Abstract
NASA's experience with electrochemical sensors in a hand-held toxic gas monitor serves as a basis for the development of a fixed on-board instrument, the Contingency Gas Analyzer (CGA), for monitoring selected toxic combustion products as well as oxygen and carbon dioxide on the Orion Crew Exploration Vehicle (CEV). Oxygen and carbon dioxide are major components of the cabin environment and accurate measurement of these compounds is critical to maintaining a safe working environment for the crew. Fire or thermal degradation events may produce harmful levels of toxic products, including carbon monoxide (CO), hydrogen cyanide (HCN), and hydrogen chloride (HCl) in the environment. These three components, besides being toxic in their own right, can serve as surrogates for a panoply of hazardous combustion products. On orbit monitoring of these surrogates provides for crew health and safety by indicating the presence of toxic combustion products in the environment before, during and after combustion or thermal degradation events. Issues identified in previous NASA experiences mandate hardening the instrument and components to endure the mechanical and operational stresses of the CEV environment while maintaining high analytical fidelity. Specific functional challenges involve protecting the sensors from various anticipated events—such as rapid pressure changes, low cabin pressures, and extreme vibration/shock exposures—and extending the sensor lifetime and calibration periods far beyond the current state of the art to avoid the need for on-orbit calibration. This paper focuses on lessons learned from the earlier NASA hardware, current testing results, and engineering solutions to the identified problems. Of particular focus will be the means for protecting the sensors, addressing well known cross-sensitivity issues and the efficacy of a novel self monitoring mechanism for extending sensor calibration periods.
Introduction
In the confined environment of a spacecraft like Orion, air quality is a matter of life and death. Beyond monitoring $O_2$ and $CO_2$ for life support, the crew needs to detect "surrogate" gases—specifically Carbon Monoxide ($CO$), Hydrogen Cyanide ($HCN$), and Hydrogen Chloride ($HCl$)—that indicate fire or thermal degradation of electronics.
While electrochemical (EC) sensors are ideal for this due to their low power and weight, standard commercial-off-the-shelf (COTS) versions are not "space-ready." They face three major hurdles:
- Pressure Extremes: Rapid depressurization or low cabin pressure can cause electrolyte leakage or desiccation (drying out).
- Calibration Drift: COTS sensors typically require calibration every 1–3 months, which is impossible during a long-duration mission.
- Cross-Sensitivity: Metabolic hydrogen ($H_2$) produced by the crew can "trick" $CO$ sensors, leading to false fire alarms.
Methods
The researchers at Hamilton Sundstrand and NASA implemented a multi-layered engineering approach to address these challenges:
- Hardware Isolation: Instead of "passive" sensing, the CGA uses a small pump to draw air into a flow cell. This cell is protected by solenoid valves that snap shut if cabin pressure drops too low or changes too fast, shielding the sensors from physical damage.
- Enhanced Screening: To make sensor behavior predictable, the team implemented "enhanced acceptance criteria." They evaluate physical traits like electrode spacing and surface area to ensure the sensors age in a mathematically consistent way.
- Algorithmic Compensation: The team developed a "built-in state-of-health" test. By subjecting a sub-sample of sensors to accelerated aging (high heat and low humidity), they created a drift model that the CGA software uses to automatically adjust its own calibration.
- Dual-Sensor $CO$ Correction: To solve the $H_2$ interference problem, they paired a high-selectivity $CO$ sensor with a second sensor sensitive to both $CO$ and $H_2$. By solving a simple linear equation between the two, the system can "subtract" the hydrogen signal.
Key Results and Discussion
Drift Reduction
Using the primitive drift compensation model, the error in sensor sensitivity was significantly reduced. In accelerated aging tests, uncompensated sensors showed a drift of approximately 25-30% over 60 days. With the compensation algorithm, the error was maintained at less than 3%.
Interference Mitigation
The dual-sensor approach proved highly effective at mitigating cross-sensitivity. In tests with 35 ppmv $CO$ and 150 ppmv of interfering $H_2$, the uncompensated signal rose to nearly 45 ppmv, while the $H_2$-compensated signal accurately reflected the 35 ppmv $CO$ concentration with less than 1 ppmv error.
Ruggedness and Desiccation
$O_2$ sensors showed no adverse effects after more than 300 cumulative hours at pressures below 2 psia, thanks to the zero-headspace cell design. A key finding was that $HCN$ sensors failed at low pressure due to desiccation (evaporation of electrolyte) rather than leakage. By screening sensors strictly, the team could treat an entire batch as a single "population" with shared aging characteristics, making software-based compensation viable.
Conclusion and Future Work
The CGA represents a significant advancement in space-based gas monitoring. While the system successfully addresses pressure and drift issues, the primary remaining limitation is the response time for certain gases. $HCN$ sensors can take 5–6 minutes to reach 90% of their steady-state value ($t_{90}$), which is acceptable for post-fire cleanup but less ideal for immediate detection.
Future work is focused on:
- Validating the drift models with larger sample sizes and longer-term (multi-year) studies.
- Refining the software triggers for the protective solenoid valves.
- Investigating why $HCN$ response times become "sluggish" as the sensors age.
Summary updated