simon hurricks - genesis energy

P a g e 1 | Gas Turbines Conference 2016

Modal Balancing and Over Speed Testing of large two Pole Generators Following Modal Balancing and Over Speed Testing of large two Pole Generators Following Modal Balancing and Over Speed Testing of large two Pole Generators Following Modal Balancing and Over Speed Testing of large two Pole Generators Following RewindsRewindsRewindsRewinds

Simon Hurricks

Predictive Maintenance Engineer

Genesis Energy LtdGenesis Energy LtdGenesis Energy LtdGenesis Energy Ltd.

Modal Balancing and Over Speed Testing of large two Pole Generators Following Rewinds

P a g e 2 Gas Turbines Conference 2016

Contents Background................................................................................................................................................................... 3

Machine design considerations. .................................................................................................................................... 3

Why rewind? ................................................................................................................................................................. 4

Why Balance in a balancing machine? ........................................................................................................................... 4

Why high speed balance? ............................................................................................................................................. 4

Why over speed? .......................................................................................................................................................... 5

The risks of over speed testing. .................................................................................................................................... 6

Balance facility design. .................................................................................................................................................. 6

Risk mitigation. ............................................................................................................................................................. 7

In Conclusion. ............................................................................................................................................................... 8



Background.

Huntly Power Station has 4 X 250 MW single reheat turbo-alternators which were commissioned between 1982 and 1985. These machines are at the end of their design life but can be kept in service by carrying out major maintenance such as rewinding the generator rotors as required.

View inside turbine hall. Unit 4 is in the foreground. The Huntly generator rotors are 10 metres long and weigh 42 tonnes.

Machine design considerations.

Two pole generator rotors are made from a single forging into which winding slots are milled, as can be seen in the photograph below. The windings overhang the slots at each end and are restrained from centrifuging out by end-winding retaining rings or end caps. The end cap material must be non-magnetic and as such the rotor diameter is governed by the bursting stress of the end caps. To make the generator bigger the rotor is made longer which means that the shaft is more flexible. Typically large rotors run above the first critical speed and, as in our case, run just above the second critical speed. As a further complication the rotor stiffness across the poles is higher than that across the windings. To compensate for this there are pole slots as can been seen in the photo above. (Differential stiffness creates a 2X vibration which cannot be balanced out or removed) In order to reduce the physical size of the generator and to make it more efficient the rotors run in a pure hydrogen atmosphere. There are two reasons for this; hydrogen has less windage heating because of its very low density, and hydrogen has a higher heat carrying capacity compared to air and so it is more efficient in removing heat.

Rotor forging End windings

End caps



In our case the stator windings are hollow and are cooled by pure water again to remove heat and so make the generator physically smaller.

Why rewind? Over the life of the generator the stop/ start and loading cycling causes thermal expansion and contraction which gradually degrades the quality of the insulation. If left this can lead to one of two faults;

1. Inter turn faultInter turn faultInter turn faultInter turn fault. Where the insulation breaks down between adjacent winding layers. This causes a higher current in one side of the rotor and therefore the rotor bows from uneven thermal expansion. This is seen as an imbalance which is load related. To some extent this can be compensated for by adding balance weights but ultimately the machine will become un-runnable.

2. Rotor earth faultRotor earth faultRotor earth faultRotor earth fault. This is where the insulation between the winding and the rotor body breaks down which results in severe localised heating and will instantly trip the machine.

The ideal is to plan for a rewind just before either of these two events occurs so that the repair can be managed around other outages and load requirements. In the case of the Huntly unit 4’s generator there were subtle indications from the changing vibration patterns and a known deterioration in the insulation quality that prompted the rotor rewind which was carried out at BWE in Geelong in 2012. In the case of unit 1, the generator was again known to have deteriorating insulation quality and we had de-commissioned unit 3 and so had the unit 3 rotor rewound at Siemens in the UK in 2014 to be fitted into unit 1 during an extended outage in 2015.

Why Balance in a balancing machine?

With the rotor in a balancing machine all of the balance weight positions are accessible. On our rotors there are balance weight holes along the body of the rotor (body weights), there are weight positions in the end caps and there are weight positions in the fan hubs at each end of the rotor. When the rotor is in the stator only the fan hub weight positions are accessible and then only when the generator has been degassed, a weight change takes 24 hours for this reason. The vibration at the rotor 1st critical is driven by the static imbalance component, the 2nd critical is driven by the couple imbalance component, so it is crucial to be able to correct both of these components and to do this access to all the weight positions is required.

Why high speed balance?

