Abstract

Public Buildings like universities and educational buildings are considered among the most challenging assets to maintain and modernize. Statistics show that the non-residential buildings are prone to a significant shortage in maintenance and rehabilitation expenditure. As per the Canadian infrastructures report card (2016), 42% of Municipal buildings rank between “fair” and “very poor”. In addition, according to the American society of civil engineering (ASCE) (2017) the school sector grade, which is the largest sector of educational buildings is D+ (poor condition). Thus, a maintenance optimization methodology is essential for these building types to appropriately plan for the maintenance of systems that are competing for limited funding. Upon reviewing the literature, a gap was revealed in the area of building maintenance, repair and rehabilitation, which is the lack of consideration of the space type inside the building and how different space types affect the maintenance planning process. Considering the fact that any building is composed of different space types having variable needs, requirements and functions that help in supporting the overall function of the building facility, therefore, not including the space types and functions as part of the maintenance plan would result in a loss in the overall functionality. Hence, the main objective of this research is to develop a maintenance optimization model which takes the space type into account and accordingly optimize the maintenance actions to be implemented inside a building.
To achieve the main objective the following sub-objectives were identified: 1) review the maintenance prioritization and optimization methods for buildings maintenance, 2)develop a deterioration models for building systems and 3) establish an optimization model for buildings maintenance. The methodology encompasses three main phases. First, a space-based condition assessment, is adopted as a part of this research to determine weights of the spaces and the different systems inside each space of the building. Second, a Weibull Distribution model, is utilized to predict the future condition of building systems inside each space by modeling their deterioration over the time. Finally, a Particle swarm optimization is employed to optimize the activity selection. A case study on educational buildings is leveraged to illustrate the applicability of the proposed model. Non-dominate solutions were established considering the defined constraints. The selected compromising solution result was 11.06 building condition with a total cost of $475,000. A Sensitivity analysis was conducted showing the impact of the systems service life on the space-based assessment and the maintenance cost. The output of this study is a framework that selects the best combinations of maintenance activities to be implemented inside a building, considering the varying space types and maintenance costs.