71
UPTEC E 16003 Examensarbete 30 hp Mars 2016 Power Transformer Monitoring and Diagnosis using Transformer Explorer Svante Karlsson

Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Embed Size (px)

Citation preview

Page 1: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

UPTEC E 16003

Examensarbete 30 hpMars 2016

Power Transformer Monitoring and Diagnosis using Transformer Explorer

Svante Karlsson

Page 2: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Teknisk- naturvetenskaplig fakultet UTH-enheten Besöksadress: Ångströmlaboratoriet Lägerhyddsvägen 1 Hus 4, Plan 0 Postadress: Box 536 751 21 Uppsala Telefon: 018 – 471 30 03 Telefax: 018 – 471 30 00 Hemsida: http://www.teknat.uu.se/student

Abstract

Power Transformer Monitoring and Diagnosis usingTransformer Explorer

Svante Karlsson

Power transformers are one of the most expensive and vital components in thepower system. A sudden failure could be a very costly process for both thetransformer owner and the society. Several monitoring and diagnostic techniqueshave been developed over the last decades to detect incipient transformer problemsat an early stage, so that planned outages for maintenance and reparation can becarried out in time. However, the majority of these methods are only secondaryindicators which do not address the transformers fundamental function: to transferelectric energy between different voltage levels with turn ratio, short-circuitimpedance and power loss within acceptable limits.Transformer Explorer is a concept developed by ABB which utilizes ordinary currentand voltage signals available in the substation to extract transformer fundamentalparameters such as: turn ratio, magnetizing current, impedance and power loss, whichhas significant diagnostic value. By estimating these parameters the method should beable to detect a number of problems related to the windings and the magnetic circuitof the transformer. Transformer Explorer is expected to find it's application in twodifferent versions, either as an permanent on-line monitoring and diagnostic tool or asa short-time version for temporary measurements.The thesis could be divided into three main parts. The first one focusing on aquantitative study trying to answer questions regarding the concepts feasibility whenthe temporary version is used. The second part is about optimizing and improving theprocedure by which the fundamental parameters are estimated. In the last part, a newmethod for reducing the impact of errors introduced by the acquisition system on theestimated power loss is proposed. All the investigations related to the three topicscovered in this thesis showed interesting and promising results.

UPTEC E 16003Examinator: Mikael BergkvistÄmnesgranskare: Karin ThomasHandledare: Nilanga Abeywickrama

Page 3: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Sammanfattning

Krafttransformatorer tillhor de dyraste och mest vitala komponenterna i kraftsy-stemet. Ett oforutsett haveri innebar inte sallan en lang och kostsam process foragaren, da enheterna ofta tillverkas pa bestallning. For att forhindra detta ar detviktigt att upptacka problem i ett tidigt stadie, vilket traditionellt gors med olika ty-per av overvakningssystem som installeras pa transformatorn. Systemen overvakarsekundara indikatorer som t.ex. temperatur eller gasinnehall i transformatoroljan.Nar nivaerna avviker fran det normala indikeras detta med en varningssignal fransystemet och mer avancerade diagnostikmetoder kan sattas in for att hitta grun-dorsaken till problemet. Detta gors vanligen med transformatorn bortkopplad frannatet. Ett problem med overvakningssystemen ar just det faktum att de overvakarsekundara indikatorer och inte de parametrar som ar viktiga for transformatornsgrundlaggande uppgift: att overfora effekt fran en viss spanningsniva till en annan,med effektforlust och kortslutningsimpedans inom rimliga varden.

Pa ABB Corporate Research har ett nytt koncept for overvakning och dia-gnostik av transformatorns tillstand utvecklats: Transformer Explorer. Konceptetanvander befintliga givare sa som strom- och spanningstransformatorer for att upp-skatta varden pa kvantiteterna: lindningsomsattning, kortslutningsimpedans ocheffektforlust. Dessa ar alla fundamentala for transformatorns funktion och koncep-tet mojligor matning utan att koppla bort transformatorn fran natet. De forsta tvauppskattas genom linjaranpassning av data till transformatorns kretsekvivalent.Konceptet ar tankt att anvandas i tva varianter: en temporar korttidsmatning darkvantiteterna uppskattas utifran all insamlad data under perioden (enstaka dagar,upp till en vecka), eller som en kontinuerlig overvakningsmetod.

Detta examensarbete kan huvudsakligen delas in i tre delar. Den forsta behand-lar fallet nar Transformer Explorer anvands for korttidsmatningar. En forsta frage-stallning ar hur kort en korttidsmatning kan vara. En kortare matning innebar enviss forlust av information och foljaktligen storre osakerhet. Ytterligare en aspektsom behover undersokas ar om den teoretiska osakerheten i en matning overens-stammer med den verkliga. For att undersoka tidsintervall och osakerhet har enstorre studie gjorts med hjalp av befintligt lagrad data fran tre transformatorer,dar Transformer Explorer har anvants for att simulera matningar av tva olikalangdintervall: 2 dagar samt 8 dagar. I denna delen har aven ett forslag till enmatrapport med metoden tagits fram.

Den andra delen behandlar fallet kontinuerlig overvakning. Har har vissa av dekriterier som styr nar Transformer Explorer ska utfora sin analys studerats samtden matematiska tekniken bakom uppskattningen av parametrarna.I den tredje delen har en ny metod utarbetats for att kompensera for de systema-tiska fel som matsystemet introducerar pa uppskattad effektforlust.Simuleringarna visade att den kortare tvadagars perioden bor vara fullt tillrackligfor temporara matningar, samt att den teoretiska uppskattningen av osakerhetenkan anvandas med en viss sakerhetsmarginal. Tva forslag pa nya metoder for kon-tinuerlig matning togs fram: en som relaterar till kriterierna och en ny algoritm foratt utfora passningarna, vilken ar mindre kanslig mot storningar i data. Metodenfor forbattrad uppskattning av effektforlust visade sig fungera tillfredstallande.

Page 4: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Contents

1 Introduction 31.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31.2 Aim of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41.3 Outline of Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2 The Power Transformer 52.1 Transformer and the Power System . . . . . . . . . . . . . . . . . . 52.2 Transformer Design . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.2.1 Ferromagnetic Core and Windings . . . . . . . . . . . . . . . 62.2.2 Insulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.2.3 Tap Changer . . . . . . . . . . . . . . . . . . . . . . . . . . 82.2.4 Bushings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.2.5 Cooling System . . . . . . . . . . . . . . . . . . . . . . . . . 82.2.6 Instrument Transformers . . . . . . . . . . . . . . . . . . . . 8

2.3 Transformer Physical Fundamentals . . . . . . . . . . . . . . . . . . 92.3.1 Ideal Transformer . . . . . . . . . . . . . . . . . . . . . . . . 92.3.2 Practical Transformer . . . . . . . . . . . . . . . . . . . . . 10

2.4 Transformer Failures . . . . . . . . . . . . . . . . . . . . . . . . . . 132.4.1 Winding Faults . . . . . . . . . . . . . . . . . . . . . . . . . 142.4.2 Core Related Faults . . . . . . . . . . . . . . . . . . . . . . . 152.4.3 Tap Changer faults . . . . . . . . . . . . . . . . . . . . . . . 15

2.5 Monitoring and Diagnostics Methods . . . . . . . . . . . . . . . . . 15

3 Transformer Explorer 173.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173.2 System Description . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3.2.1 Instrument Transformers . . . . . . . . . . . . . . . . . . . . 183.2.2 Acquisition System . . . . . . . . . . . . . . . . . . . . . . . 183.2.3 Phasor Calculation . . . . . . . . . . . . . . . . . . . . . . . 183.2.4 Vector Group Transformation and Zero-Sequence Removal . 183.2.5 Extraction of Quantities . . . . . . . . . . . . . . . . . . . . 19

3.3 Transformer Explorer Software . . . . . . . . . . . . . . . . . . . . . 203.3.1 LabVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203.3.2 Program Structure . . . . . . . . . . . . . . . . . . . . . . . 213.3.3 Data Acquisition . . . . . . . . . . . . . . . . . . . . . . . . 223.3.4 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 223.3.5 Display Options . . . . . . . . . . . . . . . . . . . . . . . . . 243.3.6 Logging of Data . . . . . . . . . . . . . . . . . . . . . . . . . 27

3.4 Field Installations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283.5 Fault Detection Ability . . . . . . . . . . . . . . . . . . . . . . . . . 29

1

Page 5: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

3.6 Transformer Explorer Applications . . . . . . . . . . . . . . . . . . 303.7 Uncertainty and Accuracy . . . . . . . . . . . . . . . . . . . . . . . 30

3.7.1 Derivation of Uncertainty in a Linear Fit . . . . . . . . . . . 303.7.2 Uncertainty in the Estimated Parameters . . . . . . . . . . . 31

4 Short Time Measurements 324.1 Day and Week Comparison . . . . . . . . . . . . . . . . . . . . . . . 32

4.1.1 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324.1.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

4.2 Data rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484.3 Measurement Report with Vattenfall . . . . . . . . . . . . . . . . . 49

4.3.1 The Report . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

5 Analysis Improvements 525.1 Control of Data Distribution in Fit . . . . . . . . . . . . . . . . . . 525.2 An Alternative Fitting Algorithm . . . . . . . . . . . . . . . . . . . 59

6 Power Loss 646.1 Application Examples on Field Installation Data . . . . . . . . . . . 65

7 Conclusions & Discussion 667.1 Short Time Measurement . . . . . . . . . . . . . . . . . . . . . . . . 667.2 Analysis Improvements . . . . . . . . . . . . . . . . . . . . . . . . . 667.3 Power Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677.4 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

2

Page 6: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

1 Introduction

Power transformers are one of the most important components in the electric powersystem. By making it possible to interconnect different voltage levels, they permita economical way to transfer energy over long distances. Power transformers arealso among the most expensive components in the network and they are often notmanufactured until ordered [1]. A sudden failure which means that the transformerneeds to be taken out of service, is often associated with considerable costs forthe owner. To avoid such failures, it is important to detect problems in an earlystage so that a planned outage with necessary repairs and maintenance can be car-ried out in time. A number of monitoring techniques has been developed duringthe past few decades for this purpose. The majority of these methods are only sec-ondary indicators such as temperature measurements and gas-in-oil analysis. Whena problem is indicated by the monitoring system, the transformer is disconnectedand more advanced diagnostic methods can be adopted to find the actual problem.Very few of the conventional monitoring methods addresses the fundamental func-tion of the transformer, i.e., to convert electric energy between two voltage levelswith power losses within acceptable limits and enough impedance to limit effectsof short-circuits. Two important parameters that ensures this function are turnratio and impedance; these are traditionally only measurable with the transformeroff-line.

1.1 Background

At ABB a new monitoring and diagnostic tool [1]: Transformer Explorer, is beingdeveloped. The concept uses voltage and current signals obtained from the instru-ment transformers in the substation to estimate some of the transformers mostfundamental parameters:

Turn ratio

Magnetizing current

Short-circuit impedance

Power loss

all which are strictly related to the transformers fundamental function, and thushave a significant diagnostic value. By applying the measured quantities to a simpletransformer circuit model it is possible to estimate the three first of the aforemen-tioned parameters by means of linear fitting. The power loss is simply given by thedifference in power going in and out of the transformer. The central part of theconcept is a software with the same name, which is used for both signal acquisition,

3

Page 7: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

analysis and visualization. The current version of the program is written in thegraphical programming environment LabVIEW.

The concept is expected to find use in two different applications: either as apermanent installation where the parameters are continuously monitored, or usinga temporary installation where the parameters are estimated based on data collectedduring the measurement period.

Transformer Explorer has many advantages compared to conventional monitoringand off-line diagnostics methods. Some of the most important are that no expensiveequipment is needed apart from the all-ready installed instrument transformers andno outage is required for the installation or during the measurement, which is thecase when measuring both turn ratio and short-circuit impedance with conventionalmethods. Further the power loss can be considered as a general problem indicatorand is much faster than secondary indicators, i.e., temperature [1].

1.2 Aim of Thesis

The aim of the thesis is to evaluate the concepts feasibility for short-time mea-surement and how the result from such a measurement should be presented in areport. A comprehensive study will be conducted using field data from transformersin operation. Other topics related to the Transformer Explorer concept that willbe studied are:

The procedure for estimating the fundamental parameters in the analysis.This includes the mathematical method for making a linear fit as well as thecriteria controlling when a fit should be made.

