Wood use in nonresidential buildings: opportunities and barriers.

Abstract

The North American nonresidential construction sector represents a substantial market for structural building materials, yet it is relatively untapped by wood. This study explored wood use and perceptions held by architects and engineers with respect to the structural use of wood in nonresidential buildings, using an extensive mail survey and a series of specifier focus groups in select geographic regions. Several clear barriers for wood emerged across code, technology transfer, and research and development categories. Key problem areas identified were fire-related building code limitations, wood's cost-competitiveness with steel, wood's design difficulty, and poor training for designers and wood tradespeople. Recommendations for addressing these impediments in both the short and long term are offered.

**********

An examination of wood's potential as a nonresidential structural material in North America was performed from June 2000 to April 2001. (1) The purpose of this study was to 1) gather information on current wood usage; 2) determine what types of buildings are currently popular in this market; 3) assess the attitudes and knowledge levels of designers regarding wood; 4) understand the people and factors that influence structural material selection; 5) identify barriers to wood in this market; and 6) search for opportunities for wood in this market.

At $292.3 billion in 2000, the total value of new nonresidential construction in the United States was 110 percent that of new residential construction (USBC 2002), representing a substantial market for all construction materials. Yet, according to a recent study on nonresidential wood product usage, only 1.5 billion board feet of lumber and 1.2 billion ft.[.sup.2] of structural panels were used in that sector in 1995 (McKeever and Adair 1998). This is approximately equivalent to only 10 percent of the total wood usage in the new residential construction sector and has been in decline since 1985 (WPC 1987).

Nonresidential buildings are all those not meant for habitation: offices, schools, hospitals, warehouses, etc. Of course, wood is not always an appropriate choice for these types of buildings; indeed, building codes restrict the structural use of wood within this market to smaller buildings. However, code-related issues are only estimated to limit wood's potential nonresidential market share to approximately 50 percent (Goetzl and McKeever 1999). Given the small share held by wood to date, this represents tremendous room for growth.

There have been relatively few studies recently that assess the market potential for wood products in nonresidential construction. Two notable examples are the works of McKeever and Adair (1998) and Kozak and Cohen (1999). McKeever and Adair (1998) found that the potential for increased wood use in this sector is substantial, including the use of wood in combination with steel and concrete construction types. They concluded that the use of wood in nonresidential construction is a function of building size and type, among other factors. For example, there are possibilities to increase wood use in smaller nonresidential structures like stores, office buildings, and schools. Kozak and Cohen (1999) used survey methodologies to not only assess the market potential for wood products in the nonresidential sector, but also to address issues pertaining to why wood was or was not being used in this context. According to their results, there exist many possibilities for producers of wood products to make market inroads into this sector, especially in "wood-friendly" regions of North America. However, marketing efforts must be targeted with varying promotional messages in low-, moderate-, and high-use wood segments.

The present work further studies why wood is not a preferred structural material in the nonresidential construction market and explores means for increasing its use in this context. The results of previous work (Spelter and Anderson 1985, McKeever and Adair 1998, Kozak and Cohen 1999) required updating and were not strongly conclusive regarding the selection process and the hurdles for wood. This is largely due to the difficulty of performing scientifically valid research on the decision-making process of designers. It is difficult to pinpoint the material selection decision to a single point in time, not to mention to clearly identify all the decision makers and their reasons (Mackinder 1980). In addition, decision making in design can be described as a non-linear and non-optimized process. This lack of a systematic approach to the design of buildings makes study of the process a substantial challenge.

Methods

The methodologies used in this study employed various survey techniques in three phases. Data was collected by means of 1) a pre-survey focus group; 2) a mail survey of structural engineers and architects; and 3) a series of post-survey focus groups.

Each is discussed in turn. It should be noted that the decision to use a multipronged approach is based largely on lessons learned from similar studies undertaken previously (e.g., Kozak and Cohen 1999). Researchers felt that many of the issues relating to wood use and specification were too complicated and difficult to capture using traditional quantitative techniques alone. Thus, quantitative and qualitative methods were used in concert in order to address some of these issues.

First, exploratory research was conducted as a means of gathering preliminary information and guidance for the development of the survey instrument. In particular, researchers wished to better understand the many complexities and nuances of the material specification process prior to crafting and implementing a mail questionnaire. A focus group session with five architects, four structural engineers, and one code consultant was held in Vancouver in June 2000. The focus group participants were also asked to act as survey pre-testers prior to the questionnaire being mailed out.

