by Jeff St. John
December 15, 2015

Back in 2013, the California Public Utilities Commission set in motion a mandate that calls for the state’s big three investor-owned utilities to procure 1.3 gigawatts of energy storage by the end of the decade. Overnight, it transformed California into the world’s largest market opportunity for grid-scale and behind-the-meter batteries -- at least on paper.

But it also presented regulators and utilities with an unprecedented challenge -- setting up a regime to deploy relatively unproven technologies at megawatt scale, while building the cost-effectiveness measures into place to make sure they’re all worthwhile. That’s a complicated process, and it has come with a good deal of wrangling between utilities, independent energy storage providers, traditional power plant developers and environmental advocates, with the CPUC as final arbiter.

It’s also taken a good deal of time. It’s been a year since Pacific Gas & Electric, Southern California Edison and San Diego Gas & Electric issued their first requests for offers (RFOs) to meet the 2014 procurement round of the mandate, opening a floodgate of proposals that swamped the utilities’ needs for a collective 200 megawatts.

Since then, energy storage industry watchers have been eagerly awaiting the results of that process -- and not just to learn which companies won which projects. They’re also hoping to get a glimpse into how utilities are structuring these first-ever contracts for energy storage products and services, collectively known as energy storage agreements (ESAs), and what that will mean for the much larger procurements down the road.

Now the wait is over. On December 1, Pacific Gas & Electric, Southern California Edison and San Diego Gas & Electric filed applications seeking CPUC approval for nearly 100 megawatts of storage projects -- and unveiling lots of details on how they’ve set them up.

They’ve also given the industry some surprises. Last week, we covered PG&E’s 75-megawatt procurement, which included a lot of lithium-ion batteries in different configurations, but also 20 megawatts of flywheels from Amber Kinetics and 13 megawatts of zinc hybrid cathode batteries from Eos Energy Storage, two startups that have never deployed at such scale before.

There are also some unusual contract arrangements, such as Southern California Edison’s resource adequacy-only deal with developer Western Grid for its 15-megawatt procurement. (We’ll have more on that later.) The primary filing from SDG&E, meanwhile, details why it chose not to award a contract for one specialized project it bid out for proposals last year -- an example of how utilities may judge batteries to be cost-ineffective in certain cases. 

At last week’s Energy Storage Summit in San Francisco, I moderated a panel featuring executives from each of California’s utilities involved in its storage procurements -- PG&E’s Charles Post, SCE’s Jesse Bryson, and SDG&E’s Randy Nicholson -- as well as CPUC senior analyst Manisha Lakhanpal. While they wouldn’t discuss costs or pricing for these bids (a tightly held secret for utilities and storage vendors alike), they did cover a lot of other details of the new applications, why they ended up the way they did, and the next steps in the process. Here’s a quick overview.

Pacific Gas & Electric

PG&E’s application was the largest, calling for 75 megawatts of storage in seven projects ranging in size from a 30-megawatt transmission-interconnected project, featuring developer NextEra Energy and batteries from Tesla, to several 1-megawatt projects backing up substations.

Charles Post, PG&E’s energy storage program manager, noted at last week’s event that the utility also decided to “experiment” a little by selecting 20 megawatts of flywheels and 13 megawatts of zinc-based battery chemistries. The first round is the right time to do that, he said, given that the CPUC’s rules allow for some flexibility on cost recovery for early projects.

But besides the two 1-megawatt substation batteries, PG&E chose not to own most of the projects, he said -- interesting, given that the CPUC rules allow for utility ownership of up to half of the entire mandated procurement. On this point, Post noted that PG&E’s process allowed third-party and utility-owned proposals to compete head-to-head against one another on each of the particular use cases it was looking to fill. That would indicate that third-party ownership provides a better cost-benefit outcome at present -- although PG&E, like the other utilities, is keeping a tight lid on its proprietary energy storage evaluation model.

In lieu of direct ownership, PG&E has created energy storage agreements, or ESAs, with developers of each of the projects, he said. That’s akin to a power-purchase agreement with a large-scale renewables developer, but with the added complexities of a system that can charge as well as discharge, and which has a lot more flexibility in its uses. Here’s how PG&E's application describes the key features of an ESA:

Under the ESAs, the sellers will provide a “product” consisting of all delivered discharge energy, ancillary services, capacity attributes, and any other product associated with the project and energy storage services, which will be delivered for PG&E’s sole use. The seller is responsible for ensuring that the project qualifies as Resource Adequacy (“RA”) capacity, subject to specified operational limitations.

