0
by Jeff St. John
October 16, 2018

Everyone agrees that Federal Energy Regulatory Commission Order 841, and its commandment to create participation models for energy storage across the country, is going to be a big deal.

Brattle Group estimates there are about 7,000 megawatts, or more than 20,000 megawatt-hours, of cost-competitive energy storage potential to be unlocked by Order 841’s implementation across the capacity, energy and ancillary services markets run by the country’s independent system operators (ISOs) and regional transmission organizations (RTOs). 

But behind these grand-scale projections, there are dozens of key details being worked out that will affect how much of that energy storage potential is unlocked, at what pace, and at what cost. Order 841 implementation is now in the hands of the ISOs and RTOs. And while many of its key provisions are set in stone, many more are left up to the grid operators and various stakeholders to work out in ways that best suit their particular needs.  

These are the details that are keeping the energy storage industry, as well as state regulators and public and private utilities, finely tuned for each new Order 841 implementation plan, known as a “straw proposal,” to come from the nation's ISOs and RTOs. This month so far has seen a final straw proposal from Midcontinent Independent System Operator put before stakeholders, as well as the latest iteration of straw proposal from PJM Interconnection. As the country’s biggest grid operator and host to hundreds of megawatts of batteries playing into its frequency regulation market, PJM is a focus point for the industry. 

These proposals are coming at a relatively rapid pace for a process that took nearly two years to move from a proposal to a final rule. ISOs and RTOs have a deadline of December 3 to file the tariff changes needed to “establish a participation model for electric storage resources that ensures eligibility to participate in its market in a way that recognizes the unique physical and operational characteristics of such resources.” By December 2019, those rules are set to go into effect. 

Order 841 lays out some core concepts that each ISO or RTO must adhere to, as this February briefing (PDF) from the Energy Storage Association, Customized Energy Solutions and Brattle Group lays out. They include: 

  • Storage is eligible to provide all capacity, energy and ancillary services that the resource is technically capable of providing
  • RTOs/ISOs must account for the physical and operational characteristics of electric storage resources through bidding parameters or other means
  • RTO/ISO minimum size requirements do not exceed 100 kilowatts (an opening for distributed storage, a topic for another day)
  • Storage can be dispatched and can set the wholesale market clearing price as both a wholesale seller and wholesale buyer
  • Storage will pay wholesale locational marginal price for charging energy

But for Jason Burwen, vice president of policy for the Energy Storage Association, each of these bullet points is just the start of a series of conversations, or debates, about how best to implement FERC’s broader vision. “Remember, there are 76 directives in Order 841,” he said in an interview last week. And while ISOs and RTOs are engaged in a diligent effort to make progress on each of these fronts, “from month to month, what’s in these presentations has changed,” he said. 

What makes Order 841 implementation more complicated, though necessarily so, is that FERC has given each ISO and RTO a good deal of latitude in how to use their existing market constructs and operational capabilities to inform with their own solutions, he said. “The order obviously directs all the RTOs to do things, while recognizing that different markets have different approaches in general.” 

Some ISOs have more catching up to do than others to meet Order 841’s new requirements, he added. But even those that have made a good deal of progress in integrating energy storage into some markets and functions, as PJM has done with frequency regulation services, or the California Independent System Operator is doing with distributed energy storage integration, will need to adapt to other markets that haven’t seen energy storage playing a role as of yet. 

Building the state-of-charge parameters for a unique resource 

The first thing to note about the performance of an energy storage asset is the simplest, yet most profound — its ability to both charge, or take grid power, and discharge, or contribute it. That’s unlike any other grid resource, except pumped hydroelectric power, and it's part of what FERC Order 841 defines as energy storage. That means it excludes thermal energy storage, which uses grid power to avoid using electricity later in the day, because it doesn’t actually push electrons back onto the grid.

While ISOs and RTOs have traditionally treated its resources as either a generator or a load, Order 841 requires them to come up with ways for energy storage to play either in one of them, or in both simultaneously. That last part is the trickiest, since it’s something most ISOs and RTOs haven’t done before. But it’s also the area where batteries, with their fast-reacting up-and-down flexibility, are expected to yield some of the highest values, compared to traditional gas-fired power plants or other traditional alternatives. 

