BKnvmbpa25Mm57X

INTERIM FIKAL ERA 360-5-13-029(a) ACUTE HAZARDOUS EVENTS DATA BASE EXECUTIVE SUMMARY Prepared for The Office of Toxic Subataneea Economic* and Technology Diviaion U.S. Environaental Protection Agency undar EPA Contract Ho. 61-02-4053 Praparad by Induatrial Eeonomica, Inc. Cambridge, ma Management Technology and Cara Syateaa, Inc. Boaton, MA PEI Aaaoeiates, Inc. Cincinnati, CH December 1915 HONS 026440 FORWARD ERA publiahad critaria for ealecting aubatancea on a lietirg of aeutaly hatardoua aubataneae, aa wall aa a Hat raeulting froa tha application of thoaa criteria, in Dacanbar liS. Guidance docuaanta for making uaa of tha critaria and liat wara aada available at tha aana tiae. Theae aatariala wara publiahad aa intaria final docuaanta, with opportunity for public raviaw and coaaant. Tha data baaa that ia tha aubjaet of thia report waa devel oped aa a part of tha procaaa of aatabliahing tha critaria, ar.d lnforaation froa tha data baaa waa uaad in drafting tha publiahad critaria. ERA plana to add ethar racorda to tha hiatorieal profila praaantad in thia intaria final report, in order to aid in tha validation of tha liat-aaking critaria. Currant plane era to aupplaaant tha currant 3,121 racorda with racorda froa two additional yaare of data froa tha National Raaponaa Center and froa ona or aore Stata aoureaa. HONS 026441 ACWfOVLCOGCMENTJ This study wee conducted under CFA Contract number 11*02. 4099 by Industrial Economic*, Inc.; Management Technology and Data Systems, Inc.; and HI Associates, Inc* The principal aembere 0f the project team vara: Or. Janas Cumaings-Sexton, Ns. Laura laehnar, and ns. Anna K. crook (Industrial Economies. Inc.) Or. Semuol J. Natick, Nr. Fatriek Suck I ay. and Hr. seepnan wu (Menagoaont Technology and Data Systems, Ine.); and nr. Thonas K. Corwin, Hr. Sdvln A. Ffettlnq, Hr. Z. Fadha Krishnan, and ns. Barbara A. trace (FIX Assoeiatas, Inc.). Tha project team racaivtd valuable and generous assistance Iron Individuals In several organltstlons in acquiring and making use of tha data that vent Into tha study. Among these are the KFA Office of Solid waste and Emergency Fesponse and tha Trans* portatIon syatoaa contor of tho U.S. Department of Transportation for aceoao to tho National Faiponaa cantar data; CFA's Fegion 7, for baking availablo lta computerised data base; tho California Highway Patrol and Offlca of Emergency Sarvleos; tho now Joraoy Oopartoant of Znvlronpontal Quality; tho Ohio Environmental Frotactlon Agency; and tha Taxaa Air Control board. Several SPA employees had aetlva roles In tha project. Mr. Fradarlek n. Taleott was rasponaibla for assembling and ceding raeords frosi tha nows modi*; ha was assisted by staff members gf tho KFA Headquarters Library and tha Office of Standards and Hogulatlon*. both Hr. Taleott and Ha. Karan Seat conducted analytle atudiaa using tho data baas that have substantially contributed to thlo report, nr. Janto Cottrell of tho SFA Offleo of Toxic Svbotaneoo, Intonation Manaqosont Oivision, provided valuable asalatanco in assigning cas eodot to several hundrad listed oubavanoao. Ha would like to thenk tho EFA project officers, nr. Frsdsrlck w. Taleott and Hr. Mail Fatal, for their active guidance and participation. Overall design and direction wee provided by tho Anolysls subqroup, ltd by Hr. Taleott. and inelu* dinq Hs. last# Mr. Fatal, and Hr. David wgner. These indivi duals end other members of tho Acuta Hatsrds List workgroup provided reviews, cements. and other valuable assistance throughout tho study effort. Copioo of this Kxocutlve Summary ar.d of tha full report, Acutq Htaerdous tvonto Data bait (report number CFA 9o-s-soati are available to tno public froa tho national Technical Inforaation Sasvlea, Springfield, Virginia ansi HONS 026442 ACUTE HAZARDOUS EVENTS DATA BASE: EXECUTIVE SUMMARY Tha Acuta Hazardous Events (AHE) Data Base was assembled as pact of CPA's review of the dangers posed to tha O.S. public and industrial workers by sudden, accidental releases of