Steam trap failure statistics are some of the most-quoted numbers in industrial energy management – and some of the most misunderstood. The figures get repeated across vendor brochures and survey pitches without much attention to where they came from or what they actually measure. This page compiles the credible, primary-sourced data on how often steam traps fail, what failure costs, and what a defensible benchmark looks like, with the caveats that most published versions leave out.
What percentage of steam traps fail?
The most widely cited figure comes from the U.S. Department of Energy’s Steam Tip Sheet #1: in steam systems that have not been maintained for three to five years, between 15% and 30% of the installed steam traps may have failed, allowing live steam to escape into the condensate return system. In systems with a regularly scheduled maintenance program, the DOE states that leaking traps should account for less than 5% of the trap population.
That gap is worth sitting with. The 15–30% band is not a description of steam systems in general – it describes failure specifically after several years without a program. Read together, the two DOE figures bracket two very different operating states: a neglected system and a managed one.
The benchmark numbers at a glance
| Metric | Typical value | Source |
|---|---|---|
| Failure rate, unmaintained 3–5 yrs | 15–30% may have failed | U.S. DOE, Steam Tip Sheet #1 |
| Leaking traps, managed program | Below 5% of trap population | U.S. DOE, Steam Tip Sheet #1 |
| Average steam-using plant | ~24% leaking or blocked | TLV survey data |
| In the fourth year of a management program | ~9% failed (from ~25%) | TLV 170,000-trap dataset |
| Cost of one failed-open trap (150 psig) | ~$6,640/yr (75.8 lb/hr) | U.S. DOE worked example |
Is there an annual steam trap failure rate?
It is tempting to convert the DOE’s 15–30% range into a tidy annual percentage, but there is no reliable universal annual steam trap failure rate to cite. Failure timing depends on trap type, operating hours, pressure, steam quality, and how often repairs are made, so a single annual number does not transfer cleanly between systems. What the DOE data does say is directional: failures accumulate over time, which is why the 15–30% figure attaches specifically to the three-to-five-year neglect window rather than to any given month. A system that tests and repairs on a rolling cycle keeps that accumulation from compounding, instead of letting it build across years of inattention.
What a large vendor survey dataset finds
The DOE figures are guidance ranges. One large body of published field data comes from steam-trap survey programs – which means they come from companies that perform trap surveys and sell trap-management services, so they are not independent of that business. They do, however, represent large measured samples. TLV, which conducts site-wide surveys, reports that in an average steam-using plant roughly 24% of traps are either leaking significant amounts of steam or blocking condensate. Analyzing a sample of approximately 170,000 traps surveyed in Japan, TLV reports that the failure rate was 25.0% at the first survey and fell to 9.4% by the fourth year after a regular trap-management program was implemented.
That trajectory – about 25% at the first survey down to roughly 9% by the fourth year of a program – is one of the more useful datasets in this space. It is directionally consistent with both DOE figures: neglected systems sit in the 15–30% band, and sustained management pulls the measured failure rate down sharply over time.
Why the published numbers vary so much
You will see steam trap failure rates quoted anywhere from single digits to well over 50%. Most of that spread comes from differences in the systems and methods behind the numbers, not from a single “true” rate. The main factors:
- Maintenance history. Whether a testing-and-repair program exists at all is the dividing line between the DOE’s under-5% (leaking) target and its 15–30% neglected band.
- How “failure” is defined. Some surveys count only failed-open (blow-through) traps; others also include failed-closed, plugged, undersized, misapplied, or out-of-service traps. Broader definitions produce higher percentages.
- Testing method. Ultrasonic, infrared, and visual checks differ in what they catch.
- Trap type and application. Different designs and duties fail in different proportions.
- Operating pressure and system age. Both vary widely between facilities and between surveys of the same facility.
The practical takeaway: treat any single “X% of steam traps fail” stat as a function of the population and method behind it. The defensible benchmark is a range, not a point – under 5% leaking is the managed target, 15–30% is the neglected band, and where your measured number lands provides useful context, provided the failure definitions and testing methods are comparable.
Failure modes: open vs. closed
Two major steam-trap failure outcomes are failed open and failed closed, and the distinction matters for both cost and detection. The proportion that fails each way varies by trap design and application, so there is no single universal “most common” mode – but the two behave very differently.
Failed open (blow-through) is the mode that drives the energy-loss figures. The trap stops differentiating between condensate and steam and passes live steam continuously into the condensate return. Because the process keeps working, a failed-open trap is easy to ignore – it bleeds energy while the steam system is operating, and may not immediately interrupt the process or produce an obvious warning.
Failed closed backs condensate up into the steam space. A failed-closed trap may cause little direct steam leakage, but condensate backup can reduce heat transfer and, in some applications, contribute to water hammer.
How much does a failed steam trap cost?
