Typically, electrical power is considerably cheaper in the dead of night than during peak demand during the day. This is due to the very simple reason that demand is much higher during the day, and spare capacity must be built to accomodate it. We can see this clearly from a graph of California's energy demand:
In deregulated electricity markets, there is the possibility to use electricity storage as a means to buy low and sell high -- known as arbitrage. There are a number of different schemes designed for utility scale power storage. The oldest power storage scheme is probably pump hydro, where a volume of water is pumped back and forth between two reservoirs. Another system stores energy as air under pressure in an underground aquifer or salt cavern. These systems suffer from the disadvantage that they require geographical features in order to function.
The other option is to pursue technological solutions. I have already detailed the possibility of using large numbers of plug-in hybrids to do voltage regulation. However, static installations are also possible, such as flywheels or flow batteries. With these systems, since they are not fulfilling the energy demand for transportation, the round-trip efficiency of the storage method becomes paramount. Hydro-pump systems are typically around 80 %.
Of all the electricity storage systems, I feel that flow batteries are the most promising. There are a number of different flow battery electrolytes: Zinc-Bromine, Vanadium redox, Sodium Poly-sulfide Bromide. The Vanadium system appears to be the most popular commercial system at the moment. It was invented in Australia:
The power density of flow batteries is similar to lead-acid batteries. The true advantage of the flow battery lies in that it has a high round-trip efficiency and it separates energy storage capacity from maximum power. They also have very long lifetimes -- only the pumps and the membranes can wear out. The membrane is much simpler than that used in, say, a PEM fuel cell.
The biggest player in the North American flow battery market is VRB Power of Vancouver:
We can see from their numbers that the round trip efficiency is approximately 75 %. It's worth noting that if they could be directly connected to a DC rather than AC distribution grid, this number would be higher. The capital cost of the system is very high -- they estimate that it is approximately $325,000/MWh. However, maintenance and operating costs are claimed to be extremely low, at only $1/MWh.
The real question is, are these systems a good investment? Let's examine the numbers.
California's energy price fluctuates greatly day to day. Using the prices for Palo Verde today, we find that peak power sold for an average of $90/MWh and off-peak for $60/MWh. Given the return rate of 0.75, we see that a Vanadium redox battery could make a profit on the order $10/MWh. This in itself is not much. However, the VRB can also provide excellent ancillary services: voltage regulation and spinning reserve. The value of ancillary services can vary wildly; I will very conservatively estimate them at $15/MWh. Thus, $10 + $15 - $1 = $24 can be generated each day by a VRB system per MWh of storage.
Given the current capital cost, that works out to a payback period of 37 years. Clearly, the system is not a good investment for the California grid at this time. However, the technology is still at an early stage of development and there is no manufacturing efficiency of scale benefits at this point. With an order of magnitude reduction in the costs of a flow battery, they would suddenly become a very attractive investment, with an annual rate of return of about 25%!
Whether or not electricity arbitrage systems will become cost effective in the future probably depends on where you see alternate energy going. If you think the grid will become wind heavy, the demand for voltage regulation will greatly increase and the profitability of these systems will improve. On the other hand, the introduction of widespread solar might depress peak power prices, making the price more uniform and causing the round trip efficiency of paramount importance. Of course, if solar becomes dirt cheap, power will be cheapest during the day and one will again be able to conduct arbitrage, buying at noon and selling at dusk. It does appear clear that the introduction of 'renewable', intermittent sources of electricity will increase the need for a cost effective means of storage that is capable of providing ancillary services.