Systematic errors affecting the accuracy of measured power loss.

1.3 Outline of Thesis

Some general theory of power transformers is presented in chapter 2. This in-cludes both physical and design fundamentals as well as monitoring and diagnos-tics aspects. Chapter 3 comprehensively introduces the reader to the TransformerExplorer concept. In chapter 4-6, the main work is presented with method andobtained results. Finally the conclusions from the results are presented in chapter7 along with discussion and future work.

4

Page 8: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

2 The Power Transformer

To make this thesis understandable for a reader with limited knowledge in the field ofpower transformers – this chapter is intended to give some necessary fundamentals.

2.1 Transformer and the Power System

The transmission losses in a power system is strictly related to the voltage level [2].With a high voltage, the current can be held low for a given power level, and sincethe losses is proportional to the square of the current, the losses can theoreticallybe made as small as desired by raising the voltage level. At the end of the ninthcentury, electric power could only be transferred relativity short distances becauseof the high losses in the transmission lines. After the invention of the commerciallypractical power transformer in 1885 the transmission distances and the utility ofelectric power increased drastically [3].

The power network is mainly divided in two sub systems: the transmission systemand the distribution system [4]. The transmission system takes care of the energytransfer over large distances and the distribution grid distributes the power to cus-tomers at appropriate voltage levels. While the transformer finds it’s given key rolein both these systems, this work will mainly focus on the power transformer whichis used for transmission. Fig. 2.1 shows an example of a large power transformer.

Figure 2.1: A large power transformer in the field (ABB).

5

Page 9: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

2.2 Transformer Design

2.2.1 Ferromagnetic Core and Windings

The main active parts of a power transformer are the core and winding. The core isusually build up by thin sheets of high permeability silicon steel which are stackedtogether [4]. This is done to prevent eddy current losses.

It is the core that take care of the energy transfer between the two windings byproviding a low reluctance path for the magnetic flux. The high permeability en-sures a low leakage flux and magnetizing current. The construction of the core couldbe divided in two main categories: core type and shell type construction, where thecore type is the far most common for power transformers. All the transformersstudied in this work are of core type. A picture illustrating the salient differencebetween the two types is shown in Fig. 2.2.

Figure 2.2: A sketch showing the two core configurations: core type and shell type [4].

The other fundamental part of the transformer is the windings. They are locatedconcentrically around the core. The conductors of the transformer winding are usu-ally made of high conductivity copper. It has excellent mechanical properties andthe highest conductivity of all the commercial metals [5]. In order to optimize theusage of space the conductors are commonly rectangular shaped. The conductorsare insulated with either varnish or cellulose base paper. The transformer windingsare typically constructed in four different ways: layer, helical, disk and foil windings[4]. All these four types are similar in the aspect that they are cylindrical and havea rectangular cross section. It is mainly the number of turns and the amount ofcurrent that determines which type of winding that is preferred when constructinga transformer. For large power transformers however, the ability to withstand shortcircuit is considered of higher importance than the losses.

2.2.2 Insulation

Since various parts of the transformer operate at different voltage levels, sufficientinsulation is needed between the parts to avoid electrical breakdown. For example,the core is usually grounded while parts of the high voltage winding operates at fullvoltage.

The insulation in a power transformer could be divided in two parts: Solid insu-lation and Oil insulation. The solid insulation typically consists of paper and pressboard and is used for both electric and mechanical insulation between the windingsand core. Paper and press board are based on cellulose that consists of molecules in

6

Page 10: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 2.3: The left picture shows a winding arrangement with a helical winding closestto the core surrounded by a disk winding. The outer is a layer winding for voltageregulation. The right picture shows example of barrier arrangement between windings[4].

long chains. The cellulose materials undergo degradation over time by breaking upinto smaller molecule chains. The degradation drastically decrease the mechanicalproperties of the insulation and it becomes brittle. There is a standardized methodfor measuring the degradation process of cellulose: degree of polymerization, i.e.,the DP number. A newly fabricated cellulose fiber has typically a DP number ofbetween 1200 and 1400, whereas the value can sink as low as 200 for an old trans-former.When this happens, the mechanical strength has decreased to such an extent,that the insulation’s ability to withstand the forces during an external short circuitno longer may be sufficient. This could of course lead to metal-to-metal contact inthe winding with an internal short circuit as an result. The degradation process isa growing process that is highly affected by the environment in the transformer [4].It could be good to mention that the dielectric properties of the insulation staysalmost unaffected during the whole aging process, so without external events thetransformer may operate satisfactory even with a highly reduced DP number [4].

The transformer oil fills two functions in the transformer: cooling and insula-tion between different parts. The transformer oil itself has a rather high dielectricbreakdown voltage (20-50 kV/mm vs 3 kV/mm for air), but gets excellent dielec-tric properties when combined with paper. Between the windings and windings tocore the distances with only oil can be comparably large and the field stress can behigh enough to cause breakdown. This is solved by adding barrier systems of pressboard sheets as additional insulation. The press board has good dielectric prop-erties and also provides stability to the winding structure. Two different windingarrangements with press board sheets can be seen in Fig. 2.3.

7

Page 11: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

2.2.3 Tap Changer

Since the losses in the grid is load dependent and varies with time, compensationneeds to be done in order to sustain a stable voltage level. One way to achievethis compensation is by letting the transformers operate at different turns ratio.Therefore, most of the power transformers are equipped with tap changers.

Tap changers can be divided in two categories: Off-load and On-load tap chang-ers [4]. The off-circuit type has a more simple design and is only operated whenthe transformer is off-line. The On-load tap changer however is a sophisticatedmechanical device, which is able to change between adjacent winding blocks (1-2%)without interrupting the current flow. An On-Line tap changer should be able toperform a very large number of operations during it’s lifetime.

2.2.4 Bushings

The transformer windings needs to be connected to the power grid through theearthed tank. This is achieved using different types of bushings. The bushingsfunction is to provide isolation between the conductor and the transformer tankwhich is normally at ground potential. As the voltage could be very high, theconnection between bushing and tank is exposed to high electric stress. Sincethe insulation medium outside the tank normally is air, the bushing also needsto provide sufficient insulation distance between the incoming conductor and thetank. There are two types of bushings: solid and capacitance graded, where thelatter one is used for higher voltage ratings.

2.2.5 Cooling System

The real losses in a power transformer is typically less than one percent of ratedpower [3]. These losses leads to temperature rise, and thus the transformer has tobe cooled to prevent overheating. For small transformers the ambient air is usuallyenough, but for a power transformer a cooling system is needed. The cooling systemcollects hot oil at the top of the transformer, the oil is then cooled in a external heatexchanger and supplied at the bottom of the tank [4]. The oil circulation could beeither natural (N) or forced (F) by pumps. The oil could also be cooled in severalways, either by the ambient air, by external fans or water cooling. Their are IECstandards that divides the cooling types into different categories [4].

2.2.6 Instrument Transformers

Since the presence of faults in the electrical system and in the transformer itself cannot be fully eliminated, one has to make the consequences of an eventual fault assmall as possible by use of protection equipment. This equipment typically consistsof fuses and circuit breakers. The protection equipment should break the currentwhenever a dangerous situation occurs that could damage the transformer. Ac-curate values of voltages and currents must be known in order for the protectionsystem to fast distinguish between a fault and normal operation. The triggeringsignals for the protection system is provided by so called Instrument transformers.These are special transformers with the purpose of transforming the high voltagesand currents from the lines to lower values more easy to handle for instrumentsand relays [6]. Instrument transformers are divided in to two basic types: Voltageand Current transformers, which will be described in more detail in the following

8

Page 12: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 2.4: Circuit representation of a current transformer with a connected burden [6].

sections. Just like ordinary full size transformers, instrument transformers are alsosubjected to errors associated with magnetizing current and leakage reactance. Fur-ther, the errors for current transformers are also load dependent, with a relativelylarge error at lower loads. To specify the amount of error one can expect, instru-ment transformers are divided into different accuracy classes specified by IEC andIEEE [6].

Current Transformer

Fig. 2.4 shows a schematic diagram of a current transformer with its primarywinding connected to a power line. The primary winding usually only consists ofone turn of the main conductor wound around the current transformer core and isalways connected in series with the line. In comparison, the secondary side whichcarries the low measurement current consists of many turns surrounding the core.The current measurement equipment – in Fig. 2.4 represented with a burden – isconnected to the secondary side. The current transformer is constructed to give amax value of 1 or 5 A at rated current at the secondary side according to certainstandards.

Voltage Transformer

There are mainly two types of voltage transformers: electromagnetic and capac-itor type, where capacitor type is the most common in high voltage application(>100kV) [6]. A voltage transformer is typically connected between line and ground.An important difference to current transformers are that the errors are less load de-pendent and the variation is small even if the voltage varies with wide limits.

2.3 Transformer Physical Fundamentals

2.3.1 Ideal Transformer

The simplest way to represent a transformer is by using the ideal transformer equiv-alent circuit, which is shown in Fig. 2.5. It consists of two windings that are cou-pled to each other via a magnetic core. For the ideal transformer the followingare assumed [2]: zero winding resistance, infinite core permeability, zero winding

9

Page 13: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 2.5: A circuit diagram showing an ideal transformer together with a voltagesource and a load [7].

reactance and no active losses in the core. This leads to the relationship:

VpNP

=VSNS

(2.1)

2.3.2 Practical Transformer

A practical transformer differs quite a bit from the ideal model in several aspects[2]:

The winding resistance is not zero.

The permeability of the core is finite, i.e. the core has reluctance.

The flux is not fully restricted to the core.

Losses are present in the core.

In order to more properly describe a practical transformer a more complete modelhas to be used. The most commonly used transformer model is shown in Fig. 2.6.This is the model that will be used in this work. Here R1 and R2 is the windingresistance and represents the ohmic losses in the primary and secondary windingrespectively. X1 and X2 is included to represent the leakage reactance of eachwinding, described in the section below. R and X can be combined to the compleximpedance Z. The shunt-branch is associated with the no-load loss in the core andcarries the magnetizing current Im. Here, Rm and Xm represent the real powerloss and the reactive loss for magnetizing the core. The ideal transformer impliesgalvanic isolation and takes care of the voltage transformation.

10

Page 14: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 2.6: The practical transformer circuit.

Leakage Reactance and Short-circuit Impedance

Since the magnetic flux is not fully restricted to the core, not all of the flux producedin one winding links the other. This non-linked part of the flux is refereed to as theleakage flux. Because of this flux the windings is said to have a leakage reactance.The leakage reactance introduces a voltage drop in the winding and causes thesecondary voltage to not be directly proportional to the primary. Transformers arenormally designed with specific leakage reactance, since it also fulfills a valuablefunction by limiting currents during fault conditions in the network [4].

If one short-circuits the secondary side of the equivalent transformer circuit, theshunt branch can be neglected and the only components that will be present arethe series impedances Z1 and Z2. The very high impedance of the shunt branchwill make the magnetizing current almost negligible compared to the total currentdrawn by the transformer. Therefore, Z1 plus Z2 are commonly referred to as theshort-circuit impedance Zw of the transformer. This is the leakage reactance andthe winding resistance added for both windings. As explained above, the short-circuit impedance will give rise to a voltage drop when the transformer is loaded.This parameter is typically given on the name plate of the transformer and is therereferred to as the percentage voltage drop, which is the voltage drop that the short-circuit impedance causes in percentage of the rated voltage according to

Vr =IrZe

V100 (2.2)

where Ir is the rated current and V is the open-circuit voltage [5]. Since the leakage

11

Page 15: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 2.7: Cross section of two windings used for Rabins’s Method [4].

reactance is the imaginary part of the short-circuit impedance, this will be depen-dent on the magnetic leakage flux, and thus the parameter will be strictly relatedto the geometrical properties of the primary and secondary windings.Their are several methods developed for calculating the short-circuit impedancefrom given geometrical properties of the winding. One example is Rabins’s Method.It uses a simplified geometry including core, coils, and yokes of infinite extent[5].Thus it is ignoring the surrounding tank walls and structures. The methodis still accurate enough to perform calculations of the forces and inductance’s of thewinding. The short-circuit reactance for the windings given by

Xw = 0.827 ∗ 10−9f(t1 + t2 + 3t12)(D1 +D2)

2h

M2

Sl

α (2.3)

where t is the radial width of shells and ducts, f is the power frequency, h is theheight, M is the total ampere turns of each shell, Sl is the rated power per woundlimb and D is the mean diameter of shells and ducts. The simplified geometryused to derive the formula is shown in Fig. 2.7. The factor α is the so called Ro-gowski factor which is a compensation factor for the leakage flux. It compensatesfor the non vertical behavior of the field lines. From Eq. 2.3 it can be seen thatthe impedance is highly dependent on the distance between the shells and thus thegeometrical properties of the winding.

