How Accurate is My MFC for Multiple Gases?

Manufacturers use different methods to calculate flow rates for multiple gas MFCs. The accuracy of these MFCs can vary greatly dependent upon the MFC or equipment manufacturer and the variety of gases used. Understanding these methods can help you, the user, identify problems in the process caused by inaccurate flow measurements. We suggest the following cautions to anyone using MFCs for multiple gases:

Know the Sensor Factor

All MFC manufacturers will provide a sensor factor to users. Commonly, this is a ratio of the specific heat of the process gas to the calibration gas; however, these factors are not universal across MFC manufacturers for each individual gas. Different bypass sensor designs will utilize different flow rates. Heating elements also operate at different temperatures, and a gas’s specific heat will change with the temperature. You should only use the sensor factors provided by the MFC manufacturer.

Since it is a ratio, the sensor factor is a linear conversion that provides a fixed offset from the calibration gas curve throughout the flow range. Nitrogen is typically the calibration gas because it is readily available, very safe to use, and the flow response through the sensor’s flow range is pretty linear. Other gases do not respond in the same way when flowing through the sensor. Ask your MFC manufacturer to tell you what gas was used for calibration. Specific heat differences change the measurement resolution of the sensor. Difference in viscosity and density change flow rate through the sensor bypass.

This graph shows the typical behavior of certain gases through the MFC sensor. This will vary depending on the design of each sensor. Gases like nitrogen and argon can utilize most of the sensor’s available range, whereas gases like methane cannot. High specific heat pulls extra heat away from the sensor’s heating element, which saturates the sensor and limits the max measurable sensor flow.

When a ratio of specific heat is used to change gases, the user effectively sees the following:

The lower dashed line represents applying a conversion factor for argon to the nitrogen calibration. The upper dashed line represents applying a conversion factor for methane to the nitrogen calibration. MFC will control for non-calibration gas assuming flow follows the dashed line. At some points in the flow range, the dashed line will be closer to the “true” gas flow than at others. This will cause variable flow accuracy across the flow range of the MFC. This is troublesome to identify, as accuracy varies based on both gases and flow rates.

Ask About Different Methods with Multiple Gases

Some MFC manufacturers also use different methods to support multiple gases in a single MFC. Some have tested and developed physics-based models in an effort to provide a more accurate measurement across gases. Some condition the flow entering the MFC in an effort to reduce the difference in flow response between gases as they pass through the sensor.

Ask the manufacturer to describe the method conversion used for alternate gases and to define the accuracy specification for each alternate gas. This should be a different specification than the accuracy with calibration gas. Claims that “gas expansions are linear” means that a linear conversion factor is used.

Note that flow accuracy will vary between different gases, and may vary at different flow rates even when using the same gas. The importance of this will depend on the nature of the process that the MFC is used in and how critical the flow measurement provided by the MFC is. Purchasing a process-gas calibrated MFC for each individual gas is an option when accuracy improvements are needed.

If the varying accuracy poses a problem for the user, these MFCs can be flow checked in many of these alternate gases. For safety reasons, some calibration facilities may limit access to explosive or toxic gases. The flow standard used for the flow test should be immune from the effects of property differences between different gases. Thermal flow standards may face similar issues. If an MFC is used as a flow source, this may add additional uncertainty to the total calculated uncertainty of the flow bench. A DP-based flow transfer standard, like a sonic nozzle or laminar flow element, may contribute too much uncertainty to the flow bench’s total calculated uncertainty. This is particularly true if the DP transfer standard was calibrated in a gas other than the process gas.

The Met Lab series from Mesa Labs uses a positive displacement measurement technique that is immune to these differences in gas properties. Met Lab standards also have low minimum pressure requirements, and offer integral purge connections when being used with hazardous and combustible gases.