Upon completion of the exploratory focus group session, mail surveys were designed and sent to 6,000 architects and structural engineers throughout the United States and Canada in the fall of 2000. The survey instrument was based on previous work by Kozak and Cohen (1999) and adhered largely to methods prescribed by the Tailored Design Method (Dillman 2000). The choice to use self-administered mail questionnaires served the need to reach a geographically dispersed audience in a cost-effective manner.

The survey itself was 12 pages in length, consisting of 7 sections: 1) building design; 2) structural material use; 3) wood use in nonresidential buildings; 4) learning about building materials and design; 5) design processes and philosophy; 6) environmental issues; and 7) workplace information.

Most questions related to a comparison of wood, steel, concrete, and masonry in structural applications. Within each section, questions were ordered from general to specific as a means of minimizing bias. In order to maintain respondent interest, questions took on a variety of forms and measurement scales. Open-ended questions were used throughout to collect qualitative information surrounding more in-depth and complex issues related to the design process. Quantitative information was collected using simple dichotomy questions, determinant choice questions, checklist questions, rank order questions, numerical rating scales, and constant sum scales. This information formed the basis of the statistical analysis that was carried out in this study.

In total, four rounds of mailings were used for data collection, with each pack-age being personally addressed to individual architects and structural engineers to ensure a higher response rate (Dillman 2000). Mailings were sent out at 2-week intervals. The first mailing consisted of a cover letter explaining the objectives of the study, a number-coded survey, and a business-reply envelope. The second mailing was a reminder postcard. The third mailing was a second (replacement) survey with cover letter, and the final mailing was a second postcard.

The population under study included all North American architects and structural engineers. However, the survey was primarily targeted to design practitioners involved in nonresidential buildings four stories or less. (2) The mailing list was obtained from a local list broker as well as various professional associations, and represents the most comprehensive sample frame of North American architects and structural engineers that could be generated. All lists were randomly sampled to obtain the samples stratified by both profession and location. It should be noted that stratification sampling was used as a means of increasing the precision of estimates, creating data sets that were easily comparable between strata, and ensuring that there was adequate representation from each geographic region of North America.

Returned surveys were then entered into a spreadsheet and analyzed using a variety of statistical and graphical tools. Concurrently, five post-survey focus groups were held across North America to further explore some of the trends and questions that were emerging in the preliminary analyses of the survey responses. These focus groups were held between November 2000 and February 2001, in San Francisco, Toronto, Boston, Atlanta, and Calgary.

Every effort was made to ensure that the focus groups followed established protocols with respect to duration, the role of the moderator, the number of participants, and selected topics (Churchill 1992, Edmunds 1996, McDaniel 1996). For each focus group, an equal mix of architects and engineers was sought, with the only criteria being that each participant's current workload involve nonresidential construction and that he or she have had at least a few years of experience as a design professional. The number of specifiers in the focus groups was typically between 8 and 10, recruited with the assistance of local contacts in each city. The focus groups were held in conveniently located downtown areas during lunchtime and were no longer than 2 hours in duration. Each focus group session was audiotaped (with permission) for use in the transcription and summarizing process.

Results

Respondents

Of the 6,000 surveys mailed out, 626 were returned, for a total response rate of 11.6 percent (conservatively assuming that 10% of the respondents were not reachable). Of these, 442 respondents completed the questionnaire in its entirety, for a revised response rate of 8.5 percent. Reasons for not completing the survey related to a lack of knowledge or experience in designing nonresidential buildings four stories or less. In ascending order of importance, the most commonly cited reasons for not responding were:

* design workload is devoted to a combination of buildings and applications that were not small- to medium-scale nonresidential buildings;

* design workload is devoted to single-family dwelling units/duplexes;

* design workload is devoted to non-building structures;

* design workload is devoted to buildings five stories or more.

The presence of non-response bias in this investigation was measured by comparing the response patterns of early versus late respondents on key variables under study. In the absence of information from architects and structural engineers that did not respond to the survey, late responses (i.e., those that responded only after the fourth reminder) can be used as a valid proxy for non-respondents (Armstrong and Overton 1977). That said, non-response bias was measured by comparing structural material use patterns between the early and late respondents. Specifically, a series of two-tailed z-tests ([alpha] = 0.05) were performed between early and late respondents on the proportions of wood, steel, concrete, and masonry used in nonresidential construction. No statistically significant differences were detected, indicating a lack of non-response bias and allowing for inferences onto the populations of North American architects and structural engineers to be made.

