Risk, Uncertainty and Engineering: Best Practices of Risk Analysis on Engineering Projects.
So much information on risk analysis exist but projects still have high failure rates and people still doubt the reliability of risk analysis. Misunderstandings and insufficient tools are factors that encourage this trend, and lessons have being learned but at a high cost to companies. It is important that a good understanding of methods and applications is acquired and proper application of this knowledge will ensure attainment of objectives. Harsher environments now surround new investments prospects; hence investors and insurers now assess companies with their risk management methods.
Uncertainty put simply is the absence of certainty. It is a state at which limited knowledge of an existing state or future result is available, while risk is a state of uncertainty with multiple outcomes in which certain outcomes would lead to undesired loss (Douglas Hubbard, 2007). Risk analysis is an assessment tool used to evaluate risk and uncertainty and their inter-dependencies, with an aim of providing best scenarios (economic and/or safety). There has being increased awareness of risk analysis in the oil and gas industry with the present instability of oil prices, global financial crises, challenging environments, new technologies and fewer reserves discoveries with smaller traps. Also, after the Piper Alpha accident (1988), new regulations came into place in the United Kingdom and certain other European Union countries, of which risk analysis was of importance. Hence, risk analysis is applied from the commencement of hydrocarbon exploration to well abandonment, with integration in the enhancement of platform layout, operational procedures, safety, orderliness and cost.
Risk Analysis: How?
Risk management if applied correctly is believed to deliver set objectives on time or within set budget, but this is actually not a guarantee. Because, in reality the possibility of attaining set cost and schedule can only be enhanced by mitigating certain actions and the desired expectations are provided by quantitative analysis. Qualitative Risk Analysis defines the scope of analysis by stimulating the circumstances surrounding the uncertainties; it includes methods for prioritizing risks and applicable actions (Peterson, Wardt and Murtha 2005). It is normally a pre-requisite for quantitative risk analysis.
The qualitative analysis should be comprehensive though less detailed than the quantitative stage. According to Smith and Merritt, 2002, we should be able to identify the 6W's which will indicate that sufficient data is available to ascertain the associated risks. Here is a typical example of its application in a petroleum refinery construction.
- What is the function of this refinery?
- Who needs the refinery?
- Why do they need this refinery?
- Where will the refinery be needed?
- When is the refinery needed?
- How is the refinery going to be designed/produced?
Now, the last W (How..?) should further be analysed and more W's be ascertained. This provides more insight on the process required to achieve the set objectives (in this case producing a certain product).
- What part of the refinery will be developed internally, and what would be sourced externally?
- Who will design these parts?
- Where will the construction plant(s) be located?
- When can these parts be delivered?
After identifying the opportunities by qualitative risk analysis, the consequences of the possible decisions are quantified and the optimum results within the possible outcomes assessed, this is quantitative risk analysis. It involves design of the probability simulation model, its detail, structure and the sensitivity analysis. Probability simulation models should be properly designed to fit needs of the project. The level of understanding of the model should be meet that of the project and should cover all range of uncertainties involved; hence, mitigation approaches for identified risks and contingency allowance for unseen risks. It should be structured for easy assimilation with key equations, assumptions, defining parameters and correlations of the input distributions listed. The fully built model should undergo multiple testing to determine its sensitivity. The results of these tests will direct strategies for risk mitigation and provide allocation for contingency due to uncertainty. A typical example using the Monte Carlo simulation model; the net product value (NPV) for a refinery project in the Niger Delta is obtained with equation (i) for discount rate i and life of N years. Key parameters like oil price, construction costs, taxes, security, oil production and political stability are each represented in statistical distributions. Random values are then selected from each distribution and substituted into equation (i) to obtain a possible NPV. This is iterated thousands of times to obtain a chart of multiple project scenarios, from which the expected NPV and loss probability are calculated by integration.
Recommended Best Practices
Risk management just as any other task requires due diligence to ensure set objectives are realised. Engineers tend to sidetrack from the main objectives as they go further in the project due to certain mistakes made earlier in the analysis. The following guidelines are recommended for a successful risk analysis process:
- Guidelines should be documented and communicated to all involved parties.
- Supervision of evaluation process to prevent inconsistency, bias and misunderstandings.
- Risk analysis training for all involved parties (technical and management alike).
- Evaluation software should be easy to use, standardised and adaptable to changing technology.
- Proper understanding of inter-dependency of variables.
- Risk method should be flexible to allow integration of new techniques.
- Proper understanding of results and close monitoring to expedite adjustments to methods.
If done properly, risk analysis is worth the effort and very rewarding. Investing in risk management is expensive, hence the need to get it right. Companies should invest sufficiently in personnel training and software enhancement to enable integration with changing technology. Also, commitment and dedication from involved parties is required, tempered with good professional judgement.
HUBBARD, D., 2007. How to Measure Anything: Finding the Value of Intangibles in Business. New Jersey, NJ: John Wiley & Sons.
SMITH, P.G., and MERRIT, G.M., 2002. Proactive Risk Management: Controlling Uncertainty in Product Development. New York, NY; Productivity Press.
NEWENDORP, P.D., 1976. Decision Analysis for Petroleum Exploration. Tulsa, OK; Penn Well Books.
GALLI, G.A., ARMSTRONG, M., JEHL, B., 1999. Comparison of Three Methods for Evaluating Oil Projects. Journal of Petroleum Technology, 51(10), pp. 44-49.
PETERSON, S.K., DE WARDT, J. and MURTHA, J.A., 2005. Risk and Uncertainty Management - Best Practice for Cost and Schedule Estimates. SPE PAPER 97269 prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Dallas, Texas, U.S.A, 9-12 October 2005. Available from http://www.onepetro.org. [Accessed October 2009]
ALEXANDER, J.A. and LOHR, J.R., 1998. Risk Analysis: Lessons Learned. SPE PAPER 49030 prepared for presentation at the SPE Annual Technical Conference and Exhibition held in New Orleans, Louisiana, U.S.A, 27-30 September 1998. Available from http://www.onepetro.org. [Accessed October 2009]
CUNHA, J.C., DEMIRDAL, B. and GUI, P., 2005. Use of Quantitative Risk Analysis for Uncertainty Quantification on Drilling Operations - Review and Lessons Learned. SPE PAPER 94980 prepared for presentation at the SPE Latin American and Caribbean Petroleum Engineering Conference held in Rio de Janeiro, Brazil, 20-23 June 2005. Available from http://www.onepetro.org. [Accessed October 2009]
SOLIS, R. et al., 2004. Risk, Uncertainty and Optimization for Offshore Gas Asset Planning in Litoral Tabasco. SPE PAPER 90177 prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Dallas, Texas, U.S.A, 26-29 September 2004. Available from http://www.onepetro.org. [Accessed October 2009]
KOMLOSI, Z. and KOMLOSI, J., 2009. Application of the Monte Carlo Simulation in Calculating HC-Reserves. SPE PAPER 121256 prepared for presentation at the 2009 SPE EUROPE/EAGE Annual Conference and Exhibition held in Amsterdam, The Netherlands, 8-11 June 2009. Available from http://www.onepetro.org. [Accessed October 2009]
Net Product Value (NPV), VP
Vp = C + ? --- .........................................(i)
1 (1 + i)n
i = Discount rate
N = Number of years
Fat,n = Cash flow after tax, for year n
UKOMADU EMEKA PATRICK 0900090