# What's a reasonable exception rate for utilities?

> A framework for setting and testing a defensible exception rate on utility and energy invoices, grounded in PPI-tracked rate movement rather than an invented.

Source: https://valuexpa.com/insights/what-is-a-reasonable-exception-rate-for-utilities-and-energy
Publisher: ValueXPA (https://valuexpa.com)
Updated: 2026-09-07

---

Margin drift is the gap between what a vendor contract says and what the invoice actually charges. On a utility bill that gap hides inside rate schedules, tariff classes, and demand charges that few AP teams have the reference data to challenge.

Asking for "a reasonable exception rate" assumes a benchmark exists. It does not, not as a published figure anyone can cite. What does exist is a defensible method for setting your own threshold and testing it against a dated public index instead of a guess.

## Executive Summary

Utility and energy invoices carry a built-in excuse for every anomaly: rates move. That excuse is legitimate often enough that most AP teams stop testing exceptions past the first plausible explanation, which is exactly the gap where a stale rate schedule, a misapplied tariff class, or a demand charge computed off the wrong billing period survives quarter after quarter.

There is no published benchmark for what share of utility invoice lines should trip an exception. The engine has no dataset to produce one, and any AP automation vendor quoting a flat percentage is quoting a number with no denominator behind it. What exists instead is a public, dated reference point: the Producer Price Index tracks how much industrial electric power actually moved, so a variance against that reference is checkable, while a variance against last month's invoice is only checkable against itself.

What changes the outcome is not a target percentage. It is separating rate movement, which a published index can confirm, from rate application, which only the contract and tariff schedule can confirm. An exception process that tests both, with the index as the ceiling check and the tariff as the floor check, catches drift that a single flat threshold misses in both directions.

## 1. Why isn't there a published exception rate for utility invoices?

**No industry body publishes a standard exception rate for utility and energy invoices, because the right threshold depends on a specific meter's rate class, contract terms, and billing history, none of which generalize across companies. A benchmark implies a population of comparable invoices has been measured and averaged. That dataset does not exist publicly, and a vendor quoting a flat percentage is quoting a number with no stated denominator behind it.**

A threshold that fits one facility's utility account will not fit another. A plant on an interruptible rate schedule has a different variance profile than one on a flat industrial tariff, and a facility that recently changed its contracted demand level will show a jump that has nothing to do with billing error.

Treating a borrowed percentage as a target invites two failures at once. Set it too low and every seasonal rate adjustment triggers a review, which trains the AP team to wave exceptions through without reading them. Set it too high and a genuine tariff misapplication sits inside the tolerance band indefinitely.

The fix is not a better borrowed number. It is building the threshold from your own account's billing history and the tariff schedule that governs it, then testing movement against a dated public reference rather than against last month's bill.

## 2. What actually causes a utility invoice to vary month to month?

**Utility invoices vary for reasons that are legitimate and reasons that are billing error, and the two look identical on the page. Legitimate movement includes seasonal rate changes, fuel cost adjustment riders, and demand charges that shift with actual peak usage. Billing error includes a tariff class applied after a facility's usage profile changed, a demand charge calculated off the wrong billing window, or a rate schedule that was never updated after a contract renewal.**

Separating these two categories is the entire exception review. Neither shows up differently on the invoice format itself, so the check has to happen against source documents, not against the bill's layout.

### A. Legitimate movement

Fuel cost adjustment riders pass through the utility's own input cost changes and reset on a published schedule. Seasonal or time-of-use rates shift the per-kWh or per-therm charge by calendar period under the tariff itself. Demand charges move with the facility's actual peak draw in the billing window, which changes with production schedule, not with billing error.

### B. Billing error

A tariff class that no longer matches the facility's connected load or usage pattern keeps charging the old schedule until someone challenges it. A demand charge computed off a prior month's peak, carried forward by a metering or billing system error, overstates the current bill. A rate card left un-updated after a contract renewal charges the expired rate indefinitely, and nothing in the invoice format flags that it happened.

