Experimental and theoretical studies of turbine meters in two-phase flows.
This thesis presents the findings from a three year research
project into the performance of turbine meters in two-phase flows.
The aim was to determine the geometrical changes needed to be made to
the rotor design to allow it to register total volumetric flow when
subjected to two-phase flows.
A single-phase model is presented based on airfoil theory and
including retarding torque terms, velocity profile and blade
interference effects. Differences of less than 1% between
experimental and theoretical results, over the operating range of the
meter are shown.
The performance of a range of meters in various two-phase flows is
discussed. With the exception of the 40 degree rotor in water/air
the results may be summarized as follows. At low flow rates, of
between 5 and 20 to 30l/s, the meters tend to underestimate total
volumetric flow by up to -10% at void fractions of 25%. The greater
the void fraction the greater the underestimation. At high liquid
flow rates, above 30 to 40l/s, the meters overestimate total
volumetric flow rate by as much as 8% at a 25% void fraction.
A reasoned arguement is presented to explain the performanceof the
meters in two-phase flow. This is based on phase distribution
effects, influenced by the varying flow patterns, which alters the
magnitude of the driving torque developed radially along each blade.
Using this explanation the single-phase model is developed to
encompass two-phase flow. Agreement between experimental and
theoretical results is good, particularly in that the sense of the
meter error is correctly predicted by the model. Quantatively,
errors of less than 2% are achieved.
From this model, it was apparent that simple geometrica1 changes to
the rotors' design were unlikely to bring about the changes in
performance needed to meet the original aim of the project. It must
therefore be accepted that, in its existing form and using
established pulse total ising techniques, the turbine meter is not
suitable for use as a two-phase flow measurement device.
The author has discovered a relationship between fluctuations in
rotor angular velocity and flow void fraction. It is shown that
after suitable processing a single characteristic may be used to
describe the meters' response to all void fractions up to 25%, the
limit of the test facilities. After calibration of the meter, at a
single void fraction, it can be used to indicate both void fraction
and flow rate directly. Even at this early stage of development,
errors of less than 10% are being achieved in indicated void
This technique is covered by a patent, the rights to which are held
by the DTI as the funding body.