For many years, the Minnesota Department of Transportation (MN DOT) traditionally contracted projects utilizing the low-bid process in conjunction with the design-bid-build delivery method. Under this practice, MN DOT designs the project, either with its own or consultant staff, and then puts the plans and specifications out for bidding. Typically, estimated quantities are already computed and bidders enter their pricing for various items and respective quantities. On the letting date, the bidder with the lowest pricing is declared the apparent winning bidder. Assuming they meet all contractual requirements, they are subsequently awarded the contract.
The traditional design-bid-build delivery method tends to use prescriptive (i.e. cookbook) specifications by which contractors are instructed how to perform the work. There are advantages to this method, including a long history of acceptance, encouragement of open competition, flexibility in the outcome of the project that is retained by the owner, and the fact that it is easy to tender delivery of the project. On the other hand, innovation is not optimized; there are usually cost overruns and disputes between parties; MN DOT absorbs most of the risk and is responsible for errors and omissions; and the delivery process is linear and difficult to accelerate — an important consideration when planning projects that involve crucial transportation systems.
This final point was especially significant to MN DOT as it began planning the ROC (Rochester) 52 project, a massive undertaking that originally included expansion and improvements along an 11-mile stretch of Highway 52 in Rochester. (The project corridor was extended by two miles after issuance of the Request for Proposals.) This portion of Highway 52, which at the time was a four-lane controlled access rural/urban freeway, serves the commuting needs of a growing metropolitan area that includes industries such as the Mayo Medical Center (Mayo Clinic) and IBM and is a crucial link between Rochester and the Twin Cities.
Planning for ROC 52 began in the late 1980s and the Environmental Impact Statement was approved in 1996. At that time, it was estimated that it would take more than 11 years and 15 stages to finish the project corridor due to funding limitations for MN DOT District 6. In 2000, an Economic Impact Study performed via a partnership between MN DOT District 6 and the City of Rochester analyzed four staging alternatives along with the corresponding retail transfer and business impacts within various commercial business centers. Study results prompted recommendation of a five-year maximum construction timeline. It was clear MN DOT needed to investigate alternative delivery methods if it were to cut the estimated schedule by more than half.
Setting the Stage for Design-Build Best-Value
Through its research, MN DOT recognized that the design-build process offered solutions to a number of obstacle associated with the ROC 52 project. First, design-build would streamline the overall letting process by allowing the project to be completed under a single contract. New funding mechanisms also became available, including federal accelerated construction dollars. Another design-build advantage was a faster completion rate than the traditional design-bid-build process. It was determined that the ability of design-build to permit construction on some areas as design progressed on others facilitated enough reduction in construction time to meet the five-year goal. Besides meeting the recommendations of the study and qualifying for federal funding, design-build offered safety incentives. The much-needed additional through-lanes and major interchange reconfigurations would be in place sooner than was originally anticipated, reducing driver uncertainty and resulting in a safer, more efficient travel corridor.
That’s not to say there weren’t some challenges. Selecting design-build for the ROC 52 project was a monumental decision for MN DOT. At $237.7 million, the project represented the largest highway contract ever initiated by MN DOT and the first time the agency had employed the “true design-build” delivery method using best-value selection. The fact that MN DOT selected this project to make its foray into new contracting territories speaks to how substantial it viewed the rewards to be. Rather than awarding the contract based solely on the lowest qualified bid, best-value rates potential contractors on a number of factors, including bid price, schedule, and a displayed ability to comprehend and manage the project.
In other words, the contractor has to bring more to the table than just the ability to adhere to a prescribed design and stay within budget. The contractor must be a partner who is willing to share the role of project management, assume some of the risk responsibility, and contribute to the design. Awarding a design-build contract based on a best-value grading system can help identify the most qualified and cost-effective contractor for the project. Although price was a major factor in the best-value selection process, by definition, value means to have worth measured in usefulness or importance. Factors such as time, quality, and professional project management were deemed highly important as well.
Before selecting a contractor for its inaugural design-build best-value project, MN DOT established a core team (CT) that would be able to devote the necessary time and energy for a project of this size and complexity. The CT typically remained independent from other MN DOT functional groups and management assignments and had authority to call upon other departmental personnel for assistance and support as needed. As a result, the CT could make on-the-spot decisions for many project-level issues, which facilitated production of project documents and helped maintain the project schedule. In addition to personnel from MN DOT’s District 6 office and the Office of Bridges and Structures, the core team included a representative from the Federal Highway Administration (FHWA) and a procurement consultant, HDR. The management group comprised four of the nine CT members who were dedicated solely to the project. The CT met bi-weekly with an executive management team, which included people from MN DOT District 6, MN DOT Central Office, the City of Rochester, and Olmsted County to discuss impacts on the MN DOT program and policy issues. Because the CT and the executive management team acted as an extension of the owner on the ROC 52 project, for the purpose of this text future references to all three entities will be generalized as MN DOT or the ROC 52 project team.
