Fiber-Optic Gas Sensors in High-Speed Railway Tunnel Construction
Why High-Speed Railways Keep "Running Into" Gas
As the high-speed railway network extends into the mountainous southwest, long tunnels crossing coal-bearing strata and passing near mined-out areas are increasingly common. Under the Code for Design of Railway Gas Tunnels (TB 10120—2019), a tunnel passing through gas-bearing strata must be classified as a gas tunnel, further categorized by gas emission characteristics into one of four classes — slightly gassy, low-gas, high-gas, or gas outburst prone — with vastly different protection, equipment and management requirements for each class.
The main component of gas is methane, with explosive limits of roughly 5%–15% (by volume), yet construction control thresholds are far lower: in general sections, methane concentration in the airflow should not exceed 0.5%; if exceeded, ventilation must be intensified. At 1.0% in the working face airflow, drilling and blasting must stop; at 1.5%, construction must be halted, personnel evacuated and power cut off. In other words, the monitoring system must intercept the danger long before it can explode. This demands detection equipment that is "accurate, fast-acting, and never adds trouble."
The Two Legs of Traditional Monitoring: Adequate but Not Great
Tunnel gas monitoring currently runs a dual system of "manual inspection + automated monitoring," with the two cross-validating each other.
Manual inspection relies on dedicated gas inspectors working three shifts with portable gas detectors and interference-type methane detectors; blasting operations also require the "one blast, three checks" procedure (before charging, before firing, after firing). The code requires at least 2 checks per shift in slightly gassy worksites and at least 3 per shift in low- and high-gas worksites — but no matter how diligent, manual inspection is still "sampling," and the gap between two checks is precisely the risk window.
Automated monitoring systems (gas monitoring, alarm and power-interlock devices) provide continuous online monitoring, with catalytic combustion and electrochemical sensors as the mainstream. Both types have been used in coal mines for many years and are mature and reliable, but they have several inherent weaknesses on high-speed railway tunnel sites:
First, the explosion-proof dilemma. Catalytic combustion sensors measure methane concentration via platinum-wire catalytic combustion — the sensing element is itself a miniature heat source — so in gas worksites they can only rely on flameproof enclosures and intrinsically safe circuits as a backstop, like dancing in shackles. Second, they are fragile. Catalytic elements are poisoned and fail when exposed to sulfur, silicon and similar substances, have short service lives, and typically require recalibration every 15–30 days on site — maintenance costs are not trivial; electrochemical sensors last even less long and drift noticeably at low temperatures. Third, electrical signal transmission. Inside the tunnel, drilling and blasting, heavy electromechanical equipment and variable-frequency devices are densely deployed, and long-distance cable power and signal runs are both vulnerable to electromagnetic interference and add points of failure.
How Fiber Optics Measures Methane: Three Technology Routes
The common feature of fiber-optic gas sensors is "light in, electricity out": the probe side is passive and unpowered, with concentration information relayed back to the host unit outside the tunnel via optical signal. There are three main routes in principle:
Route 1: Spectral absorption (closest to engineering maturity). Methane molecules have a characteristic absorption peak in the near infrared around 1650.9nm. When a laser passes through a gas cell containing methane, the degree of light absorption corresponds to the concentration. In engineering practice, TDLAS (Tunable Diode Laser Absorption Spectroscopy) is commonly used, combined with wavelength modulation and second-harmonic detection, achieving sensitivity from ppm to percent levels while avoiding cross-interference from dust and water vapor. Domestic practice already includes multi-channel fiber-optic methane detection systems developed for metro tunnels — separated probe and host, intrinsically safe, essentially maintenance-free, with a single fiber carrying multiple optical probes, which fits perfectly with the "long-distance, many-point" nature of tunnels.
Route 2: Fiber grating gas-sensitive film. Coat an FBG (Fiber Bragg Grating) or LPFG (Long Period Grating) surface with a methane-sensitive material; gas adsorption changes the refractive index around the grating, which manifests as a shift in the characteristic wavelength. Combined with weak grating + OTDR technology, multiple sensing points can be daisy-chained on a single fiber for networked measurement; reported methane detection sensitivity reaches below 0.05%, meeting the range requirements of construction monitoring.
