Understanding how mechanical conditions in a generator’s brushgear system contribute to severe brush wear rates and slip ring damage can aid in solving these problems.
By Douglas E. Franklin
The performance of brushgear on large hydroelectric generators often is a vexing problem. Older machines with brushgear that has performed reasonably well for years can develop serious problems for no apparent reason. Similarly, new machines with tried-and-true brushgear designs can perform poorly from day one.
Understanding why there are problems often requires developing an understanding of the mechanical conditions affecting the performance of the brushes, brush holders, and slip rings. Parameters to consider include the tribology of composite materials as a function of current density, brush pressure, brush grade (quality), temperature, humidity, cooling air quality, and slip ring material/finish. Often, when there is a problem with brushgear, a single solution is envisioned. In reality, it is likely that more than one factor is contributing to the problem of good brushgear performance.
The design of a brushgear system usually is left up to the original equipment manufacturer (OEM). The OEM selects the brush type (grade/quality), holder type, slip ring material and design (finish and grooving), and brush holder support system. My experience shows it is not uncommon for new machines to experience brushgear problems until the best brush type is determined for the application/environment.
Three main system components associated with the control of an electrical machine are the exciter, brushgear assembly, and rotor field winding. The prime function of modern static exciters and older rotary exciters is to boost, or buck, the current in the rotor to achieve the desired voltage output on the stator. I consider exciters to be equivalent to large batteries, capable of delivering thousands of amps to the rotor winding.
For more than a hundred years, the method of getting the current from the exciter to the field winding has been with some form of carbon brushes riding on slip rings. Many different types, styles, and grades of brushes and brush holders are available. The slip rings can be made of cast iron, steel, brass, or other metals, and the rings can be solid or “grooved.”
It is important to look at all aspects of the collector-brushgear assembly system when trying to identify the root cause of problems with brushgear. In particular, the grooves in the slip ring can affect brushgear performance. If the impedance of the load on the exciter increases, the exciter boosts the voltage to maintain the required current. Because the impedance of the rotor field winding is unlikely to change, the additional voltage drop will be in the brushgear. The additional power output of the exciter will be dissipated by the brushgear in the form of heat, and the most likely place for this to occur is at the contact face between the brush and the slip ring.
Analysis of brushgear systems is basically the study of the tribology of composite materials. In other words, it deals with the design, friction, wear, and lubrication of interacting surfaces in relative motion. The two primary objectives in the design and operation of the brushgear system are to minimize the electrical contact resistance while getting power onto the rotor and to minimize the mechanical contact friction. This delicate balancing act is difficult; what is done to improve one aspect frequently has the opposite effect on the other.
Problems with brushgear often are identified by operating conditions such as overheating, rapid brush wear rates, excessive carbon dust, discoloration, sparking, brush chipping, pitting of the rings or brushes, and even a flashover and/or meltdown of the brushgear assembly. Over the years, numerous companies and individuals have investigated brushgear problems. Studies provided by OEMs recommend current densities, brush pressures, operating temperatures, ambient humidity, and slip ring surface finish. Suppliers of carbon brushes also will readily recommend a suitable brush for a specific application.
Typical concerns for maintenance personnel are brush current density, brush grade, brush/ring temperature, vibration, cooling air contaminants (such as oil, dust, or vapors), humidity, and spring tension. Brush suppliers usually look at current density, slip ring material and finish, and brush holder design, then make a recommendation on brush grade. Designers also look at the structural method of supporting the brushes and brush holders on the direct current (DC) bus. Unless there are unusual ambient conditions around the brushgear, such as is the case in geothermal units, there is little reason to vary from a standard design.
A standard design of the slip ring brushgear assembly consists of a series of identical brushes and holders mounted on the DC bus, equally spaced around a portion of the slip rings pointing radially inward. Some designs also have a slight angle of the holder to the tangent of the slip ring, and some have an axial offset of the brushes so that more of the slip ring is in contact with the brushes. As mentioned before, the slip rings are typically a high-quality carbon steel with either a solid surface or a spiral grooved surface. In the case of the spiral grooves, there can be from one to four spiral “start” grooves on a ring, and the groove edges are chamfered.
The mechanical forces on the brushes are those generated by the tangential friction on the slip ring surface, plus any radial forces given by the spring and “dynamic” forces due to out-of-radial-runout of the ring, radial motion of the generator rotor (vibration), and/or motion of the DC bus support. Any change in the tangential force on the brush will result in an equal change in the force the brush holder exerts on the sides of the brush. If the slip rings have spiral grooves, then an additional dynamic force is introduced by the wiping or threading action of the grooves in the axial direction of the rotor. This combined tangential and axial force can result in “wiggling” of the brush in the holder. This wiggling can result in intermittent loss in electrical contact between the brush and the ring similar to brushes bouncing on the ring.
Analyzing brush current
To better understand brushgear performance or, better yet, poor performance, in 2006 BC Hydro undertook a study at its 1,736-MW Mica Creek Generating Station. The purpose of this study was to look at the current distribution between individual brushes on slip rings. It would be expected that identical brushes on the same DC bus would have very similar currents because they are in the same electrical and physical environment.
The tests involved installing an AC/DC (hall effect) current transducer on the pigtail of the selected brushes, then displaying the current as a function of the angular position of the shaft where this angular position is derived from a synchronizing pulse. To average out the effect of individual transients, BC Hydro took 20 synchronous time averages and then added them together to display the typical current in a given brush for each revolution.
Figure 1 shows an example of the current trace for a well-performing brushgear system.
Figure 2: These graphs display the currents measured in several adjacent brushes, all on the same upper slip ring of a 200-MW unit.
Figure 2 shows the currents measured in several adjacent brushes on the upper ring of a 200-MW unit. On this unit, the brushes are staggered in height on the ring in a manner that causes the even-numbered brushes to follow one track and the odd-numbered brushes to follow another. Examination of these plots shows that the brushes running in each track on the ring exhibit a similar current profile. The data presented in these plots has been filtered at 40 Hz and synchronous time averaged over 20 shaft revolutions. This data has been plotted so that the current profiles in each brush are shown relative to the same shaft angle.
Applying the same testing methodology to a 60-MW machine with a high brush wear rate yielded very different results. Figure 3 shows an example of current dynamics in one brush on a slip ring at this site. With a mean current of about 60 amperes (A), there are peaks to 360 A in each rotation, which means one brush is carrying all the rotor current for a short period of time.
Slip ring grooving
The value of the grooving in the slip rings has been questioned because some machines have grooves and others do not. Little information is available, but it is generally thought that the grooves are for cooling and/or carbon dust removal. The only documented origination of the grooving of rings is from a brochure:1
“Helical grooving arose from the current distribution between brushes. When many are operating in parallel at high speed an air cushion is formed underneath the current distribution.
This photograph shows a typical setup of a brush/slip ring arrangement where the ring is grooved and the brushes are mounted on the direct current (DC) bus in holders having constant tension springs.
It was established about 1924 that a definite improvement could be obtained in cases of uneven current distribution, or “selective action,” by cutting axial slots across the contact face of each brush. The success of this arrangement was attributed to the removal of the gas layer between brush and ring which can give rise to unstable conditions and a variable contact voltage drop. With the removal of this gas layer the contact voltage drop becomes much more uniform and a great improvement is obtained in the distribution of the current between brushes operating on the one ring.
Some years later an application was made, and duly granted, for a British Patent for spiral, “helical,” grooving rings and commutators. This was a new concept on the problem and it achieved an improvement in current distribution by rendering each part of the brush conducting surface inoperative for a certain period of time during each revolution of the ring. Thus in the case of “selective action” where a particular brush collects more then its theoretical share of current, the brush is forced to shed its current and equilibrium is restored. With axial slotted brushes there is a risk that the selective action condition will persist as there is no forced shedding of current.
The presence of helical grooving gives the added advantage of precluding a gas layer under the contact surface of the brushes and thereby gives the same beneficial effect as that of cutting axial slots in the contact surface of each brush.”
The key point this excerpt makes is that the grooves were put in “high speed” (tangential) machines. In the relatively low tangential speed of hydro machine slip rings, the value of the grooves is somewhat questionable. The grooves produce a third force on the brushes that is axial relative to the rotor. If the brushes in the holders have a natural cant in the holder (axially) as a result of the spring action on the back of the brush, then the threading forces can either add to or work in opposition to this natural cant. If the threading force is in the same direction as the cant, then the result is a slightly increased force on the top or bottom of the brush holder. If, on the other hand, the threading force is in the opposite direction to the cant, a rocking or wiggling of the brush may take place. This wiggling likely will result in periodic contact break between the brush and slip ring. This breaking of contact will force other brushes to pick up the current and will likely result in sparking under the brush. This sparking can lead to spark discharge erosion of the ring and brush.
Comparing brushgear performance
Starting in the early 1980s and continuing through 2005, BC Hydro reviewed the performance of brushgear systems on 15 different machines. This review, performed on machines where BC Hydro suspected problems existed, showed several inconsistencies with the brushgear systems. Five examples are given below.
Example 1: In the early 1980s, BC Hydro tested three machines using an AC/DC current transducer. The utility found that the brushes on the machines with one, two, or three start grooves on the slip rings had, respectively, one, two, and three current “pulses” per revolution. This finding indicates the grooving plays an active part in the transport of the current.
Example 2: During other testing performed in the 1980s, BC Hydro determined that for one machine that had grooved slip rings and both brush holders mounted in the same orientation, the brushgear performed very well. Conversely, on another machine with the same setup – grooved slip rings and both brush holders mounted in the same orientation – the brushgear performed very poorly. This was traced to structural resonance of the DC bus causing brush vibration. The motion of the bus supporting the brush holders, and consequently the brushes, caused the brushes to bounce on the slip ring. After BC Hydro eliminated this resonance, the brush performance improved significantly.
Example 3: In 2003 and 2004, BC Hydro performed investigations on a number of machines that had grooved rings. In these units, the brush holders were on top of the DC bus for one ring and suspended below the other DC bus for the other ring. In general, only one ring on these machines developed a problem. To explain these results, the argument can be made that the current flow is opposite for the two rings. Another argument is that the axial force produced by the grooving on the ring is in the opposite direction relative to the base of the brush holder.
Failure of the brushgear on this 75-MW machine was attributed to oil contamination of the ring. However, heat generated during the failure may have destroyed evidence of other contributors.
Example 4: In 2005, observations of the brush motion on brushgear associated with a 465-MW unit showed that one set of brushes were visibly rocking along the rotational axis of the slip ring in the holders while the brushes on the other ring were steady. The brushes rocking were the ones performing poorly, resulting in chipped brushes and a damaged slip ring. This rocking motion results in intermittent contact of the brush on the ring and consequently poor performance.
Example 5: One machine BC Hydro tested in 2005 had a single-grooved slip ring with brushes mounted both above and below the DC bus. This means the brushes should all be carrying the same current as they are in parallel, in the same environment (temperature, humidity, contaminants, etc.), and the current is going in the same direction. However, there was a 14.6-degree Celsius difference in the temperature of the upper and lower brushes. This temperature difference must be caused by some differences in the system. The reason for this difference could be explained by the cooling air flow or by the drag of the slip ring grooves being opposite relative to the base of the brush holder. It also is possible there are subtle mechanical differences in the design of the upper and lower slip ring/brushgear.
Brushgear that is performing well should be left alone. Achieving the delicate balance of electrical and mechanical parameters resulting in good brushgear performance can take time. Often, more than one electrical or mechanical parameter is not right, so it is necessary to look at the entire design of the brushgear system.
Mr. Franklin may be contacted at BC Hydro, 6911 Southpoint Drive, Burnaby, British Columbia, V3X 4N4 Canada; (1) 604-528-1602; E-mail: doug.franklin@ bchydro.com.
1“Sliprings and Carbon Brushes on Turbo-Alternators,” Morgan Crucible, Windsor, Berkshire, United Kingdom, pages 16-18.
Doug Franklin, P.Eng., is a senior engineer in the Engineering Group at BC Hydro in Burnaby, British Columbia, Canada. For more than 30 years, he has been involved in commissioning and central maintenance of most of the utility’s large machines.