Evaluation and Grading of PF and DF Test Results
General
After the PF tests are completed and results obtained, each apparatus and equipment insulation should be evaluated to its serviceability. The evaluation criteria may be divided into four categories. They are
The recommended practice for evaluating test results is not only to assess whether the test results fall into one of the four categories mentioned above, but also to compare the test results with previous year’s results to see how much change has occurred in the condition of the insulation since the last test. This is to say that the year-to-year test results are compared for trending purposes to signify any changes in the health of the insulation due to normal aging, but as well as other causes. Any sudden and large changes in test results between two test intervals should be a cause of concern and should be investigated before putting equipment back in service.
Usually, the failure hazards of electrical equipment are expressed in terms of maximum allowable PF values, however changes in the normal dielectric losses (watts loss), capacitance, and AC resistance are also used for indicat- ing problems in the insulation. Depending upon the type of equipment, many manufacturers publish factory and operating limits for PF or capaci- tance values for their equipment which can then be used for evaluating the test results. The normal test values for various types of equipment used in the industry, and discussed in this text have been obtained from testing sim- ilar equipment in the field and factory over many years. The abnormal or unsafe limits have been established from correlation of known test values at which equipment insulation has failed in service. Insulation in deteriorated condition may operate for a period of time without a failure depending upon its exposure to abnormal operating conditions, such as voltage and current transients, short circuits, temperature, etc. However, it should be recognized that deteriorated insulation creates a definite operating hazard and if goes uncorrected will result in service interruptions and equipment damage. If deteriorated insulation is removed before failure, it may be reconditioned and restored to service with substantial savings in equipment cost and unnecessary service outages.
Whenever the test results are questionable or marginal, it is generally recommended to perform tests on more frequent basis in order to keep abreast of the condition and to establish a trend. A gradual and consistent increase in PF may be due to contamination, deterioration, or normal aging, where as a sudden increase in the PF is a cause of immediate concern even when the absolute PF value is not considered excessive. Whenever an increas- ing trend is established, then the equipment should be removed from service as soon as practical for inspection to determine the cause of the problem and to make appropriate repairs.
Analysis of the Results
The electrical characteristics of the insulation vary with temperature. In order to compare PF and DF test results on a periodic basis of given equip- ment or apparatus, the PF and DF values should be converted to base tem- perature of 20°C as was discussed in Section 3.2.3. The reader is referred to Doble Engineering Company’s Power Factor Test Data Reference Book for more details on the effects of temperature and correction factors. The analysis of the PF test results for the various types of equipment and apparatus is dis- cussed below and is condensed from the Doble Engineering Company’s ref- erence manuals.
Transformers
Typical problems in transformers that are found with PF (and DF) and excitation current tests are
• Wet insulation
• Short circuit windings and/or turns
• Corona damage and carbonization
• Contamination from sludge, varnishes, etc.
• Displaced windings and core damage
The appraisal of PF and DF test results is divided into oil-filled and dry-type transformers. Also, further reference to the term DF is not made in this dis- cussion since PF and DF are the same for evaluating the condition of the insulation.
Oil–Filled Power and Distribution Transformers
The overall PF test results of oil-filled power and distribution transformers indicate the insulation condition of the solid windings, oil, barriers, bush- ings, etc. The overall PF value for individual windings-to-ground and interwindings insulation of modern oil-filled transformers should be 0.5% or less, corrected to 20°C. Service-aged power transformers will have some- what higher PF values due to normal aging, loading (heat), and voltage stress.
A PF value as high as 1% is considered acceptable for older transformers when previous history or knowledge of PF value of the transformer, or similar trans- formers is not available. However, when the PF value of one insulation system of the transformer is higher than the others, for example, HV winding insu- lation PF is higher than the LV and interwinding insulation PF, then causes of the higher PF should be investigated. A PF value of 2% for extremely old power transformers may be considered acceptable. In the case of older transformers that utilized varnished-cambric or varnish insulation, these transformers may have normal PF values in the range of 4% to 5% at 20°C.
Transformers that are subjected to excessive internal forces due to possible through-faults or other causes may have windings that are physically distorted, that is the core–coil assembly configuration has changed from its original design configuration. If this should happen, then capacitance of the winding-to-ground and interwinding (i.e., CH, CL, and CHL) would have changed. Therefore measuring the capacitance of the individual winding is also important in judging the condition of the winding insulation. Also, the transformer excitation current test discussed in Section 3.6.3 is extremely effective in revealing damage to the core, core–coil assembly, tap changer, and short circuited turn-to-turn insulation. In evaluating excitation current test results for three-phase transformers, evaluation is based on a normal pattern of excitation current readings, that is either a two high and one low reading, or two low and one high reading depending on the transformer winding connections. A change in this pattern is a cause for investigation. Also, if benchmark data, such as factory or field acceptance data are avail- able, then the evaluation should be based on comparing the test results with the benchmark data.
