The tanker Haven sinking off Arenzo, Gulf of Genoa, Italy
Source: Frank Spooner Pictures
THE PROBLEM |
Technology-related accidents are of major concern as sourcesof impacts on human health and the environment (Chapter 18). This concern arises from three interrelated characteristics: unpredictability of when and exactly how they will occur (and hence perceived lack of control), uncertainty over environmental pathways and impacts, and unforeseen interactions (human and technical) in the source facility. There is already substantial evidence of the short- and long-term level of human catastrophe which can result, for example, following the nuclear accident at Chernobyl and explosions such as at Los Alfaques and the release of dioxins at Seveso. The complex and possibly long-term damage to environmental resources (particularly soils and water) and dependent ecosystems is cause of increasing concern.
The causes of impacts from major accidents are distinguished by the fact that, although the source activities (eg, power generation, chemical processes and transportation) are planned and generally continuous, the hazards and environmental pressures associated with accidents are neither routine nor planned. Although statistics of past accidents provide an indication of what can be expected in the future, it is not possible to predict with absolute confidence either where or when an accident will happen. Combined with the significant uncertainty attached to the nature and magnitude of the resulting impacts, this justifies treating accidents as a significant source of societal 'risk' for the purpose of assessment and management. Risk, in this sense, can be characterised as the nature and magnitude of an undesired effect in relation to the probability of its occurring.
The industrial activities which give rise to these risks (primarily production and transport of chemicals) are increasing in intensity (see Chapter 20). In addition, interactions between human society and the natural environment are showing increasing signs of vulnerability to hazardous events (eg, the effects of Chernobyl and Seveso on agricultural activity, and the effects of hurricanes on the international insurance business). The prevention and control of such accidents to minimise the risks to within 'acceptable' limits can therefore be considered a priority. This chapter focuses on how appropriate goals can be set for a problem with such diverse elements, and how the risk can be managed by the parties concerned industrial operators, regulatory and planning authorities and the public. Nuclear accidents are dealt with separately at the end of the chapter.
GOALS: SETTING ACCEPTABLE RISK LEVELS |
Most attention, and therefore data collection and research, relating to accidents has traditionally focused on human health impacts. For this reason much of the following discussion describes this experience and looks at how it can serve as the basis for addressing environmental concerns. There are good reasons for doing this; the protection of human health and the environment from the consequences of accidents involve both complexity and uncertainty. This warrants the use of risk assessment and management for addressing both problems. Furthermore, it will increasingly be necessary to integrate human health and environmental issues into risk management of hazardous operations.
Human health risk |
Most polluting activities can be characterised in terms of pollutant discharge levels (eg, sulphur dioxide emissions) over different time periods for different operating conditions. By linking these to associated impact levels (eg, acid deposition, ambient air quality) and setting impact criteria (eg, critical loads, air quality objectives), appropriate source-oriented targets can be set in terms of maximum permissible pollutant discharge levels. Plant operators can plan to achieve these targets through a range of technical and management measures. Success in achieving the targets can also be monitored relatively easily by the operator and the relevant authorities.
Goals with regard to accidents have been developed on a broadly similar basis, but with some important differences. Firstly, the emphasis to date has been on limiting direct impacts on human health and society, rather than on the environment, to acceptable levels. Mortality is usually the most frequently used measure of impact, taking account of the fact that for every death resulting from an accident there will also be a number of less seriously injured casualties, both in the short term (typically from an explosion) and in the long term (for example from radiation exposure or inhalation of toxic fumes). Secondly, because of the uncertainties associated with the occurrence of accidents and their consequences, impact targets have been expressed in terms of the probabilities of a range of impacts occurring, each with a different magnitude. This allows 'acceptable' risk to be defined as it affects both individuals and society as a whole.
Individual risk is usually defined as the probability per year that a certain individual will die as a result of an accident associated with a specific industrial plant or activity. This is because goals for the protection of individuals can be quantified as the maximum acceptable individual risk associated with a single plant. It is also possible to identify group risk, ie, the probability that a certain group of people (usually defined by size) outside an industrial facility will die due to an accident on site. Thus, goals for the protection of society can be quantified in terms of group risks by specifying the maximum acceptable probabilities per year of different numbers of people dying as a result of accidents at a particular plant.
