|
|
Most of the electricity produced in the OECD is generated in large generating stations. These stations produce electricity and transmit it first through high voltage transmission systems, then at reduced voltage through local distribution systems to reach electricity consumers.
Distributed generation (DG) plants are different: they produce power on a customer's site or at the site of a local distribution utility and supply power to the local distribution network directly. Distributed generation technologies include engines, small turbines, fuel cells and photovoltaic systems. Despite their small size, DG technologies are already a factor in electricity markets, particularly for high reliability applications, as a source of emergency capacity or to defer the expansion of a local network.
In some markets, DG is actually displacing more costly grid electricity. Worldwide, more distributed generation capacity was ordered in 2000 than for new nuclear power.
Diesel and gas reciprocating engines and gas turbines are well-established commercial distributed generation technologies. Industrial-sized diesel engines can achieve fuel efficiencies in excess of 40% and are relatively low cost per kilowatt. Engines and turbines accounted for most of the DG capacity being installed - approximately 20 GW in the year 2000, or 10% of total capacity ordered. While nearly half of the capacity was ordered for standby use, the strength of the demand for units for continuous or peaking use has also been increasing - see Figure 1.
Other DG technologies have yet to make a large commercial impact. Microturbines form a new technology that converts natural gas to electricity with relatively low emissions. But their capital costs is higher than for natural gas engines, while fuel economy is similar. Fuel cells are the object of intensive research and development, primarily for transportation applications. They have been deployed for power generation in a limited way, but their capital costs will need to fall greatly to be competitive. The cost of photovoltaic systems, while still high, is expected to go on falling over the next decade.
Distributed generation has some economic advantages compared to power from the grid, particularly for on-site power production. First, on-site power production avoids transmission and distribution costs which otherwise amount to about of 30% of the cost of delivered electricity. The possibility of generating and using both heat and power generated in a CHP plant can create additional economic opportunities. Distributed generation may also be better positioned to use low-cost fuels such as landfill gas.
Against these advantages, unit capital costs per kW are higher for DG than for a large plant. Fuel economy is lower, unless used in CHP mode, and DG uses a more limited selection of fuels. For photovoltaic systems, operating costs are very low but high capital costs render it uncompetitive with grid electricity.
The relative prices of retail electricity and fuel costs are critical
to the competitiveness of any DG option. This ratio varies greatly from
country to country. In Japan, for example, where electricity and natural
gas prices are high, DG is attractive only for oil-fired generation. In
other countries, where gas is inexpensive compared to electricity, DG
can become economically attractive - see Figure 2.
Conventional economic assessments of generating options tend to understate the value of DG's flexibility to the owner of generating plant. Many DG technologies can be very flexible in their operation, size and expandability.
A DG plant can operate during periods of high electricity prices (peak periods) and then be switched off during low price periods.
The ease of installation of DG also allows capacity to be expanded readily to take advantage of anticipated high prices. Some DG assets are portable. They can literally 'follow the market'. New analytical techniques, such as 'real option valuation', can quantify the economic value of flexibility.
In addition to this technological flexibility, DG may add value to some power systems by delaying the need to upgrade a congested transmission or distribution network, by reducing distribution losses, and by providing support or ancillary services to the local distribution network.
CHP is economically attractive for DG because of its higher fuel efficiency and low incremental capital costs for heat-recovery equipment. The size of the CHP system matters: the most economical are those that match the heat load. Economies of scale also matter. More than 80% of CHP capacity is in large industrial applications, mostly in four industries: paper, chemicals, petroleum refining and food processing. Even so, much of the CHP capacity in the OECD has been developed as a consequence of supportive government policies. Such policies have also encouraged systems to produce power for export to the grid.
Domestic-level CHP, so-called 'microCHP' is attracting much interest, particularly where it uses external combustion (Stirling) engines and in some cases fuel cells. However, despite the potential for short payback periods, high capital costs for the domestic consumer are a significant barrier to the penetration of these technologies.
The provision of reliable power represents the most important market niche for DG. Emergency diesel generating capacity in buildings, generally not built to export power to the grid, represents several percent of total peak demand for electricity. Growing consumer demand for higher quality electricity (e.g., 'six nines' or 99.9999% reliability) requires on-site power production.
The status of DG differs in each OECD country. While economics are certainly a fundamental factor, differences in government policy can also affect the role that DG plays.
DG is a viable option to many electricity consumers in
Japan because of the country's high electricity prices and limited electricity
market opening. Oil-fired engines are commonly used, because of the relatively
high price of delivered natural gas, except where excluded by air quality
regulations. Over half the systems use CHP.
A survey by the Japan Engine Generator Association (NEGA) estimates that, from 1997 to 2000, installation of DG, excluding emergency power, grew by 2418 MW, about 11% of the amount installed by the utilities during that period. In addition to many independent suppliers, eight of the ten electric utilities in Japan have established subsidiaries to offer DG services.
Several regulatory barriers have been removed by Japan in order to encourage the development of DG and, particularly, of CHP systems. These actions include adjustments to fire regulations and on-site staffing requirements. However, some regulatory barriers still remain. Selling excess DG to another electricity customer is generally not allowed. The costs of electrical protection equipment can be quite substantial: about 10% of the total cost of the facility.
