Working on the utility of the future
Written By: Stratton Report
February 13, 2017
As William Gibson says, the future is already here, it’s just not evenly distributed. One of the places where it seems to be lying a bit more thickly on the ground is in southern Arizona, where Tucson Electric Power is dealing with several challenges to “business as usual.” The utility is simultaneously moving away from retail net metering for solar, installing its first grid-scale energy storage (and thinking about getting lots more), seriously rethinking its resource mix (including tripling its renewable energy supply) and looking to promote some fairly significant changes to the standard utility-customer business model. We spoke to Carmine Tilghman, senior director of energy supply at the utility to get an update on where he and TEP see all this going.
The Stratton Report: Tucson Electric has three energy storage projects that are in the news. Could tell us a little bit about how those are going?
Carmine Tilghman: Let me start with the NextEra and the E.ON Climate Renewables energy storage projects. Those were two that were awarded via our competitive solicitation process. We went to market looking for a set amount of energy storage capacity to solve a very real frequency response issue that we were experiencing on our grid. We ended up choosing both of those proposals, each with 10-MWs and 2.5-MWhs of energy, optimized for frequency response. The NextEra project, which just went commercial on February 27, is built around lithium-ion nickel manganese cobalt batteries, technologically very traditional. It is operating as a stand-alone system tied into one of our substations. The E.ON project is a lithium-ion titanate chemistry, and it is being paired with a 2-MW solar facility. It is at the U of A Tech Park, and it is scheduled to be operational by the end of March. The associated solar facility is, for all intents and purposes, completed. So, it’s been very exciting.
SR: Have you learned any lessons, now that the first project is working?
CT: Working with our very first energy storage project was a challenge; getting it installed, getting all of the engineering team on board, and going through the commissioning process to be sure it operates properly and is visible through our energy management system—all that has provided us with a terrific learning experience. I would say that for the very first ESS project, that process, although it might’ve seemed like organized chaos at times, went relatively smoothly and was delivered on time. We were very pleased. And our second project, now that we have commissioning down, will be easier, and future ones should be easier still. So we expect the E.ON project to come online here at the end of March, and have two operating 10-MW facilities for frequency response for our system.
SR: So how about the third project?
CT: The IHI Project is a 1-MW battery storage facility, and that is more of an R&D project. IHI has asked TEP if we could facilitate putting in a storage system at one of our local solar facilities. is the project is designed to test its ESWare planning and control software and work with TEP to develop efficient control strategies for energy storage systems. That project is not operational, but still under construction. I believe it will be ready sometime this year.
SR: Does the addition of solar power to the E.ON project add to the complexity of managing it?
CT: It does, but mostly it makes things more complicated for the developer. I have said many times publicly that I’m not a fan of pairing storage with a specific renewable facility. Doing that takes away from the energy storage units’ ability to respond to the needs of the grid overall, and focuses its capabilities primarily on the needs of that individual renewable facility. And that’s not necessarily in the best interests of the balancing authority, looking at the system in its totality. It’s basically done to capture the tax incentives that are associated with renewable facilities, which can be extended to a storage facility if the storage facility is charged with a certain percentage of the renewable energy generated onsite. I view solar + storage as beating a square peg into a round hole; you are forcing a project’s design into a less than optimal configuration to take advantage of the tax incentive. Obviously, it would be far more beneficial if the policy makers created a tax incentive designed specifically to promote the energy storage industry. With a “paired” system, the control program has to be more complex, the system design and engineering have to be more complex, and the whole thing is more complex.
SR: I understand energy storage systems have multiple levels of control programs, to manage the needs of the battery pack as well as the way in which the ESS interacts with the balancing authority. Does the balancing authority need to be aware of the “internal” constraints of the ESS?
