How to write a checkable utilities contract

Executive Summary

Utilities and energy invoices are hard to audit not because the vendor is dishonest but because the contract that governs them was never written to be checked. Rate riders, tariff pass-throughs and demand charges live in supplier tariffs, ISO filings and amendment letters instead of one document, so nobody on the AP team has a single reference to match the bill against. The mechanism that causes drift is a gap between where a rate lives and where it is billed.

What changes it is writing the contract as a control document from the start: every rate component named and dated, every pass-through tied to a named published index, every demand and reactive-power charge given its own line with a formula instead of a bundled total, and every true-up given a stated calculation window. A contract built this way turns an invoice audit from a negotiation into a lookup.

The remainder of this page sets out the clauses that make that possible, where energy contracts most commonly drift from that standard, and what an AP team should ask a utilities vendor to add before signing.

1. What makes a utilities contract hard to audit in the first place?

A utilities contract is hard to audit when the number on the invoice cannot be traced back to a single clause in the agreement. Rate structures get defined by reference to a supplier tariff filed elsewhere, demand and energy charges get bundled into one line, and pass-through costs cite an index without naming which published series or which date it is read on. Each of those is a place where the invoice and the contract can drift apart unnoticed.

Most utilities agreements are short relative to the billing complexity they govern. A two-page service agreement can sit behind a rate schedule that runs to dozens of tariff pages, updated by the utility on its own schedule and only referenced, not reproduced, in the contract.

That reference is the weak point. An AP clerk checking a bill has no way to confirm the tariff cited is the one currently in effect, or that a rider added mid-term was ever disclosed. The contract technically covers it. The invoice reviewer cannot see it.

The fix is not a longer contract. It is a contract that either states the rate components directly or names, with a date and a document number, exactly where each one is published, so the current version can be pulled and compared line by line.

2. How should rate components be broken out in the contract?

Each billable component, energy charge, demand charge, capacity charge, transmission charge and any rider, should appear as its own line in the contract with its own rate or formula, not folded into a single blended rate. A blended number cannot be checked against a bill that itemizes separately, and a bill that also blends its charges cannot be checked against anything at all. Separate lines are what make separate verification possible.

A blended all-in rate is easy to negotiate and impossible to audit. If the contract states one number per kWh and the invoice later itemizes energy, capacity, transmission and a handful of riders, there is no way to test whether the sum matches the promise without reverse-engineering the blend.

The contract should instead state each component's basis: energy charge per kWh, demand charge per kW of peak demand with the demand window defined, and any rider expressed as its own rate or formula tied to a named index.

Write the demand window explicitly. Whether peak demand is measured over 15 minutes or 30, and whether it resets monthly or rolls over a longer period, changes the number materially, and a contract silent on this cannot be used to check a demand charge later.

3. Which pass-through and index clauses need to be tied to something checkable?

Any clause that lets a rate move with an external cost, a fuel adjustment, a capacity market cost, a regulatory surcharge, needs to name the specific published index or tariff docket it tracks, the date the reading is taken, and the formula applied to it. Subject to fuel cost adjustments with no source named cannot be tested against anything; a named index read on a stated date can be pulled and recalculated.

Pass-through clauses exist because utilities themselves face volatile input costs and want to move that risk to the customer. That is a reasonable commercial term. The problem is not the pass-through, it is the vagueness with which it is usually written.

A clause reading "charges may be adjusted for changes in the cost of fuel or purchased power" gives the utility discretion and gives the reviewer nothing. A clause that instead reads "adjusted monthly per the [named index], read on the first business day of the billing month, applied per the formula in Exhibit C" gives both parties the same number to check.

As one example of a real published series: the US Bureau of Labor Statistics' Producer Price Index for Industrial electric power (series WPU0543) stood at 341.966 in July 2026, up 1.2% year over year (BLS PPI, read 2026-09-07). A pass-through clause naming a specific index like this, rather than a general reference to "fuel costs," is what lets a reviewer confirm the adjustment applied was the one actually published.

4. What should the true-up and reconciliation clause say?

A true-up clause needs a stated calculation window, a stated deadline for the utility to issue it, and a stated method for applying the correction, credit on a future invoice or direct refund. Without those three elements, a true-up is a promise with no enforcement date attached, and an AP team has no way to know whether a reconciliation that should have happened actually did.

Utilities routinely bill on estimated usage and true up later against actual meter reads, or reconcile an annual capacity charge against a forecast set months earlier. Both are normal. What is not normal is a contract that mentions the true-up exists without saying when it happens or how it is applied.

State the reconciliation period explicitly: monthly, quarterly or annual, and against what baseline. State the deadline by which the utility must issue the corrected figure. State whether the correction appears as a line item on a future invoice or as a separate credit memo, because a true-up buried inside a normal-looking invoice total is the easiest kind of correction to miss entirely.

A true-up clause with all three elements gives the reviewer a checklist: was it due, did it arrive, does the math match the stated method. A clause without them gives the reviewer nothing to check against.

