Disasters as Heuristics a Case Study

7/30/2019 Disasters as Heuristics a Case Study

1/5

Australian Journal of Emergency Management32

Disasters as Heuristics?A case study

by Dr Simon Bennett, Lecturer in Risk

Studies, Scarman Centre, University ofLeicester

IntroductionDuring the early 1970s the members of alocal environmental group in Port Talbot,South Wales began complaining aboutthe environmental and health record of alocal petrochemicals plant. It was allegedthat the plant generated dust, noiseand light pollution, and odours. Concernwas also expressed about the use ofvinyl chloride monomer, a suspectedcarcinogen, as a feedstock gas. The plantsexplosive hazards were not a major boneof contention.

On June 1, 1974, the Nypro nylon-

manufacturing plant at Flixborough onHumberside exploded killing 28 workers.The deaths and property damage to localhouses were widely reported. Followingthis event, the Port Talbot group began tofocus on the potential explosive hazardsgenerated by the petrochemicals plant.

This paper attempts to explain thechanging risk perceptions of the BaglanAction Group (BAG) following theFlixborough disaster. Two discourses areemployed: Kaspersons (1992) social

amplification of risk model,

and thesocial psychological theory of heuristics(Rachlin 1989). The paper commenceswith a brief description of the twodiscourses. This is followed by a historyof the petrochemicals plant at Port Talbotand a description of the Flixboroughdisaster. Case study data is presented, andsome tentative conclusions drawn.

DiscoursesThe Social Amplification of Risk modelFor Kasperson, our experience of risk has

two dimensions: First, actual physicalharm; Secondly, the acquisition orcreation by individuals or groups ofinterpretations of hazards. The socialamplification model focuses on the latterdynamic. Specifically it addresses thepossibility that psychological, social,institutional, and cultural processescan heighten or attenuate perceptions ofrisk and shape risk behaviour. The socialamplification model is premised on theassumption that risk is part objective

threat and part a product of culture andsocial experience. Not all risks are givenequal attention. Receptivity is a functionof currently held beliefs and agendas(Kasperson 1992):

Some groups and individuals activelymonitor the experiential world,searching for hazard events related totheir agenda of concern[they] pro-cess the information [and] locate it intheir agenda of concernsSome maychange their previously held beliefsand be motivated to take action

[S]ignals that are inconsistent withprevious beliefs or that contradict thepersons values are often ignored orattenuated.Crucially, therefore, groups or indi-

viduals may amplify or attenuate risksignals. As Kasperson (1992) puts it, Eachcultural or social group selects certainrisks and adds them to its strand ofworry-beadseven as it selects out otherrisks. Kasperson uses Rayners electronictuner imagery to explain why some risk

messages are picked up and others not.Thus according to Rayner and Kaspersonpeople are pre-tuned to pick up onlycertain kinds of messages. Kaspersonelaborates by asserting that while mostindividuals are largely reactivemanyorganisations actively seek out and orderrisks. The media plays an important rolein amplification or attenuation processes.The processing of risk events by themediaprofoundly shapes the societalexperience with riskHeavy mediareporting appears to stimulate social

mobilisation, says Kasperson. The mediaresponse is roughly proportional to thephysical consequences of the event, whilerisk perception appears to incorporateextent of human exposure as well as riskmanagement performance.

HeuristicsA heuristic is a mental shortcut thatprovides a simplifying routine thatleads to approximate solutions toeveryday problems (Fiedler and Schmid;

Hewstone 1996). In todays complex worldit is sometimes difficult to come to adecision. There seems to be too muchdata to process in the available time. Suchcognitive mechanisms as heuristics

help the individual to save mentalresourcesand enables him or her tocope with multiple affordances at thesame time (Fiedler and Schmid; Hewstone1996). Take the problem of deciding thesafety of air travel. Rather than examineall available data, one might be temptedto calculate the safety of air travel on thebasis of those events that come mosteasily to mind. Unfortunately, giventhe widespread and often detailedreporting of air disasters, one mightdecide that air travel is dangerous. Highprofile air disasters offer the public an

availability heuristica means ofquickly estimating the safety of airtransportation. Air disasters constitute asimplifying routine for the processing ofinformation and decision-taking. AsPerrow (1984) puts it, If there has recentlybeen an airline crash, we focus on thatevent and ignore all the successful flightswhen we think about the probability of acrash while deciding whether to take aflight or not. Making a decision on thebasis of a limited set of data (an easily

