Condition-based assessment of on-load tap changers

March 1st, 2019, Published in Articles: Energize

On-load tap changers (OLTCs) are a crucial element of utility networks, as they must operate in a precise fashion in order to maintain a constant voltage output. This must be achieved regardless of variation on input or load. OLTCs have been a weak link in many networks, as they deteriorate over time due to mechanical problems or contact wear from repeated operation. Erosion of the contacts over time is expected due to the nature of their function.

Coking of the contacts causes overheating, which can cause thermal runaway. Regular maintenance is necessary to ensure continued proper functioning. Oil testing has long been recognised as an important tool for detecting incipient faults in the main tanks of transformers and is being applied to load tap changers.

Some of the advantages of oil testing are:

  • Can usually be performed while equipment remains in service
  • Can detect a wide range of problems in the early stages
  • Can be used to ascertain a reasonable sense of the severity of the problem
  • Has been shown to be very cost effective

Oil testing is used to provide one of the most important early warning diagnostics for on-load tap changers and, because of its effectiveness, has made condition-based maintenance a reality. Dissolved gas analysis (DGA) has been applied successfully for many years to non-current switching oil-filled power equipment. The application of DGA to OLTCs has both similarities and differences to the use of DGA in other oil-filled power equipment. It is similar in that the same processes produce the same gases. However, in terms of gas production, OLTCs are far more complex than transformers. 

Table 1: Gas-producing process.

OLTCs may or may not produce all of the so-called “fault gases” in normal operation and the gases that are produced may or may not be lost through venting. The first applications of DGA to evaluate the condition of an OLTC were based on experiences with transformers. Threshold limits were developed for the gases produced by overheating, both individually and in combination. Many factors such as design, operation, ventilation, and online filtration affect gas levels. 

Consequently, this gas-threshold approach offered limited success, but proved the potential usefulness of fluid testing for OLTC condition assessment. Since gas data alone cannot provide sufficient information to fully assess OLTC condition, new approaches were required for OLTC evaluations. The search for an effective new approach led to the development of condition codes, which provide a condition assessment of the load path components. In addition to revealing useful information for the maintenance of insulating fluid, fluid assessment tests are used in conjunction with OLTC gas data to provide diagnostic information about the condition of OLTCs. Keeping the oil free of water, arc decomposition products and other contaminants is essential for proper operation of the OLTC.

Table 2: Indication of gases.

Particle profiling provides important information about the deterioration of materials that result in particle production. This includes information about in-service processes such as fluid degradation, contact deterioration and mechanical wear of moving parts and rust formation. Two of the most important fluid degradation processes to be evaluated are charring of the oil and coke formation.

Table 1 shows the gas-producing processes that occur in oil- filled electrical equipment. The gases that are produced by these processes are listed in Table 2. Recognition of these differences between normal and abnormal gassing patterns paved the way to diagnostic assessment of OLTCs. DGA and other non-invasive tests can be combined to assess the OLTC’s condition. Diagnostic programmes have been successfully developed to achieve the goal of condition-based maintenance with the consequent saving in costs.

Non-invasive tests for OLTCs

Dielectric breakdown voltage

The oil in an OLTC should maintain a minimum dielectric breakdown voltage. In recent years on-line filters have been used for compartments containing arcing contacts in oil to better maintain the dielectric breakdown strength of the insulating materials. This approach has been effective in pushing out maintenance cycles and reducing the rate of contact wear. The dielectric breakdown voltage is a function of the relative saturation of water in oil and the amount, size, and type (conductivity) of particles.

Water content

Excessive water reduces the dielectric breakdown strength of the oil and can accelerate the ageing of the contacts.

Neutralisation number

As oils and cellulosic materials age, they will deteriorate and form ageing by-products, including acids. Eventually the ageing by-products will begin to polymerise and form sludges. Increasing acidity can be used as a guide to the ageing rate of the oil. When high values are reached, the oil should be replaced or reclaimed. Acid by-products, particularly in the presence of water, are corrosive.

Total metals in oil

The metals test, consisting of both particulate metals and those dissolved in the oil, is an extremely meaningful test. It provides an indication of the amount of material that has been worn or sublimated from the moving and/or stationary contacts and is now present in the oil. It also provides a quantitative analysis as to composition of the metals found in the oil.

Particle count and qualitative analysis

The total number of particles by size groupings is used to detect abnormal quantities of by-products and wear materials. The ratio(s) of the size groupings provides information as to the extent that a detrimental condition has progressed.

DGA diagnostics

It is undisputed that DGA plays a primary diagnostic role. The interpretation protocol for applying DGA is empirical in nature. In the case of OLTCs it is generally accepted that fault gas interpretation will be most useful if it is model specific. Individual gas concentrations-based diagnostics for OLTCs are not useful as they are operation count-and breathing configuration-dependent. DGA ratios of fault gases are fairly independent of operation count. The various gas ratios are excellent diagnostic tools. 

Table 4: Diagnostic gas ratios.

