32
A Guide To Diagnostic Insulation Testing Above 1 kV US $9.95 GBP £5.95

US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

  • Upload
    dotuyen

  • View
    229

  • Download
    0

Embed Size (px)

Citation preview

Page 1: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A Guide To Diagnostic Insulation Testing Above 1 kV

US $9.95GBP £5.95

michael ryan
MR STAMP 2
Page 2: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE
Page 3: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 1

TABLE OF CONTENTS

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

What is Insulation? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

What Causes Insulation to Degrade? . . . . . . . . . . . . .3

Electrical Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3

Mechanical Stress . . . . . . . . . . . . . . . . . . . . . . . . . . .3

Chemical Attack . . . . . . . . . . . . . . . . . . . . . . . . . . . .3

Thermal Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3

Environmental Contamination . . . . . . . . . . . . . . . .3

How Can Predictive Maintenance Help Me? . . . . . . .3

The Benefit of New Technology . . . . . . . . . . . . . . . . .4

How Insulation Resistance is Measured . . . . . . . . . . . . . 4

How an Insulation Resistance Tester Operates . . . . .4

Components of Test Current . . . . . . . . . . . . . . . . . . . .4

Capacitive Charging Current . . . . . . . . . . . . . . . . . .4

Absorption or Polarization Current . . . . . . . . . . . . .5

Surface Leakage Current . . . . . . . . . . . . . . . . . . . . .5

Conduction Current . . . . . . . . . . . . . . . . . . . . . . . . .5

Connecting your Insulation Tester . . . . . . . . . . . . . . .6

Selected Typical Connections . . . . . . . . . . . . . . . . . . . .6

Shielded Power Cable . . . . . . . . . . . . . . . . . . . . . . . .6

Circuit Breaker/Bushings . . . . . . . . . . . . . . . . . . . . . .6

Power Transformer . . . . . . . . . . . . . . . . . . . . . . . . . .7

AC Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7

Insulation Resistance Tester Scale . . . . . . . . . . . . . . . .7

Voltage Characteristics . . . . . . . . . . . . . . . . . . . . . . . . .8

Evaluation and Interpretation of Results . . . . . . . . . . . . 9

Interpretation of the Infinity Reading . . . . . . . . . . . .9

Diagnostic High Voltage Insulation Tests . . . . . . . . . . . . 9

Spot Reading Test . . . . . . . . . . . . . . . . . . . . . . . . . . . .10

Time vs. Resistance Test . . . . . . . . . . . . . . . . . . . . . . .11

Polarization Index Test . . . . . . . . . . . . . . . . . . . . . . . .12

Step Voltage Test . . . . . . . . . . . . . . . . . . . . . . . . . . . .13

Dielectric Discharge Test . . . . . . . . . . . . . . . . . . . . . .14

Different Problems/Different Tests . . . . . . . . . . . . . .15

Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

Potential Sources of Error/Ensuring Quality Test Results . . . . . . . . . . . . . . . . . . . . . . . . . .16

Test Leads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

Making Measurements above 100 GΩ . . . . . . . . .16

Accuracy Statements . . . . . . . . . . . . . . . . . . . . . . . .16

Delivery of Stated Voltage . . . . . . . . . . . . . . . . . . .16

Interference Rejection . . . . . . . . . . . . . . . . . . . . . .17

Rules on Testing and Comparing . . . . . . . . . . . . . .17

Testing Insulation Resistance of Rotating Machinery . . . . . . . . . . . . . . . . . . . . . . . .18

The Guard Terminal . . . . . . . . . . . . . . . . . . . . . . . . . .19

Effects of Temperature . . . . . . . . . . . . . . . . . . . . . . .20

Effects of Humidity . . . . . . . . . . . . . . . . . . . . . . . . . .21

Ingress Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . .22

High Potential Testing . . . . . . . . . . . . . . . . . . . . . . . .23

Current (nA) Readings vs. Resistance (MΩ) Readings . . . . . . . . . . . . . . . . . . .23

Burn Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23

Drying Out Electrical Equipment . . . . . . . . . . . . . . .24

Test Item Discharge . . . . . . . . . . . . . . . . . . . . . . . . . .25

Charging Time for Large Equipment . . . . . . . . . . . .26

Motor Driven Insulation Testers . . . . . . . . . . . . . . . .26

Megger 5 kV & 10 kV Insulation Testers . . . . . . . . . . . . 27

Page 4: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

2 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

INTRODUCTION

Electrical insulation degrades over a period of timebecause of various stresses, which are imposed upon itduring its normal working life. The insulation has beendesigned to withstand these stresses for a period ofyears, which would be regarded as the working life ofthat insulation. This often runs into decades.

Abnormal stresses can bring about an increase in thisnatural aging process that can severely shorten theworking life of the insulation. For this reason it is goodpractice to perform regular testing to identify whetherincreased aging is taking place and, if possible, toidentify whether the effects may be reversible or not.

The purpose of diagnostic insulation testing is:

n To identify increased aging.

n To identify the cause of this aging.

n To identify, if possible, the most appropriate actionsto correct the situation.

In its simplest form diagnostic testing takes the form ofa “Spot Test.” Most electrical maintenance professionalshave made spot tests where a voltage is applied to theinsulation and a resistance is measured. The diagnosis inthis case is limited to “the insulation is good” or “theinsulation is bad.” But having made this diagnosis whatdo we do about it? It’s a bit like going to the doctorwith a bad cough and the doctor simply telling you,“You’ve got a bad cough.” You wouldn’t be happy tocome away with only that information. You expect thedoctor to examine you, carry out a few tests, and tellyou why you have a bad cough and what to do about itto cure the cough.

