Leak Detection

Leak Detection

New completion based on supplied sources.

Produce a repairable record

Leak detection should identify where compressed air is escaping, document the conditions, assign a repair priority and enable a later recheck. Its useful output is a traceable action register. Photographs and detector readings help the investigation, but a collection of images without exact locations, owners and follow-up cannot demonstrate that losses were reduced.

Define the survey boundary and purpose before visiting the plant. A search for urgent safety or process defects differs from a systematic energy-loss survey. Both may be needed, but their priorities and evidence should remain clear. Detection, quantification, repair and verified savings are separate stages. Completing one stage does not justify claiming all the others are complete.

Map the areas and operating states

Divide the system into identifiable zones and record the equipment, branches and consumers included. Mark areas that cannot be accessed and states that cannot be tested. Note whether the survey occurs during production, idle operation or a controlled shutdown. Pressure and operating state can affect both leakage and the ability to detect it.

Use a route that allows coverage to be checked later. Record completed, partially completed and uninspected areas. An inaccessible elevated connection should remain an open coverage limitation rather than disappear from the report. If an important machine operates only intermittently, arrange a relevant observation or state that its active condition was not surveyed.

Establish safe access and boundaries

Survey work must follow the site's access, pressure-system, electrical, work-at-height and machinery procedures. A detector does not authorize reaching into moving equipment or loosening a pressurized fitting. Identify restricted areas and the support needed for access. Keep investigation and repair activities within the approved scope.

Where a leak presents an immediate safety or process concern, escalate through the site's responsible personnel. Its priority may exceed the largest estimated energy loss. Do not approach a damaged hose or pressure component merely to obtain a better photograph. Record what can be observed safely and arrange competent assessment and isolation where required.

Select the detection method for the task

Ultrasonic detection can help locate escaping compressed gas amid audible background noise. Acoustic imaging may assist localization over a wider scene. Suitable leak-detection fluid can support confirmation at accessible connections when compatible with the application. Pressure or flow tests can help assess a defined zone. Each method has its own limits and should be described accurately.

Detection equipment can indicate a likely source without directly measuring its flow. Background sources, reflections, access angle and operating noise can affect interpretation. Follow the instrument's method and use confirmation where needed. Do not convert every indication into an exact flow or annual rupee figure without a documented quantification basis.

Record pressure and context

Record the local or relevant system pressure, production state and whether the component is normally energized or isolated. Include the date and time so observations can be linked to other logs. A leak observed at one pressure may behave differently at another; a machine valve may leak only in a particular state.

If pressure is unavailable, state that limitation. Avoid silently using a nominal plant setpoint as though it were measured at the leak. The same applies to assumed annual operating hours. A hose connected only during occasional work should not automatically be assigned the entire plant's operating calendar.

Localize the exact component

Move from a broad indication to a specific joint, hose, valve, seal, drain or component where safe and practical. Record an unambiguous location using the asset ID and physical description. A photograph should show enough context to find the item and, where useful, a closer view of the exact source. Use a durable tag or a reliable digital reference appropriate to the site.

Avoid duplicate records for the same source detected from different angles. Conversely, do not combine several independent leaks into one vague machine-level entry if they require different repairs. A clear record supports work planning, parts identification and re-verification. If localization remains uncertain, label the finding as a suspected source requiring further diagnosis.

Distinguish leakage from intended discharge

Some equipment intentionally vents or purges during normal operation. Dryers, controls and process devices may have legitimate air use that still deserves optimization but should not be mislabelled as a failed joint. Establish the component's intended function and cycle before classifying the indication. A drain can also discharge normally for a short period or fail open continuously.

Record whether the observed release is continuous, intermittent, state-dependent or unresolved. Where an intended use appears excessive, create a separate investigation record with its own evidence. This prevents the leak register from overstating repairable defects and helps the correct specialist evaluate an inefficient but functioning process.

Quantify only with a documented method

If a detector provides a flow estimate, retain the instrument, settings, distance, pressure assumptions and applicable method. Identify the estimate's limitations and uncertainty. If a known orifice or another model is used, state its geometry and gas-flow assumptions and whether those conditions were verified. Do not treat an irregular damaged joint as a precisely known calibrated orifice.