Figure 1 shows the proportion of surveys sent to each geographical region, compared to the proportion of completed surveys received from each region. In general, Canada (especially western Canada) is overrepresented in the dataset and the United States (especially the U.S. South) is underrepresented. From what we know about wood use in these two regions, there is the possibility that overall survey responses are somewhat biased to the "wood friendly" designers. However, no significant differences were observed by constructing 95 percent confidence intervals around the proportion of surveys returned for each region and comparing them to the proportion of surveys sent out, further indicating a lack of non-response bias.

Of the total respondents, 55.8 percent were architects, 42.6 percent were structural engineers, and 1.6 percent claimed to practice both. Fully 95.9 percent of the respondents were male, with a mean age of 52 years. On average, they had been practicing architecture or structural engineering for over 26 years, worked in offices with less than 10 designers (three-quarters were self-employed), and had been part of over 200 nonresidential building projects in their careers.

Characterization of practice

Of the buildings four stories or less that were designed and built by respondents, the average designer's workload was devoted mostly to structures that were nonresidential in nature (61.3%), followed distantly by detached homes, multifamily homes, mixed commercial/residential structures, and other types of buildings. The most common types of nonresidential structures designed were industrial buildings (Fig. 2). These were followed by offices and educational buildings, both of which ranked ahead of stores/malls and hospitals/care facilities. Public buildings, religious buildings, restaurants, commercial/residential combinations, hotels/motels, recreational buildings, entertainment facilities, and farm structures each had less than 10 percent of the design share. There seems to be a fairly strong relationship between the types of buildings designed and design preferences. For example, the most preferred building types are offices, followed by industrial buildings. A relatively high degree of preference is also given to educational buildings, followed by stores/malls, hospitals/care facilities, and religious buildings.

[FIGURE 2 OMITTED]

In terms of building size, the most common structures designed by respondents were one story high and less than 10,000 ft.[.sup.2] in area (nearly 50% in each category). Significant proportions of two story buildings were also built, as were buildings with areas exceeding 10,000 ft.[.sup.2] in area. However, comparatively fewer buildings three stories or more were built. The respondent architects and structural engineers preferred to work on buildings that were one or two stories in height and between 10,000 ft.[.sup.2] and 50,000 ft.[.sup.2] in area. These preferences were relatively strong with regards to building height, but less so for building area.

Wood use

Table 1 shows that wood has an average use of 26.8 percent of nonresidential projects four stories or less, which is a fairly sizable market share (among respondents) as a structural material in nonresidential construction across North America. This is slightly less than the share of steel (34.9%) and exceeds the share of concrete and masonry (10.4% and 12.8%, respectively). (3) It should be noted that these proportions are reflected in, although slightly contradicted by, the average number of buildings designed per year (6.4 steel buildings versus 6.9 wood buildings). This anomalous result indicates that wood may be used in more buildings, but the buildings are, in general, smaller. Lastly, these average proportions do not take material combinations into account, indicating that the percentage for each structural material is somewhat underestimated. Wood use is comparable to the proportion of buildings designed using structural materials in some combination (12.9%). The most common combinations are steel/masonry (26.3%), steel/concrete (16.2%), masonry/wood (17.2%), and steel/wood (13.1%). Of the buildings that they have designed out of materials other than wood in the past, respondents indicated an average of 29.4 percent could have used wood as the main structural component.

When structural material use is broken down geographically, some very interesting, albeit expected, trends emerge. Wood use is highest in western North America: both Canada and the United States. Its market share becomes marginally lower as one moves east and south across the continent, eroded primarily by the use of steel, as well as concrete and masonry to a lesser degree. An interesting substitution pattern exists in the western part of North America where wood has a relatively healthy share of the market. In Canada west/north, steel nonresidential structures are commonplace, exceeding the use of wood, in fact. In the U.S. West, this is not the case. The use of steel is comparatively low, but the use of materials in combination is high.