recallable event, for example) is knownas satisficing. Ideally, of course, oneshould optimisethat is, spend time andeffort weighing up all available data beforemaking a judgement. In the case of airtravel, for example, this would meanavailing oneself of transportation safetydata. Given that air travel is three timessafer than rail travel and thirteen timessafer than car travel, one might well decideto go by air rather than by any other formof transport. But, as described above, thereis a tendency to use mental short-cuts in

decision-making. To save time and effortwe allow ourselves to be influenced bysuch memorable and vivid events asdisasters or the modes of death of therich and famous. As Rachlin (1989)explains:

You are driving along the highway andsee the flashing lights of police cars andambulances You drive a little slower

for the next hour or so. You hear thatPresident Reagan has rectal cancer and

you make an appointment with your

own doctor. In making judgements aswell as decisions we tend to use theinformation that is most available tous It is of course easiest and fastestto make judgements on the basis of what


2/5

Spring 1999 33

is easiest to call to mindthe heuristicof availability.

The Port Talbot PlantThe petrochemicals plant in question wasowned and operated by British PetroleumChemicals International. Sited at BaglanBay near the town of Port Talbot, itcovered about 600 acres of a 1,000 acresite. BP built the plant in the 1960s, sitingit adjacent to a low-rent housing estate of15,000 people, (Jones 1989) a school with1,700 pupils and a sports ground.

The school was situated about 500yards from BPs perimeter fence. Workon the housing estate had commenced in1950. The plant came on stream in 1963.By 1968 it employed around 600 people.The company boasted of the plants SpaceAge technology that would help theNations Balance of Payments by elimina-ting imports.(BP Chemicals International

ad. 1968). BP completed a major additionto the plant in March, 1974, by which timeit had invested a total of 140 million. Theplants incremental growth reflected abuoyant UK petrochemicals sector(Cranfield 1974). Petrochemicals plantspresent both fire and explosive hazards.The larger the plant and the greater thevolume of on-site storage, the greater therisk (Napier 1974).

There had been two major fires at theBaglan Bay plant prior to the Flixborough

disaster. On 6 February 1973,Baglans new ethylene plantcaught fire during com-missioning. Flames soared150ft above the works.Fourteen workers wereinjured. In addition to worksappliances, five municipalfire engines attended thescene. According to onenewspaper (The Port TalbotGuardian, 9/2/1973) the localpopulation seemed ignorant

of the cause and nature ofthe problem:

Thousands of people on theSandfields and Baglan

Estates watched as columnsof yellow flames belchedskywards. Many believedthe fire was all part of thecommissioning work.It emerged that the fire had

been caused by a componentfailure that allowed large

volumes of flammable hyd-rocarbon gases to escape athigh pressure (The PortTalbot Guardian, 23/2/1973).There are similarities

between this incident and the later Flix-borough explosion, where a flammablemixture escaped under pressure (Kennett1982). At Flixborough, however, the gasexploded with enormous force (seebelow).

At the end of February 1974 the styreneplant caught fire releasing a large quantityof smoke which drifted like a hugepall across the nearby Sandfields Estate.Eight municipal fire appliances attendedthe blaze, which burned for 35 minutes.The headmaster of the local school, onseeing the dense black pall, ordered thatall the schools windows be closed (ThePort Talbot Guardian, 1/3/1974).

By the end of 1977 the number ofpollution-related complaints received bythe plant had been greatly reduced (TheWestern Mail 24/11/1977), but the plantcontinued to suffer periodic fires.In March, 1983, a blast in the ethylene

plant injured two workers, one severely.As The Western Mail reported on 10March:

The blastsent thick black smokehundreds of feet into the air. The forceof the explosion rattled windows up tothree miles away. The blaze took twohours to bring under controlThirteen

fire appl iancesincluding 10 watertenders, two foam units, a turntabletender [sic] and an emergency unit

fought the blaze.

In October 1990 the ethylene plant againcaught fire. It took 12 hours to bring theblaze under control.

Despite such obvious setbacks,however, during the 1990s Baglan Bayslocal management made great efforts toinvolve the community more with theplant (Bennett 1996).