Much of the initiative relating to identification of acceptable risks and risk assessment for planning purposes was taken in Europe by the UK Health and Safety Executive and the Dutch government. The cumulative risk assessment work for chemical complexes on Canvey Island in England (HSE, 1978; 1981) was a landmark in this respect, and subsequent investigations (HSE, 1989a; 1989b) have focused on development of appropriate risk criteria for landuse planning and their use with quantified risk assessments as an input to decision making (see below). The societal risk criteria adopted by the Dutch province of Groningen (Province of Groningen, 1979) laid the basis for Dutch national limits (VROM, 1990) and these are used as an example in Box 30A.
Risk criteria have traditionally been developed for application to potentially hazardous installations. However, they can in principle be applied to transport activities in order to limit the risks to which different groups located along transport routes should be exposed. This in turn would allow the application of risk assessment to route planning for the transport of hazardous substances. To allow for the exposure of population groups to hazards from many stationary and mobile sources, it is also possible to establish limits for cumulative risks. For example, the Dutch government has set the maximum permissible level of individual risk at 1 in 105 per year, ie, ten times the maximum level due to any one source.
The importance of risk perception |
Setting risk targets for planning purposes involves three steps: deciding on a tolerable, or broadly acceptable, level of individual risk, deciding on the acceptability of accidents that affect numbers of people (societal risk), taking into account the possible aversion to large-scale and/or disruptive accidents, and a consideration of the achievability of the criteria.
Actual risks of different activities can be estimated in quantitative terms using a range of scientific approaches. However, setting appropriate criteria for acceptable or unacceptable risk is essentially a political issue. The human perception of risk is one of the most important factors which the appropriate government authority needs to consider. At an individual level this depends on whether the risk is voluntary or involuntary, whether the individual perceives he or she can control the risk, the expected benefits of the activity in question and the nature and timing of possible deleterious effects.
There is growing evidence (Reed et al, 1992) that individual risk levels in the order of 1 in 1 000 000 and below do not generally give cause for concern and that levels in the order of 1 in 10 000 and above will often result in significant concern and a desire to see the risk reduced. Within this band the 'acceptability' of a given risk often varies, depending on the nature of the cause and the individual or group affected. The more an individual feels in control of a situation the more likely the risk will be judged acceptable (eg, driving a car along a motorway). If the exposure to the risk is considered involuntary, and/or the individual feels it is beyond his or her control, a similar level of risk is more likely to be judged unacceptable. For comparative purposes, Table 30.1 gives the mortality rate of a range of activities and occurrences in The Netherlands (VROM, 1990).
Notwithstanding this, the community's perception of risk will not be sufficiently wide-ranging to form the basis of acceptable risk levels. This must be the responsibility of the relevant government authority, which is in a position to take account of other important factors such as direct and indirect economic impacts (eg, through loss of local workforce or loss of confidence of investors). The feasibility of achieving different risk criteria is also an important factor to consider, if they are to be used as the basis of planning permission for hazardous, yet economically or socially important facilities (see below). A balance needs to be achieved. Thus, whereas the Dutch government has set an unacceptable group risk level of 1 in 2 billion (109) per year for 1000 deaths (see Box 30A), the Hong Kong limit is 1 in 2 million (106), reflecting the higher population density of Hong Kong.
It is important to appreciate that risk perception is subjective at the individual level and also, on a bigger scale, at societal level, where politicians must make judgmental decisions on behalf of the society they represent. For this reason, it is difficult to determine international goals for societal risk. The appropriate levels of acceptability or unacceptability are likely to vary on a country and even a regional basis.
Risk to the environment |
Environmental impacts of accidents are not included in standard statistics on environmental quality. As discussed in Chapter 18, accidents should therefore be seen as an additional threat to the environment. However, the effects of accidents on the natural environment are usually complex and possibly far-reaching, and cannot be assessed easily by reference to a single indicator or parameter.
The impact of an accident usually has a widespread effect, for example, by damaging different parts of an affected ecosystem successively, often via the food-chains. Several attempts have been made to model the environmental damage of accidents and to quantify it. An example is the Natural Resources Damage Assessment Model for Coastal and Marine Environments used in the USA and generally known by the acronym NRDAM/CME. It was developed in 1987 to assess damages to natural resources covered by the Comprehensive Environmental Response, Compensation and Liability Act (CERCLA) of 1980.
