W. Stephen Tait, Ph.D. Chief Science Officer & Principal Consultant, Pair O Docs Professionals, LLC Corrosion Corner What is the difference between risk management and gambling? Figure 1: The empirical relationship between test length and the risk of pitting corrosion in traditional aerosol containers. testing and the risk decreases to less than 1% after 30 to 90 days of electrochemical corrosion testing. Figure 1 can be used to approximate the risk when there are no corrosion data and/or a company corrosion database available to estimate risk. For example, Figure 1 could be used to approximate risk if one is considering truncating a storage test. A properly designed corrosion test is the most robust way to reduce risk. However, another way to reduce risk—and avoid gambling—is to build a company corrosion database to use in conjunction with corrosion test data and employee experience. I strongly recommend that the database a) include both successes and failures and b) be in a written or electronic form that is easily accessible to the appropriate members of your spray product development team. We would be happy to teach our Elements of Spray Package (Aerosol Container) Corrosion short course at your R&D facility. We have also introduced the Corrosion Partnership Program, which makes our electrochemical corrosion laboratory, corrosion consulting and anti-corrosion technology conveniently available for your R&D program. Contact: rustdr@pairodocspro.com; 608-831-2076; www.pairodocspro. com. Back articles of Corrosion Corner are available from Spray. Thanks for your interest and I’ll see you in September. Spray Hello, everyone. One of my favorite graduate school proverbs on risk is the three types of engineering guesses: 1. The Wild A Guess (WAG) 2. An Engineering Wild A Guess (EWAG) 3. The Educated Engineering Wild A Guess (EEWAG) The WAG is based on no data; the EWAG is based on one data point and the EEWAG is based on two data points. In the previous May through July Corrosion Corners I discussed a) how the cost of spray package failures (leaking) is significantly higher than conducting corrosion tests, b) that derivative products also need to be tested for corrosion, c) the different types of corrosion tests and d) the strengths and pitfalls associated with each type of test. In other words, it is costly to skip corrosion testing; risky to assume a derivative product will not be corrosive when the original product is not; and only properly conducted corrosion tests reduce risk. Therefore, what is the difference between gambling and risk management? I equate WAG, EWAG, EEWAG and risk management with the following arbitrary statistical confidences: • WAG: No statistical confidence (gambling) • EWAG: 77% statistical confidence (without using a company corrosion database) • EEWAG: 80% statistical confidence (without using a company corrosion database) • Risk Management: 90% and above statistical confidence Figure 1 (also used in the June and July Corrosion Corners) illustrates how risk decreases from gambling to risk management as a function of time. The graph for traditional storage testing in Figure 1 is an empirical curve that was generated for a wide range of spray products in traditional metal aerosol containers. The two data points for electrochemical testing are empirical relationships derived from over 700 direct comparisons with actual spray package corrosion. The electrochemical data includes comparisons with all types of spray packages. Notice that the no-data-risk is approximately 62% in Figure 1. The risk decreases to around 7% after one year of storage 76 Spray August 2016
Spray August 2016
To see the actual publication please follow the link above