Baglan Action Group

BAG was formed in February 1974 inresponse to perceived inadequacies inBPs attitude to pollution. It had tenmembers. Initial concerns focused onthe medical effects of industry onpeople. While members were aware ofthe potentially explosive (The WesternMail 19/2/1974) nature of BPs activities,until Flixborough this particular risk didnot feature prominently in exchangesbetween BAG and BP Chemicals.

The Flixborough Disaster

The Nypro plant was located on mostlyflat land between two villages, Flixboroughand Amcotts. The site extended over 60acres. The population could choose towork either at Nypro, a local steel plant orin agriculture. Although Nypro employedabout 550 people, few Flixboroughresidents worked there (Clammer 1974).The plant produced Nylon-6 for thegarment industry by combining benzenewith ammonia. A Nypro chemist likenedthis process to boiling petrol (Scullion

Map of the Port Talbot region surrounding and including the British Petroleum Chemicals Internat ional plant at Baglan Bay.Reproduced from Ordnance Survey mapping with the permission o f Ordnance Survey on behalf o f Her Majestys Sta tionery Office Crown copyright. NC/99/005.


3/5


1974). Twenty-eight workers were killedand 36 injured in the disaster. Had thedisaster not occurred at the weekendmany more workers would have beenkilled (Taylor 1975). The plant explodedwith a force of between 15 and 45 tons ofTNT (Kirkwood 1997). Some witnesseslikened the event to an atomic explosion(Kennett 1982). A police report des-

cribe[d] an enormous black mushroomcloud laden with debris which rose to overa mile (1.6km) above the devastatedareaFalling debris started small fires upto three miles (4.8km) away (Kirkwood1997). The main conflagration lasted24 hours. Smaller explosions occurredthroughout the following week. Threethousand local residents were evacuated.One thousand homes lying within athree mile radius were damaged.

Out of Flixboroughs 79 houses, 72 weredamaged (Brown 1990). There were 53

recorded casualties outside the perimeterfence. The disaster was covered in themedia and discussed in Parliament. Thelocal MP asserted that the cost in termsof grief and misery my constituents havehad to sufferis too high for a so-calledcivilised society to bear (Ellis in the Times1974).

MethodologySecondary sources only have been used,specifically the reportage of three local

newspapers, The Western Mail, The SouthWales Echo (sister paper to the Mail) andThe Port Talbot Guardian. The papershave been used to tell the history of therisk debate and as a source of witness

statements. The publication dates aregiven in parentheses.

It is accepted that this methodology isnot optimal. Contemporary researcher-conducted interviews with protagonistsand participant observation would haveprovided the most reliable data.Nevertheless, on the assumption thatreporters recorded events and the views

of group members accurately, the threesecondary sources may be considered toprovide a truthful account of the terms,nature and course of the risk debate.

The risk debate as reported byThe Western Mail, Port TalbotGuardian and South Wales EchoPrior to the Flixborough disaster, the riskdebate pertaining to the BP Chemicalsplant had two major themes: environ-mental pollution and the health risk toworkers presented by vinyl chloride

monomer (VCM). VCM had been linkedto cancer of the liver. The pollution debaterevolved around four aspects of the plantsoperation: its alleged dust and odouremissions, light pollution and itsgeneration of excessive noise (Mail 15February 1974). The Baglan ActionGroupFormed to fight pollutionorchestrated the campaign against BP. Asthe leader of BAG put it:

Our complaints are well known. Theyare of smoke, noise and chemical

pollutions [sic] and we are desperatelyanxious to have some early answers toour questions (Echo 19 February 1974).BAG worried that pollution would

devalue property (Mail 20 February

1974). It is BP thats devaluingthe property said the actiongroups Secretary in late February1974 (Echo 20 February). In thesame month a local councillorcomplained that the noisenuisance is as bad now as it was12 months ago (Guardian 22February 1974).

Questions about BPs environ-mental performance had beenraised prior to BAGs formationin 1974. On 19 September 1973 theHeadmaster of the school adjacentto the site had complained tothe Port Talbot public healthinspectors office about a pungentodour. Residents complainedthat the smellmade them feeldrowsy and gave them a burningfeeling in their throats. A spokes-man for BP stated that it was not

certain the odour had come fromthe plant (Mail 21 September1973).

