'Project cost estimating' A book review

The expectations of many connected with estimating project costs were no doubt raised, writes Costimator, by the advance publicity for Jack Sweeting's book 'Project cost estimating - principles and practice', published by the IChemE* at £56.00.

After all, it is years since much new and substantial in the form of a comprehensive text was written on the subject; even the IChemE's well-thumbed (but out of date) 'A guide to capital cost estimation' leaves much to be desired. However, I do admit that a good part of the audience for the latter is the student community and, as such, it has served a very useful purpose. In contrast, Mr Sweeting's book is a disappointment.

Let us start with the aims of the book. These are stated to be the examination of four aspects, as follows:

{{* the basic mathematics associated with cost estimating;* factors that influence the accuracy of estimates;* how computers can assist in producing estimates;* the organisation of estimating data.}}

The first two of these are said to be concerned with developing a theoretical understanding of cost estimates; the others relate to practical implementation. Unfortunately, I cannot see how the average estimator can relate to the 'practical implementation' objective when there no examples, data or analysis of relevance to any real project.

Admittedly, in his preface, the author states that 'I attempt to show how to deal with some of the underlying issues and make no attempt to provide details of the cost of anything.' The trouble is, what professional estimator can afford the time or have the dedicated concentration to read 258 pages of 'underlying issues'?

Of course, a book which explores the notion of the statistical nature of estimates and demonstrates this clearly using dice (or a die), and from which some very broad and general conclusions (or rules) can be drawn, is welcome. To Costimator's knowledge, this has not been better done in any other book. But I still question the target audience.

Rather than give a chapter by chapter commentary, the panel lists the major subject headings within each chapter, which - except for the lack of illustrative cost data - are largely self-explanatory. Features of note include an excellent account of the skewed characteristics of estimates and the log normal and gamma distributions which can be used to describe them. Also, the mention of discounts from the purchase of multiple units is well worth noting, as is the point about management taking an overview of a portfolio of projects in preference to regarding each one in isolation.

While I agree with the quotation beginning Chapter 4 that 'there is nothing so practical as a good theory' (which, by the way, is due to Clausius and is not a Proverb), the average practitioner of the estimator's art will, I think, find some mathematics in Appendix 1 a difficult chore. Still, that is why they are relegated to an appendix. An example of a mathematical omission is in the derivation of the 'two-thirds' (or 0.6 as Sweeting has it) power law (see Costimator, PE June 1987, and January and May 1988).

A further omission is the Pareto Principle (again, see Costimator, PE August 1992), which states that roughly 20 per cent of equipment items account for 80 per cent of the total project estimate. Therefore, the estimator should concentrate on the 20 per cent most costly items - which is a much shorter way of making one of the points in the book.

Most of Chapter 3 is based on the Monte Carlo simulation concept. The over-brief mention of this widely familiar technique is thus, in my view, regrettable. Perhaps the Monte Carlo approach should have been taken from the start.

In Chapter 7, there is an argument that cost per tonne should not be used to calculate the cost of say, a vessel. This is certainly true in some cases, particularly that quoted in the IChemE's 'A guide to capital cost estimating'. Nevertheless, the two-thirds power law is derived from the idea that cost is proportional to the surface area of the vessel which, for a given wall thickness, in turn determines its weight. It thus seems that weight can be used - but with care and bearing in mind a 'shape factor'.

Although my review may appear to be somewhat critical, this book is the first to tackle the theory fundamental to understanding the process of estimating. In addition, Chapters 5 and 6 are very helpful and project personnel should benefit from these. All in all, 'Project cost estimating - principles and practice' is a book for departmental libraries, rather than individuals' collections.

*Institution of Chemical Engineers, Rugby, CV21 3HQ. Tel: 01788 578214

{{Summary of Contents

Chapter Subjects covered

1 Estimating Fundamentals Introduction, aims, nature of cost estimating2 Descriptive Statistics Statistics for a die, the normal distribution, comparing distribution, populations and samples3 Estimating Exercises Single item, equipment, construction, conditional costs4 Estimating Mathematics Single item, simple sums, scaling, products, conditional costs, differences5 Practical Estimating Estimating forms, uncertainty, different views, adders and multipliers, consolidation process, cost of accuracy, computers and data, MTO, data maintenance6 Computer Estimating Applications, databases, limitations, overview of computer estimating, standard modules7 Curve Fitting The six-tenths rule, variations in space and time, linear regression, power law fitting, cost curves and estimates8 Implementation Management considerations, programmers, estimators

Appendices Mathematics and calculations, normal and gamma distributions, tables of dice throws}}