One Agency, Two Thresholds: $100,000 of Pipeline Damage Counts on Only One Line

A contractor's excavator catches a buried pipeline. Product escapes, crews respond, the street is opened, damaged pipe is cut out and replaced, and some weeks later the operator's accountants settle on a number for what the day cost: roughly $100,000. Whether that release ever becomes part of the federal record of American pipeline failures — whether it is counted at all — depends almost entirely on what the pipe was carrying.

Two sections of federal regulation decide it, and the same agency administers both. Under 49 CFR 191.3, a gas release becomes a reportable incident when estimated property damage reaches a figure that, for the year beginning July 1, 2026, stands at $153,600. Under 49 CFR 195.50, a hazardous liquid release becomes a reportable accident when estimated property damage exceeds $50,000. A $100,000 release on a liquid line is an accident in the federal record. The same $100,000 on a gas line produces no report, no row, and no entry in any trend line.

The gap is not a rounding artifact. At the thresholds in force today it is a factor of 3.07. And the two numbers are drifting apart every year, because one of them is indexed to inflation and the other has not been touched since 1994.

The Chain

The distance between a physical failure and a number in a federal table is longer than it looks, and the count can move at every link without anything changing in the ground.

1. The release happens. This is the only step governed by metallurgy, soil, pressure and the excavator operator's attention. Everything after it is governed by text.
2. The operator assembles a damage estimate. This is not a neutral act of addition. Each rule prescribes a different set of things to add up, and — critically — a different treatment of the escaped commodity itself.
3. The estimate is compared against a threshold. One threshold is recalculated every spring and takes effect each July. The other is a fixed number printed in the code.
4. A report is filed, or it is not. There is no partial credit and no lesser category. Below the line, the event leaves no federal trace under the property-damage criterion.
5. The report enters PHMSA's flagged files, where the agency adds derived fields — cause and subcause classifications, a "Fire First" designation, and the property damage figure restated in both current-year and 1984 dollars.
6. Counts are aggregated into trend lines that are then read as statements about pipeline condition over time.

Steps 2, 3 and 5 are all places where a constant rate of physical failure can produce a changing published number. Step 5 is the subtlest: the same dataset carries the same damage figure in two different years' dollars, for two different purposes.

Two rules, two definitions of "damage"

The regulatory text is short enough to read directly, and the difference is in the exclusions rather than the dollar figures.

49 CFR 191.3 defines a gas incident to include an event causing "Estimated property damage of $122,000 or more, including loss to the operator and others, or both, but excluding the cost of gas lost." The section then adds that "For adjustments for inflation observed in calendar year 2021 onwards, changes to the reporting threshold will be posted on PHMSA's website."

49 CFR 195.50 requires an accident report for a release resulting in "Estimated property damage, including cost of clean-up and recovery, value of lost product, and damage to the property of the operator or others, or both, exceeding $50,000."

Read side by side, the two rules treat the escaped commodity in opposite directions. The gas rule explicitly removes the value of the gas that got away before the comparison is made. The liquid rule explicitly adds the value of the lost product in. For a release of identical financial magnitude, the gas estimate is assembled to be smaller than the liquid estimate, and is then measured against a threshold roughly three times larger.

There is a second, quieter asymmetry in the verbs. The gas rule counts damage of "$122,000 or more" — inclusive. The liquid rule counts damage "exceeding $50,000" — exclusive. A hazardous liquid release assessed at exactly $50,000 is not reportable on the property-damage criterion. A gas release assessed at exactly the threshold is. The boundary condition differs by one rule's choice of preposition.

One threshold moves every July; the other has not moved since 1994

Appendix A to 49 CFR Part 191 sets out the indexing procedure in a single formula. The revised threshold Tr equals the previous threshold Tp multiplied by the ratio of two Consumer Price Index values: Tr = Tp x (CPIr / CPIp), where CPIr is the average CPI for All Urban Consumers published by the Bureau of Labor Statistics across the most recent complete calendar year, and CPIp is the average CPI-U for the year used to set the previous threshold. The appendix specifies that each year after 2021 the Administrator publishes the update on PHMSA's website, effective July 1 and running to June 30 of the following year.

