New business models make energy storage attractive.
As policymakers and business leaders assign a higher priority to efficiency goals, companies are finding that the low-hanging fruit already has been picked in many areas. Reaching the next level might require some fundamental changes in regulatory policy and market models. Increasingly, regulators are focusing on the conflict between efficiency investments and volume-based ratemaking (see “Stimulating Efficiency”). But decoupling and negawatt rates won’t remove some of the most important barriers to system-wide efficiency improvements.
For example, distributed energy storage offers a host of efficiency benefits, including deferring or eliminating T&D investments; reducing the need for fuel-hogging spinning reserve; and increasing the reliability of variable generation sources, including wind and solar. But today, most utilities aren’t investing in grid-scale storage—not necessarily because of its capital cost, but because regulatory and market structures discourage it.
Transmission owners with assets managed by independent system operators (ISO) can’t put storage assets in their rate base, because those assets also provide generation services. Similarly, distribution utilities frequently can’t justify the cost of energy storage only on the basis of its distribution-system benefits. And generation companies struggle to make energy storage pay off, because the market hasn’t yet developed bilateral contracts that value the full range of energy storage services.
“That’s the key impediment to moving storage forward,” says Edward Cazalet, vice president with MegaWatt Storage Farms, and former member of the California Independent System Operator (CAISO) board of governors. “We need a mechanism to put some of the cost in the rate base, and to recover the remainder from the power market,” he says.
Even as end-use technologies have become more efficient, the U.S. electricity grid actually has remained relatively inefficient—largely because utilities must maintain enough reserve capacity to meet critical peak loads that might occur only a few hours a year. While base-load nuclear plants operate with capacity factors averaging greater than 90 percent, many gas-fired peaking plants operate at capacity factors in the 5 percent range.
The result is an inefficient system, both in terms of resource utilization and fuel consumption. And ironically, it might get worse as end users reduce their overall consumption but not necessarily their peak demand.
“When people talk about energy efficiency, usually they think in terms of site efficiency,” says Chris Hickman, senior vice president at energy service company Ice Energy. “Unfortunately things like CFLs [compact fluorescent light bulbs] and Energy Star appliances tend to reduce the base load, not necessarily the peak. The unintended consequence for the grid is that the thermally driven component—air conditioning—becomes a more significant problem for peak load.”
In principle, energy storage offers a tidy solution to this problem. Distributed storage in particular—in the form of deep-cycle batteries and thermal storage like Ice Energy’s technology—can reduce peak demands in a targeted, intelligent way with minimal impact on end users. By installing storage capacity downstream of grid constraints, storage can improve efficiency both in terms of T&D infrastructure and energy consumption.
To encourage development of energy-storage systems, the Stimulus Bill provides funding that may be used to develop battery technologies, and includes energy storage among the project types eligible for the bill’s $4.5 billion in matching grants. But notwithstanding this new source of money, energy- storage investments remain difficult under current regulatory structures.
A case in point: In 2006, FERC defined the 500- MW Lake Elsinore Advance Pumped Storage (LEAPS) project as an “advanced transmission technolog[y]” under the 2005 Energy Policy Act (EPAct), which directs the commission to encourage such projects. But when the project’s independent developer, Nevada Hydro, sought cost-based rate treatment and incentive tariffs for the project—as provided under EPAct—the California ISO refused, saying it “cannot support treating LEAPS differently than existing, similar generating units.” FERC upheld CAISO’s decision, effectively leaving storage in a state of limbo.
Large pumped-storage projects like LEAPS face a buzz-saw of siting, permitting and licensing requirements. Smaller, distributed storage projects are easier to site, but because their services are difficult to price in the market, they’re almost impossible to finance on a project basis.
“It’s hard to find a long-term bilateral contract for ancillary services,” Cazalet says. He explains that when a utility enters a long-term power-purchase agreement with a generator, the contract typically bundles ancillary services under a tolling contract or some similar arrangement. “Storage is a different animal,” he says. “It has a different mix of services, and there’s no established forward market for those services that would reveal the prices.” Without the ability to secure customers, battery projects can’t obtain debt financing.
Cazalet’s company is trying to overcome that problem by structuring projects and contracts in ways that quantify and apportion the costs and benefits. In some cases this might result in a hybrid partnership, in which a wires company owns the T&D assets while MegaWatt Storage Farms owns the power assets. In other cases, the company will enter bilateral contracts to provide services to a transmission company or load-serving entity. “You can structure a long-term PPA for the suite of services provided by a given storage technology,” Cazalet says. “A 15-year contract for 1,000 MW of distributed storage would look just like 1,000 MW of resources, except it could be dispatched in one second and ramped from -1 GW to +1 GW. No other machine on the grid can do that today.”
