America's leading renewables market in flux
Community choice aggregators in as many as 23 California counties, power marketers and customer-sited generation like rooftop solar could take as much as 85% of electricity load from utilities by the 2020s. The California Public Utilities Commission is in the process of changing two key constructs that are central to the economics of solar: net metering and time-of-use pricing. Storage is starting to gain a foothold and could displace gas peakers. Four key market participants had a lively discussion at the 28th annual Chadbourne global energy and finance conference near San Francisco in June about how these and other changes are transforming California.
The panelists are The Honorable Liane Randolph, a member of the California Public Utilities Commission, Jan Smutny-Jones, CEO of the Independent Energy Producers Association in California, Ed Fenster, executive chairman of solar rooftop company Sunrun, and Susan Kennedy, CEO of energy storage company Advanced Microgrid Solutions and former chief of staff to California Governor Arnold Schwarzenegger. The moderator is Todd Alexander with Norton Rose Fulbright in New York.
MR. ALEXANDER: Jan Smutny-Jones, what are community choice aggregators?
MR. SMUTNY-JONES: It is fitting that this panel follows the Latin American one. California has a lot in common with our neighbors south of the border, except you cannot get political risk insurance for doing business here. [Laughter] In 2002, after the California energy crisis, the state put a freeze on allowing electricity customers to choose their suppliers. We came up instead with the concept that communities can choose to leave the utility and form their own procurement entities.
It is important to understand that a CCA is not the same thing as a publicly owned utility. It is only doing procurement. This lay dormant for about a decade and then around 2010 to 2012, Marin County, which is north of San Francisco, formed a CCA. I think there are five now: Marin, Sonoma, San Francisco, the peninsula, and Lancaster, which is down in the high desert.
They are basically joint powers agencies that are not tied directly to a local government, so this presents a credit challenge. Are they creditworthy? They have been growing over the last couple years, and there is speculation that anywhere between 40% and 80% of the electricity customers will end up in CCAs.
One last important point: it is not really choice. If your local government decides to form a CCA, you become a customer of the CCA. You can opt out, meaning you can go back to your local utility.
MR. ALEXANDER: Susan Kennedy, you were on the California Public Utilities Commission around the time these rules were first being implemented. Was it intended that 20% to 80% of customers would procure their energy through community choice aggregators? Why were CCAs put in place? What do you think the impact will be?
MS. KENNEDY: Interesting. The market today is one in which commodity prices are going down, rates are going up, and there is an oversupply. These are the same conditions that preceded the energy crisis in 2000. But back then, as Jan suggested, you could depart the bundled service and be a direct-access customer.
Now that that gate has been shut and frozen, a CCA is the path of choice for communities that want to get out of bundled service. The concept lay dormant for a number of years because the economics were not compelling enough to use it, but we are starting to see the same conditions again where people are looking for alternatives to bundled service. A CCA is the place to go because direct access is not available.
MR. FENSTER: This is really just a longer-term trend that we see in our business and that drives our business: customers want choice. Rooftop solar is one manifestation of that. CCAs are another manifestation of that.
There is a hot debate about how this should work, but if a community leaves the local utility to form a CCA, there is a distribution charge or what is called a power charge indifference adjustment.
You are seeing a lot of CCAs supporting renewables. People want a higher renewables percentage than the local utility is currently offering. Marin has over twice the renewables procurement that PG&E has.
MS. KENNEDY: That’s right. While the original motivation for CCAs was just wanting to get out of bundled service from the local utility in an effort to lower costs, the availability and falling prices for renewable energy have really fueled the growth of CCAs. The CCAs want cleaner energy.
MR. ALEXANDER: Jan Smutny-Jones, CCAs want more renewable energy. If I am a solar developer with a 100-megawatt solar project, can I sign a 100-megawatt power purchase agreement with a CCA and use that as a basis for financing my project? How do I prove to the banks I want to lend me money that the CCA is creditworthy?
MR. SMUTNY-JONES: There are several issues buried in that question. CCAs were originally not seen as a suitable foundation for building new large-scale power plants. The thought was new generators may sell some of their output to the CCAs, but not the entire output.
