Thursday, 27 February 2020


Guest Post by David Vardy
Former CFO Derrick Sturge told the Muskrat Falls Inquiry the project and its financing were “rock solid”. CEO Stan Marshall told us two years ago the project would “finish strong”.
Premier Dwight Ball promised “Newfoundlanders and Labradorians that they will not bear the burden of higher electricity rates or taxes as a result of Muskrat Falls. We will deliver on that promise.”

On February 25, 2020 Nalcor CEO held a press conference and took questions from the media confirming the delays which had already become well known. He did not provide a cost update on the project, which appears likely to exceed the official 2017 estimate of $12.7 billion. He repeated the statement that Nalcor would “finish strong”.

Stan Marshall’s press conference followed on the heels of the announcement on February 10, 2020 by Premier Dwight Ball and Natural Resources Canada Minister Seamus O’Regan of a framework for refinancing Muskrat Falls in order to make power rates affordable.
How realistic was that plan? Will it indeed keep power rates stable and avoid massive increases? Do the two media events signal a “strong finish” as promised by CEO Marshall? Do they confirm that that citizens will “not bear the burden of higher electricity rates or taxes as a result of Muskrat Falls”? 

The elements of the refinancing plan are as follows:

1. Adoption of a target of 13.5 cents/KWh for the first year of “full” power, perhaps 2021 or maybe later, along with the April 2019 Rate Mitigation Plan
The target rate of 13.5 cents/KWh is based on the need for new revenues and cost savings in the amount of $726 million as shown in the framework document and summarized in the table below:
This shows a high reliance both on dividends ($200 million) and on help from the federal government ($200 million), both adding to $400 million of the $726 million in revenue requirements.

The following table from the PUB report page 104 at reports a smaller amount, only $193 million, toward the $726 million in 2021 revenue requirements, rather than the $529 million in the rate mitigation plan of April 2019. Clearly the PUB took a more conservative and realistic view of the options. There is a big disconnect between the April 2019 rate mitigation plan and the February 2020 report of the PUB. 
 As shown in the Technical Appendix below, the $726 million is a low estimate of the 2021 costs, known as “revenue requirements”.  It is lower than the $808 million estimate used by Nalcor in 2017 and the official cost estimate of $12.7 billion has not changed. The $12.7 billion is an understatement of the capital cost of the project, as explained in the technical appendix, and should be closer to $13.8 billion. The costs imposed on both ratepayers and taxpayers to supply the projected energy requirements of 2021 are more than $1 billion, much higher than the 2021 revenue requirements used both by the PUB and by government. This is explained in the Technical Appendix below.

The PUB was able to identify only $193 million in mitigation measures for 2021 and close to half of that amount is forgiveness of dividends, which were fictitious from the start. By 2030, 72% of the mitigation potential is forgiveness of non-existent dividends. It will be difficult to keep power rates at 13.5 cents/KWh in 2021 and to hold them at that level in subsequent years.

The main problem with the province’s rate mitigation plan is that it diverts revenues from other purposes. The additional revenues from exports and from electrification will be modest in the early years. The magnitude of the shortfall is such that a major infusion will be required from the federal government. The federal government may offer some relief on sinking funds and on pre-payment of cost overruns but that simply shifts the problem into the future. Similarly an increase in the federal loan guarantee might be used to offer some relief to the province but ultimately the burden will fall on ratepayers or taxpayers in the province.
David Vardy (Photo Credit : The Telegram)
What can the federal government do to assist the province?  They could inject new equity into the project in order to bear more of the project’s risk and to recognize that the benefits of the project, if any, will flow to other Canadians, particularly Nova Scotians. Additionally, the federal government could enhance its equalization program or other support programs. They could enhance fiscal stabilization payments or increase payments under Canada Health Transfers (CHT), and under Canada Social Transfers (CST). The burden of Muskrat Falls is best addressed through the prism of fiscal policy in recognition of the fact that the fiscal impact far outweighs any energy benefits.

