Efficiency measurements in low head hydro power plants

Abstract

The work presented in this thesis involves efficiency measurements performed with thethermodynamic method and the pressure time method. The thermodynamic method haslimitations with regards to the power plant head, as uncertainty will increase as the headbecomes low. The Pressure-Time method has limitations concerning geometrical propertiesof the waterways, as it demands a certain length of closed conduit with uniform area. Bothmethods are considered to cause relatively short downtime for turbines to be measured, andit is therefore of interest, for economical reasons, to expand the use of the methods to morepower plants.

In the thermodynamic method, a large source of errors and uncertainties comes from flowand temperature variations in the draft tube outlet. To investigate if this source of errors anduncertainties could be reduced, a setup with multipoint temperature and velocitymeasurements was installed in a low head hydro power plant. The general conclusion is thata multipoint measurement is beneficial in connection with low head measurements.Uncertainty from temperature variations in the water from the reservoir, and difference inthe energy at the inlet is also discussed.

For the Pressure-Time method, this thesis presents investigations made with shorterdistances between measurement cross sections than prescribed in the standards, and withbends between the measurement cross sections. The investigations were performed both inlaboratory and in a field measurement. For laboratory experiments it was also investigated ifdeveloped models for unsteady friction could be used to correct the flow estimates. Thegeneral conclusions are: Measurements with short distances should not introduce large errors, but theuncertainty and spread of measured points will increase. Bends give an underestimation of the flow rate that, under certain circumstances, isconstant independent of initial velocity. How large the underestimation ratio willbe can yet not be predicted. Introducing unsteady friction in the Pressure-Time calculations improves the flowestimate. However, the results are not fully satisfactorily, and more work has to bedone to investigate the correction due to unsteady friction.