Thursday, January 17, 2013

CHAPTER 10 MOTOR FAILURES




10.1     GENERAL


The squirrel-cage induction motor, as used offshore and onshore, is a very robust machine and should give little trouble in service unless it is abused.

Nevertheless, like any machine, it can be subject to failure, and this chapter deals with those failures that are most likely to be met.  Excluded from the list are such obvious things as physical damage or flooding, which can hardly be blamed on the motor itself.

Failures in a motor may be electrical or mechanical. They are discussed below.

10.2     INSULATION FAILURE


Failure of insulation of the stator windings, which carry the full system voltage, can cause a current leakage to earth or between phases.  If this is not detected and the motor taken out of service, it can quickly lead to a complete breakdown and a flashover to earth or between phases - that is, a short-circuit.  Short-circuits, even though starting between two phases, spread rapidly to envelop the third phase and produce a full 3-phase short-circuit.

Insulation breakdown can occur if the original insulating material was faulty, but it is much more likely to be due to damp or overheated insulation.

If a motor has become damp through a long period of disuse, it is always advisable to dry it out with its own heaters or a temporary heater before voltage is applied.  An insulation resistance test will indicate when a satisfactory state has been reached.

Deterioration of insulation may not at first be enough to cause breakdown, but if repeated or prolonged it is progressive.  The surge of current when starting a motor in this vulnerable state may just be enough to cause the final breakdown.  Again, periodic insulation resistance tests should indicate any progressive deterioration that may be taking place.

10.3     OVERCURRENT/OVERLOADING


During the course of its service life a motor may have become overloaded and so subject to overcurrent in its windings.  Such overcurrents may perhaps not have been enough to operate the motor’s overcurrent protection but sufficient to overheat it if prolonged, remembering that heating is proportional to the square of the current.  Thus a 20% overcurrent will cause over 40% heat generation, leading to possible and eventual insulation breakdown.  Continuous running at reduced voltage can produce the same effect (see Chapter 6).

Stalling of a motor while still connected and running can lead to a similar overcurrent (but not ‘overload’) situation, where the current rises to its starting value and is sustained without ventilation until the motor protection operates.  With a typical starting (‘locked-rotor’) current five times full-load current, the heating rate is 25 times that at full-load, aggravated by lack of ventilation.  Such a situation, if sustained even for a few seconds, can quickly lead to insulation failure and breakdown.

Repeated attempts to start, or even repeated successful starting over a short period, can also cause overheating of the windings, as explained in Chapter 5.  For this reason some operators have made a strict rule against repeated starts, which is set down in that chapter.

10.4     CABLE BOX FAILURE


One area where flashover is not uncommon is the cable entry box to the motor.  This is usually a ‘trifurcating box’, where the 3-core power cable enters through a gland, and inside it the outer cable sheath is removed and the three cores are led separately to the motor terminals.

It has been found sometimes that during installation individual cores may have been bent too sharply, so cracking their insulation and presenting a weak spot for eventual breakdown.

A cable box failure can be very dangerous, as the release of energy caused by a flashover in the confined space of the cable box can lead to what amounts to an ‘explosion’, with danger of fire and to personnel.  In high-voltage motors where the fault level is also high (see the manual ‘Electrical Protection’) the release of energy can be enormous.

10.5     ROTOR BURNOUT


Rotors of squirrel-cage motors have uninsulated conductors lying in slots in the laminated iron; there is therefore no problem of insulation failure as with the stator, and furthermore the iron rotor in direct contact with the conductor bars provides a good ‘heat sink’.  Rotors can have other problems however.

The rotor conductors carry very considerable currents when running, and even more when starting.  A rotor is susceptible to similar overcurrents as occur in the stator, and for the same reasons as given in para. 10.3.  The usual construction of the squirrel-cage itself is for the conductors to be brazed into the end-rings.  If the quality of the brazing is not good, the rotor currents may cause excessive heating at the joint due to poor contact.  This can become progressive, leading to final breakdown of the joint, severe arcing and eventually total burnout, which could well involve the stator too.

When motors are dismantled for overhaul, special attention should be given to the brazed end-ring joints for signs of overheating.

For motors with wound rotors and sliprings (very rare), the rotor would have the same insulation problems as described for the stator, to which may be added possible trouble with the sliprings and brushes.

10.6     BEARINGS


Most motors have rolling bearings, and sometimes one bearing is shared with the driven load.  Larger motors usually have journal bearings.

The calculation of the ‘life’ of a rolling bearing is a complicated business involving the maker’s data for that bearing (based on a stated speed and total running hours) modified by the actual operating speed and the total running hours required.  The designer calculates the correct bearing for the job, and it should give the life he expects of it.