## 3. How does the PPI industrial electric power index help set a threshold?

**The US Bureau of Labor Statistics publishes a Producer Price Index series for industrial electric power (WPU0543) that gives a dated, external ceiling for how much power costs actually moved. Per the US Bureau of Labor Statistics PPI series WPU0543 (read 2026-09-07), the July 2026 index value was 341.966, up 1.2% year over year. A facility's rate increase that runs well past that figure needs a specific explanation beyond "rates went up."**

The index does not tell you what your contract should charge. It tells you what the broader input cost did over the same period, which is the number a rate increase should roughly track if nothing else changed.

When an invoice's per-unit rate rises faster than the index over the same window, that gap is the actual exception worth investigating, not the raw dollar change from last month's bill. A facility with a fixed-rate contract should show almost no correlation with the index at all, and if it does move with it, the contract may not be fixed the way it was sold.

This only works as a ceiling check. It confirms whether a rate move is plausible against a broader market signal. It cannot confirm whether the rate applied to your account matches your contract, which is a separate check against the tariff schedule and the signed agreement, not against an index.

## 4. How do you separate a rate movement exception from a rate application exception?

**A rate movement exception asks whether the price per unit changed by more than the market moved, checked against a published index. A rate application exception asks whether the rate charged is the rate the contract or tariff actually specifies, checked against the signed schedule regardless of market movement. Running both catches different failures: one catches an unexplained spike, the other catches a stale or misapplied rate that never spikes at all because it was wrong from the start.**

The two checks use different documents and answer different questions, which is why a single threshold applied to raw dollar variance cannot substitute for either one alone.

Two separate checks catch two separate failure modes on a utility invoice.

| Check
| Question it answers
| What it cannot catch

| Rate movement
| Did the per-unit rate move more than the published index suggests it should have?
| A rate that has always been wrong, since a consistently wrong rate shows no unusual movement

| Rate application
| Does the charged rate match the tariff class and contract terms on file?
| A rate that is correctly applied but reflects a genuine market increase

| Demand charge recompute
| Was the demand charge calculated from the correct billing window's actual peak?
| A tariff class error that affects the per-unit rate rather than the demand component

## 5. What should trigger a manual review on a utility invoice?

**A manual review is warranted when a rate change cannot be explained by a published fuel adjustment rider, a seasonal tariff step, or index-level movement in the underlying commodity. It is also warranted whenever a facility's tariff class, contract ID, or meter number changes on the invoice without a corresponding change in the vendor master file. Both are checkable against documents already on file, not against a percentage threshold applied blindly to every line.**

Each trigger below points to a specific document the reviewer checks first, which keeps the review fast even without a numeric threshold to anchor it.

- **Tariff class mismatch:** The invoice cites a rate class different from the one in the signed service agreement or the last confirmed utility account record.

- **Demand charge outlier:** The billed demand exceeds the facility's known peak draw for that billing window by an amount the production schedule does not explain.

- **Rider stacking:** More than one adjustment rider appears on the same line item without a corresponding notice from the utility explaining the change.

- **Contract renewal gap:** The invoice date falls after a known contract renewal or rate negotiation and the rate has not changed at all.

- **Index divergence:** The per-unit rate change materially outpaces the movement shown in the relevant PPI series over the same period.

## 6. Can you build a defensible exception threshold without a benchmark?

**Yes. A defensible threshold is built from the account's own twelve-month billing history plus the tariff schedule, not from an industry average. Take the account's historical rate variance, exclude the periods explained by a published rider or seasonal step, and set the tolerance band around what remains. Anything outside that band gets checked against the tariff first and the PPI index second, so the review always starts with the document that can actually confirm or deny the charge.**

Start with the twelve months of invoices already on file for the account. Strip out the variance that a published fuel adjustment rider or seasonal rate step already explains, since that movement is expected and does not belong in the exception band.

What remains is the account's baseline noise: metering rounding, minor demand fluctuation, small billing timing shifts. Set the tolerance band around that baseline, not around a number someone else's account produced.

When an invoice falls outside the band, check the tariff schedule first. That document either confirms the charge or it does not, and it settles the question faster than any index comparison. Use the PPI series only when the tariff confirms the rate class is correct and the question left is whether the level of increase itself is plausible against the broader market.

