Passive Balancing Triggering Unprecedented Dual Pricing in Dutch Balancing Markets

Passive Balancing Triggering Unprecedented Dual Pricing in Dutch Balancing Markets

NL: Dual pricing has been the subject of many conversations over the last weeks. Our research has shown it is not happening more often that upward and downward balancing prices split, but the spreads have become bigger over time. What is causing this?

This question has been the topic of many conversations lately. And we have always replied that it is caused by "passive balancing" which is all but passive, it is parties actively dispatching assets in response to extreme prices in the balancing market. No data exists for this and the generation data that is published by TSO's is too chunky to analyze this.

We did give it a go and it turned out to be an interesting exercise!

The aFRR activation down, purple area is increasing in the beginning of a quarter, reaches a minimum level, and is positive at the end of the quarter. This patterns repeats itself for 15 times!

NRV Sensitivity Analysis

Every time the aFRR downward activation was more than 50 MW, the expected prices in that quarter hour became negative. This created a trigger to curtail renewables or ramp down power plants.

But here is the interesting fact. Just a couple of minutes after the aFRR down reaches it’s minimum level we observe an extreme upwards activation of IGGC. Which means that either Belgium or Germany is netting off an opposite balancing position to the Netherlands to solve a shortage on the grid. This means the system has become short!

We have calculated an 'implied passive balancing volume' based on the above principles and it leads to the below visualization.

Conclusion, passive balancing is too aggressive, so much so. that an oversupply immediately triggers a shortage in minutes. The implied volume is pretty high, so that it flips the balance of the grid immediately and the TSO is forced to deactivate its reserve activations.

Why is this pattern repeating 15(!) times?

  1. It’s obvious that the mistake is known at the end of the quarter (regeltoestand 2).
  2. The 'passive' action is reversed to prevent the losses.
  3. Resulting again in an oversupply situation at the begin of a next quarter hour.

We would advise caution in 'chasing' the imbalance signal with flexible assets, there are clear patterns in which the risk of dual pricing is not as high as in these examples.

*) Analysis by Jean-Paul Harreman and André Bosschaart

Julien Jomaux

Expert in energy, electricity, renewables, and markets

9mo

Great insights !

Like
Reply
Lars Weber

Partner, Business Development at Yggdrasil Commodities

11mo

Jean-Paul Harreman, André Bosschaart, great analysis. Indeed the passive balancing has become larger than the expectation was from TennetNL. In addition, there might be market participants that are reacting via passive imbalancing slower than others. This reaction could also be deliberate. I can see an algorithm saying " if there have been large downward regulation prices early this quarter, please shut down as I will make more money on the shutdown in the last minutes of this quarter than by producing". But if this algorithm has too much volume, the market flips and thus the market is short again. As this flip was done very quickly, TennetNL will request upward regulation for the rest of the quarter via IGCC. This creates a "see-saw" motion that repeats each quarter....

Like
Reply
Mayk Thewessen

Energy Markets, Batteries & PPA bij Ventolines

11mo
Like
Reply
Robert Kleiburg

Managing Director EFS

11mo

Good analysis Jean-Paul Harreman!

Like
Reply

To view or add a comment, sign in

More articles by Montel Analytics

Insights from the community

Others also viewed

Explore topics