New UK Electricity Regulatory Regime: Lessons from Year One
One year after its introduction is a good time to review how the new regime regulating the English and Welsh electricity industry has settled down and how purchasers of utilities there have sought to protect themselves from the vagaries of the market.
Has the new regime provided a useful model for regulators elsewhere? How effective have tolling agreements been at providing bankable revenue streams? This article looks at the new electricity trading arrangements — or “NETA” — from these two perspectives.
NETA became effective in England and Wales in March 2001, sweeping aside the pool system that had been in place since deregulation in 1990.
Generators used to bid to supply electricity to the pool at a certain price for each half hour, and the National Grid Company dispatched the generators on the comparative merits of those bids, then sold the electricity on to suppliers. The UK government took the view that the pool was weighted in favor of generators and served to inflate prices generally. NETA was devised with the aim of encouraging competition, promoting market liquidity, and reducing system balancing costs.
The UK regulator reported in February 2002 that, since NETA, electricity prices have been reduced by 20% to 25% and daily balancing costs by 50%, while liquidity is up by multiples of two and three respectively in the case of trades and contracts.
Balancing and Settlement Code
Under NETA, there is no bidding into a central pool, no central dispatch and no centrally determined price. Instead, generators self-dispatch on the basis of bilateral contracts they have freely entered into with trading parties. These bilateral contracts include contracts for the physical sale of electricity or non-physical trades on power exchanges. However, the contracting parties — be they generators, suppliers or traders — must be party to the “Balancing and Settlement Code,” or “BSC,” and it is in this respect that the National Grid Company retains its role as ringmaster. The grid company is responsible via the BSC for balancing the system in terms of frequency and voltage as well as in terms of generation and demand.
The BSC applies equally to consumers, traders and generators of electricity.
The grid company carries out its grid balancing functions through a subsidiary called Elexon.
Each wholesale generator, retail supplier and trader must notify Elexon how much electricity it has contracted to generate or take for each half hour period of the day. These notifications give Elexon the data it needs to manage the grid.
A generator may also indicate how far it is willing to increase or decrease the volume specified in the notice and at what price. Each offer to increase generation at a price must be matched by a bid to reduce by the same volume of electricity at another price: together these are each referred to as a “bid-offer pair.”
This enables Elexon to accept offers and bids in order to balance the system (in terms of supply and demand and frequency and voltage). To discourage gaming, failure to perform in accordance with a bid-offer acceptance results in liability for a non-delivery charge, which is the cost to the grid of covering the excess or shortfall.
Before NETA was formally implemented, many people wondered whether any generators would take the risk of trading primarily by way of this balancing mechanism of bid-offer pairs rather than through bilateral contracts. This has not occurred, at least not yet.
Any generator or other supplier and customer who fails to match metered output with contracted output must pay “imbalance charges.” Because they can be very onerous financially, imbalance charges are probably the most notable feature of NETA. The imbalance charges are not calculated against the amount of power the generator notifies the grid in advance that it will produce, but rather they are payable as against deviations from two other notices, called ECVNs and MVRNs. ECVN stands for “energy contract volume notification,” MVRN for “meter volume reallocation notification.”
ECVNs and MRVNs work as follows. Any trading of electricity by a person subject to the BSC regime must be notified centrally. ECVNs or MVRNs are the two alternative ways of trading electricity under the BSC. Each BSC party has an energy account. ECVNs or MVRNs identify the volume of electricity that is to be debited from or credited to the relevant energy account.
Under an ECVN, the two parties to a transaction agree on the specific volume of electricity that is to be accounted for. For example, if a generator is selling 100 megawatts to a retail supplier, then that 100 megawatts is debited from the generator’s energy account and credited to the energy account of the supplier.
MVRNs are used when a generator wishes to transfer its exposure to potential imbalance charges and the risk of fluctuation in electricity prices for a fixed volume or percentage of its metered output to a person responsible for another generating unit. This transfer of liability is in itself a trade. The MVRN is the method generally used in a tolling agreement, particularly one involving 100% of output. The energy account of the toller is debited with the reallocated volume. The toller is treated in all respects as if the output was its own. The result is that it takes all of the regulatory upside and downside in respect of that output (for example, for transmission losses and imbalance charges). The generator — having shed liability — ends up simply with a fixed revenue stream.
Elexon will notify BSC parties if any imbalance charges are payable. If a generator generates more than his notice said he would, Elexon will need to sell that excess by way of accepting bids from other generators or offers from retail suppliers. If a generator generates less than he said he would, Elexon will need to find another generator to generate more or a retail supplier to accept less. It does both of these things by looking at the bid-offer pairs.
Although the term “charge” is used, an imbalance charge may actually be a benefit to a generator or other BSC party. The amount of any imbalance charge is equal to the amount the grid company is able to collect for the excess or must pay to cover the shortage. There is no cap on these prices or, consequently, on imbalance charges; this may be a matter of concern to lenders to a limited recourse project. A generator who generates more than it said it would is paid as an imbalance “charge” what the grid company is able to collect for the excess power. If it generates too little, it must pay an imbalance charge to the grid. Equally, buyers will be charged for uncontracted supply and will be paid for contracted volumes that are less than those consumed. Parties should be alert to the fact that the price for uncontracted generation might at times be negative, meaning that payment is required from the generator: If generators are not inclined to reduce output, then they may well end up being charged for producing excess energy.
