Review by Peter Appel
It is interesting to observe how the science of analysis is continually developing more and more precise methods that can analyse smaller and smaller amounts of a target compound. Historically determinations obtained from analytical laboratories were first in percent, later in ppm, followed by ppb, and most recently in ppt. However, these scientific and technical advancements were not followed up by thoughts and demonstrations on how we can obtain reliable results of concentrations in the large amounts of original material, the so-called lots.
Kim Esbensen’s book is presenting a new, valuable introduction in the form of a comprehensive overview of the principles and unit operations behind documentable, representative sampling from heterogeneous lots typically 1:103–1:106 larger than the primary samples, and far larger still of the analytical aliquots. This book introduces the Theory of Sampling (TOS) in a novel didactic fashion, which will be of wide interest.
My first encounter with Kim Esbensen was several years ago when he invited himself to give a presentation of Theory and Practice of Sampling (TOS) in the department Economic Geology of the Geological Survey of Denmark and Greenland where I worked then. Together with my colleagues I listened politely to his presentation, but found that this whole Theory of Sampling (TOS) approach to sampling was way too time consuming and thus not really appropriate for our work as practical field geologists (but yes, perhaps in the sample-processing stages in the laboratory, which was, luckily, “somebody else’s problem”).
However, some years after this presentation, I embarked on a project where I realised that TOS was in fact the only way to obtain scientifically reliable results. I participated in a global project finding ways to clean millions of tons of tailings from small-scale gold mining. These tailings contain large amounts of mercury and gold, but contents are in the very low ppm range, e.g. 3–150 ppm, while tailings tonnages are enormous, 3–5 tonnes. Mercury released from these tailings account for 37 percent of the global mercury pollution! In order to evaluate the efficiency of a newly developed mobile processing plant, I had to know the exact ultra-trace concentrations in the tailings. This was crucial for evaluating the efficiency of success extraction these two metals from the overwhelmingly large tailing tonnages. This was when I suddenly found that the usual geological sampling approach was totally inadequate for this purpose, but where only the TOS could outline a valid procedure. The results of this project, and others, have been published,1 where the details of this “barefoot sampling” approach can be found.
This new introductory book has many excellent chapters explaining why and how to obtain proper representative results of the content of a particular element or group of elements in significantly heterogeneous materials, rocks, ores and tailings (typical lots in my line of work), but one of the book’s strongpoints is that is shows the common nature of proper sampling. For example, TOS is equally important in investigating the pollution impacts from mining operations or from dumping of industrial tailings of various sorts. In fact, at first it was shocking to read that “from the point of view of sampling, all materials are identical, it is only a matter of whether their heterogeneity is small, medium or large”, after which TOS’ claimed universal principles made much more sense. I am now slightly embarrassed over my first dismissal of TOS, which sadly, is rather typical within geology. Take notice!
Thus my subsequent work using TOS was really an eye-opener. Upon reflection, and the above critical use of TOS in practice, this book should be compulsory in educating not only geology students at universities, but also scientists and technicians in a number of other disciplines, not least analytical chemistry (there is a whole world outside the four walls of the laboratory). All will benefit fundamentally from addressing this book.
Leo Simon Morf
Review by Dr Leo Simon Morf
Deputy Head of Waste Management Department, Canton Zurich, Switzerland
Head of Technical Advisory Committee, Swiss foundation development center for sustainable management of recyclable waste and resources (ZAR)
Lecturer for waste management techniques at ETH Zurich, Switzerland
Investment decisions in the mining industry, or raw material delivery prices, depend critically on reliable knowledge of content of the substances that determine the value of the products and goods. In this industry sector, it has since long been known that the precision of an analytical result, whether in the exploitation of a deposit or the characterisation of a ship’s cargo, is less dependent on the analytical modality (e.g. high-precision gravimetric methods or high-resolution, accurate analytical instruments) than on the question of whether it is based on representative samples. For this reason, representative (accurate and precise) sampling and sample processing techniques have been developed by the sampling community, specifically by the Founder of the Theory of Sampling (TOS) Pierre Gy. This development started no longer ago than in 1950.