5. How should demand ratchets and minimum charges be defined?

A demand ratchet, where a single peak month sets a minimum demand charge for months afterward, needs its lookback period and its percentage stated explicitly in the contract, not left to the tariff's general language. A minimum bill or minimum commitment charge needs the exact floor amount or formula stated, so a reviewer can tell whether a low-usage month's invoice reflects the contract or an error.

Ratchet clauses are one of the least understood terms in energy contracts because their effect shows up months after the triggering event. A single hot month's peak demand can set a raised minimum for the following eleven months under a ratchet, and if the contract states only that ratchets apply per the utility's tariff, the reviewer has no way to verify the ratchet percentage or lookback period without pulling the tariff itself.

State the ratchet's lookback window and percentage directly in the contract summary, even if the authoritative source is the tariff. Do the same for any minimum bill or take-or-pay demand charge: state the floor, or the formula that produces it, so an invoice below expected usage can be checked against a known minimum rather than assumed correct.

6. Can a utilities contract require the meter data itself to be auditable?

Yes. A contract can require the utility to provide interval meter data, not just a summed monthly total, on a standard schedule and in a machine-readable format. Without interval data, a reviewer checking a demand charge or a time-of-use rate has only the utility's own calculated total to check against itself, which is not an audit.

Interval data is what lets the charge be recalculated independently.

Most utilities can provide 15-minute or hourly interval data on request, and many already generate it for their own billing systems. The contract should require it be delivered routinely, in a standard file format, rather than requested case by case when a dispute arises.

Without interval data, an invoice reviewer checking a time-of-use or demand charge has to trust the utility's own summary figures, because there is no independent way to recompute peak demand or time-of-use splits from a monthly total alone.

This single clause does more to make a utilities contract auditable than any rate clause, because it converts every other clause in this page from a paper check into a data check. A rate formula stated in the contract, tested against real interval data, is a control. The same formula tested only against the utility's own monthly summary is not.

7. What should an AP team ask a utilities vendor to add before signing?

Before signing, ask for each rate component broken out with its own formula, every pass-through tied to a named index and read date, a true-up clause with a calculation window and deadline, ratchet and minimum terms stated as numbers rather than tariff references, and a standing right to interval meter data. None of these are unusual requests, and a utility unwilling to state them in writing is telling the buyer something about how it expects to be billed.

These clauses cost nothing to add if the underlying rate structure is already fixed. Asking for them is asking the utility to write down what it is already doing, not to change its pricing.

A vendor that resists itemizing a blended rate, or that will not name the index behind a pass-through clause, is not protecting a trade secret. Published tariffs and index series are, definitionally, public.

A short checklist works better in negotiation than a long one: named rate components, named pass-through index and date, true-up window and deadline, ratchet terms as numbers, and standing access to interval data. Five items, each independently checkable, and each one closes a specific gap this page has walked through.

For the wider pattern this sits inside, start with the margin drift guide. See also accessorial charge audit: the surcharges nobody validates and duplicate freight billing and the multi-carrier consolidation problem.

8. Frequently Asked Questions (People Also Ask)

Can I just require the utility to attach the full tariff to every invoice?

You can ask, but tariffs run long and change on the utility's schedule, not the invoice cycle. A better approach is naming the specific tariff docket number and effective date in the contract so a reviewer can pull the current version and compare, rather than relying on a static attachment that may already be outdated.

What if the utility refuses to itemize a blended rate?

Ask why. A blended rate is often easier for the utility to bill but harder for you to check. If itemization genuinely is not possible for a regulatory or system reason, ask instead for a written breakdown of how the blend is calculated, so the formula, if not the line items, is checkable.

Does a ratchet clause apply to every commercial electricity contract?

No. Ratchet clauses are a specific tariff or contract feature, not a universal charge. Check whether your rate schedule includes one before asking a vendor to define its terms. If it does not apply, the contract should say so explicitly rather than leave the question open.

How is a true-up different from a regular monthly adjustment?

A true-up reconciles estimated or forecast charges against actual usage or costs over a defined period, often longer than a single billing cycle. A regular monthly adjustment applies within the current invoice. Both need a stated method, but a true-up additionally needs a stated calculation window and deadline because its effect can span several billing cycles.

Who typically owns interval meter data, the utility or the customer?

This varies by jurisdiction and utility, and is worth confirming directly rather than assuming. Some utilities provide interval data as a standard service; others require a specific data-sharing agreement or third-party access request. The contract should state which applies and how the data will be delivered.

Can these clauses be added to an existing contract mid-term?

Often yes, through an amendment or side letter that adds specificity without renegotiating price. Because these clauses describe how existing rates are documented rather than changing what is charged, a utility has less commercial reason to resist adding them outside a renewal cycle.

What happens if a pass-through clause names an index that stops being published?

The contract should include a fallback: a named successor index or a defined method for selecting a substitute. Without that, an interrupted index leaves the pass-through clause unenforceable until both parties negotiate a replacement, which can take longer than a single billing cycle.

Is a checkable contract enough, or does someone still need to run the comparison?

A checkable contract is necessary but not sufficient. Naming every rate, index and formula only creates the reference; someone still has to pull each invoice and test it against that reference on a recurring basis for the clause to do any work.