The trick is putting a dollar value on those capabilities, and structuring a contract that accurately reflects that value in a given situation. “We’re happy to do it in any way that makes sense for moving the industry forward,” Cazalet says.
One possible solution is being pursued by Ice Energy. The company’s technology faces some practical limitations—i.e., it only works during the cooling season, and its scale is limited by the amount of rooftop AC capacity in a given area. But Ice Energy’s business model suggests a potential path for expanding efficiency investments in the future.
Since 2003, the Windsor, Colo.-based company has been using ice-making machines as an energy storage device. Ice Energy connects refrigerant lines between its ice makers and the end user’s rooftop AC units. During off-peak hours, the ice makers go to work making ice, and during peak hours, the ice cools the refrigerant and reduces or eliminates AC compressor cycling. The system can shift up to six hours of cooling load each day.
Ice Energy has sold most of its equipment as a cost-saving device for end users. Utilities, including PG&E and Southern California Edison, have offered rebates for customers who purchase the equipment—as they do for other types of energy-saving gear. Now, however, Ice Energy is taking the next step, aggregating cooling capacity at numerous facilities and marketing it on a wholesale basis as fast-reacting, dispatchable power.
“We bid head-to-head in RFPs with gas peakers,” Hickman says. “But because we’re a distributed resource, we have the opportunity to affect overall system efficiency. We work with the jurisdictional utility or the city, and talk to their planning group to identify constraints on their system.”
Hickman says the company has standing agreements with about 30 nationwide commercial hosts, which allows it to identify local site prospects quickly—and to target its deployments most effectively for the utility. For example, if feeders are operating near their limits, Ice Energy will install its machines at customer facilities served by those feeders. This serves more than one goal for the utility—reducing system peak, and relieving stress on specific circuits.
Additionally, Hickman says Ice Energy’s approach saves energy overall. Because the technology shifts cooling load from hot, peak-demand hours to cooler, off-peak periods, Ice Energy claims its system is effectively “lossless”—i.e., it delivers 1 kWh worth of cooling for every 1 kWh of AC load that it offsets. Of course the ice makers themselves aren’t 100-percent efficient; Hickman says thermal losses account for about 12 or 13 percent of electricity used. But by running ice-machine compressors at night when ambient temperatures are lower, the round-trip process consumes about the same amount of electricity that the conventional AC units would use during the hotter daytime. “We use what nature gives us—a temperature differential from day to night,” Hickman says.
In the bargain, the system uses off-peak capacity that’s more energy efficient—by virtue of power plants operating with a lower heat rate, and cooler T&D systems transmitting power more efficiently. And of course the shifted load consumes less peak-priced electricity. Rate structures determine whether that’s money saved by the utility or the end-use customer.
“The capabilities of the product create a wide variety of options to help utilities manage the grid, from a very granular level to a large system level,” Hickman says. Such options include optimizing the potential of variable generation sources, such as wind turbines that reach their peak output at night, like they do in Texas. Another example is rooftop PV—which itself can reduce peaking-capacity requirements (see “Net-Zero Neighborhoods”), but only when the sun shines. As a result, several states have classified Ice Energy’s solution as a renewable resource.
The company sells this package of capabilities in two ways. First, a utility can purchase Ice Energy’s equipment outright, and contract with the company for operations and maintenance (O&M). Hickman says this approach works best for municipal utilities and cooperatives, whose low cost of capital makes buying assets more attractive than contracting. Second, a utility can enter a long-term contract as it would with an IPP.
“We’ve structured our deal so it includes all the things utilities are used to seeing in a PPA,” Hickman says. “You sign a 20-year agreement with availability and power factor criteria, insurance, O&M, maintenance reserve, etc. Our first cost is higher than it is for a gas-fired peaker, but because we have lower operating cost, we’re more cost-effective over the 20-year PPA.”
Its cost characteristics distinguish Ice Energy in the market; battery systems, by comparison, can’t compete against fossil-fired peaking plants. As a result, deploying most types of distributed storage technologies will require more complex financing structures, to accurately apportion the values and costs of ancillary services. But as utilities seek ways to manage load and optimize grid resources, ancillary services will gain value—making distributed storage projects an easier sell.
“Storage has gone mainstream in the last year and a half,” Hickman says. “People now recognize that storage is a critical aspect of what we need to manage the grid effectively.”