DIAGNOSTICS

Instrument transformers are precise measuring instruments that have to satisfy the most stringent demands in terms of operating range and insula- tion

ABSTRACT Testing and Instrument transformers for medi- um- and high-voltage applications play a key role in energy supply. Act- ing as the link between the primary diagnostics of network and the metering or pro- tection equipment connected to the secondary side, a safe and reliable medium- and high- operation without failures is essen- tial. In relation to this background, this paper discusses testing and voltage instrument diagnostic methods for instrument transformers. Different application examples are presented that help to ensure an efficient production pro- transformers cess at the manufacturing site and efficient on-site testing for site- ac 1. Introduction prerequisite for good operational manage- ceptance, commissioning or regular ment and a reliable supply of energy. ITs are maintenance. The importance of the testing and diag- precise measuring instruments that have nostics of instrument transformers (trans- to satisfy the most stringent demands in KEYWORDS formers used for protection and measuring terms of operating range and insulation. purposes) is all too often underestimated. Incorrect operation, the use of defective calibration, commissioning, diagnos- Although the cost of instrument transfor- ITs, or the failure of an installed IT can have tics, instrument transformer, testing mers (ITs) is relatively low, their correct far-reaching financial, as well as technical, and uninterrupted operation is an essential consequences [1]. In addition to these op­

6464 TRANSFORMERS MAGAZINE | Volume 5, Issue 4 Michael FREIBURG

teristic behavior of current transform­ers is However, the primary resistance and their (extreme) quasi-short circuit opera­ leakage reactance cannot be ignored [3]. tion with a low-impedance burden. Voltage transformers are operated under quasi no-load conditions and are there­ In such a case, the ideally transformed fore connected to an extremely high-im- primary current will not be present in pedance burden. the secondary side of the current trans- former. Instead, it will be reduced by an Voltage drops occur on the winding amount proportional to the magnetizing resistors R1˝ and R2 due to the burden current (error compensation is possib- current ib and the excitation current le by adjust­ing the number of turns). As i0, causing the ideally transformed the operating burden increases, so does input voltage to differ from the output the induction in the core and hence the voltage – and resulting in an error. The magnetizing ­current, which, with the leakage flux (a quantity that is geometry aforementioned simplifications, is the dependent) also causes a lower voltage to only possible cause of the transformer be induced on the secondary terminals error. Changes to oper­ational parameters than the voltage that has been induced on in terms of current and burden therefore­ the primary side as corresponding flux affect the ratio and phase accuracy of to the primary voltage (Ψ =U1HV/(2πf)). current transformers. The associated additional “losses” are modeled in the equivalent circuit with The current measurement deviation is cal- a reactance that is split between the culated according to equation 1 (related to primary and secondary sides. Fig. 1). The phase error is the difference in angle between the primary and secondary The leakage and winding capacitances current vector and will be negative when shown in Figure 2 can be ignored in or- the secondary current vector is lagging [2]. der to simplify the description of the ratio accuracy at operating frequency. As the εI = (KnIb - I1P)/I1P ∙ 100 % (1) voltage rises, however, the primary capa- citance generates an additional capacitive The transformer equivalent circuit can current due to the design of the primary also be applied to voltage transformers. winding. This capacitive current can influ- erational aspects, manufacturers endeavor to produce high-quality equipment at a reasonable cost. Carrying out the correct Primary (HV) Secondary (LV) tests during the production process will not only save time, but also improve the quality 1p I I1 Ls2 R2 Ib of the final product [1]. n1 n2 I0 UZ2 2. Fundamentals U U1 L Ub Z 1HV H RFe Ucore b Current and voltage transformers are nor- mally single-phase transformers, which, despite their differences in design, can essentially be described using the familiar Ideal transformer transformer equivalent circuit diagram. K=n1/n2 2.1. Current and voltage transformers Figure 1. Simplified, qualitative equivalent circuit for current transformers, referred to sec- ondary side In the case of a current transformer, typi- cally the primary elements can be ignored when modeling the equivalent circuit. In Although the cost of ITs is relatively low, general, the secondary stray inductivity can their correct and uninterrupted operation be ignored without introducing any major calculation errors if the closed core design is an essential prerequisite for good opera- is symmetrical in terms of winding. At the tional management and a reliable supply of nominal frequency of 50/60 Hz, the leak­age capacitances can also be ignored. A charac- energy www.transformers-magazine.com 65 DIAGNOSTICS

