Subscribe to receive more articles like this: Print/digital

Choice of MFCs for test stands must ensure accuracy at different flow rates and consider turndown and calibration

Fuel cell engineers use test stands to simulate the performance of fuel cells. The challenge is to mimic reality and determine the optimum fuel mixture and temperature, while also examining failure mechanisms. Another objective is to test the performance of anode/ cathode materials, catalysts, and membranes. Fuel cell test stands consist of systems for mixing, delivering, and humidifying gases, all of which depend on reliable flow measurement and control. A typical test stand uses several thermal mass flow controllers (TMFCs) of various flow rates to deliver the precise amount of blended gases to the fuel cell. Since the test stand must imitate normal operating conditions, these flow controllers must react to process signals quickly—much like a car accelerator—to provide a true test of the desired performance. In addition, the TMFCs must have a broad turndown to mimic low and high fuel consumption rates. Many fuel cell test stands use TMFCs that do not have the performance capabilities to adequately test fuel cells. (See Figure 1)

For research, development or manufacturing of fuel cells, engineers need test tools to evaluate and validate fuel cell technology. Test systems must provide fl exible data acquisition, monitoring and control to precisely handle the fl uid for fuel cell operation and experiments.

As research engineers constantly incorporate new measurements into their tests they need reliable, accurate and flexible test systems to help shorten development cycles, increase quality, and lower the cost for all stages of fuel cell development, from research and design validation to manufacturing.

One of the most important and complex tasks is the control of gases to the fuel cell and the selection of the best measurement and control equipment. Gas measurement and control can be done with a variety of technologies ranging from a simple variable area meter (VA), commonly know as a rotameter, to a thermal mass flow controller. A thermal mass flow controller is the preferred choice by most fuel cell developers due to its response time, turndown capability, and accuracy.

Variable Area Meters
(Rotameters)
Measurement and control using a variable area (VA) meter is accomplished with a glass tube and a float (ball) and a needle valve as shown in Figure 2. The measurement itself is volumetric and readings can be fairly accurate as long as conditions are close to normal pressure and temperature; however, wide variations in temperature or pressure will affect the measurement, often significantly. Nominal accuracy is dependent on the device selected, with a range of 2-10% of full scale. To approximate mass using a VA meter, one must assume that the temperature is 68° F and pressure is sea level, and stay constant throughout the measurement range. In the early days of fuel cell testing, the VA meter was a commonly used device; it was simple and inexpensive; however too many times when results were compared with those using a mass balance, it was apparent that the VA meter’s accuracy was unacceptably inconsistent due to wide fluctuations in temperature and pressure.

Mass Flow Controllers
The MFC is the preferred choice for controlling gas flows in today’s test stations. Two main measurement technologies are used for doing the measurement, thermal, and pressure/ temperature. Both technologies measure inferred mass. The mass flow measurement using pressure and temperature is based on the ideal gas law. The drawback with this technology is that it assumes that everything stays constant. Unfortunately, this is not the case. Accuracy is in the range of 2% of full scale, with a repeatability of 0.5% of full scale, still not a perfect choice for an accurate test and mass balance.

The most common measurement and control instrument is the TMFC. (See Figure 3) It is available in a variety of configurations and with a very wide performance range. In this case the common adage that you get what pay for is very true. While a TMFC is a more expensive technology than the VA or pressure/temperature mass flow types, the performance benefits significantly outweigh the costs.. The measurement is based on the heat characteristics of a specific gas. A small part of the flow through the TMFC is diverted into a sensor tube where a temperature measurement is made. From this sample, mass flow is automatically calculated.

It is very important to understand the accuracy and turndown limitations. (See Figure 4) A TMFC with an accuracy of 1% of full scale controlling at 40% of calibrated flow rate will have an accuracy of ± 2.5%, but at 20% of calibrated flow rate its accuracy changes to ± 5%. These accuracy ranges make the use of a mass balance almost impossible. Thermal mass flow controllers with an accuracy of 1% of rate stay as a fl at line over the entire calibrated range.

How an TMFC is calibrated is also very important. The thermal characteristics of some gases can be very complex. Mainly with H₂ and CO₂, the characteristics become three dimensional, such that the gas factor changes with flow rate, temperature and pressure. Many MFC manufacturers use surrogate gases for calibration and then apply a factor to correct for the actual gas. When choosing an MFC for use on H₂ or CO₂ it is very important to select a manufacturer who calibrates on the actual gas, that is, a CO₂ MFC should be calibrated on CO₂.

Digital vs. Analog MFCs
The internal operation of an MFC has changed greatly over the last 10 years, with advances of microprocessor-based MFCs. The zero and span adjustments have been replaced with computer interfaces. Specific gas thermal characteristics are defined as an ‘S’ curve; the linearization of the ‘S’ curve is directly proportional to the turndown and accuracy of the MFC. Older analog MFCs have a very limited circuitry for adjusting or “linearizing” the ‘S’ curve, typically incorporating only two or three points. CO2 has a very complex ‘S’ curve. A high performance microprocessor-based MFC can use up to 25 points for linearization of the ‘S’ curve, plus it will also use 4th order polynomials in the calibration. (See Figure 5)

A digital MFC will also allow for multiple calibration curves to be stored in the MFC. These are real calibrations as opposed to gas factors, which are not substitutes for real calibration curves. In reading most manufacturers’ instruction manuals, the conclusion is inescapable that applying surrogate gas calibrations or using gas factors will result in ± 5% full scale accuracy, typically inadequate for fuel cell testing and production applications.

In addition to using an MFC calibrated to the actual gas, microprocessor- based electronics will also greatly improve the valve control of the MFC. This is very important for fast response and repeatability.

Conclusion
When selecting digital mass flow controllers in gas blending/ delivery systems, look for response times of less than 1 second. While turndown (when measured as control range staying within 2% of rate accuracy) can range between 2:1 and 12:1 for many devices, 30:1 is achievable.

JAN CHRISTENSEN, SENIOR FLOW SPECIALIST, BROOKS INSTRUMENT. CHRISTENSEN IS A GRADUATE ELECTRONICS TECHNOLOGIST WITH MORE THAN 24 YEARS OF FLOW EXPERIENCE AND 12 OF THOSE YEARS AS A FIELD SERVICE APPLICATION SPECIALIST. HE HAS WORKED WITH MOST OF THE WORLD’S FUEL CELL COMPANIES OVER THE LAST FEW DECADES, ASSISTING THEM IN DEVELOPING GAS DELIVERY SYSTEMS.

To read the entire article, sign up for a free subscription and then click on
Click here for the digital edition of the current issue