Leak Detection for H₂ & Gas Turbines: Catalytic Bead vs TCD vs Optical Sensing

By Green Gas Turbines Team · Published November 23, 2025 · 11 min read

Why Leak Detection Needs Its Own Design, Not Just a Spec Line

As gas turbines migrate from pure natural gas to hydrogen blends, ammonia pilots, and higher-pressure fuel systems, leak detection becomes a core part of the safety case. It is not enough to “add some detectors” to a P&ID. You need the right sensing technology, in the right place, configured for the right failure modes.

This article compares three workhorse technologies for combustible gas leak detection around turbines and hydrogen yards:

• Catalytic bead (pellistor) detectors
• Thermal conductivity detectors (TCDs)
• Optical gas sensing (primarily infrared point and open-path)

We will focus on how each works, where each shines, and how to combine them intelligently on a power site handling natural gas, hydrogen, and blends.

How Each Technology Works (In Engineer Terms)

Catalytic Bead (Pellistor) Sensors

Catalytic bead detectors are the “classic” industrial flammable gas detector. Inside the sensor, a small bead coated with catalyst is heated electrically. When combustible gas and oxygen reach the bead, the gas oxidises on the surface, heating the bead. The electronics measure the resulting change in resistance.

Key characteristics:

• Measures: Combustible gas concentration as a percentage of the Lower Explosive Limit (LEL), typically 0–100% LEL.
• Requires oxygen: No O₂, no combustion, no signal. Not suitable in fully inerted spaces.
• Broad response: Responds to most flammable gases and vapours, including methane and hydrogen.
• Vulnerable to poisoning: Certain silicones, sulfur compounds, or lead can permanently reduce sensitivity.
• Heated element: The bead itself runs hot, but is enclosed in a flameproof or intrinsically safe housing.

Thermal Conductivity Detectors (TCD)

Thermal conductivity detectors compare how quickly a gas removes heat from a heated element versus a reference gas (usually air). Hydrogen and helium have very high thermal conductivity compared to air or methane, so they produce a strong signal.

Key characteristics:

• Measures: Change in thermal conductivity of the sample gas versus a reference.
• Excellent for hydrogen: Hydrogen’s high thermal conductivity makes TCD very sensitive to H₂ across a wide range.
• Wide range: Can measure from low concentrations up to high volume percent; not limited to 0–100% LEL.
• Not gas-selective: Any gas with different thermal conductivity than the reference contributes to the signal.
• Common form factor: Often packaged as a process gas analyser rather than a simple point alarm detector.

Optical Gas Sensing (Point IR and Open-Path)

Optical gas detectors use light absorption to detect gas along a path. Most industrial units use infrared (IR) wavelengths where gases like methane, propane, and CO₂ absorb strongly. A source emits IR light, the gas absorbs at specific wavelengths, and the detector measures the drop in intensity.

Key characteristics:

• Point IR detectors: Fixed point devices with a short optical path inside the housing. Good for localised hydrocarbon leak detection.
• Open-path IR detectors: Emitter and receiver are separated by several metres to hundreds of metres; measure average gas concentration along the beam path.
• No oxygen requirement: Works in inert or oxygen-poor atmospheres.
• Not suited to hydrogen with standard IR: H₂ has very weak IR absorption in the bands used for hydrocarbon detection. Conventional IR detectors that are perfect for methane will not see hydrogen unless specifically engineered for it.
• Fast response and low maintenance: No chemical consumption, no hot bead, less susceptible to poisoning.

Beyond IR, there are emerging optical techniques for hydrogen (e.g. fibre-optic sensors with palladium coatings, time-of-flight LIDAR for methane plumes), but these are still specialised and less common on standard GT sites.

Side-by-Side Comparison for Power Plant Use

Choosing for Natural Gas vs Hydrogen vs Blends

For Pure Natural Gas Turbines

For conventional natural gas GT sites, optical IR is already common:

• Point IR detectors near flanges, manifolds, and inside enclosures.
• Open-path IR detectors across pipe racks or between equipment rows as a “curtain” to detect larger releases.
• Catalytic bead detectors sometimes used where legacy systems exist or where broader flammable-vapour coverage is needed.

IR’s main advantages here: it is not consumed, not poisoned easily, and can operate in oxygen-poor spaces. For methane-heavy gas, this is often the default choice.

For Hydrogen-Rich or Hydrogen-Blend Systems

Hydrogen changes the picture:

• Standard hydrocarbon IR detectors do not see H₂; their bands target CH-stretch absorption, not hydrogen’s much weaker IR features.
• Catalytic bead detectors will respond to hydrogen leaks in ventilated areas, but require oxygen and are more prone to poisoning in some industrial environments.
• TCD-based detectors can give strong signals for hydrogen over a wide range, including high-concentration piping or enclosed volumes.

