Peter Kelly-Detwiler: 617.875.6575 | Leighton Wolffe: 781.547.1193,

The clock struck midnight Sunday on emergency generation for demand response, and the boom fell today: May Day was supposed to have been a bad day for many of the curtailment response providers (CSPs) – and their customers with emergency generators – that provide demand response to competitive power grids across the United States. That’s because May 1st (well, in reality May 2nd since May 1st fell on a Sunday) was the anniversary of a D.C. Circuit Court of Appeals ruling that affects the ability of emergency generation to serve as a demand response resource. The week was quiet, with no news until a few hours ago, when participants reportedly received notification that the time is now up and emergency generation is off the table as a demand response resource.

Complicated background: This is super inside baseball, so let’s back up a bit and start at the beginning: Many years ago, the Federal Energy Regulatory Commission began to promote demand response – the ability of customers to curtail energy consumption during periods of stress on the power grid. A kilowatt-hour not consumed is essentially the same as a kilowatt-hour generated (actually better, since there are no losses on the transmission and distribution lines), so demand response in essence created a virtual power plant to support the electric grid in times of duress. Some of that curtailment involved customers actually turning things off, such as pumps and motors. But a significant amount involved customers utilizing emergency generators (machines that otherwise would only be used in the event of a blackout to keep the lights on), and a good portion of that emergency generator fleet was fired by diesel fuel.

One problem with this is that diesel generators are not the cleanest beasts in the zoo – they typically emit more particulates, nitrous oxide (a precursor to ozone – especially during the hottest days when a demand response event is more likely to be called), and sulfur dioxide than the average centralized power generation fleet. However, running a number of diesel generators was seen by many as preferable to the potential health and safety risk of incurring a blackout.

So, after much discussion, the U.S. Environmental Protection Agency granted a waiver to back-up diesel generators in 2013, to an original ruling which limited back-up generation to no more than 15 hours. The waiver allowed generators to run up to 100 hours annually for “emergency demand response” or in the event of a 5% drop in frequency or voltage (and they would get paid accordingly for offering capacity to the system). In part, the EPA indicated that it had loosened an initial ruling based on PJM’s (the Mid-Atlantic power pool) comments that they needed generators available for at least 60 hours to participate in emergency demand response programs (in fact, PJM indicated that this availability was only needed in the aggregate, rather than for each individual generator). This waiver was celebrated by the CSPs, as many of them had substantial percentages of their capacity being supplied by emergency generators.

Zero sum game – I win, you lose: However, when it comes to power pools, electricity markets, and economics, one frequently encounters a zero-sum game. If one entity wins (in this case, customers with generators and the CSPs) another party loses (in this case the large merchant generators). So generating companies and the Delaware Department of Natural Resources (concerned about local air quality issues) opposed this ruling. The Delaware DNR stated that the utilization of generators in emergency demand response programs would significantly worsen ozone pollution (especially the emissions blowing to the east into Delaware from other states).

The Independent Market Monitor, charged with overseeing and critiquing the effective functioning of the PJM power pool also weighed in, commenting that allowing emergency generator participation would distort the healthy functioning of the power markets. And Calpine – a generation company – charged that some end-use customers would specifically buy diesel fired back-up generation to participate in demand response, with the result that cleaner generation resources would be displaced. In fact, the generation resources are used very infrequently so to date the air quality issue has not been very significant (and the environmental impact perhaps less so than building a new centralized power plant).

The golden carriage has now become a pumpkin – what next: On May 1, 2015, the DC Federal Court of Appeals found unanimously in favor of the plaintiffs ruling that the EPA acted “arbitrarily and capriciously when it modified the National Emissions Standards and the Performance Standards to allow backup generators to operate without emissions controls for up to 100 hours per year as part of an emergency demand-response program.”

However, the Court also noted that if vacating the provisions of the 2013 rule-making caused administrative or other difficulties, it would allow the EPA “reasonable time to develop interim standards.” The EPA requested, and was granted, a stay of the court’s mandate for one year. However, it appears little, if anything was done in terms of developing a mitigation plan, despite the fact that this generation was meant to address grid emergencies. That year is now over as of May 1, the clock has struck midnight and the carriage has now turned into a pumpkin. Presumably, these diesel engines will now be declared ineligible to participate in demand response programs. In fact, according to sources, the Court mailed that information to parties today (May 4) and an opportunity to address the situation was lost.

“It is unfortunate that the EPA did not use this past year to implement clear long-term standards for these engines.”

said Frank Lacey, president of Electric Advisors Consulting, a Pennsylvania-based energy markets consulting firm.

“By addressing the Court’s concerns, they could have cemented the benefits of low-cost energy, reduced air emissions and a more reliable grid.”

An April 15th EPA memo on the topic issued on April 15 from the EPA Office of Air Quality and Standards to Regional Air Enforcement Managers and Regional Air Directors states that regulatory requirements will depend on every engine’s horsepower and age and the criteria and requirements information can be found on the EPA website. Some engines will be subject to performance testing within 180 days of the mandate (by October 29, 2016). The memo further states that owners/operators of emergency generators will have to submit annual reports detailing the hours the engines are used.

It is unclear just how many megawatts of demand response are provided by back-up emergency generation across the country, but it has not been a small number in recent years. In PJM, the largest power pool in the country, generators represented 23% of the registered demand response in 2015, with 81% of that being diesel generation. In theory, this could potentially affect almost 19% of the DR resources available, totaling over 1,700 megawatts of emergency-only generation. In fact, many of the major demand response providers saw the writing on the wall and have reduced their exposure to this issue over the past year.

Raj Chudgar, CEO of Viridity Energy, a software company that assists customers in their interaction with the grid, including demand response, noted that the air quality issue has been minimal to date, given the infrequent deployment of emergency generators to backstop the grid. Chudgar commented that the decision was inherently inefficient from an economic perspective:

“It is clear that a significant part of the US electrical infrastructure can be backed up with emergency backup generators for either local use or grid support. It is striking that the same emergency backup generators cannot provide this service to assist with grid stability as we move to a more distributed model and the grid, which has in general become older and less stable, cannot use these resources. Why would we not explore using this steel in the ground to better manage our grid?”

Perverse Outcomes? Economists have a phrase –perverse outcomes – for results that are unintended and contrary to original goals. In the case of the Court ruling, there may indeed be one – or even two unfortunate results in PJM. A study undertaken by the consulting firm Navigant looking at this issue for the Advanced Energy Management Alliance indicated that removal of 1,750 megawatts of emergency generation in PJM will actually increase emissions while driving electricity prices up.

The logic behind this conclusion, based on Navigant’s complex model of generating resources, is that retirement of low-cost, part-time emergency generation resources will increase prices by approximately $35 per megawatt/day. That in turn will improve the economics of coal plants that night otherwise have retired. And since those polluting plants can now remain in the game, they may displace cleaner generation in some cases. Admittedly sophisticated models and complex stuff, but there is clearly the potential for both higher prices and increased emissions levels.

Impact on Curtailment Service Providers: To the extent that Curtailment Services Providers have not realigned their capacity obligations (the amount of capacity they are on the hook to deliver), there could be a mad scramble to find replacement capacity resources. When contacted, PJM indicated that it has notified the CSPs regarding the EPA’s recent memo. A spokesperson commented in an email that “PJM has also reminded CSPs that behind the meter generation may only be used if it can operate when dispatched by PJM and comply with all local, state and federal laws (includes environmental permits). PJM posted the following document to help with the communication process

It is not good news for demand response, curtailment service providers, and their customers with emergency generation, and it may be economically inefficient and result in worse emissions than otherwise would be the case. But merchant generators are breathing easy…