Engineering

Upstream Level Sensors

There are two level sensors measuring the upstream level. This is to improve the reliability of this important measurement. If one drifts out of calibration we don’t know which one is at fault but the turbine will shut down rather than unknowingly breach our abstraction licence.

Difference between the upstream level sensors (mm)

How to interpret this graph:
The margin from minimum abstraction level to overtopping the weir is 60mm so errors above 10mm become significant. Errors above 35mm will trip the turbine.

Sensor:
Type = Impress LMK307
Range = 4m
Measurement ? ± 20mm (0.5%)
Drift ? ± 4mm/year
Temp. ? ± 8mm/10K

Unexpected Loss EXPERIMENTAL

The turbine was analysed when running in a good condition and the graph below shows the difference between what would be expected from the turbine and how it is actually performing

Losses which were not predicted (W)

How to interpret this graph:

Short duration “spikes” should be ignored but a prolonged trend should be investigated.

Measurement accuracy is ?± 100W so negative discrepancies of 500W or more should be investigated to check the equipment, positive discrepancies of 500W or more should be investigated by checking level sensors for calibration.

Positive Values Negative Values Action 
Sensor drift or error Sensor drift or error Recalibrate level sensors
Blocked Exit Screen Check screen and clear if blocked
Bearing failure Check for noise (rumbling) at the bottom of the turbine
Gearbox failure Check for noise and hot spots on the gearbox
Cold gearbox If spike occurs when gearbox is cold (e.g. after a start up) – ignore
Poor software algorithm Poor software algorithm Check for unusual operating conditions (ice, spate, flood) – otherwise refer to technical support