Read the survey before the savings headline

How to Read a Steam-Trap Survey Report

A report can list many traps and still leave part of the system unassessed. Separate the inventory, valid test results, unresolved observations and estimated losses before deciding which work to approve.

Airplex Mechanical resources Reviewed November 16, 2026

Start here

What should you check in a steam-trap survey report?

Reconcile the trap inventory with the tested and untested populations. For each result, retain the asset ID, operating state, assessment method, defect classification and uncertainty. Read loss estimates with their assumptions and keep repair completion separate from follow-up verification. An untested trap is not a pass, and an inconclusive result is not a confirmed failure.

Establish which population the report represents

Find the survey date, report revision, areas covered and equipment boundaries. Check whether the inventory includes all known traps or only the accessible subset requested for this visit. If older records list additional assets, ask whether they were removed, renamed or left outside the scope. Keep that reconciliation separate from the condition count.

DOE/ORNL’s Steam System Survey Guide describes trap identification, qualified assessment, a maintained database and follow-up action. It distinguishes failed-open and failed-closed conditions and explains limitations of assessment methods and loss estimates. These are report-reading principles from a May 2002 guide; its old cost examples and generalized failure rates are not used here.

References: [1]

Worked example: reconcile 100 listed traps

The following fictional survey has 100 listed assets. Eighty were assessed during the visit; twenty were not. The assessed population contains 58 reported operating normally, 12 classified failed-open, 4 classified failed-closed and 6 inconclusive. These numbers describe an invented report, not a typical building or Airplex result.

Annotated sample population reconciliation
CategoryCountMeaning in this sampleReview action
Reported normal58Assessed under the recorded conditionsRetain method, conditions and asset-level records.
Failed-open12Reported defect classificationMatch each result to evidence and proposed corrective work.
Failed-closed4Reported defect classificationRecord service impact separately from leakage estimates.
Inconclusive6Assessment did not establish a final conditionRequest the limitation and follow-up owner.
Untested2010 inaccessible; 10 not in suitable operationKeep both reasons visible; do not mark normal.

Check the denominator behind a percentage

The classified defects total 12 + 4 = 16. In this example, 16 ÷ 80 = 20% of assessed traps were classified with those defects. That does not establish a 20% failure rate for the entire 100-trap inventory: the six inconclusive and twenty untested assets still have unresolved condition.

The same sixteen defects are 16% of the listed inventory, but that figure is a known classified-defect share, not proof that the remaining 84% work properly. Require the summary to name its denominator. Never extrapolate an inaccessible group as if it had been selected randomly and tested.

A later report must also explain changes in population. If ten traps were replaced, three renamed and a new wing added, comparing only the headline percentages can misrepresent progress. Preserve stable asset IDs and a separate change register.

Read one asset row all the way through

A useful row connects the finding to the operating context. The method name alone is incomplete without the assessor’s interpretation and limitation. Staff should read the supplied results; this guide does not teach testing or ask anyone to open, bypass or adjust steam equipment.

Illustrative asset-level annotations
Asset / resultRecorded contextMissing evidence or uncertaintyAction owner
ST-014: failed-openHeating circuit in service; acoustic and thermal findings referencedLoss calculation lacks effective-opening assumptionSurvey author clarifies estimate; qualified service scope reviewed
ST-039: failed-closedServed heat exchanger reported under demandReturn-side contribution not resolvedMechanical reviewer defines investigation and consequence
ST-062: inconclusiveIntermittent operation; observation window limitedNo final condition establishedSurvey lead defines suitable follow-up
ST-087: untestedOccupied room unavailableNo condition evidenceProperty manager coordinates access with survey provider

Separate a calculated loss from a measured saving

A loss estimate should name the trap, assumed leak condition, relevant pressure basis, estimated leak rate, operating hours and method. DOE notes that effective opening and partial obstruction can make failed-open loss estimates uncertain. A precise-looking total does not remove those limitations.

References: [1]

Worked arithmetic: the hours are an assumption too

Suppose an illustrative report assumes 5 lb/h of steam loss for 3,000 operating hours. Multiplication gives 15,000 lb for that period. This is only arithmetic on supplied assumptions. It is not a measured leak, annual fuel saving or verified consequence of replacing that trap.

If the report provides the 15,000 lb total but omits how the 5 lb/h and 3,000 hours were obtained, annotate both inputs as unresolved. Converting the quantity into a fuel-cost claim would need further stated assumptions about steam generation and cost. Do not copy a historical fuel price or apply the same leak estimate to failed-closed traps.

A failed-closed classification can matter to the served function even when a leakage-based savings table assigns it little value. Ask the qualified reviewer to describe the actual service impact. Ranking all repairs by an uncertain energy total can hide a more immediate operational concern.

Close an action without erasing its history

Assign the response to the finding: report clarification, access coordination, further assessment or approved repair. Keep work orders linked to the original asset IDs. For a replacement, record the installed identity and the responsible team’s application basis.

Then distinguish repair complete from follow-up verified. If suitable operating conditions were unavailable after the work, retain the open verification item. The building contact should be able to see what was done, what was assessed later and what remains unknown. In occupied NYC and Northern New Jersey properties, access and shared heating schedules belong in that handoff.

Put this guide to work

Steam-trap survey interpretation and action register

Use the population rows to reconcile the report, then add asset-specific findings with references. Enter existing survey evidence; do not perform tests to complete the sheet.

Open each item to add your notes, then download the CSV to keep a copy. This tool does not save or submit your entries; do not rely on your browser restoring them.

Inventoried population and scope revision
Inventoried population and scope revision
Assessed population and class totals
Assessed population and class totals
Inaccessible / untested population
Inaccessible / untested population
Inactive / unsuitable-condition population
Inactive / unsuitable-condition population
Inconclusive findings and follow-up
Inconclusive findings and follow-up
Failed-open asset and loss assumptions
Failed-open asset and loss assumptions
Failed-closed asset and service impact
Failed-closed asset and service impact
Asset ID or location discrepancy
Asset ID or location discrepancy
Approved repair / replacement record
Approved repair / replacement record
Post-work verification and remaining items
Post-work verification and remaining items

Questions & answers

Does an untested trap count as working?

No. Untested describes missing condition evidence. Preserve the reason and arrange the appropriate qualified follow-up.

Is every failed trap assigned the same steam-loss estimate?

No. The failure classification, application and estimate basis matter. A failed-closed result should not inherit a failed-open leakage calculation.

Does a lower defect percentage prove the repairs succeeded?

Check the population, test conditions and asset history first. A changed denominator or excluded area can lower the percentage without demonstrating that the original findings were resolved.

Sources & technical references

  1. DOE/ORNL: Steam System Survey Guide, May 2002

    Sections 6.2.2–6.2.2.5, PDF pages 56–61, read November 16, 2026 for assessment, database and estimate limitations. All sample counts, records and arithmetic are original hypothetical examples.

NYC & Northern New Jersey

Bring the report and its open items

Share the relevant survey pages, equipment scope and unresolved findings with Airplex Mechanical. Supported steam-system work and required qualified specialists are confirmed for the actual request.

Discuss the survey findingsCall (347) 627-7146
Call Now Request Service