This Note describes the characteristics of sensors measuring Lower Explosive Limit (LEL) percentages of combustible gases and vapors. It compares pellistor-type and NDIR LEL sensors and describes handling of sensor poisons. Lists of Correction Factors (CFs) that are used in WatchGas instruments are included, along with CFs used in EMEA countries.
Pellistor-Type LEL Sensors
• Are low cost
• Measure all volatile combustible gases including Hydrogen
• Require Oxygen
• Are subject to catalyst poisons
• Are not good for heavy chemicals like diesel
Pellistor-type LEL sensors work by oxidizing combustible vapors that diffuse into the sensor using a heated catalytic bead. The additional heat released by oxidation changes the resistance in a Wheatstone bridge circuit, which is correlated to the gas concentration. Therefore, at least 10% by volume oxygen is required in the sampled gas for pellistor-type LEL sensors to sustain catalytic oxidation, and the calibration gas should have an air matrix instead of nitrogen. Advantages of Pellistor LEL sensors is that they are inexpensive and measure all volatile combustible gases present including hydrogen. A disadvantage is that they require the sample gas to diffuse through a flame arresting frit and therefore they respond poorly to heavier compounds such as diesel fuel. Another important disadvantage is that the pellistor catalyst is subject to poisoning by certain compounds including silicones, Hydrogen Sulfide, acids, and Chlorinated compounds.
NDIR-Type LEL Sensors
• Cost more but last longer
• Do not measure Hydrogen
• Do not require Oxygen
• Are not subject to poisons but can be affected by physical deposits
• Are OK for heavy compounds like diesel
• Have extended range versions up to 100% Methane
Non-Dispersive Infrared (NDIR) sensors work by measuring the amount of light absorbed by the C-H bonds on hydrocarbons, such as Methane and Propane. Flammable compounds that do not contain any C-H bonds, such as Hydrogen (H2) and Carbon Disulphide (CS2) cannot be measured. Therefore, NDIR sensors may underestimate the explosion hazard of an atmosphere and should not be used in any situation that has a possibility of containing H2 or CS2 gas at combustible levels. However, a major advantage is that NDIR sensors are not subject to the poisoning that affects pellistor-type LEL sensors. In addition, they do not require Oxygen to function and can be used to measure hydrocarbons in inert atmospheres such as Nitrogen or Argon, and calibration gas for them can be supplied in pure Nitrogen. By the same token, NDIR-LEL sensors may overestimate explosion hazard when not enough Oxygen is present to support runaway combustion. Lastly, these sensors do not require a flame-arresting frit and thus respond faster to heavier hydrocarbons than do pellistor-type LEL sensors.
Exception: The hydrocarbon “Acetylene“ cannot be detected through NDIR-type sensor, due to strong tripe bond “C2H2“ (Ethyn/Acetylene).
European vs US LEL Values
In 2010, Europe and other EMEA countries (Europe, Middle East and Africa) who require ATEX certification switched to using IEC (IEC60079) and EU (EN61779) standards that define LEL concentrations differently than the ISO10156 procedure used in the US. The US procedure measures LEL values in a static chamber, whereas the IEC/EU procedure uses a stirred gas chamber that often results in slightly lower LEL values. The most important result is that the US still defines 100% LEL as 5.0 Vol% methane, while EMEA countries now use 4.4 Vol%. New CFs based on the IEC/EU LEL values are given in the table below.
Applying Correction Factors
The best accuracy is always obtained by using the same calibration gas as the gas to be measured. However, if methane is used to calibrate, the user can select any measurement gas from the list and multiply the reading by the CF to get the true LEL value:
True %LEL = %LEL Reading x CF
The table lists the LEL CF values that are programmed into POLI instrument firmware, based on US LEL values for pellistor sensors. If the user calls up one of these LEL measurement compounds, the factor will automatically be applied so the instrument reads correctly in units of the measured gas. If the CF for EU/EMEA or NDIR sensor is needed, it must be programmed in as a custom factor using WatchGas Suite software (see Section 6 of the POLI User Manual).
Handling Sensor Poisons
Poisons and Inhibitors: As noted above, pellistor sensors can lose response when exposed to sensor poisons that either permanently deactivate the catalyst or temporarily reduce its response. Silicone compounds, such as those present in many lubricants and cleaners are the most common poisons and should be kept away from the sensor. Acids such as Muriatic acid (HCl), Sulfuric acid, Nitric acid and Acetic acid can corrode the sensor. Some VOCs such as Chlorinated, Nitrogen-containing and Sulfur-containing compounds can form acids during combustion on the sensor and thus reduce response. Hydrogen Sulfide and Mercaptans also inhibitor sensor response, but sometimes just temporarily. Keep exposure to potential poisons as short and as low a concentration as possible. Response Recovery: In some cases, response can be recovered simply by turning on the pellistor sensor and letting it run in clean air for several hours or days. For faster recovery, place the unit into diagnostic mode (using WatchGas Suite) and expose the LEL sensor to 100 %vol Methane for a few seconds or 20 %vol Methane for 5-10 seconds. This procedure raises the sensor temperature enough to burn off the surface and recover sensitivity. Longer exposure times do not help much because Oxygen is needed for the process, but repeating the process once or twice may help.
Propane or Pentane Calibration
Sensor poisons are especially important to avoid when measuring non-Methane compounds and using correction factors. This is because poisons often de-activate response to Methane more than to other VOCs like Propane and Pentane, which means that the correction factor changes. Fortunately, if calibrating with methane, the other compounds will over-respond and the error will be on the safe side. To prevent false alarms one can calibrate with Pentane or Propane (or whatever VOC is being measured). In this case the factors in the table below need to be adjusted by dividing by the CF of the calibration gas. For example, if propane (CF = 1.6) is used to calibrate for measurements of gasoline (CF = 2.1), the new CF is calculated as 2.1/1.6 = 1.3. The WatchGas POLI multigas detector does all these calculations automatically so the user can simply select the calibration gas as propane and measurement gas as Gasoline, and the concentrations will display correctly.
Use Propane or Pentane calibration only in areas known NOT to contain Methane. Otherwise, as sensor response degrades, Methane could give false low readings resulting in an undetected explosion hazard.