The PUB report on rate mitigation focused on opportunities to raise revenue through the sale of additional electric power and by making the provincial power system more efficient. The opportunities to raise revenues through increased electrification will take time to achieve and will be modest in scale. Without major restructuring of our electric power industry the opportunities for cost savings are equally modest. The province can divert revenues, such as offshore royalty revenues and revenues from equity investment in oil and gas, but this will simply take money away from other social priorities and from debt reduction.

2. Monetization of dividends from the transmission line.
The flow of dividends from the Labrador Island Link is relatively modest. If these were somehow capitalized they would generate little revenues. If they were monetized at the front end it would simply shift the problem into the future. But what is “monetization”? It is a euphemism for borrowing against future revenues, which is no cure. It simply allows the problem to grow exponentially and to shift a growing burden to future generations.

3. Deferral of sinking fund payments until the end of 2021 and a waiver of pre-funding of cost overruns.

Deferral of sinking fund requirements provides some immediate cash relief as does the waiver of the need to pre-fund cost overruns. This tinkering does not reflect the magnitude of the problem.

4. Major changes to the Power Purchase Agreement (PPA) with its take-or-pay contract and which places 100% of the cost on ratepayers on the Island.

The PPA is an anomalous agreement between Muskrat Falls Power Corporation (Nalcor) and NL Hydro. It places the obligation to pay on Hydro. Hydro must in turn place 100% of the cost on power consumers, on industrial customers and on the customers of Newfoundland Power and NL Hydro. This 50 year contract locks ratepayers into a death spiral, where high rates force ratepayers to cut back and Hydro responds by raising rates even higher. Already the demand for power has ratcheted down in anticipation of higher power rates.

The replacement of the PPA is absolutely necessary. Ratepayers who will consume little if any Muskrat Falls power cannot be expected to absorb 100% of the costs, any more than customers in the Maritimes or New England can be expected to cover the costs of Muskrat Falls. The PPA is at the heart of a flawed business plan, which depended upon legislation to create Nalcor as an unregulated monopoly. This locks our electric power industry, along with the rest of the provincial economy, into a high cost electricity system and prevents the adoption of new, low cost technology.

5. The Premier’s letter to Federal Finance Minister Bill Morneau states that “At the core of our agreement is the requirement to transition the Muskrat Falls/Labrador Transmission Assets revenue model to a Cost of Service model, which will ensure that equity returns are redirected from Nalcor to ratepayers.”

The costs of Muskrat Falls, including the cost of building the project, are recovered through power rates, using a combination of “cost of service” (COS)  and “escalating supply prices” (ESP). COS ensures that rates cover all costs on an annual basis and this is the traditional model used by the PUB in NL and by many other jurisdictions. The ESP model tends to shift some of these costs into the future, particularly the return on provincial equity. All costs are recovered over a 50 year “supply period”. A description of these models is found in the Technical Appendix.

The Ball-O’Regan announcement, strangely, places the transition to a full COS model and the rejection of the hybrid approach “at the core” of their refinancing agreement. The hybrid approach leads to rising costs, or revenue requirements. Using the latest numbers from Nalcor they rise from $726 million in 2021 to $920 million in 2030 and to $2.5 billion in 2069. A COS model would lead to higher revenue requirements at the beginning but they would taper off over time. The Ball-O’Regan announcement writes off $30 billion in dividends (ROE) on generation assets, a write-down from about $39 billion in dividends over the 50 year supply period.  Whether the federal government will inject new funds remains unknown but such an injection is necessary. Without access to the spreadsheets underlying the public announcement it is difficult to say.

What does seem clear is that the transition to a cost of service approach alone will not bring power rates down to 13.5 cents/KWh. Nor will it keep them there. What appears to be under discussion is a modified cost of service approach plus a write-down of provincial equity. A write down of equity can produce impairment on the province’s public accounts and an increase in our net debt (see Technical Appendix).

If we have learned anything from the Muskrat Falls Inquiry it is the need for more accountability and transparency as the province moves into negotiations with the federal government. We need to understand that we are not just shifting the burden down to future generations. The hybrid approach was intended to avoid “rate shock” by just such shifting of financial burden. If we as a generation are to act responsibly to our children and grandchildren we need to avoid adding an increasing load of debt.

