Battery farms: one key to a better grid

Battery farms could help create a better grid.

An article in IEEE Spectrum discusses the case for "replacing" power plants with "battery farms."  Besides the obvious problems that batteries don't generate energy but only store them, there is some merit in implementing battery farms.  Most importantly, they serve as a stock that can help balance the differences between rates of power generation and consumption.  That systems aspect is particularly interesting to me.

Here's a quick primer on a key concept of systems thinking - stocks and flows.  Systems can be modelled as reservoirs of resource (stocks) and movements of resources between those reservoirs (flows).  The classic example is a bathtub and its faucet and drain.  The faucet and drain are flows; the bathtub itself is the stock.  As the rate at which water enters the tub via the faucet and exits it via the drain, the level of water in the tub will change.  This kind of model can be represented mathematically too, allowing for some quite advanced calculations and simulations.  If you want to learn more about stocks and flows, and systems thinking generally, I must recommend Donella Meadows excellent book, Thinking in Systems: A Primer.

Now let's say that the faucet of our tub is some power generating station, the drain is the collection of electrically powered devices and machines.  If we're talking about a conventional power plant (hydroelectric, nuclear, coal, natural gas, etc.), the output is fairly constant.  The consumption, however, is variable - we use more electricity some times and less electricity at other times.  It's useful to have a stock between the power station and the user-devices, because we can use it to balance out the total amount of electricity.

Think of the tub again.  If water is available only irregularly, there will be times when we need water but no water will be available.  We can then use the tub to store water we don't need when it's available, so that we have some when we do need it.  Similarly for electricity: during times of low consumption, we can store electricity somewhere, and use it when we need it.  Without a stock for electricity, there must be enough power generating capacity to meet peak demand, otherwise we'll have brownouts or blackouts.  But if we have a stock into which we can store electricity, then we only need enough power generating capacity to meet the average (sort of) power demand.  When we need less electricity than average, the excess is put into the stock; when we need more electricity than average, we use both the power plant and the reserves held in the stock.  This means we don't need as much power generating capacity, but we still avoid brownouts/blackouts.  And we save money as a result too, because power generating stations can be very expensive.

So the question is: where can we store electricity?  The obvious answer is in batteries.

These aren't your usual batteries, though; these are dozens of shipping-container-sized structures, each basically a fancy, giant Duracell - hence the term "battery farm."  Of course, building, operating, maintaining and decommissioning these battery farms costs money.  But if the cost of the farms is less than the cost of the extra power plants we'd need to meet peak demand, then we're ahead of the game.  Battery farms are highly modular, have nearly no moving parts, and can be easily scaled.  Power generating plants have none of these characteristics, so battery farms are definitely attractive.

That's why so many researchers and companies have been trying to figure out how to make these giant batteries for less money.  It used to be that a battery farm was simply too expensive.  (Although I have to wonder about the total life-cycle costs: for each coal fired power station we can do without, for instance, by using battery farms, we're significantly cutting carbon emissions and doing less damage generally to the environment.)  However, we're now getting to the point where battery farms are economical, and so many people are seriously looking at battery farms now as good ways to help provide a reliable electric grid without further damage to the environment.

Some people are concerned about the chemicals used in these giant batteries, and what would happen if one of them exploded or even just leaked.  Bad things would happen, that's a given.  However, I would wager that battery farms are more robust than most power plants, and the nature of the damage that would result would be far less than, say, the environmental nightmare of burning coal, or the potential radioactive damage of a nuclear plant failure.  Everything has a risk - there's no way to eliminate all risk; it's stupid to even consider living without risk.  However, we can choose technologies that lower the net risk to people, property, and the environment.  And that seems to be what battery farms can give us.

Of course, it's more complicated than this, because not only do we need to keep a constant voltage and amperage in our electric supply, we need a steady frequency.  And managing all these things is like managing not only the rate of flow of water at the faucet and drain of our bathtub, but also managing the pressure of the water in both the faucet and drain, as well as the quality of the water.  These all have analogues in the science of the electric grid.  In fact, the basic equations governing both fluids flowing in pipes (and tubs) and electricity flowing through wires are essentially the same.  If you want to learn more about that, you'll have to study system dynamics, and in particular system equivalence.

Now, things get trickier still when we consider most of the typical sources of sustainable energy, like solar, wind, or wave energy.  Unlike conventional power systems, which more or less produce constant rates of electricity, the most common sustainable energy sources produce variable rates of electricity.  The sun doesn't shine at night, and the wind blows only in the presence of pressure differentials.  We cannot change when the sun shines or when the wind blows, so these sources of energy cannot be expected to be predictable. If they're not predictable, then we have to find a new way to manage the flows of electricity through the grid.  Both inputs (the energy sources) and outputs (the rate of electricity consumption) are variable and quite unpredictable, and that makes for an analytical nightmare.

Again, battery farms can solve this problem because the farm doesn't care when or how fast (within reason) the electricity arrives for storage.  Sustainable sources, variable though they might be in the short term, do tend to produce quite steady rates of electricity in the long term.  Even though the sun shines only some of the time, and clouds can obscure it, the rate of insolation is relatively steady over, say, a year.  So we can predict what the yearly averages should be, and design battery farms that can accommodate any reasonable variation around those averages.  So one can in fact imagine a solar+battery farm that is able to deliver a fairly constant supply of electricity day and night.

Not only does this make development of battery farms more urgent, but it also makes solar/wind/wave energy more attractive.  Win-win.

So far, the battery farms that have been proposed have been pretty massive things, typically associated with specific power plants.  This makes sense in that the existing grid was designed to expect fairly steady power inputs.  It doesn't matter whether the power is coming from a hydroelectric plant (that produces nice steady power all on its own) or a solar plant augmented with a battery farm.  From the grid's point of view, the power is steady, and so it's happy.

But let's think bigger than this.  The grid was designed in a bygone age.  We know so much more and have much better technology available to run the grid.  Maybe we need to think about the use of batteries in other settings as well.

Here's my idea: Why don't we use batteries at the other end of the grid - the consumption end - to balance out the electric load?

Basically, the grid is designed to work best when supply very closely matches demand.  We put battery farms on the supply end of the grid - at the power plants - to help ensure a constant supply of electricity to the grid.  But the other end of the grid - when we use electricity - has tremendous variability.  We put batteries at the supply end to control variability of supply - why not put batteries at the demand end to control variability of demand?  If we did that, then the grid would always see both a constant supply and a constant demand.  Most of the grid's complexity comes from having to manage variability; if we can eliminate the variability, the grid will become simpler, more robust (i.e., fault tolerant), more reliable, and cheaper to operate and maintain.

What I envision are smaller battery farms at the level of individual streets, perhaps even buried in underground bunkers so as to keep neighbourhoods attractive.  Each battery would store excess electricity from the grid when demand at the houses on that street is low, and then release it when demand increases.  Again, from the grid's point of view, that street would seem to always consume electricity at a constant rate.

Not to mention, a local battery could help in cases of blackout by being able to provide enough emergency power for devices and machines to shut down gracefully.  As more and more of our technology becomes electronic and electromechanical, it will become more and more important to make sure they have a chance to react safely during power outages.

So these giant battery farms may in fact be very, very helpful to maintain the grid and our "developed" cultures, while also helping to control pollution and environmental damage.