Infrared-Based Combustible Gas Detectors:
How They Work and Why They're Better

Introduction

Optical detectors for combustible gas detection provide an increasingly popular alternative to traditional pellistor-based combustible gas detectors. One reason for their popularity is the ease of implementation in Safety Instrumented Systems (SIS) intended for high availability and Safety Integrity Levels (SIL).

This article describes the fundamental operation of optical gas detectors with the purpose of explaining the issues important to high availability and SIL.

What is Availability?

For the purpose of discussion, in an extremely simplified context, the availability of a gas detection system can be described as:

In this equation, MTBF is the overall mean time between failures of the detection system. MTTR is the overall mean time to both diagnose and repair a failure in the system.

For a gas detection system of any practical size, the MTBF will be on the order of several thousands of hours. As an example, consider a system with an MTBF of 10,000 hours. To obtain minimal SIL-3 performance (>99.9% availability), as shown below, the overall MTTR must be less than 10 hours:

The time involved in the "repair" component of MTTR is largely independent of sensor architecture. This is dependent upon having spares readily available on-site and having personnel available to perform the repair. The "diagnose" component of MTTR is most easily shortened by using sensors that actively auto-detect and indicate failures so that prompt repair is initiated.

Why Infrared Gas Detectors?

The historical method of detecting combustible gases involved sensors based on pellistor beads. Pellistor based sensors are susceptible to significant covert (undetected) sources of failure which extend the MTTR. Primary of these are catalyst poisoning and flame arrestor plugging, either of which prevent the sensing of gas. In the above example, it is easy to see how a covert pellistor failure could remain undetected for much longer than 10 hours as the only way to detect a failure is to periodically check with gas.

The absence of covert failure modes is the intrinsic characteristic of infrared gas detectors that provides a significant advantage over catalytic beads. All failure modes are overt and instantly signalled so that repair can be initiated immediately.

All infrared gas detectors use variations on the basic measurement scheme outlined in the illustration. In this scheme, there is an infrared source that illuminates a volume of gas that has diffused into the measurement chamber. The gas absorbs certain of the infrared wavelengths as the light passes through it whilst others pass through completely unattenuated. The amount of absorption is related to the concentration of the gas.

The amount of absorption is measured by a set of optical detectors and subsequent electronics. The change in intensity of the absorbed light is measured relative to the intensity of light at a non-absorbed wavelength. The microprocessor computes and reports the gas concentration from the absorption.

The signals from the IR detectors are interpreted as shown in the illustration. Panel A and B illustrates the normal state of operation of the detector. When there is no gas present the signals for the reference and active channel sensors are balanced. When there is combustible gas present, there is a predictable drop in the output from active channel sensor because the gas is absorbing light.

A fault condition is illustrated in panel C for the signal levels encountered in the case of dirty optics or a weak and failing light source. The former scenario is a trigger to perform routine maintenance and the latter is an indicator that preventive maintenance should be scheduled. In either case, the instrument continues to faithfully measure gas concentration up until the situation degrades to an untenably low signal level. These maintenance situations are easily flagged with modern digital field communication and any number of asset management and predictive maintenance programs.

Panel D and E illustrate the situation with failed components in the measurement loop. If a sensor fails or light source fails, signals on one or both sensor channels will fall to zero.

Summary

The achievement of high availability in an instrumented safety system is critically dependent upon the mean time to repair (MTTR) in the event of a failure of system components. Optical IR detectors for combustible gases have intrinsically low MTTRs because there are no covert sources of failure thus minimising the amount of time between an instrument failure and the time that it is detected.

By: Guest Author, Dr. John  Jarvis
Dr. John Jarvis started in the chemical industry where he first developed his career long interest in practical, real-time, on-line chemical measurements. His 20-year career has spanned the chemical industry, academia, and the instrument industry. He is currently employed as a principal scientist at Detector Electronics Corp. in Minneapolis, MN, USA and is responsible for the design of optical gas detection products.