Analysis of pressure measurements in Kaplan turbines

In a risk situation the flow through a hydropower turbine must be reduced under a short amount of time. The reduction in flow causes low pressures in the region called draft tube under the runner in the turbine. If the pressure drops below vapor pressure the water in the turbine may evaporate and form large cavities. The phenomenon called water column separation describes the situation when these cavities of vapor forms and eventually collapses which may damage the unit and canbe a danger to personnel at site. Different types of flow reducing maneuvers are tested before commissioning turbines in order to detect dangerously low pressures but these measurements can be difficult to interpret. The measured draft tube pressure can be filtered numerous ways and in addition other measured or calculated parameters must be studied to fully understand the consequence of a measured emergency stop. In this paper the different ways of filtering and interpreting the measurements from stops in Kaplan turbines were analyzed. The results show that the choice of filtering method can significantly impact the interpretation of a stop and that the filter parameters often used do not result in equivalent results. The standard way to filter draft tube pressure with moving average filters is shown to result in much higher draft tube pressures than low-pass filters and the filter parameter used needs to be further investigated to correctly isolate pressures caused by water column separation. The main difference between different companies when it comes to these measurements are the used cutoff frequency when using low-pass filters and the sampling frequency of measurements