Fitting the bill18 December 2005
In the mid to late 1980s the world embarked on a fascinating experiment to discover whether electricity could be delivered within a competitive market framework, as opposed to the ‘command-and-control’ model, which had dominated previously. By Malcolm Grimston
Liberalisation of power markets is still at an experimental stage. Although in the 1970s and 1980s reforms were introduced in some countries (USA and Chile) to allow independent power producers to supply electricity to existing monopoly utilities, usually on the basis of inflexible long-term contracts, widespread liberalisation only began with the introduction of competitive electricity markets in the UK in 1990, Norway in 1991 and New Zealand in 1994. The has IEA (International Energy Agency) predicted that by 2007 more than 500 million consumers in OECD countries (about half the total) would have a choice of supplier.
Experience of ‘liberalisation’, in its many and varied forms, has been mixed. In some regions, countries and states such as the UK, Germany, the Nordic region, Texas, New England, PJM (Pennsylvania/New Jersey/Maryland), Chile and New Zealand, experience has been largely positive, with lower power prices and no serious long-term disruption of supply for capacity reasons. In others – California, Alberta, Ontario, Victoria, Italy – liberalisation has been associated with problems in terms of higher power prices and disruptions in supplies. In 2001, the Center for Responsive Politics (USA) described the situation as follows:
Advocates of deregulation say reducing government control of the industry will benefit consumers. But among the twenty-four States that have enacted electricity deregulation plans, results are mixed. Rising prices, skyrocketing demand, and limited supply in some areas have raised questions about the viability of deregulation.
In a number of places – California, Arkansas, Arizona, Montana, Ontario, Queensland, Switzerland, Thailand – steps have been taken to delay, reject or reverse liberalisation. In perhaps the most dramatic example, the Dominican Republic renationalised key elements of its power industry in 2003, just four years after they had been sold off.
Whether this pause is a sign of growing disillusionment with the concept of liberalisation or a short-term response to a series of problems (very high price spikes in Scandinavia, the Netherlands, California, Alberta, Ontario, New Zealand, Argentina; blackouts in New York, Copenhagen, Italy, Victoria, Auckland, London, Birmingham (UK), Athens) is not yet clear. However, by early 2004 statements such as that of the president of the US National Energy Marketers Association, Craig Goodman, were becoming more frequent:
A utility that uses its scarce capital and credit rating to buy and sell a commodity as volatile as gas and electricity and all the risks associated therewith, that has zero upside potential, zero profit and a likelihood that they’ll never recover the costs of that function, is not acting prudently on behalf of their shareholders.
Electricity is a unique commodity. It cannot be stockpiled in large quantities and yet secure supplies on a moment-by-moment basis are enormously important. Blackouts in the developed world, as listed above, have both demonstrated that major outages are possible in developed countries and reminded us that the effects of such outages, in economic, social and health terms, can be significant.
The introduction of competition into electricity supply systems – variously referred to as ‘liberalisation’ or ‘deregulation’, although such terms are something of a misnomer – has become the predominant trend in the markets of developed and, increasingly, developing countries since 1990. Many major organisations, such as the European Union, the Federal Energy Regulatory Commission (USA) and the IEA, have encouraged countries to increase the extent to which both generation and supply of energy are open to market forces, while recognising that some
elements of the process, notably transmission and distribution, are natural monopolies and must therefore continue to be regulated. The details of liberalisation vary significantly from country to country, but broadly most models share some or all of the following features:
- Unbundling of the natural monopolistic elements of electricity provision from those elements that are amenable to competition – generation and supply.
- To a greater or lesser extent barriers on vertical integration between generation and supply and also measures to prevent single players winning too large a share of either or both of these subsectors.
- Introduction of a competitive market in generation, with a range of contracts available in the marketplace.
- Bilateral ‘over-the-counter’ trading.
- An independent system operator responsible for managing the spot market and dispatching plant in real time.
- Competition in the retail market, at least for consumers with large power demands.
- A regulator to oversee such issues as fair competition and mitigation of market power, monitoring of capacity margins, and so on.
It is only now that the capacity margin in some developed countries is reaching the point at which major new investment will be required in the near future. Even in the unusual case of the UK – where liberalisation was accompanied by a major increase in investment in gas-fired generating capacity – investment slowed significantly at the turn of the century, though in that case not before considerable new reserves of capacity had been added.
