So first off my opinion
This one sounds like something I’d come up with after several pints.
We’ve got too much electricity.
Right.
Let’s use it to compress an enormous amount of air.
Where are we going to put the air?
Underground.
Then when we need electricity again we’ll let the air back out through a turbine.
Right…
Except Compressed Air Energy Storage is a real thing.
And once you get past how slightly mad it sounds, the idea is actually quite interesting.
So how does it work?
The acronym is CAES.
Compressed Air Energy Storage.
When electricity is plentiful, electric compressors compress air.
That compressed air is stored.
Potentially in a very large underground cavern.
When electricity is needed later, the compressed air is released and used as part of the process of driving a turbine and generating electricity.
So in its simplest form:
Electricity → Compressor → Compressed Air → Storage → Turbine → Electricity
Basically a battery.
Except the storage medium is compressed air.
And the battery might be an enormous hole underground.
I love energy engineering.
Why underground?
Because storing serious amounts of compressed air requires serious amounts of space.
Building a pressure vessel big enough would get rather expensive.
But geology can potentially give us the container.
Salt caverns are particularly interesting.
We already use underground geological formations to store things such as natural gas.
So rather than manufacturing the entire storage vessel, suitable geology can become part of the energy infrastructure.
That’s a recurring theme I’ve started noticing.
Energy systems don’t just depend on engineering.
Sometimes they depend on what happens to be underneath your feet.
Compressing air creates heat
This is the bit where my school physics started coming back to me.
Compress a gas and it gets hot.
Let compressed gas expand and it cools.
That matters because the heat represents energy.
If we compress the air and simply throw that heat away, we’ve lost some of the energy we originally put into the system.
Then when the air expands later, we may need to deal with the temperature dropping.
Different CAES designs deal with this in different ways.
Some systems use additional fuel during the generation process.
More advanced ideas try to capture the heat created during compression and reuse it when the air is expanded.
Which seems much more satisfying.
Store the air.
Store the heat.
Put them back together later.
Obviously considerably easier to write in a sentence than engineer in the real world.
So why not just use batteries?
This question keeps appearing whenever I look at energy storage.
And I think I’m gradually learning that it’s the wrong question.
Different storage technologies have different characteristics.
How much energy can they store?
How quickly can they deliver it?
How long can they store it?
How much does capacity cost?
How much energy is lost?
How quickly does the system degrade?
What geography does it require?
Lithium-ion batteries are very good at some things.
Pumped hydro is very good at others.
Compressed air might potentially make sense where we need very large amounts of storage and have suitable geology available.
The question isn’t necessarily which technology wins.
It’s what problem we’re trying to solve.
And then there’s the cavern
From a data point of view, this bit fascinates me.
If I’ve stored an enormous quantity of compressed air underground, I’d quite like to know what it’s doing.
Pressure.
Temperature.
Air mass.
Injection rate.
Withdrawal rate.
Cavern condition.
Compressor performance.
Turbine performance.
Electricity consumed.
Electricity generated.
Now add market data.
Electricity price.
Renewable generation.
Forecast demand.
Weather forecasts.
Suddenly the question becomes:
Should we charge the cavern now?
That’s a data problem.
Efficiency becomes an end-to-end measurement
It would also be very easy to measure the wrong thing.
The compressor could be running efficiently.
The turbine could be running efficiently.
And the overall storage system could still be performing badly.
What matters is the complete journey.
How much electrical energy went in?
How much useful electrical energy eventually came back out?
What happened in between?
That means data from multiple pieces of equipment needs joining together into one operational story.
That’s exactly the kind of problem that makes data engineering in energy interesting.
The equipment doesn’t care that its data lives in different systems.
Physics has already joined the datasets together for us.
Our databases just need to catch up.
So where is my head at?
Compressed Air Energy Storage is one of those technologies I initially thought sounded ridiculous.
Now I think it’s rather clever.
I’m not convinced it makes sense everywhere.
You need appropriate storage.
You’ve got conversion losses.
And dealing with the heat generated during compression appears pretty important to making the overall system efficient.
But the fundamental idea is wonderfully physical.
Take electricity you don’t currently need.
Use it to squeeze air into a geological formation.
Then get some of that energy back when you need it.
It’s difficult to imagine a more literal way of storing pressure for later.
That’s where my head is at the moment.
Gareth Winterman