The great advances in computer modelling over the last two decades have brought a desire in many professions to rely exclusively on the past technology for problem solving and design. In the field of geotechnical engineering, modelling has indeed made some significant contributions, however, an over-confidence in the ability and outcomes of numerical methods of analysis would be misguided and ill-advised.

It is important to re-state some engineering fundamentals that appear to be getting lost in the high-tech electronic world we live in. As a starting point, it is worth defining the term “engineering design” as this is often at the heart of the problem.

The Institution of Engineers Australia has published several excellent texts on this subject, with the following basic quotes being taken from “Are You at Risk” 1990:

Engineers project an image of dealing in “hard” models, whereas in the main they deal with “soft” models.

We know (or should know) that our models are limited in their ability to represent real systems, and we use (or should use) them accordingly. The trouble is that we are so inordinately proud of them that we do not present their limitations to the community and leave the community with the impression that the models are precise and comprehensive.

Design is the central engineering activity. It is a process which combines knowledge with judgement to obtain a desired outcome. Our mathematical models contribute only partially to the process, but we often give the impression that they contribute all.

Judgement is the key to the engineering method. It is the only skill that can appropriately manage a heuristic environment (ie. one that is not fully defined or understood, yet still requires engineering design to be applied to it, and demands that we endeavour to increase our level of knowledge as we go about our work).

Powerful words indeed, that should be central to the core activities and philosophies of every practising professional engineer. This is particularly so for geotechnical engineers as the properties and behaviour of rock masses are so hard to pin down.

However, the key question that emanates from the previous statements is: “how does one acquire judgement?” It is certainly not a taught subject, yet it is apparently the critical aspect of engineering design. Engineering judgement is almost entirely borne from experience and this brings us to a critical theme. In the field of mining geotechnics, the potential experience base is huge.

For example, many longwall panels are mined each year and each one is a full-scale test of a pillar design and ground support systems in a specific geotechnical environment. We need to try to learn the history lessons the real world is endeavouring to teach us. The key is to gather and quantify mining experiences whenever and wherever possible.

As geotechnical engineers we have to work with a rock mass that is an extremely complicated natural system. The ground comes with a large variation in material properties and reactive behaviour. Other sciences also deal with complex and variable systems; medicine, for example, has to work with the human body.

There is still no satisfactory “numerical model” for the human body. Instead, medical scientists use empirical methods based on quantitative and qualitative data analysis using statistics. Empirical techniques have resulted in many of the phenomenal accomplishments of modern medicine.

New drugs are approved every day based on empirical studies and controlled trials. Medicine does not question the benefits of this basic methodology; it simply understands the limitations and applies the results accordingly. Why, therefore, would we as rock mechanics engineers deny ourselves the use of the same successful scientific method?

Perhaps design based on “experience” has got a bad name in some mining circles because it used to mean “that’s what we’ve always done, so it will be right this time”. Obviously, that kind of “experience” runs into trouble the first time conditions change significantly.

Scientific empirical methods rely instead on large databases of mining case histories that reflect a broad range of possible conditions. Statistical analysis can then help to provide the most appropriate design guidelines. These in turn are not the “final answer”, but rather a sound basis for engineering judgement.

Moreover, empirical methods have a worldwide proven track record in the mining industry. Concerns that there aren’t any reliable methods for tailgate design simply because numerical methods have not been successful are entirely misplaced.

In the USA nearly every longwall gateroad panel design over the past decade has used the Analysis of Longwall Pillar Stability (ALPS) method, precisely because it has proven to be so reliable. In Australia, there is a reliable, fully engineered, empirical technique that is being used by mine operators, consultants and the inspectorate to assess the same issues, known as ALTS II or Analysis of Longwall Tailgate Serviceability (Version 2).

The database on which ALTS II is based brings wide-ranging experience, from all Australian longwall operations, to assist in solving the complex issue of tailgate support design.

Empirical methods are also routinely used to develop guidelines for a variety of other issues, from the suitability of cut-and-flit mining to the support required to mine through a longwall recovery room and many others.

Mine operators generally relate well to empirical methods because improvements in mining techniques have traditionally been based on learning from experiences and they can easily see and understand where answers come from. In overall terms, it is absolutely vital that persons charged with making geotechnical decisions at an operational level have a basic knowledge and understanding of the engineering design methods in use.

This is readily achievable with empirical methods due to their simplicity and transparency, but is obviously more difficult when complex mathematical codes are used. To assist in this process, the originators of empirical techniques in Australia and the USA have regularly run workshops to clearly explain the methods and associated limitations to all interested parties. Continues.