April 22, 2016 by Roger Andrews at euanmearns.com

“Net metering” allows anyone with a solar installation to sell surplus solar power to the grid when the sun is shining and to purchase power back from the grid when it isn’t. Net metering has been described as the lifeblood of solar in America, and it’s probably true to state that without it there would be few, if any domestic rooftop solar installations anywhere in the country. However, the program is now coming under attack, with Hawaii and Nevada recently rolling back net metering benefits and with a number of other states also considering changes. What happens if enough states impose similar rollbacks, or maybe do away with net metering altogether? This post reviews this question and concludes that domestic solar in the US will slowly wither and die.

The Nevada decision

On December 23, 2015, the Nevada State Legislature passed Senate Bill 374, following which the state Public Utilities Commission cut the rate payable to owners of domestic solar installations who sell surplus power to Nevada Energy. The rationale was that intermittent solar power sold to the NV Energy grid “differs from” the dispatchable power the grid sells back and that domestic solar owners were getting paid too much for the former and not paying enough for the latter:

The order separates the prices of energy and related services provided by NV Energy, and the intermittent renewable energy provided to NV Energy by net metering customers. This approach is fair because it recognizes that the energy and suite of energy services provided by NV Energy to net metering customers differs from the intermittent excess energy delivered to NV Energy’s system.

This decision will be welcomed by all who recognize that solar is incapable of providing more than a small fraction of total electricity supply because of prohibitive storage requirements and that it’s presently getting a free ride on the back of grid generation that substitutes for storage. Certainly my rooftop solar panels would be totally uneconomic if I couldn’t use grid power at night and had to use storage batteries instead.

The Nevada solar industry, however, was not amused. Three solar companies – SolarCity, Sunrun and Vivint – announced they would have to cease operations in the state and local installers have been forced to cut staff. Also not amused were Nevada’s 18,000 existing rooftop solar array owners, who thought they were “grandfathered” but found that they weren’t. Their response was to launch a class action lawsuit against NV Energy alleging the utility “conspired to unlawfully reduce incentives” and NV Energy caved in, announcing that it would file a proposal to keep existing customers on the old rates, recognizing the desire for a “stable and predictable cost environment.”

“A potentially worrisome precedent”

But still the outcome in Nevada sets a potentially worrisome precedent for the US solar industry, with roughly half of all U.S. states currently studying or changing their net metering policies. States are taking action now because domestic solar in the US has grown so fast that several of them are now approaching or have already reached their net-metering caps. (A net metering cap is a target set by state authorities and it’s usually related to some fraction of peak demand or to capacity. But each state uses different criteria and some of them are extremely complicated. Details for anyone who might want more information are available here and here).

Two states other than Nevada have already revisited the question of how much intermittent solar power is really worth and how much of it their state can really use. The first was Hawaii, where some of Hawaii Electric Company’s grids were getting swamped by rooftop solar to the point where solar generation exceeded total demand at daytime solar peak. An example is given in Figure 1, which shows “backfeed” conditions between 10.30am and 2pm on August 8, 2013:

Figure 1: Average transformer load showing “backfeed” conditions, Hawaii utilities

Because of growing problems of this type the Hawaii Public Utilities Commission shut the net metering program down for new participants in October last year. As was the case in Nevada this shutdown was also accompanied by weeping, wailing and lawsuits from the local solar industry and rooftop solar owners, but the situation was obviously unsustainable. And it arose with less than 1% overall annual solar penetration in the state, not the 10% commonly assumed. More about this later.

Another state on a collision course with net metering is California, the home of the “Duck Curve”: (The Hawaii curve is known as the “Nessie Curve”, although the resemblance is less obvious.)

Figure 2: The California “Duck Curve”

At expected rates of solar growth California will also have a potential overgeneration problem by 2020, and the ramp rates needed to cover the period between about 5pm and peak load at 9pm reach potentially alarming levels. California’s solution has been to mandate the installation of 1.3GW of storage capacity (again no “h” given) by 2020, but this is just a drop in the bucket by California standards.

Current Status of the US solar industry:

One of the remarkable things about the US solar industry is how insignificant it is. Figure 3 plots percent solar penetration in the 36 states for which solar data are available (estimated as total solar generation divided by total generation using 2015 data from the EIA detailed state generation data base). The average level of penetration in 2015 was only 0.6%, and many states generated effectively no solar at all:

Figure 3: Solar generation by state as a percentage of total generation.

Only California is anywhere close to 10% solar penetration. Solar penetration in Nevada is less than 5% and in Hawaii less than 1%. (I checked this number and found that according to Hawaii Electric Company it’s correct). The implication is that solar may begin to stress grids at levels of penetration much lower than 10%, particularly at the local level.