12

Page 16: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 2.8: Failure location distribution obtained in a survey from Cigre [8]. It is basedon 536 major failures in sub stations transformers. Note that the winding and tap changerare major failure sources.

2.4 Transformer Failures

Like most other electrical equipment, power transformers are subjected to failuresduring their lifetime. The latest transformer failure statistics survey conducted byCigre is shown in Fig. 2.8. This indicates that winding related faults are responsiblefor nearly half of the total failures.

13

Page 17: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 2.9: Buckling of the outer winding of a three phase transformer.

2.4.1 Winding Faults

Because of the leakage flux, the current carrying windings are subjected to mechan-ical forces during load due to the Lorentz force. Since the current is passing zeroevery period, the forces are pulsating. The forces are comparably small during nor-mal operation, but since the force is proportional to the square of the current theycan reach high values during short-circuit faults, typically 100-400 times normalforces [4]. The transformer is designed to withstand these forces, but especially inold transformers, the barrier system could have been degraded to such an extent,that a bolted short-circuit can cause a permanent deformation of the winding, socalled buckling as shown in Fig. 2.9. Buckling can be bad for the transformer inseveral aspects. First, as discussed in section 2.3.2, a geometrical deformation ofthe winding will affect the short-circuit impedance of the transformer, which willfor example change the system aspects. More important, a deformed winding is notmechanically sound, and hence the withstand ability is significantly reduced belowthe guaranteed, therefore, next short-circuit can easily lead to complete breakdownof the winding.

14

Page 18: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

2.4.2 Core Related Faults

Another failure, however not so common as winding faults is problems related to thecore. Since the purpose of the core is to carry the magnetic flux, a problem couldbe visible as an increase of the losses. Core problems often takes it’s visibility as anincreasing in the magnetizing current or additional heat development. The cause offailure could have many reasons, e.g., damaged sheet lamination that increase eddycurrent losses.

2.4.3 Tap Changer faults

As can be seen in Fig. 2.8 tap changer breakdowns are one of the major failuresources which transformers are subjected to. These are mainly of mechanical originsince the tap changer is the only moving main part of a transformer and serviceshould be carried out regularly to minimize the risk of failure [4].

2.5 Monitoring and Diagnostics Methods

Monitoring refers to on-line collection of data from measurements on the trans-former. Diagnostics defines interpretation of data and all test performed with thetransformer off-line. The main reason for using on-line monitoring are to preventfailures, optimize maintenance and extend the life time of the transformer [4]. Opti-mize maintenance means in this context to change from periodic to condition basedmaintenance.

Most of the large power transformers are equipped with some kind of monitoringsystem for protection purpose (additional to ordinary instrument transformers).The typical information includes [4]:

Temperature

Gas in oil

Partial discharge

Moisture

Tap changer operation

By knowing for example the number of tap changer operations performed; the wearcan be calculated and maintenance can be carried out when needed.Different diagnostic methods could be adopted to more deeply investigate the prob-lem cause when a warning signal has been received from some of the monitoringsensors. Some widely used methods are described in the following sections:

Short-circuit Test

These test are carried out by measuring winding resistance and reactance with thetransformer decoupled from the grid. Could give indications of contact resistanceand winding deformation. In addition to this the magnetizing current can be mea-sured in the same procedure.

15

Page 19: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Dissolved Gas Analysis

Analyzing samples of the transformer oil is a valuable source of information for thetransformers health condition and several analysis methods has been developed overthe years for this purpose. The main principle behind all the methods is that, whentransformer oil is subjected to electrical sparks or severe heat; chemical reactionsoccur that generate substances in the oil [9]. One method that has gained muchattention is the Dissolved Gas Analysis (DGA). Samples of transformer oil is thenanalyzed in order to find deviant contents of different gases. Gas production inthe oil could be a result of a variety of faults, e.g., disruptive discharges, partialdischarges and core heating, to mention a few. DGA should even be able to diagnosewhich type of fault that is present in the particular case and the severity. DGAcould be considered as a standard test method that is used regularly on most powertransformers.

Frequency Response Analysis

One relatively common diagnostic method for discovering deformations of the trans-former windings is frequency response analysis (FRA). The principle behind themethod is that geometrical displacement or deformations of the windings cause achange in the leakage reactance [10]. By injecting a pulse on one side of the wind-ing and measure the output, a transfer function could be obtained by dividing thetwo Fourier spectra of the input and output signal. Changes in the geometry couldthen be detected by analyzing the obtained transfer function. This method however,requires that a fingerprint of the transfer function is available for comparison.

Partial Discharge Test

Partial discharge problems can change the content of the oil and thus be indicatedby monitoring the oil. To localize the source, different type of acoustic methodsbased on triangulation can be used [10]. These tests can be performed in interruptedservice.

16

Page 20: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

3 Transformer Explorer

3.1 Introduction

As an answer to the markets increasing interest in on-line monitoring and diagnos-tics of power transformers; a new on-line monitoring and diagnostic method is beingdeveloped at ABB Corporate Research: Transformer Explorer. The method usesfundamental frequency signals of currents and voltages obtained from ordinary volt-age and current transformers available in a substation to estimate the transformersperformance. Some of the most important information, the so called transformerfundamental parameters: turn ratio, short circuit impedance and power loss, areextracted from the transformer equivalent circuit by means of the current and volt-age signals. The method was first tested and verified using numerical computingprograms such as MATLAB with data collected from several transformers in ser-vice. Later, a complete program was written in LabVIEW for both data acquisition,analysis and visualization. It is this version of the program that will be used in thisthesis.

3.2 System Description

A simplified flow chart of the transformer explorer system can be seen in Fig. 3.1.The chain starts with the collection of analog current and voltage signals frominstrument transformers in the substation. The signals are then acquired, sampledand necessary phasors (for definition of phasor see [2]) are calculated along withvector group transformation to obtain the right winding quantities depending onthe transformers vector group (see below). Finally the fundamental quantities arecalculated using linear fits.

Figure 3.1: An overview of the Transformer Explorer system.

17

Page 21: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

3.2.1 Instrument Transformers

The start of the signal chain is the current and voltage transformers. The analogsignals are taken from the secondary circuit of the current and voltage transformers.As described in the theory part, instrument transformers are divided into differentaccuracy classes, according to IEC and IEEE standards [4]. The instrument trans-formers used in most substations is of protection class, which means that theyusually have an accuracy around 1 %.

3.2.2 Acquisition System

There are mainly two ways to access the analog signals in the substations controlroom. With the first option, the analog signals are measured at the secondarycircuit by means of additional sensors, i.e., current clamps and voltage dividers.The signals are then collected using a acquisition board with a built-in analog todigital converter, which can provide the sampled waveforms needed for the lateranalysis. This is the method of choice for older substations, equipped with electro-mechanical relays. With the second option the signals are obtained from the distur-bance recorder, which is an integral part of many numerical protection relays, e.g.,ABB REx670, installed in many modern substations today. Then, there is no needfor signal sampling. If an even more modern relay is available, like the REx670V2;this can also estimate accurate phasors, which will be described more in the nextsection.

3.2.3 Phasor Calculation

Transformer Explorer uses phasors of the voltage and current signals to do itsanalysis. Since Transformer Explorer is used for monitoring and analysis purpose,no immediate response is needed compared to protection systems. This means thatanalysis of relatively long and steady waveforms can be allowed, which result invery accurate phasor estimation, even at lower sampling rates. The TransformerExplorer software uses an algorithm: high precision alternating current (HPAC) toestimate the phasors, which is a type of interpolated FFT with high accuracy. Inthe case an numerical relay such as ABB REx670V2 is used, the HPAC algorithmis built in and the phasors can be directly acquired from the relays .

3.2.4 Vector Group Transformation and Zero-Sequence Re-moval

Most electrical systems needs to have a connection with earth, for functionality andsafety reasons. In case of a transformer, earth connection is provided by connectingthe star point to earth, either by an solid conductor or through an impedance, socalled impedance grounding [5]. The network should also provide a path for thethird harmonic current in order to suppress or eliminate harmonic components inthe voltage waveform. This is achieved by connecting either the high or the lowvoltage side in delta configuration. Since Transformer Explorer uses the one phaseequivalent circuit, line quantities measured by the instrument transformers, in caseof a delta winding, needs to be transformed to the right winding quantities beforethey can be applied to the model. This is because the delta winding introduces aphase shift in the voltage. A transformer can have many different types of vectorgroup configurations and Transformer Explorer must be able to deal with every

18

Page 22: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

one of them. Depending on the winding configuration of the analyzed transformer,vector group transformation might need to be performed to obtain the right voltageand current phasors for the transformer model.

Transformer Explorer does also have the option to remove zero-sequence com-ponents (see [2] for definition of sequence components) from the phasors, beforethey are used for fitting. Whether they should be removed or not depend on theparticular transformer configuration. It is however convenient in most cases whenestimating the impedance, since the transformers zero-sequence impedance is dif-ferent from the positive and negative equivalents. The presence of zero-sequencecomponents could be for two reasons: there exist an actual phase imbalance, eithercaused by the feeding grid itself or by a load imbalance on the supply side of thetransformer, or mismatch between the sensors, e.g, instrument transformers, caus-ing an artificial imbalance. The zero-sequence component could also be excludedfrom the current phasors when estimating the turn ratio for some transformer types.This is due to the fact that a delta winding blocks the zero-sequence component onone side.

3.2.5 Extraction of Quantities

The main principle behind transformer explorer is to fit observed data to equationsderived from the equivalent transformer circuit. These equations are of the linearform

y = ax+ b (3.1)

where y and x are the phasor data, e.g., primary and secondary current phasors, aand b are quantities derived from the transformer model. The line fit entities usedare summarized in table 3.1 and are described in more detail in the sections below.

Table 3.1: Line fit entities used for the fits.

x y a bn I1 I2 1/n I0Zw I2/n V1 − nV2 Zw Z1I0

The Turn Ratio Fit

Since the magnetizing current I0 is independent of the load current, a linear relationbetween the input and output current of the transformer can be obtained as

I1 =I2n

+ I0 (3.2)

This means that by fitting the input current against the output current, two impor-tant quantities can be obtained: The turns ratio as the slope and the magnetizingcurrent I0 as the intercept. Since two currents are used, the slope will be dimen-sionless and I0 will have units of A. Note here that the turn ratio should be a purereal quantity. Any imaginary part obtained should thus be a indication of eventualsensor mismatch.

19

Page 23: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

The Impedance Fit

The voltage drop ∆V is the voltage difference between V1 and V12, i.e., the inputvoltage and the output voltage multiplied by the turns ratio. This is mainly de-pendent on the load current, since I0 is relatively low. ∆V could be used to writethe next linear equation, which ia basically obtained by writing ∆V as a functionof the load current I2. This gives the resulting equation, c.f., Fig. 2.6,

∆V = V1 − nV2 =(R1 + jX1 +R12 + jX12)

I2n

+(R1 + jX1)I0

= ZwI2n

+ Z1I0 (3.3)

The important quantity from this fit is the short-circuit impedance Zw which is givenin Ω. The short-circuit impedance is usually given on the transformers nameplateas the percentage voltage drop Uk at rated load. This could easily be converted toΩ using

Zw = UkVnominal

Snominal

(3.4)

This quantity has both an imaginary and a real part. The imaginary part reflectsthe leakage flux and the real reflects winding and eventual contact resistance. Anydeviations from this fit is given in kV .

Power Loss

The last quantity that can be obtained is the total power loss. This is not obtainedusing linear fits, instead it could simply be obtained by taking the difference be-tween the power going in to the transformer and the power going out. Describedmathematically using the same notation as in Fig. 2.6, this is

Sloss = V1I∗1 − V2I∗2 (3.5)

One problem with this equation is that, since it formed as a difference, it is verysensitive to errors in the measured quantities. This results in that the power lossis difficult to measure with any absolute accuracy without applying some sort ofcorrection to the sensors (see more in chapter 6).

3.3 Transformer Explorer Software

In this section, the structure and the functionality of the transformer explorer soft-ware is presented. The section starts with a brief introduction to LabVIEW.