Protest against BPs environmentalperformance took a number of forms.In September 1973 three residents on alocal private housing estate held a ratesstrike (they refused to pay their propertytax). One complained that pollution fromthe plant was affecting property values:

Our lives have become intolerable as aresult of the noise and pollution coming

from the BP plant over 14 months andwhich has continued despite comp-laints. When our rates were increasedunder the new assessments it was thelast straw, because the conditions havereduced the value of our properties.Another rates striker complained about

the inhuman noise pollution producedby the plant (Guardian 21 September 1973).

At the beginning of 1974 The WesternMail reported that the deaths of threeworkers in America who had beenexposed to VCM gas were being investi-

gated. On 31 January, The Western Mailannounced that Medical records of about700 men at BP in South Wales are to bechecked after it was found they have beenin contact with [VCM].

The themes of environmental pollutionand workers health dominated the publicdebate about the plant between Januaryand May 1974. On 9 May, The Western Mailreported that the British Society for SocialResponsibility in Science (BSSRS) hadurged local residents [T]o demand that

BP Chemicals at Baglan Bay publish theirpollution figures. BAGs Secretary sup-ported the BSSRS call for more infor-mation on pollution to be supplied toresidents. Acknowledging the campaign,

The British Petroleum Chemicals International plant at Baglan Bay, Port Talbot, South Wales in January 1993. Photographedfrom the western bank of the River Neath. The town of Port Talbot and the Sandfields Estate are located behind the plant.


4/5

Spring 1999 35

BP revealed the extent of its investment inremedial measures to the press. As TheWestern Mail explained on 16 Feb 1974:

The company is spending about 60,000a year checking on pollution...30,000on cutting out smells...and a further

225,000 on installing silencers to reducenoise.Following the Flixborough disaster of

1 June 1974, however, the focus of concernof the members of BAG shifted frompollution and VCM to explosive hazard.As The Western Mail put it; Until theFlixborough disaster at the weekend, theBaglan Action Group had channelled theirprotests towards noise and pollution atthe plant (4 June 1974). BAGs Secretarydescribed the mood shift thus:

I think everybody was a little stunnedand shocked when the news came out.But what has alarmed us is that theexperts have said this sort of thing could

happen again (Echo 4 June 1974).(This comment bears certain simi-

larities to statements (Kennett 1982) madeby Flixborough residents on realisingthat they had been living next to a plantwith the capacity to explode with greatdestructive force).

Interestingly, in a statement to TheWestern Mail, the groups Secretary hintedthat BAG had been conscious of the plantsexplosive risks for some time but hadchosen to remain silent:

Some of us have had anxieties aboutthis type of incident which we have notpressed publicly to avoid being alarmist(4 June 1974).In a letter to The Echo, BAGs Chairman

made the same assertion:Wishing to alarm no one the Baglan

Action Group played down the all-too-real parallel which exists here. Nobodycan accuse us of selfishly exploiting thetragedy for our own ends(17 August1974).It seemed to BAGs Secretary that the

situation at Port Talbot was potentiallymore serious than that at Flixborough:

There were acres of farmland sur-rounding Flixborough, but at BaglanBay there is a school on the perimeterof the siteIf a similar explosionoccurred at Baglan Bay, it would ripthe school from its very foundations(Echo 4 June 1974).BAGs Chairman had no doubt that a

catastrophe was inevitable:Unfortunatelythe question is not will

it? but when will it go up and howmany will go with it? In such anevent both workers and BPs closestneighbours would stand nochanceShould they survive an

explosion the poison gaswouldrapidly dispatch them (Echo 17 August1974).BAG called for a review of the plants

explosive risks, as did The Port TalbotEnvironment Society, Port Talbot NewTenants Association (PTNTA) and theAberavon North Ward Labour Party.Because of the concentration of housingand schools nearby People have theright to know what the dangers are, saidthe Chairman of PTNTA (Mail 5 June1974).

In mid-June, BAG published a reporton the risks presented by the petro-chemicals complex. It contained thefollowing unequivocal statement ofconcern:

In the light of the Flixborough disasterthe dangers of explosion and fire areobvious (Mail 15 June 1974).It would appear that Flixborough had

made obvious to BAG the dangers ofexplosion. BAGs report went on:

The two major fires [see above] havecaused much concern, but untilFlixborough the public did not appre-ciate the horrifying potential of such a

plant At this t ime it would be wrongnot to press for an inquiry into the

general security of the plant (Echo 14June 1974).