The DOE’s worked example is the standard reference. A trap on a 150 psig line with a 1/8-inch orifice stuck open loses an estimated 75.8 lb/hr of steam. At a steam value of $10 per 1,000 lb running continuously (8,760 hr/yr), that is:
75.8 lb/hr × 8,760 hr/yr × $10/1,000 lb = $6,640 per year, per failed trap
A lower-pressure example – a 15 psig line with the same 1/8-inch orifice – loses about 13.7 lb/hr, or roughly $1,200 per year. Both figures assume a 1/8-inch orifice, continuous operation (8,760 hr/yr), and steam valued at $10 per 1,000 lb. Larger openings, higher pressures, longer operating hours, and higher steam costs increase the estimated loss; the relationship with opening size and pressure is not linear. They are illustrative reference points, not a fixed range.
At the facility scale, the arithmetic compounds: a system with hundreds of traps and a meaningful failed fraction can reach six figures in annual steam loss. The additional steam demand generally increases fuel use and associated emissions, depending on the steam source and boiler efficiency.
How often should steam traps be tested?
The DOE’s 2012 Steam Tip Sheet #1 recommends testing intervals that scale with operating pressure:
| System pressure | Recommended testing interval |
|---|---|
| High pressure (150 psig and above) | Weekly to monthly |
| Medium pressure (30–150 psig) | Monthly to quarterly |
| Low pressure (below 30 psig) | Annually |
This is one DOE recommendation. The DOE/FEMP Operations & Maintenance Best Practices Guide also addresses steam trap testing as part of a broader O&M program. In practice, many sites survey on a rolling basis – more frequently while the failed population is high, then easing the interval once it is brought under control. The goal of any interval is the same: keep the failed fraction from compounding between surveys.
How to read your own numbers
If you run a survey, the DOE benchmarks give a simple frame of reference. Leaking traps under 5% of the population is the level the DOE associates with a system under scheduled maintenance. A measured rate in the 15–30% range is consistent with the DOE’s description of systems that have gone three to five years without active maintenance. The DOE also notes that a steam trap survey of any system with more than 500 traps will probably reveal significant steam losses.
Frequently asked questions
What percentage of steam traps fail?
In steam systems left unmaintained for three to five years, the U.S. Department of Energy reports that 15–30% of installed traps may have failed. In systems with a regular maintenance program, the DOE says leaking traps should account for less than 5% of the population.
What is a normal or acceptable steam trap failure rate?
The DOE indicates that leaking traps should account for less than 5% of the trap population in a system with scheduled maintenance. A measured rate of 15–30% is consistent with the DOE’s range for systems that have gone several years without maintenance.
How many steam traps fail each year?
There is no universal annual steam-trap failure rate. DOE guidance instead reports that 15–30% of traps may have failed in systems left unmaintained for three to five years, while regularly maintained systems should keep leaking traps below 5%. Failure timing varies with trap type, operating hours, pressure, and repair frequency, so the range does not convert cleanly into a fixed annual percentage.
How much does a failed steam trap cost?
The DOE’s standard example – a 150 psig trap with a 1/8-inch orifice stuck open, losing 75.8 lb/hr – costs about $6,640 per year. A lower-pressure trap (15 psig, same orifice) costs roughly $1,200 per year. Both assume a 1/8-inch orifice, continuous operation (8,760 hr/yr), and steam valued at $10 per 1,000 lb; larger orifices, higher pressures, and higher steam costs increase the loss.
What is the most common steam trap failure mode?
Failed open (blow-through) is the mode that produces direct, continuous steam loss. The trap passes live steam into the condensate return and may not immediately interrupt the process or produce an obvious warning. The proportion of traps that fail open versus closed varies by trap design and application, so there is no single universal most-common mode. Failed-closed traps can reduce heat transfer and, in some applications, contribute to water hammer.
How often should steam traps be tested?
The DOE’s 2012 Steam Tip Sheet #1 recommends testing high-pressure traps (150 psig and above) weekly to monthly, medium-pressure traps (30–150 psig) monthly to quarterly, and low-pressure traps (below 30 psig) annually. The DOE/FEMP Operations & Maintenance Best Practices Guide also addresses steam trap testing as part of a broader O&M program.
Why do steam trap failure rate estimates vary so much?
Estimates vary because of differences in the systems and methods behind them: maintenance history, how “failure” is defined (failed-open only versus also counting failed-closed, plugged, or inactive traps), testing method, trap type and application, and operating pressure.
Related guides
- Industrial Energy Waste From Poor Maintenance
- Predictive Maintenance ROI Benchmarks
- Unplanned Downtime Frequency Benchmarks
- Maintenance & Reliability Glossary
Sources
- U.S. Department of Energy, Advanced Manufacturing Office – Inspect and Repair Steam Traps, Steam Tip Sheet #1 (failure-rate ranges, testing intervals, cost example)
- TLV – Steam Trap Survey field data (average-plant and 170,000-trap program datasets)
- U.S. Department of Energy / FEMP – Operations & Maintenance Best Practices Guide, Release 3.0 (steam trap O&M)
Figures reflect U.S. Department of Energy guidance and published field-survey data. Actual failure rates and energy costs vary with system age, operating pressure, trap type, steam cost, and maintenance history.