The NRDAM/CME contains submodels for physical fate, biological effects and economic damage, which are supplemented by a series of databases containing physical, chemical and biological information. The biological submodel uses lethal toxicity data for various compounds to assess the mortality of both adult and juvenile animals in the receiving habitat. However, there is a particular lack of information on the long-term effects of accidents on the environment. This is due particularly to the paucity of baseline information available. It is virtually impossible to assess the long-term ecological damage from a spill of toxic chemicals into a river if the original state of the affected ecosystem had not been previously examined.
Considering the complexity of the problem and also the substantial uncertainties involved, it is likely that the concept of environmental risk could be a potentially useful tool for setting goals in relation to the environmental impacts of major accidents. This is a relatively new concept and undeveloped compared with application of risk to human health impacts. However, it is now receiving serious attention, for example, in the Dutch Environmental Policy Plan. This provides for the setting of ecosystem risk limits for large accidents aimed at safeguarding populations of plants and animals rather than preventing damage to individual species. This might involve the setting of ecosystem protection levels (expressed as the percentage of species in an ecosystem) to act as maximum permissible impact levels. Setting ecosystem risk limits will require the development of reliable methods for assessing such risks in order that relevant standards can be set. Initially attention is being focused on the use of 'EC50' and 'LC50' values to allow the characterisation of disasters in relation to the likely exceedance of the critical limit for the most sensitive test organism or the loss of half a population, respectively.
A similar risk approach might be feasible for the protection of environmental functions of land and waters affected by major accidents (eg, the recreational value of natural areas, agricultural value of soils, use of surface waters for supply of drinking water). To some extent this might be assisted by expressing damage in monetary terms, though this has yet to be fully investigated and would be subject to the traditional judgmental problems associated with monetarisation of environmental resources.
MANAGING THE RISKS |
Basic elements |
The human health and environmental risks associated with accidents vary according to the location and inherent safety of industrial installations or other hazardous activities (eg, transport of dangerous substances over land or by sea). This reflects the variation of the two basic elements of risk: the magnitude of the consequences of an accident (strongly dependent on location) and the probability that these will occur (mainly dependent on the safety of the operation). Accident risk management therefore focuses on these two variables.
In practice, risk management requires a combination of source-oriented and impact-oriented measures, involving the operator of the hazardous activity, the relevant public authorities and services and members of the public. The basic elements are outlined below.
Source safety |
The owners and managers of potentially hazardous operations are in a position to determine how likely it is that a particular accident, of a given type and size, will occur. This results from their influence over design technologies, management systems and human performance during normal and abnormal operations. The system design of a chemical process, the selection of appropriate plant, the incorporation of adequate safety measures into both automatic and manual (human) procedures and appropriate staff training at all levels will all contribute to reducing the probability of abnormal operating conditions occurring (eg, excessive operating temperatures or pressures). They will also help ensure that the magnitude of the immediate consequences of any such occurrence (eg, release of a toxic gas) will be minimised.
Since the 1950s the safety of industrial operations has been progressively improved by the introduction of a series of techniques relating to design standards, plant inspections, technical safety and safety audits (eg, HAZOPs). HAZOP (Hazard and Operability) studies are often undertaken on industrial sites to assess their safety and identify potential hazards. Under the group of EC Directives on Major Hazards (82/501/EEC, 87/216/EEC, 88/610/EEC), commonly referred to as the Seveso Directive, certain sites with an inherently high hazard potential need to prepare a safety study and have to do a HAZOP study.
Lately, industry has started to pay attention to 'human factors'. This relates not just to individual performance in doing a job or responding in an emergency situation, but also, and more importantly, to management, organisational and system aspects. It is now recognised that organisational and management failures are often an important contributory factor in major accidents, such as the Flixborough explosion in 1974 (Parker, 1975), the Piper Alpha offshore disaster in 1988 (Cullen, 1990) and Bhopal in 1984 (Bellamy, 1986). The report of the Zeebrugge ferry disaster (HMSO, 1987) identified a serious failure to appreciate responsibility for ship safety from director level downwards. These types of observations are leading to the development of integrated safety management systems and audit tools (eg, see Bellamy, 1986), which focus on all aspects of human factors in risk assessment to complement the existing technical focus.
However good such integrated approaches might be for assessing and reducing the impact from industrial accidents, any consideration of this topic cannot be separated from the socio-economic and political dimensions. For any safety management system to be wholly effective, two political/economic questions need to be addressed (Bourdeau and Green, 1989):
In the former case, particularly relevant to the countries of Central and Eastern Europe, there is an acute scientific/technological problem of assessing the hazards and quantifying the risks, especially where the plant has been modified over time with inadequate documentation of the changes from the original design. In both cases, the ultimate resolution of the question is likely to be encouraged by financial inducements by government.
