Every answer and its explanation appears here once you have finished the path. Each one then links to the matching glossary entry, where the concept is set out in full with its worked example.
1. A fifteen year availability guarantee given by a supplier carries amounts comparable to a policy's. What nonetheless keeps it from being insurance?
It covers a promise rather than a fortuity: the supplier committed to a measurable result and owes the difference, with no chance transferred from its own standpoint
The difference is one of nature, not form, and it fits in a sentence: insurance indemnifies damage suffered, a performance guarantee pays the gap between a promised result and a measured one. The three other answers name accurate consequences of that nature and mistake them for the cause. The absence of authorization is an effect: no authorization is needed to promise an output. The absence of pooling and of premium describe the arrangement correctly, and it is precisely because no fortuity is transferred that there is no pool to build and no premium to charge. Taking any of those three signs for the test leads to believing that a guarantee large enough, pooled enough or priced enough would end up being insurance, which never happens.
Glossary entry →2. A guaranteed output is measured under reference conditions that reality never reproduces, so measurements must be corrected. Where does the difficulty this mechanism creates lie?
The correction curves are supplied by the very party that owes the guarantee, so the debtor provides the instrument that measures its own debt
The measurement mechanism decides everything and is more disputed than the principle, because correction brings a real measurement back to contractual conditions through curves the supplier established. There is no dishonesty here, only a structural asymmetry: nobody else knows the machine well enough to produce them. The answer entrusting reference conditions to an independent body describes an arrangement that exists for other purposes, type certification for instance, and applies it where it does not operate. The one invoking statistical uncertainty generalizes into a rule something met on very small gaps. The answer requiring twelve months of continuous measurement imports a requirement specific to AVAILABILITY guarantees, which are indeed measured over a long window, whereas an output is measured on a test.
Glossary entry →3. An operator has received availability penalties two years running on a fifteen year contract. What separates its position from that of an insured having suffered two losses in two years?
The cap of a performance guarantee is spent for good, whereas an insurance limit reinstates or is renegotiated each year: the operator has spent an asset, the insured has used capacity
This is the least understood feature of these guarantees: a cap that gets spent looks like a limit until it has been drawn on. Two years of penalties on a fifteen year contract are not two years of losses on an annual policy, because the former permanently remove protection from the thirteen remaining years. The answer seeing no structural difference holds a market truth, losses are always paid for, and misses that one is paid in a negotiable future premium and the other in protection already lost. The one putting the operator ahead names a real advantage, no cancellation, and sets it against a drawback that weighs far more. The answer favoring the insured states an automatic and free reinstatement that is not the rule: reinstatement is stipulated, and it is paid for.
Glossary entry →4. An availability guarantee runs fifteen years, longer than the construction policy, the maintenance period and the decennial. What risk does that duration create, and how is it handled?
Counterparty risk: a fifteen year promise is worth only the balance sheet of whoever carries it, hence frequent backing by a bond or a parent company guarantee
A guarantee often outlives whoever gave it, and that is what turns it into a promise rather than a protection. The remedy is known and leads back to the module on bonds: the undertaking is backed by a third party whose balance sheet carries it. The answer on obsolescence describes a real commercial nuisance, with no bearing on the enforceability of what was promised. The one on limitation confuses the duration of the undertaking with the period to sue, which runs from each breach and not from signature. The answer about accumulation invents a mutual cancellation between two regimes bearing neither on the same facts, availability on one side, impairment of structural integrity on the other, nor on the same debtors.
Glossary entry →5. A machine repaired after covered damage does not recover its output. The supplier refuses its guarantee, the insurer objects that underperformance is not damage. What must have been done not to be left in between?
Having recorded performance before and after each repair, those measurements being the only usable evidence to attach the shortfall to the damage rather than to the design
Both refusals hold on their own, which is why the way out is evidentiary rather than legal: one must be able to say where the shortfall comes from, which requires a before state. The step costs an engineer a day and can never be recovered afterward. The answer on a jointly appointed expert describes good practice that settles the record of the works, not the link between the shortfall and its cause, and it assumes besides that both agree to take part. The one on prior written agreement is the most attractive because it looks prudent, and it obtains only an undertaking whose scope will be argued exactly like the rest: maintaining a guarantee does not say at what level of output. The answer notifying a second claim recharacterizes a disappointment through notification, which does not change the nature of the fact notified.
Glossary entry →