building performance engineering during construction - CiteSeerX

BUILDING PERFORMANCE ENGINEERING DURING CONSTRUCTION T. Michael Toole1 and Matthew Hallowell2 Published in the Proceedings of the 2005 ASCE Construction Research Congress, April 5-7, San Diego, CA. Iris Tommelein, Editor.

ABSTRACT Exploratory research was performed into the hypothesis that a substantial number of building performance engineering tasks on design-bid-build projects are typically provided by entities associated with the construction phase, not with the AE of Record. The project specifications, drawings, and submittals were analyzed and employees from key project participants were interviewed for five $5-45M building construction projects. Twenty four building performance engineering tasks were required by the project specifications to be performed by entities associated with the construction of the buildings. Many of these tasks had to be completed by a licensed engineer retained by subcontractors or material suppliers. This large number of delegated design tasks indicates the increasing fragmentation of the design and construction process, which has implications for design-bid-build, design-build and lean construction processes. The shift in engineering from AEs to construction entities is proposed to result from decreased design time and profit margins, attempts to minimize liability, increased specialization and increased prefabrication. The compelling preliminary findings suggest additional research is warranted. Keywords: Fragmentation, Engineering design, Design and construction process, Design delegation. INTRODUCTION For the past half-century, the design and construction process for buildings has been understood as a three-step process: Architects and engineers (AEs) design the entire buildings, bids are solicited from contractors, and contractors construct the building. The process has traditionally been viewed as being linear and compartmentalized, that is, the engineering design is completed by the AE before the construction begins, and the contractor (in conjunction with specialty subcontractors and material vendors) merely implements the

1

Ass’t Professor, Civil & Env. Engineering, Bucknell University, Lewisburg, PA 17837. Phone 570-5773820. Fax 570-577-3415. [email protected] 2

Graduate Student, Civil & Env. Engineering, Bucknell University, Lewisburg, PA 17837. [email protected].

1

AE’s design. In other words, it has been assumed that engineering and design are only performed by AEs, that general contractors and subcontractors only provide construction services, and that material vendors only manufacture and deliver materials. This paper represents exploratory research on whether the conventional notion of the design-bid-build process is accurate today. The paper’s hypothesis is that a substantial portion of building performance engineering tasks is typically provided by entities associated with the construction phase, not with the AE of Record. The empirical research being conducted by the authors indicate that engineering during construction extends well beyond the example of structural steel connections being designed by the steel fabricator, not by the structural engineer of record. In other words, exploratory research indicates there has been a gradual, but significant, shift in the engineering of buildings from recognized design professionals to entities associated with the construction of designs, namely constructors and material component manufacturers. The idea of construction-related entities becoming increasingly involved in engineering design may not seem like news to the reader. After all, this is the premise underlying designbuild, which is becoming increasingly common in all market segments. Rather, the paper’s primary contribution to the literature is that this shift is shown to be occurring in traditional design-bid-build projects. The paper’s contribution also stems from the belief that other researchers have not investigated the shift in multiple portions of construction nor fully considered its implications for design-build or lean construction. Before proceeding with the paper, it is appropriate to first provide some basic definitions. Engineering tasks are defined as discrete processes that require the application of mathematics, science and engineering principles. The simplest, but not all inclusive, test is whether the task involves the consideration—whether explicit or implicit—of forces and stresses. The output or work product resulting from task completion are decisions, analysis, drawings and specifications. There are two types of engineering tasks. One set will be referred to as building performance engineering tasks. Such tasks affect the functioning of the completed building; that is, the building will not perform as intended if the task is not completed effectively. In other words, building performance engineering tasks have consequences for the building’s end user. The second set of engineering tasks will be referred to as construction engineering tasks. This set of tasks only affects the characteristics of the construction process (duration, cost, inherent safety level, etc.). They therefore affect all of the firms and individuals associated with the construction of a building, but not the end users. Differentiating between the two types of engineering tasks is important because constructors have long been responsible for construction engineering tasks, the reason being that such tasks are interdependent with constructors’ choices of means and methods. This paper will focus on building performance engineering tasks. These tasks are hypothesized to be increasingly performed by construction-related entities rather than by entities associated with the AE. The paper is divided into four sections. The next section reviews the relatively small body of literature relevant to the paper’s focus. Next, the methodology and findings of the