Emergency response planning |
In some cases, appropriate measures on site or at the scene of an accident will ensure that pollutant releases are prevented or minimised and that significant impacts are avoided. In general, however, once an accident has happened, some level of impact is most likely, requiring clean-up measures in the case of environmental impacts, and possibly evacuation of the local population in the case of potential human health impacts. The scope for limiting impacts through response measures varies significantly, depending on whether the effect and impacts are instantaneous and simultaneous (eg, as with an LPG explosion following a road tanker crash), or whether there are delays between the time of the accident, the release of the pollutant and the impacts themselves (eg, as with an oil spill at sea).
Limiting impacts after an accident is usually very expensive (see Table 30.2) and, in the case of environmental impacts, rarely successful. For these reasons emphasis is put on prevention as the most cost-effective and efficient means of risk management. This involves both source safety measures (see above) and landuse and transport planning measures (see below).
Despite the necessary emphasis on prevention, it is recognised that accidents will continue to happen, often with considerable potential for disaster. In this context important initiatives are under way to coordinate international approaches to short-term accident response and follow-up activities with regard to clean-up of the environment and longer-term human health effects (eg, the UNECE convention on Transboundary Effects of Industrial Accidents, the UN Centre for Urgent Environmental Assistance).
The aim of an emergency or contingency plan is to localise any accident, to contain it and to reduce the harmful effects on health and the environment by rapid and responsible action. It should provide the necessary guidance to allow for a flexible response to a range of possible circumstances. Close coordination between the site operators and the local authorities and the strict definition of responsibilities in the case of an accident are essential (eg, Bourdeau and Green, 1989). In the EU, the Seveso Directive requires preparation of emergency plans and liaison between operators, local authorities and emergency services.
Emergency response plans must necessarily involve operators, local authorities and emergency services and members of the public who may be affected by an accident. They must start with an early warning of potential problems and provision of all relevant information to the relevant public authorities by the owner and/or manager of a hazardous operation. If the emergency plan is significantly to minimise the adverse impacts of any accidents, it needs to be continuously reviewed, taking into account any previous experiences and both the present as well as the future state of the industrial systems to which it is to be applied. Industrial audits serve to support the review of emergency plans.
As with other areas of environmental management, the transboundary aspect of this problem is becoming more important, as demonstrated by accidents such as Chernobyl and the Sandoz plant in Basle. This is emphasised by the UNECE (UNECE, 1993) along with the important role of industrial associations, such as CEFIC, in implementing programmes concerning awareness and preparedness for emergencies at local level (APELL) (UNEP, 1988).
Landuse and transport planning |
Regardless of any steps taken by the owner and/or manager of an operation to ensure its safety and reduce the probability of a hazardous event occurring, there is an equal responsibility on the part of public authorities to ensure that potentially hazardous installations or traffic are separated adequately from population centres and environmentally sensitive areas. This will determine, ultimately, whether a human-made disaster is possible. As Reed et al (1992) have indicated, this underlies the very different casualty figures associated with the Flixborough explosion in the UK in 1974 and the isocyanate release at Bhopal, India, in 1984.
There are two complementary aspects to this element of risk management:
If an operation is inherently highly hazardous, the decision on where to site it is of great importance. Environmentally sensitive as well as heavily populated areas should be avoided. There are similarities between this aspect of rural and urban planning and that concerned with protection of human health and the environment from routine industrial activities, involving the use of industrial zoning and environmental impact assessments (eg, the EC Directive on environmental assessment of projects, 85/337/EEC). This is an important feature of the Dutch environmental policy plan and is gaining favour elsewhere in Europe and beyond (eg, see Reed et al, 1992).
Within the context of landuse planning, industrial zones are often designated to avoid conflicts of use by predetermining the siting of industrial installations. Nevertheless, it is useful to quantify the off-site risks, particularly with regard to the cumulative risks associated with all facilities located in one area. Requiring new facilities to be subject to an environmental assessment will not ensure that cumulative environmental impacts will be avoided in an area of intensive development. For this reason the concept of strategic environmental assessment is gaining popularity as a tool for landuse planning. Similarly, managing the cumulative risks associated with accidents requires a more strategic approach than considering hazardous operations only on a case-by-case basis.