"Resource adequacy" is the mix of generation and demand-side assets California utilities are required to procure to cover the peaks and troughs in statewide energy supply and demand. For the most part, all of the contracts we’ve discussed so far are meant to meet these RA needs, including four hours of capacity per day.

Beyond that, however, PG&E can schedule the dispatch of that asset as if it owns it, as long as it doesn’t exceed any of the restrictions placed on battery charge-discharge cycling, depth of discharge, or other parameters that can affect an energy storage system’s lifespan, Post said. That opens up the opportunity for additional revenues in the state’s energy markets, or to manage the utility’s grid capacity needs in specific areas.

PG&E is also asking to recover the costs of these projects in its general rate case, expected to get underway in 2020. Specifically, it’s looking at recovering the costs of projects through the Energy Resource Recovery Account, which allows the state’s utilities to adjust projected energy costs due to shifts in fuel costs and other variables. It’s also asking for the “recovery of above-market energy storage costs associated with the ESAs using the Power Charge Indifference Adjustment,” or PCIA -- the state’s unique mechanism that allows utilities to impose charges for “departing load.”

Southern California Edison

Southern California Edison’s application asked for a lot less new energy storage than PG&E’s did -- only 16.3 megawatts, to be precise. But that’s only because SCE has procured so much energy storage already, as part of its long-range effort to shore up its West L.A. and Orange County regions in the wake of the unexpected closure of the San Onofre nuclear power plant.

In a groundbreaking local capacity resource (LCR) procurement last year, SCE contracted for about 260 megawatts of energy storage, including some 160 megawatts of behind-the-meter batteries, ice-making air conditioners, demand response and other distributed resources. While that’s part of a different CPUC proceeding, the utility has been allowed to count parts of it for its 2014 procurement target of 90 megawatts.

But what SCE’s 16-megawatt application lacked in size, it made up for in innovative contract approaches, as Jesse Bryson, principal manager of power origination, outlined at last week’s Energy Storage Summit. The first, a 1.3-megawatt lithium-ion project contracted to General Electric, will be tied to the Stanton Energy Reliability Center, a natural-gas-fired peaker plant -- an “integrated peaker,” he said -- to reduce fuel costs and increase its ability to ramp up and down more quickly.

The second, a 15-megawatt project featuring Eos’ zinc hybrid cathode batteries, is an interesting “RA-only” model, Bryson said. That means that SCE will buy the RA credits for the project, but turn the future market values of the storage system back to developer Western Grid.

That’s not a unique concept -- distributed energy resources being procured under the Demand Response Auction Mechanism, or DRAM, pilot project launched last month also split up utility RA credits and developer market revenues. If SCE’s contract is approved by the CPUC, it could become part of the menu of options for meeting the state’s energy storage mandate as well.

SCE’s contracts for its LCR procurements will also feed into the mandate, which brings a lot more flavors of contract into the mix. Bryson noted that the utility has seven different forms of contracts available at present for these distributed energy resource procurements, starting with the specialized contracts it created for Stem, Ice Energy, NRG Energy, Advanced Microgrid Solutions, AES Energy Storage, and the other companies it’s tapped for local capacity.

The utility expects more distributed storage to come on-line in future months as part of its Preferred Resources Pilot (PRP), which is seeking megawatts of distributed energy resources to support two stressed-out Orange County substations. The utility has already acquired energy efficiency, demand response and two rounds of solar PV projects under the pilot, and expects to bring energy storage on-line as well, he said.

San Diego Gas & Electric

SDG&E is much smaller than the state’s other big IOUs, giving it a much smaller storage mandate target -- 165 megawatts total, with 20 megawatts due in 2014. It also had about 51 megawatts of existing or under-construction energy storage projects, including 40 megawatts of pumped hydro at Lake Hodges, 7 megawatts of Smart Grid Storage Demonstration projects, and about 4 megawatts of customer-connected storage or permanent load-shifting technology, putting it well ahead of compliance with its first round.

Despite this, SDG&E sought out another 16 megawatts in its January 2014 RFO, split into one 10-megawatt transmission-connected project and two smaller distribution-connected projects, at 4 megawatts and 2 megawatts apiece. The utility promised exciting things in its solicitation document (PDF), with “offers from a wide range of bidders” that have led to shortlist negotiations for “first-of-their-kind energy storage contracts."

But SDG&E has a three-month extension until March 1, 2016 to pick winners of that RFO, which means that Randy Nicholson, SDG&E policy manager in charge of its storage mandate efforts, wasn’t able to say much about it at last week’s conference.

