What do the months of March 1979, January 2001, and February 2008 almost have in common? In March 1979, a loss-of-coolant accident at the Three Mile Island nuclear station brought the expanding nuclear power industry in the United States to a dead stop. Only a handful of reactors were completed post-TMI, with the majority of projects abandoned or converted to an alternative technology.
In January 2001, the recently deregulated power market in California suffered a complete meltdown requiring the state to begin buying needed power. That in turn sent paralysis through other progressing Regional Transmission Organizations. Quickly discarded were names such as SETRANS, Desert Star, and GridSouth, as well as other would-be centralized dispatch markets.
February 2008 could have been a similarly defining moment for the U.S. windpower industry. However, in this case ERCOT averted a crisis, preventing that date from becoming a sort of Waterloo for windpower.
At just after 6:30 p.m. on February 26, the Texas power market experienced a sudden frequency drop, initiating second-stage emergency procedures. Operating reserve resources known as LAARs (Load Acting As a Resource) were utilized to rapidly add about 1,100 MW of windpower capacity to the strained system. ERCOT, the Texas system operator, cited a steep drop in windpower output as a contributing cause. ERCOT restored stability to the system over the course of a three-hour period, and in mid-March issued a preliminary report on the event (see, What Happened in ERCOT). But a nagging question remains: If this event had caused system transients, significant power outages and equipment failure, could it have become a “defining moment” for the windpower industry leading to a significant shift in public and political attitudes toward wind expansion?
To suggest the February ERCOT incident could have been windpower’s Three Mile Island is admittedly an over-dramatization. But current projections of wind expansion in West Texas show that windpower capacity in the region likely will double over the next two years, from the current level of roughly 5,000 MW to upwards of 10,000 MW by the summer of 2010. What would have been the consequences of losing twice as much output on February 27th when the wind died down more than expected in West Texas?
Current state-by-state Renewable Portfolio Standard targets for installed wind capacity pose a major challenge. Many states have mandates for 20 to 25 percent of demand to be met by windpower over the next 15 to 20 years, up from only a fraction of a percent today. So the wind industry might soon face a defining moment that will, at a minimum, force the utility industry to reconsider the pace of windpower development.
Many windpower promoters contend that such an event can be avoided by improved forecasting, better operating procedures, aggressive demand-management programs, increased reserve requirements, and other techniques. These changes are being examined in ongoing wind-integration studies, and most show potential for the ability to buffer wind volatility. Meeting current wind-penetration targets, however, ultimately might require the development of an economically viable means of storing windpower.
The advantages of implementing storage technology for windpower are huge. Storage virtually eliminates the need to curtail wind output, guaranteeing high utilization of potential wind energy. The stored power can be dispatched during peak periods, reducing the need to run older, expensive units that are often high polluting. Storage capacity reduces the need for investment to expand the transmission grid, avoids the cost of increasing operating reserves as a buffer for wind variability, and actually might reduce the cost of carrying reserves.
Using windpower as a firm peaking resource allows a much greater percentage of installed wind capacity to be credited toward system-reserve margins, reducing the need to build new peaking resources. Being able to store and dispatch windpower renders wind integration into a non-issue, and removes almost all operational obstacles confronting the wind industry.
So if energy storage makes so much sense, why isn’t it happening? The answer is clear: Up until now it just hasn’t made economic sense to install. When there is a relatively small amount of wind being produced, the regional power system can absorb it quite easily even at full utilization. And storing power is expensive. The primary means involve construction of pumped hydro or compressed air storage plants, or via use of battery-storage technology.
Pumped hydro has the advantage of being a fairly well-known technology, where cheap off-peak power is used to pump water to an elevated reservoir, then released to generate power during peak demand hours. There are currently 31 operating pumped-storage plants in the continental United States, supplying roughly 2.5 percent of total electrical demand. Pumped storage units, however, require a large up-front capital investment and have specific siting requirements than can be problematic. West Texas, for example, is notably short on water supplies, which is obviously a key requirement for locating a pumped-storage plant.
