D. Aldwin Vogel
Technical & Quality Director, Commodities Global Service Lines, Bureau Veritas, Rotterdam, the Netherlands
From management’s perspective the cost of sampling must be as low as possible: samples are “just” a necessity to enable the laboratory to do its tests. Once the lowest cost sampling method has been identified and implemented—either by the in-house quality department or through a Testing, Inspection, Certification (TIC) service provider—management is done with sampling… Well, except for the occasional slap on the wrist to the samplers when there is a complaint on quality, or a dispute: “Our client does not get the same control results as our own”. This must clearly be the samplers’ fault; they took the wrong sample!
Anyone with Theory of Sampling (TOS) knowledge will disagree with this scenario and will go through fire and water to try to explain that there is no such thing as a right (or wrong) sample. When there is no representative sampling process, there are only specimens… those pesky lumps of matter collected uncontrollably from a lot: specimens are not representative by TOS definition.
Sampling experts always find themselves explaining the “risk of being wrong” and love to bring up the hidden cost of using a non-representative sampling process. These confident boffins happily and relentlessly illustrate with numerical examples, or graphs with error margins, precision and accuracy … that a non-representative sampling process is very likely to significantly reduce, e.g., life-of-mine or result in a financial loss during a transaction (they have an endless array of horror stories from all over industry to tell).
Yet often the experts are met by a yawning manager, or by a manager having a trader mind set, who is feeling lucky that he or she may also benefit. The “risk of being wrong” may just as well flip into “the 50 % possibility of being favoured”. Especially when we TOS illuminati throw in statistics, standard deviations, variances, use “±” signs and may top it all off with a normal distribution graph etc., then the managerial thinking still goes: “Even in the worst case, on balance I will be okay!”
The real world
The process of representative sampling depends on two critical success factors: 1) elimination of Incorrect Sampling Errors (ISE) and 2) reduction of the Correct Sampling Errors (CSE) to an acceptable level.
Here, in order to avoid the yawn, we will completely skip all further explanations, those dull “technical explanations”, but leave the reader with sufficient references (should the interest develop) for proper sampling access to how to make sure every particle can and will be included in the sample, and how to decide on the necessary-and-sufficient number of increments to select (thereby also fixing the all-important question about the optimal sample mass); for references, just look at all other contributions above and below.
The technical truth
Thus, for now, we can refer to what is easily understood by managers—Murphy’s Law, which states that that there cannot be an overall “on balance” when representative sampling is addressing significantly heterogeneous materials and lots, as when compromised by the desire to involve the least expensive sampling approach (grab sampling), which unfortunately is tantamount to allowing a significant sampling bias. This is a single-sided effect that is always a cost and never a benefit; again, just look at all other contributions above and below.
The magnitude of this cost?
The costly truth
Well, let Murphy’s Law decide that for you, instead of us experts trying to make “reasonable” assumptions about inherent heterogeneity and shaky, but dead-cheap, sampling procedures (again grab sampling) in order to quantify a monetary amount or build the resource model for the new mine for example, you know much better yourself!
But, by the way… now that you know this critical issue in these simple terms, imagine how your shareholders will react next time the results from a non-representative sampling process interfere with the bottom line of your annual reports!
What to do—how to go forward?
Simplified there are just three phases for representative sampling.
- The planning phase, prior to sampling
- The actual sampling
- Making managerial, inter alia decisions based on the sample (results)
The TOS’ focus is overwhelmingly on phase 1) and phase 2), e.g. to determine essential stuff like heterogeneity which is needed for better planning. Phase 3) is only for the user… e.g. the manager.
The economic impact
The economic impact of representative sampling is abundantly clear: it is essentially neutral and does not favour, nor prevent, a specific wishful thinking. Ironically representative sampling delivers exactly what a manager expects from a sample: something that can be considered as factual and true… as fully representative of the bulk from which it was taken from, and for which reason one can have complete faith in the corresponding analytical results.
How to tell it to management
So, no big Dollar or Euro amounts to be presented here, no complicated statistical results, no graphs, no error margins. Just you, your imagination and the knowledge that representative sampling is a process that can remove all your fears of a financial claim, or of upsetting your shareholders, or the fear of prosecuting regulators.
Ultimately the economic argument for representative sampling is just that, the most coveted position regarding all business risks: “peace of mind”.
Just a warning though: If the adjective “representative” is removed from any sampling process—all the above goes away in a blink!
D. Aldwin Vogel
Duncan Aldwin Vogel (born in the Netherlands, 13 October 1973) is a global expert in weighing, sampling and testing of traded commodities. Already during his study in business management at the International School of Economics, Rotterdam, Aldwin started building his pedigree in the renowned family inspection business Hoff & Co. Services BV that became part of Bureau Veritas in 2010. From September 2011 to August 2013 Aldwin was based in Houston, USA, seconded as acting Director, Steel and Energy Products. Returning to Europe and the Metals & Minerals Trade Business Line in September 2013, Aldwin is now responsible for Technical Governance of Bureau Veritas’ Commodities Trade services globally. His expertise covers all aspects of inspection, sampling and analysis starting from green field prospect requirements to fully implemented turn-key projects. Embracing augmented inspection services through IoT and smart communication, Aldwin recently also came out as inventor and patent holder of several novel inspection solutions. He is highly experienced at all aspects of testing for Transportable Moisture Limit and was leader of the TML workgroup of the TIC Council. Aldwin is a delegate of the Netherlands on ISO Technical Committee 102 (Iron ore and direct reduced iron) and TC183 (Copper, lead, zinc and nickel ores and concentrates) where his focus is on sampling, sample preparation, moisture determination and TML.