This article is an excerpt from The Art of Tendering: A Global Due Diligence Guide, which is available for purchase.
In its spring 2018 report entitled Replacing Montréal’s Champlain Bridge—Infrastructure Canada, the Auditor General of Canada determined that delayed decision-making resulted in $500 million (CAD) in unnecessary maintenance costs for the improperly designed original bridge. The report also found that the government failed to properly develop a business case to justify the use of a public-private partnership (P3) model for building the new bridge.
As the report explained, in “October 2011, the Government of Canada announced construction of a new bridge to replace the existing Champlain Bridge, which links the island of Montréal with the south shore of the St. Lawrence River.” The government deemed this project necessary because premature deterioration in the original bridge had increased maintenance costs and created serious traffic congestion:
The existing bridge was less than 50 years old, but it had deteriorated badly. Heavy investments were required to repair and maintain it (Exhibit 4.1). If a structural problem forced the bridge to close, the four other river crossings in the area could not accommodate the displaced traffic without significant congestion. Even partial closures for brief periods or load restrictions could significantly affect the flow of people and goods through the region, and also affect the economy.
The Auditor General found that the government’s failure to properly monitor lifecycle costs for the original bridge led to inappropriate planning delays and significant, avoidable direct and indirect costs:
This finding matters because the planning, procurement, and construction of a bridge of this size generally takes about seven years, from the initial decision to the date the bridge comes into use. The delay entailed avoidable expenditures of more than $500 million for the government, as well as economic costs for the Greater Montréal area resulting from truck load limits and lane closures. It is therefore important to make timely decisions in order to avoid significant delays and costs at the end of the service life of a bridge or comparable infrastructure asset. To ensure cost-effectiveness, a best practice is to plan life-cycle costs according to the asset’s remaining service life.
The report also found that the premature deterioration of the original bridge was attributable to faulty designs. Those designs failed to properly account for environmental conditions on the St. Lawrence River, to integrate corrosion monitoring systems into the bridge, or to accurately anticipate future traffic volumes over the bridge:
The existing bridge came into use in 1962. It deteriorated more quickly than expected, for several reasons:
It was not resistant to salt corrosion. The use of road salt for deicing began soon after the bridge opened, accelerating deterioration of the concrete and steel components.
It was extensively damaged by salt water. It had no drainage system to keep salt water away from the beams.
It contained key elements that could not be inspected for deterioration. The bridge girders had been designed with embedded pre-stressed cables, but there were no interior sensors to determine the condition of the cables.
It had not been designed for the volume of heavy truck traffic using the bridge since the 1990s.
The Auditor General found that the Jacques Cartier and Champlain Bridges Inc. (JCCBI), the body responsible for managing the bridge, failed to adequately notify the responsible Minister of the premature bridge deterioration, which delayed the decision to ultimately replace the bridge using a P3 construction model:
For the infrastructure that it owns—including the existing Champlain Bridge—The Jacques Cartier and Champlain Bridges Inc. (JCCBI) is responsible for planning for life-cycle costs and requesting the necessary funding from the minister to whom it reports. The planning must take into account costs of maintaining and operating the infrastructure, and replacing it when the end of its service life is approaching.
As a subsidiary of The Federal Bridge Corporation (FBCL) from 1998 to 2014, the JCCBI was required to make use of the FBCL’s corporate plan to officially request funding for life-cycle costs. The plan went for review to the minister responsible for the FBCL (then the Minister of Transport), who in turn recommended to the government to approve it.
In October 2011, the government announced its decision to replace the existing Champlain Bridge. It also stated that the chosen procurement model would be a public-private partnership (P3). A P3 is a contractual agreement between government and the private sector. Under the partnership, the private-sector partner delivers public infrastructure and assumes a major share of the risks in terms of design, construction, operation, and maintenance. This is one of a number of procurement models used by the government. Others include more traditional models such as design-build or design-bid-build.
After delaying its decision to build a new bridge, the Auditor General then found that the government rushed to judgment in deciding to use the P3 model for its construction. In fact, the audit found that the business case supporting the P3 approach was developed two years after the initial decision to use the P3 approach was made, and that the P3 approach proved to be potentially costlier than more traditional procurement models:
We found that Infrastructure Canada analyzed procurement models two years after the government had decided in favour of a public private partnership (P3) model. The Department did not base its analysis on reliable data and assumptions, and did not consider all key risks. More thorough analyses would have yielded results indicating that a P3 model could be more expensive than a traditional procurement model.
More specifically, the Auditor General found that the decision to adopt a P3 model was made prior to the required qualitative and value-for-money analyses.