The same pros and cons hold for compressed-air storage, which requires well-defined natural geological structures, usually salt mines. This technology has been utilized heavily in Europe in concert with wind generation, but has much less natural application in the Great Plains region of the United States. So while these traditional, large-scale technologies may work well in certain areas, it is unlikely they can be utilized at the magnitude needed to effectively deal with the anticipated large-scale wind energy expansion in the Midwest.
These obstacles leave battery storage as the perhaps the most feasible option for storing wind generation for peak demand use in many regions. In fact, there have been two recent announcements of initial tests of megawatt-scale batteries being installed at wind farms in Minnesota and in Ireland. Battery arrays can be located at each individual wind tower and even utilize the electronics of the wind equipment to provide grid connectivity, thus lowering the storage cost. This also makes the battery-storage option extremely scalable, because power can be diverted from the grid into the battery as needed to reduce off-peak maximum wind output to avoid curtailments, even if only a portion of the wind towers have batteries installed. If the published operating characteristics of large-scale batteries are accurate, they likely could qualify to provide spinning-reserve support during periods when the batteries are charged, potentially adding new revenue from battery installation. Wind-farm owners negotiating power-purchase agreements could also offer a firm capacity deliverable to enhance contract payment terms.
There are, however, some negative aspects of megawatt-scale battery-storage technology. First, the capital cost to buy and install batteries, generally quoted at around $200/kWh, still is quite expensive. Current battery technology also has a 20 to 30 percent loss of energy during the charge and discharge cycle, meaning for every 100 MW of wind output stored in a battery, only 70 to 80 MW reaches the grid for consumption. Because the development of large, megawatt-scale batteries constitutes new technology, there are questions about long-term durability and maintenance requirements. There also is uncertainty on how battery storage will affect renewable energy credits under different state-run RPS programs, as well as a lack of clarity on the status of battery storage as a qualifying technology for energy tax credits.
Finally, not all windpower advocates consider batteries a “green” resource, particularly since most batteries use large vats of chemical electrolytes.
Many of the challenges facing battery storage technology will diminish over time. As in other industries, better economics derived from large-scale production and technology advancements inevitably will lead to better, cheaper battery products. In addition, there are currently some compelling market issues that make battery storage an increasingly attractive alternative.
For example, in ERCOT, recent wind expansion has begun to create significant zonal energy price differences across the system due to transmission congestion in exporting windpower from West Texas. ERCOT congestion charges from the West Zone to the North Zone, the corridor for moving windpower from West Texas into the Dallas-Fort Worth area, have skyrocketed in the last few months as wind generation has increased (see Figure 1).
In 2009, ERCOT will move to a nodal market with locational marginal prices calculated at individual buses across the ERCOT system. One element of the market is the auctioning of congestion revenue rights (CRRs), where participants can buy the financial rights to congestion between their generation and load to hedge themselves against anticipated congestion costs.
In studies performed by Ventyx Energy Advisors, projected costs for purchasing the congestion rights for moving windpower from West Texas wind sites to Dallas load have been as high as $40 million for 2009. And congestion costs likely will continue to increase until at least 2012 when the first significant new transmission lines are slated for completion to export power from West Texas.
High congestion costs may provide a new role for battery storage in the ERCOT market, as a way to hedge against congestion. With battery storage in place, a wind-farm operator can choose to charge the battery during highly congested hours rather than pay the steep costs of exporting power to serve load. This congestion-cost savings may go a long way toward offsetting the high installed cost of battery-storage technology in the upcoming ERCOT nodal market. And if the economics of the ERCOT market can make sense for battery-storage manufacturers, the technology might get a needed technological boost toward improvements and scaled production economies, leading to applications in other less constrained markets.
West Texas is simply a precursor to what will happen in other wind-development regions if the current rate of wind-capacity growth continues. In the absence of energy storage, markets will continue to struggle to integrate significant new levels of wind generation.
History is rife with examples of nature proving more unpredictable than mankind can imagine or prepare for. Without a method for reducing the volatility of wind output, the windpower industry sooner or later will be faced with an event that will become what Feb. 26, 2008 was not—the defining moment that could stop the industry dead in its tracks. Just one such event could lead to wind moratoriums, investigations, and possibly a shift in public and political will that may take years to reverse. And that would be a shame, given the significant promise windpower holds for contributing to a safer, cleaner future.