Taking stock of community solar
A panel of five community solar experts talked in late February in Boston about new trends in the community solar market, including evolving contract terms, consolidated utility billing, customer attrition rates, state requirements for a certain percentage of low- and moderate-income customers, where to probe on diligence when buying community solar projects, potential inflection points that would affect the future trajectory of the market, and other topics. The community solar industry resumed its annual conference with a large audience after a two-year interruption due to COVID. The following is an edited transcript.
The panelists are Laura Stern, co-CEO of Nautilus Solar, Richard Keiser, founder and CEO of Common Energy, Taymaz Jahani, chief operating officer of OYA Solar, Tom Matzzie, CEO of Clean Choice Energy, and Myles Fish, vice president of business development for Perch Energy. The moderator is Keith Martin with Norton Rose Fulbright in Washington.
MR. MARTIN: This is a challenging year with lots of headwinds. In addition to rising international political tensions with Russia invading Ukraine, we have tax law uncertainty, broken supply chains, inflation, Customs blockages of some solar panels due to forced labor concerns and a threat of anti-circumvention duties.
Laura Stern, are there any new trends this year that are unique to community solar?
MS. STERN: Most of the new trends are really a microcosm of what you see in utility and rooftop solar. One consequence of our success in community solar is that we are now facing many more interconnection issues that we have to address in order for the industry to reach the growth projections for which it is aiming.
MR. MARTIN: You can build the project. You can’t get the electricity to market.
MS. STERN: Or you can’t even build it because many community solar markets have regulatory cliff dates and deadlines. Everything from interconnection studies to actually tying into the grid at the end of construction has taken much longer than most developers anticipated.
MR. MARTIN: Richard Keiser, is interconnection the number one issue? Is there another new trend?
MR. KEISER: We are a subscriber management organization. One of the most important trends that will emerge this year relates to collections from consumers.
The periodic payments that subscribers make and bill credits they receive are collected on the developer’s behalf by subscriber management organizations. If you run a sensitivity analysis, the return of the developer is about 10 times more sensitive to the collection percentage than it is to the amount the developer pays a subscriber organization.
What is increasingly clear is that for most companies, the collection rates are in the 50% to 60% range, sometimes as high as 80%. And yet, when we are asked to bid on an RFP, we are never asked about our collection rates. We are only asked for customer acquisition costs and customer management costs, but those are irrelevant if you are unable to collect the money.
We are very focused on the collection percentage. I think that will become an emerging trend as people figure out that their returns are highly dependent on collection efficiency.
MR. MARTIN: Did I hear correctly that the collection percentage is only 50% to 80%?
MR. KEISER: Yes. Let me explain why it is so difficult to track. You start a period, which would be generation period X, let’s say January to February. Then the utility must calculate how much in bill credits to allocate to subscribers over the next period. That will take a few days. Then a few weeks later, the utility will tell you the number of bill credits that were distributed. Then you start collecting from subscribers. Three cycles end up being mixed together.
If Laura were to look at her bank account at the end of the 30- to 45-day bill cycle for customer payments and see a large pool of money, she might not know whether it means there was a 99% or only a 40% collections rate for the first cycle. When the auditor connects all of the dots, she may be disappointed with the results. We are very focused on trying to push collections rates into the 90% range.
MR. MARTIN: So there is potential for improvement, but any financier should discount the revenue stream?
MR. KEISER: Correct.
MR. MARTIN: TJ, what is a new trend for community solar?
MR. JAHANI: It is taking longer to develop projects.
MR. MARTIN: Because of supply chain difficulties?
MR. JAHANI: The design and engineering costs are going up. Developers are learning from earlier projects. The supply chain is also a challenge.
MR. MARTIN: Bankers have told us that 20% to 30% of projects that were supposed to fund at the end of last year flipped into 2022, and they are already seeing delays into 2023. Is that your experience as well?
MR. JAHANI: Yes, but we are also seeing manufacturers decide to ship their products to other markets. Solar installations are increasing in lots of other places besides North America. If prices are higher elsewhere, they will turn their ships around and send the panels to another country. That has been very challenging for us.
MR. MARTIN: Is this a US problem or do you have the same problem in Canada?
