At an early stage an employer will want to know the probable cost of his intended project. Usually no realistic figure is possible until a feasibility study of the project has been completed; before that only an order of magnitude figure or budget estimate can normally be quoted. Three main methods of producing this are as follows:
by reference to the cost of similar projects;
by sketch layout and component costing;
by use of cost curves if available.
The first assumes a record is available of the cost of past projects undertaken by the employers engineer, or perhaps costs taken from the technical press.
The reference costs need to be accompanied by data, such as project size, project components and distinctive features, dates of construction, and whether the price includes land, legal and engineering costs. Inflation factors may have to be applied to update the costs. By comparing the principal features of the proposed project with those for which past costs are available, a probable order of magnitude total cost may be derived.
The second method is the most reliable. Even before a feasibility study is undertaken it should be possible to sketch out the proposed project on some notional site if the actual site is not yet decided, so the layout and sizes of the various components required can be judged. The components can be roughly sized so that their possible cost can be estimated by comparison with price data held for similar structures. This procedure can also reveal costs for items which might otherwise have been missed.
The third method, using published cost curves, is not very reliable, because the data on which such curves are based is so frequently absent, and virtually every civil engineering project has some unique feature substantially affecting its cost. Hence costs expressed per unit of size or output can vary greatly.
However, a cost curve can be used to show whether costs developed by the other methods seem realistic.
While any of the above methods will involve uncertainty, they can be useful in comparing different options for a scheme, provided uniform parameters are used. The final estimate of cost drawn up by the engineer should be based on current prices and include a substantial contingency sum. It need not include for possible future inflation of prices, because this is a matter for the employers financial advisers to deal with, but the basis of the estimate should be clear. The possible range of the cost should be shown; but whether the employer chooses to quote the highest or lowest estimate is up to him. Many a major project providing a major benefit (including the Channel Tunnel) would probably not have been built if the initial estimate quoted for it by the employer had not erred on the optimistic side.