The plan does not appear to inject the massive quantum of resources that is needed to stabilize power rates and to create a sustainable electric power system. Muskrat Falls was intended to be a solution to a perceived energy problem. It has instead exacerbated a growing fiscal crisis and has to be treated with the instruments of fiscal policy, bringing to bear federal support programs designed to mitigate not only power rates but the other fiscal pressures upon us. Realistically such support can only come as part of an agreement which imposes fiscal discipline on the province, as would be the case if we were a fiscally beleaguered nation seeking succor from the World Bank or the International Monetary Fund. We cannot expect other Canadians to share the burden unless we show we are ready to make sacrifices.

Our citizens know that the financing of this project is anything but “rock solid” as described by the former Nalcor CFO. They know that a “strong finish” is wishful thinking, rooted in delusion. Citizens are not gullible. They cannot accept the promise they will be saved “from the burden of higher electricity rates or taxes as a result of Muskrat Falls.” They are willing to bear their share of the burden along with other Canadians.

David Vardy

Cost of service (COS) Model vs. Escalating Supply prices (ESP)

The data used in this post are illustrative and not definitive. All of the numbers are estimates for a project yet to be completed and whose schedule and costs remain highly uncertain.

This Technical Appendix below is intended to clarify some of the concepts and measures that are used in the public dialogue on Muskrat Falls. My hope is that it will unravel some of the complexity and also create some understanding as to how high the stakes are in these negotiations. In the appendix I show how both the capital costs of Muskrat Falls and the size of our provincial equity investment are underestimated. The technical appendix also explains how “economies of scale” impact on the project and how such a large project, whose scale far exceeds the power needs of the province, makes it an inappropriate fit for the province. I conclude with a discussion of net debt.

“Cost of service” (COS) was adopted for recovering costs from the transmission assets. COS requires that costs be recovered from ratepayers each year. The capital costs are primarily the cost of interest on debt, the cost of debt repayment, the cost of equity capital raised from shareholders and the return of equity to shareholders. Depreciation is a surrogate for the repayment of bond and equity investment. To this must be added the cost of operations and maintenance, fuel costs, water power rentals and payments to the Innu Nation.

The 2017 cost update for Muskrat Falls reported a capital cost of $10.1 billion in direct costs plus $2.6 billion in financing costs, for a total of $12.7 billion. This official estimate has not yet been revised. The annual revenue requirements on which power rates are based were estimated in 2017 at $808 million for 2021, the first year of full power. In 2019 this estimate was revised downward to $726 million and it is this estimate which has been used both by the PUB and by government as the incremental cost of Muskrat Falls.

This is the amount which must be found through other revenues or cost savings if power rates are to be stabilized. It is difficult to accept the notion that revenue requirements can realistically be expected to be lower than those estimated in 2017 in light of known cost escalation which has materialized since that time, including the problems with synchronous condensers and the costs arising from GE’s inability to produce transmission software on schedule. For this reason I am disinclined to accept an estimate lower than the $808 million for 2021 revenue requirements.

The power rates which will recover costs are based on a hybrid of “cost of service” and “escalating supply prices (ESP)” or “PPA rates” (based on a power purchase agreement). The pattern of revenue requirements for the cost of service model is generally one which declines over time while the ESP or PPA model shifts certain costs into the future and leads to revenue requirements which rise over time. Since generation costs for this project are higher than transmission costs the ESP model applied to generation assets dominates the revenue requirements and produces rising costs, growing from $808 million (using 2017 estimates) to $2.5 billion in 2069. This is referred to as “back end loading”. This is in contrast to the tendency under COS cost recovery for “front end loading” of costs.

In the public utility business there are frequently pressures to reduce costs today and to shift them into the future. Expenses are capitalized so they are not expensed or recovered in the year in which the costs are incurred. Such shifting mechanisms have led to controversy in British Columbia and Ontario in particular. In the case of Muskrat Falls the shifting of costs was intended to avoid rate shock through the adoption of an unorthodox ESP approach. This hybrid model has been in place since the PUB reference of 2011, if not before.

