Based on a paper presented at the recent IMIA conference, Roger Cottell and Andrew Morton discuss human factors in engineering risk.

Each year IMIA1 members present papers considered to be at the forefront of risk engineering. At last year's conference there was a focus on human factors in engineering risk, pushing forward some traditional perceptions that risks are largely quantifiable as a consideration of the human elements, and the consequences these may have on an insurer's portfolio in a global context.

For some time, a number of techniques have been used to reduce risk during the life cycle of the risk. These may include quantitative methods such as fault tree analysis or more qualitative analysis, typically Hazop (hazard and operability study) or FMECA (failure modes effect and criticality analysis) with input from specialists who ideally have first-hand knowledge of the plant to which the technique is to be applied. For maximum benefit, these quantitative methods, which have a direct link with the human element, should be undertaken throughout the life cycle of a risk from original design through to maintenance and modification and eventual decommissioning.

But there is more to reducing risk than merely following prescribed methods, taking into account local cultural factors such as ‘power-distance (PD)' – how far a culture encourages people to exert power – and ‘uncertainty avoidance (UA)' – the degree to which a culture copes with novelty and encourages risk taking. So it would seem that different approaches would aid risk reduction in countries such as the UK, perceived to be a ‘village market' (low UA and PD) to a ‘well-oiled machine' such as Germany (high UA, low PD) or, say, a ‘pyramid of people' culture such as France (high UA and PD).

The UK Health and Safety Executive gives some useful guidance on human factors, defining them as referring to the environment, organisational and job factors, and human and individual characteristics which influence behaviour in work in a way which can affect health and safety. So there is a focus here on: the job – task, workload, procedures, environment, ergonomics; the individual – competence, skills, risk perception, personality, attitudes; and the organisation – culture, leadership, communication, resources.

A key aspect in reducing risk is therefore seen as intervention at an appropriate level, considering behaviour (what's happening now) as low level intervention, structure (what is the desired system) as medium level intervention and context (the setting) as high level intervention. At an organisational level, for example, there may be a poor climate with anxiety and lack of response to environmental change (behavioural level); perhaps organisational mirroring may improve this. But there may be more deep-seated structural problems such as misunderstood or poor strategy where a change of structure would be appropriate.

These ideas are developed into a three-dimensional intervention matrix with the axes ‘degree of intervention', ‘level of intervention' and ‘behavioural taxonomy' – a framework in which the human effects may be measured throughout the life cycle of a risk. This will assist in providing an indication of where a risk lies and what would be the most appropriate action in the short, medium and long term to reduce the risk.

This analysis need not be overly complex. As an example of where loss of plant and injury has been minimised, legislation including an inspection regime has reduced the number of reported power press incidents in the UK from over 500 in 1943 to a consistent 20 to 30 annually in recent years. While some of this reduction is connected to the decrease in power presses, statistics clearly show a link between intervention and risk improvement. Further improvement is anticipated as the impact of the risk-based 1998 PUWER (Provision and Use of Work Equipment Regulations) becomes apparent.

So how does this impact on an insurer's portfolio? A number of cases illustrate. Perhaps there are design inadequacies that are similar to those which have been experienced previously elsewhere but, despite these deficiencies, such as the inadequate radiusing of the blade root of a Kaplan-type hydroelectric turbine which led to failure, problems should have been picked up with monitoring equipment prior to the loss. Or, what would be the effect of running a black liquor recovery boiler in a pulp and paper mill with reliance on limited safety equipment? Eventually, particularly if the human elements had not been fully considered, there would inevitably be a failure. And yet these circumstances arise.

Risks may be improved significantly however with limited external intervention. During extension of an Eastern European refinery with western project management, difficulties were overcome by employing a Polish-speaking safety manager who had lived in the west. He therefore had an appreciation of the values of both sides and was able to lever the excellent project management skills of the team and apply them locally. Or, there may be a long-term partnership between the client and the insurers where trust is built and more substantial intervention is undertaken with guidance from insurers' risk managers.

What is being proposed then is a reduction of risk by turning engineering into people rejuvenation. In building a total risk map with environmental, financial, market and organisational factors, the constituent elements in this map will themselves comprise a balance between discipline and stretch (for example, balancing safety policy with new technology) underpinned by an atmosphere of trust and strong level of support. In the situation where a long-term partnership is developed, this principle may be applied to improve the organisational risk (by a focus on the human factors) and environmental risk with, for example, improved safety equipment and procedures. This would also impact on market risk, as there would be greater confidence that the required quality of product could be supplied, and hence the financial stability of the company.

There is a clear benefit here for both the client and the insurer. But originally the insurer was placing itself at considerable exposure. Is this wise? Portfolio theory suggests it may be and, if the desire were to have an optimum portfolio of risks, each of which concentrates on return, risk and correlation coefficient, then human factors would likely have a direct impact. By carefully choosing a portfolio it is possible to improve risk without reducing return, or reduce risk without sacrificing return. That is, it is possible to adjust a portfolio to increase return by taking on a proportion of less-than-ideal risks.

But of course an insurer's portfolio is constantly evolving. The aim though is to improve risks until they are as close as possible to an ‘efficient frontier'. By the building of loyalty and trust with clients it becomes possible to rebalance portfolios by periodically choosing further less than ideal opportunities. Conversely, where there is little commitment to a partnership resulting in reduced opportunity for risk improvement, or where there is an indication that a risk is reverting away from the efficient frontier, by this regular scanning of the risk map it will be possible to identify the point where the risk becomes unacceptable.

As engineering insurers, IMIA members are in a privileged position in being able to assess risk first hand. The application of the intervention matrix and risk map indicates areas where this may be enhanced, but the key is appropriate intervention by the insurer's engineering facilitator who is able to win trust. This therefore is sound risk management built on trust in which there is rejuvenation developed through an openness and willingness to commit.

1. The International Association of Engineering Insurers recently changed its name from the International Machinery Insurers'Association (IMIA)