D.W. Ellis & Associates Ltd. offers consulting services in risk management education and training, risk management project facilitating, and risk management modeling software.
This information on Risk Management is designed to be read sequentially. You can, however, go directly to any of the following sections:
Decomposition is the key to understanding complicated business risk. It is typically too complicated to try to understand, for example, the risk of operating a chemical plant in an urban area. By looking at the risk separately as the risk to:
the situation is immediately more easily comprehended. Even then, more decomposition is required to fully understand it. For example, the following decompositions may be helpful:
Data on each aspect of each component of risk must be collected so the exposure to risk can be objectively assessed.
It is up to you to define what is at risk for you.
For example:
Unfortunately most risk analysis has been financially based (investment risk analysis).
Usually there are more than one potential injury or loss type. Thus the different risks must somehow be melded - multi-criteria decision making. Whether it is done in our heads or done analytically, it needs to be done.
It is usually not enough to simply consider whether an undesirable has occurred or not.
For example:
The proper use of risk analysis will require you to consider not only the undesirable event but also the magnitude of the event.
For example:
It is usual to use a probability distribution or simply probabilities of various magnitudes.
PROBABILITY OF OCCURRENCEProbability of an event occurring is the chance the event will occur.
Probability of an event occurring is the long term frequency that the event has occurred (assuming no substantial changes to ‘things’ that affect the event).
Example:
These definitions assume there can be no more than one event in the time period defined. In general, the time period can be reduced to ensure this.
Another approach is to provide a probability distribution for the number of occurrences in the time period. For example, the number of accidents at a particular intersection in a year may be from 0 to 15. Provide a probability for each number that sum to 1.0.
The probability may be determined from theoretical considerations or historical evidence may be used. For most business risk analysis, the historical experience must be used.
For example:
Interventions are the key to risk management. Select them carefully, assessment them thoroughly, and manage them rigorously.
Typical Risk Analysis would determine the risk for the current situation or the status quo (exposure analysis).
This risk would be assessed corporately to determine if it is acceptable.
It is perfectly acceptable to accept this level of risk.
If it is unacceptable, an intervention on your part is required.
Action Alternatives:
Interventions can:
Interventions:
Reduce the probability of an undesirable event:
Reduce the magnitude of the undesirable event:
Reduce the exposure of the undesirable event:
You must be able to estimate the impact of the intervention you are proposing.
Whenever possible, data should be used to determine these changes.
In all likelihood, you will have to use more judgement in estimating the impact of these changes.
The Delphi Method can be used to determine a group’s perception of the estimated impacts of interventions.
EXAMPLE 1 - PIPELINE RISK ANALYSISThis case is the analysis of the risks and consequences of the rupture of a pipeline carrying volatile products.
The risk to be analysed is the risk to people, assets, environment, and public image.
The pipeline is divided into segments that are relatively homogeneous relative to each of the 4 types of risk. Thus for each segment, the distribution of people, assets, environment and public image items are about the same along the segment. The segments do not have to be of equal length.
For example, there may be long segments in the country where the pipeline passes through farm land, a short segment where the pipeline crosses a river, a short segment in a city through residential areas, and a short segment through industrial areas.
An inventory of people, assets, environment, and public image items is taken in each segment and the data is collected within 100 meters of the pipeline, 100-200 meters of the pipeline, 200-400 meters of the pipeline, and 400-800 meters of the pipeline. An earlier assessment has determined that the risk outside 800 meters is very small.
Data on pipeline failures under various situations is collected. Because failures are very infrequent, 20 years of history over an extended area is required. The industry keeps such records. Various severities of pipeline ruptures will be considered and the inventory by distance from the pipe line will be used to consider different exposures.
Some scientific data is available to assess different impacts at various distances. The historical data on ruptures will be used to determine the frequency of the various severities of rupture.
With this information, measures of risk were developed for each of the 4 types of risk (details are not discussed here).
This level of risk was considered unacceptable.
The following interventions were considered:
Level 1:
Level 2:
These 2 cases were evaluated. The results show reductions of risk in the 4 segments with overall risk reductions of:
Intervention People Assets Environment Public Image
1 19% 18% 13% 14%
2 49% 49% 33% 36%
EXAMPLE 2 - RAILWAY RAIL FAILURE RISK ANALYSISThis case is the analysis of the risks and consequences of the failure of a rail in a railway system.
The risk to be analysed is the number of failures, hours of train delays, number of derailments and the cost to manage the rail failures.
The railway system is divided into segments that are relatively homogeneous relative to each of the 4 types of risk. Thus for each segment, the weather, age of the track, volume of traffic, freight/passenger mix, ease of access, and topography are about the same along the segment. The segments do not have to be of equal length.
For example, in the prairies, the railway is relatively straight, has uniform track bed, and has easy access. In the mountains, there may be a segment with old track, high volume, and difficult access.
The history of rail failures and defects found during test runs was studied relative to the timing of the tests runs of the track. The relationship between defects found during testing and rail failures was found.
With this information, measures of risk were developed for each of the 4 types of risk (details are not discussed here).
The following interventions could be considered:
Many of the interventions could be tested in the model to see the expected changes. Costs and results are shown graphically. In addition, an optimization is included in the model. For a given budget for track testing, the best locations for the testing are shown.
D.W. Ellis & Associates Ltd. develops customized risk management models for clients. Because of an existing structure for the models, they are very quick to develop. They are Windows based, easy to use, and provide for multiply case analysis. Be sure to see some of the features illustrated in Example 2 above. Please contact us for more details and examples.
Please see our information on the Public Perception of Risk.
If you have questions or comments, please e-mail <.
Back to D.W. Ellis & Associates Ltd.'s management science page