In insulation testing, a spot test on its own is theequivalent of the doctor telling you that you are well oryou are sick. It’s minimal information. This is the sort oftest that is typically applied to low-voltage circuitswhere the cost of a failure is low and equipment can bereplaced easily and inexpensively. Since the equipmentbeing tested is low voltage equipment, these tests aretypically performed using a 500 or 1000 V test voltageand will be familiar to all electrical maintenancepersonnel.

However, if the doctor records the results of hisexamination and compares them with those fromprevious visits, then a trend might be apparent whichcould lead to medication being prescribed. Similarly, ifinsulation resistance readings are recorded andcompared with previously obtained readings, it may bepossible to see a trend and to prescribe remedial actionsif such are called for.

Diagnostic insulation testing at voltages above 1 kV isan area that is less familiar to many electricalmaintenance personnel. The purpose of this booklet,therefore, is to:

n Acquaint the reader with making diagnosticinsulation resistance tests.

n Provide guidelines for evaluating the results ofthese diagnostic insulation resistance tests.

n Introduce the benefits of multi-voltage testing athigher voltages.

A series of appendices are included at the end of thebooklet to provide the reader with additionalinformation related to diagnostic insulation testing.

This booklet is based on the principles established in thebooklet “A Stitch in Time… The Complete Guide toElectrical Insulation Testing” first published in 1966 bythe James G. Biddle Company.

WHAT IS INSULATION?

Every electric wire in a facility, whether it’s in a motor,generator, cable, switch, transformer, or whatever iscovered with some form of electrical insulation. Whilethe wire itself is a good conductor (usually made ofcopper or aluminum) of the electric current that powerselectrical equipment, the insulation must resist currentand keep the current in its path along the conductor .Understanding Ohm’s Law, which is expressed in thefollowing equation, is the key to understandinginsulation testing:

E = I x R

where,

E = voltage in volts

I = current in amperes

R = resistance in ohms

For a given resistance, the higher the voltage, thegreater the current. Alternatively, the lower theresistance of the wire, the more current that flows forthe same voltage.

No insulation is perfect (has infinite resistance), so somecurrent does flow along the insulation or through it toground. Such a current may be insignificantly small formost practical purposes but it is the basis of insulationtesting equipment.

So what is “good” insulation? “Good” means arelatively high resistance to current flow. When used todescribe an insulation material, “good” also means “the

Page 5: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 3

ability to maintain a high resistance.” Measuringresistance can tell you how “good” the insulation is.

What Causes Insulation to Degrade?

There are five basic causes for insulation degradation.They interact with each other and cause a gradual spiralof decline in insulation quality.

Electrical Stress

Insulation is designed for a particular application.Overvoltages and undervoltages cause abnormal stresseswithin the insulation, which can lead to cracking ordelamination of the insulation.

Mechanical Stress

Mechanical damage such as hitting a cable whiledigging a trench is fairly obvious but mechanical stressesalso may occur from running a machine out of balanceor frequent stops and starts. The resulting vibrationfrom machine operation may cause defects within theinsulation.

Chemical Attack

While you would expect insulation to be af fected bycorrosive vapors, dirt and oil can also operate to reducethe effectiveness of insulation.

Thermal Stress

Running a piece of machinery in excessively hot or coldconditions will cause over expansion or contraction ofthe insulation which might result in cracks and failures.However, thermal stresses are also incurred every time amachine is started or stopped. Unless the machinery isdesigned for intermittent use, every stop and start willadversely affect the aging process of the insulation.

Environmental Contamination

Environmental contamination covers a multitude ofagents ranging from moisture from processes, tohumidity on a muggy day, and even to attack by rodentsthat gnaw their way into the insulation.

Insulation begins to degrade as soon as it is put inservice. The insulation in any given application will havebeen designed to provide good service over many yearsunder normal operating conditions. However, abnormalconditions may have a damaging effect which, if leftunchecked, will speed up the rate of degradation andwill ultimately cause a failure in the insulation.Insulation is deemed to have failed if it fails toadequately prevent electrical current from flowing inundesirable paths. This includes current flow across theouter or inner surfaces of the insulation (surface

leakage current), through the body of the insulation(conduction current) or for a variety of other reasons.

For example, pinholes or cracks can develop in theinsulation or moisture and foreign matter can penetratethe surface(s). These contaminants readily ionize underthe effect of an applied voltage providing a lowresistance path for surface leakage current whichincreases compared with dry uncontaminated surfaces.Cleaning and drying the insulation, however, will easilyrectify the situation.

Other enemies of insulation may produce deteriorationthat is not so easily cured. However, once insulationdegradation has started, the various initiators tend toassist each other to increase the rate of decline.

How Can Predictive Maintenance Help Me?

While there are cases where the drop in insulationresistance can be sudden, such as when equipment isflooded, it usually drops gradually, giving plenty ofwarning if tested periodically. These regular checkspermit planned reconditioning prior to service failureand/or a shock condition.

Without a periodic testing program all failures willcome as a surprise, unplanned, inconvenient and quitepossibly very expensive in time and resources and,therefore, money to rectify. For instance, take a smallmotor that is used to pump material, which will solidifyif allowed to stand, around a processing plant.Unexpected failure of this motor will cost tens maybeeven hundreds of thousands of dollars to rectify ifdowntime of the plant is also calculated. However, ifdiagnostic insulation testing had been included in thepreventive maintenance program it may have beenpossible to plan maintenance or replacement of thefailing motor at a time when the line was inactivethereby minimizing costs. Indeed, it may have been thatthe motor could have been improved while it was stillrunning.

