An Empirical Perspective on Building Durability


This paper presents results from a recent Athena Institute survey of the reasons for demolishing 227 commercial and residential buildings in St. Paul , Minnesota from 2000 to mid-2003.

Despite the dearth of statistical data for North America , there is an increasing tendency to make assumptions or claims about the relative durability of different structural materials. In an attempt to bring some facts to bear on this critical issue, the survey focused on the age of the demolished buildings, the main structural materials, and the reasons for demolition. When building condition was cited as a reason, the survey probed for details about specifics, such as the state of maintenance or defined structural problems. Seventy percent of the buildings were in the 51-100+ age category, but the remaining 30% were all less than 50 years old, with 6% in the 0-25 category. Respondents cited physical condition as a reason for demolition for only 31% of the buildings, with lack of maintenance cited as the main underlying problem for the majority of those buildings. In only eight cases was a specific problem with structural or other materials or systems cited.

The survey results challenge the emerging conventional wisdom about durability as a function of material selection, and shift the spotlight toward construction practices and maintenance. The study also supports the view that we should do more to develop building systems that are flexible, and that can be readily deconstructed for reuse in different locations if future land use is in question for economic or other reasons.


Very little published, statistical data exists regarding the actual service lives of buildings, despite the relevance of this kind of data to decisions about financing, insurance, depreciation, and other aspects of the development process. At the same time, there is an increasing tendency to make assumptions or claims about the relative durability of different structural materials. In fact, based on the results shown in Figure 1 from surveys of North American architects, structural engineers, builders, and developers, it is clear that many practitioners in the building industry believe that buildings last a long time, and that longevity is a function of the structural materials used in their construction. [1,2]

In an attempt to bring some facts to bear on this critical issue of durability, the Athena Institute undertook a survey of buildings demolished in the City of St. Paul, Minnesota, during the period 2000 to mid-2003. Survey data was collected for 227 buildings, of which 135 were commercial and the remainder residential. We collected information about the age of the buildings, the main structural materials and the reasons for demolition. When the building condition was cited as a reason for demolition, additional details were gathered to probe what aspects of the building condition were most associated with its demolition.

Our hypothesis was that no relationship exists between the use of specific structural materials and building service life. Based on anecdotal evidence, we further hypothesized that buildings are often demolished before the end of the useful life of the structural systems. In other words, we felt that building industry beliefs that some structural materials last longer than others are most likely confusing how long a building could last with how long it is actually kept in service. In fact, a few previous studies indicate that service lives of most buildings are probably far shorter than their theoretical maximum lives. For example, a large study of U.K. residential buildings found 46% of demolished structures fell in the 11-32 year age class [3]. Another large study of office buildings in Japan , found the typical life span to be between 23 and 41 years [4]. It seems clear that these buildings were demolished for reasons that had little or nothing to do with the materials used in their construction.


Average Expected Service Life for Non-Residential buildings by Primary Material



The City of St. Paul was selected as the survey location for two reasons: the city was likely to have a mix of structural materials in its non-residential buildings; and the state of Minnesota is currently undertaking a large building database project, which offered the potential for additional data in the future. We obtained city demolition permit records for the years 2000 through mid-2003, which provided contact information for the building owners and demolition contractors. The City's database includes both residential and commercial properties, including garages. For the purposes of this study, the garages were excluded, leaving a total of 302 records. We eventually obtained data for 227 of the buildings.

The survey respondents were either private sector building owners or representatives of the City of St. Paul in the case of City-owned buildings, particularly the Code Enforcement Office. The respondents were therefore decision-makers or their advisors, and presumably knowledgeable with regard to the reasons for demolition. Private building owners were first contacted with a mailed questionnaire, followed up with a second mailing and then telephone interviews as necessary. In the case of buildings owned by the City, the survey questions were all answered by telephone.