2

exploratory empirical research performed by the authors are discussed. The paper then theorizes why the shift in engineering might be occurring. The final section theorizes the implications for design-bid-build, design-build and lean construction. EXISTING LITERATURE While the authors are not aware of any literature that explicitly addresses the hypothesized shift in engineering from design professionals to constructors, three sets of literature address, sometimes indirectly, the roles of the various entities in the design and construction process. These three sets are guidelines and articles published by professional trade organizations on their specific trade, scholarly articles on design-build construction, and scholarly articles on lean construction. Well known examples of the first set of documents include the American Concrete Institute’s Manual of Concrete Practice and the American Institute of Steel Construction’s Code of Standard Practice for Steel Buildings and Bridges and Manual of Steel Construction, which are intended to establish custom and practice within their subindustries. Given the detail and size of these documents and the fact that both design engineers and constructors play major roles in the membership of the organizations, one would expect to find language indicate the respective roles of design professionals and constructors. Several sections (namely 318 and 347) within the Manual of Concrete Practice (ACI 2003) clearly indicate that the contractor is required to perform construction engineering tasks including design of formwork, bracing and shoring. The Manual does not, however, indicate that contractors are expected to engineer the reinforcing steel within the concrete, which is building performance engineering. Similarly, the AISC documents clearly assign all construction engineering tasks to constructors and shop drawings to fabricators, but do not identify building performance engineering tasks that are to be performed by contractors. An AISC document called Construction Management of Steel Construction (Mrozowski et al 1999) dictates, in detail, the steel design and construction process and indicates that all building performance engineering is to be performed by the Structural Engineer of Record. Several articles in practitioner-oriented trade journals also indicate custom and practice roles for the steel construction phase. Two articles on steel design roles published in AISC’s Modern Steel Construction (Carter 2000, Andrews 2003) contained no references to construction-related entities undertaking building performance engineering. A second set of literature, including Quatman (2000), relates to design-build construction. This literature points to the many problems that frequently result from the disjointed nature of the design and construction processes in design-bid-build projects. The premise of designbuild is that by having one legal entity responsible for both design and construction, seasoned construction professionals may interact with design professionals to ensure the design is practical and cost effective. The premise of design-build is not that constructors would assume the role of performing building performance engineering design, however. Ideally, construction entities would work with designers, not substitute for them. A third set of relevant literature is associated with lean construction, an operations management-based approach involving radically transforming the design and construction

3

process (Ballard 2000, Ballard and Zabelle 2000, Tsao et al 2004). As is true for literature on design-build and construction management, this literature identifies the serious problems that result from a fragmented design and construction process. Some of the principles underlying the Lean Project Delivery System include set based design, sharing complete information, simultaneous product and process design, and work structuring, all of which are not possible with the current design-bid-build or design-build processes. Although lean principles have been the focus of journal papers and annual conferences for over a decade, the implementation of lean construction in the U.S. has been limited to date. (An anonymous reviewer pointed out that the Last Planner and other lean construction tools have been adopted in projects located in the United States, United Kingdom, Brazil and Chili.) METHODOLOGY The exploratory research performed to date included five case studies, all institutional building projects built in Central Pennsylvania over the past four years. Four of the projects were associated with Bucknell University (an engineering building, a building housing the offices and laboratories associated with geology and psychology, a dormitory, and a recreation and athletics center) and the fifth was a large addition to a community hospital. The projects ranged from $5 to $45 million. While all of the projects were design-bid-build and were ostensibly designed by a traditional architectural and engineering firm, some diversity existed between the project delivery methods. Two of the projects involved an atrisk construction manager, one involved an agency construction manager and the final two involved traditional general contractors. For each project, the specific engineering tasks performed by entities other than the AE were identified primarily by reviewing the project technical specifications. Specifically, each specification section was read closely to identify language requiring the contractor to perform design and/or engineering, often through an engineering consultant, subcontractor, or material vendor. (Example text is provided in Table 1.) The following information was recorded for each task: ♦ The text in the specifications that referred or alluded to the engineering task. ♦ Which entity (GC, engineering consultant retained by GC, subcontractor, or material vendor) was required to or likely to perform the task. ♦ Whether the specifications required the task be performed by an individual with specific qualifications, such as a professional engineer (not associated with the AE of Record). ♦ The result or work product of the task (calculations, shop drawings, decisions, etc.) ♦ The engineering principles underlying the task (engineering mechanics, hydrology and hydraulics, geotechnical, electricity, etc.) Project contract drawings and submittals were then analyzed to confirm building performance engineering tasks that the technical specifications required to be performed by entities other than the AE. Engineering responsibilities identified through these archival