Of course, reducing the costs of power cannot be the only aim of electricity policy. Modern economies require secure supplies for both business and residential purposes. Energy
production and use have major environmental implications, notably for atmospheric challenges such as climate change and acid rain. Personal safety, public perceptions and political issues (especially the wide range of social issues associated with the availability and pricing of energy) also shape energy policy in important and sometimes unpredictable ways. The relative weight given to such issues will vary from country to country. While it is difficult to generalise, it is fair to say that in the developing world liberalisation has been driven more by a recognition that the state is often incapable of providing the capital necessary to upgrade and maintain electricity supply systems against a background of rapidly increasing demand and a consequent desire to attract private capital and more efficient working practices into the marketplace.
Herein lies perhaps the central tension within the liberalisation of electricity supply systems. If private investment is to be attracted into the industry there must be a reasonable prospect of companies being allowed to profit from good business decisions without the impression that governments and regulators are forever poised to intervene in the marketplace in unpredictable ways. However, governments have a responsibility to ensure that the wider implications of electricity supply (industrial, environmental and social) are taken into account and so will always feel an urge to regulate in order to ‘guide’ the market towards acceptable outcomes.
There is so far little empirical evidence of how liberalised markets will develop when the initial circumstances in which they were instituted have changed. Nonetheless, a number of themes are emerging, particularly within developed countries, which require careful consideration and which may point the way towards energy policies of the future. These involve a number of issues.
New investment signals
What will happen when capacity margins become sufficiently tight to threaten chronic power cuts?
Since electricity cannot be stockpiled, there tend to be long periods, when demand is relatively low, in which prices (especially for power traded through spot markets) fall towards the avoidable costs of the marginal generator.
To compensate, in order to provide a reasonable return on capital employed, there will need to be periods in which prices are significantly above avoidable costs. However, periods of very high prices bring political risk and loud cries for an end to the ‘profiteering’ of the power companies. There is evidence from a range of countries that when the political temperature rises in this way, governments, either directly or via regulatory bodies, do indeed intervene to cap price rises or otherwise ‘guide’ the market. In California, since 2000, there have been price caps at seven different values, ranging from $55 to $750 per MWh. For someone considering making the sizeable investments involved in power generation, the extra risk associated with second-guessing the actions of regulators will inevitably delay, or perhaps even drive out, necessary investment, or (which is in effect the same thing) increase the required rate of return.
One model for attracting investment is one in which future customers provide the initial investment and then buy power, in proportion to that initial investment, at or near avoidable cost. TVO’s new nuclear plant in Finland is being financed on such a basis,
as is Intergen’s Rijnmond CCGT in Rotterdam. However, investment in ‘merchant’ power plants, selling electricity on an open market, may be more problematic. Long-term power contracts between generators and consumers also bring potential problems.
Investors, consumers, society
How can the inherent tensions between the needs of investors, the needs of consumers and the wider needs of society be accommodated within a liberalised power market?
Different communities require different things from an energy system. Governments, for example, will wish to intervene in order to ‘guide’ electricity markets to fulfil political requirements such as secure supplies and low prices. Potential investors, by contrast, require a stable investment environment in which governments have credibly divested themselves of powers to intervene in such fashion. It is not clear that these inherent tensions are easily managed within a competitive electricity supply system.
Is competition possible (or desirable) in a commodity with the unique characteristics of electricity, without determined regulatory effort?
Except where regulators have made major efforts to prevent it, mature liberalised power markets are increasingly characterised by growing concentration, both through consolidation at generation and retail levels and through growing vertical integration across generation and supply, in many cases involving local distribution networks as well. There has also been a growing trend towards integration of gas and electricity industries. Larger companies have greater market power which can be used to erect barriers to new entrants as well as to manipulate prices to higher levels than they might have been in a more competitive environment. In 1998 the five biggest generators in Europe commanded 46% of the market, but by 2002 this had increased to 62%.
It can be argued that, with its associated risks, the nature of investment in electricity generation is such that it can only be funded by large, probably cross-border, operators such as those that are emerging within western Europe. However, it is not yet clear what is the appropriate level of competition (or the appropriate number of competitors) to fulfil the dual goals of creating downward competitive pressure on prices while also retaining entities large enough to contemplate timely investment in new plants.