But with this flexibility comes a key limitation. “Energy storage has [the metric] 'state of charge,'” Burwen said. “You’re always somewhere between zero and full, and that changes what you choose to do.”

Again, this is not really the same as any other resource in wholesale energy markets. Generators largely run on fuel that can support all-the-time operations, as with natural gas pipelines, coal piles or nuclear rods, or comes intermittently from the sun and wind. Demand response and other load-side resources have their own known limitations, but only in one direction. 

Batteries can make a huge difference in those resources' value, both to the grid and to their investor-owners, Burwen said. In the most obvious terms, asking a lithium-ion battery to discharge itself completely, multiple times a day, will shorten its effective lifespan to the point where it’s not worth the initial investment. But in subtler terms, an energy storage project’s cost-effectiveness is determined through carefully managing the state of charge against a panoply of market opportunities — a set of complexities largely lacking from most other grid resources today. 

In Order 841, this is expressed in a series of state-of-charge parameters that FERC has asked each ISO and RTO to incorporate into their tariffs. These include minimums and maximums for a battery’s state of charge, how fast and how often it charges and discharges, and its run time and ramp rate — an important set of rules for grid operators and energy storage providers alike.

“How is energy storage going to participate on a regular basis in energy markets? That’s where these parameters are going to show up; that’s where these rules are being developed,” said Burwen. “From the energy storage industry’s perspective, we’re looking for a compliance approach that enables the most flexible use of storage for system benefit, in a manner that takes account of the physical characteristics and capabilities of energy storage resources.” 

So far, ISOs and RTOs have started out with the core concept of allowing energy storage to operate in several modes, he said. In the case of PJM, it’s proposing continuous, charge and discharge modes for energy storage assets, a move intended to allow for pumped hydro, which does face limits in switching between “charge” and “discharge” modes, as well as for batteries that don’t face those limits, he said. 

But PJM hasn’t yet laid out its ideas for incorporating the various state-of-charge parameters for batteries and other flexible resources, said Burwen, adding, “It's not clear yet whether they’re actually ready to implement state-of-charge parameters” based on its proposal.

“That’s one of the questions here. PJM’s intent is that storage managers can manage their state of charge using these different modes and setting different limits throughout the day," he said. The question is, will their ultimate compliance approach include parameters set specifically by PJM, or simply by the choices of the storage operators themselves?

The state of charge is important because it will help avoid infeasible dispatch schedules, as well as avoid impacting the O&M costs for these resources. There are limits to harmful dispatches of this kind. “Order 841 said storage asset managers will have the right to manage their own state of charge,” Burwen said. 

But “if PJM is not going to have these parameters for day ahead and real-time scheduling, then most folks will have to self-schedule” he said. That, in turn, could expose them to the difficult-to-foresee risk of being ordered to do something harmful for the battery or face some kind of financial penalty. 

Burwen was quick to note that PJM hasn’t explicitly said it won’t be building these parameters into its software for optimizing the bidding and dispatch of resources across its markets. “They have not indicated that they are going to not do it. But they also haven’t said, 'Here’s how we’re going to use them,'” he said.

“That’s not just specific to PJM," said Burwen. "Each RTO is incrementally figuring this out as we go.” 

Finding the right durations for capacity markets

One critical operating parameter for batteries and other energy storage assets is its effective, or optimal, duration. Put another way: How many minutes or hours of energy can it provide before it needs to recharge again? For lithium-ion batteries, which make up the vast majority of grid-scale batteries today, this limit tends to be about 2 to 4 hours — go any longer than that and you risk degrading the battery’s performance and lifespan. 

Many energy markets, like PJM’s frequency regulation market, can make effective use of resources with less than an hour of duration, making them a natural fit for lithium-ion batteries. But the capacity markets run by many grid operators do come with multi-hour durations — and finding the right way to translate these rules in ways that include energy storage is another key point of contention over Order 841 implementation. 