4

documents were confirmed through interviews with key personnel associated with the owners, AEs, general contractors and subcontractors on the projects. FINDINGS The documents associated with the five projects and the interviews of the key player involved with the projects indicate that the number of specific portions of building projects that have building performance engineering performed by entities associated with the construction phase is significant. Table 1 includes twenty-four distinct construction components that were designed by entities not associated with AE for at least one of the five projects. Many of the items had building performance engineering performed by construction entities for at least two projects. Although these twenty-four items are associated with only eight phases of construction identified in the far left column, the majority are provided by a different firm. For example, while both concrete formwork and concrete reinforcement shop drawings are provided by a professional engineer retained by the concrete subcontractor, the engineering required for structural steel connections, steel joists, steel trusses and steel decks are likely to be provided by engineers employed or retained by each of the separate fabricators/manufacturers. This point is relevant because, as will be discussed shortly, the greater the number of entities involved in engineering, the more fragmented the process. Table 1 provides sample text from one of the project’s technical specifications to illustrate the requirement for most of the twenty-five items listed. Note that the majority of items explicitly state that a professional engineer be responsible for the design associated with the item. Furthermore, most items specify that the PE must be registered in the state in which the project is built. This point is important to dispute the assertion that Table 1 mostly contains manufactured items that have always been pre-engineered. In fact, many manufactured products or assemblies that clearly require engineering in the product design process were omitted from Table 1. For example, windows are manufactured items that require engineering as part of the product design process. But it would be inappropriate to include them in Table 1 because integrating the specified windows into the project does not require project-specific engineering. FACTORS DRIVING THE SHIFT If Table 1 provides tentative indications that significant portions of commercial buildings are being engineered by entities not associated with the AE, it is appropriate to briefly consider possible explanations for why this shift is occurring. It is the authors’ perceptions that the shift in engineering from AEs to construction-related entities reflects four factors that are already known by many industry professionals. One factor is the increasing pressure on designers by owners to complete the design in the shortest possible time and for reduced profit margins. It is not difficult to reduce design fees below the traditional percentages of construction if portions of the work are only taken just past the concept design stage. A second factor is the increased effort by design professionals to minimize liability. Increased litigation associated with design and construction projects has resulted in risk