The questionnaire asked for the age class of the building in 25-year increments, the primary structural material of the building (concrete, steel or wood), and the reasons for demolition (area redevelopment, too expensive to maintain, etc.). When the building condition was cited as a reason, the survey probed for details. The questionnaire is included as an appendix to the full survey report, which can be viewed and downloaded from the Institute's web site at

The analysis focused primarily on buildings with discrete structural systems — wood, steel or concrete — as opposed to a mixed system — steel and concrete, for example — in order to better isolate the building longevity implications of different structural materials. We examined results for all buildings types (excluding garages) and for commercial buildings alone.

Survey Results

Age Profile and Structural Systems

As indicated in Figure 2, 70% of the 227 buildings were in the 51-100+ age category, with 51% in the 76 and over bracket. The remaining 30% were all less than 50 years old, with 6% in the 0-25 category. Wood buildings accounted for two-thirds of the buildings in the survey, which is not surprising given the geographic region. Concrete buildings accounted for 25% and steel structural systems for 4%. The remaining 5% used mixed structural systems. Figure 3 shows the age distribution by structural material.


Percentage of surveyed buildings by age (years)


Figure 3

Age by structural material

Houses accounted for more than half of the sample, which partly explains the prevalence of wood structures. The longevity of houses is likely driven by different factors than those which affect the service lives of non-residential buildings. For example, older homes have characteristics valued by many people, and older homes are often more affordable. We therefore looked separately at the results for the 105 non-residential buildings.

Figure 4 shows a much younger age profile for the non-residential buildings, with a concentration in the 26-50 year group.



Distribution of Non-Residential Buildings by Age Class
Excludes 2 buildings of unknown age)


In Figure 5, the age distribution for non-residential buildings is shown by structural material for the 94 buildings that had one primary structural material as opposed to a mixed system — 54 concrete, 10 steel and 30 wood buildings.

These results are in sharp contrast to the impressions held by design professionals (Figure 1). Not only is the age profile of demolished non-residential buildings much younger than assumed, but the relationship between age and structural material is also quite different. About 56% of the concrete buildings were 26-50 years old at demolition, while 63% of the wood buildings were older than 50 years, and the largest group of those fell in the 76-100 year age class. Meanwhile, 80% of the steel buildings were 50 years or younger, with half of those no more than 25 years old.

Reasons for Demolition

Looking again at the full sample, the four main reasons for building demolition given by the survey respondents were “Area redevelopment” (35%), “Building's physical condition” (31%), “Not suitable for anticipated use” (22%) and “Fire damage” (7%). Figure 6 shows these results for the full sample.

Lack of maintenance was cited as the specific problem for 54 of the 70 buildings where ‘Building's Physical Condition' was given as the reason for demolition. However, lack of maintenance primarily relates to the building envelope or shell, roof systems for examples, and a specific problem with structural or other materials or systems was cited in only eight of those cases. All eight had foundation problems, along with other concerns, and all but two (one of which was of unknown age) were in the 75+ age categories.


Distribution of Non-Residential Buildings by Age Class and Structural Material

(excludes buildings with mixed structural systems)



Reasons for Demolition

(number of buildings)


Figure 7 segments the distribution over the top four demolition reasons by structural material. Surprisingly, given that they are the youngest in the data set, most of the steel buildings were demolished because they were no longer suitable for their intended use or due to their physical condition. For the wood buildings, physical condition was the dominant reason. This almost certainly is due to an age effect, since “physical condition” as the main reason for demolition predictably correlates with age of the building (Figure 8), and the wood buildings are the oldest in the data set.


Distribution of Buildings by Material and Top Four Demolition Reasons



Age Class Distribution of Buildings Demolished for ‘Physical Condition' Reasons

The Figure 7 pattern is similar for the ‘area redevelopment', ‘suitability for future use' and ‘fire damage' categories when residential buildings are removed from the analysis. In the case of the ‘physical condition' category, steel is the most dominant structural system, followed by wood and concrete, which are close to the same levels. This is not a surprising result given the younger age profile for commercial buildings and the dominance of steel and concrete structures in the younger buildings (Figures 4 and 5).

