If California wind operators could have more accurately predicted their production during the state's energy crisis four years ago, they would have made a mint. Instead, some lost small fortunes when their production did not meet their forecasts. Recognising that a repeat of that chain of events represents an unacceptable level of risk for further wind investment, California has taken steps to prevent it from happening again -- and possibly set the stage for a nationwide market model that could facilitate far greater sales of wind power.
Under the rules of the California Independent System Operator (ISO) in 2001-2002, generators were required to forecast their production 24 hours in advance in ten-minute increments. If a generator did not meet its forecast, it paid to supply the energy it failed to deliver at the spot price of the moment. Prices normally fluctuate somewhat during a day, but during the energy crisis they fluctuated so wildly that there were "tens of thousands of dollars" of payments that wind operators faced because their forecasts were off, says Nancy Rader of the California Wind Energy Association (CalWEA). "It really hurt. That's why we ran screaming to the ISO saying something has to be done."
The California ISO formed a work group of representatives from the ISO, CalWEA, developers, operators and others. From their work was born the California ISO's Participating Intermittent Resource Program (PIRP).
Rader says PIRP's primary benefit is removing penalties for not meeting a scheduled forecast. "At a minimum, that's what the PIRP program is about," she says. "If you penalise us for not meeting the schedule, you are penalizing us for being wind. That's what the PIRP program recognizes." At the same time, wind operators are not exempt from the scheduling requirements on which California's electricity market functions. "This is really a great program," says Rader. "It works for them, and it works for us."
The PIRP has two primary components. First, it provides a way to compress the forecasting window using automated equipment. Second, it employs a monthly -- rather than ten-minute -- balancing of over and under production. The result is to reduce costs for inaccurate forecasting.
On the first point, the California ISO determined that if it could get wind production forecasts that were 85% accurate or more within a two-hour ahead timeframe "we could make the adjustments to the system on an economical basis," says Randy Abernathy, the ISO's vice-president of marketing. PIRP allows wind generators to provide forecasts two hours ahead, but with a condition. The operator has to install monitoring and reporting equipment that automatically provides that forecast to the ISO. "For us to net you over the month you have to give up the right to tell us what you'll produce," says Abernathy, That prevents a generator "gaming" the program by systematically over or under producing a forecast, he adds.
The equipment also enables the system operator to build a database of energy production for each wind project, allowing it to better predict wind production state-wide. "The forecast is based on actual production," says Gregg Fishman, also with the ISO. "As we develop more data, we can tweak that and work it into the profile." Then, for instance, if a wind farm historically drops 10% at a certain hour of the day, that drop can be incorporated into the forecast. Instead of 100 MW forecast, the automated forecast would be for 90 MW. "That's the kind of fine tuning we're engaged in," says Fishman.
The database of wind conditions is maintained by a third party, AWS Truewind. Generators pay a set fee of $0.10/kWh for the service, which the ISO says helps provide cost certainty for wind generators at a level that should be lower or at least comparable to imbalance costs under the old system. "So far the program's been surprisingly accurate," says Abernathy, "Initial equipment problems that biased the forecasts slightly in favour of wind generators appear to have been corrected."
When it comes to balancing charges, PIRP allows participating wind operators to calculate on a monthly basis the amount of kilowatt hours they were under or over their forecast amount. A generator can be a little high or low in the amount of wind energy produced in any hour, but the highs and lows are balanced against each other once each month. Typically, says Fishman, those highs and lows balance out to almost zero over a month's time.
CalWEA's Rader says that imbalances are not as much of an issue when the market is on a relatively even keel. "The problem is that it might get out of whack, and we can't tolerate that. The real value of the PIRP program is that it eliminates the risk that those charges are going to get out of whack," she says.
System operators in other regions of the US are starting to look seriously at the California ISO's solution to the scheduling challenge. Recently, the ISO presented PIRP to other independent system operators at a wind scheduling workshop organised by PJM (for Pennsylvania, Jersey and Maryland), a regional transmission operator (RTO) serving all or parts of 12 states and the District of Columbia in the eastern US. It is the largest centrally dispatched electricity grid in the world, but has only 260 MW of wind as yet.