If advanced insulation degradation goes undetectedthere is an increase in the possibility of electrical shockor even death for personnel; there is an increase in thepossibility of electrically induced fires; the useful life ofthe electrical equipment can be reduced and/or thefacility can face unscheduled and expensive downtime.Measuring insulation quality on a regular basis is acrucial part of any maintenance program as it helpspredict and prevent electrical equipment breakdown.

This is particularly appropriate now when we considerthat large parts of the electrical network in the USA andEurope were installed in the 1950s in a burst of postwar

Page 6: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

4 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

investment. Some equipment is approaching the end ofits design life, while some has already exceeded it but isstill operating satisfactorily.

Since diagnostic testing is generally reserved for morecritical items we normally, but not always, find thatdiagnostic testers have voltage outputs of 5 or 10 kV ,these voltages being more suitable for testing the assetswhich themselves are usually medium voltage machines,cables, transformers, etc.

The Benefit of New Technology

Insulation testers date back to the early 20th centurywhen Sidney Evershed and Ernest V ignoles developedtheir first insulation tester (which developed in 1903into the Megger® range of testers).

In the early days, most instruments were hand-cranked.This limited their ability to carry out tests which took anextended time to complete, and limited the voltagestability to the operator’s ability to crank steadily. Later,these same instruments were capable of having anexternal motor drive added which helped with longduration tests but did very little to improve the voltagestability. However, the range of these instruments rarelyexceeded 1000 MΩ. The analog movements were veryheavy and actually served to damp out any transientevents.

The appearance of electronics and the development ofbattery technology revolutionized the design ofinsulation testers. Modern instruments are line orbattery-powered and produce very stable test voltagesunder a wide variety of conditions. They are also able tomeasure very small currents so that their insulationresistance measuring range is extended severalthousandfold into the teraohm (TΩ) range. Some caneven replace the pencil, paper and stopwatch, whichwere formerly used to manually collect results, byrecording data in memory for later download andanalysis. It is fortunate that these astonishingenhancements were made since the manufacturers ofinsulating material have been working hard also, withthe result that modern insulating materials now exhibitmuch higher resistances than those in the early 20thcentury.

Newer technology offers enhanced performance so thatestablished procedures can yield greater insights andnew methods can be made available. Moderninstruments deliver stable voltage over their fullresistance range, with microprocessor sensitivity in themeasuring circuit enabling measurements in the T Ωrange. The combination of stable voltage and enhancedsensitivity enables the tester to measure the minuscule

amounts of current that are passed by quality insulationin new, capital equipment. Accordingly, sophisticatedprocedures that rely on precise measurement have beendeveloped and may be easily implemented.

Now that the insulation tester isn’ t limited to valuesassociated with faulty or aged equipment, it can beused to pinpoint the test item’s position anywhere alongits aging curve. The “infinity” indication that is a delightto the repair technician represents a void to thediagnostician. Some instruments have diagnostic testspreprogrammed into their software and can run themautomatically, filling that void with valuable analyticaldata.

HOW INSULATION RESISTANCE IS MEASURED

How an Insulation Resistance Tester Operates

The Megger® insulation tester is a portable instrumentthat provides a direct reading of insulation resistance inohms, megohms, gigohms, or teraohms (depending onthe model chosen) regardless of the test voltageselected. For good insulation, the resistance usuallyreads in the megohm or higher range. The Meggerinsulation tester is essentially a high-range resistancemeter (ohmmeter) with a built-in dc generator.

The instrument’s generator, which can be hand-cranked,battery or line-operated, develops a high dc voltagethat causes several small currents through and oversurfaces of the insulation being tested. The total currentis measured by the ohmmeter, which has an analogindicating scale, digital readout or both.

Components of Test Current

If we apply a test voltage across a piece of insulation,then by measuring the resultant current and applyingOhm’s Law (R=E/I), we can calculate the resistance of theinsulation. Unfortunately, more than one current flows,which tends to complicate matters.

Capacitive Charging Current

We are all familiar with the current required to chargethe capacitance of the insulation being tested. Thiscurrent is initially large but relatively short lived,dropping exponentially to a value close to zero as theitem under test is charged. Insulating material becomescharged in the same way as a dielectric in a capacitor .

Absorption or Polarization Current

Absorption current is actually made up of up to threecomponents, which decay at a decreasing rate to a valueclose to zero over a period of several minutes.

Page 7: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 5

The first is caused by a general drift of free electronsthrough the insulation under the effect of the electricfield.

The second is caused by molecular distortion wherebythe imposed electric field distorts the negative charge ofthe electron shells circulating around the nucleustoward the positive voltage.

The third is due to the alignment of polarized moleculeswithin the electric field applied. This alignment is fairlyrandom in a neutral state, but when an electric field isapplied, these polarized molecules line up with the fieldto a greater or lesser extent.

The three currents are generally considered together asa single current and are mainly affected by the type andcondition of the bonding material used in theinsulation. Although the absorption current approacheszero, the process takes much, much longer than withcapacitive current.

Orientational polarization is increased in the presenceof absorbed moisture since contaminated materials aremore polarized. This increases the degree ofpolarization. Depolymerization of the insulation alsoleads to increased absorption current.

Not all materials possess all three components and,indeed, material such as polyethylene exhibits little, ifany, polarization absorption.

Surface Leakage Current

The surface leakage current is present because thesurface of the insulation is contaminated with moistureor salts. The current is constant with time and dependson the degree of ionization present, which is itself

dependent on temperature. It is often ignored as aseparate current, being included with the conductioncurrent below as the total leakage current.

Conduction Current

Conduction current is steady through the insulation andis usually represented by a very high value resistor inparallel with the capacitance of the insulation. It is acomponent of the Leakage Current, which is the currentthat would be measured when the insulation is fullycharged and full absorption has taken place. Note that itincludes surface leakage, which can be reduced oreliminated by the use of the guard terminal (to bediscussed later).