5

management pervading the entire design process. It is easier for an AE to deflect liability when the AE can point to the project specifications as explicitly assigning design responsibility for a portion of a project to an entity associated with the general contractor. A third hypothesized factor is the increasing specialization of the engineering and construction process. On the construction side, it is rare that general contractors will selfperform much of the work. Instead, commercial building projects typically involve forty to seventy subcontractors and material vendors. The increase in number of specialists has also increased on the design side. Perhaps fifty years ago, it was not uncommon to have one design firm design an entire building project. Today, it is recognized that the architect retains design specialists such as structural engineers, geotechnical engineers, MEP engineers, and site/civil engineers. Often, design firms specializing in acoustics and vibrations, earthquake engineering, clean rooms, manufacturing, etc. are also often part of the AE team. Perhaps AEs are increasingly delegating the design of specific portions of the project to construction-related entities is because it is easier to offload a portion of the design than it is to manage so many specialists. The fourth factor that may be driving the shift in engineering is prefabrication, that is, the shift in location of significant portions of the work from the project site into the factory. Accompanying this shift in location is an increase in the complexity of prefabricated systems. That is, prefabricated assemblies often are based on systems design principles and feature innovative materials (Toole 2001). As such, many prefabricated systems, particularly involving HVAC, fire protection and electrical work, are proprietary to some degree. Because these systems still need to be installed by constructors, it may be easier for the AE to postpone the design of work associated with these systems to the construction stage. IMPLICATIONS The apparent shift in engineering from design professionals to constructors and manufacturers has implications for the alternative contractual arrangements on design and construction projects. If fragmentation is defined as the extent to which a thing or process is broken up, the tentative findings clearly indicate the traditional design-bid-build process is becoming increasingly fragmented. Not only are more entities providing building performance engineering than before, engineering entities are also becoming increasingly split between three different groups: design professionals who produce contract documents, engineering consultants hired by constructors, and component manufacturers. There has traditionally been insufficient communication between these groups, so the increased fragmentation will likely lead to increased inefficiencies in the design and construction process. Two specific examples of project inefficiencies were mentioned during the interviews conducted for the research. One example involves steel connections. Structural engineers hired by architects often size the wide flange beams and columns but assign the design of connections to steel fabricators. Notes on the drawings or the specifications often direct the fabricator to “max out the connections,” that is, to design the connections such that the beam will fail before the connections fail. Yet because structural steel members are commercially

6

available in discrete sizes, the members are typically overdesigned. The structural engineers’ directions to “max out the connections” rather than designing connections to meet the actual design loads result in additional inefficiency in the structural steel frame. Another example of inefficiency came from a project in which the specifications did not require the mechanical contractor to have the shop drawings stamped by a professional engineer. The AE’s contract drawings and specifications provided partial design of an expansion loop needed in one part of the building. The contractor then chose a combination of pipe, fittings and layout that he thought would result in acceptable performance. Several months after the building was completed the expansion loop failed. An investigation revealed that both the designer’s and contractor’s design decisions contributed to the failure and both parties agreed to split the cost. The reference to the lack of sufficient communication between design and construction entities and the two specific examples mentioned above would seem to point towards the inherent superiority of design-build over design-bid-build. In addition, one could argue that the tentative findings reported in this paper indicate that design-bid-build is slowly being transformed into design-build because construction-related entities are increasingly being associated with the design of specific parts of the building. However, one could also argue that the findings that an increasing number of entities are involved in the design process suggest that design build will never achieve its potential. Design-build is based on the sufficient exchange of information between design and construction entities. Yet there will always be at least some communication barriers between individuals in different firms. The barriers that design-build is supposed to break down between design and construction will be replaced by barriers between the sets of entities performing design and construction of specific portions of the completed project. The negative implications of the tentative findings for design-bid-build and design-build suggest that it is time for a radical reengineering of the design and construction process, as proposed by the lean construction literature. Yet the findings of increasing fragmentation also suggest problems for a lean construction delivery system. Design professionals and manufacturers provide only building performance engineering. Constructors, through permanent in-house staff and through hired engineering consultants, provide some building performance engineering and all construction engineering. The lean construction literature suggests that the design and construction process will be most efficient if the two types of engineering are coordinated and simultaneous. Yet the findings of this exploratory research indicate the engineering and construction industry is moving away from the ideal process, not towards it. That is, the increasing fragmentation of the engineering process due to increasingly technical specialization and complex prefabrication will make it increasingly intractable to implement lean construction principles. ADDITIONAL RESEARCH NEEDED This paper has reported the tentative finding, based on a limited exploratory empirical study, that building performance engineering is increasingly being performed by entities associated with construction, not with AEs. This tentative finding, while interesting and potentially