The pivot for shifting costs into the future is the province’s investment in generation assets, currently $4.2 billion. The return on equity of 8.4% is built into the costs and recovered over a 50 year “supply period” and is “back end loaded” with most equity returns in later years, 2041 and beyond. The $726 million in 2021revenue requirements used by the province and by the PUB provides $57 million in return on the province’s $645 million in provincial equity but no return on equity for the province’s $4.2 billion in generation assets.

Effectively the province is borrowing the money and absorbing the cost ostensibly on an interim basis until such time as rising demand and rising nominal revenues allow repayment to be made to the shareholder. Using 8.4% as the cost of equity capital and applying this to the $4.2 billion in equity invested by the province results in an additional cost of $353 million, which needs to be added to the $726 million in revenue requirements targeted in the province’s rate mitigation plan of April 2019. If we instead use 3% as the cost of borrowing this additional cost becomes $126 million. To this we add $210 million to repay borrowed funds over 50 years we get $336 million. It is difficult to see how the cost to ratepayers and taxpayers in 2021 would be less than $1 billion, much higher than the $726 million target.

The PPA model assumes that the revenues will be generated to pay the 8.4% return and to return the original equity investment to the province. This is hardly credible. At the outset it was highly questionable whether the dividends would materialize. With more realistic demand projections, based on realistic assumptions on population growth and demand elasticity, along with doubling of capital costs, the prospect of dividends is remote.
The rate mitigation plan of April 2019 and the refinancing announcement of February 10, 2020 are an admission that the business plan for the project is fatally flawed and that dividends are fictitious. The Ball-O’Regan Accord recognizes that $30 billion in dividends over the 50 year supply period is not realistic and must be written off.

Government has to be concerned with the full annual costs of the project, both those paid by ratepayers and those transferred to taxpayers. For this reason it is necessary to reflect the full financing arrangements and to include the full cost of the project, including the cost of financing the funds borrowed by the province to invest equity into the project.

Understatement of Capital Cost
The latest quarterly Oversight Committee Report Annex A  dated December 23, 2019 shows that direct capital costs are estimated at $10.1 billion dollars, with $3.7 billion for the Labrador Island Link (LIL), $5.5 billion for the generation site (MF) and $0.9 billion for the transmission line between Churchill Falls and Muskrat Falls, known as Labrador Transmission Assets (LTA). The generation site (MF) and Labrador Transmission Assets (LTA) are referred to as “generation assets” and their direct capital cost together are estimated at $6.4 billion, representing 63.3%,  vs 36.7% for the Labrador Island Link (LIL).

The financing costs of $2.6 billion must be added to this, resulting in the official cost estimate of $12.7 billion. These financing costs are the subject of sections 2.3-2.6 of the  Oversight Committee report for the period ending September 2017 and dated November 3, 2017. Section 2.6 reports that for the LIL the allowance for funds used during construction (AFUDC) is estimated at $440 million, which is included in the financing costs. No such estimate is included for AFUDC on the generation assets, which means that the estimate assumes zero opportunity cost for $6.4 billion in direct capital costs. To estimate how much should be added to correct this omission I have taken the LIL AFUDC of $440 million as a proxy and made two adjustments.  The first reflects the fact that generation assets are 72% higher in direct capital costs ($6,404million/$3,714 million) than the costs of the LIL. The second reflects the fact that the minimum equity in generation costs, as required in the 2012 federal loan guarantee, is 35%, vs 25% in the LIL or 40% higher  (35%/25%). Accordingly I have taken the $440 million in LIL AFUDC and multiplied it by 2.41 (1.72 x 1.4), providing my estimate of AFUDC on generation assets of $1,060 million. Adding this AFUDC for generation assets to the official cost estimate of $12.7 billion produces an adjusted cost estimate of $13.8 billion.

Estimating Provincial Equity
The province’s equity investment in generation equity, disclosed in PB-519-2019 Table 4, is $3.1 billion. To this must be added the AFUDC of $1.1 billion ($1,060 million rounded) to produce an estimate of provincial equity investment, at the time of project commissioning, of $4.2 billion. To this must be added the $645 million in LIL equity invested by the province to produce an estimate of total provincial equity investment of $4.8 billion.

