This article was generated with AI assistance from cited sources and has not been individually reviewed by an editor.
How you pay for service determines who carries the availability risk — the risk that the asset is not earning when it should be — and who carries the augmentation cost as the battery degrades. Here are the three structures and what they actually mean.
You pay for each intervention as it happens. It looks cheap at signing because there is no standing fee. Who carries availability risk: you. Every minute of downtime is your loss, and the vendor has no financial incentive to prevent failures — each one is revenue for them. Who carries augmentation cost: you, at market rates, when the day comes. Per-call works for small, non-critical assets with low downtime cost. It is the wrong structure for a revenue-generating C&I asset, because it misaligns every incentive.
A fixed recurring fee buys defined scope: scheduled maintenance, a response-time commitment, and typically a performance guarantee. Who carries availability risk: shared, but moving toward the vendor if the agreement includes availability or performance guarantees with penalties. Who carries augmentation cost: defined in the agreement — the best LTSAs pre-agree the augmentation price and chemistry. This is the structure most institutional owners and investors will require, because it converts an unpredictable operating cost into a predictable one and gives the vendor a reason to keep the asset healthy.
Not a separate commercial model but a clause that can sit inside an LTSA: the vendor guarantees a minimum availability or a minimum retained capacity over time, with a defined remedy — typically a rebate or a free augmentation — if they miss. Who carries availability risk: the vendor, within the guaranteed envelope. Who carries augmentation cost: the vendor, if degradation exceeds the guarantee. The catch is the small print: the guarantee’s baseline cycling assumption. If the vendor assumes a gentle cycle profile and you run the asset hard, the guarantee may not bite. Demand a guarantee written against your actual duty cycle, not a lab profile.
The EPC (engineering, procurement and construction) contractor is built to build, not to operate. Its in-house service offer often looks attractive at handover — the same team, the same site knowledge — and then thins out: the construction team moves to the next project, the spare-parts stock was never really stocked, and the response-time “commitment” was a handshake, not a contract. By year two, the owner discovers the EPC’s service is an afterthought bolted onto a construction business. The lesson is not that EPCs are dishonest; it is that service is a different business from construction, with different incentives, and you should contract it as such — either a dedicated service provider or an EPC whose service arm is a separate, capitalised, contractual entity.
Ask three questions and the structure chooses itself. One: what does an hour of downtime cost this asset? If the number is material, you need an LTSA with a performance guarantee, not per-call. Two: who do I want carrying augmentation risk as the battery degrades — me, at market rates in year five, or the vendor, at a pre-agreed price? Three: can the vendor show me a service bench and spare-parts stock in-country, or am I buying a promise? In Belgium, the Netherlands and Germany — where the capacity tariff, congestion contracts and the AgNes grid-fee reform are all live variables in 2026 — the answer to question three is the one that tells you whether the other two can be answered honestly at all.
No manufacturer or EPC reviewed this guide before publication. Corrections are published, marked, within 48 hours of verification. Sources: Fluvius tariff sheets (2026); TenneT/ACM Sequoia announcement (April 2026); BNetzA AgNes timetable (2026).