PHMSA's April 2026 notice supplies the inputs for the current period. The 2025 threshold of $149,700 is multiplied by the 2025 average CPI-U of 321.9 over the 2024 average of 313.7. Working that through gives 149,700 x 1.026140 = 153,613, which the agency rounds to the published $153,600 for July 1, 2026 through June 30, 2027. The same arithmetic reproduces every prior step in the chain: $145,400 x (313.7 / 304.7) rounds to the published $149,700, and $129,300 x (292.7 / 271.0) rounds to the published $139,700.

Against that staircase, the hazardous liquid figure is a flat line. PHMSA's own 2025 rulemaking states the position plainly: "The current $50,000 threshold has not been adjusted since 1994." The agency's history of incident reporting criteria traces the liquid figure from $1,000 in 1969 — third-party property only — to $5,000 in 1981 and to $50,000 in 1994, where it has stayed. The gas figure travelled from $5,000 in 1970 to $50,000 in 1984, sat alongside the liquid number for decades, and then separated from it in 2021.

Since that separation, the gas threshold has risen from $122,000 to $153,600 — an increase of 25.9 percent in five annual adjustments. The liquid threshold has risen by nothing.

The number printed in the code is not the number in force

An engineer who opens the Code of Federal Regulations today to check the gas reporting threshold will read $122,000. That figure is five years stale. The operative number is $153,600, and it does not live in the code at all — it lives in a PDF notice on the agency's website, as the regulation itself instructs. The difference between the printed figure and the enforceable one is $31,600.

This is a deliberate design, not an oversight: indexing a number in regulatory text would require an annual rulemaking, so the text points outward to a posted value instead. But it means the authoritative public document is, by construction, wrong about its own threshold every day since July 1, 2022, and wrong by a widening margin.

A third yardstick, denominated in 1984 dollars

Reportability is only the first gate. PHMSA also flags a subset of reported events as significant incidents, and that flag uses a third dollar standard. The agency's PHMSA, National Pipeline Performance Measures lists events including "Fatality or injury requiring in-patient hospitalization"; "$50,000 or more in total costs, measured in 1984 dollars"; "Highly volatile liquid releases of 5 barrels or more or other liquid releases of 50 barrels or more"; and "Liquid releases resulting in an unintentional fire or explosion." The same list appears on PHMSA's Pipeline Incident Flagged Files page with the note that "Fire First incidents are excluded" — events where a nearby fire or explosion damaged the pipeline rather than the reverse.

So a single PHMSA dataset carries damage figures measured against three separate denominators at once: current-year dollars at $153,600 for gas reportability, current-year dollars at $50,000 for liquid reportability, and deflated 1984 dollars at $50,000 for the significance flag. The flagged files carry the damage restated in both current-year and 1984 terms precisely because two of these comparisons must be run on the same record. Anyone comparing "incidents" to "significant incidents" across years is comparing counts produced by yardsticks that move at different speeds — the same structural problem that appears whenever a published figure depends on a bracket rather than a measurement, as it does when one hundred AQI points cover 35.4 micrograms in one band and 90.0 in the next.

The rule that was withdrawn seven days before it took effect

PHMSA has tried to close the gap. On July 1, 2025 the agency published a PHMSA direct final rule, 90 FR 28050 that would have set both the gas and hazardous liquid property-damage thresholds at $149,700, with an effective date of October 9, 2025. Its stated rationale was consistency: the rule "corrects this discrepancy... while also ensuring consistency across gas and hazardous liquid pipeline reporting requirements."

It never took effect. On October 2, 2025 — seven days before the effective date — PHMSA withdrew it: "Due to the receipt of adverse comment, PHMSA is withdrawing the DFR." The agency restarted through ordinary notice and comment, publishing a PHMSA notice of proposed rulemaking on April 24, 2026 with the same $149,700 figure and a comment period that closed June 23, 2026. That proposal would also end the commodity asymmetry: its proposed § 195.50(e) excludes "the cost of lost product," matching the gas rule's exclusion of gas lost, and both proposed sections add exclusions for permit costs and for removing and replacing non-operator infrastructure not damaged by the release. As of the eCFR text current to September 11, 2026, 49 CFR 195.50(e) still reads $50,000.