Flow measurement or a controlled zone test may provide a stronger basis for some decisions, but it also has boundary and operating requirements. Select the method appropriate to the significance of the conclusion. A qualitative priority can be useful when precise quantification is unavailable, provided it is not disguised as a measured numerical loss.

Use controlled zone tests carefully

A defined zone can sometimes be evaluated by measuring its non-production flow or pressure decay under approved conditions. Confirm which consumers remain active and which boundaries are isolated. Storage, temperature change, passing valves and essential uses can affect the result. Record the test setup and exclusions so the result is not attributed entirely to leaks without justification.

For a pressure-decay calculation, vessel and connected volume, absolute pressure, temperature behaviour and test duration matter. No isolation test should interrupt a critical service or exceed the site's approved operating procedures.

Prioritize by consequence and recoverability

Assess personnel safety, process reliability, product quality, accessibility, repair complexity, estimated loss and recurrence. A small leak on a critical control line may outrank a larger non-critical leak. Record the reason for priority and the person responsible for action. This allows maintenance to plan a sensible sequence rather than simply sorting by an uncertain cost estimate.

Consider whether the proposed repair is feasible during operation or requires a shutdown, replacement assembly or specialist work. An inaccessible or deferred leak remains part of the known condition. Record its risk owner and review trigger. Do not count it as recovered capacity or savings until the relevant repair and verification have occurred.

Prepare a useful repair instruction

The finding should identify the asset, exact source, component type, observed state, evidence and required planning information. It should not prescribe an unverified part or improvised tightening method. Use the parts compatibility process where replacement is needed. The maintenance team must confirm the repair method against the actual component and site procedures.

Include access restrictions and any requirement for isolation, depressurization or process coordination. If the defect may involve a damaged pressure-containing component, escalate appropriately instead of reducing it to a routine seal change. The survey's value is to make the next action clear while preserving the boundary between detection and competent repair design.

Keep estimated airflow separate from energy savings

Leak airflow can be expressed on a stated reference basis when a suitable method supports it. Converting that flow to energy requires the compressor station's response, operating hours and a compatible performance basis. A reduction in demand may produce more unloaded operation rather than the proportional power reduction implied by a simple average ratio.

For a preliminary scenario, clearly label the assumed relationship and all inputs. Under the owner's instruction, unsupported savings are excluded from project conclusions. A measured reduction in leakage or zone flow is a useful result even before an energy saving is verified. Do not inflate that result with a precise annual financial figure based on unverified hours and tariff assumptions.

Worked register example

Consider an illustrative finding tagged L014 at machine M07's inlet coupling. The survey records a continuous indication during production at a measured local pressure, a context photograph and the exact coupling location. The repair owner confirms that replacement requires a planned isolation window and a compatibility check. The record remains open until the repair occurs.

After replacement, the same location is rechecked under the relevant pressure and machine state. The indication is no longer detected within the method's capability. That supports verified local closure. It does not, by itself, establish annual electrical savings. A separate station or zone measurement is needed to support that additional outcome.

Worked example of a false savings claim

A survey lists fifty indications and assigns every one the same annual operating hours and energy rate. Several are intermittent process vents, some are duplicate observations and some are in areas that could not be confirmed. Summing the assumed values produces an impressive total but not a defensible project result.

The corrected approach resolves duplicates, distinguishes intended use, records coverage limitations and quantifies only where the method supports it. Repairable findings receive owners and priorities. Energy claims remain withheld until the relevant conditions and post-repair response are measured. The improvement is in decision quality and traceability, not in making the headline total larger.

Recheck the same location and condition

Verification should return to the exact tagged source and the operating condition that made the leak observable. A check while the branch is depressurized cannot prove the leak is closed under service. Record pressure, state, method, date and verifier. If the source was intermittent, include the relevant cycle or identify the remaining limitation.

Where repair creates a new leak nearby or the original indication remains, keep the item open or reopen it with a linked record. Do not erase the first finding. The history helps identify poor fittings, vibration, unsupported hoses, unsuitable materials or repeated assembly problems.

Measure the system response after repairs

Compare suitable pre- and post-repair periods with production, pressure and control state accounted for. Confirm whether airflow, compressor loading and electrical energy changed. If other interventions occurred at the same time, record them and avoid assigning the whole outcome to the leak programme. The verification boundary should be agreed before the repair campaign where possible.

