Application of Bromilow's Time - Cost Model to Predict Construction Project Duration for Multi-Storey Commercial Building Projects in Kampala Central

Abstract

Practitioners and researchers have long recognized the significance of time and cost of construction as crucial to project feasibility assessment and budget allocation decisions. Therefore it is imperative for clients and contractors to obtain reliable and accurate estimates of both time and cost at the planning phase of a project. This study examined the application of Browmilow’s TimeCost (BTC) model for the prediction of commercial building projects’ duration in Kampala central using a set of completed projects database. Linear regression was performed to investigate timecost relationship of the data. The result revealed a positive correlation between the variables with R and R2 values of 0.8427 and 0.7101 respectively. A log-log regression in the form of the BTC model was developed to determine the value of K which showed general level of time performance and β which demonstrated the sensitivity of time performance as affected by project size measured by cost. The result indicated a K value of 2.25 which is very low compared to values obtained in previous studies for other categories of construction. The β value of 0.523 indicates a greater influence of project complexity on the prolongation of time of construction. The model demonstrated a good fit to the data with R value of 0.6832 and R2 value of 0.4668. The model also showed a weak prediction efficacy with a mean absolute deviation (MAD) of 503.8, mean squared error (MSE) of 253207.8 Mean Absolute Percent Error (MAPE) of 46.25% over a test sample which is considered inadequate for application in practice. The potential limitation of the model is that it does not consider other factors apart from cost that have effect on construction time.