I previously posted an analysis authored by electricity and gas consultant Bruce Sharp addressing the costs to consumers of relocating the Oakville gas-fired power system. What follows is an updated version of that analysis. As previously, this analysis considers the costs specific to the power plant only and does not address other impacts. One example of a cost not included is the transmission costs required to meet the reliability needs the original Oakville siting was intended to serve.
Mr. Sharp’s argument is that $280 million was wasted by the government’s decision to move the TransCanada power plant from Oakville to Lennox. The government’s Long Term Energy Plan, issued after the Oakville plant was cancelled but before the Mississauga plant was cancels explains the impact of the Oakville cancellation saying, “a transmission solution to maintain reliable supply in the southwest GTA will be required.” (p. 34)
Documents released this week in the legislature disclose one estimate of $200 million in transmission impacts directly associated with Oakville cancellation.
Critics of Mr. Sharp’s assumptions and/or methodology are invited to present their own analysis. However, I would be very surprised if someone can make a convincing argument for a lower bottom line than Mr. Sharp presents.
A key lesson to take from Mr. Sharp’s work is to observe the complexity of the commercial issues involved.
Consider this complexity in the context of the negotiations that took place. Through every action he has taken in the electricity field throughout his term as premier, it is clear that Mr. McGuinty cares not a whit about the long term cost of electricity to Ontario consumers. Rather, his orientation has consistently been to look only for the happy stories. When forced to respond to criticisms, or even an electoral embarrassment, his orientation is to seek a quiet way to smooth things over.
TransCanada is a top tier international energy company. The company has been phenomenally successful in many fields. The company’s asset management record has only slipped a few times with major assets (e.g. Mainline gas transportation system and possibly also TransCanada’s recent investment in Bruce nuclear refurbishment.) Otherwise, TransCanada’s record, particularly in electricity, is a long win column dating back to the Ocean State power plant built in Rhode Island in 1990.
TransCanada has direct experience negotiating with Mr. McGuinty dating back to the original Bruce refurbishment contract dating from October 2005. In that case, Mr. McGuinty decided to first announce the accelerated closure of coal power and then to cut a deal for nuclear refurbishment. Not surprisingly, ratepayers got a rough ride that time, although in hindsight the Green Energy Act makes the ratepayer harm of 2005 look small. When the former provincial auditor was brought in to review the contract, he raised many concerns (although not some key ones) but one was that the structure of the deal resulted in a higher realized cost for the new power than reported in the government’s press release.
When it came to the Oakville negotiations, TransCanada knew that all they had to do was to structure the deal so that the initial hit looked small. With a figure of $40 million as the headline, McGuinty was willing to sign any deal, no matter how crazy the real price tag.
Mr. Sharp’s analysis suggests that the biggest ratepayer hit is not in the $40 million site costs, or the $210 million turbine purchase and the associated Net Revenue Requirement adjustment, but in the gas management component. Folks outside the power industry should appreciate that gas management costs are complex legal, operational and financial arrangements.
While Mr. McGuinty was rushed in his negotiations, TransCanada sharpening its gas model. Nobody was protecting the ratepayer.
The legislature is now seized with the issue of the government’s cover-up of the Oakville cancellation and its costs. Although there has been much productive debate, I don’t see any realistic possibility that the legislatures will be able to understand the intricacies of the underlying commercial issues in the current debate over the censure of Minister Bentley.
By contrast to both Mr. McGuinty and the legislature, TransCanada would have had a clear understanding of the issues from the beginning.
As in the case of the Bruce nuclear refurbishment, the Ontario Provincial Auditor should be brought in. The Auditor has recently shown an interest in electricity rate issues and has done a reasonable (though not perfect) job of bringing vast amounts of government waste to light. The Auditor can get outside advisers with appropriate experience to go through the intricacies.
The auditor’s mandate should be broad enough to check all the details on the costs of moving the Mississauga gas plant too. The only information available on the cost of moving the Mississauga plant has come from the government and the government’s admissions have escalated from $180 million to $190 million.
Rate Hit in Context
Let’s assume for a moment that no new cost impacts are discovered and that the impacts described in this post turn out to be reasonably accurate. If we assume a low discount rate such as the long term Canada bond (2.4%) plus 30 basis points (typical Ontario vs. Canada bond spread) and a 20 year amortization, the annual cost of a one time $480 million hit is about $31.38 million per year. Spread across the entire power system, this cost would translate into a cost per kilowatt hour in the range of 0.2 cents/kWh.
To put this in context, the government’s own estimate of the increase in household electricity rates is 33% (note an early edition of this post incorrectly reported this as 37%) in inflation adjusted terms over the period 2011-2015 or 4.4 cents/kWh in 2010 current value. (This forecast was issued before several big discount programs for industry were announced that will jack residential rates substantially.) Not to diminish the importance of the TransCanada power plant fiasco, but the overall electricity crisis ongoing in Ontario is more than 20 times as significant to ratepayers.
For all McGuinty’s defenders who claim it is unfair to blame him for rising power rates, note that power rates in the U.S. are declining.
Please join me in thanking Bruce Sharp for contributing his expertise to expanding the public’s understanding of this issue.
TransCanada Energy (TCE) 900 MW Natural Gas – Fired Generating Station
Analysis of Cost of Moving From Oakville to Lennox
29 September 2012 (Version 3)
Bruce Sharp, P. Eng.
