The module on the testing period deals with time: when the window opens, how it is counted, when it closes. It does not say what is being tested, and that is a gap because the word testing covers three operations with neither the same object, nor the same risk, nor the same fate in insurance. Confusing them leads to believing covered what is not, and above all to placing in the wrong spot the moment when the project becomes dangerous.
The first tier is mechanical testing. One verifies that the installation is correctly erected: circuit continuity, cold pressure tightness, no-load rotation, response of safety devices. Nothing is produced yet, the energy involved stays low, and the nature of the risk is that of erection. It is the only tier where an incident generally stays local, and it is also where the least costly defects are found, which explains why a mechanical campaign rushed to save three weeks almost always costs more later.
The second tier is commissioning testing, where the installation receives for the first time what it is built to transform. That is the tier the testing period module described: protections partly inhibited, provisional settings, the first ramp-up. The risk changes nature there because the energy becomes real, and a difficulty of its own is added: at that stage the installation still belongs to the contractor in risk terms, but it is operated by mixed teams in which the operator's staff are learning. Responsibility for an incident is then argued between two organizations neither of which has exclusive control.
The third tier is performance testing, and it is of an altogether different nature. One no longer verifies that the installation works, one measures what it produces: a flow rate, an efficiency, a specific consumption, an availability over a period. The stake is no longer safety but the contract, since those measurements trigger acceptance, release payments, and determine whether performance liquidated damages are due. A failed performance test is almost never a loss: it is a contractual non-performance, and the policy has no part in it.
That distinction is the module's most useful, because it separates two flows of money that projects mix up. Equipment that breaks during a test comes under the policy, on the conditions described in the testing period module. Equipment that does not break but produces ten percent less than guaranteed comes under the supply contract, and is settled in performance damages or a price reduction. The first is damage, the second a disappointment. Claiming the second from an insurer is the same mistake as claiming unforeseen physical conditions, described in the public project owner module.
A gray zone nonetheless exists, and it carries serious amounts: equipment that fails to reach its performance because it was damaged during an earlier test. The performance shortfall then has a covered material cause, and the question becomes one of linkage, that is, exactly the reasoning of the causal role module applied in the other direction. It must be established that the shortfall arises from the damage and not from a design that would not have kept its promises anyway, and that demonstration rests on measurements taken before the damage, which nobody thinks to preserve.
The link with delay cover deserves stating plainly to finish. A failed performance test requires rework, modification, then retesting, which pushes commissioning back exactly as damage would. But delay cover requires covered physical damage at the origin of the slippage, and underperformance is not damage. The project therefore suffers a real, quantifiable delay for which no indemnity will come, and whose only possible compensation is contractual, from the supplier. It is the same configuration as the bad weather described in the penalties module: the delay exists, it costs, and insurance has nothing to do with it.
A separation unit must reach 940 tonnes per day and a guaranteed specific consumption. Mechanical testing ends in January. In March, during commissioning testing, an overpressure damages an exchanger, repaired for 1.2 million euros paid by the policy, and testing resumes in June. In September, the performance test measures 880 tonnes per day, that is 6.4% below guarantee, and specific consumption exceeds by 3%. The supplier argues that the repaired exchanger has not recovered its original characteristics. The project owner suffers six further months of slippage and claims from its insurer both the production shortfall and the delay. What can it obtain?
The three tiers must be sorted before answering, or everything blends together. The March overpressure is physical damage occurring during commissioning testing: the policy responded, 1.2 million settled, that point is closed. The 6.4% shortfall measured in September is of another nature: nothing is broken, the installation works and produces less than promised, which is a contractual non-performance and not a loss. Claiming it from the insurer is the file's mistake, and the compensation, if any, lies in the supply contract's performance damages. That leaves the gray zone, which is the real subject and which the supplier itself opened: if the repaired exchanger has not recovered its characteristics, then the shortfall has a covered material cause, and one falls back on the causal role reasoning, applied in the other direction. The demonstration requires comparing performance before and after the damage, therefore having the January and March measurements; without them, nobody will be able to tell a badly repaired exchanger from a design that would not have held its 940 tonnes anyway. As for the six months of slippage, they arise not from the March damage but from the September test failure: delay cover requires physical damage at the origin of the slippage, and underperformance is not damage. The delay is real, it costs, and no indemnity will come.
- 01Three tiers under one word: mechanical testing, where nothing is produced yet; commissioning testing, where energy becomes real; performance testing, where what the installation produces is measured.
- 02A failed performance test is almost never a loss: it is a contractual non-performance, settled in performance damages and not with an insurer.
- 03Breaking and underproducing are two different flows of money: the first is damage, the second a disappointment, and projects constantly mix them.
- 04The gray zone is equipment that underproduces BECAUSE it was damaged: the demonstration rests on measurements taken before the damage, which nobody thinks to preserve.
- 05A performance failure pushes commissioning back without opening delay cover, for want of physical damage at its origin: the delay exists, it costs, and insurance has nothing to do with it.