Predicting required maintenance and repair funding based on standard facility data elements



Journal Title

Journal ISSN

Volume Title


Texas Tech University


Government entities and educational institutions have billions of dollars invested in facility portfolios designed to supply services to those that they support. Maintaining these portfolios requires continuous investment to keep them viable in order to meet their intended mision. In the past fifteen years, owners of these portfolios have realized that the facilities have degraded to the point that they may not be usable, they may require a significant investment to return them to full service, and they require a continuous financial commitment to maintain them. Both government and educational institution managers have realized that they have allowed this situation to occur due to chronic underinvestment in annual maintenance. Now they are faced with a large backlog of deferred maintenance and potential loss of mission.

This research investigates the underlying cause of chronic underfunding of the annual maintenance and repair of large facility portfolios, reviews the related literature for existing methods for estimating annual maintenance and repair funding, and develops a model that can be used by a facility portfolio manager based on facilty attributes commonly found in a condition assessment program. In addition, the research determines the effect on the developed model from varying facility portfolio size and facility model types, and compares the developed model to three models most often cited in the related literature.

Using multiple regression analysis, a prediction equation has been derived for the research portfolio, and is found to have good correlation to one of the models cited in the literature. It does not have good correlation to two of the models cited in the literature. Further, the research found that "fine tuning" a prediction equation to a specific facility portfolio yields the best results, although a more generic model is useful for an order of magnitude estimate.