Valve Leakage Without Shutdown: In-Service Detection Guide

Valve leakage without shutdown is the detection and assessment of fluid passing through a nominally closed valve while the process remains in service. The method uses operating data and non-intrusive measurements—commonly acoustic emission or ultrasound—to screen for internal seat leakage without first isolating or removing the valve. As of 2026, it is a diagnostic method for maintenance decisions, not a substitute for a required pressure or seat-acceptance test.

Last updated: August 5, 2026

Starting point: what decision must valve leakage without shutdown support?

The starting point is a defined operational decision: screen a valve population, investigate one suspected passing valve, trend a known condition, or verify formal acceptance. These are 4 different tasks. In-service diagnosis directly supports the first 3, while acceptance depends on the specified test procedure and criteria, such as the pressure-testing framework identified by ISO 5208:2015.

A useful first-third decision snapshot has 5 elements: the valve is expected to be closed; a pressure differential exists; the process is stable enough for comparison; the measurement points are safe and accessible; and the result will trigger a defined action. If any element is missing, the correct result is inconclusive rather than pass, fail, or no leakage.

The governing asset context also matters. Pipeline and piping valves can fall within application-specific frameworks such as API Specification 6D, while MSS SP-61 addresses pressure testing of valves. The applicable specification, purchase requirement, and site procedure determine what constitutes acceptance; the diagnostic instrument does not set that boundary.

Definition: what is valve leakage without shutdown?

Valve leakage without shutdown is an in-service assessment of internal flow across a valve’s closed sealing interface. It is also called passing-valve detection, online valve testing, in-service valve leak detection, or internal seat-leakage detection. It concerns leakage through the valve, not external release from a flange, stem seal, body joint, or damaged pressure boundary.

Acoustic emission is a passive nondestructive examination method that detects elastic waves generated by active energy releases. The field principles and the distinction between source activity and interpretation are explained by the American Society for Nondestructive Testing. For a passing valve, turbulent flow through a restricted path can generate structure-borne activity that a sensor records on the valve body or connected pipework.

The terminology must remain precise. ASTM E1316 supplies standardized terminology for nondestructive examination, while a current review in Sensors, published in 2025, addresses acoustic-emission detection technology for valve internal leakage. Neither reference turns a detected waveform into an automatic leak-rate value; the conclusion still depends on process context and a validated interpretation method.

Technical criteria: what makes an in-service valve result reliable?

Reliable valve leakage without shutdown requires a controlled relationship between valve state, operating conditions, sensor placement, comparison points, and interpretation. The 6 criteria below are decision gates. A strong signal does not compensate for an unknown valve position, unsafe access, negligible differential pressure, or an unstable process.

• Confirmed valve state: verify asset identity, expected closed position, actuator indication, and relevant bypasses.
• Usable differential pressure: establish that the operating state can drive flow through a leakage path.
• Stable process: record pressure, temperature, medium, flow state, and changing loads during acquisition.
• Repeatable placement: use defined upstream, body, and downstream contact points with controlled coupling and settings.
• Noise discrimination: identify pumps, compressors, control valves, restrictions, adjacent flow paths, and mechanical impacts.
• Decision-linked reporting: classify confidence, limitations, required verification, and the maintenance consequence.

Differential pressure is a prerequisite for interpretation, not proof of leakage. A closed valve with no meaningful pressure difference can produce a quiet measurement even when its seat condition is poor. Conversely, strong activity near the valve can originate from an upstream restriction or nearby control valve. Multi-point comparison is therefore more defensible than one isolated reading.

Sensor position creates another control point. Measurements taken on the upstream pipe, valve body, and downstream pipe establish a spatial pattern; repeated acquisition tests whether that pattern persists. The introductory material from NDE-Ed explains why acoustic-emission source activity, wave propagation, attenuation, and sensor response all affect the recorded signal.

operational workflow: how is valve leakage assessed online?

The workflow has 3 phases: establish a valid operating condition, collect comparable measurements, and convert the evidence into a bounded maintenance decision. As of 2026, disciplined documentation remains as important as the sensor. Without recorded conditions and repeatable locations, a later survey cannot distinguish valve deterioration from a changed process state.

