Cost Benefit Analysis
Cost Benefit Analysis reports were carried out by Nichols Applied Management for the NWT Department of Transportation on the Bridge project. The original report was dated September 2002. An update was provided in February 2003, to take into account the newly revised Traffic Forecasts that had become available. Another update was provided in December 2007, after the Concession Agreement had been signed.
Original 2002 report
The original 2002 report examined costs and savings for the main users of the river crossing (community re-supply, mine supply), projected traffic, costs and savings for ferry and ice bridge operations, and other elements associated with the project. It was based on an estimated capital cost of $55 million to build the bridge. It concluded that the project "generates net benefits within the normal range of acceptable returns". The net benefits were estimated at around $32 million (2002 dollars). In general, mine-resupply generally saw a net loss, as much of their traffic was over the winter bridge at a lower cost than proposed toll rates, while communities such as Yellowknife and other commercial traffic (and non-commercial traffic, which would not be tolled) saw an overall benefit.
The 2003 update was mainly to incorporate the newly available traffic forecasts. These were slightly higher than those used in the original report, and produced a slightly higher benefit analysis, but otherwise did not change the nature of the analysis significantly. The net benefits estimate increased to over $38 million (2002 dollars).
The 2007 update updated the capital costs for the Bridge. Between the original planning in 2003-2004, and the date the project was re-launched in 2007, the estimated cost had risen from around $55 million to the $160-$170 million range. The report used $155 million (although the budget would actually be $159 million plus a $10 million contingency fund, for a total of $169 million). At these levels, the net benefits became negative, estimated at negative $50 million (2007 dollars). The report noted that "At 3.2% the IRR, which is the discount rate that equalizes the stream of costs and benefits, is low relative to what is often deemed appropriate for public sector projects", but also stated that "Generally, higher construction costs and lower traffic volumes in the NWT contribute to lower IRRs of projects as compared to the rest of Canada."
As the report mentioned, the entire cost increase was being absorbed by the GNWT, leaving the other relative impacts largely unchanged. It went to on to recognize a number of "not quantified benefits, defined by the Department of Transportation", and concluded that in light of these, the project "supports the Department of Transportations twin objectives of creating opportunities for economic development and connecting communities."
There hasn't been an update since the 2007 report. However, I took a quick look to see if I could do anything to update them. Nichols costs included the Net Present Value (at 5% discount) of a $155 million construction budget spread over three years. Since the Bridge is actually financed by 35 year bond debt, this would seem to require changing the cost equation to account for that, and there are other potential changes. Warning, Ravens are not bankers or cost accountants, and I may easily have made some silly obvious error in trying to do some of this stuff!
First, there is a small issue with the Nichols 2007 numbers, which show a $50 million negative benefit for the project. The capital cost is given as $155 million, but the construction budget actually started at $159 million, although this number may not have known to Nichols at the time. This would increase the amount of negative benefit by $3-4 million, to the -$54 million range. I also wasn't sure originally if they had included the Pre-Funded Contingency Amount in their calculations, since their Table 2.1 just shows the $155 million construction cost, but it looks like they used the full $165 million to come up with their adjusted $160.6 million Total Costs.
Since the capital costs aren't being paid over the 3 years of construction, but instead over the 35 years of the bond repayment, it would seem necessary to recalculate the Total Costs to account for this. My idea is to consider the costs as the NPV of the tolls and annual GNWT contribution for the 35 years of lease period, plus an additional 40 years of operating costs afterwards. I used the same 2002 "Conservative" Traffic Forecast, and kept the benefits untouched. Replacing their costs with this (and adjusting to 2010 dollars) seems to reduce the disbenefit in half, to around $27 million, for the original scenario.
Ferry cost adjustment
The Minister of Transportation indicated in debate that the annual ferry/ice bridge costs were now around $3 million. This is significantly higher than the $1.7 million or so that Nichols were using from 2002, even accounting for inflation. However, if the $3 million figure is correct, the benefits due to avoided costs increase dramatically. I removed the ferry/ice bridge benefits from the Nichols total benefits, and replaced them with the NPV of 75 years of $3 million costs. Also, the Minister indicated that Annual Operating Costs as a GNWT project would be reduced from the $600,000 in the Concession Agreement to something like $350,000. So I adjusted the 40 post-lease years of operating costs from the above to reflect this reduction.
These two changes seem to reduce the disbenefit even further, to right around the break-even point, or perhaps slightly positive if one takes the GNWT returns into account.
Of course, the GNWT has had to approve additional funds for the project, which adds an immediate $16 million cost, putting things that much back in the hole. However, this money will slowly be repaid (with interest?) over the lease period, which is a benefit. Given the 5% discount rate used by Nichols is greater than the likely interest rate, the benefit wouldn't fully offset the cost, but it would tend to bring it back towards break-even.
Toll collection costs
The Nichols analysis doesn't seem to consider the cost of collecting the tolls, which is borne by the GNWT. The 2004 report on this posited roughly $1.4 million in capital costs and another $700,000 or so annually. This seems high, and even taking those numbers as 2010 dollars produces an NPV cost of over $12 million, sinking the project that much further into the hole.
Other traffic forecasts
I've used the 2002 Prolog "Conservative" traffic forecast for this, as it was the one used in the 2003 and 2007 Nichols benefit calculations. There might be a slight improvement with the 2006 Prolog "Conservative" forecast, as those traffic numbers are a bit higher, and may more closely represent current traffic. But the benefits would have to be recalculated, and I'm not sure if there is enough detail in the Nichols report to do that.
|This article is currently incomplete. BridgeWiki will eventually improve by expanding it.|