For a hydrogen-capable turbine site, a common strategy is:

• Use catalytic bead or TCD sensors in local enclosures and near hydrogen equipment (valve skids, electrolyzer rooms, manifolds) for point detection.
• Keep optical IR for methane (if you are blending natural gas and hydrogen) in pipe racks, compressor halls, and fuel-gas areas.
• Supplement with process analyzers (e.g. TCD or gas chromatographs) on key fuel headers to measure H₂ fraction and provide composition/Wobbe signals to the GT controls.

Design Principles for Sensor Layout on GT / H₂ Sites

Start With Release Scenarios, Not Device Types

Before picking sensor technologies, define the credible leak scenarios:

• Small flange leak, low rate, in a ventilated area
• High-rate rupture inside an enclosure or below a platform
• High-pressure gas release in a yard or pipe rack, forming a buoyant or semi-buoyant cloud
• Hydrogen leak inside an equipment building or electrolyzer room

For each scenario, consider where the gas will accumulate (ceiling/roof for hydrogen, lower areas for heavier-than-air gases) and how fast you need to detect it to trip, ventilate, or isolate.

Use the Strengths of Each Technology

• Catalytic bead: Robust, relatively low-cost coverage for many flammable gases. Good for general area coverage where poisoning risk is manageable and oxygen is present.
• TCD: Ideal for hydrogen measurement over a wide range; good for process streams, vents, or high-concentration areas where LEL-based devices are not appropriate.
• Optical IR: Excellent for methane and hydrocarbon detection, especially as open-path systems to form detection “curtains” around pipe racks and tank farms.

Avoid Common Mistakes

• Relying only on hydrocarbon IR for a hydrogen yard: You might detect a tiny methane carrier gas but completely miss pure hydrogen.
• Installing catalytic bead sensors in oxygen-poor zones: They will appear “quiet” even if gas accumulates.
• Ignoring plume dynamics: Hydrogen jets will rise and collect near roofs; methane may linger at mid-levels in poorly ventilated structures. Place sensors where gas will actually pass, not where it is easiest to mount them.
• Underestimating maintenance: Leak detection is a safety function; no calibration or bump testing means you are flying blind.

Specifying Sensors: What to Ask Vendors

Regardless of technology, ask for:

• Response time: T₅₀ and T₉₀ in seconds for the target gas, including at low temperatures.
• Cross-sensitivity: Which other gases cause response or interference?
• Poisoning and inhibition: Which chemicals degrade catalytic beads or coatings; what is the expected lifetime?
• Environmental ratings: Temperature, humidity, IP rating, vibration limits.
• Hazardous area approvals: Ex d / Ex i, gas group (IIC for hydrogen), temperature class.
• Self-diagnostics: Does the device detect optical path blockages, lamp failure, filament breaks, or internal faults and expose them via fault relays?
• Calibration intervals and tools: How often, what gas, and is there a standard procedure that your team can realistically follow?

Operations, Testing, and Human Factors

Even the best sensor mix fails if operations and testing are weak.

• Define clear alarm actions: For each alarm level (e.g. 10% LEL, 20% LEL) specify what happens: ventilation start, fuel isolation, turbine trip, muster, etc.
• Regular bump tests: Use test gas or test lamps to verify sensor response; log failures and trends.
• Training: Operators should understand the difference between technologies—what a “fault” vs “alarm” means, and where each detector is located.
• Change management: If you reroute piping, add new hydrogen equipment, or change ventilation patterns, revisit your leak detection strategy.

Recommended Patterns for GT + Hydrogen Projects

For a modern gas turbine project transitioning to hydrogen blends, a pragmatic pattern is:

1. Use IR open-path detectors across natural gas and blend manifolds, pipe racks, and compressor areas for hydrocarbon leak coverage.
2. Deploy catalytic bead or TCD point detectors in hydrogen-specific locations: electrolyzer rooms, hydrogen valve skids, storage manifolds, and turbine enclosure roof spaces where H₂ could accumulate.
3. Add process TCD or GC analyzers on fuel headers to measure hydrogen percentage for controls and safety interlocks.
4. Combine detection with ventilation and isolation logic, so alarms are actionable (ventilation start, valve closure, safe shutdown) rather than just indications on a panel.
5. Design testing and maintenance into the layout: Make sure technicians can safely access detectors for calibration and replacement without special scaffolding every time.

Conclusion: Pick the Right Tool for the Job, Then Combine Them

No single sensor technology is “best” for every leak. Catalytic bead, thermal conductivity (TCD), and optical IR-based detectors each solve different parts of the problem:

• Catalytic bead is a robust workhorse for general flammable gas monitoring where oxygen is present.
• TCD is a powerful ally for hydrogen measurement and process analytics over a wide concentration range.
• Optical IR is hard to beat for methane and hydrocarbon coverage, especially in open-path configurations.

The most resilient hydrogen-ready gas turbine sites treat leak detection as a system: credible scenarios, correct placement, complementary technologies, and disciplined testing. Done right, leak detection becomes a quiet, reliable layer in your hydrogen safety case—one that you can defend to regulators, insurers, and your own operations team.