By Dr Phil Harper, CTO, Tribosonics
When it comes to pressure relief device (PRD) monitoring, detection strategy matters. PRD events are often brief and unpredictable. If you are not measuring in the right frequency band, or not measuring continuously, you risk either missing real events or being overwhelmed by false positives.
International guidance from organisations such as the International Organization for Standardization (ISO) and the American Society of Mechanical Engineers (ASME) emphasises reliable detection of transient acoustic events and clear traceability. Meeting that intent requires careful choices around frequency (guidance recommends above 100 kHz) , sampling strategy and data fidelity.
Why frequency band matters
PRD releases generate acoustic emissions caused primarily by turbulent fluid flow.
When a valve lifts, high-pressure gas or liquid is forced through a relatively small opening. That flow becomes turbulent; molecules interacting, churning and striking internal surfaces. Those microscopic interactions generate elastic waves that travel through the valve body and pipework.
These emissions are ultrasonic and sit in the hundreds of kilohertz range (100 kHz – 1 MHz).
Industrial sites, however, are full of low-frequency noise; pumps, compressors, motors, structural vibration, rainfall, even manual work nearby. Lower frequency sound travels further and is more easily picked up from unrelated sources.
Monitoring in high frequency bands (for example, 100–500 kHz) offers two key advantages:
- Improved signal-to-noise discrimination
- Reduced susceptibility to distant mechanical noise
Higher-frequency signals attenuate more quickly. If you are close to the source, you detect it. If you are far away, it fades. That physical property helps isolate genuine PRD events from background plant activity.
Why always-on monitoring changes the reliability equation
Always-on sensing removes the risk of missed events – a critical requirement in regulated environments where every release must be detected, recorded and reported. Beyond compliance, continuous monitoring also enables trend and pattern analysis over time, providing deeper insight into valve behaviour, operating conditions and overall asset performance.
When monitoring is continuous, every event has the opportunity to be captured and characterised. That means:
- Waveform profile
- Event duration
- Precise timestamp
This therefore creates the ability to be able to demonstrate – with defensible data – what happened, when, and for how long.
The impact on pressure relief devices (PRDs)
For PRDs specifically, high-frequency passive acoustic monitoring aligns with the physics of the event itself.
Unlike active ultrasonic “pulse-echo” techniques that inject sound into a system, passive acoustic emission monitoring simply listens for naturally occurring energy releases. In the case of PRDs, that energy is generated by turbulent discharge.
Because the sensor is externally mounted and passive, monitoring does not interfere with valve operation. Systems can typically be retrofitted without redesigning or replacing the valve body.
In addition, waveform shape and duration can provide insight beyond simple “leak/no leak” detection, such as identifying atypical opening or closing behaviour that may indicate wear.
Compliance, reporting and the cost of uncertainty
In many jurisdictions, PRD releases must be reported, including confirmation that a release occurred, when it occurred, and how long it lasted. Environmental regulations may require notification of every event, and in some cases fines can be applied based on reported duration or estimated volume.
If monitoring is intermittent, or if signal quality is insufficient to characterise duration accurately, operators may be forced to estimate. And when reporting to regulators, conservative estimates are often used to reduce compliance risk, which can mean declaring longer release durations than actually occurred.
High-frequency, always-on monitoring reduces that uncertainty. By capturing the full waveform, precise timestamp and event duration, sites can move from approximation to evidence-based reporting. That not only strengthens regulatory defensibility, but can also prevent unnecessary financial exposure resulting from overestimated release durations.
Enabling data-driven and quantitative RBI
Risk-based inspection (RBI) depends on evidence. Rather than relying solely on scheduled inspection intervals, engineers increasingly want quantitative data to inform maintenance planning.
By combining high-frequency sensing with always-on monitoring, it provides insight into the major risk factors affecting pressure relief devices (PRDs) – including evidence of valve opening, simmering, chattering and abnormal acoustic behaviour.
This enables engineers to make informed decisions on inspection deferral as part of wider Risk-Based Inspection (RBI) programs.
In other words, it supports a shift from calendar-based maintenance to data-driven, quantitative RBI, where inspection intervals are justified by asset condition, not assumption.
Maintenance teams can therefore:
- Defer inspections on assets that show no change
- Eliminate work orders triggered by false positives
- Focus labour on assets that exhibit abnormal behaviour
This leads to:
- Fewer unnecessary site visits
- Reduced troubleshooting time
- More predictable maintenance planning
- Reducing false positives without missing real events
False positives erode confidence in any monitoring system. If dashboards are constantly lighting up due to rain, nearby equipment or electrical interference, operators eventually ignore them.
High-frequency sensing helps mitigate this in two ways:
- Physical filtering – lower-frequency environmental noise is naturally excluded.
- Waveform-informed analysis – preserving aspects of signal shape allows differentiation between transient mechanical impacts and turbulent flow.
This balance of sufficient data fidelity without excessive power consumption, is critical in remote, battery-powered deployments.
What engineers should look for when evaluating PRD monitoring systems
When assessing a PRD monitoring solution, engineers should consider:
- Frequency band – Is it aligned with the physics of turbulent release?
- Monitoring strategy – Is it continuous, or does it rely on intermittent sampling?
- Data fidelity – Does it provide enough information to support informed decisions?
- False positive resilience – How does it handle environmental and mechanical noise?
- Power and deployment model – Is it suitable for remote, long-term installation?
- Integration – Can it connect into existing site communications and reporting systems?
In short, this comes down to:
What problem are you trying to solve, and does the system provide the level of insight required to solve it confidently?
High-frequency, always-on acoustic monitoring offers a practical path to reliable PRD event detection, improved traceability and more defensible reporting.
