Today, most Battery Energy Storage System (BESS) installations and data centre facilities are equipped with UL 2075 certified or third-party certified gas detectors as part of their thermal runaway monitoring strategy.
UL 2075 certification is widely regarded as a baseline indicator of product performance and compliance and it is often relied upon by facility owners, insurers and authorities having jurisdiction (AHJs) as evidence of adequate risk mitigation.
However, an increasing number of field reports and incident investigations indicate a growing gap between certified gas detectors and real-world BESS operating conditions.
NFPA 855-2026 edition introduces the concept of off-gas detection after the industry found that catalytic sensors don’t perform well in BESS, while the fact remains that many off-gas detectors on the market rely on metal-oxide (MOX) VOC sensors.
MOX sensors have inherent technical limitations and may not be well suited for BESS installations.
MOX sensors are widely utilised in the small household appliance industry rather than the life safety sector.
Leading BESS manufacturers found that off-gas detectors are easily poisoned by silicone in the deployed BESS environment and failed to survive even though those detectors and sensors are third-party certified.
In practical installations, BESS containers and energy storage rooms supporting data centres present conditions that differ significantly from standard certification test environments:
• Heavy silicone outgassing from sealants, cables, potting materials and structural adhesives inside cabinets.
• Sustained high ambient temperatures, often exceeding 70°C – 80°C for weeks or months after installation, and prior to commissioning.
• Limited ventilation during early project phases.
• Long idle periods before the system is energised and monitored continuously.
Under these conditions, many commonly deployed gas detectors, despite being UL 2075 certified, have been reported to experience:
• Silicone poisoning of gas sensors, leading to permanent sensitivity loss.
• Thermal degradation of sensing elements.
• Silent failure modes, where the detector remains powered but no longer responds accurately to target gases.
These failures are not random. They’re fundamentally linked to the underlying gas sensing technology used in the detector.
This situation raises an important question for BESS and data centre stakeholders.
How can facility owners, insurance companies and AHJs justify that installed gas detectors are functioning as intended under real operating conditions?
UL 2075 certification confirms that a product meets defined requirements under specified test conditions. It does not automatically guarantee performance in:
• Silicone-saturated environments.
• Extended exposure to temperatures approaching or exceeding the sensor’s design limits.
• Pre-operational phases where monitoring is assumed but rarely verified.
From a risk management perspective, relying solely on a certification label, without validating technology suitability and application robustness, introduces uncertainty into safety assumptions.
Different gas detectors may carry the same certification but rely on very different sensing principles, each with distinct vulnerabilities:
• Some sensing technologies (refer to table 1) are inherently susceptible to silicone contamination, which irreversibly blocks active sensing sites.
• Others experience accelerated drift or failure at elevated temperatures, even if short-duration tests are passed.
• Few technologies are designed for long-term exposure to both high-temperature and chemical contaminants simultaneously.
In figure 1, C：Offset in clean air at room temperature. An offset drift greater than 100 ppm was observed over 30 days. The sensors in table 2 are sealed together with silicone (10g in 4L chamber) for 7 days.
In table 3, C0: Sensor offset in clean air at room temperature. C1: Sensor response to 200 ppm H₂ at room temperature.
After silicone exposure, the response to 200 ppm H₂ decreased to approximately 50% of the initial value, indicating a significant impact on detection accuracy.
In figure 2, “C” denotes the TC sensor offset.
The offset is significantly affected by temperature (t) and relative humidity (RH), with variations ranging from +800 ppm to –600 ppm. In figure 3, “C” denotes the TC sensor offset.
In this evaluation, humidity was adjusted using N₂ to further assess the effects of temperature (t) and relative humidity (RH) on sensor offset.
Offset variations ranging from –500 ppm to –2300 ppm were observed.
Figures 4 and 5 illustrate the sensitivity drift of the H₂ sensor, Model FC-H2-20000, under storage at 85°C and 85°C/85% RH, respectively. S：Sensitivity during the test. S₀：Initial sensitivity prior to exposure.
S/S₀ represents the normalised sensitivity ratio. During the evaluation, samples were stored under the specified environment and the sensitivity at room temperature was periodically checked.
The red curve represents a reference sensor continuously stored at room temperature.
The data demonstrates stable sensitivity performance under both storage conditions.
In table 3, I0: Offset current before test; I1: Offset current after test. S0: Sensitivity before test; S1: Sensitivity after test.
During this test sensors 1-5 are sealed with 0.9g of 9661 silicones in a 0.35L test chamber for a period of 42 days while the reference sensor is always stored in clean air.
From this test we can see that the sensor is immune to silicone.
Figures 6 and 7 show the sensitivity drift of the FC-CO-5000 CO sensor when it is stored at 85°C and 85°C and 85%RH. In the above two figures, S: The current sensitivity S0: The sensitivity before the test.
S/S0 refers to the drift ratio of the sensor. During this test the sensor is stored in the corresponding environment and is taken out to check its sensitivity at room temperature periodically.
The red line in the above figures 6 and 7 is the reference sensor stored at room temperature.
These results indicate that the sensor remains stable under the tested conditions.
In table 4, I0: Offset current before test; I1: Offset current after test. S0: Sensitivity before test; S1: Sensitivity after test.
During this test sensors 1-5 are sealed with 0.9g of 9661 silicones in a 0.35L test chamber for a period of 42 days while the reference sensor is always stored in clean air.
The test shows the sensor’s immunity to silicone.
In BESS and data centre applications, where early detection of thermal runaway gases is critical, these limitations directly translate into latent safety risk (refer to table 5 Performance of UL 2075 Certified and Third-party Certified Gas detectors used in BESS).
For owners, insurers, and AHJs, the discussion should move beyond “Is it certified?” to include:
• Has the gas detector technology been validated for silicone-rich environments?
• Can the gas detector and gas sensor maintain performance during extended exposure to 80°C or higher?
• Is there evidence of long-term stability, not just short-term compliance testing?
• Are there mechanisms to verify the performance of gas detectors before system commissioning?
Only by addressing these questions can stakeholders confidently assess whether installed gas detection systems provide meaningful protection, rather than symbolic compliance.
Recognising these challenges, ProSense BESS gas sensing solutions have been developed specifically to address the gap between certification requirements and real-world BESS operating conditions.
Our advanced CO sensor Model FC-CO-5000, FC-CO-5000L and H2 sensor Model FC-H2-20000 and FC-H2-20000L are certified by UL Solutions.
The performance of the ProSense gas sensor is far beyond UL 2075 certification requirements.
By providing early detection, our solutions significantly enhance the safety resilience of BESS and data centre infrastructure and mitigate the risk.
The ProSense approach is based on gas sensing technologies with inherent resistance to silicone poisoning, high-temperature endurance (-40°C ~105°C) suitable for prolonged exposure in BESS containers and energy storage rooms and design validation aligned with actual BESS installation and pre-operation environments, not just laboratory test profiles.
Rather than following traditional standard testing conditions, ProSense solutions are working closely with standard organisations and stakeholders to complement certification frameworks by ensuring that gas detectors continue to function reliably in the environments where they are expected to protect life, property and critical infrastructure.
As data centre and large-scale energy storage systems continue to expand, the industry must recognise that certification alone is not the endpoint of safety assurance.
It is the starting point.
For stakeholders responsible for risk acceptance – facility owners, insurers and AHJs – the focus must shift toward the right gas sensor technology selection, environmental robustness and verifiable and demonstrated gas detection performance throughout the full project lifecycle.
