A model for forecasting the running costs of commercial buildings in Sri Lanka

Abstract

Conventionally, early-stage investment decisions on buildings were purely based on initial capital costs and simply ignored running costs and total lifecycle cost. This was basically due to the absence of estimating models that yield running costs at the early design stage. Often, when the design of a building, which is responsible for 10 to15% of its total cost, is completed, 80% of the total cost is committed. This study, therefore, aims to develop a model based on building characteristics, which enable an early-stage determinant of running costs of buildings, to predict the running costs of commercial buildings. While positioning this research in a positivist research paradigm, a survey research strategy was adopted along with a questionnaire survey and a documentary review for data collection. The study involved 135 respondents for identifying factors influencing the running costs of commercial buildings and 46 commercial buildings were accessed to collect running costs and building characteristics data. A Pareto analysis, relative significance index (RSI), bivariate correlation analysis, regression modelling, and hedonic price imputation index were performed on collected data. The RSI confirmed, eight categories of running costs factors: environmental, maintenance, managerial, building characteristics, building design and construction defects, social, tenant, and political, respectively. Among 48 sub-factors identified, the study confirmed, natural deterioration, failure to identify the true cause of a defect, lack of preventive maintenance, insufficient fund, building services, building age, occupancy, vandalism by tenants, misuse of property, expectation of tenants have a substantial impact on running costs. According to the Pareto analysis, utilities (39%), services (19%), admin work (14%), and cleaning (8%) are four main cost constituents, responsible for 80% of running costs, which can be represented by highly correlated building characteristics of the number of floors (0.950), building height (0.945), and building size (0.943). Approximately 94% of the variance in annual running costs/GIFA (sq.m) is expressed by variables of net floor area, the number of floors, and working hours/day together with a mean prediction accuracy of -1.6%. The index constructed revealed, there is an increasing trend of running costs of commercial buildings in Sri Lanka, in office and bank buildings particularly, over the last recent years. Further, a noticeable increase in running cost can be observed during the first quarter while there is a slight reduction in the second quarter of each year. Early-stage supportive running costs estimation model proposed by the study would enable construction professionals to benchmark the running costs and thereby optimise the building design. The developed hedonic model illustrated the variance of running costs concerning the changes in characteristics of a building. While facilitating early-stage running costs estimation, the study findings collectively support building owners, designers and constructors, and facilities managers to optimize the in-use phase costs of a commercial building in terms of designing and constructing cost-effective, sustainable facilities by altering building characteristics during its’ design stage as well as carefully considering the significant running costs factors during its’ in-use phase. Keywords: Building characteristics, Commercial buildings, Hedonic price imputation