This analysis shows that opportunities exist to improve the competitive position of wood products in North America, but regional differences must be taken into account. These results were echoed by the discussions that took place in the post-survey focus groups. Not surprisingly, regions in which steel and/or concrete dominate are generally less amenable to the use of wood in a nonresidential context. One of the major reasons for this seems to be related to the abundance of steel and concrete in certain regions and the availability of skilled construction labor for these structural materials.

Using rankings to query designers about their preferences, structural material use was also broken down by application. In general, steel is the most preferred material in structural applications, while wood has, by far, the highest preference ratings for interior trim/detail applications. Wood is also fairly extensively used in interior partitions (equal to steel) and roof systems (slightly less than steel). It is used to a lesser degree in floor systems (slightly less than steel and concrete) and exterior wall systems (less than masonry and steel).

[FIGURE 3 OMITTED]

Respondents were asked to what degree they intend to continue using wood products. Only 11.2 percent of the specifiers surveyed intend on using more solid sawn wood in future designs, although 52.1 percent intend on using more engineered wood products. However, 71.0 percent and 45.1 percent of the specifiers surveyed intend on using the same amount of solid wood and engineered wood products, respectively, in future designs.

Favorability to wood use

In developing competitive strategies for the wood products industry, the salient question that needs to be answered is, "How favorable are architects and structural engineers to the structural use of wood in nonresidential applications?" The survey gauged favorability to wood use by asking specific questions regarding comfort levels, design philosophies, and wood use. Even though 74.3 percent of the specifiers surveyed feel comfortable designing with wood in this context, 41.4 percent currently do not use wood and 21.9 percent feel that their "design philosophies" are not even conducive to wood use. Over 34 percent of the designers surveyed like seeing wood used in this context only sometimes, with 17.6 percent not liking to see it used at all, and 15.5 percent saying that wood cannot deliver adequate performance. As shown in Figure 3, favorability varies by region.

Not surprisingly, wood use is favored in buildings that have historically often been made of wood: smaller buildings in general, as well as farm structures, religious buildings, restaurants, and offices (Fig. 4). Unfortunately, designers do not generally like wood for some of the more frequently designed building types. For example, industrial buildings have the highest proportion of designers' workloads but are second only to hospitals in negative favorability rating for wood.

In terms of building size, specifiers are favorable to wood use in one- and two-story applications or buildings less than 10,000 ft.[.sup.2] in area, and moderately favorable to wood use in three stories or areas of between 10,000[.sup.2] and 50,000 ft.[.sup.2] In general, they are not favorable to wood use in larger structures, i.e., buildings exceeding three stories or 50,000 ft.[.sup.2]

Qualitative results and discussion

Open-ended questions were used to elicit responses with respect to why specifiers were or were not favorable to the use of wood in various nonresidential applications. Respondents identified many positive and negative attributes of wood that influence favorability to the material. On the positive side, wood is familiar and easy to work with, has a low cost, and can be assembled by readily available trades. "Ease of use" was rated as wood's greatest positive attribute (Fig. 5).

Wood has a higher environmental friendliness ranking than the alternatives, although this is by a surprisingly small margin (Fig. 6), primarily due to concerns over forest management. The implication of this lower-than-expected ranking is probably not substantial at the moment, as environmental issues were not identified by respondents as a major factor in material selection. However, this is anticipated to change due to rapidly escalating interest among designers and the public in "sustainable building design."

[FIGURE 5 OMITTED]

Wood's appearance is a positive attribute; however, this is only recognized in the small proportion of buildings that expose their structure. Where wood is chosen for its appearance attributes, as in some heavy timber

buildings, the value of the appearance attribute is recognized by the project team and the associated cost premium is accepted. This motivation to use wood is well understood and requires no further study. The concern in this study is with applications where the structure is hidden, in which case wood's "beauty and warmth" have little value.

The remainder of this discussion focuses on the negative attributes of wood identified by respondents, which can be viewed literally as marketplace impediments. The open-ended questions and the follow-up focus groups were critical to understanding the complex nature of these barriers. Code issues top the list of barriers, followed by structural performance concerns, durability problems and design difficulty (Fig. 7). For discussion purposes, it is useful to break the barriers down into four categories: codes, costs, performance and infrastructure of the design and construction industry.

Building codes. -- Building codes are typically viewed by both the wood industry and by designer practitioners as responsible for one of the largest hurdles to wider uses of wood in nonresidential construction. Codes provide a mechanism for local jurisdictions to regulate construction activity and to protect public health and safety. Because fire is one of the biggest safety risks in buildings, codes apply major restrictions on the use of wood as a structural material due to the fact that it is "combustible." Codes contain tables that dictate allowable building heights and floor areas as a function of structural system and building use. Structural use of wood is limited to smaller buildings.