2.2. Model-based transformer

Primary (HV) Cps Secondary (LV) testing

R1'' Ls1'' I1'' Ls2 R2 Ib n1 n2 Model-based accuracy tests have been de- I0 veloped [6, 7] based on the equivalent cir- UZ1 UZ2 cuits shown in section 2.1 and the opportu- U1HV U1'' C '' Cs Ub Z p LH RFe Ucore b nity they provide to calculate the accuracy of the transformer (for example, [3]). Car- rying out various measurements of the equivalent circuit parameters with small Ideal transformer K=n1/n 2 signals compared with the nominal quanti- ties allows the calculation of the non-linear dependent accuracy of current (CT) and Figure 2. Simplified, qualitative equivalent circuit for voltage transformers, referred to voltage transformers (VT), which depends secondary side on the voltage, current and burden, Fig. 3.

Using this approach, the measurements of the excitation properties take place, regard- Measurment of short- Calculation of less of the device being examined, from the circuit impedance from transformer error secondary side using the single-winding secondary side method (compare [6] for VTs). The volt­ (VT only) age drop across the secondary elements is taken into account when calculating the Ratio measurement induction. The input voltage is sinusoidal. K=n /n Measurement of 1 2 Any non-linear induction caused by the winding resistance on CT: Secondary injection VT: Primary injection non-linear excitation current is ignored. secondary side Model-based accuracy verification ena- bles all the relevant transformer parame- Excitation measurement Modeling of ters to be determined without applying a fromsecondary side missing parameters nominal current or nominal voltage. The model-based verification also permits all the parameters specified in the standard Figure 3. Block diagram for model-based accuracy verification of current and voltage to be calculated as shown in Table 1 (the transformers table may not be fully comprehensive).

3. Tests and diagnostics Model-based accuracy verification enables carried out by the all the relevant transformer parameters to manufacturer be determined without applying a nominal 3.1. Preliminary testing of the current or nominal voltage transformer

Preliminary testing (testing before the IT ence the transformer error (especially the difference in angle between the primary is completely manufactured) is unavoid­ phase-angle error) [4]. and secondary voltage and will be nega- able if an efficient and high-quality manu­ tive when the secondary voltage vector is facturing process is the aim. In particular, The voltage measurement deviation is lagging [5]. preliminary testing ensures that any er- calculated according to equation 2 (con- rors that occur during critical production nected to Fig. 2). The phase error is the εU = (Kn∙Ub - U1HV)/U1HV ∙100 % (2) phases are not carried over the whole manufacturing process into the final type test or routine test (compare [1]). The mo- Table 1. Overview of transformer parameters del-based approach enables adherence to the specifications to be checked at various Type of transformer Application Parameters stages of the production process. This test goes beyond the ratio testing that is nor- Metering εI, δI, FS CT mally carried out. The transformer can be Protection εI, δI, ALF, Kr, Ts, RCT, EK, Ie, εt, Ktd, Kssc tested using low-level test signals, even if Metering εU, δU VT they do not yet fulfill all the requirements Protection εU, δU regarding insulation.

66 TRANSFORMERS MAGAZINE | Volume 5, Issue 4 The following critical process steps can Testing at critical production phases ensures be identified, for example: selection of the right type of core material for the applica- that any errors that occur are not carried tion in question (testing after winding or over the whole manufacturing process stacking of the core); the winding process (testing after applying the primary and/or secondary winding); and casting or instal- lation in the housing (testing before final installation). If, for example, corrections can still be made before casting, then any Layout and transformer that has already been casted transformer design will have to be produced again if it fails to Layout and meet the specified (limit) values. transformer design Production of transformer core Figure 4 shows an example of the process Production of A involved in the manufacture of medium- transformer core Production of voltage current and voltage transformers. A secondary winding Application of B Testing of the magnetic properties can be secondary winding carried out at point A. As there is no win- Production of ding that can be used for the measurement D primary winding at this point, a temporary winding must Any necessary C be used. The connecting cables required to corrections Assembly of core carry out a four-wire measurement using E and coil the single-winding method are connected Installation in D via a “flexible coil” in the form of a cable. If housing / casting Any necessary the core dimensions are known, the mag- F mold and casting corrections netic parameters can be determined. The E relative permeability µr, the saturation in- Installation in ductance, and the losses are of particular housing / casting interest at this stage. F mold and casting An insulation test can be carried out at points B and C after producing the wind­ Figure 4. Critical production steps for medium-voltage current (left) and voltage (right) ings for a voltage transformer; this will transformers highlight any damage to the winding wire.