One of the early orders of business for the ROC 52 project team was to construct a system for evaluating and scoring proposals. A major challenge in a review process such as this is preserving the integrity of the process and ensuring that a fair and unbiased selection process is followed. Otherwise, the process could be subject to challenge from unsuccessful bidders. To address these issues, an independent process oversight committee was instituted to oversee the entire review process. The committee had the authority to write a formal report at the end of the technical review process; however, on ROC 52 they did not determine that this effort was necessary.
A standardized technical proposal evaluation manual was prepared by the Department and utilized by a technical evaluation committee in scoring the proposals. All technical proposal scoring was required to be completed without any knowledge of the contents of the financial proposals and prior to the project letting date, when the price proposals were opened. Under the best-value selection method, the winning team is the one with the lowest adjusted score, determined by dividing the price by the technical proposal score. The last step in the evaluation process was to share the final scoring of each of the voting members. Prior to this final tallying of scores, committee members did not share or discuss individual scores.
The technical proposal phase of ROC 52 comprised 100 points divided into four categories, each with a predetermined significance:
- Project management — 40 percent
- Project understanding — 30 percent
- Schedule — 25 percent
- Innovation — 5 percent
There were 89 criteria and, based on the project cost, each technical point was essentially worth $2.5 million in the best-value computation.
After the technical proposal evaluation was completed, a public letting was conducted, during which the financial proposals were opened and the adjusted score determined for each team. The winning team, Zumbro River Constructors (ZRC), had both the highest technical score and the lowest price.
MN DOT’s intent was to create a fair and uniform basis for final proposal evaluation in accordance with Minnesota’s design-build legislation. The consensus among those who participated in the process was that this had been achieved well beyond satisfaction.
Another factor in the success of the best-value selection process was a Minnesota legislative measure that encourages companies to put forth their best effort when bidding for design-build projects in the state. By law, MN DOT is required to offer each unsuccessful entrant a stipend of 0.2 percent of the estimated project cost. It is thought that this stipulation encourages more teams to bid on the projects and results in a higher quality of proposal packages by providing some level of compensation for the costs associated with submitting a proposal. Each of the three unsuccessful teams on ROC 52 was offered and accepted a $454,000 stipend. It was felt that several of the teams far exceeded this amount in their efforts in submitting a proposal.
In accepting the stipend, the unsuccessful teams agreed to allow MN DOT to assume ownership of any alternate technical concepts included in their proposal materials. MN DOT could then negotiate with the winning team, ZRC, to implement these concepts. As it turned out, several alternative technical concepts were acquired in this manner and were utilized on the ROC 52 project. Alternative technical concepts are discussed in greater detail later.
As complex as best-value selection is, MN DOT’s first experience with the process progressed from initiation of the design-build delivery method in early November 2001 to letting on November 1, 2002 — a total of 51 weeks. During this time, MN DOT issued a Request for Qualifications, evaluated four State-ments of Qualifications, short-listed four teams, issued a Request for Proposals, implemented an alternate technical concept process, and evaluated and scored technical proposals from the four finalists. This design-build effort already was setting an accelerated pace for such a large project.
Alternate Technical Concepts
One concept typically endorsed in a design-build environment is the use of performance specifications in lieu of traditional prescriptive specifications. In a paper titled “The Multiple Roles of Specifications in Lean Construction,” Patrick T.I. Lam, Mohan M. Kumaraswamy, and S. Thomas Ng said the following about performance-based specifications:
“The recent proliferation of specialist works has seen an increasing use of performance-based specifications replacing the mainstream ‘prescriptive’ specifications, which are characterized by detailed descriptions of material and workmanship requirements. In order to give flexibility and encourage innovations in the use of materials, systems and methods, performance specifications state the required end results and leave the contractors to come up with means to achieve those results.”
In response to this trend, MN DOT developed an alternate technical concept process, whereby proposing teams can submit concepts that do not meet the requirements of a contract. Rather than telling the potential design-build contractor how to complete the project, submitting teams now have an expectation that they will be allowed to introduce innovation and infuse their own expertise in the form of performance-based specifications. That said, MN DOT remains the decision-maker and has the right to approve, conditionally approve, or reject any alternative technical concepts.
