Over the past nine years HSE has been engaging with various stakeholders on the issues arising from societal risk levels in the vicinity of major hazards, including a consultation exercise. The Seminar is part of the engagement process and focuses on the technical issues associated with the estimation of societal risk levels for assisting with the provision of LUP advice to Planning Authorities and decisions on measures at sites. Prior to the Seminar HSE published a research report (RR703) summarising the background and issues; a critique by the IChemE on HSE’s approach to societal risk and HSE’s response to the critique. The Seminar programme and speaker presentations are available on the HSE societal risk web page. This short note captures the discussion points raised and summarises planned work.
The consultation document CD212 explained that risks from on-shore major hazards include the contribution from catastrophic accidents with potential for tens and possibly many more fatalities (on and off site). The likelihood of such major accidents is extremely low owing to the prevention and control measures put in place by companies, as required by the COMAH1999 regulations. However, it is not possible to eliminate all potential major accidents entirely and there remains a very small ‘residual risk’ of significant harm to workers and neighbouring population. An estimate of the associated ‘societal risk’ provides information that can be used by decision makers to make judgements on:
These judgements are often very difficult and information on safety is one factor amongst other socio-economic considerations. Consequently, the overall aim of HSE’s developmental work on societal risk is to ensure that suitable and sufficient information is available to decision makers (whether they be local planning committees, HSE inspectors, the operators of hazardous installations or other key stakeholders) so that they can make sound judgements. The key objectives of the work programme are:
The main discussion points and related topics are summarised below with some comments by HSE.
There were a number of requests for continued engagement with Stakeholders. HSE is committed to involving Stakeholders in developing and finalising its approach to societal risk and concern. HSE plans to hold a follow up seminar in autumn 2009 and a wider stakeholder seminar early in 2010 to coincide with consultation on the updated risk criteria document for decisions in land use planning. Prior to the 2010 seminar, we will publish some of the HSE research relating to the technical issues associated with the assessment of societal risk.
During 2008 and into 2009, HSE has undertaken a programme of work to address some of the key technical issues (23 in total and described in RR703 (link)) in the estimation and assessment of societal risk. The Technical Advisory Group, whose terms of reference are described in RR703, is providing HSE with views on which of these issues should be taken forward. In some cases the issues will be clarified by research and advice. Others will necessitate the adoption of a policy position; the rationale behind any new policy will be made clear and published.
The criteria and approach on which HSE bases its current LUP advice were described in the HSE publication known as the 1989 “Risk Criteria Document” (Risk criteria for land-use planning in the vicinity of major industrial hazards). This document is being updated to take account of changes in methods and to define the approach HSE will adopt to assess societal risk and advise LPAs. Stakeholders will be offered an opportunity to comment on the final draft prior to revision and publication.
HSE’s current LUP advice to local planning authorities on new development proposals does not take account of local societal risk levels; it is based predominantly on individual risk and case societal risk codified in the PADHI methodology. Following the outcome of the Consultation exercise, Ministers agreed that the current advice should be supplemented by more explicit consideration of societal risk at about 55 sites where HSE has already identified that supplementary policies are needed. The companies and LPAs involved have been notified. The existing LUP advice system, including use of PADHI, will continue for all sites. It is envisaged that most advice on local societal risk will be provided at the strategic plan stage but there will remain a need to advise on a case by case planning application basis. Where PADHI indicates ‘advise against’ (AA) there will be no additional consideration of societal risk. Where PADHI does not AA for the top 50+ sites, additional advice based on societal risk will be given, essentially a societal risk overlay. There may be a need to modify the way PADHI deals with development in the IZ as the contribution to the expected number of fatalities per year (EV) from a site is significantly greater than for the same size of development in the MZ or OZ. This is particularly true where the hazard is omni-directional.
We are working towards greater consistency but this will take some time. For providing LUP advice the inconsistencies may not be that significant as there is a degree of correlation between risk of dangerous dose and risk of death. The new risk criteria document will improve transparency here and justify overlaying societal risk onto the current IR approach.