In the second half of the previous century, a large proportion of the valuable substances in geogenic deposits has been extracted and are now to be found in the material inventory of industrialised countries, whether it be in infrastructure, consumer and investment goods (sometimes characterised as “Urban Mines”). On the other hand, our modern, highly-developed society demands ever more complex products with more and more components; these include not only valuable, recyclable materials such as precious metals, rare earths, but also increasing amounts of new, potential pollutants, often also toxic, e.g., flame retardants, nanomaterials. It is therefore not surprising that waste flows out of these mines contain both potentially hazardous substances and potentially recyclable materials that warrant development of new technologies for their utilisation. In this contemporary waste management context, in essence quite similar to geogenic raw materials, these resources should be explored and characterised using suitable analytical procedures, and the finished preparation products derived therefrom (such as metal fractions from municipal solid waste incineration plant MWIP bottom ashes) should absolutely be considered for emerging commercialisation, while hazardous substances should get destroyed or immobilised. In this challenging field of interrelated urban mining, sorting, refinement, recycling and safe waste deposition representative sampling play a critical key role also in the emerging paradigm of circular economy.
Surprisingly, in the global waste management sector, the Theory of Sampling has yet barely been considered or applied consistently. The Danish Standard DS 3077 for representative sampling, the facto international standard, is not known or applied in practice in this sector so far. Combined with potential errors due to incorrect sampling techniques and inadequate sample preparation (as may arise from “simplifications” in both steps), this will unavoidably lead to unnecessary errors in the determination of the quantities of pollutants, toxic components or recyclable materials, and will thus hamper optimal estimation and assessment of both ecological risks and economic opportunities. Applying TOS approaches it has recently been demonstrated that highly significant, but quite unnecessary, uncertainties accumulate along the “lot-to-aliquot” pathway, which is extraordinarily complex within waste management and circulatory economy.
Prof. Esbensen’s new comprehensive book Introduction to Theory and Practice of Sampling closes an important general gap, also for users in the waste management sector. This book is critically necessary for application of professional sampling procedures concepts and techniques in modern waste management systems—and beyond. This book shows in an unusually refreshing, simple and clear way all the relevant concepts and connections between the General Principles of sampling and the powerful concept of Sampling Unit Operations, and it demonstrates the often large quantitative influences on the final uncertainty of the tiny analytical samples that ultimately are delivered to the laboratory. A very useful principle is that all mass-reduction, sample splitting, sub-sampling and sample preparation operations can be treated identically as primary sampling, only at smaller and smaller scales. This understanding helps building a necessary overview of TOS.
One of the book’s major advantages is the lavish use of carefully designed didactic diagrams which help the reader to form his/her own understanding of what turns out not to be so “complex” subject-matter as many claim. The many practical examples are vivid and valuable with which also better to understand TOS’ systematic interconnections.
I am convinced this book will find great interest and application among individual actors, companies, organisations and other stakeholders in many application sectors in science, technology and society.
Review by Allan Trebbien
B.Tech Management & Marine Engineering
For a reader like me, a newcomer to the field, this new book gives the reader a well-crafted overview of how to get your sampling right from the start—and what it will cost you if you don’t. Combined with a first foray of practical highlights from the theory behind it all, the Theory of Sampling (TOS), it also supplies a lot of interesting background.
From the start, the author delivers a book written in a somewhat unorthodox fashion that makes it very easy to read, complete with many detailed examples and case histories. This book may well end up being the standard introduction sourcebook for representative sampling.
Having read the entire book, the reader is well equipped to start performing sampling in an effective and professional way. The reader will be fully able to appreciate the Danish standard DS3077 “Representative sampling – Horizontal standard (2013)”, the only international standard on this complex subject and a very large trove of literature (and internet) sources as well. This book delivers the necessary understanding for how to perform representative sampling across a great many types of materials, under nearly all conditions of interest in technology and industry.
It comes highly recommended.
