Investigation of resonance in propeller turbines: The case of units 14/15 at Kiira power station

Abstract

This dissertation presents results of an investigation which focused on determining if resonance is the cause of vibration on two propeller turbines at Kiira Hydro-Power Station, located in Jinja, Uganda. The main goal of the investigation involved finding out whether the interaction of frequencies from wicket gate wakes, runner and inner cover in operating conditions would result in a resonance situation. Draft tube performance was not considered since there was no unusual noise or vibration found and the tools to assess it’s behavior were not readily available. This method was selected because a review of vibration measurement results from existing online vibration monitoring system against vibration standards ISO 7919-5 and ISO 10816-5 indicated that the problem area was around the turbine section. A 3D CAD model of the turbine runner was developed and it’s natural frequencies were evaluated using Finite Element Analysis (FEA) in dry (air) and submerged (running) conditions. The results of the runner from FEA were validated against measurements of vibration frequencies which were obtained from within the turbine pit. The inner cover was the only stationery part which was modelled and investigated in both dry and running conditions since it is the part which forms the turbine pit where the unwanted vibrations are being experienced. Natural frequency results of the inner cover were matched against the results from the runner in running conditions so as to detect conditions of resonance. The wakes from the wicket gates and their interaction with the turbine runner were calculated using modified equations for pelton turbines. Frequency measurements within the turbine pit from startup to full load indicated that the dominant problem frequencies were 67Hz and 85/87Hz. The outcome of the FEA investigation showed that the inner cover appears to be resonating with the runner around 85Hz. It was also found out that the dominant frequencies were not related to the wicket gate wake frequencies and hence turbulence and vortex shedding within the runner chamber was not the possible root cause. It was found out that most suitable field mitigation solution to the problem was the modification of the inner cover by removal of the water trapped inside by blanking off the aeration holes and widening the hole directly under each vacuum breaker. This solution is expected to reduce the “felt” vibration within the turbine pit substantially.