When approaching the challenge of cost-justifying investments to enhance reliability of plant equipment, reliability professionals sometimes struggle to quantify the benefits. Although costs are typically rather easy to define, the benefits often hang us up and sometimes result in “analysis paralysis.” In the worst instances, analysis paralysis causes us to lose the opportunity to act and/or the initiative to lose its momentum.
Sometimes, well-intentioned analysts attempt to precisely calculate that which can’t be deterministically defined. For example, if I asked you to define the mathematical constant pi to its last digit right of the decimal point, you’d probably reply that it can’t be done. Theoretically, you’d be correct. The last attempt to define pi was truncated at 33 trillion digits to the right of the decimal point.
The fact is, for most reliability investments, we don’t require that benefits be deterministically defined with absolute precision. By contrast, we’re usually okay with a probabilistic estimate, which includes sensitivity bands that give us the high-, likely-, and low-case-estimate scenarios. In other words, “pi = 5” is usually adequate.
As a rule, when you set out to estimate costs or benefits for a reliability investment, utilize whatever deterministic values you have available. But when data runs short and you need to make estimates, consider the Delphi Method. This great tool was developed by the RAND Corporation, during the Cold War, as a process for creating values based upon the collective wisdom of experts and informed stakeholders. IEC 31010-2019 (formerly IEC 60300-3-9), the international standard for risk-assessment techniques, identifies the Delphi Method as a suitable tool for those instances when we must employ estimates.
The way the Delphi Method works is for a question (or a group of questions) to be posed to an informed stakeholder group. The group may openly discuss the context of the question, but no individual is allowed to throw out a number. Then, each group member privately submits his or her response to the facilitator who calculates the average and standard deviation and shares the results without revealing any individual’s estimate.
If the coefficient of variation is low, the first estimate is accepted. If there is a great deal of variation, the group may further discuss the question and repeat the process, which normally narrows the band of variation. The stakeholders may then quite easily calculate the low, likely, and high cases for the sensitivity analysis. (I normally take the average case as the likely scenario, and subtract or add one standard deviation for the low and high cases.)
The Delphi Method is fast, easy, consensus-based, and representative of the collective wisdom of an informed stakeholder group. I’ve utilized it to estimate the benefits for a range of different projects, including calculating the earnings before interest, tax, depreciation and amortization (EBITDA) impact of a comprehensive multi-year enterprise-wide, transformative, asset-management initiative.
Go ahead and give the Delphi Method a try in your organization. After all, as the title notes, “Sometimes, π = 5 is good enough.” Just as important, though, it’s almost always better than languishing in analysis paralysis and inaction.TRR
ABOUT THE AUTHOR
Drew Troyer has 30 years of experience in the RAM arena. Currently a Principal with T.A. Cook Consultants, he was a Co-founder and former CEO of Noria Corporation. A trusted advisor to a global blue chip client base, this industry veteran has authored or co-authored more than 250 books, chapters, course books, articles, and technical papers and is popular keynote and technical speaker at conferences around the world. Drew is a Certified Reliability Engineer (CRE), Certified Maintenance & Reliability Professional (CMRP), holds B.S. and M.B.A. degrees, and is Master’s degree candidate in Environmental Sustainability at Harvard University. Contact him directly at 512-800-6031 or [email protected].
Tags: pi, reliability, availability, maintenance, RAM, cost-justification, Delphi Method, RAND Corp., IEC 31010-2019