Dry–Type Power and Distribution Transformers
The PF of dry-type transformers varies over a relatively wide range due to in part to the insulation of support insulators, bus work, and insulation materials. Corona is a greater possibility in HV dry-type transformers and the test procedure should include provisions for checking it. This can be done by making PF tip-up tests on the dry-type transformers. PF tip-up test is described in Section 3.6.7.1 under rotating machine testing. An abnormal increase in the tip-up value may be an indication of excessive corona or voids in the insulation. A comparison of the PF and PF tip-up test values with benchmark data if available, or with test results of similar units tested under similar conditions is recommended in evaluating the insulation con- dition of the dry-type transformers. It is not unreasonable to expect a PF of 2% or less for new modern dry-type transformers. However, PF may increase with age of the transformer, and may increase to 5%–8%. PF values substantially higher than the values discussed here should be investigated to determine the cause of the high PF. A better approach for appraisal of transformer insulation is to use the PF data recorded during the initial tests on the transformer, such as acceptance testing, as a benchmark for com- parison with subsequent test results.
Transformer Insulating Fluids (Oil)
The PF of good, new oil is expected to be 0.05% or less at 20°C. Used oil in good condition should have a PF of 0.5% or less at 20°C, and if the PF exceeds 0.5% then the oil is considered questionable for continued use. Oil having PF greater than 1.0% at 20°C should be investigated or reconditioned, or replaced. A high PF of the oil is an indication of the presence of contamination, such as moisture, carbon, acids, polar contaminants, etc.
Bushings
The problems found in bushing are
• Cracks
• Dirty bushings
• Loss of oil or compound
• Short-circuited condenser bushings
• Wet or deteriorated bushings or tap insulation
• Dirty tap insulation
• Corona in bushing insulation system
The PF tests including the hot-collar tests performed on similar types of bushing under the same test and weather conditions should test similarly, and be within acceptable limits. When PF of a clean bushing increases significantly from its initial value, it is usually due to the effect of con- tamination, such as moisture which lowers the dielectric strength of the bushing. The possibility of the failure of the bushing in service increases as its dielectric strength decreases due to the effect of contamination. PF tests, made on a regular basis, have been used in assessing the service- ability of the bushing over the years. To decide whether a bushing should be removed from service because it has a slightly higher PF than normal depends upon the magnitude of the overall PF and hot-collar test results. However, a bushing that shows a substantial increase in PF each year is an indication of potential failure hazard. It is recommended that the bushing insulation should be evaluated based on the results of PF, capacitance, and hot-collar test results.
With regard to hot-collar test, higher than normal losses are indicative of contamination or deterioration of bushing insulation. Any bushing differing significantly from others by few milliwatts (up to one-tenth of a watt for the 10 kV test) should be investigated. The watts loss limit in bushing for the 2.5 kV test is approximately 0.15 W. The loss of oil or compound may be detected by comparing the hot-collar test current rather than the PF value. Abnormally low test current (10%–15%) may indicate absence of compound or oil. Testing under successively lower petticoats (skirts) will show normal current reading when compound or oil is reached.
Modern oil-impregnated paper insulated condenser type bushings have PF of 0.5% or less at 20°C. Any such bushing which shows a significant increase should be investigated. Usually the capacitance and/or PF value of a bushing are provided by the manufacture on the nameplate of the bushing which should be used for grading the PF and capacitance test results. Also, the previous year’s test results if available may be used in evaluating the field test results.
Lightning and Surge Arrestors
The insulating quality of lightning and surge arrestors is affected by con- tamination, such as moisture, dirt, and/or corrosion. Other problems experienced with lightning and surge arrestors may be mechanical defects such as broken shunting resistors, broken preionizing elements, and/or misassem- bly. PF tests are effective in detecting these problems in lightning and surge arrestors. The lightning and surge arrestors are evaluated on the basis of dielectric loss (milliwatts or watts) of comparable units or previous year’s benchmark test data. The PF is normally not calculated for these devices since their capacitance is very small. Abnormal dielectric losses can be divided into higher-than-normal and lower-than-normal and which are indicative of the problems as listed below:
1. Higher-than-normal losses
a. Contamination by moisture and/or dirt or dust deposits on the inside surfaces of the porcelain housing or on the outside sur- faces of sealed gap housing
b. Corroded gaps
c. Deposits of aluminum salts apparently caused by the interaction between moisture and products resulting from corona
d. Cracked porcelain housing
2. Lower-than-normal losses
a. Broken shunting resistors
b. Broken preionizing elements
c. Misassembly
Medium-Voltage Circuit Breakers
The problems that may be identified with PF testing of medium-voltage circuit breakers are wet bushings, wet or damaged arc chutes, tracking across insulators, bushing and arc-chutes, etc. The PF test results evaluation for these breakers should be based on the basis of dielectric loss (milliwatts or watts) and not PF. These types of breakers have low capacitance and that small changes in the low test currents and losses can result in misleading changes in calculated PF values. The analysis of the test results should be based on a comparison of the test currents and losses of similar units or previous years test data on the same breaker.
