With the proper structure, demand response can shave peak, lower wholesale prices, decrease investments in new capacity and transmission, and reduce CO2 emissions. To accomplish these goals, market rules have been established to encourage participation of demand response providers in energy and capacity markets.
But are these market rules promoting the development of a real demand response market? Or are they enabling artificial bids that prevent real reductions from being competitive while undermining grid reliability?
Market rules that treat demand-side resources the same as supply-side resources are a relatively new development, and not necessarily what the electricity market was designed to accommodate. It is very clear that including demand-side resources, like efficiency and demand response, in capacity markets has clear advantages. What is less clear is how to ensure that the rules governing the ways these resources compete can prevent market manipulation.
Market manipulation hurts real demand response because all bids in the market reflect the marginal cost of supplying that resource. Real demand response has real costs: technology, sales, marketing, operations, and overhead. However, if an entity can make bids in the market that are lower cost by manipulating the market rules, this squeezes out real demand response that has to bid its true cost.
The most interesting and heated discussion about market manipulation revolves around financial arbitrage in capacity auctions.
Capacity auctions are held in PJM every three years to ensure enough resources will be available to meet demand in advance to secure the long-term reliability of the system. These are called “base residual auctions.” Every year up to the delivery year, there are also auctions called “incremental auctions” that deal with any balancing that may occur.
For example, if a power plant is bid in to the base auction, but then cannot deliver due to mechanical failure or environmental regulation, they have to purchase replacement capacity in the incremental auction so that someone else can provide the generation they promised. If they don’t, they get hit with a heavy penalty, since participating in the auction requires posting a lot of money (basically, a bond) that says they will deliver.
Since demand response can bid into the capacity market as if it were traditional supply, and because it is almost always lower cost, it pushes some of the more expensive supply out of the market. Generators do not like this, and have pointed out that demand response doesn’t have to meet all the requirements faced by traditional suppliers, who have to prove that they have capacity and promise they will deliver when called upon. Demand response has other options, including delivering 10 times a year for 6 hours at a time.
The problem, said John Shelk, president of the Electric Power Supply Association, is that “demand response providers are essentially bidding in their business plan.” Demand response does not necessarily need customers signed up to participate, just a plan for acquiring customers.
Shelk pointed out that, as a result, a lot of replacement capacity is purchased by demand response providers in the incremental auctions, a position supported by the recent market monitor report, which notes that after adjusting for some replacement due to rule changes, "the remaining replacement capacity megawatts constituted about 27 percent of cleared capacity for the 2012/2013 Delivery Year."
This is the heart of the arbitrage issue. Since the prices in the incremental auctions are much lower than the prices in the base auction, demand response providers can bid in capacity at a higher initial price and then purchase replacement capacity for cheaper later -- and pocket the difference. While it is clear why this bothers generators who lose out on that revenue, it is important to see that this could also hurt the development of a real demand response market.
If these bids are being made with no intention of delivery, the bid price would not reflect the costs to operate -- meaning that the real demand response that requires technological changes and demand reductions would be priced out of the market.
Gregg Dixon, Senior Vice President of Marketing and Sales at EnerNOC, thinks this fear is overblown.
“The idea that demand response providers have no intention on delivering on their bids is, on its face, ridiculous,” Dixon said. “When we make a commitment, our intent is to deliver physically. If we arbitrage, we lose out on our long-term value.”
Dixon pointed out that the entire demand response provider business is premised upon customer acquisition that creates real savings for those customers.
That view holds weight with independent analysts, including the Market Monitor, which noted: "The IMM has identified no evidence that any Curtailment Service Providers (CSPs) are purely financial entities that sell DR positions in capacity auctions with no intention of providing a physical resource and fully buy out of those positions every year.”
Dixon also pointed out that the beneficiaries of these incremental auctions are usually supply-side generators, because when replacement capacity becomes available, it is purchased by power plants that would not otherwise be getting capacity payments. As for the reliability question, Dixon explained that demand response is fundamentally more reliable than traditional supply because it aggregates customers. In other words, when demand response is short, it is short 10 out of 100 megawatts if some customers do not curtail. When a 100-megawatt power plant goes down, you lose 100 megawatts.
The truth is probably somewhere in the middle. The data shows that demand response has purchased a disproportionate amount of replacement capacity compared to generators. The cost differential in the auctions means that demand response providers are likely making some money off these plays.
If true, this would to some extent suppress prices and inhibit demand response from really maturing. However, it seems unlikely that this is part of a broad plan to game the system, since demand response providers need to save customers money in order to acquire new ones.
What can be done to ensure that these market rules promote real, physical deliveries on demand response commitments?
This issue is being addressed in PJM now, but there is actually a clear path forward that all sides seem to agree on: better measurement and verification techniques.
EnerNOC has long supported this idea and believes that this process should come in the form of technology standards and better visualization of demand reductions across markets (which would fall on PJM to implement). The Market Monitor report supports these changes as well, stating that “a rule requiring that DR providers demonstrate that they are actually in the business of providing DR resources would be an appropriate part of any package of rule changes related to this issue."
There is a delicate balance to be struck. On the one hand, if the market rules are too strict, there will be little incentive for companies to make the investments in smart technology and demand aggregation to be real players. Conversely, if the rules are too lax, real demand response players stand to lose out on a big slice of the pie.
Adam James is a Research Assistant for Energy Policy at the Center for American Progress and the Executive Director of the Clean Energy Leadership Institute. You can email him at firstname.lastname@example.org and follow him on Twitter @adam_s_james.