Skip to main content
Pallara
Menu
← Back to overview

Inspect a sample finding

Synthetic records. Actual engine results.

A power bill.
An answer you can trace.

Follow four fictional billing reviews from source records to a useful next step. Inspect a discrepancy, a clean result, a contract minimum, or the missing evidence that stops a calculation.

Explore four recorded outcomes

Supported line differences

Two differences disappear in the total.

Energy is billed too high. Demand is billed too low. The audit preserves both directions so the review starts with the whole record.

Invoice total
$223,000
Reconstructed total
$223,000
Scope
September 2026 / Fictional month / 2,880 quarter-hour readings

What this lets a reviewer do

Bring both differences to the supplier.

Use the charge trace to ask why energy is above the contract calculation and demand is below it. The matching total alone cannot resolve that question.

Reconciliation / Four charges

Select a charge to trace it
Billed and reconstructed charges. Positive difference means billed above the calculation; negative means billed below.
ChargeBilledRebuiltDifference
$150,000$144,000+$6,000

Trace / Energy

2,880 MWh × $50/MWh = $144,000

Contract §1 + meter intervals

$66,000$72,000−$6,000
$1,000$1,000Agrees
$6,000$6,000Agrees

Difference = billed minus reconstructed. Both overbilling and underbilling remain in the record.

No net arithmetic benefit to claim. Neither difference is recovered cash.

Take the evidence into your own review. Synthetic sources and actual engine output, together.

Four recorded engine outputs from fictional contract, meter, and invoice inputs. The controls select retained cases; they do not run an audit of your site. These unsigned examples do not demonstrate document extraction, curtailment-credit calculation, or recovered cash. Verify this case ↗ Source versions ↗

Turn the example into your question.

Describe the charge you need explained, the decision it affects, and when the answer would help. Prepare a brief for your team or Pallara.

Discuss a similar question →

Start with a simpler, one-charge calculation

One fictional day, 96 meter intervals, and a fixed rate. Follow the calculation from source to finding.

The accepted source records

Metered energy

96 readings / 15-minute intervals

Inspect meter data ↗
96,000kWh

Contract rate

Fixed energy charge

Inspect the term ↗
$50per MWh

Invoice charge

Separate from the calculation

Inspect the invoice ↗
$5,000billed

Reconstructed charge

$4,800

96 MWh × $50/MWh

Follow the calculation ↗

$200 supported difference

The invoice exceeds the reconstructed charge. The finding carries the source records and calculation for review. It is not evidence of a refund or recovered cash.

One synthetic energy charge for one day. This example does not calculate curtailment credits or represent a complete customer audit. The preview is unsigned. Its calculation, source hashes, and workbook were checked for reproducibility. Inspect the replay receipt ↗

When does the evidence fail to support a finding?

Same arithmetic. Different evidence.

This separate synthetic example uses a monthly fixed service charge. The bill is $1,200 and the comparison input is a prepared $1,000, not a calculation performed by this example. Change the contract evidence to inspect four recorded audit outcomes.

Change the contract evidence

Source 01 / Invoice

Same in every case

Monthly fixed service charge

Aug 1, 2026 to Aug 31, 2026

$1,200

Observed billed amount

Inspect invoice source ↗

Source 02 / Contract term

Fixed monthly service term

Effective from
Aug 1, 2026
Effective through
Aug 31, 2026
Prepared comparison input
$1,000
Inspect contract source ↗

Audit result

$200 supported difference

A supported discrepancy, not recovered cash.

The cited term covers the full invoice period. The audit accepts the prepared comparison and records a supported difference. Recovery still requires a separate resolution and posted cash.

Contract-supported comparison: $1,000

What this sample proves and how to verify it

One normalized fixed-service-charge fixture with a prepared comparison amount. No source extraction, automatic charge calculation, interval analysis or curtailment-credit calculation is exercised. A supported discrepancy does not establish a refund entitlement, customer savings or cash recovery.

These are four recorded outputs from Pallara’s audit engine. Selecting a case changes the evidence and result shown; it does not run a new customer audit.

Each exported case passed input replay, workbook byte comparison, and signature verification. The sample signing key establishes file integrity, not customer identity or independent assurance.

Explore another question: does a lower peak mean a lower bill?

Try the evidence / Fictional scenario

Lower peak. Lower bill?

Same monthly energy. A load shift cuts the peak from 8 to 6 MW. Change the historical peak to see what the contract changes.

Prior 11-month peak

The assumed contract bills at least 80% of that history.

Modeled change in specified monthly charges

$9,272.79lower

The lower peak reduces the billed demand.

Energy charge
$160.00 higher
Demand charge
$9,432.79 lower
Why the answer changes

7.5 MW × 80% sets a 6 MW floor. Billing demand falls from 8 MW to 6 MW when the load shifts.

Specified charges / 7.5 MW history
ChargeOriginalShifted
Energy$287,760.00$287,920.00
Demand$37,731.15$28,298.36
Total$325,491.15$316,218.36

Both schedules use 2,884 MWh. Shifting 2 MWh from $20/MWh to $100/MWh raises energy charges. Demand is priced at the assumed $4.716394/kW rate. Other charges, taxes, operating costs and workload feasibility are outside this comparison.

Download accepted inputs ↓

Recorded engine outputs for a fictional month and assumed contract. Selected charges only, not a complete bill, customer savings or an operating recommendation.

Which terms explain your demand charge?

Bring this question to an audit →