This is general information, not legal or tax advice. A tariff or contract dispute with a utility provider should go through the account's own contract counsel.

For the wider pattern this sits inside, start with the [margin drift](/guides/indirect-spend-audit-categories) guide. See also [the six categories drift hides in](/guides/indirect-spend-audit-categories) and [accessorial charge audit: the surcharges nobody validates](/guides/accessorial-charge-audit-the-surcharges-nobody-validates).

## 7. Frequently Asked Questions (People Also Ask)

### Is there an industry-standard exception rate for utility invoices?

No. No published dataset measures exception rates across utility accounts industry-wide, so any flat percentage offered as a standard has no stated source behind it. Build your threshold from your own account's billing history and tariff schedule instead.

### What is the PPI series for industrial electric power?

WPU0543, published by the US Bureau of Labor Statistics. It tracks producer price movement in fuels and power for industrial electric power. Per the series read on 2026-09-07, the July 2026 index value was 341.966, up 1.2% year over year.

### Can the PPI index confirm my utility bill is correct?

No. It confirms whether a rate change is plausible against broader market movement. It cannot confirm whether the rate applied to your account matches your specific contract or tariff class, which only the signed agreement and tariff schedule can do.

### What causes a demand charge to be wrong on a utility invoice?

A demand charge computed off the wrong billing window's peak usage, often carried forward by a metering or billing system error rather than recalculated for the current period, produces an overstated charge that a document-level check against the actual peak reading catches.

### Why do fuel adjustment riders make utility invoices hard to audit?

A fuel adjustment rider legitimately changes the per-unit rate on a published schedule, which looks identical on the invoice to an unexplained rate increase. Separating the two requires checking the rider's own published schedule, not just the dollar change from the prior bill.

### Should every utility rate increase trigger a review?

No. A rate increase explained by a published rider, a seasonal tariff step, or movement consistent with the relevant PPI series does not need escalation. Reserve manual review for increases that outpace those explanations or that coincide with a tariff class or contract change.

### How often should a utility tariff class be reconfirmed?

Whenever the facility's usage profile, connected load, or contract terms change, since the tariff class on file with the utility does not update itself when those conditions shift, and a stale class can persist on invoices indefinitely without triggering any automatic flag.

### Does a fixed-rate utility contract still need this kind of audit?

Yes. A fixed-rate contract should show minimal correlation with market index movement, and if the billed rate does track the index anyway, that itself is worth checking against the signed agreement to confirm the rate is genuinely fixed.

### Is contract complexity quietly draining your operating margin?

A small systematic drift between your negotiated contracts and your actual vendor billing compounds quietly across a year of invoices. Stop guessing at your exposure and run a targeted audit.

**[Take the Free Screener → https://valuexpa.com/margin-drift-screener](https://valuexpa.com/margin-drift-screener)**

## Executive Summary

Utility and energy invoices carry a built-in excuse for every anomaly: rates move. That excuse is legitimate often enough that most AP teams stop testing exceptions past the first plausible explanation, which is exactly the gap where a stale rate schedule, a misapplied tariff class, or a demand charge computed off the wrong billing period survives quarter after quarter. There is no published benchmark for what share of utility invoice lines should trip an exception. The engine has no dataset to produce one, and any AP automation vendor quoting a flat percentage is quoting a number with no denominator behind it. What exists instead is a public, dated reference point: the Producer Price Index tracks how much industrial electric power actually moved, so a variance against that reference is checkable, while a variance against last month's invoice is only checkable against itself. What changes the outcome is not a target percentage. It is separating rate movement, which a published index can confirm, from rate application, which only the contract and tariff schedule can confirm. An exception process that tests both, with the index as the ceiling check and the tariff as the floor check, catches drift that a single flat threshold misses in both directions.