The Auditor General determined that the government’s qualitative analysis was incomplete because it failed to account for the significant construction challenges faced by the project:
In 2014, Infrastructure Canada completed a high-level qualitative analysis for the new Champlain Bridge project. Completing an appropriate qualitative analysis was important because it was supposed to provide a first indication of whether a P3 was the most suitable procurement model. The analysis supported a conclusion in favour of the P3 model, identifying several advantages. From the viewpoint of the Department, one advantage was that the private partner would assume responsibility for more construction- and operation related risks, such as technical defects, cost overruns, and delays.
We found that the qualitative analysis was incomplete because the Department did not examine previous construction projects, nor did it analyze other possible forms that a P3 model could take, with varying levels of federal government involvement. Furthermore, the Department did not assess and consider some aspects of the project, particularly the difficulties related to undertaking a large construction project in an urban area, involving numerous discussions with municipalities, provincial and federal organizations, and stakeholders; and the potential impacts of project changes on the project completion date and related cost increases.
Additionally, the Auditor General also found serious flaws in the government’s value-for-money analysis, which it found was significantly biased towards the selection of the P3 construction model:
In January 2014, the Department finalized a value-for-money analysis to quantify the savings of a P3 model, compared with the traditional procurement model. It conducted the analysis before the financial proposals of the bidders were available. The analysis indicated that a public-private partnership would generate estimated savings of $227 million, compared with a public sector approach. PPP Canada reviewed the analysis and communicated issues and recommendations to Infrastructure Canada. However, the Department did not make all necessary adjustments. We reviewed key variables of the value-for-money analysis and found the following weaknesses:
Construction costs. The estimated construction costs for the project had a high variability, due to the Department’s use of comparatively imprecise estimates, which were based on a design that was only 5% completed. PPP Canada was concerned about the low level of design completion. Best practices recommend the use of a design that is at least 30% completed for more precise cost estimates, especially when project complexity is high. In other words, the project had not been sufficiently advanced to provide a well-based understanding of the costs. With a design that is only 5% completed, the construction costs may vary up to 30% above the estimate.
Efficiency rate. An efficiency rate represents cost savings from efficiencies in construction, operation, and maintenance. In a P3, savings usually come from leveraging private-sector experience and expertise. The estimated efficiency rate of 10% was high, compared with the 5% rate recommended by PPP Canada. The higher rate favoured the choice of the P3 model. Experts consulted by PPP Canada did not believe that an efficiency gain would be possible with an accelerated work schedule.
Project management costs. In 2014, the government estimated that its project management costs would be $15.9 million. In 2015, it revised that estimate to $158.6 million.
Risk evaluation. The risk evaluation was based on opinions obtained from experts during a workshop, a recognized industry practice. However, their opinions were based on their own expertise and in most cases were not supported by historical data from previous projects. The support of historical data is important for the calculation of plausible risk values, and is also recognized as a best practice. In the risk evaluation, we found some flaws that favoured the P3 model. For example, the risks of late completion and construction cost overruns were not properly evaluated for the P3 model. This is important because, under a P3 model, these risks are, for the most part, transferred to the private sector.
Discount rate. A discount rate determines the present value of future cash flows. The discount rate of 3.15% used in the analysis was higher than the then-current rate of 2.95%. This is important because a value-for-money analysis varies with different discount rates. In the analysis, the higher discount rate favoured the P3 model.
After recalculating key variables in the government’s value-for-money analysis using more conservative values and assumptions, the Auditor General concluded that “with more appropriate values for assumptions, a value-for-money analysis might show higher costs associated with the use of a public-private partnership instead of a traditional model.” In summary, the Auditor General found that the assumptions used by the government in its analysis were skewed to support the predetermined decision to use the P3 model:
In our view, the value-for-money analyses were of little use to decision makers because they contained many flaws favouring the P3 model. The project involved significant risks, was of unprecedented size, and required an accelerated schedule. Despite these factors, the values used in the analysis were not sufficiently conservative. The Department indicated that the low construction and financing costs in the private partner proposal could explain higher savings than the original estimates. However, it was unable to fully explain the $1.5-billion difference. In our view, the Department’s analyses indicated savings that were unrealistic.
The Auditor General concluded that in its view, “the private partner will not deliver the new Champlain Bridge within budget.”
As this case study illustrates, public institutions should be careful to properly manage the life cycle of major assets and to avoid unnecessary maintenance costs by proactively anticipating the need for replacement projects. They should also build adequate time into their life-cycle management to allow for proper procurement planning. And furthermore, when planning their major projects, public institutions should avoid rushing to judgment in their selection of a procurement strategy and should instead engage in an objective analysis of their various options to better ensure defensible planning decisions.