MR. JAHANI: It is more of a US problem. Manufacturers tend to have separate allocations for Canada. It may be easier to get panels in Canada.
MR. MARTIN: Tom Matzzie, what is a new trend for community solar?
MR. MATZZIE: State community solar programs are more likely than the last time we had this conference in 2019 to have a low- and moderate-income component. This has become a bigger part of our industry.
We are in the high 90% range in our collections. It would be unusual to see collections rate in the 50% to 80% range, in my view. However, the LMI component will introduce a collections challenge, and so the industry needs to innovate.
I think we are going to see more pre-pay models. If you walk into an Apple store, every consumer is qualified to buy an iPad and get a cellular subscription because it is an entirely prepaid product.
MR. MARTIN: Pre-pay over what time period? Five years? A year?
MR. MATZZIE: Pre-pay at the beginning of each month so that you remove the receivables risk. You still have a contract risk that the consumer could default.
Apple has democratized access to cellular service by having essentially an entirely pre-pay business model. I think you are going to see adoption of similar models. In the retail electricity sector in Texas, all the credit-disabled customers are on pre-pay products.
MR. MARTIN: Why is that considered an innovation? I remember living in London in the late 1970s, and to get hot water for a bath, you had to keep feeding five-pence coins into the gas meter. This is going backwards.
MR. MATZZIE: It is a way to allocate credit risk to the actual places where there is credit risk. It is a more data-driven approach to risk allocation rather than moving backwards, and what it will do is credit-enable more consumers. You will still have contract default risk, but you will address receivables risk.
MR. MARTIN: So it is a way to get around the need for FICO scores?
MR. MATZZIE: It depends on every counterparty’s view of what risk the FICO score is addressing. Is it addressing default risk on the contract, or is it addressing receivables risk? There is evidence that people with lower FICO scores do not have the highest level of payment defaults, but you still have contract default risk.
MR. MARTIN: Myles Fish, you gave a compelling presentation immediately before this panel about the need for collective action to expand the market by getting more states to sign on to community solar programs. What new trend would you add to the list you just heard?
MR. FISH: I think that is part of the new trends. There is an opportunity to expand the potential addressable market.
More and more states are considering community solar, but what that also means for us as market participants is we have more states to monitor and to think about where to invest. It is encouraging to see new states considering this, but it is also then incumbent on us to weed through new opportunities to avoid speculative investments.
MR. MARTIN: Laura Stern, going back to you. Talk to us about the basic business model. The community solar company builds a small solar facility and supplies the electricity to the local utility in exchange for bill credits that can be used against electricity bills. It basically sells those bill credits to local businesses and residents whom it signs up as subscribers. How does the money flow through that circle?
MS. STERN: It depends on the utility and the state, so different states have different programs. Consolidated billing is an important advance in terms of trying to facilitate not only the flow of payments, but also the level of understanding among our customers.
In places without consolidated billing, the customers receive two bills: one from the local utility for the electricity they use and one from the community solar company. The utility bill credits the bill credits.
With consolidated billing, there is one bill, and the subscriber payments on which the community solar company relies are collected in the first instance by the utility.
MR. MARTIN: So the customer makes a payment to the utility on a consolidated bill, and the utility splits it. What percentage does the community solar company receive? Tom Matzzie.
MR. MATZZIE: Our retail electricity business operates under utility consolidated billing in 34 markets, and there is a cash flow waterfall. Utility consolidated billing only works if there is also a purchase of receivables by the utility. The utility applies a discount rate of 1% or 2%. It owns the receivables risk. Then it just pays you.
If it does not purchase the receivables — and we operate in a few of those markets where there is utility consolidated billing with no purchase of receivables — there is a cash flow waterfall. It is a little bit of a nightmare to be honest with you. The utility always gets paid first, so if there is a shortfall in customer payments, the utility gets its share first and you get what is left. Then the next bill comes in and there is another customer shortfall. The utility pays itself first and pays you what is left of your prior month first and then your second month. The cash flow waterfall becomes a really important thing.
MR. MARTIN: So 34 states with consolidated billing?
MR. MATZZIE: Thirty-four markets.