Some of the CCAs have started recently to enter into long-term contracts. It is unclear whether this type of buying will become the norm or represents a few one-off situations.
I have been doing this for a very long time and my understanding of the wall of money that we have heard the investment bankers say is searching for projects has always wanted creditworthy counterparties. CCAs do not have credit ratings. Perhaps Marin is different. It has been in existence for a little longer than the others. It is a wealthy community. Maybe relying on it to pay is a low risk. However, creditworthiness is very much a live issue for how this phenomenon gets any momentum.
MR. FENSTER: It would be malpractice for anyone trying to start a power company today not to realize we are in a state of deep change in the electric industry. We have the rise of electric cars, which might add to load. There is an energy efficiency movement, which might reduce load. More people are opting for choice. More people want rooftop solar. Projecting exactly what load will look like five or 10 years out is as hard as it has ever been, and so you need to adopt an athletic position and be well capitalized to be able to move with the changing market.
MR. ALEXANDER: Perfect transition to the next question. Liane Randolph, how do the utility regulators plan in such a market? The state has a goal of getting to 50% renewable energy by 2030, and there is talk of pushing the target to 100%. How does the state plan in a market where it is not clear who will buy electricity from whom and who is obligated to do what?
MS. RANDOLPH: Ed said there needs to be an athletic approach. The PUC is exactly the opposite of that. [Laughter] We are a little slow, unwieldy and cumbersome, but we are trying to reframe our approach to long-term procurement in an integrated resource planning process that we are in the middle of launching. We are trying to be a little more resource-neutral from the supply and demand respect and trying to optimize a portfolio. The CCAs are saying they plan to procure as many renewables as possible, which is great, but we also need to look at the full landscape of reliability and costs.
The PUC has over the years really driven development of this great renewables market. We had the ability to tell the investor-owned utilities, “Thou shalt go do this, thou shalt go do that.”
As load departs and goes more toward CCAs and customer choice, it creates a regulatory challenge to make sure that all of those entities are procuring the right mix of resources going forward.
MS. KENNEDY: I think it is the customers who are becoming more athletic. They are the ones who are driving much of the change. They are the cause of much of the difficulty for any load-serving entity to figure out how best to supply its load among gas, wind, solar and storage. The customer is now capable of responding to any price signal in a dramatic fashion. You have every load-serving entity, whether it is a CCA, bundled service or a direct access provider, trying to chase a moving target.
MS. RANDOLPH: You hit the nail on the head. Are we sending the right price signals to get what we need? Are customers being incented to do the right things for the grid and for reliability?
MR. SMUTNY-JONES: This raises a serious problem. The legislature, in its wisdom, decided that we need to do an integrated resource plan, which is fine if you have three entities that are doing integrated resource planning. We have done this in the past, not successfully, but we did it. Now, if you divide PG&E up into eight different pieces with different CCAs, what is integrated about that? How is that going to work? You will be shocked to hear that the CCA is telling the PUC, “You are not the boss of me.”
It is an interesting question how to optimize the system while atomizing it into smaller and smaller pieces.
MR. ALEXANDER: Liane Randolph, I imagine the investor-owned utilities are not very happy. They have large stranded costs. They were obligated to enter into long-term contracts to buy renewable energy when it was still expensive, and now everyone wants to move to CCAs to buy renewables at the cheaper prices on offer today. How do you deal with that?
MS. RANDOLPH: Yes. That is the big issue before the commission.
MS. KENNEDY: The customers are following the price signals given to them through tax credits to encourage installation of rooftop solar systems, net metering, and feed-in tariffs. They are putting a bunch of solar on their roofs. The regulators are trying to catch up to what is happening in the market.
You have bundled utilities trying to push the time-of-use rates to later in the day in order to recover the system costs of trying to manage the electrons flowing back and forth on the grid. You have customers who put solar on their roofs saying, “Whoa, don’t make me pay for somebody else’s problem when I did exactly what policymakers asked me to do, which is go to clean energy by putting solar on my roof.” This is a case where you have the regulatory bodies and utilities woefully behind on ability to catch up to where the market is.
