While NASA facilities already use huge volumes of hydrogen as a propellant/fuel, many other federal and state programs across the country are looking at potentially expanding the use of hydrogen. There are, however, significant challenges associated with hydrogen use. These include a tendency to leak through seals (due to the very small size of the hydrogen atom) that ordinarily would efficiently stop most other materials, a very high diffusion rate, a huge explosive mixture range, and the fact that hydrogen burns with an invisible flame. Therefore, hydrogen leak detection is an important capability associated with a safe and operational work environment for NASA facilities, as well as at any other location/site that would potentially use this fuel. At the time of this reporting, there is not a technology that provides simple, inexpensive, and wide-coverage methods that enable large quantities of hydrogen to be monitored. Sensors that are available tend to be limited to either a very short range, or have no directionality. For instance, a standard hydrogen sensor detects an increase in hydrogen, but cannot determine if that increase is from a small leak nearby or a larger leak some distance upwind. A technology that is capable of alerting and providing emergency detection information about hydrogen leakage would be beneficial and increase overall safety.

Therefore, the Intelligent Flame-finder Detection and Alert System (IFDAS) was developed for Stennis Space Center (SSC) to provide a method to detect and localize both leaks and flames of hydrogen and other gases that may be otherwise difficult to detect. The sensing and localization capability of this system can be performed from a considerable distance and over a significant field of regard, thereby reducing the necessary number of sensors required to safely monitor some given volume. Compared to other flame detectors that use infrared/heat detection or chemical means to sense leaks, the IFDAS uses a unique combination of smart imaging and acoustic methods that provide additional benefits including reduced false alerts and increased accurate detection. Additionally, IFDAS uses comparatively inexpensive components to provide an overall affordable total system that achieves accurate invisible gas leak detection results.

Other unique features of the IFDAS include the selection of a very specific combination of cameras and filters that can provide low-cost near-infrared (NIR) coverage of the target area at high enough resolution to permit wide-area coverage, as well as a specific combination of microphones and amplifying designs based on actual leak acoustic signature characteristics. The optical system comprises a camera or cameras fitted with particularly selected NIR filters that pass light at a band characteristic of hydrogen flames. It is then wired or wirelessly conveyed as an image from the camera or cameras to the computational component. The acoustic component comprises a microphone or microphones with amplifying/focusing components (e.g., parabolic dish, etc.) to detect sounds characteristic of leaks, and are wired or wirelessly conveyed to the acoustic data from the microphone or microphones to the computational component. The computational component is supplied with software that can analyze the imagery from the camera or cameras to detect flames according to various characteristics (excluding distractors such as reflections of the flame, sunlight, etc.) and localize the flames in the image. The software can then analyze the acoustic data and detect the acoustic characteristics of a leak, which localizes the sound within the area of regard, and provides the capability to trigger alerts based on any leak detections. Other features of the software include: (1) the algorithms and heuristics to recognize flames within the field of view while excluding all potential distractors/sources of false positives; (2) the ability to recognize leaks and/or flames via acoustic signatures; and (3) the capability to determine the physical location for the detected phenomena.

Any facilities that utilize large quantities of hydrogen could increase their safe use and worker safety by implementing a smart sensor monitoring system such as the IFDAS. In summary, IFDAS consists of a “smart” sensor system, which can include both an optical and/or acoustic component, plus a computational component with specialized software to perform the “smart” data analysis to improve hydrogen monitoring capability and safe use.

This work was done by Zack Mian, Ronald Gamache, and Nick Glasser of International Electronic Machines for Stennis Space Center. For more information, contact IEM at (518) 449-5504. Refer to SSC-00385.