Transformer Condition Assessment
Power transformers are key components in the power train at hydroelectric powerplants and are appropriate for analysis under a risk assessment program. Transformer failure can have a significant economic impact due to long lead times in procurement, manufacturing, and installation, in addition to high equipment cost.
According to the Electric Power Research Institute (EPRI): “Extending the useful life of power transformers is the single most important strategy for increasing life of power transmission and distribution infrastructures starting with generator step-up transformers (GSU) at the powerplant itself” (EPRI Report No. 1001938).
Determining the existing condition of power transformers is an essential step in analyzing the risk of failure. This chapter provides a process for developing at a Transformer Condition Index. This condition index may be used as an input to the risk-and-economic analysis computer model where it adjusts transformer life expectancy curves. The output of the economic analysis is a set of alternative scenarios, including costs and benefits, intended for management decisions on replacement or rehabilitation.
The transformer condition assessment methodology outlined in this chapter applies to oil-filled power transformers (> 500 kilovoltampere [kVA]) currently in operation.
This guide is not intended to define transformer maintenance practices or describe transformer condition assessment inspections, tests, or measurements in detail. Utility maintenance policies and procedures must be consulted for such information.
3. Condition Indicators and Transformer
This guide describes four Condition Indicators generally regarded by hydro powerplant engineers as a sound basis for assessing transformer condition.
♦ Insulating oil analysis (dissolved gas analysis [DGA] and furan)
♦ Power factor and excitation current tests
♦ Operation and maintenance (O&M) history
These indicators are based on “Tier 1” inspections, tests, and measurements conducted by utility staff or contractors over the course of time. The indicators are expressed in numerical terms and are used to arrive at an overall Transformer Condition Index.
The guide also describes a “toolbox” of Tier 2 inspections, tests, and measurements that may be applied, depending on the specific problem being addressed. Results of Tier 2 may modify the score of the
Transformer Condition Index.
After review by a transformer expert, the Transformer Condition Index is suitable for use as an input to the risk and economic analysis model.
4. Inspections, Testing, and Measurements
The hierarchy of inspections, tests, and measurements is illustrated in figure A-1, Transformer Condition Assessment Methodology. Table A-14, Transformer Condition Assessment Summary (later in this appendix), summarizes these activities. Inspections, tests, and measurements (“tests”) performed to determine transformer condition are divided into two tiers (or levels). Tier 1 tests are those that are routinely accomplished as part of normal O&M or are readily discernible by examination of existing data. Results of Tier 1 tests are quantified below as Condition Indicator Scores that are weighted and summed to arrive at a Transformer Condition Index. Tier 1 tests may indicate abnormal conditions that can be resolved with standard corrective maintenance solutions. Tier 1 test results may also indicate the need for additional investigation, categorized as
Tier 2 tests. Tier 2 tests are considered nonroutine. Tier 2 test results may affect the Transformer Condition Index established using Tier 1 tests but may confirm also or disprove the need for more extensive maintenance, rehabilitation, or transformer replacement.
Inspection, testing, and measurement methods are specified in technical references specific to the electric utility. This guide assumes that Tier 1 and Tier 2 inspections, tests, and measurements are conducted and analyzed by staff suitably trained and experienced in transformer diagnostics. More basic tests may be performed by qualified staff who are competent in these routine procedures. More complex inspections and measurements may require a transformer diagnostics “expert.”
This guide also assumes that inspections, tests, and measurements are conducted on a frequency that provides accurate and current information needed by the assessment. In some cases, it may be necessary to conduct tests prior to this assessment to acquire current data.
Transformer condition assessment may cause concern that justifies more frequent monitoring. Utilities should consider the possibility of taking more frequent measurements (e.g., oil samples) or installing online monitoring systems (e.g., gas-in-oil) that will continuously track critical quantities. This will provide additional data for condition assessment and establish a certain amount of reassurance as transformer alternatives are being explored.