MR. ALEXANDER: Ed Fenster, explain what California is doing with time-of-use pricing to try to balance the problem the utilities face with stranded costs with the interests of customers who put solar on their roofs.
MR. FENSTER: Let’s talk separately about price signals and rate design. There is how we as a solar company deploying storage react, and there is a consumer education component. Time-of-use pricing is a good policy. We support it. What is it?
There are a few ways to set rate structures. There is a flat volumetric rate where every kilowatt hour you price costs a certain amount. You can do that by time of day where you say, “Between these hours, it is more expensive than between those hours.” Maybe the prices also vary by whether it is summer or winter.
You can assess a demand charge which is based on the 15 minutes of your highest consumption during the month. Then there could be a fixed charge or a minimum charge.
In the residential sector, simplicity is really important. Demand charges do not work for homeowners because if I am using the microwave and someone upstairs is blow drying her hair, I don’t know my electricity usage during that period. There is just no way to manage it.
It also should not matter because any customer turning on a couple appliances does not create system-wide grid cost. It is not like you have homeowners with aluminum factories in their basements. What matters is the aggregate system load, not the individual customer usage.
If it is the case that power is more expensive between four and seven in the evening, or whenever the hours are, then it is appropriate that a price signal is sent and we as a company can say, “Then we will start installing storage, and the batteries will shift the mid-day power so that it can be used later in the day.” Maybe you install more west facing solar as a different example. We are actually nimble in responding to that.
One thing we learned, as net metering transitioned to time of use in California, is that it is a major effort to reeducate the customer base to adapt. There are tens of millions of people to whom you are effectively launching a new product. You have to think of the marketing costs necessary to explain the shift.
So we think these sorts of changes are appropriate, but they should be slow and well telegraphed because otherwise the marketing expense that we and our peers face ends up really high.
The alignment to socialized cost is important. We will see, particularly with the rise in storage, that the value of distributed resources on the grid is really high. The argument on the utility side about storage has been, “You are shifting costs to everyone else.” Then the solar companies have said, “Yes, but every time we install a solar system on the grid, we are deferring investment in your transmission and distribution infrastructure. Less transmission needs to get built. The distribution system lasts longer.”
In California, a lot of power plants are on the coast. I don’t know how many of you are from California, but if you have looked outside, you can probably tell the coast would be a bad place to site solar systems because of all the fog. Our solar projects are on the other side of the state. If you then think about how you get to 80% renewables, our transmission grid is not really set up to accommodate that. Because the rooftop solar installations move renewable generation to the customer site, we are saving a lot of money as a state. Then when you layer in storage, which is beginning to happen, the numbers get even more exciting.
It should be obvious to everybody that storage is worth more on the customer side of the transmission grid than on the generation side of the grid because, if your pipe in between the two is full, it does not matter how much storage you have on the generation side.
For every 20 customers to whose system we add storage, that is one megawatt hour of storage. It is a 40-foot shipping container if you want to try to visualize it. Where are you going to put that in urban and suburban environments? We will need to see a lot of storage deployed. A lot of it will have to be distributed on customer sites because there is not a lot of other real estate in these places to do it.
So those are the components that I think make the value of solar and solar plus storage effective from a socialized perspective.
MR. SMUTNY-JONES: Now for something completely different. [Laughter]
There are some customers where what has just been described actually applies and may make sense. I am not disputing that. But the vast majority of people in my state, the 39 million people here, are not interested in this. They want the lights to go on. They want electricity to be affordable. They want it to be clean.
The group I represent also includes utility-scale solar generators. We heard a panel talk yesterday about how solar is being bid into utility auctions at less than 3¢ a kilowatt hour. That is not what is happening at the distribution level. My first job was with Western Solar Utilization Network in 1980. If you told anybody back then we would have a problem with the duck curve, that we would have too much electricity in the middle of the day, that would have been unbelievable, yet that is where we are today.
In 2008, we had 300 megawatts of utility-scale solar in this state, and it was all solar thermal. Today we have almost 10,000 megawatts of utility-scale solar and another 5,000 megawatts on roofs. Things have changed significantly.