toxic chemicals. concern over such releases was heightened by tha tragic incident that occurred on Deceaber 1, 19S* in Bhopal, India. EPA formed an Acute Hazards List Workgroup to investigate the safety-related characteristics of U.J. industry vlth regard to accidental chaaical releases that could pose exceptional risks to huaan health, and to identify the eheaicals or situations that appear to represent unusually high risks. In late Kay IMS, the decision vas aada to asseable an Acute Hazardous Events Date Baee as part of the Acute Hazards List review process. PVRPOSE AND APPROACH The aaln purpose of the AHE Oats Base is to characterize tha kinds of events releasing acutely toxic substances in the C.S., the aubataneaa Involved, end the causative factors leading to their release. Tha AHE Data Base provided an historical perspec tive to members of the Acute Hazards List workgroup, in order to select criteria for a priority list of acutely hazardous substances, EPA vantad to ensure that the resulting priority list was relevant to recent accident history in tha U.S. Tha AHE Data Base was not constructed to serve aa a basis for nation-wide estlaatas of frequencies of events, qusntities released, or their consequences. Emphasis was placed on acquiring a aaasura of tha scope of events, rather than on estiaating quantities precisely or frequencies of releases. An laportant factor in the AHE Data Base developaent vas the need for it to ba assembled quickly in order to support other aspects of tha analysis process. Because of tha rigorous tins constraints, several criteria were employed to screen data for inclusion in the AHE Data Bass. Emphasis was placed on Incidents HONS 026443 that -had acuta hazard potential. Zncidants that Involved dea*h or injury war* given highest priority. ERA further directed that priority be given to incidents involving chemicals released lr.to the air. Chemical raleasas were given priority over petroleuaralated and other types of releases. Releases involving large quantities of aatarial vers given priority, and events releasing lass than one pound of eaterials vers not coded. Incidents involving evacuations also were given priority. Although the data base Includes evsnts of many typea and with a vlda range of consequences, the AHZ Data Base has been purposely biased (through these data collection criteria) toward events considered to have the greatest potential for sudden, large-scale hare to the populace. Thus, the suaaary statistics obtained froa the data base are not reflective of the totality of potentially hazardous doaestie aaterials. ' DATA SOURCES AMD COVERAGE The data base contains 3,121 records which, through saaplinq, represent S.92S separate events. The records eeae froa thirteen public sources chosen because of their accessibility, their potantial to yield a large number of events aeating the sslaetion criteria described above, and their geographic coverage. The events occurred mostly froa 1910 to aid-1995, with 1993 and 1994 most heavily represanted. One nation-wide government source, the National Response Center, contributed 1,22S records. A multi-state regional aourea, ERA'S Region 7, provided S13 records, rive offices of four state governaents were the source of 1,029 records. Pour daily newspapers end one vire service contributed 290 records. Flnelly, 64 records vara obtained froa a suaaary prepared by a.-, engineering consulting firs specializing in industrial accidents. Sea Exhibit A. EVENTS WITH INJURIES OR DEATHS Human casualties occurred in fewer then seven percent of the recorded events. These events a total of 461 -- lad to 139 deaths and 4,717 Injuries, ranging in severity froa teaporary respiratory problems treated on-site to critical injuries ar.d extended hospitalisations. Information on causation was scanty in aost records. See Exhibit B. Tour high-volume industrial inorganic chemicals (chlorir.s. aaaonia, hydrochloric acid and sulfuric acid) together were reported to have been raleeead in over 23 percent