PJM's Ray Dotter says the RTO organised the workshop because it is looking ahead to the time when more wind is on its grid. "Most of our region is not the right environment for wind," says Dotter. "With technological improvements, we're seeing more interest and, especially as states move to renewables standards, it creates a greater impetus to do that." Dotter notes that like California, PJM offers wind operators a chance to sell into a large wholesale market. "Another of the reasons we're seeing more interest in our area from alternative technologies is because the spot market gives you a ready market. You're not tied to one buyer [and] there are lots of buyers out there."
Developers of projects totalling 3900 MW of wind have expressed varying degrees of interest in the area PJM oversees, Dotter says. "We are neutral in favouring a type of generation," says Dotter. "But we want to ensure a level playing field. Through this workshop, we're trying to make sure that wind has a fair opportunity to participate. If this is a technology that makes sense to us and helps facilitate wind generation, that's a good thing, because more wind is going to be needed. We need rules in place that treat everyone fairly and don't discourage resources from being built."
Praise from users
For wind generators using California's PIRP, the program's rules draw praise. "From my perspective, it is the most rationale market design in the United States," says Mark Smith, of FPL Energy, one of the largest wind developers in the US. "It doesn't penalise us over things over which we have no control. Equally important, it provides the California ISO with state-of-the-art, best available information on what our intermittent generators are going to do over the next hours or day, so it's a win-win for both sides."
Financially, Smith says, the monthly netting of deviations is a significant improvement over the highly volatile ten-minute settlement. "They addressed both operational concerns and financial viability of projects in one swift move with this program," Smith says. "And they've done a lot from my perspective to support intermittent resources."
Although approved by the Federal Energy Regulatory Commission (FERC) in early 2001, the program was not implemented by the California ISO until June 2004, in part, Abernathy says, because new projects are more likely to participate and wind development in the state had stalled while market players waited to see whether wind's federal production tax credit (PTC) would be extended.
Fishman adds that with the energy crisis still fresh in everyone's minds, few felt inclined to rush out and embrace a new program. Furthermore, many existing wind generators cannot participate because they already have long term contracts with utilities for their output. Utilities typically schedule such enormous amounts of energy, that the relatively small amount of wind they are also scheduling is incidental -- most deviations get absorbed in the bigger pool. They also have alternative resources to bring to bear when wind is above or below forecasts.
"As those contracts run out, some may want to come into the program," says Fishman. Currently, there are ten wind farms with a combined capacity of 400 MW participating in PIRP, with another 600 MW expected to join the program this year. "PIRP solves the control point plus the financial piece," says Abernathy. "And it provides clear, transparent and reliable information." The California ISO has a 40,000 MW peak load and 2000 MW of wind. "To have that many megawatts swinging back and forth could be a problem if you couldn't see it," claims Abernathy.
Regulators like it
Federal regulators have also taken a shine to PIRP. When the California ISO filed its proposal with FERC to implement the program, says Abernathy, FERC not only approved PIRP, the agency also used it as a model for developing recommendations for other systems seeking to foster and accommodate wind development.
Last month FERC officially proposed to begin the process of identifying reforms to account for imbalances of intermittent resources such as wind power with the aim of removing potential barriers to their development. PIRP appears to be a likely model, at least in part. In an earlier report on the program, the federal agency noted that PIRP's concept of "exempting wind from hourly imbalance penalties and substituting monthly netting of imbalances in return for centralised wind delivery forecasting is an example of the type of tariff reforms that could facilitate wind development."
"California's PIRP program is a positive step in the right direction," says FERC's Bryan Lee. "It's one example of the type of tariff reforms that could facilitate wind development." Lee adds, however, that although the approach seems to be working well in California, it might not work in regions of the US where transmission resources are limited.
Perhaps the biggest problem in the eyes of CalWEA is in the slow pace of acceptance. "Most California wind projects are scheduled by the utilities because their contracts require that," says Rader. "Utilities have not opted in to this program. That's a problem because the ISO needs those projects participating in order to have better data to understand where the wind is going -- whether it's coming up or going down. They can't get a good sense with the limited wind enrolled."