Figure A-1 – Transformer Condition Assessment Methodology
A severely negative result of ANY inspection, test, or measurement may be adequate, in itself, to require immediate de-energization or prevent re-energization of the transformer, regardless of the Transformer Condition Index score.
Transformer Condition Indicator scoring is somewhat subjective, relying on transformer condition experts. Relative terms such as “results normal” and “degradation” refer to results that are compared to:
♦ Industry accepted levels
♦ Baseline or previous (acceptable) levels on this equipment
♦ Equipment of similar design, construction, or age that operates in a similar environment
6. Weighting Factors
Weighting factors used in the condition assessment methodology recognize that some Condition Indicators affect the Transformer Condition Index to a greater or lesser degree than other indicators. These weighting factors were arrived at by consensus among transformer design and maintenance personnel with extensive experience.
7. Mitigating Factors
Every transformer is unique; therefore, the methodology described in this chapter cannot quantify all factors that affect individual transformer condition. It is important that the Transformer Condition Index be scrutinized by engineering experts. Mitigating factors specific to the utility may determine the final Transformer Condition Index and the final decision on transformer replacement or rehabilitation.
Substantiating documentation is essential to support findings of the assessment, particularly where a Tier 1 Condition Indicator score is less than 3 or a Tier 2 test results in subtractions from the Transformer Condition Index. Test results and reports, photographs, O&M records, or other documentation should accompany the Transformer Condition Assessment Summary Form.
9. Condition Assessment Methodology
The condition assessment methodology consists of analyzing each Condition Indicator individually to arrive at a Condition Indicator Score; then the score is weighted and summed with scores from other condition indicators. The sum is the Transformer Condition Index.
Apply the Condition Index to the Alternatives table to determine the recommended course of action. Reasonable efforts should be made to perform Tier 1 inspections, tests, and measurements. However, when data is missing, to properly score the Condition Indicator, it may be assumed that the score is “Good” or a mid-range number such as 2.
This strategy should be used judiciously to prevent deceptive results
10. Tier 1 – Inspections, Tests, Measurements
Tier 1 inspections, tests, and measurements routinely accomplished as part of normal O&M are readily discernible by examination of existing data. Tier 1 test results are quantified below as Condition Indicators that are weighted and summed to arrive at a Transformer Condition Index. Tier 1 inspections, tests, and measurements may indicate abnormal conditions that can be resolved with standard corrective maintenance solutions. Tier 1 test results may also indicate the need for additional investigation, categorized as Tier 2 tests.
10.1 Condition Indicator 1 – Insulating Oil Analysis
Dissolved gas analysis is the most important factor in determining the condition of a transformer because, being performed more frequently than other tests, it may be the first indication of a problem. Insulating oil analysis can identify internal arcing, bad electrical contacts, hot spots, partial discharge, or overheating of conductors, oil, tank, or cellulose. The “health” of the oil is reflective of the health of the transformer itself. DGA consists of sending transformer insulating oil samples to a commercial laboratory for analysis. The most important indicator is the individual and total dissolved combustible gas (TDCG) generation rates, based on International Electrotechnical Commission (IEC) and Institute of Electrical and Electronic Engineers (IEEE®) standards. Although gas generation rates are not the only indicator, they are reasonable for use in determining the Condition Indicator Score.
Furanic analysis may indicate a problem with the paper insulation, which could affect transformer longevity. A baseline furanic analysis should be made initially and repeated if the transformer is overheated, overloaded, aged, or after changing or processing the oil. Physical tests such as interfacial tension (IFT), acidity, moisture content, and dielectric strength usually indicate oil conditions that can be remedied through various reclamation processes. Therefore, they are not indicative of overall transformer condition that would lead to replacement. Such tests do not affect the Insulating Oil Condition Indicator score.
Results are analyzed and applied to table A-1 to arrive at a Condition Indicator Score.
Choose one furan number from table A-1, based on a transformer rated 65 °C or 55 °C rise above ambient temperature.