3.3.1 LabVIEW

LabVIEW is a graphical programming environment developed by National Instru-ments. It is commonly used by scientists and engineers for fast implementation ofdata acquisition and measurement systems. A program consists of two windows:the front panel and the block diagram. The front panel works as the interface fromwhere the user can control the program. The programming code is written in theblock diagram, by connecting individual icon boxes together in a proper way. Eachbox represent a block of underlying executable code that performs a certain task.

20

Page 24: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

LabVIEW has several features that makes it an excellent programming languagewhen it comes to writing prototype programs for scientific applications. Its graph-ical interface and large library, letting the user to quickly build a program and it’sability to divide the program into separate so called subVI where the code can beexecuted separately, makes it excellent for debugging.

3.3.2 Program Structure

Transformer Explorer uses an event driven program structure and the current ver-sion is written in LabVIEW. An example of the code structure are given in Fig 3.2.The program is build around an endless while-loop. The while-loop is wrapped witha string array with commands that gives instructions to the main case structure.The case structure is a labVIEWs graphical counterpart to else if statement com-mon in other programming languages. The program starts in the INIT case whichis where the initial settings for the program is set. When running the programit continuously searching for instructions in the command array and whenever itreceives a new command it goes to the addressed case to perform its task.

Figure 3.2: The program structure, the outermost frame is a never ending while loopfollowed by an inner case structure. This is the main part of the program which in thisfigure shows the default case. This case contains a event structure which responds to useraction.

21

Page 25: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

3.3.3 Data Acquisition

The data acquisition is performed by the getNewData.vi, and this provides a numberof possibilities for the user. As described earlier there are three main options to feedtransformer explorer with data: from an ADC, from a waveform file, directly froma protection relay or by loading previously analyzed files. Using the second optionit can acquire wave form files stored in Comtrade or ManyScope (ABB’s privateformat). When acquiring data from a relay it both has the possibility to downloadwave-form files, or in case of a newer protection relay, e.g., REx670V2, only phasorsneed to be communicated.

It also have the possibility to read from pre analyzed table files in ASCII format.Each measurement is then a line in this file, containing information of the phasorgiven by two quantities: frequency and complex amplitude. The vi uses the afrequency acceptance test in order to not let trough any phasors with too muchdeviation leading to deceptive results when later analyzed. The amplitude could begiven in either complex form or with amplitude and phase angle in degrees. Theuser could specify settings for the acquisition in the transformer explorer settingswindow shown in Fig. 3.3.

Figure 3.3: The settings window with the input tab selected. Here the user can choosewhich type of files that should be analyzed together with settings for reading intervalsand nominal frequency for the frequency acceptance test. Note that the input selector inthis example is set for reading table files.

3.3.4 Data Analysis

The analysis part can also be divided in two separate main pars: action on everynew data set and actions that are performed only when enough data has beenloaded. It is in the latter the most central part of transformer explorer takes place;namely the fitting of data. The action on new data could be divided in three stepsthat are performed on all new data that enters the system:

22

Page 26: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Transformation from line voltages and currents to winding voltage and cur-rents, as discussed in earlier section. This is performed by a sub.vi, whichtransforms the quantities according to the right vector group. This is spec-ified in the settings window tab Transformer. The vector groups that couldbe chosen are YY, YD, DD, DY and the clock options are 1, 5, 6, 11, whichincludes the most common configurations.

Since the turn ratio and impedance change with tap position, and the powerflow direction in the transformer could be different; the data has to be clas-sified according to this properties when analyzed. This is performed by asub.vi, which can determine the tap belonging to each class in two ways: ei-ther from currents or voltages. The latter one is preferably used since thecurrent method is used for determine the ratio

Transformer explorer will also find the deviations of each new data point fromthe fitted fundamental quantities. This is considered as one of the main resultsand is important when determining the uncertainty of the fit.

Actions that are performed only when enough data are collected includes of coursethe fitting itself. The fitting could be divided into running fits and reference fit. Therunning fit is done whenever enough new data are contained in the class and this isdone only with data newer than those included in the last fit. In order to simplifythe monitoring ability, a reference fit is also performed on all data included in theclass after a certain number of running fits have been done. A number of criteriamust be fulfilled in order for the program to make a fit, here the most importantare:

Minimum readings for fit: The class must contain at least this much data inorder to be able to make a fit.

Minimum load range in fit: The class must contain data where the load hasvaried a certain specified minimum percentage value of the nominal power.

Minimum load in fit: The minimum load should not be lower than a certainvalue, e.g., 10 percent. This is because effects of the magnetizing current willbecome more significant at lower loads.

Minimum fits required to freeze reference: A specified number of fits requiredto freeze the reference. The value should be set so that the reference fitcontains significantly more data than the running fits.

Exclude the most deviating data in fit: The program excludes and throwsaway a specified percentage of the data that deviates most from the fit andmakes a new fit. This is done to reduce the impact of disturbed data on thefit.

The criteria are administrated from the settings window, Criteria tab shown in Fig3.4.

23

Page 27: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 3.4: The Criteria tab, letting the user apply settings regarding the fitting pro-cedure.

3.3.5 Display Options

The analysis and monitoring in transformer explorer can be carried out using fourdifferent display options: Status, winding values, fundamental quantities and dif-ference to fit. The various display modes are accessed from a pull-down menu onthe front panel. The functionality for each one of them will be presented in thefollowing sections.

Status

The status display serves as an overview of the monitoring status. As default displayit is also the first thing the user sees when start running the program. The displayis divided in two sections: present status and fit status. The present status givesinformation about the latest measurement with corresponding winding values, e.g.,voltage and currents for each phase. Latest difference to measurement fit is alsopresented for each of the quantities. A picture of the status display is shown in Fig.3.5.

24

Page 28: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 3.5: The status display.

Winding Values

This display is intended for analyzing the values that are used for the fits, but alsoto obtain directly calculated quantities such as power and power loss. The display iscontrolled by four pull-down menus: value, component, view and parameter. Thefirst menu is the most central for this display and the following options can bechosen:

1. Voltages

2. Currents

3. Voltage ratio: ratio between the high and low voltage side.

4. Compensated voltage ratio: The same as above, but this ratio is compensatedfor the internal voltage drop.

5. Current ratio: ratio between the corresponding currents in one winding phase.

6. Voltage drop: The internal voltage drop, i.e. the difference between the highvoltage and the low voltage multiplied by the turn ratio.

7. Current difference: difference between the high side current and the low sidecurrent divided by the turn ratio.

8. Power: The power supplied to the high voltage side and the power deliveredby the low voltage side.

9. Power loss: the difference between the powers on each side.

25

Page 29: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

For all of these values the choices component and view could be varied. The compo-nent menu allows the user to view the sequence component for each value, and theview menu makes it possible to show absolute, real, imaginary or the phase angle.The last pull-down menu is basically the x-axis on the graph and makes it possibleto plot the values against other parameters than the default: time, e.g., current orreal power.

Fundamental Quantities

This is probably the most valuable display option since it shows the fit results. Fromthe fit pull-down menu the three obtained fits can be chosen: ratio, impedance andpower, where only the first two will be treated in this work. The second menuresult, where the results from each type of fit can be chosen. Slope and interceptare the most important here. The view menu is the same as for winding quantitiesexcept for that the phase angle can’t be shown. Both the running and reference fitis displayed which are shown in Fig 3.6. The reference fit is always displayed as afixed value.

Figure 3.6: The fundamental quantities display.

Difference to Fit

This display shows the difference between the latest measurement and the referencefit. It is a fast indicator or eventual problems and can thus be used for changedetection. The Fit and the parameter menus are the same as for the fundamentalquantities. In the pull-down menu it is possible to chose between the deviationunits: std. deviation, % of nominal and actual value.

26

Page 30: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 3.7: The difference to fit display.

3.3.6 Logging of Data

Transformer explorer has the ability to create a number of different files with resultsfrom performed monitoring and analysis. In order to obtain a consistent and struc-tured storage of the files it uses a specific folder structure This consists of a mainpath, which by default is set to a folder created on the start up disk. A specificfolder for each analyzed transformer is then created inside the main, to log obtainedresults.

There are mainly three type of log files generated during a run: reference fitlog, fit log and difference to fit log files. The reference to fit log files stores dataassociated with the reference fit. Every time the program has enough data to makea reference fit, new entry is written to the file. Each line contain information aboutthe class, load variation, and fit results for each type of fit. Each fit is in turngiven in slope (sloR & sloI), intercept (intR & intI) and fit deviation, where R andI refers to the real and imaginary part. The fit log files are similar to the referenceto fit but instead a new entry is made each time the program makes a new fit. Thedifference to fit log file logs the difference between each data and the reference fit,and is written to every time new data enters the analysis. Apart from the othertwo files the difference to fit log also contains the winding values for the voltageand current associated with each data. The general structure for all of the log filesare tab separated text files with a time column at its right. Settings concerningthe folder structure, as well as the log files is controlled from the logging tab in thesettings window shown in Fig. 3.8.

27

Page 31: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 3.8: The logging tab in the criteria settings window. From here, different optionscan be set for logging of the results. Note the Max memory which by default is set to1000 data points, this and the Max Files setting serves to avoid memory overflow duringlong monitoring sequences. Max memory limits how many measurements that are keptin active memory and Max Files will activate file deletion if too many files are found.

3.4 Field Installations

Transformer explorer has been tested on several transformers in service and a largedata base based on the measurements has been built up. Some of the installationshave been temporary while one transformer still is observed on-line. The threeinstallations used for this work will be presented below. They will be refereed toas Field installation 1,2 and 3 respectively. The name plate data for the threetransformers are shown in the tables below.

Table 3.2: Name plate data for Field installation 1

Fabrication year: 2000Power rating 63 MVA

Voltage 140 ± 8 x 1.67% / 55 kVConnection YN yn 00

Cooling type ONANImpedance 36.7111j Ω

Table 3.3: Name plate data for Field installation 2

Fabrication year: 1968Power rating 56 MVA

Voltage 140 ± 9 x 1.75% / 11 kVConnection YN yn 00

Cooling type ONANImpedance 51.7402j Ω

28

Page 32: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Table 3.4: Name plate data for Field installation 3

Fabrication year: 1979Power rating 30 MVA

Voltage 140 ± 9 x 1.67% / 11 kVConnection YN d 11

Cooling type OFAFImpedance 84.4223j Ω

3.5 Fault Detection Ability

By investigating the fundamental quantities – Transformer Explorer should be ableto detect a number of faults affecting the transformer. Some of the problems themethod should be able to detect includes:

Turn to Turn Faults

Faults caused by short circuit between adjacent turns in the winding should bedirectly visible as a change in the turn ratio fit. This quantity could be observedfrom several displays calculated with different methods such as: current ratio, com-pensated voltage ratio and the value obtained from the turn ratio fit. The latterone should be the most accurate of course. However, if an actual turn to turn faulthappens the change should be consistent for all of the methods making the otherones important as verification if a problem is detected. If a change only is visiblein one of the displays, it is probably due to a sensor problem, or another type ofchange, such as increased magnetizing current.

Tap Changer Function

Since the change in turn ratio between different taps should correspond to the per-centage value on the nameplate, a deviation from this value could be an indicationof lack of function in the tap changer.

Core Problems

Core problem could take its visibility in a change in the magnetizing current orthe active power loss. The magnetizing current is obtained as the intercept of theturn ratio fit, and should thus be sensitive to core related problems. It is howeverimportant to know that the magnetizing current also is dependent on the voltagelevel and this must hence be checked before any actions is taken on an observedchange. A core problem could also be seen in the active power loss. Since the corelosses is not load dependent any change is observed as an decrease or increase inthe offset of the power curve. See section 2.4.2 for more information on core relatedfaults.

Winding Deformation

As discussed in section 2.3, the winding geometry is directly related to the shortcircuit impedance. A winding deformation, e.g. caused by a short circuit, will causea permanent change in the short-circuit impedance. Such a problem could thus bedetected by the impedance fit.

29

Page 33: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Contact Problems

Contact problems, such as loose connections and coking tap changer contacts withadditional heat development as result, will cause a change in the winding resistanceand should thus be visible in the real part of the impedance.

Eddy Currents

Could be present both in the core, surrounding support structure and tank; in thelatter case they may be load dependent. Eddy currents could therefore be detectedby observing the magnetizing current and active power loss.

