6 ms·
The problem with renewables lies in unpredictability of wind power and photovoltaics, and in transmission infrastructure designed for predictable (and controlla
by ajuc 12y ago
The problem with renewables lies in unpredictability of wind power and photovoltaics, and in transmission infrastructure designed for predictable (and controllable) power sources.
We don't have good storage for electricity at the scale that is needed, not within few orders of magnitude, so at all times energy produced must be equal to energy used (the available storage is insignificant globaly, it is very expansive and is limited by geography - basicaly we can pump water upstream and let it flow back through turbines - this is only possible when you have a lot of water and some place much lower, where it can safely go, and it's still not cheap).
Additionaly each transmission line has maximum capacity and it will destruct itself if you try to send more.
With old-style energy sources we had baseload produced by water dams, coal, gas and nuclear plants, and peak load produced by plants that can be quickly (in less than 30 minutes) switched to produce more or less energy depending on demand - these are usually gas or coal powerplants. So with average demand for X, and peak deamd of Y you need X of stable powerplants and (Y-X) of controllable powerplants. Network is designed around that Y, using assumptions that power is divided into smaller and smaller lines from source to destination.
Only thanks to that, and to accurate predicting of demand by each level of power distribution hierarchy (I know about details of Polish system, but I guess it's similar everywhere) - the system works.
If you just swap 1000 MW of coal/gas/nuclear baseload plants with on-average 1000 MW of photovoltaics or wind turbines - you will literaly destroy your energy network. Half of the time it will produce too much energy (which results in blackouts and costly repairs), the rest of the time you will have blackouts because of not enough energy produced. You need to prepare infrastructure for that.
You need to adjust the whole network to bigger load, and to change the network topology from "connected stars" to peer-to-peer (or at least orders of magnitude more stars, connected orders of mganitude more intimately), and you need to expand the transmission lines a lot, because the only way to provide baseload with renewables is to average out production from a lot of places with different weather.
Right now electric networks aren't designed for that, and upgrading the whole infrastracture is going to be very expansive.
Germany is trying to do that - props to them, but in the meantime they are kinda problematic to their neighbors, because when they cannot deal with excess energy produced by renewables - they dump in on Polish, Czech,etc networks. Germany needs to pay for that (too much power is exactly as bad as not enough power - Polish and Czech controllable powerplants need to adjust production because of that and it costs both ways, also routing the power through the lines need to change because some lines may exceed capacity with that additional energy, and that cascades through the whole network. Energy distribution is organised in such a way, that when somebody predicted demand or production wrong - they pay for the rebalancing of network caused by that.
Sorry for wall of text...
- andy_ppp 12y agoThanks for the explanation. Britain seems ideally placed to use pumped storage; it may be a lot more difficult in a country as big as the US. Interesting about German energy production.
- upofadown 12y agoFrom the article (which I suspect you didn't read): >CSP power plants create the steam needed to turn the turbine by using sunlight to heat molten salt. The molten salt can also be stored in thermal storage tanks overnight where it can continue to generate steam and electricity, 24 hours a day if desired, a significant advantage over photovoltaic systems that stop producing energy with the sunset.
- ajuc 12y agoI skimmed it, but I missed that part, thanks.
- VLM 12y agoThink how handy something like this would be in Saudi Arabia to run de-sal plants. As long as you have big enough drinking water tanks to handle the occasional weather related or maintenance related you're golden. For example, if you have tanks big enough to hold a week or so of drinking water, then as long as you produce more than a weeks drinking water per week, you're all good. This applies to almost any process that requires energy that is either a long term average or you can store in a tank. Electro-refining copper (good luck keeping aluminum cells warm, lol) or pumping irrigation water into fields. I would imagine you can solar distill ethanol whenever you feel like it, more or less. Given big enough tanks, you could generate chlorine for anything from swimming pool size to municipal water system size. A country needs X megatons of ammonium nitrate fertilizer per season and yes all of it is made today from natgas, but you could theoretically make it from solar power, air, and water, as long as the long term average power generated is high enough to make enough fertilizer per year. Some industries like aluminum refining and some chemical plant work have a batch or shutdown time longer than the average weather related issue, so for basic chemistry scientific reasons they're screwed and will always require coal powered electricity, but most others theoretically only need a couple minutes of UPS power to shut down. Because electrical power is now super cheap, there are industries that have not put the slightest engineering effort on it... think CNC machinery, where there's no scientific reason a machine can't enter "sleep mode" within seconds, and return to work in seconds after sleeping, although in practice no one has ever made or shipped a machine like that, that I know of. As a simple example there are very few gcode programs that don't involve a dwell or toolchange where it could simply sleep in the middle of it, aside from the more difficult task of interrupting a cut. I guess this is an awesome industrial startup idea, not just for CNC machines but numerous other industrial machines. (And edited to add, there are numerous 3rd world areas overseas and California where the coal powered electric grid is totally unreliable, and machines that respond cooperatively to power interruptions would be quite handy even if there's no solar or wind generation involved.) The biggest challenge to solar powering those industries is going to be non-renewable energy is very cheap compared to long term capital costs, so if its cheaper to run on diesel than to let it sit while continuing to paying the bank loan, or if your competitors can out produce you because they're not sunlight limited, then they'll have to run on diesel or coal-based electricity for financial department reasons, even if its no engineering challenge at all to run on intermittent solar power.