Demand
scales the load the system has to serve. Raising it increases residual load directly and pushes the marginal technology up the cost order.
European power markets
Stress-test European power markets under changed assumptions
Move demand, renewable generation, thermal availability and cross-border stress, and read the direction and size of the resulting system pressure. Illustrative pressure — not a price forecast.
Every view of a power market rests on assumptions: that demand will behave as it usually does, that the wind will blow about as much as it normally does, that the thermal fleet will be about as available as it was last month. A stress test is what you do when you want to know which of those assumptions the market is actually leaning on.
PowerUnits lets you move four of them at once and read what happens to system pressure.
Definition
What a power market stress test is
A power market stress test is a structured analysis of how the balance between electricity supply and demand responds when assumptions about demand, renewable generation, thermal plant availability, fuel and carbon costs, or cross-border exchange are moved to extreme but plausible values. It describes the direction and the relative size of the resulting pressure on the system.
A stress test is not a forecast. It carries no probability, and it does not produce a traded clearing price. Its purpose is to make the consequences of an assumption visible, so that the assumption itself can be argued about.
PowerUnits implements this as an interactive readout across 17 European markets. Four assumption levers move a heuristic pressure band and a set of directional indicators. The output is deliberately unitless.
Worked example
A worked stress test
The configuration below is a single, deliberately severe assumption set. It is not a description of any real market event, and it is not dated to a market period, because the readout does not depend on one — the stress test compares a configuration against its own neutral baseline, not against a historical outcome.
| Assumption | Change from baseline |
|---|---|
| Demand | +8 % |
| Renewable generation | −15 % |
| Thermal availability | −10 % |
| Import and export stress | +5 % |
Heuristic readout
- System pressure band
- Stressed
- Residual load
- ↑ Up
- Renewable share
- ↓ Down
- Thermal dependency
- ↑ Up
Evidence type: Illustrative heuristic · Source: First-party PowerUnits model · Period: Not period-bound · Analytical status: static but dated, manually reviewed · Methodology version: sandbox_toy_v1 · Generated 2026-08-03
Reading the result: the band summarises how far the configuration sits from a balanced system, and the indicators say which way each component moved. Neither is a quantity. Two configurations that produce the same band are not equivalent markets — they are two assumption sets that this heuristic scores similarly.
Assumptions
The four levers and what each one moves
Renewable generation
scales the output of variable renewables. Lowering it raises residual load without changing demand — which is the distinction that makes residual load a more useful stress variable than demand alone.
Thermal availability
scales how much of the dispatchable fleet is assumed to be usable. It does not change what the system needs; it changes what is available to meet it.
Import and export stress
represents constrained cross-border exchange. A market that can normally lean on its neighbours is a different market when it cannot.
Fuel and carbon anchors
sit alongside the four levers. They change the order in which thermal technologies would be called, not the pressure band. They are labelled as illustration throughout, because the cost ordering they produce is a comparison, not a market outcome.
Coverage
Which markets this covers
The pressure readout is available for 17 European markets: Germany, France, the Netherlands, Belgium, Spain, Italy, Poland, Sweden, Austria, Czechia, Finland, Hungary, Slovakia, Romania, Switzerland, Norway and Great Britain.
Additional fleet context — a hypothetical capacity stack you can adjust alongside the stress levers — is available for five of those markets: Germany, France, the Netherlands, Belgium and Spain. Where it is not available, the stress levers still work; the fleet context simply does not appear.
Coverage is surface-specific. A market being available here does not mean historical analysis, scenario history, or asset data is equally complete for it.
| Market | Stress readout | Capacity context |
|---|---|---|
| Germany | Available | Available |
| France | Available | Available |
| the Netherlands | Available | Available |
| Belgium | Available | Available |
| Spain | Available | Available |
| Italy | Available | Not available |
| Poland | Available | Not available |
| Sweden | Available | Not available |
| Austria | Available | Not available |
| Czechia | Available | Not available |
| Finland | Available | Not available |
| Hungary | Available | Not available |
| Slovakia | Available | Not available |
| Romania | Available | Not available |
| Switzerland | Available | Not available |
| Norway | Available | Not available |
| Great Britain | Available | Not available |
Methodology
How the pressure readout is produced
The readout is a deterministic heuristic evaluated in the browser from the assumption values you set. It does not query market data, and it does not look up what actually happened in any period. That is a deliberate design choice: a stress test should answer what does this assumption imply, not what happened last Tuesday.
Relaxed, Neutral and Stressed are heuristic bands, not probability classes and not risk ratings. The directional indicators for residual load, renewables and thermal report the sign of the change the configuration produces.
Where the workspace shows a cost ordering in €/MWh, it is labelled illustrative €/MWh_el and is explicitly not a clearing price. PowerUnits does not publish a calibrated price effect for a stress configuration, because it does not have one.
Limitations
What a stress test cannot tell you
It will not give you a price.
No configuration produces a forecast €/MWh, an expected value, or a traded clearing price. If you need a number to put in a position, this is the wrong tool and PowerUnits will not pretend otherwise.
It will not give you a probability.
Nothing in the readout says how likely the configuration is. Plausibility is your judgement, not the model's.
The fleet context is hypothetical.
Where you adjust capacity alongside the stress levers, you are adjusting an illustrative stack, not the real registered fleet of that country. A hypothetical fleet is not the real country baseline.
The bands are ordinal, not metric.
Stressed is more pressured than Neutral. It is not twice as pressured.
Two markets' bands are not directly comparable.
The heuristic scores a configuration against its own baseline, so identical bands in two countries do not describe identical physical conditions.
FAQ
Common questions
What is a power market stress test?
A structured analysis of how the balance between electricity supply and demand responds when assumptions about demand, renewable generation, thermal availability, fuel costs or cross-border exchange are moved to extreme but plausible values. It reports the direction and relative size of the resulting system pressure, not a price.
Does a stress test predict electricity prices?
No. The PowerUnits stress test produces an ordinal pressure band and directional indicators. It does not produce a forecast price, an expected value, or a traded clearing price, and no calibrated price model stands behind it.
How is a stress test different from a scenario?
A stress test pushes a small number of assumptions to extreme values and reads the immediate pressure that results. A scenario builds an internally consistent set of assumptions and follows it over time. Stress tests answer "what if this breaks"; scenarios answer "where could this go".
Which European markets can be stress-tested?
Seventeen: Germany, France, the Netherlands, Belgium, Spain, Italy, Poland, Sweden, Austria, Czechia, Finland, Hungary, Slovakia, Romania, Switzerland, Norway and Great Britain. Additional hypothetical fleet context is available for five of them.
Is the stress test based on live market data?
No. The readout is a heuristic evaluated from the assumptions you set. It does not query market data and it is not tied to a market period. PowerUnits' historical surfaces, which do use market data, are separate and are not real-time either — data reaches them with a pipeline lag.
Can I see the numbers behind the pressure band?
The band is deliberately unitless, because a number would imply a precision the heuristic does not have. The workspace shows the directional components and, where available, an illustrative thermal cost ordering that is explicitly not a clearing price.
Stress Test
Stress-test European power markets under changed assumptions
Move demand, renewable generation, thermal availability and cross-border stress, and read the direction and size of the resulting system pressure. Illustrative pressure — not a price forecast.
Open the Stress-Test workspace