A statement regarding the environment, from a Pogo comic book, by Walt Kelly, circa 1969
We have met the enemy, and he is us.
Translated to the electricity industry, Ontario, 1999
We have met the shareholder, and he is us.
I’m going to talk about three things today.
- Methods for calculating stranded debt in other jurisdictions as compared to the method used by the Ministry of Finance.
- The Nuclear Asset Optimization Program and its relationship to stranded debt.
- Treatment of nuclear decommissioning costs, NUG contract costs, and other regulatory assets.
But first, a few back-of-the-envelope calculations.
- It doesn’t take an economist to know that something is radically wrong. The entire value of Ontario Hydro Genco, after spending $6 billion to fix its nuclear plants, is $5 billion.
Hydro today has 27000 MW of working generation, plus 5000 MW of dead nuclear generation that is allegedly on the road to resurrection. Assuming that Pickering A and Bruce A can be revived cost-effectively, as the government and Ontario Hydro assume, $5 billion divided by 32,000 MW is $156 per kilowatt.
Let’s assume that all of Ontario Hydro’s nuclear fleet is simply worthless – has no value. Then Ontario Hydro has roughly 18000 MW of non-nuclear units – 6500 MW of hydraulic, 9000 MW of coal, and 2500 MW of oil and gas. The $5 billion divided by 18000 MW is $277 per kilowatt. This figure is less than the net book value of hydraulic and fossil units (plant plus inventories).
The financial information presented by the Ministry suggests that Genco will have a value of about 7 times earnings. In the stock market, it would likely sell at a premium to book value based on 10-15 times earnings.
Stranded Cost in Other Jurisdictions
There are two ways to set stranded costs: administrative determination by a regulatory body, or market-based, preferably using divestiture, occasionally using appraisals.
Ontario is using an administrative method, but a very different method than most places. It is different in two key ways. First, Ontario does not clearly separate past costs from future costs. Decommissioning costs
are lumped in with future costs to determine value. Second and more importantly, Ontario uses an aggregate corporate-wide calculation.
In most places, administrative stranded costs are determined on a plant-by-plant or station-by-station basis. There is a good reason for this. To explain, I need to introduce a term "going forward costs."
These are the future costs of fuel, OM&A, and capital additions. If the expected present value of revenues from the plant at market prices are less than its present value of going forward costs, we have a
problem. The plant is worth more dead than alive and should probably be closed. We need to look for this problem in close cases. Stranded cost analyses of a number of nuclear plants in the U.S. have shown that the
plants have negative values and should be closed. Some of the oldest fossil units are also not economic.
If we include negative stranded cost values for some plants, we are giving them an operating subsidy. Operating subsidies clearly harm competitors. Operating subsidies also clearly harm ratepayers. The plant is
worth more dead than alive, but it lives on and ratepayers pick up the tab.
The outputs of an administrative calculation of stranded debt are incredibly sensitive to the assumptions used by the analysts. The Ministry of Finance states that Ontario Hydro would have no stranded debt if market
prices were 1 cent per kWh higher. One penny swings plant valuation by $8 billion. I am not saying this because we should wish our problems away with rosy scenarios, but just to point out sensitivity of results to
assumptions. In another example, Virginia Power claimed that its generators had a stranded cost of $857 million. I changed two assumptions in the utility’s analysis – the life expectancy of fossil plants
and the cost of a combustion turbine – and I got a $3 billion swing. The units are actually worth at least $2 billion more than their book value.
Let me make one more comment about administrative determination of stranded costs. Too many people across North America want to do confidential, black box analyses, although a few utilities, to their credit,
including General Public Utilities, owner of the Three Mile Island nuclear plant as well as blocks of coal and other fossil generation, made their cost and performance assumptions public. The public needs to
understand the basis for these cost calculations. After all, the public is going to pay the freight.
As an example of the perniciousness of confidential regulation, a New Jersey utility justified a $180 million steam generator repair at the Salem nuclear plant using one set of rosy market price assumptions,
and later said that Salem had a total value which was less than the cost of the repair using a different, lower set of market price assumptions. In fact, when administrative costs were properly assigned to the
nuclear plant, its value was negative, and it should have been closed. Sound familiar? The details cannot be shared with the public beyond this high level gloss. In addition, the same utility included millions of
dollars of anti-competitive costs as part of stranded administrative costs. I can’t tell you, or the people of New Jersey, what those costs are.
