Instrument Air Leak Audit: Survey Methods, Cost Calculation, and Repair Prioritization

How to Conduct a Leak Survey

There are three primary methods for finding instrument air leaks: ultrasonic detection, soapy water testing, and audible/visual inspection. Each has its place, and a thorough survey typically uses all three.

Ultrasonic leak detectors are the primary tool for systematic surveys. These handheld instruments detect the ultrasonic noise (20-100 kHz) generated by air escaping through a small orifice. They work in noisy environments because the detector filters out audible frequencies. Point the detector at tubing connections, fittings, actuator housings, and positioner exhausts while watching the signal level on the display. Most units have a directional cone attachment that narrows the detection angle for pinpointing leak locations in congested areas.

Soapy water (or commercial leak detection fluid) is applied to suspected leak points. Bubbles indicate a leak. This is the most positive identification method - if it bubbles, it leaks. The limitation is that it requires physical access to each connection and is slow for large surveys. Use soapy water to confirm leaks found by ultrasonic detection and to pinpoint exact locations (which fitting, which thread, which O-ring).

Audible and visual inspection catches the large leaks that are obvious: hissing sounds from tubing connections, ice formation on large leaks (from Joule-Thomson cooling), and actuators that visibly cycle or bleed air continuously. These are the high-priority leaks that waste the most air and should be repaired immediately.

Organize the survey by area (units, buildings, pipe racks) and tag each leak with a unique number, location description, estimated severity, and date found. Use a standard leak tag that can be attached to the tubing or fitting near the leak. This creates a repair list that maintenance can work through systematically.

Equivalent-Orifice Leak Rate Estimation

The app uses the DOE/CAC compressed-air leak chart rather than deriving a private choked-flow equation in page copy. The source chart gives approximate leak rates by equivalent orifice diameter and pressure. The app then applies the DOE sharp-edged-orifice factor of 0.61 and linearly interpolates pressure between source rows.

That basis is useful for planning but it is not a measured leak survey. Real fittings, cracks, tubing damage, valve packing, regulator vents, temperature, and pressure stability can change flow. Ultrasonic detector readings also require detector-specific charts, distance and angle control, background-noise handling, or calibrated reference leaks before being converted to SCFM.

For decision use, pair the equivalent-orifice estimate with measured survey notes: leak location, pressure during measurement, detector model, method used, accessibility, isolation requirement, critical instrument served, and whether the apparent leak is normal device bleed or a repairable fault.

Cost Calculation: From SCFM to Dollars

The app cost formula is: Annual cost = SCFM × kW_per_CFM × annual runtime hours × $/kWh. The default specific power is 0.18 kW/CFM, matching the DOE example value of 18 kW per 100 cfm. Replace that default with measured compressor power/load data or current OEM/CAGI data before using the result for savings claims.

The load factor accounts for the fact that not all leaks are active 24/7. Leaks on headers and main distribution lines are continuous (load factor = 1.0). Leaks on branch lines that are isolated during shutdowns have a lower load factor (0.85-0.95 depending on the operating schedule). Leaks on equipment that is only pressurized during operation may have a load factor of 0.5 or less.

Example: A 1/32-inch equivalent leak at 80 psig comes out near 0.79 SCFM after the DOE sharp-edge factor. At 0.22 kW/CFM and $0.08/kWh for 8,760 hours, that planning cost is about $122/year. A survey with 200 similar measured leaks would total near $24,400/year, before any compressor-control, pressure, tariff, repair-cost, or verified-savings adjustment.

The indirect effects are harder to quantify. Lost flow can reduce reserve compressor capacity and increase loaded runtime, but the amount depends on receiver storage, compressor controls, dryer/filter pressure drop, header pressure, and demand profile. Do not claim avoided compressor runtime or shutdown of a standby compressor without trend data.

Reliability Impact for Instrument Air

Instrument air leaks have a reliability dimension that plant air leaks do not, because instrument air feeds control valves, I/P converters, analyzers, and other pneumatic devices. However, the cost calculator only provides a local lost-flow heuristic. It does not predict valve stroke time, device accuracy, or safety-system performance.

If instrument air header pressure drops below actual device requirements, consequences can include slower response, inability to reach output pressure, calibration issues, or failure to meet a required stroke-time test. Those conclusions require measured header pressure at the device, receiver/dryer/filter/regulator data, device manuals, stroke testing, and qualified controls or process-safety review.

A pressure excursion study involves measuring the header pressure at multiple points during peak demand periods (startup, large compressor trip, simultaneous valve stroking during a process upset). If the pressure drops below the minimum instrument requirement at any point, the leak repair priority increases because the leaks are directly compromising process control and safety system reliability.

Some plants prioritize leak repairs based on proximity to critical instruments, but that priority is a site risk decision. A leak on a branch feeding an ESD valve or critical control loop should be reviewed against the plant safety program, operating mode, isolation needs, proof-test requirements, and repair procedure before it is treated as urgent or safe to repair.

Planning Repairs Without Overclaiming ROI

Leak cost can help rank follow-up work, but repair timing is controlled by accessibility, isolation requirements, line service, instrument criticality, process-safety risk, labor availability, and whether the leak is normal bleed or a fault. A calculator result is not a repair authorization.

High-cost equivalent leaks deserve measured confirmation and repair planning. Leaks serving critical devices deserve controls/process-safety review even if their energy cost is modest. Low-cost or inaccessible leaks may wait for outage work if isolation, scaffolding, or device disassembly would create more risk than immediate repair justifies.

Before repair, verify lockout/tagout, depressurization, hot-work, area classification, bypass/permissive needs, alarm suppression, proof testing, and return-to-service steps. After repair, verify the leak is gone and record the measured result so future savings claims are based on evidence.