So first off my opinion
Battery storage feels like one of the easier bits of the energy transition to understand.
Charge battery.
Wait.
Discharge battery.
Job done!
Except obviously it isn’t.
A battery in my phone is one thing.
A battery installation capable of pushing tens or hundreds of megawatts into the electricity grid is quite another.
And the more I’ve looked at electricity generation, the more obvious the need for storage becomes.
The problem is time
Electricity has this slightly annoying characteristic.
We need generation and demand to remain balanced.
But renewable generation doesn’t particularly care when we want electricity.
The wind doesn’t start blowing because everyone has put the kettle on half time at the world cup.
Solar doesn’t stay around for the evening peak because we’d find it convenient.
So storage lets us move energy through time.
Generate it now.
Use it later.
That’s a surprisingly powerful ability.
What is a BESS?
BESS means Battery Energy Storage System.
It’s more than just a giant collection of batteries.
You’ve potentially got:
- battery cells;
- battery modules and racks;
- inverters;
- transformers;
- cooling;
- fire detection and suppression;
- battery management systems;
- energy management systems;
- grid connections.
That means there is also an enormous amount of operational data.
Which obviously gets my attention.
Batteries don’t just store electricity
This was probably the more interesting bit for me.
My original view of grid batteries was basically:
Electricity is cheap → charge battery.
Electricity is expensive → sell electricity.
And they can do that.
But batteries can respond extremely quickly.
That makes them useful for helping stabilise the electricity system as well.
Frequency changes.
Grid conditions change.
A battery can react very quickly.
So the value isn’t simply how many MWh the battery can hold.
It’s also about what the battery can do while connected to the grid.
MW and MWh
This is one distinction that is worth understanding.
MW tells us how quickly the battery can deliver power.
MWh tells us how much energy it can store.
So a 100 MW / 200 MWh battery could theoretically deliver 100 MW for around two hours.
Ignoring losses and operating limits for the simple example.
That’s quite different from a 100 MW / 400 MWh battery.
Same maximum power.
Twice the energy.
Again, knowing what the numbers actually represent matters.
Then we get degradation
Batteries don’t stay new forever.
Charging and discharging them changes them.
Temperature matters.
How deeply they’re cycled matters.
How quickly they’re charged matters.
Age matters.
That means the asset you bought isn’t exactly the same asset several years later.
And that’s where I think the data becomes really interesting.
State of Charge tells us roughly how full the battery is.
State of Health attempts to tell us what condition it’s in.
Then you’ve got temperature, voltage, current, cycle count and potentially huge amounts of cell-level information.
This isn’t just data for a dashboard.
It potentially determines how much usable asset you’ve actually got.
So the data team needs to understand the battery
Imagine somebody asks:
Give me battery capacity for every asset.
Simple enough.
Except what do they mean?
Installed capacity?
Current usable capacity?
Maximum discharge power?
Available capacity right now?
Capacity after operating constraints?
Suddenly the engineering definition matters to the data model.
That’s exactly the sort of thing I mean when I say data engineers benefit from understanding the industry they’re working in.
You don’t need to be the engineer designing the battery.
But knowing why apparently similar numbers mean different things can stop you building something technically correct and practically useless.
Where is my head at?
Battery storage makes enormous sense to me.
But I think it’s easy to hear “battery” and mentally scale up the thing in your phone.
Grid batteries are infrastructure.
They’re power electronics, software, chemistry, cooling, safety systems and electricity-market participants wrapped together.
And from a data point of view they’re fascinating.
Because the data isn’t just recording what the battery did.
Increasingly, the data helps decide what the battery should do next.
That’s where my head is at the moment.
Gareth Winterman