The graph in Figure 2 shows the nature of each of thecomponents of current with respect to time.

The total current is the sum of these components.(Leakage current is shown as one current.) It is thiscurrent that can be measured directly by amicroammeter or, in terms of megohms, at a particularvoltage by means of a Megger insulation tester. Someinstruments offer the alternatives of displaying ameasurement in terms of current or as a resistance.

Because the total current depends upon the time thatthe voltage is applied, Ohm’s Law (R = E/I) only holds,theoretically, at an infinite time (that implies waitingforever before taking a reading). It is also highlydependent upon starting from a base level of totaldischarge. The first step in any insulation test is,therefore, to ensure that the insulation is completelydischarged.

-+

-+

-+ -

+

+

+ ++ +

--- --

+

+-

-+

+- -

+

- +-

+++

-

--

+

StoredCharge

SurfaceLeakage

Dipoles

(A) (B)

Figure 1: Alignment of Polarized Molecules

1

10

100

1 10 100

Capacitive ChargingCurrent

TotalCurrent

Conduction or Leakage Current

AbsorptionCurrent

Cu

rren

t—

Mic

roam

per

es

Seconds (x10)

Figure 2: Components of Test Current

Page 8: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

6 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

Please note:

The charging current disappears relatively rapidlyas the equipment under test becomes charged.Larger units with more capacitance will takelonger to be charged. This current is storedenergy and, for safety reasons, must bedischarged after the test. Fortunately, thedischarge of this energy takes place relativelyquickly. During testing, the absorption currentdecreases at a relatively slow rate, dependingupon the exact nature of the insulation. Thisstored energy, too, must be released at the endof a test, and requires a much longer time todischarge than the capacitance charging current.

Connecting your Insulation Tester

With modern insulating materials there is little, if any ,difference in the reading obtained, regardless of whichway the terminals are connected. However, on olderinsulation, a little known phenomenon calledelectroendosmosis causes the lower reading to beobtained with the positive terminal connected to thegrounded side of the insulation being tested. If testingan underground cable, the positive terminal wouldnormally be connected to the outside of the cable sincethis will be grounded by contact with the soil, as shownin Figure 3. Please note that you do not connect directlyto the insulation but rather to the cable’ s neutral orground.

Selected Typical Connections

Shielded Power Cable

Connected to measure the insulation resistancebetween one conductor and ground.

Circuit Breaker/Bushings

Figure 3: Simplistic Connection to a Cable

Figure 4: Connection to a Shielded Power Cable

Figure 5: Connection to a Circuit Breaker

Page 9: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 7

Power Transformer

AC Generator

Keen observers will note that the hookup to measurethe circuit breaker bushing included the connection ofthe third, or Guard, terminal. The use of this terminal isexplained in greater detail later in this booklet.

With an analog display, the movement of the pointermay provide information to an experienced operator. Isthe pointer traveling smoothly, or “stuttering?” Is itrising steadily or intermittently dropping back? Thisvaluable supplementary information would be dif ficultor nearly impossible to discern from the dancing digitsof an LCD. A few examples are listed here:

n As the test voltage increases and the item undertest approaches breakdown, corona discharge willcause the pointer to “jitter,” indicating to theoperator that the maximum voltage that the itemcan withstand is being approached. This warninghappens in time to terminate the test before actualbreakdown, and possible damage, occurs.

n To the experienced operator, the speed at which thepointer travels imparts information on thecapacitance of the item under test. This is a usefulproperty in high-voltage cable testing, and relatesto the theoretical basis of the more sophisticateddielectric discharge test that is described elsewherein this booklet.

n If the pointer alternately rises and drops back, itcould indicate arcing in the item under test that istoo small to cause the automatic shutdown of thetester. Such information helps direct the operator inpinpointing a problem.

n Observing a pointer as it slows to an apparent halt(it may still be moving, but at a “speed” likened tothat of a clock hand) can be more agreeable totaking a quick or spot reading than trying to decidewhen a digital display has reasonably stabilized. Nodigital display “freezes” on a precise numberwithout at least some fluctuation of the leastsignificant digit.

This kind of detail is dif ficult or impossible for the eyeto extract from the scrolling digits on an electronicdisplay. But whereas pointer travel may be desirable,when it stops, the operator is left to interpolate thereading between the scale markings, introducing an

Figure 7: Connection to an AC Generator

Figure 8: Megger MIT520 Display

Figure 6: Connection to a Power Transformer

Insulation Resistance Tester Scale

Most modern insulation testers offer displays thatprovide the operator with both a digital readout of theresult and some form of analog readout. Figure 8 is arepresentation of the Megger MIT520 display.

When an insulation tester is “hooked up” to the item tobe tested, and a test is started, several things occur . Thethree different currents, capacitive charging, dielectricabsorption, and conduction/leakage are flowing. Thesum of these three currents will cause the instrumentdisplay to vary with the reading increasing, initiallyquickly and then more slowly with time.

Page 10: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

8 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

element of judgment, which can be a source of error .Digital models present no such problem, as they informthe operator exactly (within the unit’s accuracyspecification) what measurement has been taken. Andremember, most will give you a value of capacitance atthe end of the test.

Most Megger insulation testers above 1 kV come withan analog/digital display. One of the advantages of thisdisplay is that the analog portion of the meter will swayand oscillate, indicating to the operator that the itemunder test has not yet reached a steady state and is stillunder the influence of the absorption and chargingcurrent. This indication means that the item should betested longer or that there is a problem. When theanalog portion of the display becomes steady, theinstrument displays the result in an unambiguous digitaldirect reading form, with no multipliers or math toperform.

