Bigger Solar in Minnesota
How much can the CSG model be scaled up? (Part 3)
Author: Ralph Jacobson
What have we accomplished with the CSG model?
By the end of 2019, about 650 MW of CSGs had been installed and brought online, with another 200 MW under development. For comparison, Xcel is proposing to bring 4,000 MW of solar online by about 2035 and would prefer that most of it not be under the Community Solar Garden (CSG) model. Given that the CSG model appears as a place holder for larger scale solar in the Solar Potential Analysis report, as mentioned in Part 1, we need to consider the issues that exist now in that market segment.
Then we should ask whether it makes sense to stretch the CSG model into duty for a much bigger role than it was designed for, or do we set a course to develop other models for larger solar? One thing to keep in mind is how much uncertainty attaches to the CSG model, and how likely this is to hinder the flow of capital into vigorous deployment of larger solar in the crucial decade ahead.
Moving a half-step away from the net-metering paradigm
The CSG model was the first opportunity in Minnesota to develop larger-scale solar. There was no size cap per Garden site, and no overall planning target, until the regulators agreed to end the practice of collocating one-megawatt Gardens, which effectively limited the size to one-megawatt per site. Through this program, the local solar industry, Excel Energy, and the many affected communities have become familiar with the complex of land-use, engineering, financing, permitting, and other issues which pertain to larger-scale solar. We have all gone up a steep collective learning curve!
Although it was intended to be a step away from the “early market” net-metering framework, addressing the issue of cross-subsidization, some aspects of that paradigm have followed into the CSG model for Xcel Energy. First, it creates an open market where solar developers can originate and propose projects, instead of responding only to utility RFPs. As Gardens become more numerous and/or larger, Xcel will experience more uncertainty in its ability to plan for the timing, location, and financing of new generation.
Second, the regulators set subscriber rates close to retail-level, using the Value of Solar (VOS) tariff for the first time in the state, as set forth in the 2013 Minnesota Solar Jobs Act, to enable the financing of one megawatt CSG installations. The intent of the VOS had originally been to address the tendency of net-metering programs to “cross-subsidize” one class of ratepayers at the expense of other classes and non-participating customers – to democratize solar by spreading the benefits over all rate classes. But has the ability to finance Gardens under the VOS adequately addressed the democratization issue? Most CSG developers find it easier and more profitable to find subscribers among municipalities, schools, churches, and corporations than among the multitude of residential customers. Although the issue is still unresolved, Minnesota is now host, over Xcel’s protests, to the most vigorous CSG market in the nation. More about the VOS further on.
Uneven access to the CSG model
The language to enable the CSG model, under the 2013 legislation, was aimed only at Xcel Energy. While other utilities in the state have experimented with CSGs, such Gardens are only built as a service to those members who wish to participate. This will not support a market the way that the Xcel program does. Even in Xcel’s service territory, because of the requirement that subscribers reside in the same or an adjoining county as the CSG, farms which are located on slivers of Xcel territory in very rural counties of Minnesota do not qualify to host Gardens because there are not enough potential subscribers who qualify under that rule. Again, this seems to undermine the original intent of the program for broader access to solar.
Repeal of the contiguous county requirement would correct this uneven access with a “two-for” benefit: not only would those farmers be able to host gardens, but inner-city residents would have broader access to those solar gardens. This is because most of the CSG activity now is in a “doughnut” of counties which ring the metro area but are two counties away from residents of Hennepin and Ramsey Counties. Such a rural-urban connection could be constructed to help the program better meet the original intent.
Impact on utility business model
Unlike net-metering, where most of the power is utilized onsite, the CSG delivers all of its power to the utility as a “merchant system.” Being on the large end of the net-metering framework, the CSG has more impact on the utility electrically and on the “wholesale to retail” business model. Without a better coordination regime to address hosting capacity and congestion, the utility loses control of the ability to target new generation siting in its planning, because solar developers are free to originate projects wherever they can. Furthermore, gardens capped at 1 MW are difficult to finance beyond a short distance from substations; this limits the number of good sites for solar gardens, clusters them near substations, and keeps the cost to build relatively high.
Two very different market structures
The CSG is a policy-driven market structure which the utility has had to accept from its regulators because, through the subscriber arrangement, consumers (ie: unsophisticated investors) are involved in the transactions. This complicates the model because it is illegal to take money from unsophisticated investors, and so capital has to be provided by third parties who expect to get a much quicker return on the investment than the utility is allowed to get on its investments. Furthermore, the cost of that power to the utility is set by a stakeholder process which the utility does not control; it is currently starts at just over 9¢ per kwh.
