An American lesson in positive pragmatism

The imperative for establishing how much it costs to integrate wind into the power systems of today has been much greater in the market economy of the United States than in mainland Europe. The emerging consensus in America is that wind's fluctuating supply adds very little cost -- so little that in some cases it is being ignored

A consistent policy across the US for how wind generation is integrated into utility systems, and how balancing markets treat wind resources, does not exist. A commendable attempt by the Federal Energy Regulatory Commission (FERC) to introduce a Standard Market Design (SMD) for the entire US power industry was stopped in its tracks last year when strong utility opposition killed the FERC's proposal. Policies for integration of wind, however, are being introduced, though in piecemeal fashion at regional levels.

The SMD, a market-friendly plan for open access to transmission and retail competition, would have adopted a California-like method of integrating wind power generation into power grids by removing the energy imbalance penalties from intermittent resources and averaging imbalances over a month. It also required the use of wind energy forecasting to ease the assimilation of wind on the power system.

Without such a singular policy, the owners of transmission lines across the US are left with a variety of rules -- mostly their own -- for integrating wind. Many are now struggling to understand the impact wind generators have on their system.

Size matters

While most US system operators agree that intermittent resources place some burden on the power system, they do not always agree what penetration of wind is needed before impacts are significant. Jim Caldwell of the American Wind Energy Association (AWEA) says a system will not begin to feel a significant presence of wind until it is close to supplying 20% of demand, unless the system is small, inflexible and weakly linked to surrounding transmission control areas. Up to that 20% mark, says Caldwell, reserve requirements for wind should not be much different than for other resources on the system.

The Northwest Power Pool (NWPP) agrees. It recently estimated its reserve requirement for the 620 MW of wind in its region at 5% of the wind capacity. The pool operated by NWPP, a loose grouping of generating utilities serving a huge area covering the US Northwest and the Canadian provinces of British Columbia and Alberta, has a winter peak of about 50,000 MW. It assesses other resources making up the bulk of its capacity, including hydro and thermal generators, as needing reserves in the range of 5-7% of their load.

Conclusions about the cost of intermittent resources on a system, however, range widely depending on the parameters, from a negligible impact on the California system to estimates of $5.50/MWh -- $2.50 for incremental reserves and $3 for imbalance costs -- at a 20% penetration for wind power for PacifiCorp in the Northwest. Other studies fall between these two marks:

o The impact of adding 2000 MW to the WeEnergies 8000 MW Midwest system is $2-$3/MWh.

o The Utility Wind Interest Group (UWIG) determines the impact of an existing 280 MW of wind generation on Xcel Energy's 7200 MW system in Minnesota is $1.85/MWh, which includes intra-hour load-following reserves, intra-hour load-following energy and regulation reserves.

o Consultant Eric Hirst arrives at a cost of $1.47 to $2.27/MWh for the same three components when adding 1000 MW of wind to Bonneville Power Administration's 14,000 MW system.

o A first phase study completed in January for the New York Independent System Operator and New York State Energy Research and Development authority found that a 10% penetration of wind generation has little impact on New York's 33,000 MW system.

These studies have led UWIG to conclude that the higher the penetration of wind on a system, the higher the costs, that wind capacity does not need to be matched by reserves of an equal volume of dispatchable power as some system operators assert, and that the cost of reserves "is significantly less when the combined variations in load and wind plant output are considered, as opposed to considering the variations in wind plant output alone." UWIG also concludes that wind forecasting would mitigate some of these costs.

carrots and sticks

Without nationwide rules like SMD to govern balancing markets, there is a wide variety of carrot-and-stick financial incentives and penalties across the US for encouraging generators to do their best to match output with likely demand -- and for dealing with wind in particular. In its treatment of wind, the Electric Reliability Council of Texas (ERCOT), which oversees 85% of transmission in the state, simply excludes the existing 1305 MW of wind generators from its current balancing market. This is because the state's renewables portfolio standard requirement of up to 2000 MW of wind capacity by 2008 requires ERCOT to get the renewable generation on the system, which peaks at 60,000 MW regardless of the impacts.