OCB
The results of OCB depend upon the insulation condition of the bushings, oil and tank members. Therefore, the PF test results of an OCB should be evalu- ated for the condition of the bushings, tank insulation and oil.
The bushing are graded based on the analysis of open and closed circuit breaker tests including supplementary hot-collar test results. Tank insula- tion is evaluated based on TLI test. Oil is evaluated on PF and dielectric strength tests, and by a battery of other tests including visual inspection for carbon and other particulate matter.
The open breaker test provides information on the insulation condition of bushings, interrupters, lift rod guide, upper lift rod guide, oil (some), and tank liner (some). Whereas the closed circuit breaker test provides informa- tion on the insulation condition of bushings (both), oil, lower part of the lift rod assembly, and tank liner. What is then the difference between the open and closed circuit breaker tests. The major difference is that in the closed cir- cuit breaker test, a larger dielectric field is established, and therefore the dielec- tric losses of the closed circuit breaker test will be different from the sum of the two open circuit breaker dielectric losses. In addition, in the closed circuit breaker test the cross head is energized, therefore a stronger dielectric field is established on the lift rod assembly. Also, a stronger dielectric field is established in the tank in the closed circuit breaker test compared to the open circuit breaker test. However, the dielectric field is decreased between the interrupters in the closed circuit breaker test. In summary, the closed circuit breaker test basically measures the lift rod assembly and tank insulation, where as the open circuit breaker measures bushing insulation.
Since open breaker test basically measures bushing losses, the PF test results should be compared to the results obtained on other bushing of the same breaker, or bushing results of other similar bushings. If PF results are high, then perform additional tests on the individual bushing by using the UST method and operate the breaker several times and retest. If the PF is still high, then check the interrupters and lift rod guide. The criteria used for grading interrupters is that if the PF is between 0% and 35% then the inter- rupters are good; if the PF is between 35% and 50%, the interrupters are marginal; and if the PF is above 50% the interrupters are wet, dirty, or just bad. The criteria for grading lift rods are that the watts loss obtained from the UST test for the lift rod should be 0.1 W or less.
SF6 Breakers
The analysis for the SF6 breakers is similar to that of the oil circuit breakers. However, it should be noted that the SF6 breakers have very low dielectric loss, and therefore they should be evaluated based on only watts loss and
capacitance measurements. All test results should be corrected to 20°C and compared with the data recorded for similar tests on the same breaker or breakers of similar model, make, manufacture and vintage.
Rotating Machines
The PF tests are made on rotating machines to detect contamination, such as moisture, dirt, dust, of the stator winding insulation and materials, and pres- ence of corona at operating voltages. The typical PF values are of the order of 1.0% or less for large modern machines. The typical PF tip-up values are usu- ally between 0.5% and 1.0%. A better way of evaluating field test results are to compare with previous year’s results or benchmark test results from the factory, or acceptance tests which were conducted when the equipment was commissioned and was relatively new.
Cables and Accessories
The overall PF value of a cable is a function of insulation type, length, size, installation (whether it is installed in metallic or nonmetallic conduit), and voltage. Therefore, the evaluation of the field PF test results should be based on comparison with previously recorded tests on the cable when it was put in service and was relatively new, such as during acceptance testing. Typical PF values at 20°C for cable insulation system as listed in the Doble Engineering Company reference book are:
It should be noted that cables without metallic sheath or grounded shield, but are installed in metallic raceway may have PF higher than the values listed above.
The results of the hot-collar tests on potheads are evaluated by comparing field test results on similar type of potheads, or previous years recorded test results. The evaluation is based on watts loss and current, and PF. Abnormally high dielectric loss and current indicate the presence of moisture. Below than normal test current indicates the absence of compound or oil in the pothead. Increase in watts loss with increased test voltage (PF tip-up) indicates the presence of corona.
Capacitors
The main capacitance of the PF and surge capacitors is quite large and may be tested at reduced voltages. The PF of these capacitors should be of the order of 0.5% or less, and the capacitance should compare with the nameplate information. The ground-wall insulation of a two bushing PF capacitor is in the order of 0.5%. The ground-wall insulation measurements and hot-collar test results should be compared with previously recorded data on these units.
Coupling capacitors are evaluated on the basis of capacitance (charging current) and PF which are compared with the nameplate data and those recorded with similar units. Typically the PF values are around 0.25% and units with PF value of 0.5% are recommended to be removed from service. An increase in the capacitance above 2% is a cause for concern and may indicate shorted elements. A decrease in capacitance may indicate an open circuit or high resistance connection between condenser foil layers.