## 1. Why isn't there a published exception rate for utility invoices?

No industry body publishes a standard exception rate for utility and energy invoices, because the right threshold depends on a specific meter's rate class, contract terms, and billing history, none of which generalize across companies. A benchmark implies a population of comparable invoices has been measured and averaged. That dataset does not exist publicly, and a vendor quoting a flat percentage is quoting a number with no stated denominator behind it. A threshold that fits one facility's utility account will not fit another. A plant on an interruptible rate schedule has a different variance profile than one on a flat industrial tariff, and a facility that recently changed its contracted demand level will show a jump that has nothing to do with billing error. Treating a borrowed percentage as a target invites two failures at once. Set it too low and every seasonal rate adjustment triggers a review, which trains the AP team to wave exceptions through without reading them. Set it too high and a genuine tariff misapplication sits inside the tolerance band indefinitely. The fix is not a better borrowed number. It is building the threshold from your own account's billing history and the tariff schedule that governs it, then testing movement against a dated public reference rather than against last month's bill.

## 2. What actually causes a utility invoice to vary month to month?

Utility invoices vary for reasons that are legitimate and reasons that are billing error, and the two look identical on the page. Legitimate movement includes seasonal rate changes, fuel cost adjustment riders, and demand charges that shift with actual peak usage. Billing error includes a tariff class applied after a facility's usage profile changed, a demand charge calculated off the wrong billing window, or a rate schedule that was never updated after a contract renewal. Separating these two categories is the entire exception review. Neither shows up differently on the invoice format itself, so the check has to happen against source documents, not against the bill's layout. ### A. Legitimate movement Fuel cost adjustment riders pass through the utility's own input cost changes and reset on a published schedule. Seasonal or time-of-use rates shift the per-kWh or per-therm charge by calendar period under the tariff itself. Demand charges move with the facility's actual peak draw in the billing window, which changes with production schedule, not with billing error. ### B. Billing error A tariff class that no longer matches the facility's connected load or usage pattern keeps charging the old schedule until someone challenges it. A demand charge computed off a prior month's peak, carried forward by a metering or billing system error, overstates the current bill. A rate card left un-updated after a contract renewal charges the expired rate indefinitely, and nothing in the invoice format flags that it happened.

## 3. How does the PPI industrial electric power index help set a threshold?

The US Bureau of Labor Statistics publishes a Producer Price Index series for industrial electric power (WPU0543) that gives a dated, external ceiling for how much power costs actually moved. Per the US Bureau of Labor Statistics PPI series WPU0543 (read 2026-09-07), the July 2026 index value was 341.966, up 1.2% year over year. A facility's rate increase that runs well past that figure needs a specific explanation beyond "rates went up." The index does not tell you what your contract should charge. It tells you what the broader input cost did over the same period, which is the number a rate increase should roughly track if nothing else changed. When an invoice's per-unit rate rises faster than the index over the same window, that gap is the actual exception worth investigating, not the raw dollar change from last month's bill. A facility with a fixed-rate contract should show almost no correlation with the index at all, and if it does move with it, the contract may not be fixed the way it was sold. This only works as a ceiling check. It confirms whether a rate move is plausible against a broader market signal. It cannot confirm whether the rate applied to your account matches your contract, which is a separate check against the tariff schedule and the signed agreement, not against an index.

## 4. How do you separate a rate movement exception from a rate application exception?

A rate movement exception asks whether the price per unit changed by more than the market moved, checked against a published index. A rate application exception asks whether the rate charged is the rate the contract or tariff actually specifies, checked against the signed schedule regardless of market movement. Running both catches different failures: one catches an unexplained spike, the other catches a stale or misapplied rate that never spikes at all because it was wrong from the start. The two checks use different documents and answer different questions, which is why a single threshold applied to raw dollar variance cannot substitute for either one alone. Two separate checks catch two separate failure modes on a utility invoice. | Check | Question it answers | What it cannot catch | | --- | --- | --- | | Rate movement | Did the per-unit rate move more than the published index suggests it should have? | A rate that has always been wrong, since a consistently wrong rate shows no unusual movement | | Rate application | Does the charged rate match the tariff class and contract terms on file? | A rate that is correctly applied but reflects a genuine market increase | | Demand charge recompute | Was the demand charge calculated from the correct billing window's actual peak? | A tariff class error that affects the per-unit rate rather than the demand component |

