In today's highly competitive environment, the decision to initiate or progress a research program is an expensive undertaking which receives increasingly detailed scrutiny. Projects with a clear potential for high returns and perceived low risks will generally pass this scrutiny with few or limited problems. However, most organizations - particularly those in mature industries - will find that the bulk of their potential research programs lie in areas with less clear, lower potential gains and/or higher perceived risks. Hence evaluating which of the numerous proposals to fund becomes more critical. Return on investment is generally a key if not commonly over riding criteria in this selection process. Since the costs to design, construct and start up the pilot plant or laboratory equipment is such a large part of the initial cost outlay, it has a large effect on determining the overall economics and hence the decision to progress the research program particularly if annual costs versus program life costs are used for the evaluation. Hence the effect of errors or uncertainty in the pilot plant estimates can become a critical factor in the decision to progress (or not) a research program.

Most research programs are evaluated in stages starting with a rough conceptual idea and progressing to a fairly detailed final program definition. The pilot plant requirements and design information start out as very ill defined and only sharpen as the process continues. However, the review process usually requires multiple cost estimates as it proceeds, many of which are at points in the process where adequate information to develop the pilot plant design much less the estimate is not available. This creates numerous problems which often result in significant estimating inaccuracy and uncertainty. The effects of this inaccuracy and uncertainty can range from difficulty in deciding if adequate economic information is available to make a final decision to inadequate funding and/or premature cancellation.

This presentation will discuss the different types of cost estimates commonly required and highlight the information required at each stage. The problems with their development and use will also be discussed. Suggestions for improving the quality of the cost estimates at each stage will be presented as well as some suggestions as to how to avoid the most common pitfalls.