Review by Gary Ritchie, MS
Principal Consultant, GER COMPLIANCE
Council for Near Infrared Spectroscopy News Editor
QC/QA GMP Remediation, Training, Audits, Data Analytics, Chemometric/Spectroscopy Remote Analysis
Former Director of Operations and Regulatory Affairs Analyst for Dynalabs, LLC, St Louis, MO, USA
Former Director of Scientific Affairs for Infratrac, Silver Spring, MD, USA
Former Scientific Fellow for Process Analytical Technology (PAT) with the United States Pharmacopeia (USP)
International expert in multivariate spectroscopic analysis, pharmaceutical analysis, Good Manufacturing Practices compliance, laboratory management, design, quality, and process analytical technology using spectroscopic methods and multivariate analysis
How to read this book?
The reader is encouraged to read the first two chapters, followed by Chapters 22–24, and only then continuing from Chapter 3. In this fashion the reader is promised guaranteed optimal conditions with which to achieve all the promised learning goals. Having read the book in the manner recommended above, I have the following comments.
Does this book work as intended?
I thoroughly enjoyed the conversational style in which the book is written. I imagine that I could very well be listening to Pierre Gy himself as he presents his Theory of Sampling for the very first time. If the author intended for the reader to experience an introduction to the TOS as if he/she was the first to ever hear it, the book has achieved its objective well.
What is good?
Far from being a conventional cut and dry textbook, the informal and didactic style adopted allows the student-reader to learn at an easy and suitable pace. I followed the recommendation on how to read the book; I read the first two chapters, followed by Chapters 23–25, and only then continuing from Chapter 3; again. This works well because once read in this manner, a self-paced cadence for further instruction and insight into the references provided was easily achieved.
There is a vast amount of additional text offered in the book (directly clickable sidebars),which can at first make it somewhat difficult to determine what is essential versus what is more for extra information. However as one moves along, these sidebars prove to be important locations for necessary reinforcement and references for further reading. The main text provides much scientific insight and color and additionalities as well, but it is precisely this mix that gives the reader-student the sense of discovering the TOS for the very first time.
The take away lesson from this book is clear, and very unsettling.
“...hidden economic losses due to faulty decisions based on faulty analytical results due to non-representative sampling.”
These words should give every analytical chemist pause for concern. After all, we have been taught and we believe that our analytical results are based on a sample that must be representative, for why else are we testing it? And then, when/if the result fails the acceptance criteria, we are told, after a proper investigation, to test it again. It has become the focus of discussions: just what is a sample really and what is it representative of? This delves right into the realm of the Theory of Sampling (TOS). While analytical testing is in and of itself a science rooted in the natural sciences and mathematics, here now comes the TOS that teaches us that homogeneity of the aliquot isn’t the goal, but representativity of the primary sample (and all sub-samples hereof) is, and that sample size matters and sampling error goes beyond conventional statistics. It is all about the sampling process, and here the book’s strongpoint is that all major points made can easily be understood within the framework of TOS’ six fundamental sampling principles and four sampling unit operations.
In Introduction to the Theory and Practice of Sampling, by Professor Esbensen, with contributions from several well-known and respected experts on the subject, the TOS is presented in an easy, comprehensible style that is assessible to everyone at all academic levels that allows readers, students and practitioners to learn very quickly a fundamental fact of nature that can no longer afford to be unknown or ignored: homogeneous materials do not exist in the real world; every material in technology and industry is heterogeneous. Hence, the flaw in our current thinking and practice of treating “analytical grab samples” as if they always are homogeneous and representative, turns out to be the single biggest source of economic loss to most of the modern world’s industrial endeavors.
I recommend this book to all newcomers to TOS, but especially also to those who want to go beyond the TOS basics and further explore its numerous literature sidebars and background references. For in depth coverage of the quantitative nature of sampling theory and practice, this is the place to start. I feel the opportunity to review this book is an extension, and culmination, of my role when as a scientist at the United States Pharmacopeia about ten years ago, as the perfect time and place to get the TOS in front of the key committee members who would go onto incorporate the Theory of Sampling into the United States Drug Compendium.