## 5. What should trigger a manual review on a utility invoice?

A manual review is warranted when a rate change cannot be explained by a published fuel adjustment rider, a seasonal tariff step, or index-level movement in the underlying commodity. It is also warranted whenever a facility's tariff class, contract ID, or meter number changes on the invoice without a corresponding change in the vendor master file. Both are checkable against documents already on file, not against a percentage threshold applied blindly to every line. Each trigger below points to a specific document the reviewer checks first, which keeps the review fast even without a numeric threshold to anchor it. 1. Tariff class mismatch: The invoice cites a rate class different from the one in the signed service agreement or the last confirmed utility account record. 2. Demand charge outlier: The billed demand exceeds the facility's known peak draw for that billing window by an amount the production schedule does not explain. 3. Rider stacking: More than one adjustment rider appears on the same line item without a corresponding notice from the utility explaining the change. 4. Contract renewal gap: The invoice date falls after a known contract renewal or rate negotiation and the rate has not changed at all. 5. Index divergence: The per-unit rate change materially outpaces the movement shown in the relevant PPI series over the same period.

## 6. Can you build a defensible exception threshold without a benchmark?

Yes. A defensible threshold is built from the account's own twelve-month billing history plus the tariff schedule, not from an industry average. Take the account's historical rate variance, exclude the periods explained by a published rider or seasonal step, and set the tolerance band around what remains. Anything outside that band gets checked against the tariff first and the PPI index second, so the review always starts with the document that can actually confirm or deny the charge. Start with the twelve months of invoices already on file for the account. Strip out the variance that a published fuel adjustment rider or seasonal rate step already explains, since that movement is expected and does not belong in the exception band. What remains is the account's baseline noise: metering rounding, minor demand fluctuation, small billing timing shifts. Set the tolerance band around that baseline, not around a number someone else's account produced. When an invoice falls outside the band, check the tariff schedule first. That document either confirms the charge or it does not, and it settles the question faster than any index comparison. Use the PPI series only when the tariff confirms the rate class is correct and the question left is whether the level of increase itself is plausible against the broader market. This is general information, not legal or tax advice. A tariff or contract dispute with a utility provider should go through the account's own contract counsel. For the wider pattern this sits inside, start with the [margin drift](/guides/indirect-spend-audit-categories) guide. See also [the six categories drift hides in](/guides/indirect-spend-audit-categories) and [accessorial charge audit: the surcharges nobody validates](/guides/accessorial-charge-audit-the-surcharges-nobody-validates).

## Common questions

### Is there an industry-standard exception rate for utility invoices?

No. No published dataset measures exception rates across utility accounts industry-wide, so any flat percentage offered as a standard has no stated source behind it. Build your threshold from your own account's billing history and tariff schedule instead.

### What is the PPI series for industrial electric power?

WPU0543, published by the US Bureau of Labor Statistics. It tracks producer price movement in fuels and power for industrial electric power. Per the series read on 2026-09-07, the July 2026 index value was 341.966, up 1.2% year over year.

### Can the PPI index confirm my utility bill is correct?

No. It confirms whether a rate change is plausible against broader market movement. It cannot confirm whether the rate applied to your account matches your specific contract or tariff class, which only the signed agreement and tariff schedule can do.

### What causes a demand charge to be wrong on a utility invoice?

A demand charge computed off the wrong billing window's peak usage, often carried forward by a metering or billing system error rather than recalculated for the current period, produces an overstated charge that a document-level check against the actual peak reading catches.

### Why do fuel adjustment riders make utility invoices hard to audit?

A fuel adjustment rider legitimately changes the per-unit rate on a published schedule, which looks identical on the invoice to an unexplained rate increase. Separating the two requires checking the rider's own published schedule, not just the dollar change from the prior bill.

---

ValueXPA runs a fixed-scope Margin Drift Diagnostic that validates every service vendor invoice against contract terms, for $100M+ US industrial manufacturers and distributors. Two to four weeks. The client retains 100% of recoveries. https://valuexpa.com/contact-us