7

important to the industry, is limited by the fact that the sample of projects studied was a convenience sample of relatively few, geographically close and similar building projects. Empirical research on a larger and more diverse sample is clearly warranted. If this shift in engineering from design professionals to construction entities is indeed occurring, two critical research questions arise. One question is whether the shift has resulted in engineering being performed by individuals or organizations that do not possess the qualifications traditionally associated with engineering design. Based on the fact that the vast majority of engineering tasks performed by construction-related entities in this study were required to have a professional engineer supervise the design, the apparent answer is no. However, several interviewees hinted that the qualifications on smaller projects, especially design-build projects with tight timelines, are not always as strict. Another issue that warrants additional research is the risk that one or more engineering tasks will fall through the cracks on individual projects. This possibility seems real given two factors. One factor is that the fear of litigation has caused entities to attempt to reduce the boundaries of their responsibilities. Both design and construction entities may attempt to avoid legal responsibility for an engineering task unless forced by a contract to assume it. A related factor is that the primary instruments for forcing construction-related entities to perform building performance engineering design are project specifications. Specifications, however, are often the weakest aspect of contract documents because creating project specifications is a mundane, thankless and underpaid activity. It seems likely, therefore, that some AEs may produce plans that in effect delegate engineering of some portions of the project to construction entities, yet fail to adequately assign this responsibility within the specifications. The two issues raised above and the implications of the findings for alternative design and construction processes suggest the apparent shift in building performance design from AEs to construction entities is important and warrants further study. REFERENCES American Concrete Institute (2003) Manual of Concrete Practice. Farmington Hills, MI. American Institute of Steel Construction (2000). Code of Standard Practice for Steel Buildings and Bridges. Chicago, IL. American Institute of Steel Construction (2003). LFRD Manual of Steel Construction, Third Ed. Chicago, IL. Ballard, G. (2000). “Lean project delivery system.” Downloaded on 11/11/04 from http:www.leanconstruction.org/lpds.htm.” Ballard, G. and T. Zabelle (2000). “Lean design: process, tools & techniques.” Downloaded on 11/11/04 from http:www.leanconstruction.org/lpds.htm.” Carter, C. J. (2000). “Revised code brings together full design team.” Modern Steel Construction. 40(8). Johnson, A. (2003). “It doesn’t have to be that way!” Modern Steel Construction. 43(1). Mrozowski, T., M. Syal, and S. A. Kakakhel (1999). Project management of steel construction. American Institute of Steel Construction, Chicago, IL.

8

Quatman, W. (2000). Design-Build for the Design Professional. New York: Aspen Publishing, Inc. Toole, T. M. (2001). “The technological trajectories of construction innovation.” Journal of Architectural Engineering. 7(4): 107-114. Tsao, C. C., I. D. Tommelein, E. S. Swanlund and G. A. Howell (2004). “Work structuring to achieve integrated product-process design.” Journal of Construction Engineening and Management. 130(6): 780-789.

9

Table 1: Portions of projects in which entities other than the AE performed buildingperformance engineering design Phase of Construction

Portion of Building

Comments and Sample Specification Text

Foundation or Superstructure

Concrete formwork1

03300 Concrete Work 1.04.A.8. Submit a letter of certification by a Professional Structural Engineer registered in Pennsylvania hired by the Contractor that all formwork and reinforcing shop drawings will be prepared under his direction…


Concrete reinforcement

See sample text provided for concrete reinforcement.


Precast concrete

03410 Structural Precast Concrete - Plant Cast B. Engineering Responsibility: Engage a fabricator who assumes undivided responsibility for engineering structural precast concrete units by employing a qualified professional engineer registered in the state of Pennsylvania to prepare design calculations, fire-resistance calculations, shop drawings, and other structural data.

Superstructure

Structural steel connections

05120 Structural Steel 1.3.B. Engage a fabricator who utilizes a qualified professional engineer to prepare calculations, shop drawings and other structural data for structural steel connections. 05120 Structural Steel 1.5. B. Design connections not detailed on contract drawings under the direct supervision of a Professional Structural Engineer” licensed in Pennsylvania.

Superstructure

Steel joists

PE registered in local state required.

Superstructure

Steel trusses

05400 Cold-Formed Metal Framing C. The roofing system as indicated on the drawings shall be designed by a Professional Licensed Engineer to verify the framing assemblies’ ability to meet or exceed design requirements as required by local codes and authorities. The calculations shall indicate compliance with structural loading requirements indicated, indicate connections, attachment to structural steel and bridging.

