Durability and the factor method of ISO 15686-1

Abstract

Interest in the service life of buildings and components and the future need for maintenance is growing, and the methodology for service life planning is starting to appear in standards. The standard ISO 15686 prescribes a factor methodology for deciding about the expected service life of a component with a prescribed probability of earlier failure. The methodology is based on extensive knowledge about materials and building technology, and examples of the use of this methodology on building components are of value for the discussion on service life planning. Results from a condition survey of timber windows are used to define the service life and effect of the factor method in the Standard. Wood is a very variable material, and for wooden windows maintenance plays a key role in the durability of the component. In this instance, condition assessment is, therefore, not sufficient to estimate the service life. Information about former maintenance and refurbishment is also needed. The results from a condition assessment and the house owner's answers to a questionnaire are combined to evaluate the estimated service life of the windows. The factors for the standardized method for estimating service life with a given confidence limit are shown to have a wide range in values that give considerable uncertainty to the practical use of the standardized methodology. For example, it is shown that the estimated service life with an 80% confidence limit is much lower than the average service life. Thus, the question is posed whether there is any meaning in estimating the lower limit of service life.

Keywords:

• building inspection
• building stock
• components
• condition assessment
• durability
• maintenance
• performance
• service life
• whole life costs

Keywords:

• inspection des bâtiments
• parc de bâtiments
• composants
• évaluation de l'état
• durabilité
• maintenance
• performance
• durée de vie de services
• coûts pour l'ensemble de la durée de vie

The condition survey, the results of which this paper is based upon, was done by The Icelandic Building Research Institute (Jónsson and Marteinsson, 1997) with financial support from the State Housing Board and The Icelandic Research Council. The support and the help from the author's colleague Benedikt Jónsson is greatly appreciated.

Notes

¹Properties of materials are seldom exactly the same for all specimens tested, i.e. there is a spreading in the values. A graph showing the number of specimens with each value of a property is called the frequency distribution. A property value limit that is chosen such that a specific number of specimens will have a better value is termed a ‘confidence limit’, i.e. 80% of all specimens will have a better value than the 80% confidence limit.

²Unpublished measurements from Iceland.