Unlike the analog/digital display mentioned above, an“average sensing” bar graph meter does not provide areal-time indication of insulation resistance. Someinstruments offer a curved bar graph in place of agenuine logarithmic arc, in which the low end of thescale is expanded relative to the high end. The bargraph takes readings over time, performs calculationsand then displays the results. The problem with thistype of meter is its principal of operation. If an eventoccurs when the bar graph is not taking readings, it willbe missed and not shown on the display . Additionally,bar graph simulations of pointer travel may not appearto the eye the same as the familiar pointer travel andmay not replicate a mechanical movement to theexpected degree.

When doing insulation testing, the more the operatorknows about the results (during and after the test), thebetter his/her decision on how to correct the problem, ifone exists. If something is missed during a test becausethe instrument had a bar graph style meter, importantinformation could also be missed.

Voltage Characteristics

The output voltage of an insulation tester depends onthe resistance it is measuring. At low resistances, saytens of ohms, the output voltage will be close to zero,maybe a few volts. As the resistance load is increased sothe test voltage will increase until it reaches therequested voltage. As the resistance increases further,the test voltage will slowly increase until a steady valueis reached. This value will probably be slightly in excessof the requested nominal voltage (e.g. 5104 V when5000 V was selected).

You should always ensure that an insulation tester isprovided with a “load graph” that indicates outputvoltage characteristics against load resistance or,alternatively, an integral voltmeter that actuallymeasures the terminal voltage during a test and displaysit continuously. By this means you can ensure that anadequate voltage is produced over the resistance rangeof interest.

A quality insulation tester will have a voltagecharacteristic that exhibits a sharp rise in voltage up to alevel of resistance commensurate with good insulation.A fast rise time ensures an effective measurement. Thevoltage characteristic shown in Figure 9 represents agood characteristic. In this example, the output voltagewill have reached 500 V at a load as low as 500 k Ω and1000 V by 1 MΩ. These values are legislated byinternational standards for testing wiring in houses,shops, etc. While this is hardly a typical use for typicaldiagnostic insulation testers, it does provide a goodbenchmark for the serious manufacturer. Similar figureswould apply at higher voltages. Voltage should risesharply up to anywhere from one to five megohms,depending on the voltage selection, and maintain thatvoltage at all higher resistances.

With lower quality insulation testers, voltage ramp is farslower. The instruments typified by the poor curveshown in Figure 10 do not produce the rated voltageuntil much higher resistances have been reached. Thustests could produce results that provide pass levels ofinsulation but have only been subjected to half thedesired test voltage.

Resistance

100 k0 1 M 10 M 100 M 1 G 10 G 100 G

Figure 9: Good Load Curve

Page 11: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 9

EVALUATION AND INTERPRETATION OF RESULTS

Interpretation of the Infinity Reading

One of the most important features of an insulationtester is the range that the instrument can measure.Testing goals determine whether basic function is allthat is needed, or enhanced range is recommended.Simple proofing applications, such as an electriciansigning off a job, can be met with a basic range of athousand megohms (MΩ). Admittedly, new equipment,if not defective or damaged during installation, willover-range all but the most advanced testers, however,this is okay. In such cases, the electrician is not lookingfor an actual value, but rather wants to see a high valueand “infinity”certainly meets that criterion. However,“infinity” is not a measurement; it is an indication that

the insulation being tested has a resistance that exceedsthe measuring capabilities of the tester and shouldalways be recorded as “greater than 1000 M Ω” orwhatever is the highest available number on yourinsulation tester. Usually this is adequate since theminimum acceptable value of resistance is likely to bemuch lower than the maximum reading available.

But for maintenance of capital equipment, a tester withonly a limited range is “shortchanging” the operator.For preventive/predictive maintenance, infinity readingsare of no use. The operator knows that the test item is“good”, but not much more. Testers with extendedrange, up into teraohms (1 T Ω = 1,000,000 MΩ), affordactual measurements right from the time of installation,thereby establishing a long time line that gives themaintenance professional plenty of “breathing room.”

Significant changes in insulation quality can occur athigh levels of insulation resistance, beyond the range ofmore limited instruments, as shown by graph in Figure 11.

In this example, a limited range tester would notcapture this valuable data. We can clearly see that,although the last recorded insulation value is in excessof 10 GΩ, the rate of decline is increasing; something iswrong. An instrument with a range limited to 2000 M Ωwould miss this totally. By the time the readings haddegraded into the instrument’s range, the maintenanceperson would be left with comparatively little time toschedule routine off-line maintenance. (It may even betoo late to rectify the fault condition.)

DIAGNOSTIC HIGH VOLTAGE INSULATION TESTS

Diagnostic insulation tests electrically stimulate theinsulation and measure the response. Dependent uponthat response, we can draw some conclusions about thecondition of the insulation.

Diagnostic insulation testing covers a very wide range oftechniques, some of which involve portable equipmentand some that require considerable fixed equipment.Here we shall consider only those tests that may beperformed with a readily portable dc insulation tester.These are:

n Trending spot tests

n Time constant

n Polarization Index (PI)

n Step Voltage (SV)

n Dielectric Discharge (DD)

Resistance

100 kΩ0 1 MΩ 10 MΩ 100 MΩ 1 GΩ 10 GΩ 100 GΩ

1 kV

0

2 kV

3 kV

4 kV

5 kV5000 V

2500 V

1000 V

Vo

ltag

e

Figure 10: Poor Load Curve

Year #1 Year #2 Year #3 Year #4 Year #5 Year #6 Year #70

1 M Ω

10 MΩ

100 MΩ

1 GΩ

10 GΩ

100 GΩ

1 TΩ

1 kV, 2 GΩ testerwould indicate “∞”

Insu

lati

on

Res

ista

nce

Figure 11: Changes in Insulation Resistance at High Values

Page 12: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

10 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

Each test gives a different view, or window, into thecondition of the insulation; the whole picture is onlyavailable when all required tests have been completed.