The proposal also quantifies its own reach, and the numbers are small. PHMSA "estimates that approximately five incidents per year would meet the current reporting requirements but not the proposed requirements," against "an average of over 370 incidents per year" — under 1.4 percent. For hazardous liquids, the 2025 direct final rule "identified 57 hazardous liquid pipeline accidents reported between 2010 and 2024 involving total property damage between $50,000 and $149,700 that did not result in a fire, the release of five gallons or more, death, or injury" — roughly 3.8 per year across fifteen years. Estimated annual cost savings: $5,235.

The proposal characterizes that band directly: accidents "with between $50,000 and $149,700 of property damage that do not meet the other criteria are definitionally the least-consequential releases reported to PHMSA." That is an honest description of what the threshold is for. It is also an admission that the boundary of the federal dataset is drawn by a dollar figure whose two versions have been out of step for five years and counting.

What a moving gate does to a trend line

The practical consequence is not that any published count is false. It is that a count is only interpretable alongside the gate that produced it. If the physical rate of small gas releases held perfectly steady from 2021 to 2026, the reported count would still have fallen, because the bar those releases must clear rose 25.9 percent over the same window. PHMSA flags the problem in its own documentation of the all-reported-incidents series, noting that "Changes to PHMSA reporting regulations have caused large shifts in the trend line."

This matters most for exactly the question the data is usually asked to answer. Where the case for replacing aging pipe rests on evidence that it fails more often, that evidence is assembled from counted incidents — and small incidents are the ones nearest the moving line. An indexed threshold systematically thins the low-damage tail of the record over time, which is the part of the distribution where deterioration shows up first and most cheaply.

The counts are also not final when first published. Under 49 CFR 191.15, an operator has up to 30 days after detection to file the initial report, and paragraph (d) adds that "Where additional related information is obtained after an operator submits a report under paragraph (a), (b), or (c) of this section, the operator must make a supplemental report as soon as practicable, with a clear reference by date to the original report." A year's totals therefore keep settling after the year ends, and a figure read in January is not necessarily the figure that will be read two years later. Exclusion rules and revision cycles reshape published infrastructure numbers routinely — removing major event days takes 2024's outage average from 662.6 minutes to 131.6 — and the pipeline record is no different in kind, only in which clause does the work. As with red flag warning thresholds that are local numbers differing by thirty percentage points, the threshold is not a detail attached to the measurement. It is the measurement.

Where This Doesn't Apply

The threshold gap governs one criterion, and it is not the criterion that captures serious events.

• Dollar thresholds do not gate consequential releases. Both rules require a report for a death or an injury necessitating in-patient hospitalization regardless of cost. The liquid rule additionally requires one for any explosion or fire not intentionally set by the operator, and for a release of 5 gallons or more. The gas rule requires one for unintentional gas loss of three million cubic feet or more, and for an emergency shutdown of an LNG facility or underground natural gas storage facility. Fatal and injurious events are counted the same way on both sides of the gap.
• The affected population is small by the agency's own estimate. Roughly five gas incidents per year out of more than 370, and 57 hazardous liquid accidents across fifteen years. This is a story about the boundary of a dataset, not about a large body of hidden failures.
• Operator judgment remains a catch-all. The gas definition includes any event "that is significant in the judgment of the operator, even though it did not meet" the other criteria — a discretionary channel that no threshold analysis can predict.
• Nothing here establishes which threshold is correct. A case exists for indexing and a case exists for stability; adverse commenters on the 2025 direct final rule evidently made one of them. The point is only that two numbers are being applied to one network.
• Federal reporting is not the whole record. Intrastate pipelines are largely regulated by state authorities operating under their own reporting requirements, and state numbers are not necessarily gated at the federal figures.
• Threshold effects are not unique to pipelines, and not evidence of concealment. Any counted series with a dollar cutoff behaves this way. The distinguishing feature here is only that one agency runs two cutoffs at once, plus a third for its significance flag, in three different years' dollars.

The next adjustment is already scheduled. Unless the April 2026 proposal is finalized first, Appendix A will produce a new gas figure next spring, effective July 1, 2027, and the liquid threshold will remain $50,000. The distance between the two federal answers to "was this release big enough to count" will widen by whatever the Bureau of Labor Statistics reports for 2026.