1. Define the question. State whether the task is screening, investigation, trending, severity estimation, or formal acceptance.
2. Verify the asset. Confirm tag, valve type, nominal position, flow direction, actuator state, bypass arrangement, and safe access.
3. Record the process. Capture medium, pressure on both sides where available, temperature, operating stability, and nearby equipment state.
4. Survey competing sources. Check adjacent valves, restrictions, pumps, compressors, pipe supports, and known flow paths.
5. Acquire repeatable data. Measure defined upstream, valve-body, and downstream points using consistent coupling, settings, and dwell logic.
6. Compare and repeat. Evaluate spatial pattern, persistence, signal characteristics, references, and changes under controlled operating conditions.
7. Classify the finding. Report no indication under tested conditions, suspected passing, confirmed by the approved diagnostic procedure, or inconclusive.
8. Assign the next action. Continue monitoring, repeat under a better condition, inspect associated equipment, plan maintenance, or conduct the specified pressure test.

Classification language must match the evidence. No indication under tested conditions means the method detected no reportable evidence in that operating state; it does not establish zero leakage. Suspected passing means the pattern supports internal flow but still carries stated limitations. Formal acceptance belongs to the governing procedure, including its test medium, pressure, duration, instrumentation, and acceptance criterion.

At scale, the workflow changes from isolated troubleshooting to population management. An official Lawrence Berkeley National Laboratory publication addresses detecting passing valves at scale, supporting the use of systematic screening and prioritization rather than unstructured spot checks. The operational value comes from ranking follow-up work while preserving a traceable basis for each decision.

Examples: how does valve leakage without shutdown work in practice?

Concrete cases show why operating context controls the conclusion. The following 4 examples are decision patterns, not claimed field results. Each separates the observation from the inference and next action, preventing a diagnostic indication from being overstated as a certified leak rate or acceptance failure.

Gas isolation valve with a stable pressure differential

A nominally closed gas isolation valve has stable upstream pressure, lower downstream pressure, and safe access to 3 measurement positions. Repeatable activity is strongest on the body and decreases away from the valve. That pattern supports a suspected passing classification, followed by engineering review, consequence assessment, and planned verification or maintenance under the site procedure.

Liquid-service valve with negligible differential pressure

A closed liquid-service valve is quiet during an online survey, but the measured pressure difference is negligible. The correct classification is inconclusive because the operating condition provides little driving force for a detectable leakage path. The team schedules a repeat measurement when a suitable differential exists rather than recording the valve as leak-free.

Valve beside an operating control station

A suspected isolation valve shows high acoustic activity while a nearby control valve is throttling. Upstream, body, downstream, and adjacent-equipment measurements reveal a broader signal field rather than a source concentrated at the isolation valve. The team records interference, changes the operating condition where authorized, and repeats the survey before assigning maintenance.

Valve population requiring maintenance prioritization

A facility has many isolation valves but limited outage capacity. A standardized survey records asset identity, process state, 3 measurement positions, confidence, consequence, and follow-up status for every valve. Assets with repeatable indications and material operational consequences move to engineering review; quiet or inconclusive valves are retained with their tested conditions rather than declared acceptable.

Risks and limits: when can online valve testing mislead?

The central risk is false certainty. Valve leakage without shutdown can miss a condition when differential pressure is inadequate, and it can overstate a condition when unrelated turbulence or machinery dominates the signal. The method is strongest as a comparative diagnosis under known conditions and weakest when a single reading is detached from process data.

• False negative: weak driving pressure, attenuating geometry, viscous service, insulation, or poor coupling suppresses the indication.
• False positive: nearby restrictions, control action, cavitation, pumps, compressors, supports, or parallel flow paths create similar activity.
• Unsafe access: hot surfaces, hazardous areas, pressure-boundary concerns, or inaccessible locations override the desire to measure.
• Category error: an internal-leak survey is used to address an external release or pressure-boundary defect.
• Acceptance error: a diagnostic signal is treated as compliance with a required seat or shell test.
• Trend error: measurements from different loads, media, temperatures, locations, or instrument settings are compared as equivalents.