Issues with NETA
NETA is an apparent success in that liquidity is up while prices and balancing costs are down (although prices might have gone down in any event); but has the new regime thrown up any major problems? The short answer is yes.
Accidental Imbalance Charges. There have been concerns about imbalance charges that have been accidentally incurred. For example, London Electricity is reported to have lost £7.5 million because of faulty reporting on ECVNs and MVRNs. Scottish Power lost £10 million. Having reviewed this, the BSC panel recommended in late 2001 that BSC parties should be able to recover losses resulting from failure of contract reporting systems. This is a problem that has not been wholly resolved and may in any event re-emerge in different contexts.
Electricity “Spilled “During Commissioning. Commissioning of a power plant necessitates generation of electricity. The volumes and timing are unpredictable and can lead to imbalance charges. The upside is that the generator will be entitled to be paid for the volume spilled. However, the generator will not be paid anything by the grid if aggregate generation exceeds aggregate consumption. In such circumstances, recipients will have to be paid to accept more electricity or other generators will need to be paid to shut down; and, in either case, that cost will ultimately be passed on to the generator who is spilling. Generators have looked for ways to mitigate this risk, and there have been proposals for derivatives to cover it.
Renewable Energy Sector. The UK government has made much of increasing the renewable energy sector’s share of the UK electricity market. This applies in particular to wind farms. Wind farms do not lend themselves to the NETA regime. In order to reduce the potential for imbalance charges, the possibility has been raised of wind farms clubbing together and selling their product in aggregate through an agent. While this should mitigate the problem, it will surely not solve it given the relatively small size of the UK.
Combined Heat and Power. The same type of problem affects cogeneration facilities as affects wind farms, given that cogeneration plants are designed to be variable while NETA puts a premium on guaranteed output. In the first part of 2002, Powergen dropped the anticipated sale of its cogeneration portfolio, purportedly citing NETA as the source of its difficulty. Perhaps, given that they are less likely to behave uniformly, clubbing together to sell power in aggregate would be more appropriate for cogeneration plants than for wind farms.
Base Load to Ramping. Most of the UK’s stock of power plants has historically been designed for base load use rather than ramping up and down. The balancing mechanism of NETA with its bid-offer pairs gives considerable financial incentive to plants being flexible and able to ramp up and down. NETA would therefore appear to favor peaker plants and to be least desirable to nuclear plants.
There have been concerns among developers about the “bankability” of independent power projects in the UK under NETA. These concerns may be exaggerated. Plants previously financed on a merchant basis are still running. In any event, tolling agreements appear to offer lenders protection and have been used in limited recourse financings in the last year. Some points on how tolling agreements have needed to be adjusted in a number of ways to cater for NETA are worth highlighting.
The toller pays the generator a fixed capacity charge in return for an availability commitment. The toller looks to the generator to be responsible for imbalance charges arising from forced outages. The principal mitigation available to the generator is that its liability for imbalance charges should last no longer than 3 1/2 hours. The reason is that that is the minimum prior notice period for any ECVNs and MVRNs. Thus, as soon an outage occurs, the generator will notify that output from 3 1/2 hours onwards will be nil. However, given that the amount of imbalance charges changes every half hour and is potentially uncapped, the generator might still find this liability unmanageable and require a megawatt-hour financial cap on its liability.
Ramping up and down can cause problems. A toller who is entitled to 100% of capacity will want to direct the generator what output should be from time to time. That way, the toller can fulfill its obligations under its bilateral contracts as well as maximize revenues by playing the balancing mechanism. The real money to be made is in being able to generate more — or reduce output — on short notice; premium prices will be paid for this. This means the toller will be directing the generator to ramp the plant up and down. Meanwhile, the generator will be reluctant to push the plant to perform to its limits with the consequential extra maintenance risk and will therefore be keen to exercise its right to reject any direction that requires the plant to be operated outside of its dynamic parameters. The interests of the parties diverge. Perhaps the best way to resolve this is for the toller to share some of the benefit from playing the balancing mechanism and give the generator a financial incentive to enhance the flexibility of the plant.
Problems arise with notices. An ECVN or MVRN is sent to Elexon’s agent who acknowledges receipt and may then either accept or reject it. The acknowledgement does not of itself constitute acceptance. The BSC party may not be alert to a problem. For example, a rejection may go unnoticed because it has gone to a different “folder” than that in which the acknowledgement was received. In such circumstances, liability for imbalance charges could follow. This type of problem raises the issue of what level of due diligence should be required of people responsible for managing the notification process. In the context of a tolling agreement, it shows the desirability of defining precisely the boundaries of responsibility between the parties for notification risk instead of, for example, relying on concepts like willful default which may be difficult to apply in practice.
Force majeure has been used as a way of managing notification risk. In such circumstances, the toller may assume the imbalance charges risk and still be obliged to pay the generator a capacity charge, thus keeping both lenders and equity financially whole.
The answer to the two questions raised at the start is NETA appears to have worked fairly smoothly — but not without hitches — and to be cost-efficient. Of most concern are the level of imbalance charges and the lack of flexibility of the regime imposing them. Tolling agreements have been negotiated that have apportioned NETA risks to the satisfaction of lenders, which will be of comfort to potential investors.