Renewable energy is intermittent. Solar is in the day time, wind is available when it is available. But, one fine sunny cool breezy day our renewable capacity is going to meet total demand. What is a regulator to do? The renewables will be too dispersed to tell them to shutdown. All the rapid response generating capacity will already be shutdown because the renewables are free and who wants to compete with that. Hydro is in the middle of a mandated water allocation flow.
So, rather than blow the grid, the regulators will call the nuclear plant and tell it to go off line. But to do that, it has to shut down so it won't be up again for three days. A week later it happens again, and so it goes that spring and the next fall and all of a sudden, the cost of operation of the reactor just went through the roof. The nuclear industry whines about base load and all that but shortly the economics take over and that plant is decommissioned because it just isn't flexible enough to work in a renewables dominated grid.
At this point, or a little sooner, it is realized that what we really need is energy storage, fast in to handle over production, and slow out as a reservoir to handle night time. Any thoughts on what that technology would look like would be appreciated.
Check this to see how quickly renewable energy might happen.
Wednesday, January 17, 2007
Why Renewables Displace Nukes First
Subscribe to:
Post Comments (Atom)
11 comments:
I think that you are a little off on the nuclear plant shutting down bit: plants will only shut down if it makes economic sense (or if there is an immediate safety concern). If there is an excess amount of energy for a few hours but it is probable that there won't be for the next three days, then it does not make economic sense to shut down the nuclear plant. (It would be cheaper to disconnect "the renewables," throwing the excess energy away. It may be true that the renewables are dispersed, but a few nuclear plant shutdowns would justify the cost of remote relays. Such relays would be good in a green sense too, as running an old coal plant for two days while the nuclear plant comes back on line is probably not the best idea.)
Fortunately, however, there are always uses for extra energy. For large scale storage, for example, consider pumped hydro (working a hydroelectric dam in reverse), which is about 70% efficient.
Well, yes, if there is an auto-off switch
installed at your house to disconnect your solar array from the grid, or on your neighbors wind mill and at 100,000 other locations (that's about what you need to make this senario go) then you could tell the renewables to go off line and keep the nuclear plant running. It is hard though to tell a nuclear plant to go off line without having it shut down as well because the cooling system is designed to handle waste heat arising from regular operation, not the full blast of the reactor. As much heat as possible is sent out as electricity to the grid. So, if you are operating but off-line you max out your cooling system and have to shutdown.
I like you're idea of reverse hydro, but I think we need to build that out. A big capacity now only works with favorable geography. I've also heard of building up air pressure in caves or old mines and then running that out. All this kind of stuff is pretty large. What could we do to store energy compactly at substations? Could heating the ground electricly and then pulling steam out later work or would there be too large of a loss?
Demand side management is already big business. Most likely an aluminium electrolysis plant could adjust their current usage to match the localised demand.
It is true that industrial demand is large and somewhat flexible. At what point though do the discounts provided for this kind of cooperation eat so far into the operating costs of the nuclear plant that it becomes a losing game? Adding extra flexiblity elsewhere to compensate for the infelxibility of nuclear power eventually becomes pointless and it is better to just decommission the dinosaurs. Similarly, if we have power storage to absorb excess production, why worry any further about base loads? You've already stored what you need.
Demand side management isn't just for businesses. It can be very useful at home, too.
Since a nuclear plant puts out as much power as 2-3 coal plants, I assume there were none to go off-line, so this must be futuristic. As such, I'll assume some slightly futuristic appliances.
First of all, any appliance connected to the grid can, with proper hardware, detect the voltage level of the grid, and thus the load or lack thereof. So you'll have some appliances that have some "slack" in their settings. For instance, your fridge may detect that electricity is cheaper now, and freeze things in the freezer another degree or two colder. If we're off fossil fuels at this point, your water heater must be electric, and it could heat its water a few dgrees higher. Your electric car, whether at home or in a parking lot, detects the cheap energy and tries to either increase its charging rate, or top off its battery. Finally, your washer and electric dryer, which you've set to run "sometime" during the day decide that this is the best power condition of the day, and start their cycles.
Finally, if it's that hard to restart a nuclear plant, I wonder why it couldn't just shunt some steam around the turbines, and waste some heat up the cooling towers?
A very smart grid will probably be needed by the time we get to this futuristic situation for other reasons: Anti-islanding mechanisms work OK if there are only a few renewables around, but what if together they power a neighborhood and begin to sync with themselves? How do they tell that the grid is not there in case of a line break?
So, there may be ways to shut the renewables down remotely, but why do that since they will really be the cheapest source of power?
It may be also that nuclear plants could be made to throttle, I've heard of this from one source. But throttled, they're not making money so their overhead goes up. I think though that running in normal mode but not sending power to the grid would require more cooling towers, likely twice as many at least. Then the plant isn't making money and it also has to build more. Nuclear power isn't really cheap and it isn't going to get cheaper. On the contrary, its inflexibility will increase its cost.