Escalating Supply Prices
Escalating supply prices (ESP) as a model for cost recovery of the capital costs of Muskrat Falls is based on the concept that the real unit cost per kilowatt hour will remain constant over the 50 year supply period.  Costs are levelized over the 50 year period and then increased by 2% annually to reflect inflation.  This inflation rate is applied to the Levelized Unit Energy Cost (LUEC). Cost recovery depends upon escalating demand for power at a rate which is held constant in real terms. If demand does not escalate then the revenues produced will fail to cover costs, let alone generate dividends on provincial equity.

Risk Bearing Equity
The federally guaranteed debt is secured by mortgages held by the federal government on the assets financed. The power purchase agreement also forms part of the collateral. Provincial equity on the other hand, is not secured. It is fully at risk. Dividends can only be paid if there is a return on equity left after all other costs have been covered. Dividends are not guaranteed. Building an allowed rate of return of 8.4% into the model does not provide an assurance that the shareholder, GNL, will be compensated for the cost of the borrowed funds used to finance the equity, let alone dividends on top of that.  Shareholders take the risk that dividends may fall short of expectations and may turn into losses.

Economies of Scale
The project depends upon economies of scale and can achieve its full potential at lowest cost when demand rises sufficiently to absorb the full 4.641 TWh of net energy produced after line losses are deducted from the plant’s rated output of 4.9 TWh. Stan Marshall’s presentation of February 15, 2018 at Memorial University, entitled “Understanding Muskrat,” showed us a unit cost of 17.42 cents/KWh, assuming the full output could be sold at cost-compensatory rates. The presentation shows only 1,324 GWh out of 4641 GWh  projected to be used on the Island in 2021, costing 61 cents/KWh, based on 2021 revenue requirements of $808 million. If we add back the return on equity of $353 million and revise the revenue requirement up to $1,163 million then the cost per KWh becomes 88 cents/KWh. Unit costs are high when demand is low.

In his report to the Muskrat Falls Inquiry, Pelino Colaiacovo of Morrison Park (MFI Exhibit P-04445.) showed the economies of scale from Muskrat Falls in his following chart 7:

Chart 7
The blue line represents the unit costs of the isolated Island option which dip for a period and then rise, exceeding the unit costs of the interconnected Island option shown in red up to the crossover point, beyond which they rise and then remain stable in real terms. The unit costs of the Interconnected Island system decline and then stabilize for the rest of the period. This is consistent with the information from Stan Marshall’s presentation. It depends upon the growth in demand which had been projected by Nalcor. If demand does not materialize unit costs will remain high and dividends will be elusive.

Higher rates than predicted will threaten the prospects for growth in future demand. As we have shown previously the cost overruns pose a threat to dividends early in the planning period and will also impact on future dividends. Despite the long 50 year planning horizon or “supply period”, going well beyond the end of the Churchill Falls contract in 2041, there is a very high risk that the shareholder will experience negative returns. This in turn will create pressure for a write-down of the equity investment in the project.

Net Debt
In her report for the year ended March 31, 2019 (page 11) the Auditor General defined net debt as follows. “Net Debt represents all the liabilities of the Province less its financial assets and indicates whether there are enough financial assets to cover the liabilities for future generations. Net Debt is a commonly used indicator to measure the financial health of the Province.”

She defines financial assets as “amounts that the Province has available to pay its liabilities or finance future operations. Financial assets consist of cash and temporary investments, amounts receivable from third parties, investments, inventories held for resale and equity in Government Business Enterprises (GBEs)…” Nalcor Energy is such a GBE and the government’s equity in Muskrat Falls appears as a financial asset, valued at cost.

Each year the value of the financial assets must be assessed to determine whether a write down is required. Currently the equity investment in Muskrat Falls is used to offset the debt incurred by government to make the equity investment. If the financial assets are written off completely then this could potentially add $5 billion or more to the net debt of the province. Such an increase in our net debt could have a detrimental impact on our fiscal ratings and may impact on our borrowing cost and on our access to financial markets to finance our operating deficits, our capital account spending and our refinancing of existing debt.