Simulating disaster sharpens judgement of crisis managers

Threat events such as natural and man-made disasters can result in significant environmental, social and economic disruption. Hence there is the requirement, by law, in the Civil Contingencies Act 2004 for a well-organised and prepared response to mitigate the effects of such disasters.

There are a number of tools and methods to ensure that the correct response is undertaken to preserve safety of human life or to protect the environment. Threat events, ranging from high-impact, low-probability episodes such as chemical, biological, radiological, nuclear (CBRN) attacks to low-impact, high-probability events such as a tanker incident on a motorway, are currently managed or tested through the use of table-top, live and synthetic environment exercises.

What is vitally important for crisis managers to understand is the way in which people perceive such incidents. Society has become immune to certain types of events, but fearful of others. For example, random sniper attacks caused fear in the US city of Washington, DC, despite the fact that fatalities from car accidents or gun crime were daily events.

Similarly, the Hatfield train crash in the UK in 2000 caused widespread economic damage. It affected timetables, train information phonelines and journey times, and many working days were lost because of delayed train journeys and the fact that people preferred not to travel. Understanding the wider picture is all-important, but many traditional methods of training for threat events have inherent problems, making managing incidents all the more difficult.

Table-top exercises are a common method used to test plans but not to select decision makers. These exercises range from a structured review of response plans to full command-and-control exercises. However, they can be difficult and time consuming to prepare, and it is not easy to represent or implement decisions involving equipment and people.

Furthermore, information that is presented is done so in an artificial format and is not always representative of what would happen in the real world. Live exercises are also difficult and expensive to organise. They require many participants; they can disrupt normal business and can put reputation at risk from media attention if unsuccessful. Some scenarios are difficult to implement, as they require visual cues such as toxic vapour clouds.

Synthetic environments, namely interactive computer simulations, can overcome these limitations. They provide a controlled environment for selecting, training and testing decision makers. Artificiality can be reduced by providing players with information that is more representative of the real event.

The implementation of decisions and the subsequent actions are also better simulated. An added dimension that can be presented is that human decisions can be used to change the outcome of any simulated event.

Exercises using synthetic environments to replicate the spilling of industrial chemicals have found that emergency response plans are only effective if they fully integrate all responding organisations. Human error can exacerbate the effects of an accident. In one exercise, decision makers were overconfident from an earlier exercise and sent resources to the wrong location.

The same exercise also found that terminology needs to be clearly defined and understood by all organisations. It became apparent that the emergency services required detailed information for conducting their risk assessments prior to dealing with the accident.

In these simulations, it was observed that decision makers became frustrated at the time it took for their decisions to be implemented. This increased their stress levels and significantly limited the information available to them, which in turn created a far more realistic set of circumstances than could be achieved in a table-top exercise.

Responding to a CBRN attack

There are some 300,000 chemical and biological compounds currently manufactured, of which approximately one per cent are toxic. The production of hazardous chemicals comes under the Control of Major Accidents and Hazards (COMAH) legislation.

In the UK, the greatest number of 'category-one' COMAH sites are located around Ellesmere Port, near Runcorn. Many of these chemicals are transported by road, and accidents and incidents involving tankers are a not-infrequent event. Their impact is usually limited, but a combination of system failures or deliberate acts of terrorism could potentially lead to disaster.

Figure 1 (above) shows some of the complexity of the situations facing decision makers responding to a CBRN attack.

When an incident has been detected, it will be necessary to assess its nature and magnitude. Areas of contamination have to be determined and weather conditions understood to ascertain dispersal rates. Tasks facing decision makers include organising and managing the decontamination and restoration of the environment.

The timescales for environmental clean-up will depend upon the incident location, the type of agent used, the method of delivery (which may not necessarily be obvious), agent persistence in the environment and the severity of contamination. The recovery strategy will depend upon the persistence and the extent of the agent in the environment. All affected victims will have to be contained for decontamination and this will involve the use of water to clean them. However, there may only be limited capacity for storing run-off water and, at some point, there is likely to be spillage or even rain. Run-off water containing suspended particle agents may well then escape into the water supply.

Effective communications are essential between the responding organisations and this includes getting the message out to the public so as to avoid panic. The response to dealing with a flood is very similar to that of dealing with a CBRN attack, especially when dealing with flash floods. Warnings of flash floods are given, but they occur suddenly and their location cannot be easily predicted. Planners can use a similar model to that depicted in figure 1. For example, where the model is being used by flood planners, the blank area in figure 1 would be occupied by the term 'air flights'.

Floods have enormous impact, affecting people's livelihoods and property and the local environment, often within seconds. It is often the case that long-term economic damage can occur and a lengthy period of recovery is needed.

Figure 2 shows some of the complexities required to represent a real-time dual-use synthetic environment to help train decision makers in managing flood and CBRN incidents. These prediction models will be highly integrated, using Distributed Information System protocols and High Level Architecture. The models, together with sensor and data inputs, will drive the simulation, thereby creating a real-time immersive, virtual threat event. With development, the BAE Systems SIMTEC man-in-the-loop modelling and simulation framework is a possible candidate for creating such an environment.

SIMTEC modelling

SIMTEC electronically recreates the command-and-control structure of the participating organisations. Simulated radio and telephone nets replicate the actual communications available, imitating only the resources provided in real incidents.

The same is true for information; therefore, decision making will be based on evolving situations with incomplete information. The frustration factor of having to wait for resources or individuals to arrive at their designated points will likewise be included. Tools will be required to quickly identify patterns, trends, irregularities and anomalies as decision makers will need to make sense of complex, abstract data and resolve conflicts of information.

There will also need to be the capability to measure qualities such as leadership, adaptive thinking, situation assessment, individual and team performance, and command, control, communications and information gathering skills when dealing with threat events. Performance could also be measured on factors such as the number of casualties and deaths resulting from stress, human error and/or inadequate resources. Looking at how external communications are managed is essential, as dealing with the media is now all-important.

Of particular interest is how a multi-organisational team will respond during the first minutes or the first hour of a threat event, especially when there may be incomplete information and delayed, corrupted or even lost communications.

Once the scenario has been worked through in terms of key decision points or the resources required, the scenario can be divided into a series of vignettes.

Using tools such as SIMTEC for building simulations and for operational analysis, the analyst can then undertake probabilistic modelling in order to analyse and then optimise outcomes in terms of decision, event and response timelines.

These can then be adopted by the decision makers and incorporated into the relevant plan. The management of threat events, regardless of their type, depends on the performance and effectiveness of decision makers, their equipment and those resources that are available.

Poor management, inadequate testing of plans, lack of adequate training and failure to take account of the possibility of human error are usually to blame when things go seriously wrong.

Synthetic environment

The requirement is therefore to create a real-time, multipurpose synthetic environment that can integrate predictive models, remote sensing and data inputs for simulating threat events.

This synthetic environment can then be used to select decision makers on their ability and not seniority, for training, testing and measuring their response to, and management of, threat events.

The use of synthetic environments will remove the need to put people and equipment at risk in practical scenarios. It will enable the testing of all possible outcomes and provide an identification of areas for correction and improvement. Once the scenario has been worked through in terms of key decision points, resources and outcome, decision, response and event timelines can be optimised and incorporated in the relevant plan.

Janusz Adamson is a consultant in homeland security for BAE Systems CS & S Design Services.