Cost-significance applied to estimating and control of construction projects

A study of the price distribution of bills of quantities has shown that about 80% of the value is attributed to about 20% of the items. It has also been shown that items of greater value than the mean value of all items contribute about 80% to the total bill value. This relationship, from which the principle of cost-significance derives, has been used to develop a new estimating technique - iterative estimating - which enables the prediction of the total value of a bill of quantities by pricing only about 30% of the items. Iterative estimating has been tested and shown to give accurate results using historical bills. The technique has also been tested in practice on six 'live' tenders. Iterative estimating produced tenders which were on average 0.7% lower, and more consistent than conventional tenders. Moreover, a saving of at least 50% in pricing time has been achieved using iterative estimating. A relatively stable relationship has also been shown to exist between any price level less than twice the mean price, and the value of items with higher prices. This relationship has made it possible for iterative estimating to be applicable not only at the mean price level, but also at all price levels less than twice the mean. The choice of the price level would depend upon the required level of accuracy, estimators' skill and experience, estimating turnover sought, and the time available to produce estimates. The effect of iterative estimating on final accounting has been considered and, using historical data, four methods of dealing with variations have been proposed and shown to yield very accurate long-term results. In the short term, errors of the order of + 4.5% in 2/3 of the times have been predicted. The use of the principle of cost-significance in construction control has also been considered. In order to identify the favourable features of control systems, a review of both integrated and non-integrated cost and time control systems proposed for and/or practiced in the construction industry has been carried out. Based on the findings of that review, combined with the application of the principle of cost-significance, a new control system has been proposed. The system uses an operational model of construction, with a built-in bias towards the controllable cost-significant items of work, to integrate the various functions of control. The main features of the system are a hierarchical project breakdown structure, integrated cost and time control through the use of project network and sub-networks, diagnosis of progress information to detect variances from plan and causes of such variances, prediction of the effect of current progress situation on future progress, and the provision of advice on optimum courses of remedial action. The proposed system is in skeleton form and needs further development.