SAIDI and SAIFI look simple: minutes and interruptions per customer served. In practice both numerator and denominator are contested. Which outages count? Planned or unplanned, transmission-caused or distribution-caused, above or below the reporting threshold? And how many customers were actually on the affected feeder at the time?
While these indices were internal management figures, the ambiguity was tolerable. Under a concession or private-operator arrangement they become performance obligations with financial consequences, and every ambiguity becomes a dispute.
Traceability is the requirement
The standard that matters is traceability: any reported index should decompose into a list of outage events, each tied to a feeder section, a cause code, a restoration time and the customer count actually connected downstream. If the asset register cannot supply that customer count per feeder section, the index cannot be audited — only asserted.
This is why reliability reporting is downstream of the asset register. An operator inheriting a district needs the register verified at handover, because it fixes the baseline against which its performance will be judged for the life of the concession.
Regulators benefit from the same discipline. A monitoring regime built on traceable indices lets the regulator engage with causes — conductor condition, transformer loading, vegetation — instead of arguing about arithmetic. The conversations get harder and more useful at the same time.