Dec 12 2016

The Energy Storage Gap

hourly_bigWe live in a time of rapid technological advance, so much so that we take it for granted. This also means we are more sensitive to situations in which a particular technology seems to be lagging.

Right now arguably the most critical technology that is lagging behind other related technologies is energy storage. As we try to update our energy infrastructure, specifically with the goal of incorporating more renewable sources, energy storage appears to be the main limiting factor.

Why Is Energy Storage So Important?

Even with our current energy infrastructure, having the ability to efficiently store large amounts of energy for later use would be a huge advantage. Energy production needs to be matched to energy demand, and demand fluctuates throughout the day. The chart above shows energy demand for New England, which peaks around 7:30 pm and bottoms out  at around 4 am. Peak demand is about 50% more than trough demand.

Energy producers have to constantly match production with this demand. This means they need to have additional power production on standby which they can bring online as needed. This causes inefficiency. For one thing, energy producers will use the most efficient energy production methods for baseline production, then bring on the less efficient production just for peak demand. Peak energy is therefore more expensive than off peak. This partly derives from the fact that peak production needs to come from sources that can be turned on quickly, while baseload sources do not need this feature.

In an optimally efficient system energy demand would be a flat line. This would mean we would not need any more capacity than average use, and energy production would not have to be constantly adjusted.

Adding renewables to the mix complicates the issue further, because they are intermittent (not on-demand) energy sources. Solar production is during the day, which does match the daytime high demand, but often missed the early evening peak. Any source that does not produce energy during peak demand does not reduce the need for peak capacity. You still need all those inefficient (and perhaps dirty) sources of power to meet peak demand.

Wind does not have any time of day restriction, but it is intermittent. It is possible to make predictions about wind power generation based on weather forecasts. These predictions get more accurate the larger the aggregated area that is being covered. So as more and more wind power is installed in the system, it becomes more predictable. Still, it cannot be counted on to offset peak demand.

Demand can be smoothed out using various methods. It is difficult to count on people changing their habits, so the best solutions will be technological. Water heaters, for example, can be placed on timers so that most of the heating takes place during off peak. As the use of electric vehicles becomes more popular, this is another opportunity to smooth out demand, by timing the recharge cycle to off peak.

Consumers do have a motivation to shift their use of energy to off peak as well, because off peak energy is less expensive. You could turn on the dishwasher right before you go to bed, for example, rather than running it right after dinner during peak demand.

Massive Storage is a Gamechanger

Massive energy storage would radically change the situation. If there were sufficient energy storage built into the grid then we would only need baseload production. Peak production could be entirely eliminated. Intermittent renewable sources could also contribute to baseload production. It essentially would not matter (much) when you produced the energy. Unused energy would go into storage and then come out of storage on demand.

I say it does not matter “much” because it still matters some. It matters because every time you convert energy from one form to another a little energy is lost entropy (damn thermodynamics again). You are better off using the power directly rather than storing then using it. This is not insignificant, but is minor compared to the massive advantages.

So – what are the options for mass energy storage? There are lots of things on the drawing board, but no clear winners yet. Massive energy storage needs a large capacity, the ability to store and release energy quickly, low energy loss with storage and release, a long lifetime (charge-discharge cycles), and it needs to be cost effective. Size and weight are not generally an issue, since storage does not have to be portable.

One interesting method for mass storage, which I just learned about and was the inspiration for this post, is using energy to chill air to liquid state (essentially liquid nitrogen). It’s funny to think of storing energy by making something cold, but the idea is clever. Allowing the liquid nitrogen to warm up causes it to expand greatly as it turns back into gas, and that expanding gas can turn a turbine to generate electricity.

There is a prototype plant near Manchester which can store and produce enough power for 5,000 homes for 3 hours. Three hours does not sound like much, but it is just enough to cover peak demand. The idea would be to have enough of these plants to store energy produced by other means during trough demand and then produce energy during peak demand. That could go a long way to flattening that curve above. The current plant is relatively small, 5MW. The company has plans for a 200MW plant that they say can power a small city for 6 hours.

According to

Highview Power Storage claims a 50% “round trip” efficiency, which they hope to increase to 80%. By comparison, batteries are 60 to 70% efficient, pumped hydro is 75% to 85% efficient, and compressed air energy storage is 45% efficient. Although this system is less efficient than some batteries, it has a virtually unlimited number of charging/discharging cycles with no loss of capacity due to excessive depth of discharge.

The Highview Power Storage is a cryogenic storage plant. I don’t think 50% efficiency is going to be good enough for mass storage. That would mean doubling energy production for any energy that is stored. An efficiency of 80% or higher seems more realistic.

Pumped hydro seems like the best option right now. This method simply pumps water from a lower reservoir to a higher reservoir to store the power and potential energy, and then releases the water to generate hydroelectric power when needed. It is already up to 80% efficiency. In the US there is already enough pumped hydro storage to meet 2% of demand. (Europe has 5%, Japan 10%.) So unless cryogenic storage can get to that promised 80% efficiency I don’t think this is going to be any major contributor. Perhaps it could be used in locations where there isn’t an available water reservoir.

Compressed air is an interesting option. For massive storage we would need large underground caverns into which we could compressed large amounts of air for later release to drive turbines. However, at 45% efficiency this does not seem like a great option.

Batteries have a decent efficiency, 60-70%, but not great. They are also expensive, have a limited lifespan, and have a small capacity. There one advantage is that they can release energy extremely quickly, on the millisecond scale, and so might have a limited use in leveling off minute to minute fluctuations in demand.

Yet another option is flywheel energy storage. These systems, which already exist, use a rotating cylinder, floating on magnetic bearing and in a vacuum, to store power and kinetic energy. The cylinder can rotate up to 50,000 rpm. Optimal systems claim 85% to 90% roundtrip efficiency. Systems using mechanical bearing are much lower, around the 50% efficiency level. The efficiency of flywheels depends greatly on the time between storage and use, as the flywheel slows down over time due to friction.

Hydrogen is another method of energy storage. Essentially you use energy to break water up into oxygen and hydrogen and then burn them back together to generate energy (in a hydrogen fuel cell). Current systems have a roundtrip efficiency of 30%, which is way too low. The advantage of hydrogen is that it is fairly portable, which is why it is being considered for use in cars. The limiting factor now is safely storing enough hydrogen with reasonable size and weight limits.

Even low efficiency systems could be useful if they are attached to intermittent renewable energy production methods. If we get, for example, to the limit of adding solar energy to the current system, we could use incremental solar power for water electrolysis to make hydrogen for use in fuel cells in homes or at other points of use. Even if this is inefficient, it could still be useful in offsetting peak demand and it’s coming from a renewable and clean source.


It seems like there is already a range of acceptable options in terms of massive energy storage, with pumped hydro appearing to be the best current option. However, there is room for additional methods of massive energy storage.

We don’t yet have the perfect method, with high efficiency, long lifespan, portable, and rapid response to demand. A highly efficient and high capacity battery made from cheap and non-toxic materials with a long lifespan would do the trick. Making hydrogen much more efficient might do the trick also.

I suspect we will end up with a hybrid system using multiple storage systems optimized to size, location, and application. I am not sure why we don’t have more pumped hydro storage. The references I found said that they are being built, but I am not sure why the delay. Was this a lack of vision, or is there some downside to this technology I am missing? It probably comes down to cost effectiveness.

In any case, energy storage is increasing important to our technological infrastructure. This may seem boring to the average user, since it is something that operates in the background. But is is critically important as we try to update our energy infrastructure.

14 responses so far