Big battery packs: Solution to renewable-power challenge?
It's not news to engineers that one of the biggest impediments to the use of non-polluting energy sources such as wind, solar, and even water is their inherent intermittent availability (although many journalists and other starry-eyed pundits don't seem to grasp this fact). In short: storage is as important to successful implementation of this type of energy system as the generation itself.
As a result, efforts to reduce dependence on traditional sources such as conventional power plants means that these facilities still need to be sized to handle the maximum load, which may occur when the cleaner source isn't available. Further, there are locations where the grid isn't available as a “backup” for when the sun, wind, or water are unavailable.
There's a reverse consideration, too: when the clean source is generating power but not needed, it is wasted unless it is connected to local storage or the grid; but the grid wouldn't be available in remote locations.
There are some proposals to have the batteries in an all-electric car (EV) or perhaps a hybrid HEV car act as the reservoir for unused power. While that may be technically feasible, it means you have to buy a car first (and all that entails). It also means you can't have storage access while your vehicle is in use; nor is it a solution for less-developed, remote regions.
One solution, of course, is to have a backup generator, but that gets you back to where you didn't want to be in the first place. So when I saw two articles in The Wall Street Journal, one saying that Sony was planning to build a factory to manufacture battery packs for such applications ("Sony's Latest Bet: Lithium-ion Batteries For Energy Storage") and the other about the fairly substantial battery backs in the Tesla all-electric cars (see my previous column "Big batteries, big issues"), I started to think: maybe a solution to the storage problem is to develop and leverage similar high-volume packs as standard storage modules, for use both as a back-up supply and storage of excess generated power?
I began with my favorite first step: a “back of the envelope” assessment of the run time and costs, based on the numbers in the article. The larger Tesla car battery pack has a capacity of about 85 kWHr, and currently costs $21,000 to $25,000 (estimated); and Tesla hopes to build a mega-factory to drive that down.
On the load side, some figures I found online (obviously, to be taken only as very rough) is that a typical American home requires about 500 to 1,000 kWHr/month, or 20 to 35 kWHr/day (obviously, a remote location might need far less). That means that the basic Tesla-sized pack could support a house for several days. I wild-guess estimate the cost of the charging, discharging, and management electronics at another $10,000. So we're talking serious money here, no question of it, and certainly more than a typical home can afford.
But that doesn't mean there is no answer. After all, our industry is full of examples of products that cost a small fortune at first, and now are almost "throwaways" (you can make your own list, of course). I realize that these batteries, connectors, and cables do not lend themselves easily to the cost-reduction path of ICs, but there may be ways to at least start on that path.
Perhaps part of the solution to the storage challenge is to look at lowering cost by using the mutually reinforcing tactics that have worked so well in so many other cases: develop standards, run high volumes, and leverage developments elsewhere. By supporting just a few storage-size options (small, medium, and large); by having common standards for key parameters such as power connectors, footprint, safety mechanisms, DC/AC inverter, and control interface; and by using technologies and manufacturing techniques being developed by auto vendors such as Tesla, it might be possible to substantially cut the price of battery-based energy storage. For example, it might make sense to use the same connector and some variation on their charging/control subsystem.
Does this approach make sense, at all? Or am I being oblivious to the reality of providing battery-based standby power for mass-market, lower-cost consumer use?