Table 4 lists the gas ratios proposed by Weidmann-ACTI (F Jakob, K Jakob and S Jones).

The primary test for OLTC diagnostics and condition assessment is that of dissolved gas-in-oil, as this detects most of the problems. There are three main types of OLTCs:

  • Reactive with arcing contacts in oil
  • Resistive with arcing contacts in oil 
  • Arcing contacts in a vacuum bottle

There should be differences in the gassing behaviour between resistive and reactive types, as the shorter time of arc extinction of the resistive type (5 to 6 min after contact separation) should lower the concentrations of gases generated. However, the gassing behaviour of different models of OLTCs is so different that generic rules for reactive and resistive OLTCs are not adequate.

The primary diagnostic gases used to develop condition codes are methane, ethylene and acetylene. In addition, three ratios are used:

  • Ethylene/acetylene: distinguishes between thermal and electrical discharge activity in oil
  • Methane/acetylene: distinguishes between thermal and electrical discharge activity in oil and can also detect partial discharge activity as a predominant gassing pattern. For example, localised overheating of contacts or the reversing switch will generally show increasing combustible gas generation with ratios of gassing going from an arcing pattern to characteristics of high temperature overheating of oil. Excessive arcing between contacts is most likely to develop high gas concentrations until the later stages when heating occurs (causing the combustible gas ratios to change). Examples of causes of overheating include:
    • Excessive contact resistance due to the formation of organic films and carbon deposits result in metal fatigue causing poor contact pressure.
    • Loss of direct contact surface area from misalignment or loss of contact material.

Excessive combustible gas build-up can result when the vent becomes plugged. This eventually leads to low oxygen content as it is consumed in oxidation reactions and is not replenished. Typically, the ratios will remain normal unless another problem is present at the same time. Various other thermal and electrical problems can also be detected depending upon the model of OLTC.

New DGA interpretation scheme

In an attempt to improve the understanding of the significance of DGA results, the “Kohonen net” cluster analysis technique has been relatively successful in grouping together DGA results of an apparently similar significance and identifying when a result appears to move to a different state. Although such techniques can be argued to provide a better basis for identifying an unusual condition than the purely statistical approach, they still rely very much on the human expert to ascribe some significance to the clusters identified.

The new approach proposed makes use of the relative proportions of the six combustible gases H2, CH4, C2H4, C2H6 and C2H2, which are displayed as a bar chart to illustrate the gas signature. The novel aspect of the approach proposed here is that this method is used to investigate and illustrate the clear difference that exists between “normal” and “abnormal” results. The graphical plotting of these ratios gives a clear indication of a heating problem (thermal runaway) when compared to the normal arcing process.

Fig. 1: Relative proportions of six combustible gases under normal and abnormal conditions.

The Duval Triangle for the interpretation of DGA in OLTCs

The Duval Triangle using CH4, C2H4 and C2H2 gases in a single graphical representation has been developed for tap changers, the basic Duval Triangle can be used to detect faults in compartment-type OLTCs.

Fig. 2: The Duval Triange for the interpretation of DGA in OLTCs.

When DGA points move with time from the normal to a fault zone, this means a fault is appearing. What makes the use of DGA in OLTCs different from its use in transformers is that during the normal switching operation of OLTCs, arcs in oil or hot spots in various OLTC components are produced, generating gases such as C2H2 and C2H4 which may interfere with the detection of abnormal faults. The “normal” gas formation of OLTCs must therefore be identified first as precisely as possible, in order to use DGA for the detection of faults in OLTCs. Triangle 2 has been developed to distinguish between normal and abnormal gas formation in OLTCs of the compartment and in-tank resistive-type OLTCs that, without any fault, generate large amounts of heating gas, especially ethylene.

Case studies

The following three case studies illustrate how DGA was used to successfully predicted imminent failure of the OLTCs, which were then removed from service and repaired before catastrophic failure occurred.

Fig. 3: Mozal smelter, Mozambique case.

Fig. 4: Durban electricity.

Fig. 5: Sasol synfuels generator step transformer.

Confirmation and complementary tests

There are several non-oil tests that have been employed to confirm or identify OLTC problems. These include the following:

Infrared thermography and temperature differential

A frequently-used method to detect or confirm overheating of contacts or the reversing switch in OLTCs is to determine the temperature difference between the main tank and the OLTC. Normally, the main tank should be operating at a higher temperature than the OLTC compartment except for the occasional transient, such as when the pumps initially come on to cool the main insulation. As thermal problems develop in the OLTC compartment, the oil temperature will consistently be higher than the main tank. This difference in temperature can be detected using continuous temperature monitors mounted to the tank wall or by periodic inspections using infrared thermography.

Electrical tests

There are several electrical tests that can be used to confirm or help identify the source of OLTC problems before entering for visual inspection.

  • Exciting current tests on all OLTC tap positions can be used to detect shorted turns and core problems in the preventive autotransformer, contact problems and connection problems in the preventive autotransformer or in taps.
  • Turns ratio can detect shorted turns in the preventive autotransformer.
  • Power factor tests are used to detect insulation deterioration such as from moisture and partial discharge activity, including tracking and carbonisation of solid insulation structures.
  • Contact resistance is used to detect excessive contact wear, poor contact pressure, and coking and polymeric films on contact surfaces.
  • Sweep Frequency Response Analysis (SFRA) and leakage reactance (short circuit impedance) are both used to detect winding movement or deformation and contact problems.
  • Tap changer diverter resistor and contact analysis provides a means for condition assessment of tap changers using dynamic contact resistance measurement techniques. A signal is injected into the phase under test with the secondary and tertiary windings shorted. The tap changer is tapped through its sequence and a trace is produced of the complete sequence. The detail of each tap can be viewed, allowing the timing of the transition to be measured. The test results provide the necessary information to establish the state of the tap changer to make planned decisions or take corrective actions minimising the need for intrusive inspections.

Acoustic and vibration analysis

Some investigators have developed a database of OLTC signatures using acoustic analysis to complement diagnostic programs for OLTCs.

The advantages of the tap changer analysis programme are:

  • Units that require maintenance are identified.
  • Units that do not currently require maintenance are identified.
  • Maintenance activities can be better focused.
  • Operators and planners can assess loading capabilities.
  • Failures and collateral damage to transformers can be reduced. 
  • Reliability is enhanced.
  • Costs are reduced.
  • Safety is not compromised.

Percent savings are based on a comparison of the percentage of units requiring maintenance as given by the condition code versus the number that would be maintained on a fixed-time interval.

Table 5: Percentage savings in maintenance.

Conclusion

The ability of utilities and other power transformer operators to convert from time-based maintenance to condition-based maintenance for on-load tap changers has been greatly enhanced by the advancement of oil diagnostic testing. These tests have gained acceptance in the industry and will be standard practice in the near future. We have found that once oil diagnostics programmes are started for OLTCs, problems that were not detected by other methods are revealed, and there is acceptance of this approach. After using oil diagnostics for all the OLTCs in a system, the number of problems found the second time around is reduced, showing the effectiveness of the programme.

References

[1] P Griffin and L Lewand, Doble Engineering; R Peck and N Letendre, Connecticut Light and Power; D Frieze and J Johnson, Empire District Electric Company: “Load Tap Changer Diagnostics Using Oil Tests – a Key to Condition-based Maintenance”.

[2] R Youngblood, C Baker, F Jakob and N Perjanik: “Application of Dissolved Gas Analysis to Load Tap Changers”.

[3] W Rose: “Using Dissolved Gas in Oil Analysis to Determine Need for Internal Maintenance of Load Tap Changers”.

[4] R Youngblood, F Jakob, K Jakob, and S Jones: “Use of Gas Concentrations to Interpret LTC Dissolved Gas Data”.

[5] Mark Sampson: “Oil Analysis of Load Tap Changers”.

[6] R Youngblood, F Jakob, K Jakob, and S Jones: “A Nomogram for Interpretation of LTC-DGA Data”.

[7] F Jakob, K Jakob, and S Jones: “LTC Condition Assessment by Diagnostic Testing”.

[8] RG Asche: “Dissolved Gas Analysis of Load Tap Changers, Subcommittee Project Report”, Proceedings of the 2002 Annual International Conference of Doble Clients, Transformer Test and Maintenance, 2002.

[9] E Finnan: “Case Studies”, Proceedings of the 2001 Annual International Conference of Doble Clients, 2001.

[10] HU Schellhase, EA Hall and EC Korolenko: “A Critical Assessment of Fouling and Coking of Load Tap Changers”, Proceedings of the 2003 Annual International Conference of Doble Clients, 2003.

[11] R Youngblood, F  Jakob and T Haupert: “Application of DGA to Detection of Hot Spots in Load Tap-Changers”, Proceedings of the 1993 Annual International Conference of Doble Clients, 1993.

[12] R Youngblood: “An Update on Load Tap Changer Hot Spot Detection Through the Use of DGA”, Proceedings of the 1994 Annual International Conference of Doble Clients, 1994.

[13] D Pomi: “Discoveries of LTC Problems which Support the Success of New Technology Monitoring and Test Devices”, Proceedings of the 65th Annual International Conference of Doble Clients, 1998.

[14] LR Lewand and PJ Griffin: “Transformer Case Studies”, Proceedings of the 66th Annual International Conference of Doble Clients, 1999.

[15] FM Deskins, SP Patel and TR Bode: “Use of Infrared Thermal Image Tests and Dissolved Gas in Oil Analysis to Detect Abnormal Conditions in Transformer Load Tap Changers”, Proceedings of the 1995 Annual International Conference of Doble Clients, 1995.

Contact Wear Check, Tel 031 700-5460, support@wearcheck.co.za

Related Articles

  • To supply and service offshore wind turbines
  • To bridge the dispatchable energy gap
  • Certified to class C2 capacitive switching
  • Inverters comply with RFG code
  • To automate thermal image processing