PJM’s latest straw proposal, for example, calls for maintaining a minimum 10-hour duration in the capacity market. As it lays out in its example case, that means that a 10-megawatt-hour energy storage resource could bid 1 megawatt of available capacity. This concept is called derating the resource, or changing the relationship between its rated capacity (i.e., its ability to pump out 10 megawatts in 1 hour) and its effective capacity to provide the resource as it’s defined by rules on things like minimum duration. 

The move to establish minimum durations for capacity is, again, not specific to PJM, Burwen said. There are other RTOs that are specifying this. The Midcontinent ISO, for instance, has a defined duration for capacity participation, and New York ISO is "now having a conversation about what that should be.”

But in the Energy Storage Association's view, PJM’s decision to propose a 10-hour duration for batteries in order to play in its capacity market (the country’s largest, although one that’s currently in the midst of a FERC-ordered reorganization) is problematic for a couple of reasons, he said. 

“There is disagreement on this point,” he said. ESA has previously pointed to the fact that peak-hour periods in PJM are no longer than four consecutive hours, which would appear to provide a good guide for what it might require from energy storage participants, he said. And PJM’s current capacity market rules allow providers of intermittent resources like solar and wind power, or demand response or energy efficiency, to submit offers based on their expected output during peak periods of three hours in late summer afternoons, or two separate two-hour periods in winter mornings and evenings.

“Order 841 says storage should be able to derate its capacity to meet any duration market requirement,” he said. But lengthening the required duration of a capacity resource beyond whatever number of hours it’s actually needed risks diluting the value of the resource, according to Burwen. 

What’s more, the process of derating an energy storage asset opens up complications in market rules known as must-offer obligations, or MOOs, he said. Simply put, a MOO is a requirement for a power plant or other capacity resource to show up for as long as it has promised to, or face “physical withholding penalties” for disrupting the effective operation of the market. And many of the power plants that serve capacity today are committed to 24-hour MOOs, meaning they have to be available to run all the time. 

Order 841 specifies that batteries shouldn’t be exposed to physical withholding penalties based on a poorly designed implementation of MOO rules. “There’s generally a recognition that storage resources are energy-limited, so they won’t have a 24-hour must-offer obligation — no ISO is requesting that, so that’s good,” he said. But “if battery storage is derating to meet that duration requirement, therefore not operating at its full capacity, we need to take that into account,” he said. “If we take market compliance rules that are made for a generator and apply them to energy storage, that’s going to be a problem.” 

FERC Order 841: "The start of a much bigger conversation"

These two issues are the most clearly defined of the dozens now being hashed out in straw proposals from each grid operator. Other hot-button topics include transmission charges, or more specifically, how to apply them across the range of reasons that batteries are charging with electricity provided by the grid. 

In simple terms, FERC Order 841 sets the wholesale locational marginal price as the price that energy storage should pay for charging themselves up — and this price includes transmission charges. But “there are certain circumstances where charging resources are not charged for transmission charges,” said Burwen, including when they’re providing services while also charging. For example, the battery absorbing grid energy to meet a “reg-down” frequency regulation dispatch.

“The order is pretty explicit on that," he said.

But Order 841 is less explicit about determining the difference between “an ISO effectively directing the charging of that asset, versus the decision of the market participant on their own” to charge the asset, he said. There may be circumstances in which the ISO directs that storage to charge, which the Energy Storage Association feels should be exempt from transmission charges as well. 

On a more fundamental level, Order 841 is facing rehearing requests from utilities that have challenged FERC’s federal authority over state energy regulation — a conflict that’s playing out in the arena of state incentives for renewable and nuclear energy as well. 

Finally, FERC’s decision to give energy storage assets a role as small as 100 kilowatts in ISO and RTO wholesale markets has opened up the challenge of how to aggregate, operate and manage batteries that are connected to the distribution grid, not the transmission grid. FERC has delayed implementation of this part of the order, 

“Order 841 is the start of a much bigger conversation about energy storage participation in energy markets,” Burwen said. “No one expects implementation of Order 841 to be the end of the discussion.” Instead, “This is creating all these other conversations that are starting to spin out of the Order 841 conversation." 

That includes how storage could serve as the foundation for enlisting all manner of inverter-connected distributed energy resources into the transmission grid.