Spot Reading Test

The spot reading test is the simplest of all insulationtests and the one most associated with lower voltageinsulation testers; the test voltage is applied for a short,specific period of time (typically 60 seconds as usuallyany capacitive charging current will have decayed bythis time) and a reading is then taken. The reading canthen be compared to the minimum installationspecifications. Unless the result is catastrophically low, itis best used when trended against previously obtainedvalues.

However, insulation resistance is highly temperaturedependent, and thus the results should be corrected toa standard temperature, usually 40º C. Whiletemperature effects will be covered later, a good rule ofthumb is that for every 10º C increase in temperature,the current doubles (resistance halves). The key tomaking the spot reading test valuable is consistenttimekeeping, effective record keeping, and trending ofresults.

As noted previously, the increased sensitivity available inmicroprocessor-based diagnostic insulation testersallows the operator to identify insulation problems intheir early stages rather than when those problemsbecome catastrophic. In many cases, the trend is farmore important than the absolute value.

Compare the two traces in Figure 12. Apparatus “A”shows a high insulation resistance while Apparatus “B”shows a low value. However, when the trend isexamined, Apparatus “B” shows little cause for concern;it has been around the same value for several years andshows every prospect of continuing in the same vein formany years to come. Conversely, the curve for Apparatus“A” is diving dramatically and the apparatus will, ifnothing is done to prevent it, fail within the next fewyears.

While Apparatus “A” has much higher absoluteresistance values than Apparatus “B,” the trend is quiteworrying. Apparatus “B” has a fairly consistent flattrend, indicating that the insulation quality is probablyacceptable.

Insulation resistance readings should be consideredrelatively rather than absolutely. They can vary widelyfor one motor or machine tested three days in a row ,yet not mean bad insulation. As mentioned, theimportant information is the trend in readings over atime period, showing lessening resistance and warningof coming problems. Periodic testing is, therefore,critical to preventive maintenance of electricalequipment. The interval between tests (monthly, twice ayear, once a year, etc.) depends upon the type, location,and importance of the equipment. Evaluating a series ofreadings taken over a number of months or years movesthe operator toward being a diagnostician.

Periodic tests should be made in the same way eachtime. Use the same test connections and apply the sametest voltage for the same length of time. T ests shouldalso be made at about the same temperature, or theoperator must correct them to the same temperature. Arecord of the relative humidity near the equipment atthe time of the test is helpful in evaluating the readingand trend as low temperatures and high humidity mightsuggest condensation on the surface of the insulation.For this reason it is essential to ensure that equipmentto be tested is at a temperature in excess of the dewpoint, as otherwise, condensation will form which willdistort the readings unless the measurement is well“guarded.” More on this later.

The following table contains some general observationsabout how to interpret periodic insulation resistancetests and what should be done with the result.Year #1 Year #2 Year #3 Year #4 Year #5 Year #6 Year #7

0

1 M Ω

10 M Ω

100 M Ω

1 G Ω

10 G Ω

100 G Ω

1 T Ω

Insu

lati

on

Res

ista

nce

Apparatus "B"

Apparatus "A"

Figure 12: Comparison of Trended Test Results

Page 13: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 11

Time vs. Resistance Test

Familiar, standardized test procedures that have beenemployed for years benefit from the improvedcapabilities of enhanced diagnostic testing. Most basicof these is the time-resistance method. A valuableproperty of insulation, but one that must beunderstood, is that it “charges” during the course of atest thanks to the movement of electrons as explainedpreviously. This movement of electrons constitutes acurrent.

Its value as a diagnostic indicator is based on twoopposing factors; the current dies away as the structurereaches its final orientation, while “leakage” promotedby moisture or deterioration passes a comparativelylarge, constant current. The net result is that with“good” insulation, leakage current is relatively smalland resistance rises continually as current decreasesfrom the effects of charging and dielectric absorption.Deteriorated insulation will pass relatively largeamounts of leakage current at a constant rate for theapplied voltage, which will tend to mask the chargingand absorption effects.

Graphing the resistance reading at time intervals frominitiation of the test yields a smooth rising curve for“good” insulation, but a “flat” graph for deterioratedequipment. The concept of the time resistance test is totake successive readings at specified times. It is based onthe relative magnitudes of leakage and absorptioncurrents in clean, dry insulation compared to that ofmoist or contaminated insulation. Good insulationshows a continual increase in resistance over time. Withcontaminated insulation, the leakage current is muchlarger and the effects of the absorption current are,therefore, much less apparent.

The benefits of the time resistance test are that it isrelatively independent of temperature and can giveconclusive information without the records of past tests.

0 10 min

Time

0

Insulation

probably okay

A

B

T 1 T 2 T 3 T 4 T 5 T 6

Meg

oh

ms

Moisture/dirtmay be present

Figure 13: Time Resistance Test Graph

Condition What To Do

a) Fair to high values and n No cause for concern. well maintained

b) Fair to high values but n Locate and remedy theshowing a constant cause and check the tendency towards downward trend.lower values

c) Low but well n Condition is probably maintained all right but cause of

low values should be checked. May simply be the type of insulation in use.

d) So low as to be unsafe n Clean, dry out, or otherwise raise the values before placing equipment in service (test wet equipment while drying out).

e) Fair or high values pre- n Make tests at frequent viously well maintained intervals until thebut showing sudden cause of low values is lowering located and remedied

or,

n Until the values have become steady at a lower level but safe foroperation or,

n Until values become so low that it is unsafe to keep the equipment inoperation.