He was able to talk about the “post-solicitation report” the utility filed in December, however. That’s an interesting document, because it’s the first time a utility has laid out its decision to pick a “none-of-the-above” option under the mandate. Specifically, SDG&E chose not to pick any of the 12 offers for a 4-megawatt, 12-megawatt-hour storage system to relieve peak conditions on a specific set of substations, he said.

That business case is known as distribution deferral -- replacing traditional poles, wires and transformers upgrades with energy storage that can inject power at times of peak load, reducing strain on old or overloaded assets to keep them in the field for a few years longer.

These are far more complicated and site-specific than the broader RA and capacity contracts that have made up the bulk of storage procured in California so far, Nicholson said. They come with the requirement of utility ownership, and the ability of bidders to follow standard distribution equipment procurement procedures (i.e., selling batteries as if they were transformers or other grid gear) and integrating with utility’s SCADA systems, distribution management software, and other control platforms.

SDG&E hired consultancy DNV GL, which used its ES-GRID software to measure each storage bid against other options, such as building a new substation, or installing “transformer bank overload triggers” to reduce the risk of damaging expensive equipment. Out of 36 combinations of technology and financial options, only 12 came back with a positive net present value (NPV).

Half of those were only cost-effective if one assumed SDG&E’s substation replacement costs ran over 20 percent of its budget -- unlikely, given the 30 percent contingency adder already built in. Four more were cut because they “contained warranty options that were significantly less than the asset’s useful life,” the analysis found. Of the two left, one just broke even on costs and benefits, and the last one showed an estimated NPV savings of $7 million.

That’s about 5 percent of the total installed cost of its top-ranked storage proposal, which means “if the actual costs exceed the estimated costs by 5 percent or more, the immediate value to customers is entirely eroded,” the utility wrote. Compared to the option of building a new substation, which is completely risk-free from the utility's perspective, SDG&E chose not to pick batteries at all.

This doesn’t indicate that SDG&E plans to start rejecting energy storage projects, Nicholson said -- only that each project has to make economic sense compared to the alternatives. AB 2514, the state law that set the energy storage mandate in place, requires that utilities pick cost-effective storage. But as SDG&E noted in its report, “neither the Storage Framework nor the 2014 Procurement Plan Decisions outline a process for communicating the results of solicitations that do not result in contracts.”

California Public Utilities Commission

As a senior analyst for the CPUC’s energy storage mandate, Manisha Lakhanpal is one of the few people who’ve seen each of the utility’s December applications, along with the top-secret data on how much they’re proposing to pay per kilowatt and kilowatt-hour.  Speaking on last week’s panel, Lakhanpal shied away from discussing these price points. But she did explain how the CPUC is building them into its “consistent evaluation protocol,” or CEP -- an important term for storage mandate watchers, since it encompasses many of the cost-effectiveness issues facing the commission.

Here’s a July CPUC presentation (PDF) that explains how CEPs are constructed from both quantifiable data, such as costs and prices, and qualitative data, such as which of the 20 different storage use cases the project is seeking to fill. The state’s utilities have already proposed their own versions of CEPs, which are being applied to their current round of projects.

But CEPs are strictly reporting tools at this time, without any bearing on contract selection, PG&E noted in a July CPUC workshop (PDF). They also don’t contain any of the proprietary data that utilities are using to value and compare projects -- although groups including the Sierra Club and the Environmental Defense Fund have asked the CPUC to open more of that data to public review.

During the question-and-answer section of last week’s panel, James Fine, senior economist with the Environmental Defense Fund, asked again for some of this data to be made public in order to allow third parties to assess its accuracy. The utility execs on stage replied pretty strongly in the negative, saying that the CPUC has already closed the books on the subject.

But this data will be making its way into the CPUC’s next tracks of work on the storage mandate, Lakhanpal said. Track One will be handled in a proposed decision due soon from the commission, dealing with “lessons learned” so far, and whether any early changes to the mandate need to be put in place, she said. 

Track Two deals with the next round of procurements for 2016, and is set to start in March, she said. That proceeding will look at revising the round’s megawatt targets, consider new eligibility rules for different technologies, and open the tricky subject of assessing the value of storage systems with multi-use applications, she said. 

As for where the storage industry can look for clearer data on costs and values for energy storage, it's hard to see how proprietary data on utility-scale projects will emerge. But Lakhanpal noted that the CPUC's distribution resource plans proceeding, which is asking the state's big three IOUs to find the value of distributed energy resources for their distribution grid investment plans, could start to reveal more data on how storage can meet utility needs on the edge of the grid.