Could Texas happen here?
American wind power was in the headlines in more ways than one this week.
From Saturday 13 to Wednesday 17 February, winter storm Uri raged across the United States, northern Mexico and parts of Canada, resulting in the largest blackouts in the U.S. since 2003, affecting over 5 million people in the U.S. and 4.7 million in Mexico.
The U.S. state of Texas was particularly hard hit, with its energy system clearly not prepared for a winter storm of this magnitude.
"We saw coal plants, gas plants, wind, solar, just all sorts of our resources trip off and not be able to perform," CEO Bill Magness of the Electric Reliability Council of Texas (ERCOT) was quoted as saying.
At the same time that energy supply systems began to fail, demand peaked as consumers tried to power up electric heaters en masse. When the imbalance in demand and supply started to distort the frequency of the grid, ERCOT saw itself forced to carry out controlled, rolling outages, effectively to cut demand.
While one Republican Texas senator sought comfort by flying to the southeastern coastal Mexican city of Cancún - thereby inviting significant public backlash -, other conservative operatives rushed to seek political gain by blaming renewable energy sources for the grid’s failure.
Investments of a different nature
For the most part, those claims were adequately debunked in mainstream media. Renewable energy sources such as wind and solar make up around 10 percent of Texas’ power supply and were not disproportionately more hindered by the storm than coal, gas or nuclear.
Some have pointed at overdue infrastructure upgrades. Despite recommendations to bolster the energy system after a wintry experience in Texas a decade ago, such investments were never made.
A decade later, however, those investments may well need to be of a different nature.
For one thing, due to climate change, exceptional and unpredictable weather phenomena such as heavy storms are likely to become more common. Power systems need to become more robust in withstanding extreme cold as well as extreme heat. At the same time, in more extraordinary circumstances, peak demand could reach new heights.
Renewable energy production may not be less robust in the face of extreme weather. But the fact that wind, solar, and battery reserves are about to make for an ever larger part of the equation does pose new challenges to how the grid is managed.
Where the data comes in
Not just climate change but also smart consumption could cause faster and higher fluctuations in demand. Next to that, renewable electricity causes lower inertia in the grid. This means that the system needs to be able to respond faster to changes in demand.
Oh and by the way, electricity has started flowing in both directions: from incumbent producers towards consumers, and from small local producers, or “prosumers”, back into the grid.
To top it off, energy markets are evolving towards a model whereby the price of electricity, which is based on demand and supply in the network, is being calculated with ever shorter intervals.
Yes, it’s important to insulate gas pipes, de-ice windmills, and remove snow from solar panels. But we also need a complete overhaul of intelligence in the grid. That’s where the data comes in.
Across the infrastructure we need sensors to collect data about weather conditions, frequency, supply, and demand. We need machine learning and artificial intelligence at the edge, in the cloud and on computing resources in-between in order to observe system failures, predict hardware maintenance, predict and compute the price of electricity, and actuate energy resources in real time.
What is our current state?
We address these challenges with RAIN, our horizontal, distributed computing platform.
But that’s not my main point here. What I want to know is: could Texas happen here in Europe? Could it happen in the Nordic-Baltic energy market?
Just a few weeks ago, the price of electricity in southern Sweden went through the roof as a result of a cut in nuclear power supply. Something similar happened in Finland in December.
Last month, apparently, a frequency disruption and a lack of operating reserves in France nearly caused a Europe-wide blackout.
I’ve been building a bit of a network of energy professionals lately. So here is my question to y’all:
What is the current state of distributed computing capabilities in the European and Nordic-Baltic power grids? Where are the bottlenecks? Who needs to infuse that data intelligence most urgently?
And if you don’t know, do you know who might?