The ROC 52 project team researched design-build projects and found that processes for innovative concepts had been developed previously. However, they found that once an innovative concept was submitted and subsequently reviewed and approved by the owner, an addendum was issued to each proposing team identifying the approved innovative concept. MN DOT felt this approach was not conducive to encouraging proposal teams to submit innovative concepts because it really didn’t give them any incentive or financial advantage as a reward for the time, effort, and money that went into developing and submitting the concept.
For this reason, MN DOT developed its own process for innovation and termed it the alternate technical concept process. Via an alternate technical concept, each team could submit concepts that were outside of the RFP’s scope of work requirements. All innovative concepts submitted would be kept strictly confidential to the submitting team throughout the evaluation process. Only after the letting could a concept submitted by one team be shared with another team. In that case it would be shared only with the winning team and only if the concept developer had accepted the previously mentioned compensatory stipend.
Each design-build team would be allowed up to five one-on-one meetings with the MN DOT’s project management staff to discuss ideas and obtain feedback. The design-build teams were encouraged to submit any and all innovative concepts for formal review. If an alternative technical concept was approved or conditionally approved, the team submitting it could include the innovative concept in their financial proposal, as long as they agreed to any and all conditions set forth in MN DOT’s approval.
The project management staff identified four areas that were deemed off-limits to alternative technical concept submittals, as MN DOT had no intention of changing the RFP requirements in these areas:
- Right-of-way
- Intelligent transportation systems
- Pavement structure
- Aesthetics
MN DOT received 100 alternative technical concepts from the four contractors submitting proposals; nine were approved and 39 were conditionally approved. The proposed innovations encompassed a variety of design areas, including but not limited to: roadway/geometrics, structures and bridges, maintenance of traffic, and partnering.
Some of the most visible examples of the benefits gained through the alternative technical concept program are the retaining and noise walls that now border Highway 52. The basic concept for retaining walls in the RFP’s scope of work was a poured cast-in-place cantilever wall. Zumbro River Constructors saved nearly $4 million in its bid price and claimed the innovative retaining wall concepts also reduced the construction schedule by one year. As mentioned previously, MN DOT negotiated with ZRC to include additional changes stemming from alternative technical concepts submitted by other proposal teams.
Risk Allocation
One of the potential strengths of design-build is assigning a specific realm of risk to the party that is better able to manage it. Several key risk areas should be evaluated during the risk allocation phase so that both partners — the owner and the design-build contractor — enter the project with a clear understanding of who is responsible for each element.
For the ROC 52 project, there were eight primary areas of risk assignment:
- Quality (design and construction)
- Design
- Environmental
- Innovation (use of alternative technical concepts)
- Geotechnical
- Right-of-way
- Schedule
- Construction
Who should assume what risk? That can be a complicated decision process, and each area of risk must be examined individually. For example, asbestos abatement and contamination mitigation might be carried out by the contractor, but because a thorough and complete asbestos survey can only be completed once a building is no longer occupied (i.e., after the contract has been let), it would be difficult for the winning design-build team to provide an accurate estimate of the cost to perform such work.
ROC 52 required removal of 79 buildings, for which there were no asbestos surveys available at the time ZRC was awarded the contract. MN DOT retained the financial risk of asbestos surveys and abatement itself rather than asking the contractor to commit blindly to a potentially costly situation. The costs associated with the actual building removal, however, remained with the design-builder. Had the contractor been asked to submit an estimate for asbestos abatement based on purely speculative information, the number likely would have been in the $4 million range. The actual cost of abatement was $1.3 million.
Similarly, MN DOT worked directly with the Minnesota Pollution Control Agency to mitigate contaminated materials found within the project site. The familiarity between the two state agencies fostered a spirit of cooperation, resulting in a mitigation plan that eliminated the need to add considerable expense hauling contaminated materials to approved facilities in either the Twin Cities or Iowa. Instead, contaminated material was diluted by mixing it with non-contaminated material. It was then used on-site as fill material.
A more innovative approach was needed when assigning risk for maintenance of the project site. For the design-builder to assume this risk, it would be required to fix items damaged by traffic accidents. MN DOT collaborated with the Minnesota Attorney General’s office to develop an agreement that would subrogate to the design-builder its authority to seek compensation from the responsible party’s insurance company. Zumbro River Constructors could keep track of costs associated with repairing damaged items on the project and submit the expenses to the appropriate insurance company. This process worked well and meant the contractor did not have to increase its bid price to account for these unknown costs.