3.6 Transformer Explorer Applications

Transformer Explorer is expected to find its application in two different scenarios:either as a diagnostics tool during a short time installation (a single day or a week),or as an on-line monitoring tool using a permanent installation. In the first case thenecessary signals are obtained using existing disturbance recorders or by secondarysensors as a temporary installation as mentioned before. The collected data fromthe measurement are analyzed and the results are communicated in a report. Theshort time installation could be seen as an on-line version of the conventional off-line measurements, with the obvious benefit of not taking the transformer out ofservice.

In the latter case the acquisition equipment is permanently installed and the dataare continuously collected and could be monitored on-line using a web-interface.The program is then able to trigger on abnormal events and save associated datafor later analysis. With this option the ability to detect developing faults in an earlystage is even higher than with the short-time option. An example of this could bean observed change in the short-circuit reactance after a close-up short-circuit fault,which has the possibility to become a complete winding breakdown during the nextshort-circuit event.

3.7 Uncertainty and Accuracy

Whenever a measurement is done in practice there is always some uncertainty as-sociated with the measured value. The uncertainty could be divided in two types:systematic and random errors. The systematic error causes the measured valueto constantly deviate from the correct value by a certain offset. It is said that ameasurement system with high systematic error has a low absolute accuracy. Ran-dom uncertainty is variations in the measured value that does not follow a regularpattern, e.g., signal noise. Since transformer explorer is directly dependent on theaccuracy of the sensors, e.g., instrument transformers, the accuracy can becomerelatively low in both the estimates and the power loss. The noise level is how-ever comparably low and thus also the random uncertainty. The HPAC algorithmprovides phasors with a noise level less than 0.01% in the normal case.

3.7.1 Derivation of Uncertainty in a Linear Fit

Consider an arbitrary straight line of the form

y = kx+m (3.6)

30

Page 34: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

How is the uncertainty of the estimated quantities affected if there is noise presentin the measured values x and y? This is rather straight forward if one considersthe exact solution for only two pairs of measured values: (x1, y1) and (x2, y2). Theslope and intercept is given by:

k =y2 − y1x2 − x1

(3.7)

and

m =x1y2 − x2y1x1 − x2

(3.8)

If one of the y’s is increased by an infinitesimal change, the change in k will be

|∆k| =∣∣∣∣ dy

x1 − x2

∣∣∣∣ (3.9)

Thus, the uncertainty in k is directly proportional to the change in y and inverselyproportional to the range in x.

It can also be shown that; by adding dy to y1 and y2 and taking the average ofthe change in m one can see that this is directly proportional to the change in m,i.e., the uncertainty.

3.7.2 Uncertainty in the Estimated Parameters

Given the information from the previous section, an expression for the uncertaintyin the turn ratio and impedance fits can be derived.The uncertainty in the turn ratio could be estimated as standard deviation of thedifference δILV = IMeas

LV − IFitLV and the range of the abscissa. This results in the

expression

∆n =σ(δILV )

max(IHV )−min(IHV )(3.10)

Using the same principal the uncertainty in the impedance fit is given by

∆Z =σ(δ∆VHV )

max(IHV )−min(IHV )(3.11)

where ∆VHV is the voltage drop over the short-circuit impedance.

31

Page 35: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

4 Short Time Measurements

A possible application for transformer explorer that could give valuable informationfor a transformer owner is short-time measurements using a temporary installation.Two main questions that needs to be adressed are then:

How and what results should be presented to provide the most valuable in-formation for the transformer owner?

What performance could be expected when the tool is used for short-timeanalysis? Is the assumed uncertainty valid?

The first question was planned to be answered by constructing an example mea-surement report using existing collected field data, together with a power utility:Vattenfall.

It was decided that the second question should be answered by conducting asurvey where field data was analyzed corresponding to two measurement scenariosof different length, one shorter and one longer measurement. The results for eachscenario could then be analyzed and compared to each other, and hopefully somevaluable information regarding the tools performance extracted. In this chapterthe method used for answering the two questions will be presented along with theresults.

4.1 Day and Week Comparison

To construct a report for short-time measurements, some parts concerning the per-formance of transformer explorer needs to be further investigated, namely:

How long should a measurement period be?

Is the derived uncertainty in the measurements correct?

4.1.1 Method

The data used for the study was taken from the already existing data-base. Tocover eventual differences between different transformer types – data from threetransformers were used: Field installation 1, Field installation 2 and Field instal-lation 3 as mentioned in section 3.4. The collected data available for the surveyconsists of so called FundLog files, where each file corresponds to data for a periodof two days. For simplicity, one FundLog file was used for the short period and 4for the longer period. That corresponds to data collected during a two day, and a8 day measurement (roughly one week). The two time periods will here after berefereed to as day and week.

32

Page 36: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

The outcome of a short-time measurement with Transformer Explorer should ofcourse contain the three fundamental quantities: turn ratio, impedance and powerloss. This survey will deal with the first two, since the power loss needs special atten-tion (see chapter 6). Other results that could be extracted and should be includedin the study is the magnetizing current. This because the magnetizing current, inaddition to the power loss, also could provide valuable information concerning thetransformers condition.

A result of any measurement should always be presented with a proper uncer-tainty – a measurement value could be considered worth nothing unless the un-certainty is properly specified [11]. The uncertainties in the fitted quantities werederived analytically (see Section. 3.7) and thus, their validity for short-time mea-surements needs to be checked.

The entire period used for each transformer could be seen in Table 4.1. Thesub-periods (days and weeks) were chosen randomly between these intervals. 10sub-periods of each kind were used to get some statistical validity.

Table 4.1: Time periods from where the data was taken for each transformer.

Transformer Day Week1 14-04-16 - 14-08-07 14-04-16 - 14-07-232 14-05-24 - 15-01-02 14-05-24 - 15-01-103 13-06-06 - 14-02-13 13-06-03 - 14-03-22

One fundamental difference when the program is used for short-time measure-ments instead of continuous monitoring, is that the program should be able tomake a fit using all data that has been collected. This functionality is not neededwhen the program is used for monitoring since then a fit should only contain datanot older than the latest fit. Since the concept initially was developed for monitor-ing purposes this functionality was missing. The problem was solved by adding anextra case to the program structure: doFullFit. The user should be able to makea full fit when all data has been read. Therefore a button was added to the frontpanel which will be visible when the continuous acquisition has been turned off.Another feature that was added was the ability to read a single file instead of anentire file list, which was the only option in the initial version of the program. Inorder to collect and analyze all the results from all the measurements in an efficientway a MATLAB1 program was built. The program should be able to accomplishthe following:

1. Read the desired information from each RefFitLog file produced by Trans-former Explorer.

2. Calculate results based on this information, e.g., uncertainties for each fit.

3. Write the results for each measurement to different tabular files which canlater be plotted and analyzed.

The presented procedure needs some further explanation. When Transformer Ex-plorer writes to the RefFitLog the following parameters, except from date and time,are included: tap position, power direction, total low load & high load, ratio slope &intercept, ratio deviation, impedance slope & intercept and impedance deviation.The quantities are further given by phase, each having a real and an imaginary

1MATLAB version R2015b was used in this thesis work.

33

Page 37: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

part. Since the output files is classified by tap position and power direction notall of the information in the RefFigLog files are needed. As explained earlier theresults need to be presented with its corresponding uncertainty. Thus, the programmust calculate the uncertainty for each quantity and write in separate columns inthe tabular file. For the magnetizing current this corresponds to the standard de-viation of the ratio fit and for the quantities: the standard deviation divided bythe range (see section 3.7). The resulting tabular files produced by the programcontains the following information:

Real and imaginary part of the turn ratio.

Ratio difference to nominal.

Ratio uncertainty.

Real and imaginary part of impedance.

Impedance difference to nominal.

Impedance uncertainty.

Magnetizing current.

Magnetizing current uncertainty.

Number of data used in fit.

High and low load.

and again, the information was given for all of the three phases. Some of theinformation was later considered as less valuable and not used in the study, i.e.,number of data and difference to nominal for the quantities. Difference to nominalfor the magnetizing current was not included since no information about the nominalmag current existed for any of the studied transformers.

4.1.2 Results

In this section the resulting plots and tables for each transformer will be presented.To facilitate the comparison, the average of the actual variations between the mea-surements and the average uncertainty obtained from each one of them (the errorbars) are shown in tables. They are referred to as several measurements and singlemeasurement, since the first column relate to the average uncertainty based on allmeasurement and the second column relates to the average uncertainty estimatedfrom each of the single measurements. This is shown for all of the investigatedparameters. If the derived uncertainty is correct, these two should at least be ofthe same order. Some of the parameters was also plotted against average load cur-rent, this to investigate if any correlation with load exists. This is justified sincecurrent transformers are slightly load dependent. Error bars with the uncertaintywas added to the plots to visualize if the expected uncertainty corresponds to thevariation between each measurement. When available, the nominal values were alsoplotted with the quantities for comparison. It is important to mention that onlyone tap position was studied: the one most commonly used. This is because notenough load variation was observed for some of the day periods to perform a fit forother tap positions. This is an important result itself since it implies that there is

34

Page 38: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

no guaranty to obtain results from more than one tap position when a short timemeasurement is carried out. However, it is very little chance to cover the majorityof the available tap positions even during a much longer period, as this depends onthe transformer usage. Generally, extreme taps are rarely used; most transformersuse only a few taps around the nominal.

Field Installation 1

The real part of the turn ratio looks rather stable for both the day and weekmeasurement with no significant deviations as can be seen in Fig. 4.1. Note herethat the individual phases differ so much from each other that the error bars isbarely visible in the graphs. Table 4.2 shows that the ratio variation correspondswell to the single estimated uncertainties. Phase 0 shoes however a slightly highervariation for both the periods. Fig 4.2 shows the magnetizing current as a functionof average load current. The variation looks also here to be in the same ordercompared to the error bars. Table 4.4 and 4.5 reveals the same information and nosignificant difference is observed. The imaginary part of the impedance is shown inFig. 4.3. Here the predicted uncertainty is higher than the actual variation and theimpedance varies slightly less for the week interval. An intriguing observation is thewave shape pattern observed for both the day and weeks within roughly the sameload interval. Fig. 4.4 shows the real part of the impedance (the resistance) as afunction of average load current. Also for this the uncertainty seems well predictedand no significant difference between the time periods can be observed. Note herethat the resistance is negative for Phase 2. This is of course not physically possibleand is a result of the low accuracy described in section 3.7. The variations betweeneach single measurement are however significantly lower and this is a good exampleof the difference between systematic and random error.

10 12 14 16Average load current [%]

2.5

2.55

2.6

2.65

Rat

io r

eal p

art

Day

Phase0Phase1Phase2Nominal

10 15 20Average load current [%]

2.5

2.55

2.6

2.65

Rat

io r

eal p

art

Week

Phase 0Phase 1Phase 2Nominal

Figure 4.1: Real part of the turn ratio as a function of average load current, Fieldinstallation 1.

35

Page 39: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Table 4.2: Turn Ratio Uncertainty2, day-period, Installation 1

Phase: Several measurements Single measurements0 14.201e-04 4.1256e-041 8.8104e-04 3.8433e-042 8.7846e-04 2.2615e-04

Table 4.3: Turn Ratio Uncertainty, week-period, Installation 1

Phase: Several measurements Single measurements0 13.868e-04 4.4078e-041 6.5666e-04 4.3228e-042 8.5841e-04 2.5309e-04

10 12 14 16Average load current [%]

0

0.1

0.2

0.3

0.4

0.5

0.6

Mag

netiz

ing

curr

ent [

A]

Day

Phase 0Phase 1Phase 2

10 15 20 25Average load current [%]

0

0.1

0.2

0.3

0.4

0.5

0.6

Mag

netiz

ing

curr

ent [

A]

Week

Phase 0Phase 1Phase 2

Figure 4.2: The magnetizing as a function of average load current, Field installation 1.

2The left column ”Several measurements” is here a measure of the actual uncertainty from allthe measurements obtained by taking the standard deviation of the turn ratio values. The rightcolumn ”Single measurements” is the mean value of all the estimated uncertainties from eachsingle measurement ( the error bars).