As expected, building codes were identified by respondents as the primary barrier for wood, especially with respect to fire (codes related to structural loads were much less frequently identified as a barrier for wood). This problem has three aspects: 1) codes directly restrict wood use based on building end use and size; 2) codes indirectly restrict use when designers are inhibited to use wood due to difficulty of applying codes; and 3) codes allow fewer choices (i.e., less design flexibility) in wood systems compared to steel and concrete. A potential fourth road block due to codes could be a perception that code restrictions are worse than they really are, leading designers to assume wood is not an option for a given project. Many survey respondents shared comments illustrating the magnitude of the code problem: "I truly doubt there are any other issues worthy of consideration."

[FIGURE 7 OMITTED]

Costs. -- For long-span or complex structures, steel is considered more cost effective than wood. The timber products (solid or engineered) required for long spans are not competitive with steel systems. This cost concern is reflected in the "economics" bar in Figure 7. In addition, and perhaps more importantly, wood design is not cost effective for the engineer: "We do not make money designing in wood due to complex detailing issues." Many of the respondent engineers highlighted difficulties of working with wood. This is not due to a lack of comfort with the material, but a lack of design support to help streamline the engineering tasks. More generic (not proprietary) design data, span tables, off-the-shelf connections, and pre-engineered systems would save the engineer time, thereby providing a better financial incentive. Similarly, existing design manuals may not be aimed at optimizing the engineer's time: "The (wood design guide) has gotten more and more complicated, making the design/fee ratio untenable." The engineer's cost concern is reflected in the "difficult to design" bar in Figure 7.

Performance. -- Many respondents, primarily the engineers, cited a variety of structural concerns about wood (the "performance" bar of Fig. 7). Span limitations were the most frequently mentioned. Other comments suggested a lack of faith in wood's general strength, its ability to withstand lateral loads, and its lack of stiffness. In addition, the "durability," "quality," and "shrinkage" bars of the graph also reflect performance concerns.

Infrastructure. -- A substantial and difficult group of hurdles can be described as inherent to the entire design and construction culture. These are not problems with wood as a material. Rather, these issues reflect a mismatch between wood and the nonresidential construction climate.

Lack of skilled wood trades was identified as a large problem that affects an engineer's confidence: "It appears that the persons marketing wood do not appreciate the huge difference between the regimented use of steel and concrete construction on the one hand and the free-for-all field construction of wood." Although "labor skill" (Fig. 7) is outweighed by other concerns, this singular item appeared frequently in respondent comments as well as during focus groups in all geographic areas.

Although three-quarters of respondents feel comfortable designing nonresidential buildings in wood, most designers learned little or nothing about wood in school (Fig. 8). This was frequently cited as an issue, along with a lack of post-professional education in wood and other forms of knowledge transfer to practicing design professionals, especially in comparison to the activities of competing material industries. It is difficult to reconcile this seeming contradiction between a high level of comfort alongside complaints about lack of education. Comments from the focus groups and in the open-ended survey questions may suggest that designers are comfortable with their general knowledge, but not with their familiarity with wood in this particular end use; "increase awareness" was a frequent recommendation. The potential benefits of increased education are unclear: "I think it would be great to have more of a balance and know more about timber coming out of school, but I don't know how much influence that may have on selection." Information support while on the job is a widely cited problem: "There doesn't seem to be ONE agency you can go to." There is a "seriousness" accorded steel and concrete, in contrast to wood, which has a knowledge base founded on conventional wisdom and folklore. This leads to image problems for wood and a good deal of misinformation.

[FIGURE 10 OMITTED]

Industry inertia is a well-known problem. There is little incentive in the construction industry to stray from established practice, because there are substantial cost and risk implications with no clear payoffs. The building industry is change resistant for good reason; the high price of failure, low profit margins, and tight schedules are all inhibitors to innovation or other deviation from standard practice. There is no reason to change what is seemingly working fine: "In my opinion, wood use does not need to increase." "Contractors have the 'I never did it this way before' mentality." Designers, builders and owners all operate with a substantial level of inertia.

Promotional activities need to target the decision makers, but who selects structural materials? As with many design decisions, this one has multiple influences. Only 27.2 percent of respondents feel that this choice is made by a single individual; the remainder clearly demonstrate that material selection is a multi-disciplinary process (Fig. 9). On average, architects are considered to have the largest say, followed closely by engineers. Architects lead by an even greater margin according to the respondents who identified only one party as responsible for the material selection. The focus groups strongly contradicted this data, identifying the engineer as more influential except in special cases where the structure is being used for aesthetic effect. This is probably an illustration of the difficulty in trying to characterize the design process in a general manner, as previously discussed.

It would be important not to overlook the objections that structural engineers have regarding wood in nonresidential construction, as they are significantly less favorable to wood use than architects (particularly for larger buildings) (Fig. 10) and possibly have a greater voice than suggested by the quantitative data. Engineers have very clearly defined issues with wood that include a high level of design inconvenience, a low level of confidence in the material, and an extremely low level of faith in the ability of contractors to implement the design correctly: "The reason an engineer prefers to work with steel is because it's well-coded and the fabricators and erectors are reliable, predictable, and accurate. Basic connections are understood and built correctly with little effort from the engineer." Irrespective of the exact role that structural engineers may play in material specification, they are key participants in the design and construction process.

The relative influence of the architect versus the engineer may actually matter little, if the choice is essentially pre-determined based on precedent or availability: "The traditional building materials for (my) area are concrete and steel for commercial and institutional construction. The wood is primarily used for housing." "I think it has a lot to do with availability of materials and then it always comes down to time. Time and money." Decisions in the building industry are very much influenced by what is considered standard practice. This is typical behavior for a risk-averse industry.

Both the survey and especially the focus group data highlighted the growing influence of owners and builders. The relationship between owner and builder is often established before the design professionals are hired, and the preference of the builder may be a more dominant factor than suggested by the quantitative data: "More and more work comes with a contractor attached ... they tell you what's cheapest for them to build and that's the way it usually ends up."

Conclusions and recommendations

Wood is a material that enjoys a high level of familiarity and comfort in North America, which is an excellent first step in gaining acceptance for new structural uses in nonresidential buildings. Wood has a long tradition as the dominant material in North American vernacular buildings due to its availability, ease of use, low cost, a knowledge base founded on conventional wisdom, and its forgiveness for lack of precision in application. A short-term course of action would capitalize on those benefits and aim for incremental improvements in already-established applications for wood. For example, wood has proven successful in smaller nonresidential buildings that can use light framing as in residential design: medical clinics, strip malls, daycares, schools, and so forth. Promotional activities should target a greater market share of these types of buildings, perhaps starting in "wood-friendly" regions.

Three very clear trends emerged as disincentives for designers to specify wood products and systems in nonresidential buildings. The primary hurdle seems to be related to real or perceived code issues, especially those addressing fire safety, with many designers stating that wood is precluded for most projects simply due to code limitations. Secondly, costs seem to be a major driver in material specification decisions for designers. Specifically, respondents highlighted total installed costs as important, stating that, even in cases where wood materials are inexpensive, the choice of wood often means that time will be added to a project due to increased complexities with respect to engineering and construction. Lastly, there seems to be a general lack of experience in using wood in medium- to large-scale applications, both on the part of designers and builders. This hurdle is exacerbated by the fact that there is little precedent for wood in certain types of nonresidential applications regardless of their size; examples include hospitals and big box warehouses. Moreover, there is inertia in the building industry whereby many of the players have little or no incentive to make changes from the standard practice of using steel and concrete.

Large nonresidential buildings, particularly those types that have rarely been built in wood in the past, present particular challenges for increasing wood's market share. When loads and spans increase, designers are more hesitant to use wood for various reasons. Contributing factors include particularly stringent code limitations for wood in many of these structures, a perception that wood systems are not cost competitive with steel and concrete for larger, long-span applications, and a belief that steel and concrete construction workers are more accustomed to the needs of engineered construction than those working in wood. Increasing market share for wood in larger nonresidential buildings therefore requires more than promotion. The hurdles for wood in these building types include technical barriers, codse and standards gaps, image problems, a construction culture that favors other materials, and educational shortcomings.

A long-term action plan would require work on several fronts simultaneously:

* Address true fire-related code restrictions either by advocating code changes to allow more wood, or developing wood systems that can meet existing code requirements for non-combustibility. Address perceived code restrictions or perceived difficulties in meeting code through better assistance or education to designers and code officials.