MR. MARTIN: In 34 markets, the utility buys the right the solar company has to subscription payments.
MR. MATZZIE: It purchases the receivable.
MR. MARTIN: It discounts the revenue stream by 1% to 2%?
MR. MATZZIE: I have seen anywhere from 0% to 3.5%. In some markets, like in Pennsylvania, the utilities have claw-backs for some retailers that have really bad-performing receivables rates, but it is 0% in places like Maryland because the utilities there are actually collecting more in late fees than they were on defaults on receivables.
MR. MARTIN: If everything works as planned, what percentage of the revenue goes to the community solar company?
MR. MATZZIE: It should be the full revenue minus the discount, so if it is a 1% discount, then 99% of the revenue goes to the community solar company.
MR. MARTIN: Richard Keiser, do those numbers square with you?
MR. KEISER: Just to be clear, Tom is talking about something very different. He runs a retail electricity business. What percentage of your customers are retail electricity customers versus community solar companies?
MR. MATZZIE: Twenty times more.
MR. KEISER: I think the point that Laura was trying to make was community solar consolidated billing does not always function exactly that way, and the models vary by utility.
MR. MATZZIE: I was explaining what you should expect after utility consolidated billing gets rolled out. It can be a great tool, but you need to get the purchase of receivables in there. Otherwise, it becomes a cash-flow-waterfall nightmare.
MR. MARTIN: Myles, you touted consolidated billing in a presentation immediately before this panel. Has it worked as you hoped?
MR. FISH: I think the model for the community solar industry is New York. Focusing on the cash waterfall, the revenue comes from the utility directly to the owner of the solar project at a discount that is communicated to the customer ahead of time. The customer will be allowed bill credits equal to the amount it pays, less the discount.
The alternative, which you can still do in New York if you prefer, is the customer pays the owner of the project directly, and the full payments translate into bill credits for the customer. The customer receives a separate bill from the utility for the electricity usage, and it applies the bill credits toward that bill.
Consolidated billing is a lot more streamlined.
MR. MARTIN: Laura Stern, when you don’t have consolidated billing, what is the split of revenue? The community solar company receives revenue solely from the subscriber?
MS. STERN: Yes.
MR. MARTIN: The utility has to receive some revenue for actually supplying the electricity. What is the split?
MR. KEISER: Let me break that down for you.
Let’s assume that the customer is buying $200 worth of electricity currently before community solar comes into the picture. If we were partnering with Laura, we would formulate an allocation to that customer so that it would receive, let’s say, a $180 credit on the bill. Now the net amount owed to the utility is $20. If the community solar company is offering customers a 10% discount from the retail electricity rate, the customer will keep $18 of that $180 credit, which would leave $162 for the community solar company. Our job in that scenario would be to make sure that Laura gets paid her $162.
What Myles was explaining was that if there is utility consolidated billing in the same format that New York has developed, then the utility just pays Laura $162 all the time with no question. She loves that model. But there are other models for retail electricity suppliers that Tom was explaining that work slightly differently, and that is why he is so focused on getting the purchase of receivables. Does that unify the responses?
MR. MARTIN: Yes, I think so. Thank you for that.
TJ, you heard in the example Richard Keiser just gave that a 10% discount was offered to the customer against the retail electricity rate to get the customer to subscribe. Is that where discounts are today generally in the market?
MR. JAHANI: That is in line with what we are seeing today. The discount that you could offer the subscriber depends on the economics of the project. As the cost of interconnection increases, as EPC costs increase, there is less ability to offer a 10% or higher discount.
MR. MARTIN: Myles, you are out soliciting business. Where do you see discounts currently?
MR. FISH: It depends on the market. In some states like Illinois, discounts tended to be a little higher in past years because of the way bill credits worked in that state. In most markets, 10% seems to be most common.
MR. MARTIN: Our last conference was in July 2019 in Philadelphia before COVID hit. It was standing room only. At that point, it seemed like most community solar companies, with the exception of Nexamp here in Boston, were having to enter into 20-year contracts with commercial customers, and they were aspiring to get down to five to 10 years with residential customers.
