Why capacity-market rules decide if your battery ever gets financed

I have a simple bias: if a market design produces a result that looks irrational, the design is usually doing exactly what it was set up to do. Not loudly. Silently.

The recent contrast between Belgium and Germany’s capacity market auctions is a perfect example. In Belgium’s latest capacity market auction, battery energy storage systems (BESS) won nearly 40% of contracted capacity. In Germany’s first capacity market auction, all awarded capacity went to gas, and the clearing price was nearly eight times higher than Belgium’s.

People will argue about “politics” or “readiness” or “grid needs”. That misses the practical point. Capacity markets do not just procure megawatts. They manufacture a cashflow shape. And that cashflow shape decides whether a battery is financeable long before anyone debates chemistry, augmentation, or inverter features.

If you develop projects, sell systems, or finance them, you should read these auction outcomes as a contract lesson: market rules determine which clauses you must demand upfront from your EPC and OEM to keep the asset bankable for ten-plus years.

The auction is a spreadsheet, and the spreadsheet is a verdict

Most capacity mechanisms look simple from the outside: offer capacity, clear a price, commit to availability, get paid. The trap is that “capacity” is not a physical property. It is a market-defined label that comes with conversion factors and punishments.

Four rule families quietly decide whether storage can compete, and whether lenders will believe the revenue:

• De-rating and deliverability rules: how the market converts your nameplate MW and MWh into eligible capacity.
• Duration requirements: how many hours you must sustain output to count as capacity.
• Availability measurement: what gets measured, when, and under which grid conditions.
• Penalty design: how missing performance translates into real cash losses and termination risk.

The Belgium vs Germany contrast is not “batteries vs gas”. It is “a set of rules that expresses batteries as dependable capacity” vs “a set of rules that expresses them as a discount and a liability”. The technology did not change at the border. The spreadsheet did.

Financeability is fragile, and the rules know it

Developers tend to model merchant upside carefully and treat capacity revenue as the stable leg. Financiers do the opposite. They want the stable leg to be stable even when reality misbehaves: forced outages, communication loss, temperature derating, warranty exclusions, auxiliary loads, late augmentations, grid events, curtailment.

Market design determines whether capacity revenue is “bond-like” or “option-like”. If it is option-like, you will get equity-style pricing on debt, if you get debt at all.

Here is the part many teams underweight: penalties are not just a performance incentive, they are a refinancing risk. The moment penalty exposure is unbounded or poorly correlated with what you can control, lenders stop treating the contract as infrastructure cashflow. They treat it as trading.

And batteries already carry enough endogenous uncertainty: degradation, augmentation timing, replacement lead times, BMS limits, and operational drift across seasons. If the market adds exogenous uncertainty through measurement and penalty rules, your “capacity contract” becomes a volatility product.

Prequalification is where markets quietly pick their winners

Most debates focus on the clearing result. The earlier filter is prequalification: what you must prove before you are allowed to bid, and what you must keep proving to stay compliant.

Prequalification sounds technical. It is commercial.

If the tests assume a generator mindset, storage gets squeezed in ways that are hard to explain to a credit committee. Typical pressure points include:

• Energy limitation treated as unreliability: a battery’s finite duration is a known boundary, not a surprise outage. If rules penalise it as if it were a trip event, you get punished twice: first in de-rating, then in penalties.
• State-of-charge obligations that ignore real operations: markets sometimes assume you can be full whenever needed. In real systems, you may be following dispatch, stacking revenues, or recovering from a prior event.
• Demonstration windows that select for conventional assets: if performance is measured in periods that correlate with worst-case temperature or constrained charging, you can be correct technically and still fail financially.

This is why the outcome matters. Belgium’s auction produced a material BESS share while Germany’s first did not, and the price divergence was extreme, nearly eightfold. That is not a footnote. It is a signal that rules can swing both technology mix and consumer cost.

The practical implication is uncomfortable: you cannot “engineer” your way out of an auction that has encoded a preference against your risk profile. You can only (1) avoid it, (2) price the risk so high you will not clear, or (3) push the risk into someone else’s contract, which leads to the next point.

The clauses you must demand before you order a single container

When market design makes revenue contingent on specific performance definitions, your project contracts must mirror those definitions. Otherwise you end up with the classic gap: the market penalises you for behaviour your OEM warranty does not cover, or for measurements your SCADA cannot prove.