CT: From the standpoint of the balancing authority, what the grid energy management system interacts with is the top-level control program. That top-level interface program has been written to provide the frequency response we need and to do it autonomously. In our original solicitation, we identified the specifications that we needed the battery to operate to and the bandwidths. So the ESS developers took those parameters and wrote their program to our specs, to tell their storage system when to charge or discharge, when to provide frequency response, and support the grid as necessary in response to signals coming from the EMS. The interface program for the storage facility senses certain grid characteristics, specifically frequency, and will respond at pre-programmed set points. The balancing authority operator can step in and provide a manual override signal to force the charge or discharge in the event that perhaps there’s a system outage that we need to respond to, especially one that may not show up in the overall system frequency.
SR: How’s all that working?
CT: We’ve only been operational for a few days. But it tested fine. It worked fine. It met all of the specifications. So the balancing authority folks are very happy and very excited to be getting not only the first energy storage system online, but the next one in another couple months.
SR: I understand that over the next decade or so you guys are planning about a 300 percent increase in the amount of renewable assets that you’re going to have on your system. What’s going to be necessary for your utility to accommodate that?
CT: We’re right in the midst of a resource plan, it’s due April first, and which has obviously changed dramatically since the company has taken the position that we will attempt to achieve 30 percent renewables by 2030. We have shown presentations on our plans to the regulators and to the public. We, like everybody else in the utility world, are familiar with the California “duck curve” showing how system load will change over the course of a typical day, and we often reference that daily California load curve in our presentations. And I often joke when I reference it in my presentations that if I only had California’s duck curve to deal with, I’d be a far happier individual. The California utilities’ task is far easier to solve than the task we in Arizona are going to have once all those renewables come on line. With all those renewables, and a significantly higher percentage of solar among those renewables, we are going to have a disproportionate amount of renewable generation in the middle of the afternoon. And while this issue is significant during our peak load season in the summer, much like California it gets drastically worse during the “shoulder” months. When we look at the spring, which is our lowest-load period, we see that within the next decade, we will have a system load at times that would only require one-tenth of our traditional generation resources while the other 90% would be served by the renewable energy. In the spring, during the middle of the day, we expect our system demand levels will be about 1000-MWs. Because of the total amount of renewables that will be generating then, our conventional baseload generation plants will only need to serve 100-MWs. The minimum generation for our traditional resources is about 500 MW, so we have a 400-MW gap in the afternoon—because even if we turn down all of the conventional plants to their minimums that would need to be running then, we will still be putting out some 500-MWs, and we only need 100-MWs to serve the remaining load. What will we do with the excess power? We can’t avoid generating it. And then in the three-hour period immediately following, we will see an 1100-MW ramp in demand for electricity to meet the evening peak. So we’re looking at how we can redesign our system. There will need to be a number of solutions; including storage, flexible contracts, fast-starting gas technologies, and possibly even coal-plant cycling.
SR: Will you replace that coal-fired capacity with different generating technology?
CT: Yes. It is likely that we will acquire additional natural gas facilities that have higher ramping rates than our traditional units. We will be looking, more than likely, at several hundred megawatts of reciprocating engines, which can both start and shut down within a matter of minutes, without any significant wear and tear. We will be looking at altering the contracts that we will utilize for large-scale renewable projects to insert curtailment provisions. We will be looking at large-scale energy storage capacity, potentially 100 MW or more. Fortunately, technology is moving pretty fast right now, and we’re hoping to take advantage of some of the newest technologies sooner rather than later.
SR: Why so much energy storage?