Richard Keiser, you sent me an email last night that said this is no longer true. Where do you think contract terms are today, and how relevant are FICO scores for residential customers?
MR. KEISER: In the vast majority of markets where we work, residential subscribers are being offered a one-year, auto-renew contract, but the subscriber can cancel at any time or with notice. It is a perilous endeavor to try to collect money from consumers who do not want your service any more.
With commercial subscribers, the contract term varies based on the financing terms that the community solar company has behind it. If you have sophisticated financiers like Laura does, then they might be more comfortable with more flexible terms. If you have a bank or tax equity investor who is new to the market, it might insist on a 20-year contract at least for the anchor customer that has some teeth to it if the anchor wants to cancel. There is more variety around the commercial agreements.
MR. MARTIN: What contract length do you think is standard for commercial?
MR. KEISER: We typically do 20-year agreements. Really the most important thing is not the term length, but the termination provisions.
MR. MARTIN: How easy it is get out. How do the termination provisions work?
MR. KEISER: Some such contracts have significant penalties to terminate early. We try to be more flexible and ask for one year’s notice, and then we ask for a termination fee for any period less than one year with the logic being that we will be able to replace the customer within that time period since the customer is being offered a significant discount from the price it would otherwise have to pay for electricity.
MR. MARTIN: So the commercial customer can walk at any time, but has to give you one year notice or pay a fee. Myles, where do you think contract terms are, and when are FICO scores required?
MR. FISH: It varies. It depends on risk tolerance of the financing party and the market. For example, Maine is offtake constrained and customers have to be offered larger discounts and more flexible terms to sign up. Illinois is less so.
We try to be flexible. Not every client prefers the same approach. We tend to be driven by any preferences the customer or financier has.
MR. MARTIN: What terms are required in the current market to be able to raise financing?
MR. FISH: We try not to get into a conversation where we are convincing customers to do one thing or another. We listen to what their preferences are and we can give advice on what the market might bear, but if their preferences are in line with what we think the market can bear, those are the terms that we will deliver.
MR. MARTIN: TJ, are financiers insisting on long contract terms and residential FICO scores?
MR. JAHANI: FICO scores were very important five years ago. They are less so today. We have managed to explain to our financiers that eliminating FICO scores lets us reduce our customer acquisition costs by going after bigger market segments. Over time, if a customer leaves, we are able more easily to replace that customer with a newer customer.
We are mostly active in New York, so with full consolidated billing, it is a no-brainer for us to dispense with FICO scores and go after bigger market segments.
MR. MARTIN: Laura Stern, Richard Keiser teed you up. You are dealing with sophisticated financiers. Are you able to have residential customers who can walk at any time and commercial customers who can walk with one year notice?
MS. STERN: Yes. The banks have come to accept consumer-friendly contract terms. This trend towards flexibility for the customer is not just something to which we have been driven by market forces; we are very comfortable with it as well. Flexible contract terms ultimately lead to reduced acquisition costs, increased customer engagement and higher retention. What we focus on more than FICO scores and contract length are the mechanisms that each state creates to mitigate the financial effects of churn, particularly features like the ability to bank bill credits and to flex up customers’ allocations of electricity output of a given project.
These program structures can be equally or more important than the credit score of any customer. It is also important to distinguish between default and churn and to implement internal policies that prevent customers from lingering with an aging balance for several billing cycles. Customer acquisition is very expensive. Our goal is to minimize churn and defaults through active customer engagement, education and communication.
MR. MARTIN: When you talk about policies, are you talking about the mix of residential and commercial customers, low and moderate income customers, or something else?
MS. STERN: No, I was referring to the mechanisms that the regulatory authorities have established to help mitigate the revenue impact of customer churn and defaults, such as consolidated billing. Some states allow the community solar generator to reallocate credits to other customers or to size one customer’s share of the electricity output at a much lower percentage of the customer’s overall usage so that when other customers terminate or default, the generator can reallocate the bill credits.
MR. MARTIN: These are state policies to help project developers mitigate the effects of customer credit problems and churn.
MS. STERN: Exactly.