36

Page 40: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Table 4.4: Magnetizing Current Uncertainty, day-period, Installation 1

Phase: Several measurements [A] Single measurements [A]0 0.0281 0.01641 0.0326 0.01512 0.0264 0.0089

Table 4.5: Magnetizing Current Uncertainty, week-period, Installation 1

Phase: Several measurements [A] Single measurements [A]0 0.0244 0.02321 0.0278 0.02252 0.0414 0.0130

10 12 14 16Average load current [%]

36.4

36.6

36.8

37

37.2

37.4

37.6

37.8

38

38.2

38.4

Impe

danc

e im

agin

ary

part

[Ω]

Day

Phase 0Phase 1Phase 2Nominal

10 15 20Average load current [%]

36.4

36.6

36.8

37

37.2

37.4

37.6

37.8

38

38.2

38.4

Impe

danc

e im

agin

ary

part

[Ω]

Week

Phase 0Phase 1Phase 2Nominal

Figure 4.3: The imaginary part of the impedance as a function of average load current,Field installation 1.

37

Page 41: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

10 12 14 16Average load current [%]

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

Impe

danc

e re

al p

art [Ω

]

Day

Phase 0Phase 1Phase 2

10 15 20 25Average load current [%]

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

Impe

danc

e re

al p

art [Ω

]

Week

Phase 0Phase 1Phase 2

Figure 4.4: The real part of the impedance as a function of average load current, Fieldinstallation 1.

Table 4.6: Imaginary Impedance Uncertainty, day-period Installation 1

Phase: Several measurements [Ω] Single measurements [Ω]0 0.0528 0.06941 0.0504 0.06852 0.0586 0.0674

Table 4.7: Imaginary Impedance Uncertainty, week-period, Installation 1

Phase: Several measurements [Ω] Single measurements [Ω]0 0.0445 0.06341 0.0398 0.06432 0.0512 0.0621

38

Page 42: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Table 4.8: Real Impedance Uncertainty, day-period, Installation 1

Phase: Several measurements [Ω] Single measurements [Ω]0 0.0681 0.06941 0.0621 0.06852 0.0684 0.0674

Table 4.9: Real Impedance Uncertainty, week-period, Installation 1

Phase: Several measurements [Ω] Single measurements [Ω]0 0.0549 0.06341 0.0474 0.06432 0.0649 0.0621

Field installation 2

Also for this installation no significant change in the results between the two periodsare observed. Note here that the load profile is much less varying – the load variationis similar for both periods. The turn ratio is shown in Fig 4.5. Table 4.8 and4.9 shows that the uncertainty corresponds well to the variation for both periods.Also for the magnetizing current shown in Fig. 4.6, the uncertainty seems in theright order. There is one measurement around 14% for the shorter period, wherethe currents are comparably higher but this day also has higher uncertainty. Theimaginary part of the impedance is shown in Fig. 4.7. Here it seems like thevariation is slightly smaller for the week measurements as also indicated by Table4.12 and 4.13. The resistance can be seen in Fig. 4.8. One could possibly see adecreasing trend for the resistance with higher average load current. The resistanceobtained below 5% deviates significantly from the other day measurements. Thisis probably due to high error in the current transformers. This is probably also thereason for the lower variation for the week periods seen in table 4.17.

5 10 15Average load current [%]

13.16

13.18

13.2

13.22

13.24

13.26

Rat

io r

eal p

art

Day

Phase 0Phase 1Phase 2Nominal

8 10 12 14 16Average load current [%]

13.16

13.18

13.2

13.22

13.24

13.26

Rat

io r

eal p

art

Week

Phase 0Phase 1Phase 2Nominal

Figure 4.5: Turn ratio as a function of average load current, Field installation 2

39

Page 43: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Table 4.10: Turn Ratio Uncertainty, day-period, Installation 2

Phase: Several measurements Single measurements0 0.0060 0.00691 0.0035 0.00722 0.0059 0.0065

Table 4.11: Turn Ratio Uncertainty, week-period Installation 2

Phase: Several measurements Single measurements0 0.0046 0.00491 0.0038 0.00512 0.0045 0.0045

0 5 10 15 20Average load current [%]

1.8

2

2.2

2.4

2.6

2.8

3

Mag

netiz

ing

curr

ent [

A]

Day

Phase 0Phase 1Phase 2

8 10 12 14 16 18Average load current [%]

1.8

2

2.2

2.4

2.6

2.8

3

Mag

netiz

ing

curr

ent [

A]

Week

Phase 0Phase 1Phase 2

Figure 4.6: Magnetizing current as a function of average load current, Field installation2

40

Page 44: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Table 4.12: Magnetizing Current Uncertainty, day-period, Installation 2

Phase: Several measurements [A] Single measurements [A]0 0.0986 0.06651 0.0885 0.06882 0.1042 0.0629

Table 4.13: Magnetizing Current Uncertainty, week-period Installation 2

Phase: Several measurements [A] Single measurements [A]0 0.1172 0.07271 0.0813 0.07552 0.0760 0.0665

5 10 15Average load current [%]

52

53

54

55

56

57

Impe

danc

e im

agin

ary

part

[Ω]

Day

Phase 0Phase 1Phase 2Nominal

8 10 12 14 16Average load current [%]

52

53

54

55

56

57

Impe

danc

e im

agin

ary

part

[Ω]

Week

Phase 0Phase 1Phase 2Nominal

Figure 4.7: Imaginary part of impedance as a function of average load current, Instal-lation 2.

41

Page 45: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Table 4.14: Imaginary Impedance Uncertainty, day-period, Installation 2

Phase: Several measurements [Ω] Single measurements [Ω]0 0.4238 1.03401 0.5300 1.04822 0.3141 1.0258

Table 4.15: Imaginary Impedance Uncertainty, week-period, Installation 2

Phase: Several measurements [Ω] Single measurements [Ω]0 0.1425 0.84211 0.1510 0.84342 0.2930 0.8263

5 10 15Average load current [%]

-2

-1

0

1

2

3

4

5

Impe

danc

e re

al p

art [Ω

]

Day

Phase 0Phase 1Phase 2

8 10 12 14 16 18Average load current [%]

-2

-1

0

1

2

3

4

5

Impe

danc

e re

al p

art [Ω

]

Week

Phase 0Phase 1Phase 2

Figure 4.8: Real part of impedance as a function of average load current, Field instal-lation 2.

42

Page 46: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Table 4.16: Real Impedance Uncertainty, day-period, Installation 2

Phase: Several measurements [Ω] Single measurements [Ω]0 0.8419 1.03401 0.6202 1.04822 0.7962 1.0258

Table 4.17: Real Impedance Uncertainty, week-period Installation 2

Phase: Several measurements [Ω] Single measurements [Ω]0 0.2830 0.84211 0.4204 0.84342 0.2906 0.8263

Field installation 3

The last investigated transformer was field installation 3. This is the smallest of thetransformers in this survey, in terms of rated power. The turn ratio as a functionof average load current is shown in Fig. 4.9. For both measurement periods, onecan see a convergence with higher loads. Further the uncertainty is decreasing forhigher currents. Fig 4.10 showing plots for the magnetizing current one can insteadobserve a divergence pattern for the week measurements. A similar converging asin Fig. 4.9 can also be seen for the imaginary part of impedance shown in Fig. 4.11.Finally the resistance is shown in Fig. 4.12. Also here the converging for highercurrent values can be seen but towards lower resistance. All the observed patternsfor this transformer is most likely load dependent errors introduced by the currenttransformers. It was later discovered that the primary side current transformer forthis unit has very high current rating compared to the average currents normallyflowing, i.e., the currents through the current transformers are around 5% for thelower loads seen in this survey. The 1% maximum error are only guaranteed whenthe current transformer are operated with a current above 10% of rated.

43

Page 47: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

10 15 20 25 30Average load current [%]

7.5

7.52

7.54

7.56

7.58

7.6

7.62

Rat

io r

eal p

art

Day

Phase0Phase1Phase2Nominal

10 15 20 25 30Average load current [%]

7.5

7.52

7.54

7.56

7.58

7.6

7.62

Rat

io r

eal p

art

Week

Phase 0Phase 1Phase 2Nominal

Figure 4.9: Turn ratio as a function of average load current, Field installation 3.

Table 4.18: Turn Ratio Uncertainty, day-period, Installation 3

Phase: Several measurements Single measurements0 0.0240 0.00701 0.0100 0.00662 0.0166 0.0069

Table 4.19: Turn Ratio Uncertainty, week-period, Installation 3

Phase: Several measurements Single measurements0 0.0267 0.00621 0.0080 0.00612 0.0170 0.0059

44

Page 48: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

10 15 20 25 30Average load current [%]

0.5

1

1.5

2

2.5

Mag

netiz

ing

curr

ent [

A]

Day

Phase 0Phase 1Phase 2

10 15 20 25 30Average load current [%]

0.5

1

1.5

2

2.5

Mag

netiz

ing

curr

ent [

A]

Week

Phase 0Phase 1Phase 2

Figure 4.10: Magnetizing current as a function of average load current, Field Installation3.

Table 4.20: Magnetizing Current Uncertainty, day-period, Installation 3

Phase: Several measurements [A] Single measurements [A]0 0.1259 0.07731 0.1727 0.07382 0.2905 0.0763

Table 4.21: Magnetizing Current Uncertainty, week-period, Installation 3

Phase: Several measurements [A] Single measurements [A]0 0.2077 0.08161 0.1819 0.08042 0.4028 0.0770

45

Page 49: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

10 15 20 25 30Average load current [%]

82.5

83

83.5

84

84.5

85

Impe

danc

e im

agin

ary

part

[Ω]

Day

Phase 0Phase 1Phase 2Nominal

10 15 20 25 30Average load current [%]

82.5

83

83.5

84

84.5

85

Impe

danc

e im

agin

ary

part

[Ω]

Week

Phase 0Phase 1Phase 2Nominal

Figure 4.11: Imaginary part of impedance as a function of average load current, Fieldinstallation 3.

Table 4.22: Imaginary Impedance Uncertainty, day-period, Installation 3

Phase: Several measurements [Ω] Single measurements [Ω]0 0.3136 0.24741 0.0736 0.23622 0.2580 0.2451

Table 4.23: Imaginary Impedance Uncertainty, week-period Installation 3

Phase: Several measurements [Ω] Single measurements [Ω]0 0.3167 0.19811 0.0570 0.18702 0.2373 0.1940

46

Page 50: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

10 15 20 25 30Average load current [%]

2

2.2

2.4

2.6

2.8

3

3.2

3.4

3.6

3.8

Impe

danc

e re

al p

art [Ω

]

Day

Phase 0Phase 1Phase 2

10 15 20 25 30Average load current [%]

2

2.2

2.4

2.6

2.8

3

3.2

3.4

3.6

3.8

Impe

danc

e re

al p

art [Ω

]

Week

Phase 0Phase 1Phase 2

Figure 4.12: Real part of impedance as a function of average load current, Field instal-lation 3.

Table 4.24: Real Impedance Uncertainty, day-period, Installation 3

Phase: Several measurements [Ω] Single measurements [Ω]0 0.1602 0.24741 0.1833 0.23622 0.3178 0.2451

Table 4.25: Real Impedance Uncertainty, week-period, Installation 3

Phase: Several measurements [Ω] Single measurements [Ω]0 0.1608 0.19811 0.1521 0.18702 0.3136 0.1940

47

Page 51: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

4.2 Data rate

Transformer Explorer can acquire data for analysis with several different data rates.The minimum rate that can be used is 1 minute rate since this is needed to obtainstable voltage and current phasors. A similar study as in section 4.1 was conductedin order to find out if any difference between the used data rate exists. Here thesame set used for the 2-day study at field installation 1 was used, but with twodifferent data rates: 10 min and 30 min. A difference is especially observed for theturn ratio where the variation seem slightly lower when the higher rate is used. Nodifference is observed between the 10 min and the 1 min rate for the same periodsobtained in section 4.1.2. No significant difference can be seen for the impedance.