The Local Planning Authority focus group will advise upon how information on societal risk will be considered in land use planning decision making. They will take account of the outcome of three societal risk based LUP trials which involve LPAs and industry. The Trials are providing information to help decide how advice is best delivered and considered by decision makers.
These involve 3 volunteer planning authorities and will provide an opportunity to test some of the issues discussed at the seminar. The trials are not for producing formal societal risk information for the volunteer LPAs: the aims of the trials are for learning lessons and getting feedback from planners and industry to help develop a workable societal risk decision framework for the UK. The first trials will be completed in May; the remaining two will be completed over the summer.
The Department of Communities and Local Government (DCLG) have engaged consultants to look at key aspects of the LUP advice system. The study will concentrate on how various aspects of the regulations work e.g. Hazardous Substances Consents, not on how risks are assessed. For LUP, HSE’s intention is to use consented substances and quantities as the basis (rather than actual) for the risk calculations supporting LUP advice due to the long-term nature of development planning. The substance(s) that are dominant in setting the LUP zones will be used in the calculations. Where possible, it is preferable that entitlements are reduced so that CDs can be reduced. Smaller CDs brings benefits to both sides: LPAs – reduced constraint on development; reduced consultation with HSE; better for off-site emergency planning etc; Industry – better Public Relations, better understanding from LPAs why they should better control land development close in. HSE will discuss with DCLG and ask them to confirm that the application process to reduce entitlements is as straightforward as a company applying on Form 1, with a covering note explaining a reduction (in quantities or substances or both) is being requested.
HSE has no plans to have one societal risk assessment for both LUP and COMAH enforcement. HSE’s role is to do the assessment for LUP advice and industry’s role is to do the assessment for demonstrating ‘all measures necessary’ under COMAH. The assessments have different purposes. LUP is looking to the future over a time frame of many years, assuming that the residual risk remains ALARP; while COMAH assessments have a shorter time scale and focus on the adequacy of risk reduction measures for meeting the requirements of the regulations. Providing societal risk information to LPAs should raise their understanding of the implications of granting development-permission close in. Research is ongoing to provide insights into the implications of land development within the societal risk zone e.g. how much land development is needed to tip the cost/benefit balance? More stringent development control advice in the inner zone could be a future possibility
We have been focussing on informing decisions on land development around sites but we are also working on taking SR explicitly into account when evaluating risk reduction measures on site. Research has started on disproportion factors and a framework for taking SR into account throughout the risk assessment process and when evaluating risk reduction measures on site. We are not expecting to suddenly find risk reduction measures that need putting in place but we need a system in place to provide a challenge function to check what more could be done and if it’s reasonably practicable, why it isn’t being done.
The initial view from the Technical Advisory Group is that all affected in an incident should be included in societal risk calculations. So yes, onsite could affect advice for off site development. However this is more of a policy decision than a technical decision. The trials should provide more insight into this.
Currently we are looking at sites individually. Where domino effects may be relevant they will be considered in COMAH Reports and we will be considering whether there are implications for LUP assessments.
The likelihood of this occurring is likely to be greater for toxic substances, as the area in which to monitor for societal risk may be double the size of the current consultation distance (CD). Where CDs are based on flammable hazards, the CD may well be the distance to which societal risk may be monitored. Neighbouring authorities will have to be kept informed, as they are with current land-use-planning zones. The trials should provide relevant information to help finalise the arrangements.
Currently we are considering fixed sites – pipelines raise additional complications. We are aware that the Institute of Gas Engineers have societal risk criteria and we will be considering these in due course
The lead on balancing our advice against other factors is with DCLG. (See also paper at http://www.hse.gov.uk/economics/research.htm)
Land development in the UK predates the current arrangements which were initially introduced in 1972. There is therefore a legacy of land development in preceding years which, in some situations, is far from ideal. Despite this, we will not retrospectively apply LUP criteria. An aspirational principle is to not increase societal risk where levels are already relatively high, and to reduce it over time (as opportunities arise). In the future, HSE will endeavour to be involved in strategic planning by providing advice on framework plans. Deviations from the plan (windfall development) will be handled under arrangements similar to those currently in place.