Governments would never impose secret taxes, without explaining what the taxes were for or how they were calculated. But across North America, many governments have not hesitated to impose secret electricity stranded
costs without clearly articulating the basis for them and allowing an open public debate.
Comparing Divestiture with Administrative Determination
The divestiture method has tended to yield higher prices than administrative stranded costs in places where both were used. The Ministry of Finance has cautioned us not to
extrapolate high asset sales prices from the U.S. to Ontario. I agree in principle that this warning needs to be taken seriously.
However, one can compare administrative estimates to prices received for divested fossil generation for the same utilities. This comparison takes into account and controls for the price environment.
The working hypothesis of the California PUC was that gas-fired generators would sell somewhere between zero and book value. In actuality, buyers paid nearly twice book value. General Public Utilities sold its fossil
generation at 2.5 times book value. That generation was originally viewed by GPU as being worth slightly above book value.
Administrative Stranded Costs for Ontario Hydro: More Back of the Envelope (er, Spreadsheet) Calculations
On a very preliminary basis, I calculated an administrative value for hydro’s generation. I used a discounted cash flow method at the level of hydraulic, fossil, and nuclear,
similar to that used by the utilities in New Jersey and Virginia, although much simplified. These valuation methods that I used account, in rough terms, for the income taxes paid by a private sector company, similar
to the Ministry’s method. I used a composite cost of capital of 9% for hydraulic and fossil and 12% for nuclear. These numbers are not perfect and merit significant refinement after closer review of the
limited amount of public data that are currently available. But they are indicative.
This very preliminary estimate is that hydraulic assets are by themselves worth $7-$8 billion, and fossil assets are worth $2-$3 billion.
Leaving aside decommissioning as a separate stranded cost component, nuclear assets are worth about $2.5 billion under the Run Till They Break Scenario. Under this scenario, Ontario Hydro would give up on Bruce
A and Pickering A, but commit funding for OM&A and routine capital additions to Bruce B, Pickering B, and Darlington. It would be planned to close units prior to retubing, and they might close sooner if major
problems arise. For purposes of this scenario, I assume an average remaining life of 15 years and a 70% capacity factor. Average nuclear production is slightly over 55 TWh per year.
The sum of these numbers yields an administratively determined value of $11-$13 billion plus some dedicated revenue in lieu of taxes.
Divestiture has demonstrated that administrative methods clearly undervalue fossil units. Coal-fired units have been sold for $400-$1000 U.S., which would yield high-end fossil values as much as $7 billion, assuming
a price of $800/kW Canadian for all coal units and (for conservatism), a zero value for Lennox and combustion turbines. This could add $4-5 billion to the estimates above, for a total of $11 to $17 billion plus
payments in lieu of taxes.
Nuclear Asset Optimization Program
The Nuclear Asset Optimization Program (NAOP) involves the return to service of Bruce A and Pickering A and plans to increase Hydro’s nuclear generation to an 88% capacity
factor. That is nearly twice as much generation as in the Run Till They Break scenario at a cost of $6.3 billion from Hydro’s 1997 Annual Report. While NAOP includes some costs of replacement coal-fired
generation that are not true incremental costs, it does not include all the capital spending identified for Pickering A and Bruce A.
NAOP is highly risky. By spending $6 billion, NAOP may add $6-$8 billion in value to Hydro’s nuclear fleet, in very rough terms, if it achieves 88% and if prices are unaffected.
However, the NAOP is likely to have other effects. First, if an additional 55 TWh of nuclear power can be delivered, it is likely to depress the market price, at least temporarily. This flood of nuclear energy would
require even more heroic efforts to exercise market power to prop the price up to 38 mills. This reduces the value of the nuclear generation itself. If a 10% reduction in market price were to arise from this flood
of nuclear energy, it would mean that Ontario Hydro would get only $4-5 billion in added nuclear value from its $6 billion expenditure.
A lower market price will also reduce the value of all other Ontario Hydro assets and increase stranded NUG contract costs. For example, the back of my envelope suggests that the value of Hydro’s hydraulic
assets would be reduced by $1 billion if the market price falls by 10%, and NUG stranded costs would increase by nearly $500 million.
There will be a double-whammy on the value of Hydro’s fossil units. If NAOP is successful, it will not only reduce the price paid for fossil generation but will also reduce the amount of energy produced
by Hydro’s fossil units. These two factors will significantly reduce the value of those assets by giving them both a lower price and a lower kWh output over which to recover fixed costs. We may end up
re-mothballing some coal and oil units.