* Provide engineers with design guides, performance data, off-the-shelf connections, product standardization, and pre-engineered systems in order to simplify the job of engineering in wood and thus match the ease of designing with steel or concrete. One successful example of this approach is demonstrated by timber "design-build" firms. These businesses perform engineering and deliver the product to the site completely pre-manufactured and ready for bolting, exactly as with pre-engineered steel buildings.

* Develop easier and cost-effective long-span solutions. Unless wood can be cost competitive with steel, long-span buildings will not be a viable market for wood, except in the niche market of exposed heavy timber structures.

* Improve product quality issues of warp, durability, strength, and shrinkage, or change users' perceptions of those attributes and support users' ability to design around those circumstances.

* Provide specifiers with more training and inspiration in school and in practice, more information about products, more performance data, better-informed vendors, better design aids, and more design examples. This can be part of a broad marketing plan to expose specifiers to the nonresidential application potential for wood and to change the low quality/residential image of wood. In particular, provide engineers with the data that meets their requirements for assurances and precision to compete with steel and concrete.

* Improve the skill of wood tradespeople.

All of these activities can combine to transform wood's image from a solely residential material to one that is suitable for more than just houses. Wood has a low-tech image, which is certainly to its benefit in a residential market that does not rely on skilled labor for construction. This image also permits the existence of a large do-it-yourself industry. However, when competition with steel and concrete becomes necessary, a low-tech image is a hindrance. These other two materials are supported on a much firmer technical basis than wood, allowing them to easily meet the requirements for engineering and construction precision in the nonresidential and civil engineering sectors, where concrete and steel predominate. It has been suggested that wood's low-tech image may be due to a relative lack of extensive performance data (compared to steel and concrete) that would provide the necessary assurances to engineers--this is one gap that needs to be addressed in the long-term. A major dilemma for wood today is how the light-frame can evolve to meet the technical needs of architects and engineers while simultaneously retaining the beneficial low-tech imagery, which helps it dominate the residential market. In other words, the same material needs to be viewed as appropriate for both engineered and non-engineered construction.

Feasibility, priority ranking, and detailed recommendations for each of these long-term tasks will require further study. The survey and focus group methods employed here are useful in capturing a generalized picture of the material selection process; however, the design process is best understood in detail through case study observation. The authors intend to continue into a second phase of this project to gain a better understanding of the major parties involved and their interactions, as there remain some questions regarding the relative influence and relationship between architect, engineer, builder, and owner. This would help refine promotional targets and methods. In addition, a deeper understanding of the barriers will illuminate which problems require a technical solution and which may be more an issue of information transfer. Finally, further study may help indicate which building types within the nonresidential sector may be the best candidates for wood.

                         Sent (6000)  Completed (442)

Canada West / North       7.8%        15.0%
Canada Central           10.8%        12.6%
Canada East / Maritimes   6.4%         9.4%
US West                  25.9%        24.9%
US Midwest               12.1%        10.4%
US Northeast             14.5%        10.4%
US South                 22.5%        17.2%

Figure 1.--Regional distribution of surveys sent and completed.

Note: Table made from bar graph.

Table 1.--Predominant structural material in nonresidential buildings
four stories or less designed by respondents, by region and for all
regions.

       Region      Steel   Concrete  Wood  Masonry  Other  Combination
                                         (%)
Canada West/North  33.8    16.4      29.9    7.8     0.4      11.7
Canada Central     35.4    11.2      24.3   10.3     1.5      17.4
Canada East/
  Maritimes        40.3    10.1      27.5    5.1     1.3      15.7
U.S. West          31.1     9.2      31.5   11.2     0.1      17.0
U.S. Midwest       38.5     9.0      24.9   16.1     0.7      10.8
U.S. Northeast     40.5     8.6      28.5   14.4     0.4       7.6
U.S. South         32.8     8.8      27.7   21.9     1.0       9.3
All regions        34.9    10.4      26.8   12.8     0.7      12.9

Farm structures                 $5.2   1.2
Religious buildings             $8.1   1.0
Restaurants                        *   0.8
Offices                        $55.7   0.6
Indoor recreational buildings      *   0.3
Hotels/Motels                  $16.4   0.2
Stores/Shopping plazas             *   0.1
Entertainment facilities           *   0.1
Educational buildings          $54.9  -0.1
Public buildings               $28.3  -0.2
Industrial buildings           $33.3  -0.5
Hospitals/Care facilities       $5.2  -0.7

*Value not available, however stores and restaurants together are
estimated at a combined value of $61.5 billion.