Once the core-and-coil assembly is com- -εu Δcos β plete (in the case of a voltage transformer) or the secondary winding applied (current transformer), a ratio measurement and the jIb(X''σ1 +Xσ2) first model-based testing can take place. It ΔSn n) will still be possible to do corrections at this ct i o stage if certain parameters are not met. The Ib(R''1+R2) orr e

c ε F ratio measurement or model-based test can then be repeated after the correction indin g jI X'' (E). After installation in the housing and w 0 σ1 ( U'' casting, repeating the previous tests will es- 1 tablish what (mechanical) effect the casting -δu 0 I0R''1 δu has had on the accuracy of the transformer. A dielectric test can also be carried out at δ this point to assess the quality of the potting and/or insulation, see Table 4. Ub

3.2. Model-based testing as a I''1 development tool Ib

I0 β Current and voltage transformers are usu- IR ally individually configured and/or ma- ρ IL Φ nufactured according to the customer’s specifications. The electrical calculation Figure 5. Voltage transformer phasor diagram [6] www.transformers-magazine.com 67 DIAGNOSTICS

Table 2. Results of ratio measurement with primary injection

Primary current Ip in A Secondary current Is in A Ratio K Phase displacement δ in °

100.00 0.05004 1998.4:1 0.22

200.00 0.10007 1998.6:1 0.21

400.00 0.2 2000.0:1 0.23

600.00 0.30011 1999.2:1 0.2

799.99 0.40028 1998.6:1 0.18

Table 3. Results of ratio measurement with primary injection (nominal operation and transient behavior)­ of the transformer employs a range ­of de- Ratio K Phase error δI in min Ratio error εI in % Polarity sign criteria [1, 3, 4] based on the equiv­alent circuit and the Möllinger and Gewecke­ 2045.5:1 0.543 -2.2627 OK phasor diagram, as exemplified for the volt­ age transformer shown in Figure 5.

The dependency between the fault vector F and the individual parameters in the equi- Installed bridge valent circuit can best be illustrated using a for primary switching phasor diagram. lectrical Ip connection As the accuracy of the transformer depends not just on the properties of the materials used for the windings and core but also on the geometry (leakage reactance Xσ), it can be affected by the choice of material, the de- Unintentional sign, and the production technology. The connection Transformer use of model-based testing provides an op- to housing housing portunity to compare the previously calcu- lated variables with the measured variables taken from the prototype (design verifica- Figure 6. Wiring diagram showing current transformer with unintended housing contact (red) tion), as well as to determine various other parameters, such as the distribution of the leakage inductance.

As well as the (distribution of the) leakage Laboratory On-Site inductance Lσ, the excitation curve (Ucore, I0) is also of interest, as this determines the volt­ Conventional Burden age-dependent no-load error and plays an calibration measurement important role regarding ferro-resonance.

B) 4. On-site testing of current Model-based Comparison Model-based accuracy verification accuracy verification and voltage transformers Nominal values Nominal values 4.1. Acceptance and commissioning tests Documentation Model-based of results Checking the transformer ratio and the and dierences B) accuracy verification Consideration Operational values polarity is part of the acceptance and com- of di erences missioning test and enables any transport damage or installation errors to be detec- ted. The testing of current transformers, for example, involves injecting a primary cur- Figure 7. Procedure for on-site calibration of current and voltage transformers using rent, which is measured on the secondary model-based calibration side. The following example demonstrates