Original media portrayals of the release of information on the cost to move this project focused only on the (sunk) relocation costs of $ 40 million. The turbine payment of $ 210 million to TCE was generally ignored, with reporters apparently taking at face value the government’s assertion that the payment would be directly offset by lower, future Net Revenue Requirement (NRR) payments to TCE. Other, potentially quite significant costs were also ignored, likely because they were not clearly identified as value transfers, let alone quantified.
This analysis estimates that the cost to move the project will exceed $ 280 million.
The following additional cost (in current dollars) elements attributable to moving the project were provided or were determined from information provided.
- Relocation $ 40 million
- Turbine payment, excess $ 88 million
- Gas Delivery and Management Services (GDMS) $ 152 million
Total $ 280 million
Other, unknown value transfers may include some or all of gas pipeline and electrical interconnection costs, payments made to Ontario Power Generation (OPG), services provided by OPG, line loss impacts, and network transmission requirements.
a) Relocation costs paid to TCE = $ 40 million.
b) Turbine payment of $ 210 million made to TCE.
c) Net Revenue Requirement (NRR) drops from $ 17,277/MW/month to
d) Ontario Power Authority (OPA) to fully reimburse TCE for Alternative Project’s “Gas Delivery and Management Services … on a flow-through basis and without an adjustment to the Net Revenue Requirement”.
e) “The turbine payment as well as covering the gas management costs for the new plant reduces the Lennox Net Revenue Requirement”.
f) TCE will receive payments to offset some or all of gas pipeline and electrical interconnection costs and payments made to OPG.
Discounted Cash Flow (DCF) Assumptions
g) Nominal generating capacity: 900 MW
h) Construction cost, overnight: $ 1.25 million/MW
i) Fixed OM&A cost: $ 20,000/MW/year
j) Debt: 70% of costs, 20-year term, 7.0% rate
k) Equity: remaining 30%
l) Inflation: 2.25%
m) 20% of inflation applied in escalating NRRs
n) Capital Cost Allowance: 8%, declining balance basis
o) Corporate tax rate: 25%
p) Turbine payment not repaid to OPA.
q) No deemed energy market revenue impact arising from higher contract heat rates.
r) GDMS encompasses the basis differential between Dawn and the Union Gas (East) delivery area Lennox falls in, plus charges for: storage demand, storage deliverability/injection demand, injection/withdrawal and distribution.
s) GDMS cost: $ 1,000/MW/month
t) 100% of inflation applied in escalating GDMS costs.
u) GDMS Net Present Value (NPV) calculation discount rate of 6%.
Findings – Turbine Payment
This part of the analysis evaluated to what extent the up-front turbine payment from the OPA offsets the impact of TCE receiving lower future NRR payments.
The up-front payment of $ 210 million was likely arrived at by looking narrowly at the stream of project cash flows. However, from a customer perspective, it appears that the $ 210 million payment was over-estimated.
The original Southwest GTA Evaluated Cost Model used a NPV approach with a nominal discount rate of 7%, to allow the OPA to compare competing bids. Taking the same NPV approach with the same nominal discount rate and evaluating the two scenarios in a very simplistic way – “Oakville” with a higher NRR and “Lennox” with a lower NRR and an up-front payment, an up-front payment of $ 254 million produces equivalent NPVs. If the analysis and negotiation were viewed through this lens, an up-front payment of $ 210 million might be viewed as a minor victory.
Alternatively, the analysis might have been done with a discount rate in the order of 10%, to approximate a typical regulated utility equity rate of return. Taking this approach, the two scenarios were equivalent at a discount rate of 9.7%.
Whatever the discount rate used, the above approach was flawed — as it left electricity ratepayers’ money on the table. The narrow, simplistic discount rate approach apparently taken focused only on the NRRs and an up-front payment did not accurately portray TCE’s differing financial results under the two scenarios. The Ontario government and the OPA should have approached the negotiation by determining the payment that would have left TCE indifferent between the two outcomes. A more appropriate approach would have been to perform a broader and more detailed discounted cash flow (DCF) analysis that looked at the NPV on TCE’s equity contribution to the project.
If TCE’s equity rate of return (EROR) is assumed to be 11-12% then an up-front payment of $ 210 million would leave TCE generally indifferent to the lower NRR. However the more detailed DCF analysis of the original project with the higher NRR indicates that TCE’s actual EROR would have been in excess of 20%. At a conservative (i.e. low) EROR of 20%, an up-front payment of $ 122 million would leave TCE indifferent to the lower NRR. The turbine payment therefore could have been lower by at least $ 88 million and TCE would have been no worse off with the lower NRR.
Findings – Gas Delivery and Management Services (GDMS)
Offloading the GDMS costs – at no further reduction of NRR – is a significant value transfer to TCE. The costs associated with the GDMS remain to be determined but due to their magnitude they deserve close scrutiny.
The Union Gas distribution demand charge alone is in the order of $ 600/MW/month and to that are added costs for: the Dawn – Union Gas Eastern Delivery Area transport or basis differential, storage, injections and withdrawals. The assumed, total, NRR-equivalent monthly unit cost of $ 1,000/MW/month is therefore thought to be reasonable or conservative.
For 900 MW and considering the 20-year term of the project, the NPV for the GDMS costs is $ 152 million. The relationship between the monthly unit cost and the NPV is linear, so once a more accurate GDMS monthly unit cost is known a more accurate NPV can be determined.