Route 3: Laboratory frontier. Approaches such as gas-sensitive films coated inside photonic crystal fiber (PCF) holes, evanescent wave sensing, and surface plasmon resonance (SPR) offer higher sensitivity and better selectivity, but remain distant from the tunnel worksite — better tracked as a technology reserve for now.
Why Tunnel Sites Need It
Placing fiber-optic gas sensors in the context of high-speed railway tunnel construction, the value is structural:
Intrinsic safety is non-negotiable. The probes carry no power and generate no heat, and optical signal transmission is inherently spark-free — what gas worksites value most is "no electrical sparks." The fiber-optic approach removes ignition risk at the level of physical principle, dramatically lightening the design burden of electrical explosion protection.
The host unit can sit outside the tunnel. The interrogator is placed in the fresh airflow zone at the portal or in the duty room, leaving only fiber and probes inside the tunnel. Optical transmission loss is extremely low — links of several kilometers are routine — which is precisely the biggest headache for copper-cable-powered solutions.
One fiber, many points. The code-mandated sensor locations — 5–10m behind the working face, return airflow cross-sections, arch crown positions of the secondary lining trolley and waterproofing membrane trolley, auxiliary fan outlets, and dead corners under the arch where gas accumulates — can all be chained along a single fiber trunk line, greatly simplifying the system architecture.
Interference-immune and environmentally robust. Inside the tunnel there is high humidity, dust, blasting vibration and electromechanical interference; fiber-optic sensors are inherently immune to electromagnetic interference, corrosion-resistant, and offer long-term stability superior to electrochemical devices.
Landing It in the Field Is Not So Simple
The physics checks out, but several hurdles remain before field deployment:
Sensor placement matters. Methane is lighter than air and tends to accumulate at the arch crown; sensors should be mounted about 20–25cm below the crown, and the working face, return airflow and dead corners must all be covered — get the placement wrong, and even the most accurate sensor is wasted.
It must be interlocked with power cut-off. TB 10120—2019 sets explicit requirements for gas automatic monitoring, alarm and power-interlock systems: fault interlock, gas-electric interlock and fan-power interlock are hard requirements. Once the fiber sensor outputs a concentration signal, it must be integrated into the automatic monitoring system, forming a closed loop with power cut-off actuators and audible-visual alarms, acting immediately on threshold exceedance — it cannot merely "alarm" without "cutting."
It runs in parallel with manual inspection. However continuous the fiber-based automatic monitoring is, it cannot replace the gas inspector system. Automatic and manual data must cross-validate each other; when discrepancies exceed limits, investigate whether it is a sensor issue or a ventilation issue. The point of automatic monitoring is to fill the gaps in manual coverage, not to replace people.
Field data is still accumulating. Fiber-optic gas sensors have track records in coal mines and metro systems, but under the dust, humidity and vibration of drill-and-blast high-speed railway construction, probe gas-cell contamination, long-term recalibration intervals and system service life all still need more field data to validate. When selecting a system, don't just read the lab spec sheet — look for engineering cases under comparable working conditions.
From "Measuring It" to "Understanding It"
Tunnel monitoring is moving from single-point, single-parameter toward multi-parameter, one-network architectures. The imagination in fiber sensing lies here: on the same fiber, a gas cell measures concentration while FBGs measure temperature and strain — hooking methane concentration, lining stress, ambient temperature and even microseismic signals onto the same optical network is no longer a fantasy. Once the data is integrated into tunnel monitoring and digital twin platforms, construction managers will see not isolated concentration curves but a comprehensive risk view overlaying "gas trends + ventilation status + construction sequence nodes" — moving from "measuring it" to "understanding it." That is what informatization is supposed to deliver.
For those working in bridge and tunnel monitoring and informatization, fiber-optic gas sensors are a direction worth watching closely: only when the sensing layer's road is paved can the platform's value truly land.