So first off my opinion
When somebody says hydroelectric power, my brain immediately goes to dams.
Huge reservoir.
Huge concrete wall.
Water thundering through equally huge turbines.
Basically something that looks impressive on a documentary.
But hydro doesn’t necessarily have to be enormous.
If you’ve got flowing water and a useful difference in height, you potentially have energy.
And that’s where micro hydro starts getting interesting.
So what is micro hydro?
Essentially it’s hydroelectric generation on a much smaller scale.
Water flows through a turbine.
The turbine turns.
A generator produces electricity.
Same basic principle as a huge hydroelectric station.
Just considerably smaller.
And importantly, you don’t necessarily need to build a giant reservoir.
Some systems can use the natural flow of a river or stream.
Part of the water is diverted through the generating equipment and then returned to the watercourse.
Water has a nice advantage
I’ve spent a lot of time looking at wind recently.
Wind turbines obviously need wind.
Solar needs sunlight.
Water flow can potentially be much more consistent.
Obviously that depends massively on the water source.
A stream isn’t going to produce the same flow all year.
Rainfall changes.
Seasons change.
Droughts happen.
But if you’ve got a reliable water source, hydro can potentially provide much more predictable generation than some other renewables.
And predictable generation is useful.
Head and flow
Two things seem particularly important.
How much water have you got?
And:
How far can it fall?
The amount of water moving through the system is the flow.
The vertical difference is generally referred to as the head.
More flow means more available energy.
More head means more available energy.
So a small amount of water falling a significant distance might be useful.
A larger amount of water falling a smaller distance might also be useful.
Immediately this becomes a site-specific engineering problem.
You can’t just say:
I’ve got a river, therefore I can generate 100 kW.
You need to understand the river.
Old industrial sites suddenly look interesting
Britain has spent a very long time using water for power.
Long before anyone was generating electricity from it.
Watermills used rivers to mechanically drive machinery.
That means there are places where people have already done some of the hard work.
Weirs.
Mill races.
Channels.
Existing drops in water level.
Not every old mill site is suddenly going to become a useful power station.
But I really like the idea of taking infrastructure originally built to mechanically power industry and adapting it to generate electricity.
Old energy infrastructure getting another job.
But you can’t just nick all the water
This is where the environmental side matters.
A river isn’t an industrial pipe.
It’s an ecosystem.
Fish need to move through it.
Water levels matter.
Flow downstream matters.
Removing too much water from part of the river could obviously cause problems.
So there are environmental constraints on how these systems can operate.
Which means the theoretical maximum amount of energy available isn’t necessarily the amount of energy you’re actually allowed to extract.
That’s an important distinction.
Then I start thinking about the data
I’d want:
Water level.
Flow rate.
Rainfall.
Upstream conditions.
Downstream conditions.
Turbine speed.
Generator output.
Bearing temperatures.
Vibration.
Efficiency.
Environmental restrictions.
Historical generation.
Then I’d start comparing them.
How does rainfall translate into generation?
How quickly does the river respond?
What does normal seasonal variation look like?
Can we predict tomorrow’s generation from upstream rainfall?
At what flow does the turbine operate most efficiently?
Now the little turbine sitting next to a stream starts producing quite an interesting dataset.
Small assets create a different data problem
This is something else I find interesting.
A 10 MW data problem and a 10 kW data problem can require remarkably similar infrastructure.
You’ve still potentially got sensors.
Telemetry.
Databases.
Dashboards.
Alerts.
Maintenance records.
Forecasting.
Except the economics are completely different.
You can’t build a £100,000 data platform to monitor an asset producing a relatively small amount of electricity.
So small-scale energy probably creates an interesting challenge for data engineering.
How do we make monitoring cheap enough that it’s actually worth doing?
Sometimes the clever answer might simply be collecting less data.
That’s something data engineers don’t say often enough.
So where is my head at?
Micro hydro isn’t going to power Britain.
But I don’t think that’s really the point.
If you’ve got a suitable site with reliable water flow, being able to generate electricity continuously from an existing natural process seems useful.
Particularly if existing infrastructure such as an old mill or weir can be reused.
It’s another reminder that energy generation doesn’t necessarily need to mean enormous centralised power stations.
Sometimes it’s a turbine quietly spinning next to a stream.
And I’d quite like to see the data coming off one.
That’s where my head is at the moment.
Gareth Winterman