Figure 4.--Average favorability to structural use of wood in various end
uses (-2 = not at all favorable; 2 = very favorable). Dollar figures
(US$) show the value, in billions, for each type of construction put in
place in the United States in 2000 (adapted from USBC 2002).

Note: Table made from bar graph.

Wood      7.0
Masonry   6.4
Concrete  6.1
Steel     5.7

Figure 6.--Average "environmental friendliness" rating for materials
(1 = unfriendly; 10 = friendly).

Note: Table made from bar graph.

Learned a little but needed more    45.0%
Learned a little but it was enough  16.3%
Learned a lot                       24.0%
Learned too much                     1.2%
Other                                2.3%
Learned nothing                     11.2%

Figure 8.--Average degree to which designers learned about wood and wood
systems at school. A total of 72.5 percent learned little or nothing.

Note: Table made from pie chart.

Architects            38.8%
Structural engineers  31.5%
Owners/Developers     17.4%
Building contractors   9.3%
Occupants              1.7%
Other                  1.2%

Figure 9.--Influence that various parties have on the material selection
process (does not include the 27.2% of respondents who believe a single
individual is responsible for the material selection).

Note: Table made from bar graph.

[c]Forest Products Society 2004.

Forest Prod. J. 54(3): 19-28.

(1) In this article, "North America" refers to the United States and Canada, but not Mexico.

(2) In general, four stories is the maximum height allowed for a wood building by North American building codes at the time of this study.

(3) Note that wood use is reported in terms of respondent use and not actual market share.

Literature cited

Armstrong, S.J. and T.S. Overton. 1977. Estimating nonresponse bias in mail surveys. J. Marketing Research 14(3):396-402.

Churchill, Jr., G.A. 1992. Basic Marketing Research. 2nd ed. The Dryden Press, Fort Worth, TX.

Dillman, D.A. 2000. Mail and Internet Surveys: The Tailored Design Method. 2nd ed. Wiley and Sons, Toronto, ON, Canada.

Edmunds, H. 1996. AMA Complete Guide to Marketing Research for Small Business. NTC Publishing Group, Chicago, IL.

Goetzl, A. and D.B. McKeever. 1999. Building codes: Obstacle or opportunity? Forest Prod. J. 49(9):12-22.

Kozak, R.A. and D.H. Cohen. 1999. Architects and structural engineers: An examination of wood design and use in nonresidential construction. Forest Prod. J. 49(4):37-46.

Mackinder, M. 1980. The selection and specification of building materials and components: A study in current practice and educational provision. Res. Paper 17. Institute of Advanced Architectural Studies, Univ. of York, York, UK.

McDaniel, Jr., C. and R. Gates. 1996. Contemporary Marketing Research. 3rd ed. West Publishing Co., Minneapolis, MN.

McKeever, D. and C. Adair. 1998. Wood products used in non-residential building construction, 1995. Wood Prod. Promotion Council, Tacoma, WA.

Spelter, H. and R.G. Anderson. 1985. A profile of wood use in non-residential building construction. USDA Forest Serv., Forest Prod. Lab., Madison, WI.

U.S. Bureau of the Census (USBC). 2002. Publication C30, Table 1. USBC, Washington, DC.

Wood Products Promotion Council (WPC). 1987. Wood used in non-residential construction, 1985. (now called the Wood Products Council). WPC, Tacoma, WA.

Jennifer O'Connor

Robert Kozak*

Christopher Gaston*

David Fell*

The authors are, respectively, Building Industry Advisor, Forintek Canada Corp., 2665 East Mall, Vancouver, BC V6T 1W5; Associate Professor, Dept. of Wood Science, Univ. of British Columbia, 2424 Main Mall, Vancouver, BC V6T 1Z4; Group Leader, Markets and Economics, and Market Researcher, Forintek Canada Corp. The authors are indebted to the Wood Products Council, the National Research Program at Forintek Canada Corp., and the Univ. of British Columbia, for their guidance and financial support of this project. This paper was received for publication in August 2002. Article No. 9534

*Forest Products Society Member.