How this gets managed is a pretty big deal. There is a relationship between storage and net metering. One of the arguments is that if you are benefiting from net metering, you kind of already are using the grid to store your electricity. The utilities are giving you back power at night when you need it. It is supposed just to be a swap of kilowatt hours. If you are now talking about doing something else with respect to batteries, it opens up interesting legal and financial questions.
MS. RANDOLPH: Jan raises a great point because we at the PUC have to worry about the customers that are not focusing so much on their energy usage. They are not looking at rooftop solar. They are not looking at rooftop storage. They do not read their bill inserts, and do not go to the website and read the detail. They just want to know that they will have reasonable and consistent electricity rates.
The education lift of trying to get people to respond to price signals is huge. Most people don’t think about electricity usage and price signals. All they see is the amount of their monthly bill. They don’t understand why the number is changing.
In my household, we switched to time-of-use rates because I thought we had better walk the walk. I still can’t get my 16-year -old not to turn on the convection oven the minute he walks in the door to do his California Pizza Kitchen frozen pizza, and then he eats the entire thing! [Laughter] I am dutifully waiting to turn on the dishwasher until 9 o’clock, but he still has that convection oven on at 4 o’clock. [Laughter] It is difficult.
MR. SMUTNY-JONES: The integrated resource plan will take care of it. [Laughter]
MR. FENSTER: At least it is only a convection oven. This is why storage is so important. One reason why there is some regulatory work to do is because a lot of our solar-plus-storage systems are like having a thoroughbred locked up in the barn. Our Hawaii installations involve just self-consumption. We never export to the grid, and we slowly draw power from the battery over night.
The much more societally efficient thing would be for HECO to call us in the evening, when solar is coming down and the gas stuff is coming up, and we would just blast the battery out. Then you could help with those transitions, but what we do not have yet is a rate or framework to do that.
The regulators, utilities and the distributed companies will be working together over the coming years to figure that out and to help solve these problems. As we get to higher concentrations of renewables, we are going to need the storage. Think about how helpful it would be for the solar eclipse expected in August to be coordinated in that way.
We have to get storage deployed, and then we have to figure out how to coordinate better, because there is a lot of societal value to be had.
It is fascinating in our business how our marginal cost is encroaching on the utility-scale cost. The marginal cost of increasing a customer’s rooftop solar system from six to 10 kilowatts is now in the single digits per kilowatt hour. Ultimately we want to get as much renewables as possible: not just electric, but also for transportation and heating. If we install a larger solar system, we can switch a customer’s heat from gas to electric and do it at a lower cost on the margin and then also start to strip away the greenhouse gas emissions from gas and oil use. There are things like that that I think are important to think about as well.
MR. ALEXANDER: Susan Kennedy, where do you think storage makes the most sense today? Where is it getting traction? What policies are needed to make the best use of the energy storage that is available?
MS. KENNEDY: The secret to energy storage is that nobody wants batteries. Everyone wants what the battery makes possible. It is a load control technology, and so the question is really who needs load control? Who is willing to pay for it? Where do the economics work paying for it?
It is challenging to make storage work on the residential side because of the economics of such distributed scale. The scale is a little better with large commercial and industrial customers. When you get into battery storage behind the customer’s meter, the critical issue is where is the return on that investment and who will make such an investment.
There is an artificial delineation today. The distribution system stops at the customer’s meter. Everything behind the customer’s meter is retail, and it is the customer’s problem.
The only way to get to a position of being able to use storage for load control and for system planning and distribution-level benefits in order to address some of the issues that Commissioner Randolph talked about is when the utilities have visibility into, and some control over, the consumption on a large scale behind the customer’s meter. The economics have to be able to translate behind-the-meter energy storage at a large-scale level into distribution-level benefits at the utility scale. If not, the economics of deploying storage will never pencil out.
The simple answer is that the battery is a piece of the grid infrastructure. If Congress wanted to make a significant, huge investment in infrastructure, it would make the investment tax credit available for stand-alone storage and let storage be deployed where load control makes sense and let the utilities use it for customer load control in a transactional way in order to balance the grid.