Also check service quality. Lower demand can change compressor control behaviour and pressure stability. A successful programme should preserve required service while reducing confirmed waste or improving reliability. If power does not change measurably, report that result and investigate the station response rather than claiming a theoretical saving as an achieved one.

Build recurrence prevention into the survey

Group repeated defects by component type, installation method, environment and failure mechanism. Repeated hose or coupling leaks may point to movement, unsupported loads or inappropriate operating practice. Use those patterns to target a durable correction. Simply increasing survey frequency may not address the underlying cause.

Track the age and repair history of recurring items. A new tag can be created for a new event, but it should link to the earlier asset and defect. This distinguishes recurrence from an unrelated discovery and prevents the programme from counting the same unresolved source as multiple successful repairs.

Detection and evidence matrix

Record detection limits in practical language

The survey report should explain where the method was effective and where it was constrained by access, background noise, operating state or instrument range. This does not require a speculative list of every possible failure. It requires a factual account of the conditions encountered. A reader should be able to distinguish an inspected accessible coupling from a concealed assembly that was never tested.

Where a source could be localized but not quantified, record that useful finding without forcing an estimate. Where a source could not be distinguished from intended discharge, assign an investigation action. These dispositions preserve value from the survey while preventing uncertainty from being converted into false precision or an unsupported financial total.

Coordinate surveys with maintenance opportunities

Link the survey route to planned shutdowns and access windows so high-priority findings can be repaired efficiently. Obtain the component identity and likely work scope early enough to resolve parts compatibility. A well-tagged leak discovered immediately before a shutdown may still remain unrepaired if its replacement component or isolation requirement is unknown.

After the window, reconcile the repair list with the original survey IDs and schedule the appropriate rechecks. Do not assume every item on the planned list was completed. Record cancellations, partial work and newly discovered defects. This turns the survey into a controlled maintenance input rather than a report that becomes stale as soon as the shutdown ends.

Interpret a lower non production flow carefully

A lower idle-flow measurement after repair is encouraging when the boundary, pressure and active consumers are comparable. Check whether automatic isolation, dryer operation, essential users or storage conditions changed. If several factors changed, identify them and report the combined observation. Do not assign the entire difference to tagged leaks without a defensible basis.

If the flow meter cannot resolve the remaining low range accurately, report that limitation. A display near zero is not automatically proof of zero leakage. The appropriate conclusion may be that flow fell below the reliable measurement range. This is a useful bounded result and is preferable to a claim of a perfectly leak-free plant.

Specify the survey closeout package

The closeout should contain the coverage record, tagged findings, evidence and quantification methods, repair priorities, owners and unresolved areas. Include the recheck plan and distinguish any repairs already verified during the campaign. The owner should be able to issue work without reconstructing locations from vague descriptions.

If a later energy-verification campaign is needed, define its boundary and required production context separately. That prevents a detection-only visit from being sold or reported as a completed savings project. The physical finding, maintenance result and energy outcome remain connected, but each receives the evidence appropriate to its own claim.

Questions about leak surveys

Can ultrasonic sound be converted directly to rupees

Only through a supported quantification and energy model with stated conditions, hours and costs. A detector indication alone is insufficient. Keep estimates and measured outcomes clearly separated.

Should the largest estimated leak always be repaired first

No. Safety and critical-service consequences can take priority, and repair feasibility matters. Use a documented priority process that includes both risk and loss.

Is a silent area leak free

Not necessarily. Detection depends on method, access, pressure and operating state. Report surveyed coverage and detection limits rather than claiming the entire area is leak free without adequate evidence.

Does a closed work order mean the leak is closed

It demonstrates an administrative or maintenance milestone. Technical closure requires a relevant recheck and evidence. Keep repaired and verified states separate.

Can a survey promise a fixed saving

No. The actual result depends on confirmed defects, repair completion, operating conditions and compressor response. Unsupported percentage or annual-cost claims are excluded from project conclusions.

Enquiry information and next action

Provide a system or zone diagram, operating schedule, pressure levels, access restrictions and any existing leak register. State whether the purpose is urgent defect identification, a systematic repair programme or verification of earlier work. Reddy Compressors can use this information to define survey coverage and the evidence needed for useful follow-up.

Preserve PROJECT ENGINEERED and PROJECT-SPECIFIC status. The survey should produce actionable evidence without implying that every detected indication is a recoverable financial saving.