3 ms·
If there's no sun, how long can the battery deliver? And if there's too much sun, where's the extra power dissipated? I'm wondering if it's possible to provide
by yodsanklai 3y ago
If there's no sun, how long can the battery deliver? And if there's too much sun, where's the extra power dissipated? I'm wondering if it's possible to provide a service equivalent to a nuclear powerplant today, and if so, how much it costs.
- _ph_ 3y agoThe thing is, the question doesn't really make sense. Yes you could design a closed system of solar production and battery storage, which delivers power 24/7. But that is not reasonable. Solar is only one part of the electricity supply to the grid. Wind is another. Add to that water, biomass, etc. and you have multiple sources which complement each other. You would need battery storage only for those remaining amounts of electricity which cannot be served by the combination above. And this amount is easily an order of magnitude (or more) lower than a solar+storage system alone. This also heavily depends on your local climate. The closer to the equator you get, the more a solar storage only needs to supply energy over the night. There a solar+storage only grid would be the no brainer.
- yodsanklai 3y agoThe question makes sense. When people claim that wind + solar wins over nuclear. They compare apple and orange. Nobody is able to explain what wind/solar (+ other renewable) installation is doable right night in Sweden (not near the equator), that would provide the same guarantee as a nuclear powerplant: 2GW of power, all year long. Or at least any time when the people need that power and there hasn't been sun/wind for a while.
- epistasis 3y agoThis is not specifically for Sweden, but if for some reason you want to imaging a single unit that produces continual output (which is not a desirable operation mode for any grid asset in the modern world), then imagine a solar installs distributed over a geographic areas with 18+ hours of storage at the desired power output. That's the overall design. People don't build that way but they could if they wanted. Right now, batteries deployed on the grid cost $300-$500/kWh, and typically LFP batteries are designed to last for 5000-7000 cycles, and will have warranties for 15+ years. This places the cost of storing a MWh at between $42/MWh and $100/MWh. Solar costs vary based on location, but range from $20/MWh to $80/MWh. For places with high seasonal variation, sizing the install for daily storage, using the seasonal minimum of input, in some areas might double or triple the cost for a particular season. Nuclear costs vary widely by construction competence, but in the largely incompetent west costs are going to be far north of $150/MWh. So for most areas, most of the time, solar+storage is going to be far cheaper than nuclear. But low insolation or high latitudes might make solar+storage more expensive than nuclear. In the future, nuclear's prices will continue to rise very slowly, and solar and storage prices will continue to plummet at ridiculously fast paces. This is why solar and storage are winning. For industrial processes that need heat, the storage options become even cheaper. And if the industrial site can be put somewhere with half-decent solar insolation, then the solar power can be directly connected without going through the grid, which means looping off that $60-100/MWh that utilities charge for the very very expensive grid transmission. We are just at the start of industry realizing what is possible with distributed renewables and thermal storage, and there is potential for drastically cheaper energy for energy intensive industry that be built at new sites. This is why solar and storage are "winning" even if nuclear might be better for Sweden. And with the ever increasing deployment of EVs, people are going to be parking multiple days worth of their home's energy needs in their garage. People will soon realize just how cheap storage is, and how perhaps even their vehicles will act as grid storage, even if it's only though charging at selective times when energy is cheaper.
- philipkglass 3y agoAnd if there's too much sun, where's the extra power dissipated? You can just open the circuit to disconnect a solar panel from the external load. The panel will get slightly warmer in full sunlight when it's not connected to a load, but there's no harm to it. Since there is no rotating mass or hot steam with accumulated energy in a solar farm, you can shut it down completely in seconds if necessary. Solar PV can safely go from 100% output to 0% faster than any other electricity source.