MR. FENSTER: I have a little different perspective. One hundred percent of our new systems being sold today in Hawaii have storage. We probably have the leading market share as a result in Hawaii today. Fifteen percent of Americans already have some form of back-up generation. Usually it is a gas back-up generator.
Homeowners are actually willing to pay for back-up storage and are willing to split the use of the battery with us. This is slightly more true on the east coast where the perception is the electric grid is less reliable than on the west coast. I am optimistic that customers across the country will eventually find the value proposition of having back-up power is significant enough that it will defray the cost. I am very optimistic about the long-term deployment of residential storage.
MS. KENNEDY: How much of the deployment is contingent on the investment tax credit and subsidies?
MR. FENSTER: We are first and foremost a solar company, and we do storage with solar so it qualifies for the investment credit. We suspect there is a market that is not necessarily tied to solar, but I agree with you it would make more sense if the investment credit were not written the way it currently is.
MS. KENNEDY: The point I am trying to make is that the utilities are struggling to figure out how to create reliability in a system where you have such mass deployment of solar. In such a market, attaching storage to solar is the responsible thing to do. It is the economic thing to do. However, unless it is part of the utility solution for control, for balancing supply and demand, then the cost for managing that is going to be borne by the grid or by all the ratepayers in that area.
MR. FENSTER: We are working on solving exactly that issue. We announced a partnership in January with National Grid to propose and build all the market mechanisms that would allow us to work with utilities on everything from capacity for the batteries to even frequency regulation and other forms of services. Texas is probably the only market you could do it right out of the gate because it is not regulated by the Federal Energy Regulatory Commission. It is a totally open market.
We are building our business plans around working with utilities so that if they need power during the evening solar ramp down or for any other reason, they can call us and we can guarantee to deliver it. We hope in five to 10 years that regulators will bid out in a competitive process new utility transmission and distribution facilities. We might even participate in that and see if we can do it, distributed, at a lower cost.
Tying Everything Together
MS. RANDOLPH: The California Independent System Operator had a stage I emergency last month or the month before, which is the lowest level of reliability. It basically misforecast the demand for that day. There was cloud cover. It was hot.
MS. KENNEDY: Cloud cover knocked some of the solar off line.
MS. RANDOLPH: Also some out-of-state resources were unavailable. There was no actual problem because the ISO was able to call on demand-response resources to reduce the load.
MS. KENNEDY: One of our clients is a large industrial that signed up for one of the reliability programs called “base interruptible program.” It faced a $125,000 fine because it was not able to respond quickly enough to that stage I emergency alert. There is a huge mismatch in the economic signals we are sending and the ability to count on demand response.
MR. FENSTER: California has a demand-response program. We bid into it with our storage assets this year and will participate in it next year. It is not a perfect fit for storage, but it is still ahead of what other states are doing for behind-the-meter solar-plus-storage technology.
MS. RANDOLPH: Part of the problem is defining the products and the dispatchability correctly.
MR. SMUTNY-JONES: There is also a need for a more robust grid. Many years ago, I was chair of the ISO and if there are 1,000 megawatts of demand response, the guys who sit on the floor expect only half that to show up. In this case, I think 60% of demand response showed up. If you are a drowning man, getting 60% to the surface will not cut it. [Laughter] We need some additional tools in the toolbox.
What was interesting is this was the first stage I emergency we have had in 10 years. As was indicated, it was cloud cover. It was a hot day. There were other factors. The odds of all that happening at once are remote. You design your electric system around those kinds of events. We have mentioned that we are going to have a solar eclipse on August 21, so please be riveted to the TV set to see how we handle it. [Laughter] It will be the Y2K of this generation. The ISO has hired a bunch of Mayan priests to help carry us through. [Laughter]
The point is that this is getting more and more complicated, and the good news is that you are seeing very creative alternatives to how to fix this, but in many respects it also creates an additional level of complexity in terms of trying to tie everything together.