Superstructure

Steel floor and roof deck

05311 Steel Roof Deck 1.6. Design deck layout, spans, fastening, joints and openings under the direct supervision of a Professional Structural Engineer” licensed in Pennsylvania. 05313 Steel Floor Deck 1.3 Design metal deck in accordance with the Steel Deck Institute Design Manual. 1.6 Design deck layout, spans, fastening, joints and openings

10

under the direct supervision of a Professional Structural Engineer” licensed in Pennsylvania Superstructure

Steel stairs and handrails


Superstructure

Steel (permanent) ladders


Superstructure

Wood Glulam beams

06185 Structural Glued-Laminated Timber 1.4.A. Engineer, fabricate, and install structural glulam timber to withstand structural loads shown on Drawings without exceeding the allowable design working stresses according to AITC 117—DESIGN.

Superstructure

Wood trusses

06192 Metal-Plate-Connected Wood Trusses Structural Performance: Engineer, fabricate, and erect metalplate-connected wood trusses to withstand design loads within limits and under conditions required. Engineering Responsibility: Engage a fabricator who uses a qualified professional engineer registered in the state of Pennsylvania to prepare calculations, Shop Drawings, and other structural data for metal-plate-connected wood trusses.

Building envelope

Curtainwall and storefront

05400 Cold-Formed Metal Framing 1.4.A. Structural Performance: Provide cold-formed metal framing capable of withstanding design loads within” deflection limits indicated.

Building envelope

Metal and composite wall panels

07411 Manufactured Wall Panels 2.02 Design criteria includes quantitative engineering criteria such as maximum deflection of spans, flexural strength, thermal transmittance, air infiltration, water penetration, fire endurance.

Building envelope

Metal roof panels

07412, Manufactured Roof Panels. Detailed quantitative engineering criteria provided similar to that provided for 07411. PE registered in PA stamp required.

Specialties

Louvers and vents

Specifications in one project only found to provide quantitative engineering performance criteria and to require design shop drawings be submitted by a PE licensed in the local state.

Specialties

Exterior sun control devices


Windows and Doors

Automatic swing doors


Fire Protection

Fire sprinkler

13930 Sprinkler Systems 1.4.C. Submit professionally signed and sealed shop drawings and hydraulic calculations… to the local jurisdiction and to the owner’s insurance underwriter.

11

Fire Protection

Fire alarm

16720 Fire Alarm and Smoke Detection System Part 1 of the specifications typically contain many pages of detailed performance design criteria.

HVAC

Equipment vibration isolation

15070 Vibration Isolation 1.4. Design Data. Provide engineering calculations indicating maximum room sound levels are not exceeded.

HVAC

Steam/Hydro nic System

02552 Hydronic Distribution 1.3 Design Requirements (Quantitative engineering design criteria provided, such as service temperatures, etc.) 1.4.D. Design Data: Submit computer analysis of each system layout performed by piping system manufacturer to determine stresses and anticipated movement of service pipe. Include design assumptions.

HVAC

Pipe expansion

15050 Basic Material and Methods 1.13 “Pipe expansion design criteria. Base expansion calculations on 50oF installation temperature to 210 oF for hot water piping, 140 oF for domestic hot water and temperature of steam pressure for steam systems plus 30% safety factor.”

HVAC

Automatic temperature control

15975 Automatic Temperature Controls 1.3.B. “The engineering, installation, calibration, software programming and check out necessary for complete and fully operational….”

HVAC

HVAC balancing and testing

01660 Testing, Adjusting, Balancing and Commissioning of Mechanical Systems 1.06.B. “The TAB work performed by the TAB contractor shall be under the direct supervision of a registered Professional Engineer, a full-time employee of the TAB contractor.”

1

Note: This item is clearly associated with construction engineering, but it is also related to building performance engineering. If the formwork is not engineered properly, deflections or movement during the curing process could lead to poor superstructure quality.

12