By contrast, utility-scale solar transactions are almost entirely between the utility and its solar developer vendors. The wholesale market is more economics-driven, which aligns better with an updated utility business model where, unlike the richer context of the VOS approach, the goal is to generate electricity from solar arrays in the same cost range as from burning coal. This parity has already been achieved in some places, where the wholesale price of electric power from utility-scale solar arrays is under 3¢ per kwh (kilowatt-hour).
Potential improvements to the CSG model
If claims made by us, the solar developers, at the legislature in 2019 are to be believed, the best sites for development of CSGs have already been utilized, and further sites are more expensive to develop (that was part of our argument). Adjusting the one-megawatt collocated cap on CSGs to two or three megawatts would somewhat alleviate this constraint and expand the range of sites for financeable installations; it would also somewhat lower the cost to build new Gardens.
Allowing Xcel customers from non-contiguous counties to subscribe would fill out the CSG market in very rural areas and very urban areas of Xcel territory. Xcel representatives consistently maintain that they originally believed that subscribers would mainly be residential customers. The “residential adder” incentive, which increases the subscription rate as more residential participants are included, will align the program better with the original intent. It will attract some developers to expand into cities and towns with urban rooftop CSGs. Further, having more residential subscribers could be made a requirement for allowing larger Gardens or subscriptions from non-contiguous counties.
The CSG program needs a cap for planning purposes
The trade-off is that any deal will most certainly require a cap on how many megawatts of solar gardens which Xcel will be required to greenlight each year. There is no agreement about that yet, but a realistic cap would be a useful planning tool for everyone’s purposes. The tug-of-war of the CSG pricing and rule-making process is a tough environment for utility planning. Recognizing that the utility needs a cap for good planning, the cap should be high enough to allow the solar developers to continue to build the program out over the next 5 – 10 years to an agreed-upon overall goal of perhaps 1,500 total MW of CSGs.
How will the VOS perform at higher solar deployment?
The VOS tariff will need a different baseline as more solar comes online. The VOS is based on a thorough methodology for valuing solar in terms that are relevant to utility economics. However, the basis of the valuation is a comparison of the cost of the solar build-out, against the cost of distribution system upgrades which may be deferred or eliminated because of the solar. But at higher levels of penetration of solar into the electric power market, perhaps associated with grid modernization and microgrid development, there will be new costs for system upgrades needed in order to accommodate more solar and other DERs (distributed energy resources), and deferral-based valuation must give way to valuation based on new build-out.
Would the VOS then become obsolete?
Consider that the genesis of this approach was the desire to move past a frustrating debate about the catch-all of “externalities” which could be mitigated by solar energy. By creating an itemized list of market-oriented benefits of solar to the utility and its customers, the VOS is a rational method to price solar in an already complex electric power market. Far being discarded, the VOS could be a foundation for the even more complex challenge of valuing DERs in a more interactive and time-oriented power grid, requiring optimization between solar, energy storage, curtailment, gas peakers, and load management. Let’s celebrate what the VOS methodology given us that can be applied going forward !!
A limited growth forecast for the CSG model
Here is a recap of the issues I have with the Xcel CSG program:
– CSG market only functions in Xcel territory
– Priced too high to be a major power source
– Doesn’t address utility planning needs well
– Hasn’t effectively addressed ratepayer cross-subsidization
– VOS tariff will need update for valuation of DERs
Here are some of the suggested improvements to the program :
– Remove the contiguous county requirement under conditions minimizing cross-subsidization
– Set a yearly cap for the Xcel CSG program
– Raise the Garden collocated size to at least 2 megawatts
– Firm up incentives for more residential subscriptions
We have a strong Community Solar Garden program right now, viewed by many as the best in the country. Although Xcel chafes under some aspects of the program, and the solar developers chafe about others, the process has brought us all into a better understanding. We will need that going forward. Rather than watch the program fizzle out over the next couple of years, a better strategy would be to make all of the improvements to the program that we can, including agreement on the ultimate scope, and keep the industry healthy by letting the program make a strong finish at 1,500 megawatts of Gardens by 2030. That’s about 10% of the solar we will need by 2050, as mentioned back in Part 1 of this article. And then let’s get to work on the DG solar models for the other 20% to 50% of the solar that we will need.
Next: The VOS approach for valuing DERs