As for emergency response planning and environmental assessment of new projects, this area of risk management is likely to require more of an international focus in coming years, to take account of facilities developed close to national borders and particularly with regard to transport. In the light of recent oil spills in European waters (for example, the Braer off the Shetlands) the European Commission adopted a Communication entitled A Common Policy on Safe Seas (CEC, 1993). This proposes a package of measures intended to reduce the potential for major accidents at sea and the impacts which may arise as a result. In particular it identifies traffic restrictions in environmentally sensitive areas as a priority issue for international action. From the source safety point of view it makes a number of recommendations relating to ship inspection and training.
Integrated risk assessment and management |
Integrated risk assessment and management attempts to provide a method for considering all the elements of risk management outlined above in developing the best overall means of ensuring that relevant risk criteria are satisfied for a particular installation or hazardous operation. The concept is based upon hazard identification, consequence/impact analysis, risk quantification and assessment against criteria and identification of mitigation options (either on-site or off-site).
Risk quantification is a technique which can be used to assess the implications of various siting and technical options for new developments. Frequencies of the initiating events identified during the hazard analysis (eg, a HAZOP) are normally quantified from the synthesis of operating data or equipment failure statistics (including those for the failure of safety barriers in the system). Logic diagrams, or 'fault trees', are then used to represent and quantify the sequence of events leading to a potentially hazardous outcome (eg, explosions, fires, oil spills, gas releases). Human performance data are incorporated into these analyses with an increasing degree of sophistication. Consequence models are used to assess the physical effects of the range of accidents. The consequences and frequencies for all events are then summed to give appropriate measures of individual or societal risk, which can be compared with the criteria.
There are inevitably inherent analytical uncertainties associated with this, for example, as a result of the imperfections in methodologies to model the behaviour of released chemicals under different conditions. These can be treated by undertaking probabilistic risk assessments. These are useful for understanding the confidence which should be attached to modelling outputs, but raise considerable problems as to how uncertain results should be compared with uncertain criteria.
Major industrial accident statistics in the EU are continually collected by MARS (Major Accident Reporting System) at the European Commission's Joint Research Centre ( see Chapter 18). The availability of accident statistics which can be used as inputs to quantitative risk assessments is a key factor in the capability to reduce risks through safety management. The Seveso Directive puts an obligation on Member States to exchange information on major accidents in order to improve the basis of accident prevention and to ameliorate the consequences of such accidents.
An emerging aspect of this is the need to consider to what extent environmental impacts are associated with different types of accident, and therefore require different measures, from those relevant to direct human health impacts, for prevention or mitigation. One approach that is intuitively attractive is the preference for reduction of hazard rather than of exposure or impact: for example, reduce hazardous material inventories or improve safety systems rather than depend on locating plants away from housing or sensitive environments. It is likely that in future more explicit consideration of the environmental impacts of major accidents will feature in the requirements of international safety regulations, such as the Seveso Directive (which is being revised CEC, 1994), and in combined health, safety and environment audits and management systems for industrial companies.
NUCLEAR ACCIDENTS |
The use of radionuclides for industrial, medical and research purposes, as well as the operation of nuclear facilities, has to satisfy specific regulatory measures and nuclear legislation to ensure the protection of humans and the environment against the dangers of ionising radiation (see Chapter 16). Basic radiological protection requirements have been established worldwide and are kept under permanent review by the International Commission for Radiological Protection (ICRP). They are embodied in most national legislations and in EC legislation which lays down legally binding basic standards for the protection of the health of workers and the general public against the dangers arising from ionising radiation. The implementation of these standards is within the competence of each individual EU Member State.
Nuclear safety results from the existence and correct implementation of such a regulatory framework, as well as from adequate technologies and techniques, and from appropriate training and dedication of personnel. The opening of Central and Eastern Europe, and the break-up of the USSR, has disclosed marked differences compared with Western Europe in nuclear safety matters, and significant weaknesses have been identified. The main problem is to remedy past mismanagement and accidental environmental pollution, as well as preventing and mitigating possible new accidents.
Although the Chernobyl accident (see Box 18E) has not provided in itself any strong specific lessons of relevance to the designs and to the regulatory framework adopted for nuclear plants outside the former USSR, the very fact that it happened in a technologically advanced country, with all its heavy consequences on people and the environment, has posed a new challenge in Europe.