Page 14: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

12 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

Polarization Index Test

The simplest implementation of the time resistance testfor solid insulation is represented by the popularPolarization Index (PI) test, which requires only tworeadings followed by a simple division; the one-minutereading is divided into the ten-minute reading toprovide a ratio. The result is a pure number and cannormally be considered independent of temperaturesince the thermal mass of the equipment being tested isusually so great that the overall cooling which takesplace during the 10 minutes of the test is negligible.

In general, a low ratio indicates little change, hencepoor insulation, while a high ratio indicates theopposite. References to typical PI values are common inthe literature, which makes this test very easy andreadily employed. However, we say “in general”because as mentioned previously there are materialsthat exhibit very little or no dielectric absorption.Carrying out a test on these materials would thenproduce a result very close to 1.

Note that resistance readings alone are dif ficult to workwith, as they may range from enormous values in newequipment down to a few megohms just beforeremoval from service.

A test like the PI is particularly useful because it can beperformed on even the largest equipment, and yields aself-contained evaluation based on relative readingsrather than absolute values. But no PI can be calculatedwith a tester of limited range, because “infinity” is nota number! Advanced testers reach the teraohm range,and therefore, do not run off the graph. The largestand newest capital equipment can be readily tested toyield repeatable data for recording and subsequenttrend evaluation. The following chart highlightsselected PI values and what they mean to the operator .

Values above 4 indicate excellent equipment for whichno action is likely to be necessary within the immediatemaintenance schedule. The operator may be called uponto make critical judgments, however. Some high valuesof PI (above 5) could indicate brittle or crackedinsulation; this should be fairly obvious. A suddenincrease in PI greater than 20%, without anymaintenance having been performed, should serve as awarning; insulation may hold its value for long periods,but is not likely to dramatically improve all by itself.

A benefit of the PI test is that it can provide anindication of insulation quality in ten minutes on verylarge pieces of equipment that might take an hour ormore to fully charge. With a spot reading test, theoperator would have to wait until the readingstabilized. For this reason it is normal to conduct a PItest at relatively low voltage before applying the highvoltages typically applied for a withstand test.

Although the PI value table has been used for manyyears and is well accepted, PI readings can occasionallybe encountered which are exceptional. Many years agothe freshly cooked stator of a 3750 kV A generator wastested and a PI of 13.4 was obtained. The stator hadcooled down and no doubt was still in its curing phase.Subsequent tests yielded reducing PI values until itstabilized around 4.7. During routine maintenance, PIvalues do not reach these heady heights.

Time (minutes)

1 10 50

PI=?Res

ista

nce

Figure 14: Benefit of the Polarization Test for Large Equipment

Polarization Index Insulation Condition

<1 Poor

1-2 Questionable

2-4 Okay

>4 Good

Page 15: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 13

or unusual resistance reduction is an indication ofincipient weakness. Modern electronics allows thesereadings to be captured automatically.

Following are some possible results from a Step V oltagetest on a motor from 500 to 2500 volts and what theymean to the operator:

n No appreciable difference in values - Insulation is inreliable condition.

n Appreciable difference in values - Insulationrequires more thorough reconditioning.

n Insulation fails at 2,500 V - Motor is in question;would most likely fail in service even if an attemptwere made to recondition it on the basis of low-voltage tests only.

The graphs in Figure 15 are taken from a motor thatwas damp and dirty (lower trace) and after cleaning anddrying (upper trace).

In general, if a deviation of 25% in resistancemeasurements is observed over the range of successivevoltages, it is an indication of the presence of moistureor other contamination. Localized physical damage maybe further revealed by breakdown or arcing. A“stuttering” or “jittery” pointer movement cananticipate this condition as the breakdown voltage isneared. It may be desirable to terminate the test at suchpoint before insulation breakdown further deterioratesthe condition of the test item.

Like the PI test, the Step Voltage test is a repeatable,self-evaluating test that, because of its short duration, isfree of extraneous influences like temperature ef fect.

It is also interesting to note that many people havetried to use the PI test on oil-filled transformers andcannot understand why a known good transformergives them results close to 1. The answer is simple. PItesting is not appropriate for oil-filled transformers. Theconcept depends on the relatively rigid structures ofsolid insulating materials, where absorption energy isrequired to reconfigure the electronic structure ofcomparatively fixed molecules against the appliedvoltage field. Because this process can go to atheoretical state of completion (at “infinite time,”which obviously cannot be achieved in the practicalfield, but can be reasonably approximated), the result isa steady diminution of current as molecules reach their“final” alignment. Because the PI test is defined by thisphenomenon, it cannot be successfully applied to fluidmaterials since the passage of test current through anoil-filled sample creates convection currents thatcontinually swirl the oil, resulting in a chaotic lack ofstructure that opposes the basic premise upon whichthe PI test rests.

Step Voltage Test

Since good insulation is resistive, an increase in testvoltage will lead to an increase in current with a resultthat the resistance remains constant. Any deviationfrom this could signify defective insulation. At lowertest voltages, say 500 V or 1000 V, it is quite possiblethat these defects might be unobserved, but as thevoltage rises we reach a point where ionization cantake place within cracks or cavities, resulting in anincrease in current, and therefore a reduction in theinsulation resistance. Note that it is not necessary toreach the design voltage for the insulation for thesedefects to become apparent, since we are simplylooking for ionization in the defect.

The Step Voltage test follows exactly this principle andcan be employed usefully at voltages reaching 2500 Vand upwards. The Step Voltage test may be employedas an undervoltage or overvoltage test. However, itmust be remembered that an overvoltage test can leadto a catastrophic failure if the insulation breaks downbecause high voltage test sets have a lot of poweravailable. An undervoltage test carried out by aninsulation tester has relatively little power available andit is therefore far less likely to result in a destructivetest.