MR. MATZZIE: It will be interesting to see what happens with churn once you have utility consolidated billing because the community solar provider will not have as much of a relationship with the customer as it does today. The churn rate in our retail electricity business, which is a utility consolidated billing business, is double or more what it is in our community solar business where we are billing customers directly and have a direct relationship with customers.
Why would the churn rate be so much higher with utility consolidated billing? You do not have as much of a relationship with the customer.
MR. MARTIN: Richard Keiser, how do you get comfortable in a market where customers are free to walk away from contracts that a new entrant will not come take your customers by offering a larger discount?
MR. KEISER: This is one of the misconceptions about community solar.
On the one hand, you need consumer-friendly, consumer-facing policies, starting with a website that is accessible to consumers. That is easy to build. Any high school student can build a web form that enables customers to click through and find a certain amount of information, so that is not the barrier to entry. The barrier to entry is where you have thousands of subscribers on 50 to 100 different projects. All of those subscribers have different usage, different credit rates, different discount rates, and all of that needs to be distributed into 50 different project bank accounts.
MR. MARTIN: Would you say that of the four solar market segments — residential rooftop, C&I, community solar and utility-scale solar — the barrier to entry is highest in community solar, or how would you rank the four segments?
MR. KEISER: It is a good question. There are different barriers in each segment, including ability to get access to sophisticated capital like tax equity. I don’t think I would be able to rank them.
MR. MARTIN: TJ, many states are requiring a certain percentage of the output go to low and moderate income customers. What is the range of percentages, and what challenges does that pose in trying to finance projects?
MR. JAHANI: Contrary to what we feared, our financiers are actually excited about the low- to moderate-income customers. It helps their environmental sustainability goals. We see goals of around 20%.
MR. MARTIN: 20%? Myles, what are you seeing?
MR. FISCH: In Maryland, it is 30%. In Massachusetts, it is 50%. In New York, you can pick your own flavor and there is a range. I spoke about enabling more state policies to help our market grow. In general, I think we should expose ourselves to those higher percentages when we are given the option to choose.
MR. MARTIN: Tom Matzzie, is your idea of making people pay at the start of the month the way to finance LMI revenue streams?
MR. MATZZIE: It certainly could help deal with receivables risk. You would still have contract default risk.
We are seeing the same LMI percentages. One thing that is important to understand is not all LMI consumers are credit disabled. There are LMI consumers who have prime credit scores and are FICO qualified. Part of the opportunity is to connect with those consumers. They are people with stable jobs with health care. For example, if you work for a school district and drive a bus, you are probably credit enabled because you have good benefits, even though you might not have a lot of money. You may qualify for an LMI program.
MS. STERN: Tom is absolutely correct. The challenge is verifying customers and the state requirements to verify customers. It can be incredibly challenging. I think our collective job is to work with the regulators to come up with a smoother, easier, more efficient way to verify LMI customers.
MR. MARTIN: Is any state doing it right currently?
MS. STERN: It is a challenge everywhere, especially in states like New Jersey that are starting new programs.
MR. FISH: The best practice is geographic eligibility, which avoids requiring someone, as he or she is subscribing, to upload a document to verify income. Any time you must scan a document to include in some sort of enrollment process, it creates barriers to enrollment. Geographic eligibility allows us to collect the customer addresses as part of the enrollment process, run the addresses against a database and treat the person as LMI or not based on the neighborhood in which the person lives.
MR. MARTIN: Richard Keiser, early financial models assumed that customer attrition would be about 5% a year, probably highest in the first year when the first bills start to be received. What do you think is the right percentage now that the industry has more experience?
MR. KEISER: That’s a great question. One way to think about it is US mobility is about one move every seven years, so you would expect natural housing churn of around 14%. Fortunately, the vast majority of people upgrade their houses when they move. If you assume that 70% of that churn is within a local area, and therefore 30% is true churn out of the system, 30% of 14% is 4.2%. That is not a bad natural model to assume.
This, like many of the other things that have been discussed on this panel, is sensitive to other factors. One factor is the developer’s timeline for building the project. Say you have a project that has been advertised as expected to go live in the spring 2022. I will tell you a phrase that has never been said to me: “Hey Richard, great news. We finished the project three months early.” [Laughter]
This has never happened, right? So if the completion date gets pushed out two years . . .