Low speed balancing machines (hard bearing or soft bearing) typically run at a maximum speed of 600 RPM. At this speed the rotor is rigid i.e. running below 80% of the shaft 1st critical. In rigid mode the static imbalance can be corrected anywhere along the rotor. A shaft can therefore be balanced when rigid but because of the bow when running near the critical speed will still be unbalanced. The position of the balance weights along the shaft is therefore important. Similarly for the 2nd critical where the positioning of the correction weights is best at the mode shape anti-nodes. The windings can move radially in their slots and in a low speed balance machine will not centrifuge to their operational position which can affect the balance. The other issue for low speed balancing machines, especially with the larger rotors, is that they support the rotor on rollers which can impose high local stresses in the journal and can damage the journal surface. Not a good idea if the journals have been precision machined as part of the rewind process. High speed balancing machines spin the rotor at normal operating speeds. The static imbalance is corrected just below the 1st critical speed, the couple imbalance is corrected just below the 2nd critical and the 3rd critical is corrected by trim weights which do not affect either of the two lower critical speeds. The rotor is supported in white metal bearings which do not load the journal surfaces. The added benefit of running the rotor at operating speed, while still at the rewind facility, is that the rotor will also attain operating temperature from windage heating and this allows full electrical testing under operational



conditions. These tests include insulation resistance and shorted turn testing utilising temporary slip rings fitted to the rotor.

Why over speed? The endcaps have a heavy interference fit on the inboard end and in our case are retained by a bayonet fitting. To fit and remove the endcaps they are heated to between 200 and 250°C. When the endcaps are refitted they may not be perfectly square i.e. they can be slightly skewed. The endcaps weigh around 1.6 tonnes and even a few thou” skew will cause a significant imbalance. If the skew is merely balanced out it is highly likely that the endcap, after a number of thermal cycles, will straighten and therefore the balance will change which will force the rotor to be rebalanced and there is no guarantee that the endcaps will not shift again in service.

On completion of balancing the rotor is over sped to 3600 RPM. At this speed the centrifugal force on the endcaps allows them to float off the interference fit and centralise. Once the end caps have been centralised the rotor is further trim balanced. Note that it is not possible to over speed the rotor sufficiently in the machine as this would over stress the turbine components. Fully built up, the maximum allowable speed is 3300 RPM.

Endcap removal. Endcaps removed.

Rotor showing temporary slip ring assembly.



The risks of over speed testing.

The integrity of the end caps is thoroughly checked by dye penetrant and ultrasonic NDT techniques when the end caps are removed and again when they are refitted. Although the risk of end cap failure is very low the consequential damage is very high as the photographs below attest.

There are generally two types of over speed; 1. IntentionalIntentionalIntentionalIntentional. Where the machine is deliberately, in a controlled manner, over sped to test the operation of the

over speed protection system. In our case the over speed protection was by mechanical over speed bolts. These have now been replaced by an electronic over speed system and testing is done by injecting a signal so that there is no need to do a physical over speed of the machine.

2. UnintentionalUnintentionalUnintentionalUnintentional. Where the protection has failed and the machine accelerates until something lets go as in the case above.

Catastrophic failures have occurred in both cases, whichever way there are huge amounts of energy released in a very short time. If a failure occurs in the machine then there is the added problem of large volumes of hydrogen being released as well as large quantities of oil. In this situation it is most likely that the machine will be written off and or the high risk of injury to personnel.

Balance facility design.

A high speed balancing and over speed facility has to be designed with two things in mind; 1. The bearing support structure/foundations etc to be such that the response of the rotor is as close as

possible to that of the rotor in the machine. 2. The facility is designed such that in the event of a catastrophic failure all the resulting debris are contained and

at the same time allow crane access to lift the rotor in and out of the facility. 1. Is achieved by mounting the bearing pedestals on a stiff foundation plate and can be measured by comparing

the critical speed in the balance facility with those in the machine. 2. Is achieved by building the facility in a concrete lined pit below floor level and to have a removable concrete lid

over the balance pit when the rotor is running.

BWE facility in Geelong.



Siemens facility in Newcastle UK Not so obvious is the stiffness of the mounting plate of the two facilities. The difference became apparent when running the rotors. At the BWE facility the indicated critical speeds were considerably lower than those in the machine as below;

In the Siemens balance facility the critical speeds were essentially the same as in the machine. The implication of the critical speed differences is that the Siemens mounting / bearing assembly is considerably stiffer than that at the BWE facility. Even more obvious is the size of the concrete lids at the two facilities. The lid over the Siemens facility weighing in at 2540 tonnes and makes the BWE lid inconsequential by comparison. When running the facilities it was noticeable that at the BWE facility the whole workshop structure vibrated at certain running speeds. You could hear the Siemens facility running but could not feel the vibration through the floor. The power supply at the Siemens facility was directly from the grid visa transformers.

Risk mitigation.

The BWE facility had an ingenious power source, namely four old diesel electric locomotives and two 1000 HP diesel generators. The noise precluded operation of the balance facility at night.

Installed in Unit Balance rig

1st 1150 RPM 800 RPM

2nd 2750 RPM 1600 RPM

3rd Not seen 3200 RPM



Over speed testing at the Siemens facility was done at night when there were no other people around.

In Conclusion.

When a rewind is under taken, carrying out a full speed balance and electrical testing minimises the possibility of there being a fault in the winding prior to installing in the machine. It is better to find out while still at the rewind facility. Over speed testing ensures that the end caps will not move when in operation and in the unlikely event that a failure occurs it is better for this to happen in a controlled situation which minimises the potential loss to that of the rotor rather than the whole machine and possibly other machines in the locality.

simon hurricks - genesis energy

Engineering