DaveS wrote:... can be emptied or filled in about 20 hours, the energy storage capacities of Foyers and Cruachan are about 6 GWh and 9 GWh respectively.
My own experience of the Hydro pumped schemes was being dragged round a damp and smelly cavern and, upon being asked whether I had a question, I blethered on about cost of filling the loch up again. Geezer snorted and asked what kind of idiot d'you think we are, we wait until it pisses down. From this insight 40 years ago I presume that recharging the batteries still costs an awful lot more than what you get back in charge.
For what its worth you can make a reasonable stab at how you might organise large scale energy storage using last month's consumption data. January is typically the peak demand month of the year. If you had 45GW of capacity you would need an average of 42GWHr stored capacity which would require the application of 77GWHr of energy over 12 hours to achieve and, of that 42GWHr, at least 8GWHr has to be capable of being released in 1 hour - equivalent to 4 very big power stations. I chose a base of 45GW because demand is "equidistant" in terms of time about that value in a 24hr cycle. Demand itself was a max of 53GW and minimum of 34GW. If you drop the base to 40GW then the energy cycle is unbalanced; there would be less time for regeneration than generation which would be, I think, untenable, and the stored energy soars threefold to 120GWHr.
The first observation about storage is that whatever system you adopt, it can not be simplex ie working in only one direction at a time; it has to be duplex which effectively rules out any pumped storage system. The reason for this is you will most likely be a net consumer or generator of your stored power for 12 hours in a 24 hour cycle; you would simply not have the time to regenerate a 24 hour cycle system in less than 12 hour. Its a quite sublime application of Newton's law of energy conservation (with a dab of modern logic flung in). If you were to extend the cycle, say 48 hours, then the stored energy requirement becomes an unimaginably large figure and the cost soars. Cost can only be controlled if you regenerate during the offpeak hours.
Its quite obvious that you have to site the stored energy facility close to where it is consumed. A lot of power has to be delivered very quickly and the further it is away, the greater are friction losses.
A further observation is that this only works with non-intermittent variables; at least, that is true of such a mathematical model as I propose and which would thus rule out wind power.
Well, where to start?
Firstly, I'm glad that my energy storage figures of about 6 GWh & 9 GWh based on memory agree with Webbie's quoted 6.3 GWh & 8.8 GWh.
There is some truth in your man's remark. In round numbers both generating and pumping is about 90% efficient, giving a storage cycle efficiency of 81%. However, captured rainfall adds to the upper reservoir - and a network of pipes captures burns further afield to increase the overall catchment area - so more water goes down the hill than is pumped up. The net result is an average storage cycle efficiency of around 90%, and more when there's heavy rain.
Your demand figures are, I presume, for GB (excluding? interconnectors). I must say that the idea of using storage to entirely balance the entire daily GB demand is heroic, and beyond anything I've ever seen previously considered. I would seriously question the practicability of this. IMHO there will always be the need for some load following generation which, combined with dynamically scheduled load, can go a long way towards achieving what's wanted. On a smaller scale this is certainly so. Prior to introduction of the current NETA / BETA trading system, and using these techniques, the South of Scotland winter load curve was essentially flat, with a peak around 1800 and a dip at around 0500 of similar energy volumes. Cruachan was used to transfer one into the other leaving a virtually flat generation steam line, i.e. a single set could be used for (quite limited) load following, the rest running flat out or stopped. In efficiency terms it really doesn't get much better. All gone now, of course, for the "benefits" of unrestrained market competition.
I think I understand your point re. balancing pumping & generation, but would point out that that is one reason why Cruachan & Foyers were designed with lower MW capacity than the dams could theoretically support: the margin means that there is no need to achieve perfect daily balance, so e.g. it is possible to deplete the upper reservoir Monday to Friday then fill it up again over the weekend.
I disagree with your point re. siting of storage to avoid friction. Despite persistent rumours to the contrary, electricity transmission is actually quite efficient - around 98% on average. This is why most of the big English coal burners were sited in Yorkshire - it's cheaper to move electricty by wires than coal by trains.
There is scope for increasing storage capacity through the installation of pumps at conventional hydro stations. Webcraft's source mentions Hydro's plans for Sloy. For bulk energy storage (as opposed to short term stability response) there is no requirement for the sophistication of reversible pump storage systems - a big industrial pump is quite good enough.