DG in the US is limited by relatively low electricity prices and affected by the widely varied pace of retail electricity market liberalization in the 50 states. CHP accounts for 50.4 GW, or about 6%, of total US electrical generating capacity, nearly all of it in large industrial plants. Emergency power generators have been identified as a potential source of emergency grid capacity. A detailed survey of standby generators in California, by the California Energy Commission, found 3.2 GW of such capacity, equivalent to over 6% of peak electricity demand, in the state.
The are several challenges to DG in the US beyond the question of economic competitiveness. Permitting processes make it difficult and expensive, on a per kilowatt basis, to get siting approval. The lack of a national standard for interconnection further increases transaction costs for DG companies, although such a standard is now under development. Incomplete regulatory reform has left distribution utilities competing with DG. Environmental standards have been toughened in some states, with the same standard applying regardless of the size of the generator. This approach effectively rules out fossil-fired DG in these states.
The Netherlands has an advanced liberalized market where
DG is well established principally because government policies have favoured
CHP and renewable energy sources. However, the general policy thrust of
the Dutch government is to avoid using favourable grid policies or tariffs
to subsidize the development of these technologies, relying instead on
other methods.
The substantial Dutch experience with DG has had some important advantages. Unlike the situation in the US, interconnection rules in the Netherlands are not a problem. Market rules were adjusted soon after they were introduced, so that CHP producers could more accurately predict how much electricity they would supply to the grid. Power parks have been established where the main power producer is the only customer with a direct connection to the grid. But CHP producers have faced difficulties because of rising gas prices and falling electricity prices. To help them cope, the Dutch Government has increased direct subsidies to producers and has encouraged distribution companies to ensure that the network value of DG is appropriately reflected in tariffs.
The UK, which also has an advanced liberalized market, has policies that favour the development of CHP and renewable sources of energy and considers the development of DG in general as an important way to increase competition among electricity producers. Nevertheless, the introduction of new electricity trading rules, known as the New Electricity Trading Arrangements (NETA), have nonetheless proved disadvantageous to small distributed generators because of higher transaction costs, requirements for balancing output against forecast, and, most importantly, because of the fall in power prices. As a consequence, NETA has led to greatly reduced power output to the grid by distributed generators.
In anticipation of these problems, the UK Government commissioned an 'Embedded Generation Working Group' to examine the role of DG in the liberalized market. The group's report, issued in January 2001, identified a number of practical measures to ensure that DG is integrated into the power system in an economically efficient way. A 'Distributed Generation Co-ordinating Group' has been established to follow up on the Working Group's recommendations.
Liberalization of the electricity market itself has had an impact, increasing the complexity and transaction costs for all market players but particularly affecting smaller producers. In western Europe, electricity prices have largely fallen at the same time that natural gas prices rose, resulting in great financial pressure on most distributed generators.
In certain markets where they can avoid charges on transmission, distributed generators may have an advantage over central generation. Elsewhere, in wholesale markets that are designed with large central generation in mind, smaller distributed generators may be at a disadvantage because of the additional costs and complexities of dealing with the market. Difficulties in the NETA market in the UK and in the new Dutch market suggest that further market measures are needed to make the system fair to smaller generators. Furthermore, treatment of connection charges for DG needs to be consistent with treatment of larger generators.
In fact, liberalization of the electricity market is not broad enough to take advantage of the flexibility of many types of DG. Retail pricing, by time and by location, would encourage the development of DG in locations where it can reduce network congestion and operate at times when system prices are high.
DG embraces a wide range of technologies with a wide range of both NOx and greenhouse gas emissions.
Emissions per kWh of NOx from DG (excepting diesel generators) tend to be lower than emissions from a coal-fired power plant or a typical utility system with a large proportion of coal. At the same time, the emissions rate from existing DG (excepting fuel cells and PV) tends to be higher than the 'best available' central generation: a combined cycle gas turbine with advanced emissions control.
This puts a serious limitation on distributed generation
in areas where NOx emissions are rigorously controlled, even where DG
could effect a substantial reduction in emissions compared to the generation
which would be displaced.
For example, new regulations coming into force in East Texas in 2005 will make it unfeasible to use fossil-fired DG, except when powered by fuel cells - see Figure 3.
The case of carbon dioxide emissions is similar. Emissions rates for DG are generally lower than those for coal plants, but not as low as those for new combined cycles.
If, however, DG is used in a CHP mode, there can be significant emissions savings, even compared to combined cycle power plants. Measures should be designed that encourage distributed generators to reduce their emissions.
The use of economic instruments (such as carbon emissions trading) would encourage DG operators to design and operate their facilities in ways that minimize emissions of greenhouse gases.
The energy security considerations affecting distributed power take two forms: first, its impact on the diversification of primary energy supplies, and second, its impact on the reliability of electricity supplies. The impact on primary fuels depends on the underlying technology. Photovoltaic systems help diversify supply away from fossil fuels. Most of the other technologies rely directly or indirectly on natural gas. Since much of the new investment in DG is for natural gas, the effect on increasing diversity in the power system is therefore quite limited. In the case of CHP, higher fuel efficiency would mean lower overall fuel consumption and therefore would be favourable to energy security.