In sum, can no longer use the old paradigm from the regulated world, where reductions in the market price of power clearly have value to ratepayers, because those reductions bring about equal and offsetting increases
in the amount of stranded debt.
In essence, to the extent that the NAOP is successful, it has the potential to add limited amounts to the value of nuclear power if everything goes right, but it could cannibalize the value of the rest of Ontario
Hydro if the additional nuclear energy causes the market price to decline.
To minimize stranded debt, Hydro and the MOF need to develop a strategy that maximizes the value of its entire company, not just, arguably, the nuclear portion of the Genco. NAOP does not do that.
NAOP also has asymmetrical risks. There is more downside risk than upside risk to an 88% capacity factor for a fleet of nuclear units that is passing the middle of its life. There are only 12 percentage points
between 88% and perfection. Thus an increase of 3 or 4 percentage points might indicate performance that is heroically better than expectations. Worse than average percentage performance could mean a decline of 5,
10, or even at the outside 20 points. This is a case where the expected value must, for reasons of pure statistics, be lower than the projected value. If NAOP does not meet the 88% goal, it will also subtract value
from Ontario Hydro.
Run Till They Break should make rating agencies happier because it is a less capital intensive strategy. Most of the $4.5 billion of capital spending that Genco is projected to have from 1999-2003 would be
unnecessary, thus reducing Genco’s need to raise cash, and possibly allowing other financial ratios to be relaxed.
Finally, NAOP is the last vestige of an activist Ontario Hydro. Why do we give Ontario Hydro money to bulk up and become a 900 pound gorilla for competition, only to require it to divest later? If NAOP is to go
forward, let Ontario Hydro privatize nuclear units either now or as part of the market power mitigation plan. Let the new private sector owners, whether British Energy and PECO, the Power Workers’ Union, or
someone else, take the risk and reap the rewards.
No one would ever say that building 5000 MW of new generation is the current function of Ontario Hydro in the upcoming competitive market. NAOP is no different. Why is spending $6 billion – more than the
cost of building new powerplants – to resurrect 5000 MW of dead generation the function of Ontario Hydro in the upcoming competitive market?
Future Stranded Debt True-Ups
The Ministry of Finance states that it can true up its stranded debt forecast if things go well – if assets are sold at high prices, if market prices rise, or if Ontario Hydro
becomes more efficient. True-ups could also share negative outcomes with ratepayers, but they expect the probability of negative outcomes to be relatively low under a conservative valuation standard.
However, such true-ups only make sense if NAOP does not go forward. If NAOP goes forward, we have Hydro business as usual. Hydro doesn’t take the risk that it may be wrong. True-ups transfer 100% of the
risk of NAOP to ratepayers and mean that the single biggest investment in generation in the next decade has no competitive risk. In essence, Hydro will not face the discipline of the marketplace for its $6 billion
dollar investment and will be backstopped by the government through the true-up if it doesn’t work according to plan – if it drives market prices down, costs more than $6 billion, or doesn’t
produce the 88% capacity factor.
Nuclear decommissioning and fuel disposal is very expensive and the costs will have to be paid in the future. Ontario Hydro has collected $2.8 billion for nuclear decommissioning,
accrued fuel channel removal, and fuel disposal. That money is simply gone. Canadian regulation has not been as vigilant as U.S. regulation. In the U.S, (started by the California PUC and later adopted by the
Nuclear Regulatory Commission) utilities must invest decommissioning funds in special segregated trust funds, external to the utility, that cannot be touched by the utility for purposes other than decommissioning
and cannot be attached by creditors in bankruptcy. U.S. utilities, like Ontario Hydro, wanted to invest those funds in their own operations. Regulators, consumers, and nuclear activists recognized the obvious
– that if the funds were invested in utility operations, it would be hard to get the money out, particularly in the event of an accident, but even in the event of routine decommissioning. This problem has pven
itself in spades for Ontario Hydro. It doesn’t really have this money anywhere except on its books. Its negative equity is greater than its decommissioning funds, and it would have to borrow the money.
The ratepayers paid once for this money, and now they will have to pay for it a second time in stranded debt because it has been spent and no longer is available. It is not clear where this money has been
accounted for in the stranded debt calculation because the working papers are not public. The ministry of Finance presentation speaks of $2.4 billion of unfunded decommissioning obligations, but those appear to be
future, not present obligations. In addition, it is stated that one of the reasons for Genco’s low valuation is payment of nuclear decommissioning from market revenues. So there are essentially three places
where decommissioning costs are muddled – in the valuation of Genco, the stranded debt figures, payments made by Genco included in operating expenses by the Ministry. There is no clear accounting of
decommissioning costs in the Ministry’s calculations to prove that ratepayers are paying all of the necessary decommissioning cost, but none of it is double-counted.