In July 1974 the gov-ernments Employment

Secretary announced aprogram of inspectionsof plants by the FactoryInspectorate. The Emp-loyment Secretary com-municated this news toBAG in a letter. BAGsSecretary commented:

We will be watching thesituation closely. Untilsome firm recommen-dations are made weare still going to be very

worried (Mail 18 July1974).Flixborough freed up

the local politicalimpasse that had existedbetween BAG and BPsmanagement. AlthoughBP had invited BAG tothe plant for discussions,BAG had refused togo. After Flixborough,however, BAG met BPs

management on neutralterritory at a local leisurecomplex. A two-hourmeeting saw frankspeaking. A further

meeting was arranged. BP also planned tomeet the New Tenants Association andThe Port Talbot Environment Society at alater date (Guardian 28 June 1974).

ConclusionThe most puzzling aspect of BAGs cam-paign was the way the group played downthe risk of explosive hazard. BAGsmembers attenuated the signals that

Baglan Bay might present a seriousexplosive hazard (the two fires), prefer-ring to tune in to pollution and VCM issuesinstead. This attenuation might beexplained by the nature of the emer-gencies at the BP plant. The two fires wereunlike the Flixborough disaster in anumber of important respects: Thedestruction at Flixborough continued forabout a week; the fires at BP were quicklybrought under control. At Flixboroughthere was extensive damage beyond the

perimeter fence; at Baglan there was nodamage beyond the perimeter. Workerswere killed at Flixborough; no workerswere killed at Baglan. There were off-siteinjuries at Flixborough; there were noneat Baglan. At Flixborough most of the plantwas razed; at Baglan only self-containedsections were damaged. There was exten-sive and prolonged local, national andinternational reporting of the Flixborough

A storage vessel on the 1,000 acre plant at Baglan Bay


5/5


R fereed

disaster; there was much less reportingof the two fires at Baglan. FollowingKaspersons (1992) discourse on themedia it may be this differential thatexplains the groups pre-Flixboroughattitude towards explosive hazard.

As Kasperson explains Heavy mediareporting appears to stimulate socialmobilisation. Could it be that while thegroup had always been aware of the riskof explosion (see statements by groupmembers, above) it took the world-widemedia event that was Flixborough topersuade the group to add the risk to itsstrand of worry beads? Perhaps therelatively low-key reporting of fires atBaglan had failed to persuade the groupto speak out on the issue of explosivehazard. While BAG had identified explo-sive hazard as worthy of inclusion in itsagenda of concerns, such risks were welldown the list.

Post-Flixborough the groupnow fullytuned-in to the risk of catastrophicexplosion posed by such plantsmovedthe risk up its agenda of concerns andlaunched a sustained campaign. Thatis, BAG, having sought out the hazardevent, was sufficiently concerned by thepossibility of a similar fate befalling theplant and community at Baglan thatit amplified the risk. Kasperson alsosuggests that extent of human exposureaffects risk perception. Certainly the

28 deaths and numerous injuriesat Flixborough would suggest the sensiti-sation of other at-risk populations toexplosive hazard.

Nuttals (1959) Standard Dictionary ofthe English Language defines an heuristicas something that leads to discovery.Following this definition the two firesat BPs Baglan Bay plant should haveencouraged BAG to air its views onthe subject of explosive hazard. However,while BAG placed explosive hazard on itsagenda of concerns it remained mostly

silent on the matter, preferring to pursuehealth issues instead.

Then the Nypro plant exploded, killing,maiming and destroying. The disasterwas broadcast across the globe. It evenled, indirectly, to new safety legislation.Here, then, was an event that was difficultto ignore. Rachlins analogy of the trafficaccident may provide an explanation forthe groups behaviour pre- and post-Flixborough. While, as drivers, we are allaware (in varying degrees) of the risks

attendant upon road use, it may take somedramatic and horrifying eventlikeseeing the physical and human costs of acar crash to first, convince us of thedanger and secondly, to persuade us to

modify our behaviour. The car crashthe most dramatic manifestation of thefailure of the socio-technical system thatis car travelacts as an availabilityheuristic, revealing the potential horrorsthat await the careless driver. As Rachlinputs it, You see the flashing lights and youdrive a little slower. Having been sen-sitised you change your risk-takingbehaviour (for a time). As with cardrivers, so too with the members of BAG.The lessons of Flixborough were sodramatic and obvious that, for a time atleast, challenging BP on the issue ofexplosive hazard became the groupsnumber one priority. The group modifiedits behaviour.

But this is possibly not the whole story.There may be another dimension to therisk debate at Baglan Bay: qualitativedifferences between the various risk-related events. The fires were short-term

hazards. The second blaze expired in35 minutes. While there were trans-boundary impacts (the pall of smoke)these soon disappeared.

In contrast, the dust, noise and lightpollution and odours from the plant weremore persistent. Such hazards had anirritant quality. Unlike the transienthazards of fire and explosion, they werefrequently manifest. They also generatedmuch complaint from the community.Given that BAGs primary mission was to

fight pollution, these hazards wereobvious candidates for amplification.After Flixborough, BAGs leaders insistedthat they had been aware of the plantsexplosive risks all along. They had simplychosen to remain silent to avoid panic-king the community.

This might be true. Alternatively, itmight be that it took a disaster asdramatic and visceral as Flixborough tomake obvious the attendant fire andexplosive risks inherent in chemicalprocess industries.

ReferencesKasperson RE. 1992, The Social Ampli-

fication of Risk: Progress in Developingan Integrative Framework in Krimsky, S.and Golding, D. (eds.) Social Theories ofRisk, Praeger Publishers, New York, pp153175.

Rachlin H. 1989,Judgement, Decision andChoicea Cognitive/Behavioural Syn-thesis, W.H. Freeman and Company, NewYork, pp 5861. (See also Manstead,

A.S.R. and Hewstone, M. (eds.) (1996) TheBlackwell Encyclopaedia of Social Psyc-hology, Blackwell, Oxford, pp 296300).

Fiedler K. and Schmid J., 1996, inManstead A.S.R. and Hewstone M. (eds.)

The Blackwell Encyclopaedia of SocialPsychology, Blackwell, Oxford, p 296.

Perrow C. 1984,Normal Accidents: Livingwith High-Risk Technologies, Basic Books,New York, pp 316317.

Jones R.C. 1989, Battle of Baglan Bay,The Independent Magazine, 28 October,p 34.

BP Chemicals International adver-tisement in Waller A. ,1968, Official Guideto Port Talbot: third edition, C. and D.Constaple Ltd, London, p 45.

Cranfield J. 1974, UK petrochemicals setfor massive growth, Petroleum Inter-national, July.

Napier DH. 1974, Background to aDisaster,Chemistry in Britain, Vol 10(7),p 239.

The Port Talbot Guardian, 1973,9 February.

The Port Talbot Guardian, 1973,23 February.

Kennett F. 1982, The Greatest Disastersof the Twentieth Century , MarshallCavendish, p 139.

The Port Talbot Guardian, 1974,1 March.

The Western Mail, 1977, Plant pollutioncomplaints are reduced to a trickle,24 November.

Bennett SA. 1996, Getting the public onboard, Chemistry in Britain, The RoyalSociety of Chemistry (RSC), London,Vol 32(4), April.

The Western Mail, 1974, Action groupformed to fight B.P. Pollution,19 February.Clammer J. 1974, Flixborough: a rural-

industrial disaster, New Society, 20 June.Scullion 1974, in Chartres J. and Osman

A., Angry residents will fight any attemptto rebuild death factory. The Times, 4 June.

Taylor HD. 1975, Flixborough: TheImplications for Management, KeithShipton Developments Ltd, p 10.

Kirkwood A. 1997, Flixborough Re-visited, The Safety and Health Practitioner,August, p 31.

Brown P. 1990, 16 years on, council actson Middlesborough ICI plant. Homesgo to avert Flixborough rerun, TheGuardian, 12 February.

Ellis 1974, in The Times, Rigorous andfar-reaching inquiry into chemicalfactory explosion, 4 June.

Nuttals Standard Dict ionary of th eEnglish Language, 1959, Frederick Warne,London.

Documents

Disasters as Heuristics a Case Study