Who owns performance risk after the battery is commissioned?

Most battery storage conversations still start with the box. Container count. MW. MWh. Delivery date. It sounds concrete, and it gives everyone a comforting feeling that the hard part is steel, cells, and logistics.

But a BESS is not a box. A BESS is a grid contract with performance clauses, metered in five minute slices, enforced by alarms, and punished when reality diverges from the spreadsheet.

Poland is a clean case study because the market is moving fast and the projects are big enough that the “who holds the risk” question cannot be hand-waved away. In one update, you can see both sides of the story: a developer selling a pipeline and a utility building a flagship system. Greenvolt selling 1.6GWh of BESS projects and PGE building an 800MWh battery project using Fluence technology are headlines about assets. The bankability, though, lives in the risk allocation that starts the day after commissioning.

If the OEM and EPC stop at delivery and “it passed the tests,” then the owner, utility, or aggregator inherits the messy layer: interconnection behavior, availability, degradation, market dispatch, telemetry quality, cybersecurity, maintenance discipline, and penalties when any link breaks. In storage, the party that owns round trip performance in real grid conditions ends up dictating cadence, process, and ultimately financing terms. Not because they are the most important person in the room, but because they are the one left holding the invoice when performance slips.

The container is easy. The interfaces eat your margin.

There is a pattern I have seen across electrification programs: margins do not evaporate because the core technology does not work. They evaporate because “interfaces” were treated as someone else’s problem.

Storage has more interfaces than most teams admit:

• Grid interface: compliance, protection settings, fault ride-through behavior, reactive power, ramp limits, and the endless negotiation of what “acceptable” looks like.
• Market interface: what the asset is allowed to do, how it is dispatched, and what happens when the market signal conflicts with the physical limits of the battery.
• Data interface: metering, timestamps, SCADA, latency, missing points, and whether your data is credible enough to defend a dispute.
• Performance interface: availability definitions, auxiliary consumption, temperature effects, and the gap between warranted numbers and real operation.
• Responsibility interface: when something goes wrong at 02:00, who is obligated to act, and who merely “supports.”

Everybody loves to sell the part they can control: hardware, EPC scope, a software license, a warranty document. The value, and the pain, sit in what happens between those scopes.

That is why I do not get excited when someone tells me they “shipped the container.” I want to know who priced the grid risk, and who is contractually on the hook for it after handover.

Bankability is a risk story, not a technology story

Financiers do not fund batteries because they admire chemistry. They fund them because they believe the cashflows survive contact with the grid.

The difference between a project that closes and a project that stalls is often one sentence in an agreement: who carries performance and integration risk once the asset is live.

From the outside, two deals can look identical. Same MW. Same MWh. Same supplier brand. Same site. But inside the contracts, one has a clean chain of obligations and the other has a pile of “reasonable endeavors.”

When risk is not priced, it does not disappear. It is simply dumped onto the party least able to manage it, usually the asset owner. That owner then compensates the only way they can: higher contingency, more conservative dispatch, slower ramp-up, and tougher financing terms. In other words, the project becomes more expensive precisely because nobody wanted to be explicit.

That is also why Poland matters beyond Poland. When markets accelerate, sloppy risk allocation gets exposed faster. Big pipelines and big utility projects, like those described in this Poland storage update, push the industry toward grown-up answers: not “can we build it,” but “who owns the guarantees after the ribbon cutting.”

The real product is “live performance under penalties”

If you want one reframing to use this quarter, make it this:

A storage project is a service obligation disguised as a capital asset.

The obligation is continuous: deliver energy when asked, absorb when asked, stay within grid limits, and prove it with trustworthy data. If you cannot prove it, you might as well not have done it.

So the question I press on, whether I am looking at a developer pipeline sale or a utility-owned build, is simple: what happens after commissioning, when the asset must perform every day?

Here are the failure modes that show up in real life, and why they matter commercially:

• Availability that looks good until you define it. Is the battery “available” when it is online but derated? When it cannot meet ramp rates? When auxiliary systems trip? A loose definition turns into a fight the first time revenue is missed.
• Degradation that is “normal” until it is expensive. Cycling profile, temperature, and dispatch choices change degradation. If one party controls dispatch but another carries degradation risk, incentives will clash.
• Telemetry that works until it is used as evidence. Bad timestamps, missing points, or inconsistent meters create disputes. Disputes delay payment and destroy trust. If you have ever tried to reconcile two systems that both claim to be the truth, you know how fast weeks disappear.
• Interconnection behavior that passes tests but fails in edge cases. The grid does not care that you passed commissioning. It cares what happens on a bad day when voltage dips and the network is stressed.

None of this is exotic. It is basic integration discipline. The problem is that many teams treat it as “post-project,” when in reality it is the core of the product being financed.

A practical checklist: make risk ownership explicit before steel arrives

If you are buying, selling, financing, or building BESS projects, I would force five questions into the earliest commercial conversation. Not as legal fine print, but as the spine of the deal.

1. Who is accountable for end-to-end performance?

   Not “who supplies the PCS” or “who runs commissioning.” Who is obligated to deliver the warranted round trip efficiency, response time, and availability in real operating conditions, over time?

2. What are the measurable service levels, and what is the evidence source?

   Define the KPIs, define the meter, define the clock. Decide which system is authoritative. If you do not do this, you are choosing future disputes.

3. Who controls dispatch, and who pays for degradation?

   Split control and cost, and you will get irrational behavior. Align them, and you can optimize like adults.

4. What is the escalation path when performance drops?

   Not “open a ticket.” I mean time to acknowledge, time to remediate, and who has remote access to do what. If nobody can act fast, penalties will educate you.

5. What does “handover” really mean?

   Handover is not a ceremony. It is a shift in risk. Write down what is still supported, what is warranted, and what is now the owner’s problem. Then price it.

These are not administrative questions. They decide whether the asset behaves like infrastructure or like a science project.

My bias: the winner is the party that stays on the hook

I have watched electrification programs bleed because integration risk was treated as “coordination,” which is a polite word for “unowned.” Once the system is live, unowned risk does not stay abstract. It becomes overtime, emergency site visits, strained relationships, and financial surprises.

Storage will reward the teams that stop selling batteries and start selling dependable grid performance. That can be done by a developer, a utility, an integrator, or a platform. But it requires one uncomfortable move: someone must keep responsibility after commissioning, not just a warranty document, and price that responsibility like it is the real product.

My opinion is blunt: if your commercial model ends at delivery and tests, you are not building a durable business in grid storage. You are pushing the hard part downstream and hoping the market does not notice. Poland, and markets like it, will notice quickly.

If you want a parallel mental model, ask yourself a simpler question: would you finance an asset when the critical dependency has no single throat to choke? That is why I keep coming back to risk ownership, and why I linked Would you finance an offshore data center on satellite and batteries? in other conversations. The financing logic is the same even when the asset looks different.

The closing takeaway is not “write better contracts.” It is this: decide who owns live performance, give them the tools and authority to manage it, then let every other scope align behind that. The grid does not pay for containers. It pays for kept promises.