Now, with all those fancy appliances, wouldn't the robot be able to mow the lawn with the electic lawn mower when power is especially abundant? Then go fold the clothes? We definitely want an available channel about the state of the grid.
As for Energy storage, if the environmental impact is planned for and mitigated, the creation of high altitude man-made lakes with downstream generators could work very well. On the plus side, such a plan could scale both up and down very well. Imagine something like a water tower, but taller, with a deep well sunk beneath it. The larger the distance between top and bottom, the more potential energy can be stored. When renewable energy production exceedes demand, the excess drives pumps that raise the water to your man made resevoir up in the mountains, or from the 1km deep well to the above-ground tower/storage vessel. To extract the energy, simply allow the water to flow through turbines.
I don't know about the chemistry, but I'm wondering if in the case of a closed system (such as tower/well) something other than standard H20 could be used. Something with greater density should hold more potential energy. Even water with disolved heavier elements (e.g.: salt) would increase the facility's capacity for energy storage.
I agree with your other commenters that:
* Throttling the nuclear plant would make more sense (and be cheaper, and easier) than shutting it down
* You seem to be implicitly assuming that there are no fossil fuel plants left in the system (otherwise, just shut them down)
* There is a lot more that could be done on the demand side (in addition to the ideas mentioned by others, desalination and other bulk processing, distillation (e.g. of biomass), making hydrogen, fixing nitrogen, smelting aluminum and silicon, etc.
* The system darned well better have a way to remotely / automatically disconnect producers individually and at several levels of granularity, for exactly the same reason you have a hierarchy of switches, fuses, circuit breakers, etc. to isolate consumers down to the device level. Without that, you aren't going to survive to the point where you're thinking about taken down nuclear plants on sunny days, because your grid will have slagged itself long before then.
Further, I'm not sure what exactly the concern is; if you are worried about the grid capacity, that would be a problem independent of how the energy is being put into the grid or taken out.
If you are concerned about the energy "going to waste" you should consider that (again, regardless of how it's generated) demand often is higher on hot sunny days, due to air conditioning, etc. And it also goes up when it's cheaper / plentiful. Don't fall into the 1980's trap of only looking at the supply side.
--MarkusQ
Reponding to MarkusQ, I think we need to look at new methods of power management. The hierarchy of switches and such is designed with a central distribution model in mind. In net metering laws the requirements are for anti-islanding circuits that detect when the grid is down and insulate the individual system from the grid. But, if the individual systems are cumulatively a big part of the total power generation capacity then a granular approach might make sense. I've been thinking more along the lines of distributed computing with two way communications on a high frequency channel. I've found this white paper to be a good starting point. I'm finding further reading materials here.
Perhaps I should have been clearer in the original post. I am assuming that coal, oil and gas are already off line because they are not needed on a cool sunny breezy day. These are temporarily displaced, but they'll come back on in the afternoon. The nuclear plant, when it is sent off line, is in a different situation because it needs to inspect it's fuel rods before it can start up again. The main thing is that it is not so costly to shut down coal, oil and gas because they save money on fuel and their plants are not worn out by the cycling, but for nuclear the cost is more in the operations and managing the damage that the extreme conditions in a reactor do to the reactor. So, while the fossil fuel plants will cut back, the nuclear plants will be closed.
Responding to the post about higher density liquids, I think this could get you some where in terms of making systems more compact. Salt, though, might play havoc with a turbine unless you were careful. This company has some experience with this.
To go to really high densities you might want to consider solid systems. Here I think a lot of the cost would be in the support structure but if you took advantage of the geography you might us a system of cables and rails to raise and lower a large mass.
Suppose we took the 150 foot drop below the Hollywood sign and lowered a mass of lead that is 50'x50'x300' over 12 hours. What would the power output be? Well, the potential energy is about 100 billion Joules so over 12 hours you get 2.5 MW. That's enough power for 1800 homes. Now, the mass of the lead would be 2.4e5 metric tonnes so damage it would cause if it broke loose from it's moorings would be a worry. This is comparable to the mass of an oil tanker. I think the engineering effort would be pretty substantial, but the needed volume would be 11 times smaller than for a system that used water.
Perhaps more practical than lead would be underground pumped hydro. Essentially, the same as conventional pumped hydro but between two seperated underground resevoirs, one somewhat close to the surface, the other very deep. Saves a lot of room aboveground, and is potentially very cheap. Although there's not a lot of info available. And suitable geology may also be a problem. Water must be abundant, and I don't think salt water will work very well in terms of possible ecological damage. Still, impermeable layers are common.
Another promising option is AACAES, compressed air storage with heat being stored for the compression stage. It's another wildcard; potentially very cheap, takes relatively little surface area, efficient enough, but more work needs to be done.
Post a Comment