Discussion

What we are seeing here is a conflict between on the one hand the utilities and grid operators, who view solar as a threat to their bottom line and to grid stability, and on the other the green lobby plus the residential owners, installers and PV panel salesmen who are now benefiting from the proceeds of subsidized solar and the existence of net metering. The surprising thing, however, is that this conflict has broken out even though solar still contributes a negligible percentage of the US generation mix. Why should this be? I think partly because the hundreds of thousands of homeowners who have installed solar arrays are dependent on a continuation of net metering to recoup their investment, partly because 200,000 people are now employed in the US solar industry, partly because solar can in some cases destabilize grids even at low levels of penetration (viz. Hawaii) and partly because of the claims made by some scientific organizations as to the percentage of US electricity generation solar could ultimately fill, such as:

• US National Renewable Energy Laboratory: 39% with rooftop solar PV alone
• Stanford University: 38% by 2050
• US Department of Energy: 27% by 2050
• International Energy Agency: 36% by 2050 (with solar thermal)

Numbers like this, which assume an approximate sixty-fold expansion of US solar capacity over present levels, can only be described as wishful thinking. Yet in the minds of many they are realistic targets.

But what happens if net metering benefits are rolled back? I picked an example which should be fairly close to reality – a household in Southern Nevada that consumes 11,000 kWh/year, the US average, with a 5kW solar array on the roof. I constructed a crude daily demand curve to show a peak around the breakfast hour and a larger one in the evening when everyone is at home watching large-screen TV or playing computer games and all the lights have been left on. Figure 4A shows hourly consumption and solar generation for the household during an average day (which assumes 12 hours of sunshine and a capacity factor of 19%, which is about right for Southern Nevada.) When the sun isn’t shining the household gets all its power from the grid, but for about 7 hours it gets all its power from the 3kW solar array. And over this period the array generates a healthy surplus that gets fed back to the grid, sending the electricity meter into reverse and causing it to wind rapidly backwards:

Figure 4: Demand, solar generation and consumption for a “typical” Southern Nevada household with net metering in place

Figure 4B shows the cumulative impacts. At the end of the day the household has consumed 30.3kWh, but because of the surplus solar power sent to the grid it gets charged for only 6.7kWh of grid power, which at current Nevada retail rates of $0.11/kWh works out to the princely sum of 74 cents, or an annual bill of about $270. Compared to what the bill would have been without solar (about $1,200) this gives the owner something like a ten-year payback on his or her solar investment after federal and state tax credits, which is not too bad when one considers that the solar array adds value to the house and that the PV panels will, one assumes, continue to generate electricity after payback is reached.

Nevada’s net metering rollback will, however, ultimately reduce the payment homeowners receive for solar electricity sent to the grid by 75% . How much difference will this make? Instead of saving almost $1,000/year on electricity bills the homeowner will now save only about $250/year. Even allowing for federal and state tax credits this will make domestic solar totally uneconomic in Nevada. And if other states follow Nevada’s lead it will eventually become uneconomic in those states as well.

And the problem doesn’t stop there. US utilities, with some justification, are also angling for increased charges to cover the costs of integrating growing amounts of solar power with their grids. (Nevada’s “grid connection charge” is scheduled to triple over the next five years). The end of the net metering road will of course be reached when the grids can’t physically accept any more solar, or no one will be able to afford the grid connection charge, whereupon Figure 4A will look like this:

Figure 5: Demand, solar generation and consumption for a “typical” Southern Nevada household with no net metering in place. The household is capable of powering itself for only about 8 hours.

Yet some believe that net metering rollbacks will provide a new opportunity for US solar. This article (which describes net metering as solar’s “junk food”) proposes a “value-of-solar tariff” where “solar customers are paid for the value of the electricity they produce at the specific time and place they put it on the grid.” This seems fair, but it too would probably kill rooftop solar. The California duck curve shown in Figure 2 shows how. The solar power produced in the middle of the day exceeds grid requirements and would therefore have to be sold at a low price if not wasted altogether, and at the nine o’clock peak, when power is in greatest demand, the sun has set or in in the process of setting. Another article views net metering rollbacks as an opportunity for domestic solar producers to go off-grid entirely and fill demand from energy storage, either in a utility-owned or domestic storage facility. But “to make the storage option appealing to customers … it would need to be offered using a low capital expenditures (CAPEX) business model.” “Energy storage” and “low CAPEX” are, however, mutually-exclusive terms, so that won’t work either.

It therefore appears that the future of domestic US solar depends on how far the states that are currently considering or reconsidering their positions roll back net metering benefits. And they probably wouldn’t have to roll them back very far before rooftop solar becomes uneconomic – unless of course the government jumps in with yet more subsidies. But hope springs eternal, particularly in the breast of the US solar industry.