68 TRANSFORMERS MAGAZINE | Volume 5, Issue 4 that this test is unable to pick up certain The accuracy of the transformer can be af- installation errors, with the result that fail­ ure may occur further down the line. Ad- fected by the choice of material, the design, ditional testing using the model-based ap- and the production technology proach as set out in section 2.2 is therefore recommended. Testing of the transformer ratio of a high-voltage current transformer (2000:1, class 0.2S as per IEC 61869, 5VA, The model-based testing determines The unintended connection in the above FS10) with a primary current injection the parameters and performs the ratio example has a sufficiently high impedance produces the results shown in Table 2. measurement­ from the secondary side to prevent it being noticed when the pri- using the voltage method (K=U1/U1HV, mary current is injected. With model- As the results show, the ratio K tends to be compare Fig. 1). This helps detect a prima- based testing, on the other hand, the short lower than the nominal value of 2000:1. ry short circuit, as in this example. In the circuit on the primary side makes the ra- Due to the low values, the ratio also de- case of the secondary voltage method, the tio larger, as the measured primary voltage pends on the accuracy of the current mea- core voltage Ucore=U1 is calculated using U1HV will be lower (K=U1/U1HV). surement Is. Looking at the results, it could the secondary voltage injected, and the be assumed that the test was “passed”, as induced voltage U1HV is measured on the 4.2. Calibrating current and voltage the result is close to the nominal value primary winding. transformers (it would be a good idea to check against reference values). Carrying out an additi- The differences in this example can be Current and voltage transformers are gen­ onal model-based test for nominal ope- attributed to a short circuit of the primary erally only calibrated as part of a routine ration with a primary current of 2000 A conductor. The housing on the head of testing procedure once they have been produces the results shown in Table 3. the current transformer is, under normal manufactured.­ There is nothing in any of circumstances, electrically connected at the international standards that requires The accuracy following the model-based one end to the primary conductor (same subsequent calibration on site or for the cali- test is not within the limits of +/- 0.2 % as potential), compare Fig. 6. In this example, bration to be repeated after the transformer specified in the IEC 61869 standard [5]. the housing is wrongly connected to the has been in operation for a specified period. The ratio value is also higher than that ob- primary conductor at both ends, resulting However, such a requirement is set out in tained when testing with a primary current. in a (partial) short circuit of the primary some local and internal company regula­ The polarity and phase error are acceptable. current. tions. On-site calibration of the transformer

Table 4. Sample overview of transformer diagnosis

Testing procedure Detectable faults/parameters

Shorted windings, poor connections, primary wiring (CT), open connections, incorrect Ratio measurement wiring of built-in devices (for example, in circuit breakers, transformer bushings) See above + influence of operational parameters (current, voltage, burden), loose Accuracy verification, conventional or model- laminations (in the case of HV transformers), displacement of primary winding (in based HV transformers), short circuit of primary conductor (CT, model-based only)

Transformer parameters (for example, ε, δ, ALF, FS, RCT, Lσ, ...) Overall model-based transformer testing Transient parameters (for example, Ts, Ktd, etc.) Loose laminations (in the case of HV transformers), operational inductance due Measurement of excitation curve to susceptibility to ferro-resonance, protection transformer parameters, short circuits or shorted windings, distortion of air gap Measurement of core properties Remanence (current transformers), permeability, wattage losses, saturation inductance

Polarity measurement Winding faults, connection errors

Partial discharge measurement Internal faults in insulation, cavities, cracks, floating potential Dissipation/power factor and capacitance Condition of insulation (conductivity and polarization losses), tap contact (CT) measurement Dielectric response measurement (FDS, Humidity (oil-paper insulated transformers), aging of insulation, contamination PDC) (µHz-kHz) Voltage withstand test Insulation capacity, short circuits Partial discharge activity, sparking (thermal), moisture, electrical resistance, Oil analysis aging condition of oil Insulation resistance measurement Condition of insulation www.transformers-magazine.com 69 DIAGNOSTICS