Table 4.26: Turn Ratio Uncertainty, 10 min rate, Installation 1

Phase: Several measurements Single measurements0 14.282e-04 4.4581e-041 7.4281e-04 4.1556e-042 8.9092e-04 2.3601e-04

Table 4.27: Turn Ratio Uncertainty, 30 min rate, Installation 1

Phase: Several measurements Single measurements0 14.981e-04 4.1995e-041 10.754e-04 4.3000e-042 10.281e-04 2.5093e-04

48

Page 52: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Table 4.28: Imaginary Impedance Uncertainty, 10 min rate, Installation 1

Phase: Several measurements [Ω] Single measurements [Ω]0 0.0464 0.07501 0.0670 0.07072 0.0634 0.0715

Table 4.29: Imaginary Impedance Uncertainty, 30 min rate, Installation 1

Phase: Several measurements [Ω] Single measurements [Ω]0 0.0665 0.06871 0.0501 0.06422 0.0772 0.0688

4.3 Measurement Report with Vattenfall

For a diagnostic method to be valuable for a transformer owner, the results needsto be presented in a concise and understandable way – one should be able to takeimportant decisions based on the outcome of the measurement. Most of the diag-nostic methods described in section 2.5 are most often presented to the customerin a report. Every method have some kind of standard template report where theresults are presented. The measured parameters are compared to standards pro-vided by, e.g., IEC and based on this, recommendations are provided. Most of thecustomers are not interested in the actual parameters them self; they want to knowif some actions needs to be taken or if the transformer can continue in service asbefore. For Transformer Explorer to be seen as an alternative to other more con-ventional methods, measurement results need to be presented in a similar way. Inthis chapter, the report constructed as a part of this thesis will be presented.

As already mentioned in the introduction the measurement report was draftedas a collaboration with a power utility. The initial version of the report was con-structed by the ABB part based on a 2-day measurement done with field data froma transformer. The other part was evaluating the report from a transformer ownerspoint of view and provided feedback for a updated version by ABB and so on. Thus,an iterative process was adopted.

4.3.1 The Report

The first step in constructing a measurement report is to define some general con-tents. It was decided that report should have the following properties:

The report should be generally informative and concise.

The measured quantities should be Ratio, impedance, magnetizing currentand power loss.

The uncertainty in each measurement should be proper specified.

Recommendations should be mentioned. The customer should not have toanalyze the results themselves.

49

Page 53: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Serve as a basis for future measurements.

All of these points except the last could appear to be more or less obvious for thereader. The latter one however is really essential. As shown from the results insection 4.1.2 the accuracy in a measurement can be really low. It is in that casevery important to have the results from earlier reports for comparison. The dataused to simulate the measurement was a two day period from installation 2. TheMATLAB program described in section 4.1.1 was used to calculate the uncertaintyand power loss.

The collaboration resulted in a final version of the report with the followingcontents

Executive Summary

Background.

Transformer data.

Method.

Results.

Conclusions.

In the background a brief description of what has been measured and why ispresented. In the next section available data on the transformer is presented. TheMethod describes how the measurement is carried out and what equipment thatwas used. In results the results are presented in a brief and concise way and finallyconclusions and when the next measurement should take place. In the followingsections each content will be described in more detail.

Executive Summary

Probably the most essential part of the report. The report is briefly summarizedand recommendations are highlighted. It should be possible for the customer tomake decisions from only reading this part.

Background

The background to the measurement is presented; what are measured and whatcauses have initiated the decision to perform a measurement. Since the used methodis entirely new, the reader are also shortly introduced to the concept.

Transformer Data

Here is basically the transformer name plate data presented in a table. Since turnratio and impedance is not directly visible here, these are presented in a separatetable with actual values for the tap positions during the measurement. If anyadditional information is available, e.g, Magnetizing current values from factorytests, these are also presented here.

50

Page 54: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Method

Here the measurement system is described, the measurement period and if there isany problems regarding the procedure that arose during the time. This section isdivided into the following subsections: System Description and Instrument Trans-formers. The first describes the different part of the system, e.g, specifications onDAC and current clamps. In the latter any information about the instrument trans-formers is presented in tables. Also the absolute error that could be introduced isspecified.

Any details about Transformer Explorer is not presented in the method section.This since a brief description is not enough to understand the concept. This willinstead be available in an appendix.

Results

The actual results from the measurement is here presented as condensed as possible,with one table for each measured quantity. The results are presented for two tapposition, since enough data was obtained for both tap 12 and 13 in this case. Thereis no guarantee however, that this is possible for other 2-day measurements asconcluded in section 4.1.2. In each table the measured value is presented for eachphase, together with: deviation from last measurement, deviation from nominal (ifavailable) and uncertainty. The most important parameter here is deviation fromlast measurement since the accuracy is low. The last quantity presented is powerloss. This is not provided in a table, instead a plot with power loss as a functionof current will be shown, both uncorrected and corrected. This is supplementedwith a table with the correction factors used. The method for correcting loss isstill under development and no uncertainty is thus shown (see chapter 7). Finally atable of the uncertainty and accuracy is shown. This is done as a comparison withthe IEC standard values for off-line measurement. This is the only reference valuesavailable up to date.

Conclusions

Conclusions based on the measurement results are presented together with recom-mendations. This could for example be that more advanced diagnostic methodsneed to be adopted for further investigation or just a new suggested time period tonext routine measurement.

51

Page 55: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

5 Analysis Improvements

When transformer explorer is used for monitoring purposes, the criteria controllingdecision to make a fit becomes much more critical. The program should no longerfit on all present data, instead the program must be able to judge autonomouslywhen the time is right to make a fit. This is done using a number of criteria thatmust be fulfilled. The most important from where program makes its judgment are:number of data in a fit, minimum load range in fit, minimum load in fit.

An additional factor is to exclude the 30% most deviating data. Whenever a fit ismade based on these criteria, some data points is always removed according to thespecified removal criteria. This is an efficient procedure for removing outliers thatare not normally distributed, e.g., disturbances that would otherwise affect the fit.A problem however is that the procedure has the tendency to also remove evenlydistributed data.

Another problem that has been discovered which is a special case of the first one,is that sometimes when the transformer changes tap position, the class from theprevious tap is almost but not entirely fulfilling all of the criteria settings. It couldfor example be that one or two data are missing to fulfill the minimum number ofdata. When the transformer suddenly changes back to the old tap position, the loadcould be quite different and thus the last data appear far away from the others. Ifthe last point causes the criteria to be fulfilled the resulting fit has been seen to bedeviating significantly in some cases.

5.1 Control of Data Distribution in Fit

One way to overcome the problem stated in the introduction to this chapter, wouldbe to have some kind of control over the distribution of data in the fits. There areseveral logics that could be implemented to achieve this. In this work, a methodwas proposed that simply controls the distance between each data points in thefit. The distance should not be larger than a certain part of the total distancebetween the minimum and maximum point, which could be specified by the user.This control will ensure a certain spread of the data going into the fit. It wasimplemented in a sub.vi inside the case controlling the fitting criteria (see Fig. 5.1).The algorithm calculates the difference between each value in ascending order, inthe low side current vector. The maximum obtained value is compared with thedistribution limit and if larger, no fit is performed.

A general example of the problem is shown in fig 5.2. The primary current isplotted as a function of secondary current according to Eq 3.2. Most of the datain this fit is evenly distributed except from one point at a significantly lower loadcurrent. This situation could for example happen when the transformer is returningto a previous tap position with already stored data from higher loads. Because ofthe nature of the least square method this point will have a lever effect on the slope

52

Page 56: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 5.1: The labVIEW code block controlling the distribution of data. The desiredvalue in percent is set by the user and is multiplied by the difference between the minand max value of the data corresponding to the low side current. The low side current isgiven by the ratioVar. The difference between each value in the vector in ascending orderis then calculated in the for loop and saved in a new vector. The maximum value is thencompared to the minimum distribution criteria, and if it is larger the result will be falseand no fit will be performed on the actual data set.

compared to a deviating point more close to the others. An outlier can in this casehave a large impact on the entire fit in a negative way. More evenly distributeddata will reduce the impact of the outlier. Fig 5.3 illustrates this situation withthe same data set but with distribution control introduced. The proposed methodwas tested on a known troublesome data set from installation 3 shown in Fig. 5.4,which shows the continuous ratio fits for tap 12. Here the transformer was shutdown for a few weeks and then reconnected at a different load situation operatingat tap 11. Somewhere around 23 December it returned to tap 12 and the fittingcriteria were fulfilled a few days later, resulting in significantly deviating fit. Acloser analysis of the data in this fit is shown in Fig. 5.5. The continuous fitsfor the same period but with a 15% distribution introduced is shown in Fig 5.4where the effects of the bad fit suppressed. Note that the fits in both cases seemto stabilize at another level after shut down more close to each other. A possibleexplanation is the higher load level when the transformer is reconnected showed inFig. 5.7. Thus, the effect is probably the same as in Fig. 4.3, i.e., load dependenterrors in the current transformers. Another example where the use of distributioncontrol also reduced the impact of outliers are shown in Fig. 5.8 and Fig. 5.9.

53

Page 57: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

0.05 0.052 0.054 0.056 0.058 0.06 0.062 0.064 0.066 0.068Secondary current [kA]

0.12

0.13

0.14

0.15

0.16

0.17

0.18

Prim

ary

curr

ent [

kA]

Figure 5.2: A general example of an outlier causing trouble. Note the single deviatingpoint at a significantly lower load current than the others.

0.05 0.052 0.054 0.056 0.058 0.06 0.062 0.064 0.066 0.068Secondary current [kA]

0.12

0.13

0.14

0.15

0.16

0.17

0.18

Prim

ary

curr

ent [

kA]

Figure 5.3: The data used in Fig. 5.2 but more evenly distributed, suppressing the badeffect on the fit caused by the deviating point at lower load.

54

Page 58: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 5.4: A situation where an unwanted fit was done after the transformer was shutdown for a few weeks. Note that the ratio seem to stabilize at and deviate less from eachother after the shut down. This is most likely due to the nonlinear error characteristic ofthe instrument transformers.

Figure 5.5: The data from the deviating turn ratio fit in Fig. 5.4.

55

Page 59: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 5.6: Same period as in Fig. 5.4 but with distribution control of 20% introduced.Note how the deviating fits in all three phases where suppressed but not entirely eliminatedsince the troublesome data will still have an impact.

Figure 5.7: The apparent power for the same period as in Fig. 5.4. Here the higherload after the reconnection explaining the convergence is confirmed.

Figure 5.8: A analyzed period with an observed deviating turn ratio fit in phase 1(white).

56

Page 60: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 5.9: The same data as in Fig. 5.2, with a distribution control of 20%. Note howthe deviating fit in Fig. 5.8 is avoided.

The distribution method was also tested on another data set from installation 3where a different kind of problem were discovered. Fig. 5.10 shows the continuousratio fits for the particular period. Here something significant happens to the es-timated turn ratio for all the three phases after 2014-02-07. Fig. 5.11 shows thesame period but with 15% distribution control. The overall result seem more stableand less fluctuating, however, the proposed method has no significant effect on theperiod with deviating turn ratio. A look at the reactive power delivered to thetransformer during the same period shows that significantly higher reactive powershown in Fig. 5.12 coincide with the deviating fits of Fig. 5.11. A dependence of thecurrent ratio on reactive power is however not expected from the simple transformermodel used. This observation could therefore warrant deeper investigation.

Figure 5.10: The turn ratio observed at installation during a month. Something happensto the fits for a few days which causes the ratio to deviate a lot from the reference.

57

Page 61: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 5.11: The same period with 15% distribution control introduced. The ratioseems more stable but the distribution has barley no effect on the significant deviations.

Figure 5.12: The reactive power going in to the transformer. The period of higherreactive power coincide with the period with deviating turn ratio fits.

58

Page 62: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

5.2 An Alternative Fitting Algorithm

The method which Transformer Explorer uses to exclude outliers is based on atechnique where the least square fit is performed twice. First to make an initialestimate, and then a second time, but this time the data points that deviate ac-cording to a specified criteria are removed before the final fit is made. This methodefficiently removes outliers and skewed data points in most cases. One problemhowever, arises when the fit must be made to very few data points where several ofthem are heavily disturbed (so called outliers). This is due to the fact that the leastsquare method is relatively sensitive to outliers (since it minimizes the least squaresum of all data in the fit). Thus, when the first fit is done containing disturbeddata this will heavily affect the slope of the resulting fit. Then, when the outliersare removed before the final fit some of the ”good” data will also be removed. Thisis unwanted since you always want to fit to as many data as possible when thereis noise present. A more robust technique that could be used is the Theil-Sen Es-timator [12]. This method is much less sensitive to outliers than the least squaremethod.

The method estimates the slope β by taking the median slope among all linestrough pair of sample points X and Y. It is given by

βn = Med

(Yi − Yj)(Xi −Xj)

: Xi 6= Xj, 1 ≤ i < j ≥ n

(5.1)

and the the intercept mn is estimated as

mn = Med(Y )− βnMed(X) (5.2)

where n is the number of data, and Med denotes the median.The basic idea is to use the Theil-Sen estimator for obtaining the first estimate

of the fit in the method described above. The estimate will then be unaffected byoutliers, which can be removed with minimum impact on the normal data beforedoing the final fit with the least square method.

