Measuring the flow through a pipe of dry gas with no liquid (or pure liquids with no gas) is relatively easy, but just a small amount of liquid in a gas makes the standard means of measuring flow rates significantly harder.
What causes the metering of wet gas to be trickier than for a single-phase fluid?
With a pure gas, if the properties remain roughly constant for a given spot in the pipe, the gas can be generalised to have a single flow rate and density (allowing perhaps for the flow profile across the pipe, with the flow at the centre of the pipe, typically moving faster than towards the edge). Simple equations can then relate the mass and volume flow rates to the velocity of the gas through the meter, and different metering technologies can be used to measure the flow rate with accuracy.
*Original source for photo unidentified
As soon as liquid is added, the situation is not as simple. An equilibrium between the multiple fluid phases will have to be found. The liquid will usually travel at a lower average speed to the gas (depending on the process conditions), with droplets or film condensing on the walls and moving back into the gas. The density, surface tension and viscosity of each component will cause various effects, where oil may be found in water droplets, or the two may remain immiscible. At higher energies, the gas may carry the liquid along as a homogeneous mist; decreasing the energy may show the liquid becoming an annular film or stratified flow pattern, depending also on the orientation of the flow meter, etc. It is the dynamic combination of these aspects that make wet gas more difficult to simulate and model.
A wet gas meter must therefore be immune to or take into account these effects to provide a robust calculation of flow rates. Future posts will expand on the technique and calculations used in typical wet gas meters as well as Solartron ISA’s Dualstream meters.
*Photos compliments of CEESI
