Worked examples: a ledger that closes, and one that does not

Illustrative only

Every figure on this page is illustrative, chosen to show how the method works. None is a measurement of a real system.

Example A: the ledger closes

The accounts add up to the intake, so the residual is +0.00%, inside the 1.5% combined uncertainty. Overhead is sub-metered, which satisfies the rule that one account besides Intake is measured independently. Waste is Estimated, so the whole result is labeled Estimated, not Measured.

The ratios from Example A

• Rack energy at M3 over the same 24 hours: 980 kWh (illustrative).
• Overhead Ratio: OR(M4/M3) ↓ 1.22.
• Decision Yield with 2,400,000 passing decisions: DY(M3) ↑ 2,449 decisions per kWh.

Example B: the ledger does not close

Here the accounts exceed the intake by 60 kWh, a residual of -5.26%. That is outside the uncertainty, so the result is reported as not closed.

Diagnosing the residual

1. Check that every meter covers the boundary it claims. Here, the intake meter sits after the uninterruptible power supply (UPS).
2. Check that the accounts sit inside that boundary. The UPS conversion losses, 60 kWh, were booked to Overhead, but they happen upstream of the intake meter.
3. Fix the boundary, not the numbers: move the intake meter upstream of the UPS, or remove upstream losses from Overhead and report them separately.
4. Re-run the closure check. With the boundary fixed, the ledger matches Example A and closes.

What this shows

Most ledgers that fail do so because of a boundary error, not arithmetic. That is why JL-26 makes the meter point part of every result.

Questions

Keep reading