Modeling done by researchers from the Strategic Energy Analysis Center at the National Renewable Energy Laboratory (NREL), in Sunny with a Chance of Curtailment: Operating the US Grid with Very High Levels of Solar Photovoltaics, shows how the three main power grids in the United States might run on the highest penetration solar days (90%+ of demand being met by solar), when 55% of annual electricity use is met with solar photovoltaics. The hourly model also shows how much extra solar electricity we’d have to do something with, when we’d have it (springtime), and how it might affect (lower) pricing of competing electricity generation sources.
The modeling also looks at other, lesser, solar scenarios – but not us. The 55% scenario, titled “Low-Cost PV + Storage”, considers energy storage pricing at $97/kWh capacity and solar electricity at 2¢/kWh without incentives. There is 1,618 GW of solar PV and 346 GW of storage deployed.
The crux of the story is how do we keep the system running safely, and in a relatively economically sound manner, while also adding in sources of intermittent clean electricity. The authors suggest that high levels of control and coordination with other dispatchable resources and the energy storage would be needed, and solar module level controls would be greatly appreciated (others have modeled 100% of rooftops with solar power and thought the same). In fact, as the penetrations got the highest (top row on right below) – some of the sudden shifts would get pretty extreme. Modern lithium ion technology can instantaneously react, and can handle it – it will be required.
The greatest mitigation of these ramps will occur once energy storage is deployed in massive amounts in the Low Cost PV+Storage scenario, right row above and same in image above that, where you note a flat smooth output from fossil sources.
The economically challenging days – low electricity priced days driving by heavy volumes of generation – are almost entirely caused by high levels of solar power, with a heavy amount of this being in the spring periods of temperate clear days, and lower electricity demand. And while there are hours in which 41% of all solar electricity being generated (remember 1,618 GW installed) would potentially not be used, the high solar+storage scenario only saw that in 4.9% of all hours (1.1 hours/day), representing 6.6% of all electricity generated (below chart).
There is research that says if we design for solar curtailment, we can monetize it and deploy more solar due to this. Other documents see that oversized solar plus with power could meet 80% of our electricity with 12 hours of energy storage. NREL used a commercially available hourly modeling tool here to look at the whole of the United States and figure out what hours thirty years into the future might have cheaper electricity. Makes me think that 80% renewables is cake, and we ought let the extremists argue over the rest.