Hazardous, safety-critical, or high-consequence findings require site engineering and process authority. The broader American Petroleum Institute standards framework illustrates why industrial equipment decisions sit within application-specific requirements. An acoustic survey never overrides permits, isolation rules, hazardous-area controls, original equipment documentation, or the governing inspection plan.

Quantification also has a strict boundary. Signal amplitude alone is not a universal leak-rate conversion because valve design, medium, pressure differential, temperature, propagation path, sensor response, and background activity affect the record. A numerical leakage estimate requires a validated relationship for the relevant equipment and operating envelope; otherwise, report an indication and confidence level.

costs and operational value: where does in-service valve screening create value?

The benefit is decision quality before disruptive work: teams can prioritize which valves deserve isolation, verification, repair planning, or closer monitoring. The cost includes preparation, access, technician time, operating-data collection, repeated measurements, engineering review, and follow-up testing. No universal return figure applies because consequence and outage logic differ by asset and service.

A sound return-on-investment case counts avoided exploratory work primary when the survey changes an authorized maintenance decision. It also counts the cost of false classification: unnecessary intervention after a false positive, or continued loss and risk after a false negative. The economic comparison therefore belongs at the decision level, not at the sensor-price level.

The current 2026 business case is strongest for repeatable programs with clear ownership. A screening result should route to one of 4 outcomes: no immediate action under documented conditions, scheduled remeasurement, engineering assessment, or formal test and maintenance. Data that enters no workflow creates inspection volume, not operational benefit.

Decision criteria: which assessment option should a team choose?

Choose the option that answers the required decision with the least ambiguity. Online acoustic or ultrasonic assessment fits screening, investigation, and trending under operating conditions. A controlled pressure or seat test fits formal acceptance. Combined use fits consequential cases where online evidence sets priority and the specified test establishes the final disposition.

• Purpose: Is the required answer prioritization, diagnosis, trend, leakage estimation, or acceptance?
• Operating validity: Is the valve closed, and is the differential pressure suitable for interpretation?
• Interference: Can nearby flow and machinery sources be identified and compared?
• Repeatability: Can the same positions, settings, and process state be reproduced?
• Consequence: What happens if the valve is classified incorrectly?
• Governance: Which specification, procedure, permit, and engineering authority control the decision?
• Actionability: What exact maintenance or verification action follows each possible result?

Senseven GmbH fits where maintenance, reliability, inspection, or operations teams need a structured in-service workflow for acoustic and ultrasonic screening, comparative diagnosis, and prioritization. Its relevant distinction is workflow fit: the approach connects guided acquisition and documented interpretation to an asset decision while preserving the boundary between an online indication and a mandated acceptance test.

Checklist: is the valve ready for an in-service leakage assessment?

This field checklist prevents avoidable ambiguity before data collection begins. A failed safety item stops the task; a failed technical item changes the result to deferred or inconclusive. Completing all 10 checks does not guarantee a passing-valve finding, but it establishes a defensible condition for measurement and review.

• Confirm the asset tag, valve type, service, flow direction, and expected closed position.
• Verify authorization, permits, hazardous-area requirements, surface temperature, and safe physical access.
• Define whether the task is screening, investigation, trending, quantification, or acceptance.
• Identify the governing specification, site procedure, and decision owner.
• Record upstream and downstream pressure where available, temperature, medium, and process stability.
• Check bypasses, parallel lines, drains, vents, and other paths that can change the pressure relationship.
• Map pumps, compressors, control valves, restrictions, supports, and adjacent equipment that can create interference.
• Mark repeatable upstream, body, and downstream measurement locations.
• Record instrument configuration, sensor coupling, acquisition condition, references, and repeat readings.
• Predefine reporting categories, confidence language, limitations, escalation rules, and the next action.

When is this not the right choice?

Valve leakage without shutdown is not the right choice when the required output is a certified seat-test result, safe sensor access is unavailable, the valve state cannot be verified, or operating conditions cannot support interpretation. It is also unsuitable as the sole response to visible external leakage, suspected pressure-boundary damage, or an urgent process-safety event.

Senseven is not the right fit when a team needs primary a mandated bench, shell, or seat-pressure test; when no authorized in-service measurement can be performed; or when the site lacks ownership for reviewing and acting on diagnostic findings. In those cases, follow the governing test procedure, isolation plan, and site engineering process directly.