April and May 2015: A reasonably good month of production followed by RPM sensor failure
by Kieran Monk
April saw a reasonably good month of production (figure 1) with no significant issues.
Production total April = 3,996.28 kWHrs Operational hours April = 225 Hrs
Production in May was far from ideal due to a combination of persistent low energy wave conditions and further issues with the RPM sensor. The repeated failure of the RPM sensor required more detailed examination and it was noticed that at higher turbine speeds the shaft coupling pulsated. It is believed that the coupling between the sensor and the turbine shaft and sensor is too rigid or that a sensor is not perfectly in-line. The result is forces being transmitted to the sensor resulting in accelerated fatigue of the rotating elements. Some strategies to damp the transmitted forces by using a soft shaft coupling are being considered.
Production total May = 1205.03 kWHrs
Operational hours May = 64 Hrs
With both resident engineers on site a number of maintenance activities were performed in addition to some activates for system optimisation. The following points were achieved in April.
1.)More optimised power take-off curves based on the theoretical curves given in (Falcão, 2002) where implemented into the inverter controller and testing has begun.
2.)Live statistical analysis graphics where implemented in the data acquisition suite. This shows the variance in turbine seed, the turbine stall frequency and severity in addition to a number of parameters relating to relief valve control. This allows the remote operator to more critically asses the plant operational conditions so that more educated adjustments can be made when someone is not there in person to observe the system behaviour.
3.)Full automation of the relief valve control routine (for short term pressure variance optimisation). Previously the relief valve control algorithm was implemented in a more basic way and was not robust enough to be left un-supervised. The routine was re-programmed to ensure redundancy of the system in the event that the routine fails or if the performance is weak. In this way the routine can be more confidently left un-supervised for long periods of time. Also the process of selection of the initialising conditions was also automated.
4.)Full machine maintenance including: testing all sensor electrical connections for functionality, complete pressure washer clean of the machine and re-greasing of corrodible components, Complete opening of the machine for inspection and carbon dust extractionform inside the generator, crack testing of rotor, welding repairs of cracks in the dummy guide vane section, welding of cracks in the guillotine valve housing, re-construction of the internal duct sealing’s to protect the generator and bearings, testing all sensor physical connections and condition (checking for blockages in Pitots, re-greasing linear sensorsetc..)
Falcão, A. F. d. O., 2002. Control of an oscillating-water-column wave power plant for maximum energy production,” Applied Ocean Research, vol.24, no. 2, pp. 73–82, Apr. 2002.. Applied Ocean Research, 24(2), p. 73–82.