Skip to main content

Brent Cullimore

I like coffee, and when I need it, I need it now.

I have an electric pot that is powerful enough to dim the lights a little when I turn it on. I imagine Homer Simpson snoring with his feet resting on the control panel. When I turn on the coffee pot, an alarm sounds in Homer’s control room, startling him and making him knock over his coffee as he rushes to add compensating power to the grid.

Homer Simpson'

That’s a silly image of course, but generating exactly the amount of power that is currently required is an amazing feat. You can reason that it must be an easy task on average, but what if the denominator in that average gets really small? What if you were trying to level the load on a small island, where the effects of one coffee pot might actually be measurable? Turns out, even big islands like Japan have a smaller denominator than most: their average power demands are more volatile than what a continent-based grid experiences.

Currently, power generation capacity must be sized to meet peak demand (usually late afternoon on a work day) then most of that equipment sits idle during the night. This represents a severe economic inefficiency.

Enter renewable energy. Solar power helps shave peak air conditioning loads, but the sun goes down about the same time that the demand for power peaks. The rate at which compensating power must be added to the grid is getting larger and larger each year, as demonstrated by the famous “duck curve:”

duck curve

In Hawaii, the quacking was loud enough that they had to issue a cap on rooftop solar installations.

In 2015, wind power was the largest source of new generation capacity added in the United States, even with cheap natural gas. And it is reported to now be more economical even without tax incentives. In fact, more new solar power generation was added than natural gas.

Has your utility company ever paid you to take electricity off their hands? Sound crazy? It happens on the electricity spot market because there is no place to store any excess power.

So the world needs batteries. Damn big grid-scale batteries and we need them soon. Cheap would be great. Super-safe and long lasting, while you’re jotting down my wish list.

The need for electrical energy storage (EES) might be getting acute, but it isn’t new. About 100 facilities around the world store electrical energy in reservoirs by pumping water up at night and letting it flow through turbines during the day. For every 100 units of energy added, up to 80 units are recovered: the round-trip efficiency is a very high 80%.

Here’s a sample problem with a more complete description of such Pumped Hydroelectric Storage (PHS) systems.

An alternative is to compress air into a huge geologic cavern called a salt dome. Here’s an example and description of such a Compressed Air Energy Storage (CAES) system. So far, only two such facilities have been built, but this remains an active area of research. CAES systems require gas-fired combustors, so in many ways they represent a Brayton cycle with a compressor that is separated from the turbine, both mechanically and temporally. Different shafts, different shifts. Pressurize during the night shift, burn and turn on the day shift. (If you ‘turned and burned’ you’d be putting the turbine in front of the combustor!)

The main problem with geology-based batteries is, well, geology. A new PHS facility is being built in Switzerland, but you won’t see one anytime soon in Planes States such as Kansas. (And yes I know it is “Plains States” but we are talking geometry after all! Well, geology anyway.)

If the mountain won’t come to Kansas, cut a hole in Kansas. And call Guinness World Records when you make the O-rings.

Salt domes are more likely to be found in such states, but there is still a limited supply and CAES is not yet a widely accepted option.

Why is it so hard to make a battery? Aren’t they everywhere nowadays in cell phones and laptops and Fitbits? (I have two of the three of those, since I’m all about conserving energy.) Can’t we just hook enough of them together? There are some who say you can, such as Elon Musk. Others are concerned about future lithium shortages.

In fact, any dry cell battery has a central problem: the rate at which it charges or discharges must be linked to the total energy it can contain. You can’t choose the megawatts independently of the megajoules. (OK, megawatt-hour is a more conventional unit for energy, but it just isn’t as memorable.) Dry cell batteries do have a lot of application to short term grid storage (“frequency regulation”). But they are hard to use for more than a few hours at a time, they have not demonstrated long life and have issues with discharging too far.

Nonetheless, investment and research into advanced dry cells is far exceeding that of any other technology, so expect gains there in a few years that will make me delete or edit the above paragraph.

Another approach already has proven long life, good scalability, and good safety: wet cells, better known as flow batteries. They will be the subject of a future blog.

There are many wild ideas out there for EES solutions that I haven’t covered. Trains full of rocks, salt rocks that melt and freeze, and, who knows, maybe trains full of molten rocks. The sheer number of ideas being pursued is a sign of the urgency of the need.

The race is on, and it isn’t clear who will win and who will lose. As long as someone wins, I get to brew my coffee for years to come. Oh yeah, and saving the world would be nice … add that to the plus column.