ERCOT's Ken Donohoo says the biggest problem with integrating wind into its network is transmission bottlenecks, not wind's impact on reserve requirements. He suspects, however, that wind generation does have an impact on the system and says ERCOT is completing modelling that will look at how the grid will be able to integrate the full 2000 MW of wind into its system.

In the Pacific Northwest, BPA's Steve Enyeart says his agency has removed generation imbalance penalties that amounted to as much as $0.10/kWh for wind (“uåX˜äŠÊ˜·³Ç, September 2002) and now charges the same fees to all generators, regardless of fuel source. If a wind facility, or any other generator, delivers less than what it scheduled (within 2 MW or 10%), then it will pay 110% of the cost of market power to make up the difference. Conversely, if it delivers more power, it will receive 90% of market price for the extra electricity. Because BPA allows generators to set their schedules one day ahead, however, and to true up that schedule 20 minutes before the actual hour, it has a system that is "near real time," says AWEA's Caldwell. The closer to real time that the market for balancing supply and demand is allowed to operate, the more likely that wind generators will supply the system with what it has been told to expect.

Market solutions

With all Northwest wind projects now using wind forecasting, scheduling accuracy is on the rise (“uåX˜äŠÊ˜·³Ç, December 2003). While that adds the cost of a forecasting service and the cost of an automated scheduling system to the final cost of wind generation, it saves more on imbalance charges and cuts the amount of reserves necessary.

"If you don't know what you can provide 24 hours ahead, then you would need more reserve," says Michael Brower of TrueWind Solutions. The company is a US leader in wind energy forecasting, providing forecasting services for over 1500 MW of wind generation, or about 30% of the total US wind capacity. "To the extent you can reduce uncertainty through forecasting, then you can make scheduling more optimal," says Brower.

In California, the independent system operator requires a central wind forecasting service, which is provided by TrueWind. Even with that service, the California balancing market does not require an immediate truing up of a resource's balancing account. Instead, it allows wind projects to true up their accounts by averaging imbalances over a month, an approach which avoids the economic inequities of penalising wind for under-production one day and over-production the next. It is a good solution for an imperfect market, says Caldwell.

But he prefers real-time balancing on the spot market, as operated by PJM Interconnection (the Pennsylvania-New Jersey-Maryland Independent System Operator that touches on seven states). Here each generator, regardless of type, schedules its production 24 hours ahead, but it can true up 20 minutes ahead. Any imbalance that remains at the time of delivery is dealt with on the spot market at spot market prices. The PJM approach avoids the requirement of most systems for precise advance delivery schedules. "It sort of naturally allows wind into the system at the right cost," says Caldwell. The PJM comes closest in the US to a theoretically perfect solution to pricing wind imbalance, he adds.

Because the FERC's SMD proposal is so disliked by state governments and the US energy industry, it is unlikely it will be included in a 2004 United States energy bill, if there is a bill at all this year. The SMD would establish nationwide deregulation of electricity markets, something states are hesitant to embrace after the shocking failure of California's deregulation experiment. For its part, the energy industry dislikes having to join regional transmission organisations and perhaps giving up control over transmission assets.

Without any sign of the SMD coming into being, the wind industry is scrambling for other ways to ensure wind gets equal treatment by transmission organisations. AWEA, says Caldwell, is beefing up its regional efforts to get fair transmission tariffs for wind generators. Wind advocates are seeing some gains in the strategy. California, the PJM Interconnection and to some extent BPA area already taking steps to take account of wind in their market rules. As other regional transmission organisations, such as the Midwest ISO, get on their feet, the industry will be there on the ground floor to ensure the rules do not prevent wind from being able to compete in the market.

AWEA and other wind advocates are also going back to FERC to ask for a technical conference aimed at exposing integration issues for wind resources and then recommending a "best practices" document that would guide utilities under FERC's jurisdiction to deal fairly with wind power. Caldwell favours guidance in preference to force at this time. He warns that an attempt to draft mandatory rules for integrating wind generation would likely end up being so watered down it would not be worth the trouble.