For verification, the method was first implemented and tested in MATLAB. A setof 100 data points was used to generate a y vector according to the relation y = x,i.e., a straight line with slope one and intercept zero. Normally distributed noisewas added and some significant outliers. Fig. 5.13 shows the first test with thetwo methods on the data set, i.e., one fit with the Theil-Sen and one with the leastsquare method. The plot confirms that the fit performed with the robust methodis unaffected by the outliers, while the least square method trying to minimize allerrors, resulting in an less desirable fit.

59

Page 63: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

0 10 20 30 40 50 60 70 80 90 100x

0

20

40

60

80

100

yData pointsLeast Square FitTheil-Sen

Figure 5.13: The two estimators tested on the corrupted data set. Note how the Theil-Sen estimator is unaffected by the outliers while the least square methods slope estimateis clearly affected.

Finally the most deviating data were removed from the two fits (10% in this case).The final remaining data and the fit with the least square method is shown in Fig.5.14. The figure shows precisely what was predicted. The final fit has some of thegood data removed and two outliers remaining which will have a negative effect onthe fit estimate.

0 10 20 30 40 50 60 70 80 90 100x

0

20

40

60

80

100

y

Data pointsLeast Square Fit

Figure 5.14: Least square fit performed twice with outliers removed. Most of the outliersare removed with this method, however, several of the good data were also removed whichis undesirable.

In Fig. 5.15 the same thing is shown but with the Theil-Sen method to obtainthe first fit. 10% outliers is removed and the final fit is done with the standardmethod. Here one can see that all of the true outliers were efficiently removed butunlike the original method more of the good data was kept. This is desirable sincethe fit should be performed with as many good data points as possible.

Finally the proposed method was implemented in the sub.vi controlling the fitting,as a part of the transformer explorer LabVIEW program. The block diagram is

60

Page 64: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

0 10 20 30 40 50 60 70 80 90 100x

0

20

40

60

80

100

yData pointsTheil-Sen + Least Square Fit

Figure 5.15: The proposed method where the Theil-Sen is used on the original dataset, 10% most deviating data points are removed and finally the remaining data is fittedwith linear regression. Note how all of the outliers were removed, while much more of thegood data was kept in the fit.

shown in Fig. 5.16. The proposed method should be able to find application in two

Figure 5.16: The code block for the Theil-sen estimator. The main frame is a casestructure containing both the standard least square algorithm and the new method. Themain of the code is the double for loop which calculates all the slopes for each x and ypair and saves in a new vector. The median is then calculated in the main case. Thelower small case structure in the figure calculates the intercept according to Eq. 5.2.

61

Page 65: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

versions. Either as described above with outlier removal, or by simply performingthe a fit once with the new algorithm. The latter means that no concern has to betaken on how many outliers that should be removed. The method was tested onsome of the field installation data. One improved case were a significant differencecan be seen was observed with data from installation 2 when no removal of outlierswas done. The data set was known from earlier to have significant outliers deviatingup to 70 standard deviations. The data plotted in MATLAB is shown in Fig 5.17.A plot showing the difference to fit for the particular set is shown in Fig. 5.18.Fig 5.19 shows the ratio fit for tap 12 with the standard method. Fig 5.20 showsthe same but with the proposed method. The fits for the latter case appears morestable. This fact confirms that the implemented Theil-Sen algorithm is working andis less sensitive to the outliers.

0.026 0.028 0.03 0.032 0.034 0.036 0.038 0.04Secondary current [kA]

0.35

0.4

0.45

0.5

0.55

Prim

ary

curr

ent [

kA]

Figure 5.17: The data with observed outliers plotted in MATLAB. This is the totalamount of data used for the fits in Fig. 5.18.

Figure 5.18: Difference to reference fit for tap 12 with the troublesome data set frominstallation 2. Most of the data are deviating less than 5 standard deviations from thereference fit, while several outliers deviate up to 70 standard deviations.

62

Page 66: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Figure 5.19: Turn ratio fits for tap 12, performed with the standard method.

Figure 5.20: Turn ratio fits for tap 12 with the proposed method. The fits appear morestable compared to Fig. 5.19 where the standard method was used.

63

Page 67: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

6 Power Loss

As mentioned earlier the power loss of a transformer is simply given by the differencebetween the input power supplied by the primary side and the output power deliv-ered by the secondary side according to Eq. 3.5. In reality however, the measuredvoltage and current signals are always subjected to a certain amount of error. Theerror consists of a stochastic part associated with noise and a systematic (static)part associated with the accuracy of the measurement equipment, i.e., Instrumenttransformers and additional sensors. If one assumes that each sensor gives a valuewith some error ε, then the voltage and current signals provided by the sensors canbe written as

V = (1 + εV )V (6.1)

I = (1 + εI)I (6.2)

where ε is a complex number as the sensors can introduce an error in both theamplitude and the phase angle of the phasors. Because of the errors introduced bythe sensor the power loss become something quite different than the actual, sincethe measured loss becomes

Sloss = (1 + εV H)(1 + εIH)∗VHI∗H − (1 + εV L)(1 + εIL)∗VLI

∗L (6.3)

where the subscripts H and L refer to the high(primary) and low(secondary) voltageside of the transformer respectively.

The power loss of a typical power transformer is less than one percent and theaccuracy class of the voltage and current transformers is normally in the same or ahigher order of magnitude. As a consequence Eq. 6.3 is very ill-conditioned as itis formed as a difference between quantities that are approximately equal in size.This makes the equation very sensitive to errors and thus the active power loss canappear to decrease with load or even adopt negative values. The aim of this chapteris to present a method for minimizing the impact of the systematic errors on themeasured power loss.

The technique to obtain the power loss correction is currently subject for a patentapplication and can thus not presently be disclosed here. Instead, a few examplesindicating the effectiveness of the proposed correction method are presented in thefollowing section.

64

Page 68: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

6.1 Application Examples on Field Installation

Data

The method was implemented in MATLAB and tested with field data from Fieldinstallation 2. As can be seen in Fig. 6.1 and Fig. 6.2, for both the individual andtotal power loss, the sensors were so poorly matched that the loss became negative.Further, Fig 6.1 shows that the proposed method has the ability to turn the lossto positive values. The curves also got a more characteristic shape (parabolic) andthe power loss is also in the right order of magnitude (If 1 % loss is assumed, thisgives approximately 10 kW losses per phase at 0.14 p.u).

Figure 6.1: Uncorrected and corrected power loss for each phase at field installation 2.

Figure 6.2: Uncorrected and corrected total power loss at field installation 2.

65

Page 69: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

7 Conclusions & Discussion

7.1 Short Time Measurement

A comprehensive study was carried out with field data collected from three differentfield installations. This in order to answer the main questions: should a shorter ora longer measurement interval be used, and is the analytically derived uncertaintyvalid for these type of measurements. The study showed that there is no significantdifference between the two intervals and there is no crucial motivation for usingthe longer one. One difference that was observed however is that, the possibilityto analyze more tap positions is hindered due to lack of data with shorter intervalmeasurements. There is no assurance however that the enough data is collected foranalyzing more taps with the longer interval. Further, the study also shows thatthe assumed uncertainty seems to be valid and could be used with a certain safetymargin. One have to be aware of the ratings of current transformers in relation toload current since the errors seem to be load dependent at lower loads, which ofcourse could lead to the wrong conclusions from a measurement. Further one shouldalso be observant of the reactive power since this has an effect on the estimatedturn ratio.

A measurement report for Transformer Explorer was drafted together with aparallel thesis at Vattenfall. The collaboration resulted in a brief and concise reportwhere the most important results are presented in a straight forward manner andrecommendations is offered to the transformer owner. The report should be able tofunction as a first version when the concept is commercialized.

7.2 Analysis Improvements

Two methods to improve the existing fitting procedure were proposed. The firstone being an alternative fitting algorithm that is less sensitive to outliers and thesecond one controlling the distribution of data selected a fit. Both methods were im-plemented in LabVIEW and tested on field data. The alternative fitting algorithmseems to reduce the impact of outliers in a fit and could hence prevent unwantedfalse alarms that could occur during on-line monitoring. This method could be usedeither with outlier removal as the initial version, or just performing the Theil-Senfit once. In the latter, one doesn’t have to decide how many data that should beexcluded (and thus risking to remove too many of the good). The best solutionwould probably be to have some kind of logic that checks the residual distributionand only removes the data that deviates more than a certain amount.

The distribution control also showed similar effects and in several analyzed casesthe fit became more stable. This is believed to occur due to two reasons. Firstthe lever effect of outliers far away from the good data are minimized, second the

66

Page 70: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

distribution tends to get more data in to a fit. It also seems logical to have somedistribution of the data from the statistical point of view.

7.3 Power Loss

A method was developed for compensating the systematic errors introduced by theacquisition system, especially the instrument transformers. The results show that itis possible to compensate for the errors to a large extent. The power loss sometimesturned from negative to positive values and shows a positive slope against the loadcurrent. One can se a smaller difference between the loss of the individual phases.It is however difficult to draw any conclusions from this since the actual losses areunknown and very hard to measure; one can only make assumptions based on ratedpower loss. Further the loss can differ between phases in the transformer due to DC-bias in the core and similar effects. An other important aspect is the fact that thecurrent transformer ratio deviates from nominal at low currents (<10% of rated).All of the transformers available for this work were operating at relatively low loadcompared to the ratings of the current transformers. Trying to obtain correctionwhen the errors varies with the load will of course introduce additional errors.

In order to use the method for presenting the power loss in a report, the uncer-tainty should preferably be experimentally verified. Further, compensation withinthe phase group is difficult due to different zero sequence impedance on the primaryand secondary side.

7.4 Future Work

If short-time measurements should be offered as a service to customers some thingsregarding the report has to be automated. First of all, MATLAB should not beused for calculating the parameters, e.g., uncertainty. This should be calculated bythe Transformer Explorer program along with the other results. Further, it wouldbe preferable if the report is auto generated and the writer just complements withrecommendations and additional information.

If outlier removal is used this should not be a fixed value since it is difficultto set a optimum amount that not tend to exclude outliers but also normal data.The general amount of outliers present in the data are not the same for differenttransformers and even not for different fits. Instead one should implement a methodthat remove only data that deviates more than a curtain amount, e.g, a few standarddeviations.

Finally the power loss correction which showed promising result should preferablybe experimentally verified on a transformer where the actual loss can be measuredwith high precision or could be compared to factory acceptance tests.

67

Page 71: Power Transformer Monitoring and Diagnosis using ...912439/FULLTEXT01.pdf · Power Transformer Monitoring and Diagnosis using Transformer ... Power Transformer Monitoring and Diagnosis

Bibliography

[1] T. Bengtsson and N. Abeywickrama. On-line Monitoring of Power Trans-former by Fundamental Frequency Signals. Cigre paper A2-110, 2012.

[2] J.Duncan Glover, Mulukutla S. Sarma, and Thomas J. Overbye. POWER SYS-TEM ANALYSIS AND DESIGN. Cengage Learning, Stamford, 2012.

[3] Robert M. Del Vecchio, Bertrand Poulin, Pierre T. Feghali, DilipkumarM. Shah, Rajendra Ahuja. TRANSFORMER DESIGN PRINCIPLES. CRCPress, USA, 2002.

[4] ABB AB. Transformer Handbook. Zurich, 2004

[5] Martin J. Heathcote. The J & P Transformer Book. Newnes, Oxford, 1998.

[6] ABB AB Instrument Transformers Application Guide. Ludvika, 2009

[7] Picture from Academic. Url: http://en.academic.ru/dic.nsf/enwiki/18967.

[8] Stefan Tenbohlen Transformer Reliability Survey. Cigre, Germany, 2011

[9] Cardoso, A. J. M. and Oliveira, L. M. R, Condition Monitoring and Diagnosticsof Power Transformers. COMADEM’98, Australia, 1998

[10] C. Bengtsson Status and Trends in Transformer Monitoring. ABB Transform-ers AB, Ludvika, 1996

[11] Lars Bengtsson, Elektriska matsystem och matmetoder. Studentlitteratur,Lund, 2012

[12] Wikipedia, Theil-sen Estimator. Url: https://en.wikipedia.org/wiki/Theil-Sen estimator

68