On-site calibration of the transformer is support, discussions and their co-author- ships in the conducted ­study, which was sometimes advisable, as the changed load the basis for this paper [10]. conditions or transport damage may have affected its accuracy Bibliography [1] CIGRE Study Committee A3: State of the Art of Instrument Transformers, 2009 is sometimes advisable, considering that in 5. Overview of transformer the period since the original calibration in diagnosis [2] IEC 61869-2:2012, Instrument trans- the lab the topology of the transformer (for formers: Additional requirements for example, in the case of separable current The testing and diagnostics methods listed current transformers transformers) or load conditions may have and commented on in Table 4 are those changed, transport damage may have occur- the author believes to be the most useful [3] R. Bauer, Die Messwandler [The ins- red, or other operational factors might have in respect of detectable faults and parame- trument transformer], Springer-Verlag, affected its accuracy (for example, aging of ters. The processes shown are based on the 1953 capacitive stack of CVTs). diagnosis of completed transformers. [4] E. Zinn, PTB Testing Instructions, Calibration in this context means check­ These testing methods enable an assess- Vol. 12: Instrument transformers, 1977 ing the accuracy (ratio error and ­phase ment of the transformer to be carried out error) using reference values; error on an application-by-application basis. [5] IEC 61869-3:2011, Instrument trans- thresholds must not be violated. A con- There is no all-embracing “correct proce- formers: Additional requirements for ventional calibration in an accredited dure” in terms of the application, but the inductive voltage transformers laboratory involves measuring the differ­ appropriate test for the desired purpose ence between a standard transformer and should be performed at a suitable point. [6] M. Freiburg, Ein Verfahren zum mo- the test object. Direct processes (for example, oil analy- dellbasierten genauigkeitsnachweis von sis) come with a significant risk and are Mittel- und Hochspannungswandlern As conventional calibration requires a not recommended for ITs, as often they für den Vor-Ort-Einsatz [A procedure current or voltage source in excess of the do not tell us very much (subsequent for the model-based accuracy verifica- nominal values, together with the physical measurements necessary) and necessi­tate tion of medium- and high-voltage trans- burden and a comparator for measuring manual intervention in the system [1]. formers for on-site use], Diss., TU Dort- the difference, on-site testing demands ex- mund, 2014 tensive technical and financial resources Conclusion (refer to [8] for more details). Model-based [7] F. Predl, Explore new paths with the testing as described in section 2.2 compen- This article examines and discusses the CT Analyzer, ITMF, 2013 sates for these drawbacks while providing importance of testing and diagnostic additional benefits in terms of time, safety, meth­ods applied to medium- and high- [8] A. Bergman, In situ Calibration of- and extra information, compare Table 1. voltage instrument transformers throug- Voltage Transformers on the Swedish- hout their complete life cycle, illustrating National Grid. Diss., 1994 To perform a traceable calibration, a range them with a number of examples. of steps must be carried out [9], as illus­ [9] PTB: Eichung teilbarer Stromwandler, trated in Figure 7. A conventional calib- Acknowledgement Technische Richtlinien E45, 2011 ration of the test object is carried out in an accredited laboratory. This is followed The author wants to thank Dr. Erik ­Sperling [10] M. Freiburg et al., Betriebserfahrun- by a model-based test without modifying from Pfiffner Instrument Transformers in gen aus Pruefung und Diagnostik von the transformer topology. The results and Switzerland and Dr. Michael Krüger from Mittel- und Hochspannungswandlern, any differences between the two measure- OMICRON electronics in Austria for their ETG, Germany, 2016 ments are documented.

Finally, a measurement of the operating burden and two model-based tests using Author the nominal burden and the measured op­ Dr. Michael Freiburg is responsible for instrument erating burden are then carried out on site. transformer tests and diagnostic equipment and is currently working as a product manager at OMICRON If the initial, direct comparison between electronics in Austria. Prior to that, he worked as a the conventional calibration with a stan- research and teaching assistant at the Technical University dard transformer and a comparator in the in Dortmund, Germany. His research interests include laboratory is not available, the accuracy the diagnostics of high-voltage equipment and material test can also be carried out (without trace- science. He received an engineering degree in 2010 and a ability) using the typical accuracies of the PhD degree in high-voltage engineering in 2014. model-based test.

70 TRANSFORMERS MAGAZINE | Volume 5, Issue 4