As more people discover various posts over the past year and a half, I receive a number of great questions. Here is a recent one. The specific operating parameter details have been omitted, but I wanted to share the flavor of the question and the answer.
We have a customer that uses a turbine flowmeter for natural gas metering. Based on the furnace-cycle time demand, cubic gas volume demand, no-flow shutdown time, our supply pressure at the metering station, and piping distances between metering station and furnace, we´d like to calculate the Dynamic Response Error due to the shutdown time in this system.
Jorge Gomez is an application manager in Emerson’s Remote Automation Solutions business and is located in Brazil. He also provides support for Daniel flow products. Jorge worked many years in Brazil’s national flow lab and has quite a number of contacts with flow technicians in TÜV SÜD’s NEL, Germany’s Physikalisch-Technische Bundesanstalt (PTB) and the US National Institute of Standards and Technology (NIST).
Jorge provides the following guidance:
Measurement of gas flow with turbines in a cyclic flow rate as you are asking is always a big problem–the main reason is that the turbine meter has a natural inertia in the rotor that cause a overmetering when the flow rate stops (the rotor keeps turning a time after the flowrate stops.) Usually this overmetering is not totally compensated by the rotor inertia when it starts to move when the flowrate returns. In other words, a turbine meters tends to show a positive error in a cyclic flowrate.
The estimation of this error is not easy, because it depends on the dynamic response of the meter that is variable depending on the model, design of the blades, mass of the rotor, wear of bearings and even the flow profile and how the flowrate changes (suddenly, slowly, pulsating, etc.)
There is a good study presented in ISO TR 3313 standard (measurement of fluid flow in closed conduits-guidelines of the effects of flow pulsations on flow measurement instruments). Despite this standard’s focus on orifice plates, there are sections covering turbines (6.2) and vortex (6.3)–these are meters especially susceptible to unsteady flow.
This standard presents a theoretical approach, but the main question is estimates the dynamic response parameter, that is strictly empiric (obtained from experiments). This standard suggests this parameter for turbines from 2″ up to 6″ for gas and liquid flow, but the suggested parameters can be always questioned. You can also obtain this parameter from experiments on a calibration bench, although I don’t know if this is possible in your case.
The standard also presents a very comprehensive bibliography, and you can purchase and download it from the ISO site.
From a practical point of view, maybe the best solution, especially if this is a custody transfer measurement–as it seems to be–is thinking about use of a flow sensing technology less affected by unsteady flow, like ultrasonic, Coriolis or even differential pressure measured across orifice plates.