Should there be no California-style experiences in the near future, confidence in liberalisation may well be restored
Responses to price spikes
Although the price elasticity of electricity demand is notoriously low – many of electricity’s uses are essential and no substitute fuel is available – when capacity margins are very tight a small increase in demand or reduction in available capacity can have a major upward effect on prices. If prices could be transmitted to major consumers in real time it is likely that demand reductions could be found which would serve to control price rises. There is indirect evidence for this in the major demand reductions that followed price crises in California, Norway and Brazil. By contrast, price caps and long-term contracts serve to protect the consumer from price spikes and
so dampen potential demand-side responses. Real-time metering is expensive, requiring both the installation of equipment at the point of use and more sophisticated demand measurement centrally, but may be a more effective approach to managing supply security than attempting to retain high reserve capacity margins.
Different fuel mixes
Are some combinations of generating fuels more susceptible than others to price spiking and/or outages?
Any power plant can come offline, either predictably or unpredictably, for a variety of reasons – maintenance needs, labour strikes, unavailability of cooling water (especially for thermal/nuclear plants), and so on. The need to maintain significant ‘spinning reserves’, ready to take over instantaneously if a plant should fail or demand increase unexpectedly, is one of the arguments that some commentators use against the concept of large-scale centralised generating plant as opposed to smaller-scale generation more embedded in the demand network.
However, there is tentative evidence that two power fuels – renewables (hydropower) and imported gas – have been particularly associated with problems. Periods of low rainfall have led to reductions in the reliability of output from hydro plants, for instance in New Zealand, Chile, Norway and Sweden and the western USA. The volatility of gas prices led to severe problems in areas such as California and Alberta in 2000 and 2001.
By contrast, Texas, with significant gas reserves of its own, was much less affected by the crisis while systems more reliant on domestically mined coal and nuclear power appear to be much less prone to price spiking with its associated political and economic effects. This may be of particular importance in Europe, which is likely to become more dependent on gas imported from the former Soviet Union, and new renewables which, like hydropower, tend to be intermittent in their output.
Although new renewables such as windpower will not suffer chronic problems such as prolonged droughts, and so will not be unavailable for the long periods that have been typical of hydropower in many countries, in the short term they can be considerably more intermittent. In some cases, such as tidal power, output is likely to be predictable but not constant. In the case of wind, output is not only intermittent but also unpredictable beyond a few days’ notice. Managing this intermittency in the absence of a large-scale method of storing electricity represents a challenge, especially at times of high windspeeds when wind generators may need to be immobilised rapidly to prevent damage and alternative capacity must be brought online very quickly.
During times when the renewables are not available, expensive peaking plants using fossil fuels will have to be dispatched, with potentially large effects on system marginal prices (and also potentially on emissions). Some of this capacity must be kept spinning, using fuel and emitting greenhouse gases even when the wind is blowing. E.ON estimates that spinning reserve capacity of some 50-60% of installed wind power capacity, and total ‘shadow capacity’ of some 80% of wind capacity, must be maintained.
The potential problems associated with overdependence on unreliable electricity sources were illustrated during the hot spell in June 2003. Increased demand for air conditioning was coupled with drought and a significant reduction in availability of power from other sources, notably large windfarms (and also from inland conventional power stations including French and German nuclear plants). This led not only to very high prices but also to a major reduction in exports from France and Germany, resulting in blackouts in Italy, which was highly dependent on such imports. Although on this occasion the contribution of the shortage of wind was not the crucial factor, it would have been more important had there been considerably more reliance on windpower in the area.
Considerations of this nature were behind the decision taken in December 2003 by the Irish energy regulator to halt connections of windpower to the Irish grid. Kieran O’Brien, managing director of ESB National Grid, said wind connections “pose an increased risk to the security and stability of the power system which exceed the level normally likely to be acceptable by a prudent system operator.”
Of course, any attempt on the part of governments to manage the fuel mix, to encourage greater use of indigenous energy supplies to reduce dependence on imports, will potentially conflict with the rationale of liberalising the market in the first place.
What are the implications of using cross-border interconnectors, generally created to enable well-coordinated bilateral electricity trading between neighbouring countries, as major components in international power markets in which large numbers of relatively uncoordinated trades take place on a continual basis?
The growth of regional markets has been one of the most notable developments of the liberalised era. Regional markets offer a number of advantages, not least in creating a larger market which may stimulate more effective competition and also in reducing the demand for each participating nation to retain such a large capacity margin. For example, unscheduled plant breakdown might be compensated by imports while in geographical areas which are extended over several longitudes the time of peak demand will be different in different parts of the regional market, allowing net east-west or west-east flows at different times to replace reserve capacity.