of the events recording human casualties. Ses Exhibit C. More than 203 2 HONS 026444 additional Identifiable aubetaneea were recorded ae released during events eseoelated vith death* or Injuries, with Industrial organic cheaieals, as a claas, contributing the largest nuaber of events vith huasn casualties. See Exhibit D. At least another 200 substances vers recorded as being released in events not resulting in deaths or injuries. Events reported vith huaan casualties included soae with very large quantities and soae vith very eaall quantities re leased. see Exhibit E. Similarly, the inherent toxicity of the substances released need not necessarily be particularly high for deaths or injuries to have occurred. Exhibit r illustrates this, shoving that as sany svanta vara recorded vith subetances in the two least toxic Reportable Quantity categories (i.a. RQ - 3,000 pounds and 1,000 pounds) as vara recorded for the tvo eost toxic categories (l.e. RQ - 1 pound and 10 pounds.) When the charac teristics of the released substances are exaalned, toxicity ap pears to be the causa of eost of the injuries recorded, while flaeaabillty and exploelvity ere the mechanises associated with aost of the fatalities in the data base. See Exhibit 0. Transportation releases account for 23 percent of all events in the data base, and a aoaevhat higher percent of death or injury events (33 percent.) See Exhibit H. Trucks are the predoainant aode of transport in the death and injury events. Aaong events vith huaan casualties at fixed facilities, storage vessels play a such larger role as release points than they do among events not leading to deaths or Injuries. Whsn viewed froa an industry perspective, the cheaieals and Allied Products and the Petroleua Refining industries togeeher account for 34 percent of the reported injuries, and sore than half of the reported deaths. The transportation industries account for 3< percent of the reported Injuries and about onefourth of the reported deaths. Industries or vendors that use or store ehaaiesl or fuel products account for about 23 percent of the deaths and injuries. See Exhibits I and J. OTHER ErrSCTS Information on evacuations, property damage and environmen tal damage vaa seldom provided in the records examined. In records furnishing such information, evacuations totaled sore than 217,000 individuals, vith one event accounting for an esti mated 20,000 evacuated people. Information on environmental and property damage was usually expressed in vague terms. However, one source reported S4 large-scale industrial aceldanta (over a 23 year period) having an average estimated economic cost of 330 million each (in 19(4 dollars), vith one event resulting in more than 3100 million in estimated damages. Sea Exhibit K. 3 HONS 026445 QUALITIES RELEASED AND ASSOCIATED FACTORS Eighty percent of tha avanta raportad a quantity for tha atarial ralaaaad, and tha aaount rangaa froa a fav pound* to wall ovar 100,000 pound*. Saa Exhibit L. Aaount ralaaaad axeaad* 1,000 pounds for ovar 31 pareant of tha racordad avants. Ralaasas ovar 100,000 pound* occur In lass than thraa pareant of tha avants, but thasa avants account for 93 pareant of tha total quantity of materials ralaaaad. The total quantity raportad ralaasad for all avants axeaad* *20 Billion pounds. ovar (0 pareant of tha avants In tha data basa raportad that at least on* of tha substaneas ralaasad was a liquid; 1* parcar.t of tha avants Involved the release of a gas; and 3 percent of the avants Involved tha release of a solid. Aaong avants occurring at fixed facilities, spill* are tha predominant and affect, followed by vapor ralaasas, fires, and explosions. Storage vassals, process vessels, and valves or piping are responsible for nearly equal shares of ln-plant events, (see Exhibit M) but storage vessels typically release nuch larger quantities than tha other release point* in a plant. Equlpaent failure is the esuse aest frequently reported for in plant avants, followed by operator error. So* Exhibit x. Accident causa* ara difficult to assess for aany of tha avants i.