SUMMARY AND Discussion

The findings of the survey support the starting hypothesis. Although the results do not prove that there is absolutely no relationship between the use of specific structural materials and building service life, the results certainly indicate a much weaker relationship than is generally assumed. The results also indicate that buildings are often demolished before the end of the useful life of the structural systems. Relatively young buildings are demolished because they are not suitable for a future use or because of redevelopment patterns, irrespective of the materials used in their structural systems. Indeed, the structural materials generally presumed to have the greatest durability are the most prominent in these premature demolitions.

When physical condition appears to be the reason for demolition, lack of maintenance comes to the fore as the primary contributing factor. There could be a variety of reasons for the lack of maintenance, an aspect that we were not able to probe in this survey. However, irrespective of the reasons, it is clear that maintenance is a key issue, and it is fair to speculate that well-constructed and well-maintained buildings are equally durable no matter what structural systems are used.

Moreover, with only a few exceptions, the lack of maintenance did not appear to translate directly into problems with the basic structural systems. In the case of the eight buildings for which specific structural problems were cited, there was a range of reasons including cracked and crumbling foundations, ground settling, poor roofs, and too much past remodeling. In only two cases, one of unknown age and one in the 76-100 year category, were specific problems with wood structural systems cited.

These findings challenge the tendency in green building circles to advocate the use of ‘durable' materials to create structures that will have a long service life. Instead, they support the idea that the longevity of a building may have much more to do with its adaptability to others uses than to the use of any specific material. The results also indicate that we should be much more concerned with the potential relocation of buildings and/or the reuse (not recycling, but reuse) of entire components or systems, at least in areas where land values or other pressures are more likely to lead to redevelopment.

It is clear that the matter of building longevity and material durability has many facets, and we need to be more precise about what we mean when we use terms such as ‘durable'. As noted earlier, the terms ‘durability' and ‘longevity' are not synonymous and are too often incorrectly used as if they are interchangeable. As well, we often encounter statements about durability that are really statements about maintenance requirements, in the sense that one material or system may require less maintenance to provide the same service life as another material or system. Our choices, in that case should take into account the entire range of environmental effects associated with the original production, maintenance, and eventual disposal of one material or system versus another.

These are matters that obviously go beyond the study presented here. However, the survey results highlight the fact that durability, like so many aspects of sustainable building, is not a simple matter. To achieve our ultimate sustainability goals, we have to encompass all facets, starting with the terms we use. And to be fully cognizant of the economic and other pressures that will affect the future use of buildings without regard for the good intentions of their designers. The Athena Institute has now launched a major follow-up study, currently in the exploratory phase, to examine a broader range of material durability and building longevity issues. That study may involve additional surveys in other climatic regions as well as more in-depth assessment of underlying causal relationships and of the ultimate fate of the building materials. Whatever the eventual findings of studies of this nature, it is clear that flexibility, adaptability, and design for deconstruction must be as much a part of the sustainability lexicon as the words durability and longevity.


•  Gaston, C., Kozak, R., O'Connor, J., and Fell, D., 2001, “Potential for Increased Wood-use in North American Non-residential Markets,” Forintek Canada Corp.

•  O'Connor, J., Fell, D., and Kozak, R., 2004, “Potential for Increased Wood-use in North American Non-residential Markets – Part II,” Forintek Canada Corp.

•  DTZ Pieda Consulting, 2000, “Demolition and New Building on Local Authority Estates,” Department of the Environment, Transport and the Regions.

•  Yashiro, T., Kato, H., Yoshida, T., Komatsu, Y., 1990, “Survey on real life span of office buildings in Japan ,” Proceedings, CIB 90 W55/65 Symposium.

Customer comments

No comments were found for An Empirical Perspective on Building Durability. Be the first to comment!