[exelius]

Actually; cheap storage makes the argument for nuclear energy even better. Nuclear is very bad at responding to demand because it can take weeks or months to restart a nuclear reactor.

Under any scenario, you're going to need some kind of power grid. That grid is going to have a non-zero "base" load as batteries charge, etc. Nuclear is perfect for generating this base load - it's reliable, clean and not subject to the variability that solar and wind are. Don't get me wrong, solar and wind are going to be a big part of the grid, but you want diversification of technology so that if you have a cloudy, but not very windy day, the grid can still keep up.

Also, fission reactors are being miniaturized, and at the timescale we're talking about, fusion may be an option as well. I don't think nuclear will be 50% of energy generation or anything, but it's a good, reliable technology that can supply a consistent amount of energy in any weather conditions.

[afarrell]

The problem for nuclear is that with the way the bulk electricity market works, shaving off the peaks could hurt the ROI of building a new nuclear plant significantly. In New York and New England, suppliers bid in how much power they are willing to supply at a given price, the market operator runs a giant LP solver over the bids and transmission constraints between locations in order to satisfy demand in each location, every supplier that is scheduled in a given location gets paid the price of the marginal watt in that location. So nuclear plants bid negative to make sure they get scheduled and rely on other positive-price sources getting scheduled in order to turn a profit.

[exelius]

Those economics change drastically when cheap, mass storage of electricity is available. Spend all day charging your batteries with solar and wind, then discharge them in the evening. You could afford to underbid conventional fuels with renewables at similar margins and still negative-price nuclear. But that's the great thing about capitalism: it adapts to small efficiency gains in commodities production very quickly.

[ThrustVectoring]

Storage shaves off price valleys, as well, by adding demand in off-peak hours. Every kilowatt-hour that's low enough under the peak price will convince someone to buy a battery to time-shift the power.

My model is that cheap batteries would drive on-demand power stations out, in favor of batteries + base load.

[ridgeguy]

My understanding is that nuclear power is used to serve base load, which is the demand that isn't subject to significant variation. This allows generator owners to run nuclear generators at their cost-effective maximum output.

It's true that there is great variation in electricity pricing, with peak shaving and other so-called "ancillary services" bringing as much as 100X as high a price as base rate (for a short period of time). Nuke generators aren't capable of rapid load following, reactive power correction, voltage support, or other fast-response services that bring premium power prices.

[nl]

The issue with base load contracts is that they are increasingly being undermined by power that is much cheaper, and reliable almost enough to substitute for base power.

Generally speaking most jurisdictions would have to be stupid to sign up to new long term (20 year+) base power contracts, but that's what nuclear power needs to pat for the high upfront costs.

As this article shows, new storage tech and load management tech makes it likely that base load will be less and less important over time.

[exelius]

Yeah, but the long-term outlook for fossil fuels is still rising prices. If given the choice between building a new LNG plant or a new nuclear plant, are LNG fuel prices low enough and stable enough over the long term to justify over a 20 year+ contract? This is an honest question and I'm sure the energy industry has a legion of forecasters and quant jocks on it, but I personally don't know how the economics work out.

There's also the carbon footprint aspect to consider. If the "carbon tax" is successful, nuclear could be a more cost-effective option than hydrocarbon-fired plants in areas without a more stable "natural" power source like hydroelectric or geothermal. The petrochemical industry likely has enough political pull to neuter any carbon tax law, but in theory that's how it should work.

[ghshephard]

That's my understanding as well. Nuclear Power plants are base load, and are contracted to do such, with long term guaranteed capacity. Peaker Plants take on load that exceeds the base, are paid on the margin, but aren't guaranteed any particular load.

[beat]

Right. That's the correct use for nuclear in a solar/wind + storage future - a stable and reliable base load.