-. tha data bass,' however. Ovar half of the ln-transit avants involve trucks, and another 3< pareant involve rail ears. Saa Exhibit O. ovar 3* percent of tna ln-transit releases staa froa a leak, and another 20 percent froa collisions. Saa Exhibit R. Although few events wars raportad for pipelines, those spills dwarf other in-transit ralaasas in taras of total quantity ralaasad. 4 HONS 026446 EXHIBIT A Of (nan or woe, w 1U, tacit________________ Iwwwh fmttr **0 Hmi M'* | numwm IM f MMMrii im **<# rt* CMh THfcf* H* Trk nm MtMII ii^t jwl m 1211 1211 iau Tlrt Mlf 12|o 12B total *10 *79 ? tin a t i t i tot so *4 S4 StS tu log 10f SI t|9 M Sts SSJ I s to s St so .. .. T S a. .. to '* t| S) " 5 ~ 1 1w ItltNl e*llf. ftfftaa f lat'tancir Servian Calffamfa iifAMy Ntftl VMM ilf Cmn( lwr om* m * ><>> NMftawii f ttf u HI to 00 t 00 Tt SI St s so tit too so US IS 01 001 Ml kMmi I oo so IS tt* t *7 U| IM 01? too i.ott rotai so Tt US Its 004 i.no Its S.tn MOMS 026447 EXHIBIT B ACUTE HAZARDOUS EVENTS DATABASE OCCURRENCE OF INJURIES ALL EVENTS OCCURRENCE OF DEATHS ALL EVENTS NUMBER or EVENTS = 6928 NUMBER or EVfNTS = 697R HONS 026448 EXIIIDIT C ACU1E HAZARDOUS EVENTS DATABASE DtSTRBUTKlN OF EVENTS BY CHEMCAL FOR DEATFVKJURY EVENTS Contribution of the 10 Most Frequenlly-Occurrinc Chemicals to the Events with Deaths orlnjwko OIMtR M.JX CHLORINE MX ANHYDROUS AMMONIA tax surunic ac 4.7X PCD 2.8X hyorochumdc ach> S.SX NltMC AC I.5X tOiUCNC 2.4X MCIHYI ALCOHOL 0.4X SI O9XOWM HYDROXJOC HONS 0264*9 EXHIBIT O DISTRIBUTION OF EVENTS WITH INJURIES Contribution of Cbomlcol Cotegorlaa iniMm n or rvrNis = 4G8 HONS 026450 EXHIBIT E ACUTE HAZARDOUS EVENTS DATABASE FREQUENCY DISTRIBUTION BY RELEASE SIZE FOR EVENTS INVOLVING DEATHS AND INJURES Legend EZ) NUMBER I AMOUNI HONS 026451 EXIIIDIT F ACUTE HAZARDOUS EVENTS DATABASE NUMBER OF EVENTS BY REPORTABLE QUANTITY FOR DEATH AND INJURY EVENTS HONS 026*52 EXHIBIT G ROLE of FIRE/EXPLOSION vs TOXICITY X o f Injuries or X of f'7'l F ire/Eploion JZW Toxicity HONS 0*6453 EXHIBIT H ACUTE HAZARDOUS EVENTS DATABASE IN-PUNT VS. IN-TRANSIT EVENTS ALL EVENTS IN-PLANT VS. IN-TRANSIT EVENTS DEATH/INJURY EVENTS miMorR or cvtms = r.9?n NiiMnrr or rvcriis = 468 HONS 026454 EXHIBIT I Reported Number of Injuries Distribution by Industry Sector Chemicals -- Unknown . 10 Basic Chemicals 15 me4 * ^041 4 >yculture te 2 % Other Manufacturing 13# uni'iin ine wtf mt 2 s Trade end Other Services 9 ft in - Transit and Transportation Services ... 36 % TOTAL INJURIES 4717 in five year period HONS 026455 EXHIBIT J Reported Number of Deaths Distribution by Industry Sector Ptrolgm Rifimnf 13 J5 CtomieslJ - unmown 9 5 a*1l !! *1 Cnamcai c intarmeaiatts c ana H Fimsnaa c Proaucts 4n -' 25 " Othtr Manufecturt u n . .: 7 * Uiiliim ina wish rpi 5 IS wnolasaia and Rataii Trada 12" and Othar Strviets 2 * In-Transit and Transportation Servicas 2*55 J TOTAL DEATHS 138 in five year period HONS 026456 EXHIDIT K EVENTS WITH HIGHEST DAMAGE COSTS (COVERING A 28 YEAR PERIOO) ' $10-29 $30-49 $50-69 $70-89 Ovtr $90 ____ COST [771 TRANSPORT |ES: MILLIONS OF 1984 DOLLARS CHEMICALS Vm PETROLEUM HONS 026457 EXHIBIT L ACUTE HAZARDOUS EVENTS DATABASE FREQUENCY DISTRIBUTION BY RELEASE SIZE FOR ALL EVENTS Legend za NUMKR 1 AMOUNT HONS 026458 EXHIBIT M ACUTE HAZARDOUS EVENTS DATABASE IN-PLANT EVENTS BY LOCATION ALL EVENTS IN-PLANT EVENTS BY LOCATION DEATH/INJURY EVENTS NUMOCR or tVtNTS = 5179 number or rvtNis = 304 HONS 026459 EXHIBIT N ACUTE HAZARDOUS EVENTS DATABASE IN-PLANT EVENTS BY CAUSE ALL EVENTS IN-PLANT EVENTS BY CAUSE DEATH/INJUKY EVENTS NUMBER Or EVENTS = 0179 NUMBER Of EVENTS = 304 HONS 026460 EXHIBIT O ACUTE HAZARDOUS EVENTS DATABASE IN-TRANSIT EVENTS BY MODE ALL EVENTS IN-TRANSIT EVENTS BY MODE DEATH/INJURY EVENTS number or tvrms = 1749 NUMOfR or CVfNTS = 164 HONS 026461 EXHIBIT P ACUTE HAZARDOUS EVENTS DATABASE IN-TRANSIT EVENTS BY CAUSE ALL EVENTS IN-TRANSIT EVENTS BY CAUSE DEATH/INJURY EVENTS MOMS 026462