Phase(s): Final Design
Alternatives to traditional low bid procurement are increasingly being pursued by public agencies that want to ensure successful, fast-track implementation of major projects. Next to the RFP process, design-build may be the most common method of negotiated procurement. Tri-Met has used a third method of non-traditional procurement called Construction Manager/General Contractor (CM/GC) to construct light rail facilities. Not as well understood as other methods, it is not as frequently used.
CM/GC is a middle ground negotiated procurement method. Several of the features of Tri-Met’s program help define this contracting approach:
Before awarding CM/GC or other non-traditional construction contracts, Tri-Met must demonstrate the rationale for procuring work through non-low bid methods. Under Oregon law, a public hearing must be held. An exemption from competitive bidding is allowed if the public agency finds (a) the contract award would be unbiased and not encourage favoritism and (b) substantial cost savings are likely.
Portland Tri-Met adopted CM/GC extensively to construct its $350 million Interstate MAX light rail line. Approximately $110 million of a $235 million construction program is under CM/GC contracts.
Awarding CM/GC contracts prior to start of construction allows for valuable contractor input during final design in the form of constructability reviews and second party cost estimates. Tri-Met found the assistance invaluable for reconciling conflicting capital cost estimates and in value engineering practicable cost savings that would keep Interstate MAX within the $350 million budget.
Establishing a Guaranteed Maximum Price (GMP) with the CM/GC contractor before construction actually begins provides greater certainty about the final project costs. Typically many quantities and work elements are negotiated as lump sum costs and only if there are significant changes in design or scope occur would these prices change. Change orders are still allowed under the CM/GC contract but are expected to be fewer and lower cost on average. Reducing the owner’s claims risk was especially important for Interstate MAX where the construction budget was constrained and costly claims during the course of a contract would threaten Tri-Met’s ability to complete the project as planned.
If the project owner and CM/GC contractor cannot agree on a GMP, the owner retains the option to seek alternative bids. This provides one level of owner leverage over the contractor when negotiating construction costs. Tri-Met did not see the need for such recourse, concluding that the contractors’ prices were reasonable.
Competitive bidding of subcontractor work elements, which are under the contract management of the CM/GC contractor, is another level of control on contractor pricing and provides an indicator, albeit indirect, of construction market conditions. These CM/GC-subcontract relationships are open to review by the project owner. In negotiating the GMPs for the two Interstate MAX CM/GC contracts, Tri-Met was able to inspect contractor pricing data, including subcontracted items. Tri-Met also had control of pricing through the development and approval of contracting plans that identified what work would be self-performed and what work would be subcontracted by the CM/GC contractor. These plans are intended to allocate work to the most capable, available firms, thereby controlling not just the overall price but also the work schedule, quality, safety management, and community impacts.
The approach is not without potential disadvantages, including the fact that the owner has less leverage over the contractor when pricing construction activities since there are no other competitors at this point. The owner may not have available complete information on construction market conditions, information that is otherwise revealed through the competitive bidding process. Tri-Met, fortunately, had recent bid histories on many similar work items from the Westside-Hillsboro project and could obtain additional information from the Portland Streetcar and Airport MAX construction programs. In accordance with the terms of the contracting plan, the CM/GC contractor obtained low bids for certain segments of work. Tri-Met could have requested that the contractor obtain bids for additional work items but concluded that was not necessary.
The performance to date of Tri-Met’s two major CM/GC contracts for the Interstate MAX light rail line is exemplary from the perspective of keeping the project on schedule and maintaining good quality but mixed from a budget standpoint.
The negotiated GMPs ended up significantly above the owner’s design engineers’ cost estimates, upwards of 15 to 20 percent depending upon the time the estimates were prepared (e.g., 60 percent and 95 percent of final design). Tri-Met concluded, however, that the contractors’ cost estimates were accurate and reasonable when quantities were reconciled and other oversights in the engineers’ designs and cost estimates were corrected. Tri-Met independently reconciled its final cost estimate with that of the CM/GC contractor. The negotiated GMP was actually less than Tri-Met’s final estimate and within the contingency allowance of earlier estimates.
Further, change order trends through a year’s worth of construction are positive, with owner-directed changes running under 3 percent of the GMP construction cost. Tri-Met is forecasting the project to come in at or under budget in large part due to the benefits of CM/GC. Should this trend continue, the small dollar value of changes will bolster the argument that the CM/GC method helps identify and reconcile price differences in the owner and contractor’s understanding of work scope, technical requirements, quantities, and risk allocation prior to initiating construction.
CM/GC warrants consideration along with design-build and RFP selection when legal and institution frameworks allow negotiated procurement of construction projects and when CM/GC’s advantages can be best realized. The advantages are having a single contractor responsible for a large share of the work, a contractor who can effectively manage a variety of subcontractors on a complex project. The types of projects where CM/GC is most relevant are those with multiple interfaces (agency, public, business) and high public expectations; where contractor input during design would be useful; and where the owner wants to retain control through final design and to have continuing influence during construction. Both traditional design-bid-build and design-build itself are less well suited for procurement under these circumstances.
The owner must be experienced in construction and have the internal resources to check contractor cost proposals and negotiate a contract with confidence that the agreed upon terms are fair and reasonable.
Note: This discussion draws heavily on information provided by Portland Tri-Met. The FTA would like to acknowledge the contributions of Don Iwin, Construction Manager, and Neil McFarlane, Executive Director, Capital Projects and Facilities Division.