The below DGA numbers are based on dissolved gas in oil generation rates and come from a combination of IEEE C57-104™, IEC 60599, and Delta X Research’s Transformer Oil Analysis (TOA) software
Any ongoing acetylene (C2H2) generation indicates an active arcing fault, and the transformer may have to be removed from service to avoid possible catastrophic failure. A transformer may be safely operated with some C2H2 showing in the DGA. C2H2 sometimes comes from a one-time event such as a close-in lightning strike or through fault. However, if C2H2 is increasing more than 10 ppm per month, the transformer should be removed from service. Because acetylene generation is a critical indicator of transformer internal condition, each utility should establish practices in accordance with published standards and transformer experts to monitor any increases in gas generation and take corrective action. Increasing the frequency of DGA and degasifying the transformer oil are potential alternatives to consider.
10.2 Condition Indicator 2 – Power Factor and
Excitation Current Tests
Power factor insulation testing is important to determining the condition of the transformer because it can detect winding and bushing insulation integrity. Power factor and excitation current tests are conducted in the field on de-energized, isolated, and properly grounded transformers. Excitation current tests measure the singlephase voltage, current, and phase angle between them, typically on the high-voltage side with the terminals of the other winding left floating (with the exception of a grounded neutral). The measurements are performed at rated frequency and usually at test voltages up to 10 kilovolts (kV). The test detects shorted turns, poor tap changer contacts, and core problems.
Results are analyzed and applied to table A-2 to arrive at a Condition Indicator Score.
10.3 Condition Indicator 3 – Operation and Maintenance History
O&M history may indicate overall transformer condition.
O&M history factors that may apply are:
♦ Sustained overloading
♦ Unusual operating temperatures indicated by gauges and continuous monitoring
♦ Abnormal temperatures indicated by infrared scanning
♦ Nearby lightning strikes or through faults
♦ Abnormally high corona detected.
♦ Abnormally high external temperatures detected.
♦ Problems with auxiliary systems (fans, radiators, cooling water piping, pumps, motors, controls, nitrogen replenishment system, and indicating and protection devices).
♦ Deteriorated control and protection wiring and devices.
♦ Increase in corrective maintenance or difficulty in acquiring spare parts.
♦ Anomalies determined by physical inspection (external inspection or internal inspection not requiring untanking) (e.g., incorrectly positioned valves, plugged radiators, stuck temperature indicators and level gauges, noisy oil pumps or fans, oil leaks, connections to bushings).
♦ Previous failures on this equipment.
♦ Failures or problems on equipment of similar design, construction, or age that operate in a similar environment. Qualified personnel should make a subjective determination of scoring that encompasses as many O&M factors as possible under this indicator.
Results are analyzed and applied to table A-3 to arrive at a Condition Indicator Score.
10.4 Condition Indicator 4 – Age
Transformer age is an important factor to consider when identifying candidates for transformer replacement. Age is one indicator of remaining life and upgrade potential to current state-of-the-art materials. During the life of the transformer, the structural and insulating properties of materials used for structural support and electrical insulation, especially wood and paper, deteriorate.
Although actual service life varies widely depending on the manufacturer’s design, quality of assembly, materials used, operating history, current operating conditions, and maintenance history, the average expected life for an individual transformer in a large population of transformer is statistically about 40 years.
Apply the transformer age to table A-4 to arrive at the Condition Indicator Score.
11. Tier 1 – Transformer Condition Index
Enter the Condition Indicator Scores from the tables above into the Transformer Condition Assessment Summary Form at the end of this appendix. Multiply each Condition Indicator Score by the Weighting Factor, then sum the Total Scores to arrive at the Tier 1 Transformer Condition Index. This index may be adjusted by the Tier 2 inspections, tests, and measurements described below. Suggested alternatives for follow-up action, based on the Transformer Condition Index, are described in table A-14 at the end of this appendix.
12. Tier 2 – Inspections, Tests, Measurements Tier 2 inspections, tests, and measurements generally require specialized equipment or training, may be intrusive, or may require an extended outage to perform. Tier 2 assessment is considered nonroutine. Tier 2 inspections may affect the Transformer Condition Index number established using Tier 1 but also may confirm or refute the need for more extensive maintenance, rehabilitation, or transformer replacement.
For each Tier 2 inspection, test, or measurement performed, subtract the appropriate amount from the appropriate Tier 1 Condition Indicator and recalculate the Transformer Condition Index using the Transformer Condition Assessment Summary Form at the end of this document.
12.1 Test T2.1: Turns Ratio Test The transformer turns ratio (TTR) test detects shorts between turns of the same coil, which indicates insulation failure between the turns. These tests are performed with the transformer de-energized and may show the necessity for an internal inspection or removal from service. Results are analyzed and applied to table A-5 to arrive at a Transformer Condition Index Score adjustment.
12.2 Test T2.2: Short Circuit Impedance Tests
Sometimes called percent impedance or leakage reactance test, these tests are conducted in the field and compared to nameplate information, previous tests, and similar units to detect deformation of the core or windings caused by shipping damage, through faults, or ground faults. Some difference may be expected between nameplate and field test results because factory tests are conducted at full load current, which is normally not possible in the field. Field connections and test leads and jumpers also play a significant role in test results, and it is impossible to exactly duplicate the factory test setup.
Therefore, the I2R losses may be different and cause different test results. By comparing percent-reactance to nameplate impedance, the differences caused by leads and connections can be eliminated. Because reactance is only the inductive component of the impedance, I2R losses are omitted in the test results. Results are analyzed and applied to table A-6 to arrive at a Transformer Condition Index Score adjustment
12.3 Test T2.3: Core-to-Ground Resistance Megger® Tests
The transformer core is intentionally grounded through one connection. The core-to-ground resistance test can detect if this connection is loose. It can also detect whether there are other undesired and inadvertent, grounds. If the intentional core ground is intact, the resultant resistance should be very low. To check for unintentional core grounds, remove the intentional ground and Megger® between the core and the grounded transformer tank. This test should produce very high resistance, indicating that an unintentional ground is not present. This test is to supplement DGA that shows generation of hot metal gases (methane, ethane, and ethylene) and to indicate if a spurious, unintentional core ground is the problem. Experience can help locate the source of the problem. Results are analyzed and applied to table A-7 to arrive at a Transformer Condition Index score adjustment.
12.4 Test T2.4: Winding Direct-Current Resistance Measurement
Careful measurement of winding resistance can detect broken conductor strands, loose connections, and bad contacts in the tap changer (de-energized tap changer [DETC] or LTC). Results from these measurements may indicate the need for an internal inspection. This information supplements DGA and is useful when DGA shows generation of heat gases (ethane, ethylene, and methane). These tests are typically performed with a micro-ohmmeter and/or a Wheatstone bridge. Test results are compared between phases or with factory tests. When comparing with factory tests, a temperature correction must be employed (IEEE P62™). This test should be performed only after the rest of the routine electrical tests because it may magnetize the core, affecting results of the other tests.
Results are analyzed and applied to table A-8 to arrive at a Transformer Condition Index score adjustment. 12.5 Test T2.5: Ultrasonic and Sonic Fault Detection Measurements
These assessment tests (sometimes called acoustic testing tests) are helpful in locating internal faults. Partial discharges (corona) and lowenergy arcing/sparking emit energy in the range of 50 megahertz (ultrasonic), well above audible sound. To make these measurements, sensors are placed on the outside of a transformer tank to detect these ultrasonic emissions, which are then converted electronically to oscilloscope traces or audible frequencies and recorded. By triangulation, a general location of a fault (corona or arcing/sparking) may be determined so that an internal inspection can be focused in that location. These devices also can detect loose shields that build up static and discharge it to the grounded tank, poor connections on bushings, bad contacts on a tap changer that are arcing/sparking, core ground problems that cause sparking/arcing, and areas of weak insulation that generate corona. Sonic testing can detect increased core and coil noise (looseness) and vibration, failing bearings in oil pumps and fans, and nitrogen leaks in nitrogen blanketed transformers.
Information gained from these measurements supplements DGA and provides additional support information for de-energized tests, such as core ground and winding resistance tests. In addition, these tests help pinpoint areas to look for problems during internal inspections. Performing baseline tests may provide comparisons for later tests. Experience can help locate the source of the problem. Results are analyzed and applied to table A-9 to arrive at a Transformer Condition Index score adjustment.
12.6 Test T2.6: Vibration Analysis
Vibration can result from loose transformer core and coil segments, shield problems, loose parts, or bad bearings on oil cooling pumps or fans. Vibration analyzers are used to detect and measure the vibration.
Information gained from these tests supplements ultrasonic and sonic (acoustic) fault detection tests and DGA. Information from these tests may indicate maintenance is needed on pumps/fans mounted external to the tank. It may also show when an internal transformer inspection is necessary. If wedging has been displaced due to paper deterioration or through faults, vibration will increase markedly. This will also show if core and coil vibration has increased compared to baseline information. Experience can help locate the source of the problem.
Results are analyzed and applied to table A-10 to arrive at a Transformer Condition Index score adjustment.
12.7 Test T2.7: Frequency Response Analysis (FRA)
Frequency Response Analysis (or Sweep Frequency Response Analysis) can determine if windings of a transformer have moved or shifted. It can be completed as a factory test prior to shipment and repeated after the transformer is received onsite to determine if windings have been damaged or shifted during shipping. This test is also helpful if a protective relay has tripped or a through fault, short circuit, or ground fault has occurred
A sweep frequency test signal is generally placed on each of the highvoltage windings, and the signal is detected on the low-voltage windings. This provides a picture of the frequency transfer function of the windings. If the windings have been displaced or shifted, test results will differ markedly from prior tests. Test results are kept in transformer history files so they can be compared to later tests. Results are determined by comparison to baseline or previous measurements or by comparison to units of similar design and construction.
Results are analyzed and applied to table A-11 to arrive at a Transformer Condition Index Score adjustment.
12.8 Test T2.8: Internal Inspection
In some cases, it is necessary to open, partially or fully drain the oil, and perform an internal inspection to determine the transformer’s condition. These inspections must be performed by experienced staff with proper training. Sludging, loose wedges, loose coils, poor electrical connections on bushing bottoms, burned contacts on tap changers, localized overheating signified by carbon buildup, displaced wedging or insulation, and debris and other foreign material are general matters of concern. Photographs and mapping problem locations are good means of documenting findings. Before entering the transformer, and while inside it, the Occupational Safety and Health Administration, State, local, and utility safety practices must be followed (e.g., “permitted confined space” entry practices). Results are analyzed and applied to table A-12 to arrive at a Transformer Condition Index Score adjustment.
12.9 Test T2.9: Degree of Polymerization
Winding insulation (cellulose) deterioration can be quantified by analysis of the degree of polymerization (DP) of the insulating material. This test gives an indication of the remaining structural strength of the paper insulation and is an excellent indication of the remaining life of the paper and the transformer itself. This requires analyzing a sample of the paper insulation in a laboratory to determine the deterioration of the molecular bonds of the paper. Results are analyzed and applied to table A-13 to arrive at a Transformer Condition Index score adjustment.
13. Tier 2 – Transformer Condition Index Calculations
Enter the Tier 2 adjustments from the tables above into the Transformer Condition Assessment Summary Form at the end of this chapter. Subtract the sum of these adjustments from the Tier 1 Transformer Condition Index to arrive at the total Transformer Condition Index. Suggested alternatives for followup action, based on the Transformer Condition Index, are described in table A-14.
14. Transformer Alternatives
After review by a transformer expert, the Transformer Condition Index—either modified by Tier 2 tests or not—may be sufficient for decision-making regarding transformer alternatives. The index is also suitable for use in the risk-and-economic analysis model described elsewhere. Where it is desired to consider alternatives based solely on transformer condition, the Transformer Condition Index may be directly applied to table A-15.