Calibration of hydrogen Coriolis flow meters using nitrogen and air and investigation of the influence of temperature on measurement accuracy

The performance of four Coriolis flow meters designed for use in hydrogen refuelling stations was evaluated with air and nitrogen by three members of the MetroHyVe JRP consortium; NEL, METAS and CESAME EXADEBIT. A wide range of conditions were tested overall, with gas flow rates ranging from (0.05-2) kg/min and pressures ranging from (20-86) bar. The majority of tests were conducted at nominal pressures of either 20 bar or 40 bar, in order to match the density of hydrogen at 350 bar and 20 degrees C or 700 bar and -40 degrees C. For the conditions tested, pressure did not have a noticeable influence on meter performance. When the flow meters were operated at ambient temperatures and within the manufacturer's recommended flow rate ranges, errors were generally within +/- 1%. Errors within +/- 0.5% were achievable for the medium to high flow rates. The influence of temperature on meter performance was also studied, with testing under both stable and transient conditions and temperatures as low as -40 degrees C. When the tested flow meters were allowed sufficient time to reach thermal equilibrium with the incoming gas, temperature effects were limited. The magnitude and spread of errors increased, but errors within +/- 2% were achievable at moderate to high flow rates. Conversely, errors as high as 15% were observed in tests where logging began before temperatures stabilised and there was a large difference in temperature between the flow meter and the incoming gas. One of the flow meters tested with nitrogen was later installed in a hydrogen refuelling station and tested against the METAS Hydrogen Field Test Standard (HFTS). Under these conditions, errors ranged from 0.47% to 0.91%. Testing with nitrogen at the same flow rates yielded errors of -0.61% to -0.82%.

Automated summary

The performance of four Coriolis flow meters designed for use in hydrogen refuelling stations was evaluated under a wide range of conditions. Results showed that pressure did not significantly affect meter performance, but temperature had some impact. Errors were generally within +/- 1% when operated within recommended flow rate ranges and ambient temperatures, with errors within +/- 2% achievable when temperature effects were controlled.