Other Considerations
Though there are many obvious advantages to forgoing the linear process of the design-bid-build delivery method in favor of design-build, moving to that non-linear process creates some challenges in and of itself. For instance, cost estimating the construction phase of a project is a relatively straightforward task under DBB because the scope of the project has been substantially determined prior to accepting proposals from potential builders, although at a cost of extensive time spent preparing a final design. Likewise, having detailed cost estimates in hand simplifies the process of determining cost splits — how much comes from each funding source, including local, state and federal agencies. In the design-build environment, however, designs often are advanced only to the 30 percent level before issuing a Request for Proposals. In this case, some assumptions must be made to calculate cost estimates.
For ROC 52, when possible a length x width x depth formula was used to compute estimated quantities and then factors were applied for unknown parameters. This provided a reasonable estimate, though certainly not as precise as would be achievable through traditional design-bid-build. Again, the challenge in putting together these estimates was regarded as an acceptable exchange for the amount of time gained under an already demanding schedule.
Tracking quantities provided another obstacle during MN DOT’s first design-build venture. On traditional design-bid-build projects, inspection staff from MN DOT put a great deal of effort toward measuring, tracking, and computing quantities of installed components of work. These measurements form the basis of contractor pay requests. The department generates vouchers based on the quantities measured or computed to date. Formal procedures have been established, and these procedures are routinely audited during project closeouts and reviews.
In the lump sum nature of the design-build environment there are very few, if any, quantities. For ROC 52 an elaborate process was implemented to document and justify pay requests. The design-builder submitted a cost-loaded and resource-loaded schedule in electronic format. Line items were activity-based, with each activity having a maximum duration of 60 days. The schedule showed all activities, durations, and applied costs. Project inspectors received copies of schedules and activities in their particular area of the project via hand-held personal digital assistants. With this information, field staff monitored what the design-builder was working on, how they were performing as compared to their project schedule, and whether or not the design-builder was working on out-of-sequence activities. This process was carried out biweekly. The design-builder included a percent completed for their activities, which was reviewed and documented by MN DOT.
During spring and summer 2004 the MN DOT Office of Audit audited the ROC 52 project and found no concerns regarding this process.
Conclusion
It is quite evident that the funding and timeline requirements faced on ROC 52 will continue to dominate a majority of MN DOT’s projects in the future. Now having the necessary legislative requirements and practical experience in delivering projects using the design-build best-value delivery method, MN DOT believes it has positioned itself to manage such demands.
Public officials can become focused on bid prices or internal costs of delivering projects. However, other factors such as user costs, can actually make those internal costs seem small in comparison. The challenge is to weigh all project costs and requirements and select the delivery method that most effectively meets those demands.
In the process of coordinating and completing ROC 52, the overwhelming sentiment expressed by the traveling public to the project team was that they would rather be impacted all at once for a short duration than for little bits at a time over a long period. Shortening the project from an initial estimate of more than a decade to a maximum allowable contract of five years was the only solution in this case. As it turned out, design-build delivered a finished product even sooner. As part of its winning proposal, Zumbro River Constructors estimated a timeline of 3.5 years to complete ROC 52. The actual project duration was three years.
Sidebar - A Narrative on Cost Growth in the Design-Build Environment
Based in Rochester, with a sub-district office in Owatonna and 23 truck stations, Minnesota Department of Transportation, District Six, supports the transportation network of southeastern Minnesota, includ-ing 1,436 miles of state roadway, 624 miles of bike “friendly” roadways, 504 miles of rail line, 837 bridges, 13 airports, and 17 operating transit systems.
HDR is an architectural, engineering and consulting firm that excels at complex projects and solving challenges for clients. For more information on HDR design-build services, visit www.hdr-design-build.com.
Zumbro River Constructors, LLC (ZRC) is a limited liability company composed of Fluor Enterprises, Inc., Ames Construction, Inc., and Edward Kraemer & Sons, Inc., with Fluor Enterprises serving as the managing LLC member. In addition to the ZRC equity members, URS Corporation was a major participant filling the role as the lead designer.
Terry Ward has been with MN DOT for 18 years, including roles as district resident engineer and district traffic engineer. Most recently, Terry represented the District 6 office as project manager for ROC 52. He may be reached at Terry.Ward@dot.state.mn.us.
Doug Jackson is a department manager at HDR who specializes in roadway design, construction surveying, construction inspection, project management, design-build, budgeting, and public involvement. He served as a technical advisor on the ROC 52 project’s core team. He may be reached at doug. jackson@hdrinc.com.