However, this does not always happen, especially if the bearing has been subjected to loads in excess of those calculated, or if it has undergone particularly severe conditions such as excessive temperature or vibration.

Bearings will consequently fail in service from time to time and will need to be replaced.  Failure is usually indicated by increasing noise and can often be confirmed by putting a rod in contact with the bearing case to the ear.

When a bearing is replaced it must be replaced by the correct one recommended by the manufacturer.  Not all bearings, even of the same dimensions, are interchangeable, since bearings with differing degrees of internal clearance are manufactured.

Particular care is required with greasing.  A new bearing is usually provided in its box packed with a preservative grease.  This is not the running grease and must be completely removed and replaced by the correct quantity of the correct running grease.

Bearings must be greased regularly as part of the planned maintenance routine.  They must never be overpacked with grease and never overgreased when running.  This causes viscous friction and consequent overheating, leading to melting of the grease.  One of the major causes of bearing failure is due to overgreasing.

Journal bearings, which are likely to be found only on the largest motors, should give little trouble so long as they continue to have proper lubrication.  In many large motors part of the starting sequence is to pressurise the bearings to ‘float off’ the journal before movement takes place.

The quite large stray magnetic fields in the area of the bearings induce voltage in the journal itself, and, as the whole rotor is floating on a bed of oil, these charges cannot escape.  If they build up sufficiently, spark-over can occur between the journal and the metal shell, breaking down the oil film and possibly damaging the metalled bearing surface.

Some large machines are provided with insulated bearings and with a brush running on the shaft near the bearing and connecting it to the earthed frame.  This discharges any build-up of voltage on the journal and should prevent the sparking problem.  Periodic checking of the earth continuity of this brush is desirable while the motor is running.

10.7     VIBRATION


Vibration is one of the major causes of failure in rotating electrical machines.  It may arise from a number of causes:

Mechanical unbalance
Electromagnetic unbalance
Thermal unbalance
Unbalance induced by starting and restarting.

Mechanical unbalance is due to the centre of gravity of the rotating mass not being exactly on the centreline of rotation.  It is checked by an overspeed test carried out by the manufacturer and, if small, is compensated for by adding small balance weights at the proper points on the rotor rim.  This, however, is a ‘basic’ test and is not normally repeated in production machines.

Provided that the production machines are correctly assembled with good quality control, there is no reason why their rotors should be out of balance.  If they were it should have become apparent - and been corrected - during the normal routine check tests.

A squirrel-cage rotor is very robust, and there is little scope for movement of the conductor-bars to cause unbalance and therefore vibration.  It is just possible that one of the small balance weights may work loose or even be shed; this would reveal itself by noticeable vibration.  In such a case the motor should at once be stopped and the cause found and corrected.

Electromagnetic unbalance can be due to an unequal air gap around the rotor, causing unequal magnetic pull as it rotates.  It can also be due to an electrical fault in the rotor windings, causing distortion of the magnetic field system in the air gap.

Thermal unbalance may be due to uneven thermal strains distorting the material of the rotating mass.

During the starting of a motor the currents in the three phases are asymmetrical at first and cause a directional pull on the rotor due to the d.c. components of the currents.  This is short-lived but may have a cumulative effect over a period of use.

Some of the largest motors are provided with ‘Vibration Monitors’.  There are sensors at each bearing which, through an electronic circuit, give an alarm if the vibration reaches a certain preset lower level and which trips the motor if it reaches a preset higher level.

Vibration is measured by accelerometers mounted at suitable points on or near the motor. They measure acceleration directly and, by suitable integrating circuits, can also be made to indicate vibration velocity or vibration amplitude.  All these quantities are periodic, and it is customary to express them in ‘root-mean-square’ terms: thus acceleration is expressed as ‘mm/s2 (rms)’, velocity as ‘mm/s (rms)’ and amplitude as ‘mm (rms)’.

Some operators set acceptable vibration velocity limits in their specifications for electric motors.  Typically it should not exceed 3mm/s (rms) measured on the bearing on a horizontal plane (for horizontally mounted motors) through the shaft centreline and with the completely assembled motor running at no-load.  A vibration sensed by the monitor exceeding 5mm/s is regarded as an alarm level and above 11mm/s as critical.  The vibration measured on the shaft relative to the bearing must not exceed 12mm/s (rms).

In practice it has been found that, although the vibration limits on test in the maker’s works have been satisfactorily met, they are often exceeded when the motor is erected on site.

No comments: