Logo

Condenser Bushing Air End Termination Issues

Presented By:
David M. Geibel
ABB Inc., Alamo, TN
TechCon 2017

Abstract

The purpose of this paper is to provide information regarding several known over heating issues with air end terminations of condenser bushings manufactured by General Electric Co., Westinghouse Electric, ABB, Inc. and others. These bushings, while applied to other apparatus such as circuit breakers, are considered here primarily as applied to transformers and oil filled reactors. I will refer to all of these as “transformers”. The issues addressed are known root causes, but the paper does not discuss in detail all possible effects of these issues. This is also not intended to suggest these are the only root causes for such issues.

Introduction

For the past 100 years or so, the condenser bushing has been the device of choice for allowing electric current to pass through the tank of oil filled transformers at voltages above about 25 kV. These, for the most part, have been oil filled devices themselves. At the external end of these bushings is a current carrying part which is used to connect the external conductors to the bushing. It is typically the only location in or on the transformer where current passes through a conductor in contact with a gasket. This presents some possibilities for this gasket and associated joint to be threatened by heat and mechanical loads. Therefore, this is a critical joint.

Overheating of Bushing Gaskets

Each bushing must have a seal of some kind at the air end of the bushing. Generally, this seal is the result of a carefully designed gasket installation. However, the selection of the gasket material, gasket design, and gasket application are not trivial exercises. They are a balance of compromises to accommodate criteria such as thermal expansion of parts, hot and cold temperature performance of materials, tightness of parts, and mechanical loading of the bushing. Over the decades, the gasket providing the best overall performance has been Nitrile. More specifically, a grade of Nitrile with cold temperature performance is very common today. However, Nitrile, like any polymer pliable enough to be a gasket, has temperature limits for a reasonable life. If this temperature limit is exceeded for any substantial length of time, the sealing characteristic of the material will deteriorate as a function of the time and temperature. See Figure 1 for such relationship for a high quality Nitrile typically used for this application. Keep in mind that “life” may not indicate the gasket will leak but that it no longer has its original sealing capabilities. In any event, temperature is critical to the life of the gasket.

As you can see, the life of a gasket is quite short if applied at 200°C (half an hour) but lasts long beyond the life of a transformer at 100°C; at, for example, 85 or 90°C life is more than one million hours, if not attacked by something else such as UV light or ozone. The maximum design hot spot of a bushing (and therefore of its gaskets as well), designed to the IEEE standards, is 105°C (30 ambient + 75 rise). This assumes these gaskets are located at the hottest spot of the bushing which is not generally the case. The hottest spot is often under the thermal blanket of the condenser well below the gasket. The problems begin when the issues addressed in this technical paper result in elevated temperature of the air end terminals such that premature aging occurs and the seal is compromised. You can also see from the graph that at temperatures around 100°C, as little as 5°C will easily reduce the life of the gasket by a factor of ten.

Gasket Life as a Function of Temperature

It is clear from this that gaskets are very vulnerable to overheating. Below is a discussion of some commonly known root causes for overheating of gaskets in the subject application. It is important to know that these causes, while problematic independently, produce much greater threats when combined in almost any fashion.

It is important to point out that once a gasket is aged by overheating, reducing the temperature will not restore the gasket. If, for example, a gasket is heated to 200°C for several hours, the gasket will be destroyed and the bushing compromised regardless of the future operating temperatures. Once compromised, it is only a matter of time before the performance of the bushing will be unsatisfactory due to moisture or other contamination. Gasket thermal deterioration is not reversible.

To illustrate the effects of several root causes of overheating, ABB performed thermal testing to evaluate the temperature rise of these root causes. Figure 2 shows a basic setup for these tests. A current loop was assembled which included the test piece, appropriate bus work, current drivers (reversed current transformers) and measuring current transformer. This loop was then installed into a test tank with oil to a level just below the bushing flange as it would be in service. Current was forced through the loop until equilibrium was reached and the temperatures of interest were measured with thermocouples and recorded with a data logger.

The results of these tests are included in the sections below. It may be of interest to note that this process is an abbreviated version of how bushing heat runs are done for type testing.

The experiments included the following:

  • Loose Connections
  • Undersized Spade Adapters
  • Undersized Bus Connection
  • Reduced Size Bushing Terminal Stud

Current Loop with Current Drivers

Loose or Poor Connections

As one would expect, if the connections in a current path are not properly tightened, a heating issue is likely. What might not be quite so obvious is where all these connection points are found. One joint, for example, which can create heating, is where the top terminal is screwed in place but is not under the spring load of the bushing clamping. This type of joint might be a draw-lead top terminal or other removable terminal. If they are installed without sufficient torque or with the improper gasket, they may begin to loosen with transformer vibration, thermal cycling, and line pull. The torque required to loosen these joints is not that high compared with the torque which could be applied when the spade terminal adapter is added and the lead is connected. It is easy to inadvertently reduce the torque during the installation.

To investigate the possibility, a bushing was constructed and connected as previously described. This bushing was tested for temperature rise at 3000 amperes before and after being modified to reduce the torque at the installed bushing top terminal. The results are indicated in Figure 3. It can be seen that the test results indicate a temperature rise increase of 5.6°C with the looser connection. It should also be noted that the mating materials in this test were all in like-new-condition and that if surface corrosion or contamination were present the effect would have been greater. Also note how consistent the rise is with the tight connection, but not with the loose connection. This would indicate that a non-stable condition exists at the joint. It can be expected that this joint would become problematic with time in service.

Temperature Rise - Loose versus Tight Connections

Undersize Spade Adapter

We believe that undersized terminal spade adapters are the root cause of some air end bushing heating issues. It can be seen in Figure 5 that these adapters are available in various sizes and material for the same stud size. They are not rated by current. For our testing, we selected two sizes which are referred to by weight as either 5.6 pounds (5.6#) for the large or 4 pounds (4#) for the small to be installed on the same size stud terminal. See Figure 6 for the installation. Figure 4 shows the results of the test which indicates, even when new, that the 4 pound terminal is running 6.5°C hotter at 2000 amperes.

Temperature Rise with 5.6# vs 4# Spade Adapter for the Same Application @2000 amperes
Photograph of Two Spade Adapter for the Same Size Stud
Installation of 5.6# Adapter on 1.5” Stud Terminal @ 2000Amperes

Undersized Bus Connections

The IEEE standard for transformer bushings requires a bus connected to the bushing for thermal testing with a rise of at least 30°C over ambient air at a point approximately 1 meter from the bushing connection. The reason for this is that it is known that the temperature of the bus affects the temperature of the bushing. Bushing manufacturers, for the same reason, use the coolest bus possible that meets this standard when testing. The IEEE bushing standard states that the “usual service conditions” are when the bus is less than 30 °C rise. Therefore, in application, if the bus connected to the bushing is hotter than this limit, the bushing will also run at a temperature higher than its design tests would indicate. A test was done with various bus connection temperatures to see how they affected the temperature of the bushing stud. Figure 7 shows a typical bus connection. The results are shown in Figure 8 which indicates an increase in bus temperature rise of 28°C translates to a bushing stud increase of 11°C for this example. The increased bus rise was accomplished by using bus work of reduced cross sectional areas.

Typical Bus Connections for Thermal Test
Effect of Bus Temperature on Bushing Stud Temperature

Reduced Size Bushing Terminal Stud

Reducing the size of the bushing air end terminal stud became an area of interest during this exercise. It was decided that an experiment would be performed by modifying the design of the top terminals of 3000 ampere and 2000 ampere bushings and testing them along with the traditional designs. See Figure 9 for the results of these tests. As you can see, the temperature rise increased 6.5°C and 5.0°C respectively for the size reductions of the 3000 and 2000 ampere designs. The 3000 ampere design was reduced from 3” to 2” and the 2000 ampere design was reduced from 2” to 1.5”. As you can see, at 3000 amperes this 2” stud would now approach a 95°C temperature in a 30°C environment.

Comparison of Temp vs Bushing Stud Size

Overloading of Bushing

Bushings are designed to have a loss of life less than that of transformers, assuming that the transformer is designed and operated per the IEEE loading guides. However, this “life” being referred to is, as with transformers, the life of the insulation system as a whole, and is not intended to specifically address the gaskets. Nevertheless, most bushing manufacturers design the gasket systems to be consistent with the aforementioned loading guides. If a bushing is applied at its full rating and the transformer is significantly overloaded, the effects of any of the previously mentioned issues will be amplified significantly.

Discussion

So what if the air-end termination runs a little hot? As mentioned above, it takes very little time at elevated temperatures to damage a gasket seal. If this occurs, several failure modes are possible. The most obvious is the contamination of the insulating system of the bushing. It must be kept in mind that the quantity of oil in a bushing is small in comparison to a transformer, making small amounts of moisture very threatening. But, there are some less obvious failure modes worth mentioning.

If the seals at the top of a draw-lead bushing are damaged by heat, it is possible for oil to find a path to migrate into the bushing from the transformer or vice versa. The problem of oil loss from the bushing is obvious, but the gaining of oil can also be very problematic if the bushing becomes completely full. If this occurs, the oil may reach very high pressures when the bushing is hot if the leak is too slow to relieve such pressures. When this occurs, rupture of the bushing has been known to occur.

Gasket damage can allow minute amounts of moisture to find its way into locations where corrosion can occur. For a lot of good reasons, most bushing less than 3000 amperes are constructed with an aluminum conductor. But for reasons known only to the users they require the terminals to be copper, only to be converted to aluminum to be compatible with their aluminum connectors and cables. In any event, the transition from copper to aluminum must be done in a location which will remain dry and free of oxygen to avoid galvanic corrosion. Figure 10 shows what can happen if the seal protecting this joint is damaged. Figure 11 shows advanced corrosion and heat damage.

Beginning of Galvanic Corrosion due to Damaged Seal
Advanced Galvanic Corrosion and Heat Damage

Once this process starts, the joint deteriorates quickly. The moisture causes corrosion, increasing the electrical resistance, which, in turn, produces more heat, which further damages the gasket and accelerates the corrosion. Eventually the joint acts as a heater that can cook the joint in question and adjacent joints as well. See Figure 13 for an example of an adjacent joint which was compromised by this type of heating. In this case, an overheated joint, whose gasket was not critical for prevention of moisture entry into the bushing, damaged the gasket of an adjacent joint which was critical. Fortunately in this case the redundant gasket was still functional.

Figure 12 shows a thermal image of a bushing with parts like those shown in Figure 11. This bushing is shown at over 100°C at only 60% of its rating. Clearly the heating at 60% is already detrimental, but would be much worst at full load.

Thermal Scan of Overheated Bushing Due to Corrosion
Heat Damage Escalated to Damage Surrounding Seals

Conclusion

Top terminal heating is an issue which needs to be addressed by all users of bushings to insure that the integrity of the bushing is maintained. Proper gaskets and torques are of utmost importance.

Proper sizing of connections and bus or cables to limit their rise to 30°C is very important to the life of the bushing.

Thermal scans of bushing are a means to determine if overheating is currently occurring, but it must be stressed that they do not indicate damaged gaskets until the deterioration begins to produce heat. If the loading is very low, high loss connections may not be evident.

With the revision of the IEEE standard in 2000 many terminal sizes were reduced below the sizes traditionally used by General Electric Co, Westinghouse Electric, and ABB, Inc. in the last half of the 20th century. For example, many 3000 ampere bushings are now required to have 2” studs instead of the traditional 3” studs and likewise 1.5” has replaced 2” for 2000 amperes. ABB, Inc. have been, for the past 10 years, compliant to this new standard with all new bushing designs but has seen an increase in terminal heating issues. At the time of type testing, these designs met all the thermal requirements but sometimes developed problems as described in this paper. It is also notable that the manufacturers of, for example, 1.5” spade adapters may not have had 2000 amperes in mind. Since these adapters are not current rated, most assume “if it fits it must be right”, but this may not be the case. Therefore, ABB, Inc. Alamo is now applying traditional size studs on new designs unless otherwise required by the customer. We are of the opinion that the added cost of manufacture is justified by the robustness of the design.

Acknowledgements

I would like to give credit to Robert Cottrell and Dominic Pollaro for their contribution to this paper. They performed all the numerous tests and they designed and constructed several special bushing setups to provide direct comparisons of conditions. This work was difficult, tedious, and messy not to mention they performed it in addition to their usual work responsibilities in a very limited time span. I believe they have nightmares about entanglement with dozens of thermocouple wires and mountains of little tubes of epoxy. Thanks guys.

References

[1] IEEE C57.19.00-2004, IEEE Standard General Requirements and Test Procedure for Power Apparatus Bushings

[2] IEEE C57.19.01-2000, IEEE Standard Performance Characteristics and Dimensions, for Outdoor, Apparatus Bushings

Join our email list

We use cookies to give you the best online experience. By using this website you agree with our cookie policy.