Interestingly we have had the opposite views expressed and some companies use them with their own criteria for risk management purposes. F/N plots and the information needed to derive them provide useful information such as the likelihood of realising particular incident severities. We find such information useful in discussions with industry. The information used to derive such plots is also necessary to estimate the expectation value (EV) i.e. the equivalent number of predicted fatalities per year. The EV, although useful for cost benefit considerations, has limitations. For example, it cannot convey information about the scale of potential accidents. HSE is looking into other metrics for representing societal risk, such as the EV density (hot spot maps) and the Dutch societal risk on a map approach. One of the main criticisms levelled at FN plots is the choice of criteria for making judgements. . As Ball and Floyd (1998) point out, “Societal risk criteria should not… be viewed as more than broad indicators of a desirable objective, with many other, non-technical factors needing to be weighed in any final decision. …” They also stressed that criteria (or guidelines) selected need to be done so with careful consideration of the nature of harm, those at risk and relevance of known anchor points.
Any use HSE makes of FN plots and associated criteria for LUP planning purposes will be justified in the new risk criteria document.
It is often argued that risk assessment methods and data give pessimistic estimates of the expected numbers of fatalities and those methods and data should be modified to match history. It is true that over the past 30 years or so the EV for off-site fatalities for COMAH sites would be greater than realised, but 30 years is a very small time frame when some accident scenarios with potential for very severe consequences have a frequency of the order of ~ 1in a million years. A 30 year period cannot represent the spectrum of possibilities for the UK situation. Moreover HSE’s approach to LUP assessments assumes that all measures necessary are already in place i.e. LUP assessments are based on the concept of residual risk i.e. the risk that remains after risks have been reduced ALARP.
These are complex issues beyond the scope of this Seminar. The risk from individual facilities and establishments is a COMAH matter. Management of risk through land development is a matter for DCLG and we will reflect your concerns to them and provide them with suitable and sufficient risk information to meet their needs. One of the LUP societal risk trials is a multi-site and multi-LPA situation so we should gain some insight from discussing the trial results with the companies and the LPAs involved.
Although this is a societal risk issue HSE is focussing on fixed installations. In the UK transport risks are a matter to the Department for Transport.
The questioner alluded to an EU study on risk assessment comparisons which showed that the results depended more on the consultants than the models and data used. The Dutch seek to avoid this problem by using the Purple book etc; HSE’s approach is broadly similar. HSE strives for consistency and validity by funding research to provide valid methods and data and by conducting its own assessments based on the published approaches. The assessment criteria HSE adopts are chosen in the light of the strengths and weaknesses of the assessment package. In the case of LUP advice based on societal risk, the new risk criteria document will provide the justification for consistency and validity.
Aversion and willingness to pay arguments was a significant feature of the discussion at the seminar. Willingness to pay argument is often used to argue against the use of aversion or gross disproportion in cost benefit analysis. Such studies put the value of preventing a fatality (VPF) at ~ £1.6 million. Willingness to pay studies are invariably based on an individual’s perception of the risks and benefits for that individual. It is therefore unlikely to matter whether he or she is killed at the same time as many other people, for example in an aircraft crash. However, the aeroplane manufacturer and the industry regulator may seek to make improvements that put a greater value on the prevention of a fatality, especially if a similar make of plane had crashed recently. Having said that, studies funded by the railway industry (Covey et al. 2008) support the view that there is no need to allow for aversion to multiple fatality events, any allowance is at most a factor of about 1.2 times the VPF. These studies are however based on the views of individuals. As Ball and Floyd (1998) point out “there is very little evidence for differential risk aversion by the public where this is based upon numbers of fatalities. However, there is a hint at least from the French study, and also the Dutch societal risk criteria which emanated from the Dutch political system …. that elected officials, senior administrators and risk managers may be averse to major accidents. That this group should be averse to high consequence accidents is hardly surprising…”
Evidence of scale aversion can be seen in the response to large-scale incidents. Consider the Piper Alpha incident: A sub-sea isolation valve would have prevented the disaster. From a safety point of view the costs may be disproportionate ~ 80 times the cost of preventing a fatality (see Fleishman and Hogh 1989). However, after the incident such valves were fitted to similar platforms probably because of companies’ aversion to the risks posed to their business. A more recent example is Buncefield where there were no fatalities (on or offsite), yet there was considerable public outcry because of the scale of the damage and what could have happened had the incident occurred at a different time. Many recommendations and improvements followed. The level of societal concern was such that a new LUP zone (a development proximity zone (DPZ) in which developments other than those not normally occupied are advised against) was introduced for similar sites, recommendations for an overhaul of the planning system were made, and HSE’s programme of work on societal risk accelerated, etc. The Buncefield incident illustrates that it is societal concern, rather than societal risk that is driving approaches to LUP.
Evidence of greater attention to industries with catastrophic potential (a form of scale aversion) might also be seen from the way HSE allocates its resources. Over the last 30 years the proportion of fatal accidents for the Chemical Industry has been less than 1 % of the total number of fatal accidents reported to HSE. Also in the same time frame no one off-site has been killed as a result of a MH accident. Yet the Chemical Industries Division of HSE is allocated approximately 5% of the HSE’s budget. The justification for this disproportion could reflect the concern of Government to the potential catastrophic consequences of large scale accidents.
These examples underline that aversion depends on the view of the stakeholder. Indeed, we are aware that one company has built aversion into its criteria for judging the significance of societal risk posed by each of its establishments.
Aversion is one aspect of societal concern; others include societal risk, the maximum number of fatalities possible, the proportion of vulnerable people involved etc (see R2P2)
Any large scale incident will generate public outrage. The magnitude of that outrage is likely to depend on what had been done to help the public to understand the risks and the risk management process. It was felt that more needed to be done to improve the publics’ understanding of the risks posed by MH sites and the benefits derived.
Over recent years it has become increasingly apparent to Government that disproportionate attitudes towards risk are exerting an unhealthy influence on policy-making. Policy-makers themselves have been acutely aware of this, and there has been ongoing debate at the highest levels on how to address the challenge of taking a more proportionate approach to risk. The Risk and Regulation Advisory Council was set up following the Better Regulation Commission’s 2006 ‘Risk Report’, which showed how poor responses to risk could lead to poor regulation. In response to a request from the Prime Minister, the Better Regulation Commission’s 2008 Report on ‘Public Risk’ then set out the structure and approach of the Risk and Regulation Advisory Council, designed to embed a fuller understanding of public risk in policy-making, even when facing pressure to react to events. With the establishment of the Risk and Regulation Advisory Council, the Better Regulation Commission was closed. HID is keen to learn any lessons from the work of the RRAC and apply them to its COMAH related work so that a sensible balance is maintained between the risks and the societal concern. The LUP advice HSE gives is part of national policy. It is seen as a cost effective way of avoiding future societal risk problems that have arisen through the legacy of land development prior to the current policy.
Ball, D.J.; Floyd, P.J (1998). Societal risks, Final Report, commissioned by the Health and Safety Executive, United Kingdom
Covey J; Robinson A; Jones-Lee M; Loomes G; (2008) Assessment of the Value of preventing a fatality. Rail Safety and Standards Board, RSSB T616, Oxford Risk.
Fleishman AB and Hogh MS (1989) The use of CBA in evaluating the acceptability of industrial risk: An illustrative case study. Sixth International Symposium on Loss Prevention and Safety Promotion in the Process Industries, Oslo, 1989.