MS. RANDOLPH: The eclipse is kind of an educational opportunity, right? We do not think the lights are going to go off, but it is a chance to get the attention of the people who don’t look at their electricity bills, the people who don’t look at the PUC website, the people who don’t think about their time-of-use rates. They start to hear a discussion about how the sun will go away for a little while. We have all these gas-fired assets that will ramp up. Maybe you folks who are concerned about greenhouse gas emissions need to think about turning your devices off for a couple of hours in the middle of the day. If you have a solar system with a backup battery, you will be fine for those couple of hours. It is an opportunity to have that conversation.
MS. KENNEDY: The economics become front and center during system planning. Utilities and the grid operators have to plan for peak usage during the entire 12-month period, so we have all these resources on spinning reserve that need to be available, not for the expected eclipse on August 21, but for the day when the heat is high and the cloud cover comes over. It is incredibly expensive redundancy we have to build into the system to do things the old way. Currently 30% to 40% of our daily load in the ISO control areas is solar and other renewables.
In an efficient market, you would have negative pricing in the middle of the day. But we are not seeing it. What we are seeing is increasing proposed real-time pricing in the middle of the day because of all those redundancies. That is a distorted market.
We started this panel by saying entities will respond to price signals. Send the price signals for what you want people to do, what you want entities to do.
The economics do not support storage today, so regulators have to figure out how to send the right price signals to incentivize storage to be installed so that customers have control over their electricity costs and their loads, and utilities have the ability to tap into that storage instead of paying to keep a peaking plant on reserve.
MS. RANDOLPH: One of the big questions is the point Ed Fenster made. What are the avoided costs that come from all these technologies, and how do you shift that spend from the distribution system to these other resources. That is a tough nut to crack. Trying to identify where there is the greatest value is challenging.
MR. SMUTNY-JONES: Let me add another level of complexity. The market signals are not there, but they are not there for the existing fleet either. I represent gas generators who basically produce about 44% of the power, 56% of the peak. People are beginning to shut plants down because there are no market signals to keep them around for that late afternoon ramp. Susan Kennedy is right. In the middle of the day, you do not need those plants because you have a lot of solar in the system, but between five and eight o’clock in the evening, you have these ramps that have to be met — some are as high as 13,000 megawatts — which is a lot of batteries to install in place of gas peakers.
MR. ALEXANDER: A lot of people say that energy storage is a form of virtual peaking power plant and maybe the day of the gas peaker has passed. Do you think we still need the gas peakers to support the high penetration from renewables or do you see gas peakers disappearing in the next five years?
MR. SMUTNY-JONES: Not in the next five years. Many of my member companies are in the storage space. One of them just did a storage that is at a peaker, so I think people are looking at storage as an opportunity.
The only problem I have with this storage discussion is sometimes it goes from practical business economics to magic. The solution to a problem is storage. The concern is we may not be doing things we know we have to do in the interim because of a blind faith that storage will solve the problems.
I am prepared for all of this. This is my backup battery. [Holds up a device, spurring laughter from the audience.] For those of you in the room who can’t see it, it is a unicorn. [Laughter] So this is how I am spending August 21. I am okay.
MR. ALEXANDER: Last question for Ed Fenster. Some reports predict declining growth or even flat-lining of residential solar. Where do you see the growth of residential solar across the United States, and in particular in California, for the next couple years?
MR. FENSTER: We think across the United States there are still five to 10 years of growth at 20% annual rates if what you are measuring is new installations year over year. There are a couple of forces at work underneath that. The first one is that we are starting to see states respond to the Paris withdrawal and be more supportive of solar energy generally. That is a tailwind. We have a reeducation process to go through in California because of the move to time-of-use rates, which is a little bit of a headwind. We think it is a good policy and we support it. We just have to get through it, and that takes time.
The other thing that is underlying the data is there were a couple companies in our industry that raised an enormous amount of money in the capital markets in the 2012 to 2014 time frame. They went wild without regard to unit economics and are now focused on cash flow and have really pulled their businesses back. These companies represented 40% of the market at one point in time.
The rooftop solar market in California is down year over year. But if you were to graph what the long-term 10-year growth rate looks like in the absence of these companies, you will see a pretty steady growth curve. That said, with the maturing market in California, I do not think we will see more customers going solar this year than last year, but I think the ultimate addressable market in California is probably four to five times what it is today in the fullness of time.