CT: Because I can only use a traditional generation plant to serve load, but with energy storage I can shift load. Energy storage can create a load in the middle of the day, when I have excess generation, and then help me peak shave in the evening when I need more power. So we have to do our resource planning differently going forward. The traditional stack of base load, intermediate load, and peaking generation, from the old days, will still be there, but it will have a considerably different look. It will no longer just be about items like the ramp rate capabilities and fuel costs. Almost all of the storage on the grid today is providing frequency response in some type of open market, with the economics based on some form of market payment for response. We’ll be one of the first vertically integrated, regulated utilities to go to our commission and say, “We want to put the storage in from a balancing authority perspective to manage grid operations, and it will actually be more for the purpose of pure energy storage—charging in the middle of the day and discharging in the evening for peak shaving and load balancing-as opposed to providing frequency and voltage support.” We probably won’t specify the exact kind of technology to provide such pure energy storage because the markets are moving faster than we are, so that we’ll use that solicitation to shake things out. Lithium ion energy storage systems may be able to give us the four-hour, six-hour or maybe longer time period storage we need between charging and using the power. If lithium-ion batteries aren’t the most cost efficient way to do this type of longer-term energy storage, then we might use one of the varieties of flow batteries. Or something else altogether. We’re agnostic. So this will be a big change for us. But it’s a very, very real issue and based on current projections it will start to become a significant issue for us by 2022, which is not terribly far off. Otherwise, we’re either going to have to start cycling our existing base load units, or decommission and replace our baseload units with more flexible generation—both of which are very expensive options. So there’s a very, very short window for us to find a good solution.
SR: The state of Arizona in the past had a net metering policy for residential solar, in which utilities paid as much for the excess power homeowners generated in the middle of day as for the power homeowners bought from the grid. Recently that has changed, with homeowners now getting paid a lower price for the excess power they generate. Did Arizona utilities and regulators intend to incentivize homeowners to install residential energy storage, so that they could use the excess energy they generate in the middle of the day themselves in the evening? Are you guys be willing to consider aggregated residential storage, also called virtual power plants, as part of your resource base?
CT: The short answer is: yes. We’ve had exactly these discussions, and I am a big proponent. With traditional retail net metering, when you live in an area like Southern Arizona where we don’t have many outages, there’s really very little incentive for a consumer to spend that money to promote behind-the-meter storage. Ultimately, it will be necessary to change rate design with the goal of incentivizing the market to respond. We want the residential energy market to respond with a customer based energy storage solution, because that actually does provide a much greater benefit to the utility than just the simple solar system with no associated onsite storage. I’m not criticizing solar power, but the truth is it really isn’t as peak coincident as most people believe, and it really doesn’t benefit the system in many ways that people believe it does. Being a grid operator, we see the issues associated with solar on the grid, which also allows us to also see that home-based energy storage truly can provide a tremendous benefit to the grid. It gives residential and small commercial customers the option of trying to shift their load, to reduce that evening or morning peak. That is, in fact, what we would want a consumer to do, and you make that happen by installing a home-based storage. Of course, each home needs only a small amount of storage, and as a medium-sized utility with a peak load of 2400-MWs, it would take significant numbers of home energy storage units to make an impact system-wide. So we are definitely hoping that large numbers of home energy storage units, aggregated together which could potentially be operated as virtual power plants, will be installed in our service territory.
SR: Won’t that kind of alter the standard utility model?
CT: It may well be 20 years down the road, but I think there could be a beautiful blend of utility-scale facilities providing the overall system balancing, while the consumer or the virtual power plant, operating on an autonomous set of features, or pre-programmed features, will provide that small-scale stability at the actual load point. We do think about that, and believe that the grid of the future will incorporate these technologies. The discussions need to had, and these discussions will continue to be had among the utilities, the regulators, the stakeholders, and the public. To bring it about, beyond simple pilot programs, we need to start with a regulatory policy, and rate design, to incentivize the market to want to take advantage of that rate design.
SR: In an interview from a couple years ago with TEP’s CEO, David Hutchins, he promoted the idea of having your utility finance distributed energy resources and energy efficiency upgrades for customers, who would pay for it on their electric bills. Is that still under discussion?
CT: Yes, the discussion continues and this is still an area that our CEO wants to focus on. A properly designed program can lead to many benefits – for the customer, utility, and the grid – while eliminating the fears of critics who believe the regulated public service entity would be encroaching on their business model. We are an energy service provider with tremendous resources and expertise in energy production, transmission, distribution, and consumption. Why not have the utility partner with those vendor providers? It makes perfect sense to take advantage of synergies among product and service providers, while offering new products and services to individual customers. I’m certain you’ll see proposals in the future for these types of programs.