IN 1998, vibrosystm installed an Air Gap Monitoring System (AGMS) on a 20 year-old bulb unit and conducted a series of post-refurbishment tests. These included a load rejection with overspeed to verify various parameters during this critical transient condition. This test reveals the mechanical stiffness of the rotor and stator, shaft alignment and displacement, magnetic imbalance, and the vibratory behaviour of the machine. It warrants the machine safety margin in the eventuality of sudden protection shutdown.
A load rejection happens when a machine is under load (often full load) and the main breaker is suddenly opened (load is removed). For a moment until the wicket-gates react, water still rushes freely through the turbine. With the generator free of generator load, the machine accelerates and can go into overspeed until the wicket-gates close and the machine slows down to a stop. At the instant the load is removed, the stator assembly moves outward as magnetic force no longer pulls it in. Simultaneously, any mechanical unbalance and offset of the rotor, shaft and turbine may be seen as abnormal vibration.
Figure 1 (right, top) shows air gap variation of all poles facing one sensor over approximately 50 machine rotations where the load rejection took place around turn No 16. Figure 2 (right, centre) emphasises two rotor poles facing the same sensor during the critical phase from load rejection until maximum overspeed.
A slight gap increase is immediately visible at turn No 17 reflecting the stator moving back from loss of magnetic pull. A rapid reduction of the air gap is observed as the rotor rim expands from the centrifugal force up to overspeed (150%) at turn 23. A consistent contraction of the rim follows which results in an air gap increase as the machine slows down.
As the load rejection occurs, the stator sets back by 0.09mm. The rotor roundness value varies from 0.55mm (5.5%) before the load rejection to 0.58mm (5.8%) at maximum overspeed, and down to 0.49mm (4.9%) at the end of the measurement. Overall, the air gap reduction between turns 16 and 23 is 0.13mm (1.3%). All these values are considered highly acceptable.
However, comparison of all sensor curves reveals differences typical of a rotational axis displacement when the field shuts off. Instantaneous gap increases were observed on sensors at 45°, 90° and 135°, and gap reductions on opposite sensors at 270° and 315°. The rotational axis is displaced by 0.18 to 0.25mm toward the location of the bulge on the stator (Figure 3; right, bottom). If vibration monitoring had been integrated to air gap in a ZOOM machine condition monitoring system, such displacement could have been easily confirmed by correlating air gap variation with shaft displacement at the generator guide bearing.
Further analysis of the same data estimates a stator set back in the range of 0.05 to 0.09 mm, which is typical of excellent stator stiffness.
VibroSystM results analysis demonstrates that the stator and rotor mechanical stiffness is excellent, and the rotor rim is tight and well balanced. However, the bulge on the stator, created at the initial commissioning of the machine, produces a magnetic imbalance that pulls the rotor toward the opposite direction (135°) when field is applied.
Based on observation of the generator dynamic behaviour recorded by the AGMS throughout the tests, the VibroSystM report concludes that the projected correction of the stator shape is not necessary in the short term for the following reasons: 1) the bulge is present since machine commissioning, and 2) the stator demonstrates excellent stability. Instead, regular air gap verifications are recommended to keep a close watch on stator roundness and minimum gap. The report also recommends to pay attention to the stator windings in the bulge area as the variation in slot width from the nominal width could create bar looseness or compression and affect the coils thermal expansion, ground contact and bar vibration.