That volatility can be dealt with by distributing the renewables geographically. It’ll be sunny or windy somewhere. This requires overbuilding, which is currently cheaper that storage, and the average generation is highly predictable for day-ahead or hour-ahead trading. My rule of thumb is you can supply 70% of demand economically today without long distance transmission.

Search for “overbuilding renewables,” e.g. https://energycentral.com/c/em/overbuilding-curtailing-renew...

This still leaves room for hydro and nuclear.

The negative costs are borne by the owners of generating units that generate excess power when net demand falls. Solar can ramp down in milliseconds, windmills in minutes, CCGT can take hours, and nuclear takes days.

But what if we did add long distance transmission? Look at the US west of the Mississippi. I once estimated the capital cost of powering today’s demand using 100% renewables without storage.

That’s approximately $200bn in generation + $100bn for overcapacity.

The transmission system was mainly built to serve the legacy coal generators, but that’s often not where the best renewable resources are located. I estimated $40-80 billion for transmission.

That sounds like a lot, but 25% of the generation has already been spent. The cost of fuel for the gas generators, which is part of the LCOE, is on par with those numbers.

My biggest criticism of the Lazard study you mention is that it understates the cost of hedging fuel volatility costs.