A new mathematically based model for forecasting construction project expenditure pattern

This research explores the viable use of'S' curve models for simulating construction project expenditure patterns. The research establishes a new approach to the classification of the current models, based on the analysis of their structure. This requires the identification of the parameters comprising the mathematical expression of the model, the attributes of the suhject matter contained in the data, and the mechanism facilitating the link between these two components. The latter is established through the analysis of the 'shape' of the construction project expenditure profiles which have been shown to contain certain characteristics. Some of these characteristics comply with the law of growth and are 'general' for all construction projects while other characteristics are 'specific' and distinguish one project from another. Subsequent to the above, the research identifies the need for the development of a new improved mathematical model. This process involves the development of a new mathematical expression comprising three modules; the Control module which maintains control over the properties of the underlying expenditure pattern, the Distortion module which facilitates simulation of any distortion on the underlying pattern, and the Kurtosis module which controls the rate of growth at the beginning and end portions of the expenditure curve. The above modules are simulated by an exponential and two fourth degree polynomial equations respectively. The simulation is carried out by identifying the values of the variables relating to the characteristic 'shape' ofthe expenditure pattern, namely, the dependent-variables. These variables are provided either directly by the user or they can be estimated by utilising the appropriate dependent-variable model. These models which are the product of statistical analysis of a large quantity of hl.;torical data, have been developed for a variety of categories of construction projects. This research, therefore, proposes a new mathematical model for simulating construction project expenditure patterns, thereby, provide an effective financial management tool. The model combines many of the advantageous features contained in a variety of existing models, namely, ease of use, user involvement and appreciation of the underlying logic of forecasting, and ability to generate a forecast at an early stage.