In essence the problem of nuclear accidents is how to deal with events which have potentially high consequences but which are expected to occur with only extremely low probabilities. Can further measures be taken to make their occurrence, for all intents and purposes, 'impossible', or can their potential consequences be reduced by, for example, improving containment of the plant?
Causes |
A summary of the major nuclear accidents in Europe and their causes and consequences is presented in Chapter 18. Incidents which can occur in nuclear facilities may result in the release of radioactive materials (all plants) or primarily chemotoxic materials (especially uranium hexafluoride from enrichment plants). Possible underlying causes of such events include:
Such underlying causes may appear in combination and can give rise to intermediate situations such as nuclear criticality events (especially in chemical process plants) or loss of reactor cooling, which in turn lead to the release taking place.
Incidents are frequently related to handling radioactive liquids, sludge deposition in piping, storage of combustible or/and pyrophoric material, and the imperfect monitoring and control of these operations. Deficiencies in plant design and operation, mishandling of radioactive materials, human error, mismanagement and insufficient training, and neglect of appropriate regulatory frameworks, responsibilities and procedures are the main possible causes of abnormal events. In fact, according to the level of 'nuclear safety culture' of the country under consideration, only one or a few of them may be dominant.
Most abnormal events which have occurred at nuclear power plants in Western Europe have resulted from human error during operation of the plants. The operational and safety status of reactors of Soviet design in Central and Eastern Europe have been able to be assessed more precisely since the political changes there in the early 1990s. Nearly all the nuclear power plants of Soviet design show significant technical and operational weaknesses in various degrees according to the type of plant and the country under consideration. In the plants of the older types (RBMK and VVER 230 types) the design of the circuits and components is not satisfactory in terms of safety. Deficiencies in quality assurance and control and maintenance have been observed, as well as a scarcity of qualified and dedicated personnel. For some systems and some situations, the designers relied essentially on the operator's intervention, with a high probability of a crucial human error as a consequence.
Objectives |
Since the end of the 1980s it has become fully accepted that, whereas severe accidents (ie, those having large-scale effects on humans and the environment) might occur in any nuclear power plant (even if the probability of occurrence is very low), the implementation of, and the priority given to, nuclear safety and environmental protection have been dealt with in a much less stringent manner in the former USSR and associated countries than in Western Europe. Two main objectives are therefore being pursued:
Strategies |
Strategies for enhancing reactor safety are being carried out inseveral nuclear countries, mainly by means of research and development programmes. These are concerned with advanced reactor design, from the 'evolutionary', based on present experience, to the 'revolutionary', calling for new concepts. Virtually no substantial effort in advanced reactor design is occurring in isolation in a single country. Within the European Community, the cooperation between Framatome (F) and Siemens (D) on the future European Pressurised Water Reactor (EPR) is an example of evolutionary design. The EPR project will also provide a solid basis for the harmonisation of French and German safety requirements and possibly promote an enlarged European cooperation. Joint research in Europe should provide a good basis for future nuclear power plants and lead to the required common understanding of what should comprise basic safety features.
The new relationship established in the early 1990s between the countries of Central and Eastern Europe and the rest of the continent has created favourable conditions for the further development of the international dimension of nuclear safety. The International Conference on the Safety of Nuclear Power, convened by the IAEA in 1991 following a proposal by the European Commission, led to the conclusion that there was a need for an integrated international approach to all aspects of nuclear safety, and for the preparation of a formal framework. The international Convention on Nuclear Safety, which has arisen from this, aims mainly at achieving worldwide uniform, and higher, levels of safety at nuclear power plants. The convention was adopted on 17 June 1994 and will enter into force after ratification.
The specific problems of nuclear safety encountered in the countries of Central and Eastern Europe are taken into account by a strategy of assistance carried out by a group of 24 countries, including Western European countries, Canada, the US and Japan, with a financial commitment from the EC TACIS and PHARE programmes, between 1990 and 1994, of ECU416 million.
If, despite these measures, a nuclear emergency does occur, fast, reliable and appropriate information is required. In order to provide this at international level, both the International Atomic Energy Agency (IAEA) and the European Commission have set up special, mutually compatible communications systems ('Emercom' and 'ECURIE' respectively). The aim of these systems is to transmit urgent radiological information between the IAEA, the European Commission and their respective Member States. Exercises are regularly held between them to keep the systems prepared and operational.