Below are the contract clauses I would insist on for any capacity-market exposed battery project. Not as legal phrasing, but as non-negotiable topics that must be resolved before notice to proceed.

1) A shared definition of “available capacity”, with a measurement method

Your capacity obligation will be defined by the market. Your EPC and OEM obligations will be defined by their contracts. If these definitions do not line up, you will own the mismatch.

• Define the net deliverable power at the grid connection point, including auxiliaries.
• Specify telemetry, sampling, and data retention adequate to defend performance during disputes.
• Clarify who is responsible for compliance evidence during audits and stress events.

This theme overlaps directly with the question of who owns performance risk after the battery is commissioned. If you cannot point to a party that owns each failure mode, you are the default owner.

2) Penalty pass-through, capped and attributable

Capacity penalties can be existential. Your contracts must turn them into something insurable and attributable.

• Penalty pass-through must require causality, not vibes. If an outage is on the OEM, penalties flow.
• Set caps aligned with realistic worst-case penalty exposure, not just warranty limits.
• Include liquidated damages triggers that match market penalty triggers, especially around repeated non-availability.

If the market has asymmetric downside, your EPC and OEM agreements cannot be symmetric on paper and asymmetric in practice. Otherwise the SPV becomes the shock absorber for everyone else.

3) Degradation and augmentation as a capacity obligation, not a “nice-to-have”

In a capacity market, degradation is not just a performance drift. It is a compliance drift. A battery that can no longer satisfy duration, or must curtail power to protect the cells, is not just earning less. It may be non-compliant.

• Define an augmentation plan as part of the supply scope, with triggers tied to capacity qualification thresholds.
• Lock down lead times and logistics responsibility for augmentation equipment.
• Make the warranty speak the market’s language: duration at power, not only energy throughput.

4) Commissioning tests that resemble the market’s reality

If your acceptance test proves something the capacity mechanism does not care about, you have bought theatre.

• Commissioning must include event-style dispatch consistent with capacity activation patterns.
• Include tests for communications loss, fallback modes, and recovery.
• Ensure the test report produces artefacts you can submit for prequalification and re-testing.

I have watched one capacity-relevant requirement get discovered late, after hardware was already fixed and contractual levers were gone. It turned a clean project into months of renegotiation and finger-pointing, because the market asked a question the contracts never anticipated.

5) Grid-code evolution and retrofit responsibility

Capacity contracts are long. Grid rules do not sit still. If your market starts tightening inverter behaviour, fault ride-through, or grid-forming requirements, you need a plan that is more than “we will discuss it later”.

This is exactly why I wrote about grid-forming mandates turning inverter specs into financial risk clauses. Technical compliance becomes a funding question when your revenue depends on being considered “reliable capacity”.

• Specify who pays for mandatory firmware updates and retrofits.
• Define response times and support terms for regulatory-driven changes.
• Include a mechanism for re-qualification testing without destroying your availability metrics.

A decision lens for this quarter: stop treating “market risk” as abstract

If you build batteries into a capacity market, market design is not a slide in the risk section. It is a set of numbers that will show up in your debt terms, your DSCR covenants, and your insurance exclusions.

My opinion is blunt: if a capacity mechanism does not allow storage to compete on its real strengths, it is not “technology-neutral”. It is just neutral in wording while being selective in arithmetic. The Belgium vs Germany outcome should make more teams stop blaming chemistry and start reading auction parameters like a term sheet.

Here is the checklist I would use before committing development spend:

1. Translate the rules into cashflow volatility: model de-rating, duration, and penalties as a distribution, not a point estimate.
2. Map every penalty trigger to a contract owner: OEM, EPC, O&M, aggregator, or you. If it is “you”, admit it and price it.
3. Align acceptance tests to qualification tests: if you cannot prove it to the market, do not pay for it at handover.
4. Turn regulatory drift into a clause today: retrofits, firmware, and re-testing cannot be an email thread.

Storage is worth building when it changes something real for people: fewer curtailments, fewer peaker hours, lower bills, and a grid that can absorb electrification without drama. Capacity markets are supposed to buy that reliability. If the rules instead buy an expensive comfort blanket, you will see it in the clearing price, and you will feel it in every project finance negotiation afterwards.