However, with rare exceptions, although the interconnectors and national wires involved are in effect playing the part of a grid serving a single market, there is no single regional transmission operator to ensure smooth running. In August and September 2003 transmission system failures underlay major power outages in the northeast of North America, in England (London and Birmingham), in Copenhagen and south Sweden and in Italy, while a reduction in the availability of imports (coupled in some cases with anomalous pressure to increase exports despite shortages of power at home) were important factors in crises in Italy in June 2003, in California in 2000 and 2001 and in Victoria in 2000.
As a result a number of questions are arising. Will the cost requirements needed to strengthen cross-border interconnectors and national grids outweigh the savings in generating capacity? Who should bear the cost of such strengthening and how will it be financed? Is there a danger that growing reliance on interconnection will further undermine investment in generating capacity, resulting in a very efficient grid but insufficient power to drive it? How can differences in national markets within a regional interconnection be managed (a major issue for example during the power crises in California and Victoria in 2000 and 2001)?
Following the California crisis there have been varying degrees of reversal of competition measures in countries and states such as Arizona, Arkansas, Montana, Ontario, Queensland and Switzerland. Should there be no California-style experiences in the near future, confidence in liberalisation may well be restored; but should there be another similar event it is an open question as to whether more countries would retreat from liberalisation and return to some degree of central planning.
The implications for investment in the major restructuring of electricity markets have been profound. Although liberalisation in most countries is still in its early days (when judged against the lifetimes of investments in the electricity industry) a number of themes are emerging: a reduction in the capacity margins which tended to be preserved within command-and-control electricity supply systems (caused at least in part by a shift in emphasis from the importance of preserving secure supplies to a desire to reduce costs and increase profits); a major shift from investment in sources with high capital/low operational cost profiles (notably nuclear power) to the CCGT (with low initial investment but high operating costs, especially at times of high fuel prices); growth in integration (both vertical and horizontal) where market rules allow it; increased cross-border trading of electricity; and periods of extreme volatility in wholesale power prices, often accompanied by government intervention in the marketplace.
If the next few years bring experiences which suggest that highly fragmented power markets are incompatible with early investment to prevent power cuts, then the most basic current assumptions about competitive market structures and the associated implications for investment in new generating capacity may prove to be invalid. A return to command-and-control power markets is not the only option, and indeed at present looks highly unlikely. However, a European electricity market dominated by the ‘seven brothers’ (EdF, RWE, E.ON, Enel, Vattenfall, Endesa and Electrabel), each with considerable market power and each with significant involvement in supply as well as generation, could well create an environment in which longer-term investment is attractive. A sensible approach to future electricity production might well involve planning for this scenario, for example by ensuring that technical options are ready to be deployed if and when required, rather than leaving everything to the short-term forces which now predominate.
For policymakers, then, the following steps would seem to be important.
Make a firm statement that secure power supplies are not merely a matter for the marketplace and that government must play a role in shaping market rules to encourage maintenance of appropriate capacity margins, for example by considering the option of capacity markets or capacity payments.
Recognise that a perception that government or regulators may intervene capriciously in markets may damage the confidence of investors to commit to funding new generating capacity at the appropriate time. This involves a recognition that very high power prices may not be reflective of abuse of market power but may be necessary to send the signals for new investment. Regulators should avoid price caps and should develop a scheme of compensation for power generators which lose income because of regulatory action.
Monitor more closely the mix of different types of capacity, recognising that some capacity mixes are more susceptible to interruptions in supply or long-term price spiking than others.
Encourage installation of real-time metering for major electricity consumers in order to maximise demand-side responses to high wholesale and balancing electricity prices.
Work towards developing cross-border system operators and system rules to respond to the (at present largely uncoordinated) growth of international trading in electricity.
Recognise that no single generic model of liberalisation will be appropriate for all circumstances and that schemes should be designed with the particular requirements of the country or region in question in mind. For example, different measures will be needed in countries with developed electricity systems and reasonable capacity margins to those needed in countries with rapidly growing demand and heavy social pressure to provide electricity services at affordable cost to people who do not presently have access to them.
|Should there be no California-style experiences in the near future, confidence in liberalisation may well be restored|