What lessons should we learn about decommissioning?
- Establish a segregated fund now! If Genco is a competitive business that is likely to be diminishing in size, at least in Ontario, as envisaged by the Ministry and most market players, it’s a bad idea to
allow Genco to reinvest decommissioning dollars in its own business. It will be difficult if not impossible to get the money when you need it, and it would be virtually impossible in the event of an accident.
Make Genco (and the Government if it collects money through stranded debt) put the cash in a trust fund where it will be invested in stocks and bonds. Make any new buyer of nuclear assets do the same thing.
Learn from the U.S. You are only 10 years and $2 billion too late.
Separate the past from the future. Most decommissioning and past fuel disposal costs are the responsibility of ratepayers and should be collected through non-bypassable charges, because they were created when the plant first became radioactive. We have a moral obligation to clean up the garbage. But they are not part of the costs of a going-forward Genco.
Future fuel disposal costs and any costs of fuel channel removal as part of retubing should be recovered in market prices by Genco, because they are going-forward costs. In addition, 5-10% of the future share of decommissioning costs should also be recovered in market prices.
Most U.S. states have pulled the portion of decommissioning that is a ratepayer responsibility out of the stranded cost calculations and treat it separately, allowing money to be recovered from non-bypassable charges over longer periods of time than stranded debt recovery. Consider that for Ontario.
Nuclear decommissioning needs to be brought into the sunshine. There is nothing confidential. Ontario citizens deserve a clear accounting of decommissioning and fuel disposal costs, to know whether any money is available now from the $2 billion of past contributions, and to understand how future costs will be collected (from a combination of non-bypassable charges on ratepayers and market prices), how money will be invested, and how money will be protected for decommissioning use only. As stated above, it appears that the Ministry of Finance has included part of the decommissioning costs in several different places rather than clearly and concisely stating what is owed.
The present value of uneconomic NUG contracts shown in the stranded debt report is probably overstated by several billion dollars. Ontario Hydro claims that total payments for NUGs
have a present value of $6.2 billion over 20 years in its annual report. The Ministry suggests that the uneconomic portion is $5 billion. The back of my envelope suggests that it is closer to $3 billion. Better
estimates should be made based on clear assumptions.
Ontario should also consider paying NUG costs above market, as they materialize over the remaining terms of their contracts, through non-bypassable charges, rather than trying to forecast these amounts and make
wildly different present value estimates of stranded debt that are dependent on discount rates and market prices. It should also encourage contract buy-outs entered into voluntarily between NUGs and Ontario Hydro.
Many states in the U.S. do this.
Pensions and Benefits
Finally, I hope that there are no unpleasant surprises in the assessment of regulatory assets for stranded debt that have not been disclosed in the Ministry’s initial
presentation. In particular, lurking in the recesses of Ontario Hydro’s balance sheets are $875 million of past pensions and post-retirement benefit obligations. These costs should not suddenly become part of
stranded debt when the preliminary numbers become final.
First, the portion of these costs associated with transmission and distribution can be recovered from those functions on a cost-of-service basis without affecting the value of Servco. Utilities in US are doing
this now. Servco costs are simply not stranded.
Second, and more importantly, Ontario Hydro’s pension plan at the end of 1997 contained $3.7 billion more in assets than its projected benefit obligations. While a healthy cushion is important, deferred pension
costs of $112 million and $763 million charged for future post-employment benefits can be covered from these pension surpluses rather than through stranded debt.
There are four main lessons from this preliminary analysis.
- The Ministry’s forecasts of stranded debt are likely to be too high, with the possible exception of decommissioning costs.
- More data should be made public to enable parties to make better estimates and engage in informed debate on a basis that is clearly understood.
NAOP is the last vestige of the old Ontario Hydro. Allowing Hydro to spend billions of dollars to resurrect 5000 MW of dead nuclear generation is incompatible with the new competitive market. It is also an extremely risky undertaking.
Ontario needs a clear accounting of decommissioning costs, past and future, answers as to whether any of the $2.8 billion previously paid is available to fund decommissioning and fuel disposal, and a segregation of all future payments for these purposes in a dedicated trust fund.