The reliability of electric power systems can be enhanced by DG. The availability of standby generators in the US electricity markets in the summer of 2001 helped reduce the risk of blackouts. Better integration of standby resources into system operations would further enhance security of supply. Furthermore, the use of distributed generators at selected locations helps distributors overcome local bottlenecks. Increasing DG could reduce the demand for transmission, thereby increasing margins on transmission lines. Ultimately, a power system based on a large number of reliable small generators can operate with the same reliability and with a lower capacity margin than a system with equally reliable large generators.
The main negative impact from DG would be in reducing a network's ability to supply primary reserve power if DG technologies are not capable of responding to load changes. This would be the case if most of the DG capacity was nondispatchable because of natural variability (wind and PV) or because of their operating characteristics (CHP where power output is matched to heat demand). The operators of the Nordel system have identified the expansion of wind and CHP as a design concern and are studying how best to address it. They have suggested that 'distributed network regulation', where sub-areas of Nordel each have their own system operator, may be needed.
The IEA's latest World Energy Outlook has incorporated DG explicitly into its 'reference scenario' for the first time. The key methodological step was to assume that demand for DG occupies a unique and growing niche in the demand for electricity, given the strong growth of DG over the past decade, policies in some countries that favour development of CHP, and the increasing need for greater power reliability. Reliability concerns are especially pertinent to developing countries such as India where large consumers have turned to DG to displace requirements for grid.
The reference scenario assumes that global DG will grow by 4.2% annually between 2000 and 2030. New DG, excluding photovoltaic systems, will amount to 521 GW in 2030, roughly 11% of new generating capacity. This includes about 100 GW of fuel cells - see Figure 4.
While DG might remain of secondary importance at the global level over the 30-year period (additional DG produces about 1800 TWh of electricity in 2030, versus global production of over 31,000 TWh), its role is greater inside the OECD, particularly during the last decade of the outlook period. In fact, the Outlook results suggest that over half of the net increase in generation between 2020 and 2030 in the OECD will come from either DG or renewable sources.
Despite the limited penetration of distributed power in today's OECD power markets, the future could have a power system that is much more decentralized than currently. Such a system would have potential advantages with respect to security and reliability of supply. It could emerge from the present system in three stages:
the accommodation of DG in the current system
the creation of a decentralized network system that works in tandem with a centralized generation system
a dispersed system in which most power is generated by decentralized stations with central generation playing a limited role.
There are a few signs that electricity networks are beginning this evolution. New technologies are already being used to control output from DG at several sites to respond to market conditions, creating a kind of 'virtual utility'. The exception to this is an example found with a recent California Distributed Generation Plan.
The operation of a network with large number of virtual utilities will require a much greater flow of real-time information than distribution network operators currently have at their disposal. Over the long haul, there is a need to reform distribution system design requirements to accommodate DG.
This would include upgrading the system to make it capable of accommodating two-way flow, including increased communications and control capabilities. The skills required to operate and manage a distribution system will become more complex. But the technical and institutional changes cited in the IEA report would make a more decentralized electric power system technically feasible.
Distributed generation is expected to play a greater role in OECD power generation over the coming decades. There is a growing interest on the part of power consumers for installing their own generating capacity in order to take advantage of flexible DG technologies to produce power during favourable times, enhance power reliability and quality, or supply heating/cooling needs. While much of this capacity generally contributes little to overall electricity production, it can be expected to become an increasingly important source of peak supply. In this way, DG is contributing to improving the security of electricity supply. However, the introduction of increased wind and CHP systems could increase needs for primary regulation capability.
Retail market liberalization will play the key role in opening up the economically efficient development of DG. Retail market liberalization will give consumers access to the distribution system. Structural reform will leave the distributor indifferent to, rather than in competition with, DG. Unbundled pricing will make it possible, at least in principle, for a distributed generator to capture the value it can bring to a distribution system, and to pay the costs that it imposes.
Nevertheless, there remain substantial institutional and regulatory barriers to develop DG markets fully.
Partially liberalized markets leave DG competing with the utility through its lack of legal access to the distribution grid.
The lack of standards for connection of smaller DG increases transaction costs for distributor and distributed generator.
Emissions regulations can be overly demanding for small sources.
Reforms to OECD electricity markets therefore need to ensure that distributed generators can get access to local electricity grids and do not compete with the distribution company for supply.
Regulators will play a key role in ensuring that distributors are rewarded when they encourage DG that reduces distribution network costs, rather than discriminating against distributed generators.
Standardizing interconnection rules would reduce transaction costs.
Environmental approvals should be streamlined and recognize the net benefits that DG could offer.
Prices that recognize the locational value of DG to support distribution network activities, and to relieve network congestion, are also needed.
If DG does take a large share of the generation market, the role of distribution utilities will become vastly more important than currently. There will be a need to reform distribution system design requirements to accommodate DG. Undertaking further studies to identify the technical capabilities, the operating strategies, and the skill requirements of distribution network operators would help prepare electricity markets for a more decentralized electricity system.