MR. MARTIN: The customers won’t wait around.
MR. KEISER: Correct. It’s like if you try for a second date after not contacting the person for two years. It’s not going to work.
MR. MARTIN: TJ, suppose you are out in the market looking to buy community solar projects under development from other developers. Where do you probe first on diligence?
MR. JOHANI: I would probe first into whether the project has a feasible path to interconnect to the grid. We look at curtailment. We also look at the site control documents to make sure that we do in fact have site control, the proper easements and the tile is clean and financeable. We also check whether the project has a clear path to permitting.
We start with desktop diligence and, if we like the project, we dig deeper.
MR. MARTIN: Laura Stern, where would you probe?
MS. STERN: I agree with what TJ said. Physically, community solar is just like any other solar project. Our major gating milestone is the allocation for the state’s community solar program. The allocation in the community solar program is equivalent to securing a PPA.
MR. MARTIN: Are there any audience questions?
MR. WILLCHENE: Sean WillChene, CEO of Shared Solar Advisors. Illinois just adopted consolidated billing. When do you expect that to go live?
MR. MARTIN: Anybody know the answer? [Pause] We may not have an answer.
MR. KEISER: I would expect two to three years.
MR. MICHELMAN: Tom Michelman, senior director, Sustainable Energy advantage. How do you see growth, outside of Texas, of community choice aggregation combined with community solar? We have been waiting a long time for it in Massachusetts. It looks like it is about to happen in New Hampshire.
MR. KEISER: We have only seen it in New York so far.
MR. FISH: It is only in New York today, and that is just a pilot program. New York has said it wants to be careful about the long-term implementation of opt-out CCA integration with community solar. In general, this is a positive thing for the industry because it is a streamlined way to get a lot of customers to participate in community solar. CCAs are municipalities. Residents in their areas are automatically enlisted unless they opt out.
MR. FELT: Justin Felt, director of policy analysis for Baltimore Gas & Electric. Do residential and commercial subscribers sign something that looks like a power purchase agreement where they pay a per KWh charge tied to a monthly meter reading?
MR. MATZZIE: Early on there were more esoteric contract structures. We have some customers who pre-pay each month on assets that we took over from other asset owners, and other customers have end-of-year reconciliation back to a credit rate. Nowadays, it is a discount to the bill credit rate. The esoteric kinds of contracts are mostly gone.
MR. MARTIN: Here is my last question. Investors look for inflection points in any market: things that could change the market trajectory. What should we be looking for in the next two years in community solar as possible inflection points?
MS. STERN: Our inflection point needs to be a real breakout of the installed capacity of each state’s program. We need more meaningful progress than establishing pilot programs in a few new states every year. We need to expand programs in the states that we are already in as well as promote larger programs in new markets. The programs are just too small. One gigawatt a year for all of us in this room is just not enough. An inflection point would be getting to 30 to 40 cumulative gigawatts of installed capacity.
MR. MARTIN: That is where the Coalition for Community Solar Access, the organizers of this conference, play an important role. Richard Keiser, inflection point?
MR. KEISER: I agree with that. The inflection point would be a lot more states with depth, like 400- to 500-megawatt programs.
MR. MARTIN: TJ, do you have another inflection point?
MR. JOHANI: I agree with the comments made. There needs to be a push from the federal government to streamline the processes, and perhaps to allow battery storage to be incorporated into projects to allow more penetration on the grid.
MR. MARTIN: Tom Matzzie?
MR. MATZZIE: An inflection point implies a much steeper rate of change. State programs tend to be incremental, and so I think what is required is a fundamental change in the business model to get to the type of scaling where the industry is adding tens of gigawatts. So what would that mean? It would probably be a more organized power market rather than one-off distributed energy resources.
MR. MARTIN: You are returning to your retail electricity supplier role.
MR. MATZZIE: Yes, but the thing about that model though is that there is no capacity limit. I can sell as much as I can find customers to buy. That is where you want to get eventually.
MR. MARTIN: Myles Fish, you get the last word. Inflection point?
MR. FISCH: I agree with what others have said. California would really change the addressable market in one fell swoop.