A recognized standard procedure is to increase voltagein five equal steps at one-minute increments and recordthe final insulation resistance at each level. Any marked

1000

500

250

100

50

0.5 1.0 2.5

Applied Voltage (kV)

good insulation condition

poor insulation condition

1.5 2.0

Meg

oh

ms

Figure 15: Step Voltage Step Graph

Page 16: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

14 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

Dielectric Discharge Test

The Dielectric Discharge test (DD) is a relatively new testmethod that was developed by EdF, the national powerutility of France, and based on years of research. Whilethe other methods mentioned measure the currentsflowing during the charging process, the DD testmeasures the current that flows during discharge of thetest sample. As such, it is not a pure insulation resistancetest but rather an adjunct to traditional insulation tests.

The charge that is stored during an insulation test isautomatically discharged at the end of the test whenthe insulation tester’s discharge resistors are switchedacross the terminals.

The rate of discharge depends only on the dischargeresistors and the amount of stored charge from theinsulation. However, the capacitive charge is dischargedrapidly until the voltage across the insulation hasreduced to almost zero. At that time, the ef fect ofleakage currents will be negligible. So only the reversalof dielectric absorption is left. This is known as dielectricreabsorption and is a mirror image of the dielectricabsorption.

The capacitive current quickly decays from a high valuewith a relatively short time constant (a few seconds).The absorption (or reabsorption during a discharge)current always starts at a high level but has a muchlonger time constant (up to many minutes). It is causedby the dipoles randomizing their alignment within theinsulation and the electron shell returning to anundistorted shape. This has the effect of a currentflowing if the discharge circuit is still connected, or avoltage reappearing on the sample if it is left opencircuit. Rapidly removing the effects of leakage andcapacitive currents allows the possibility of interpretingthe degree of polarization of the insulation and relatingit to moisture and other polarization ef fects.

The test item is first charged for anywhere from 10 to30 minutes at high voltage until full absorption hastaken place. (The Megger insulation testers thatautomate this test charge the test sample for 30minutes.) At this time, capacitance is fully charged andthe dielectric absorption is essentially complete. Onlyleakage current continues to flow. At this point the testvoltage is removed and the insulation is dischargedthrough the instrument’s internal discharge resistors toquickly discharge the capacitive charge. After 60 secondsof discharge, any remaining current flow is measured.At this time, the capacitance has been discharged andthe voltage has collapsed so that the charge stored inthe dipoles can be viewed independently of the“masking” currents that are dominant during thecharging phase of an insulation test.

The measured results are then entered into thefollowing formula and an index is calculated.

Current flowing after 1 minute (nA)

Test Voltage (V) x Capacitance (µF)

The measurement is temperature dependent, so it isimportant to test at a reference temperature or torecord the temperature.

Insulation in high voltage equipment often consists oflayers, each having its own capacitance and associatedleakage resistance. When insulation is built up in thisway, the aim is to make each layer such that the voltagestress is shared equally between layers. When theinsulator is discharged, each layer’s charge will decreaseequally until there is no voltage remaining.

Capacitive

Current

Absorption

Current

Total

Current

0.1 1.0 10.0 100.0

0.1

1.0

10.0

100.0

1000.0

Time (sec)

Cu

rren

t(µ

A)

Figure 17: Reabsorption Currents

+-

-+

+- -

+

- +-

+++

-

--

+Discharge

Resistor

Figure 16: Discharge of Test Item’s Stored Charge

Page 17: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 15

When a layer is faulty between good layers, its leakageresistance will decrease while capacitance is likely toremain the same. A standard insulation test will bedetermined by the good layers, and not likely to revealthis condition. But during dielectric discharge, the timeconstant of the faulty layer will mismatch the others toyield a higher DD value. A low DD value indicates thatreabsorption current is decaying quickly, and the timeconstant of each layer is similar. A high value indicatesthat reabsorption exhibits long relaxation times, whichmay point to a problem.

Typical conditions from practical research, primarilycarried out on generators by EdF, arrived at the figuresof merit in the following table. This technique wasdeveloped for high voltage generators but hasapplication on any multilayered insulation.

Different Problems/Different Tests

As we have just seen, the Dielectric Discharge T est canbe used to identify problems in a single layer ofmultilayer insulation. Other test methods might notpoint to problems on this specific type of insulatingstructure. Similarly, the Polarization Index test isparticularly valuable in revealing moisture ingress, oilsoaks, and similar pervasive contamination. Theseinvading contaminants provide convenient paths forelectrical leakage, which damages the surroundinginsulation and eventually burns through as a “short.”This type of problem is revealed at almost any testvoltage and will appear as a characteristically “flat” PI.Moisture and contaminants will also bring down thereadings, but this requires a previous value forcomparison; the PI test has the advantage of making aninternal comparison.

However, other problems may seem to “pass” a PI orsimple Spot Reading test by yielding high resistancevalues at a given voltage. Such problems includelocalized physical damage like pinholes or dry, brittleinsulation in aged equipment. Step voltage tests revealsuch problems. Increasing numbers of imperfections willpass current as higher and higher voltage is applied, andbe reflected in a declining resistance. Higher voltage willpull arcs across small air gaps, thereby providing an“early warning” of an incipient problem. As equipmentages, such gaps can narrow by accumulation of dirt andmoisture until a short to ground develops.

DD Value (in mA V-1F-1) Insulation Condition

> 7 Bad

4 - 7 Poor

2 - 4 Questionable

< 2 OK

Page 18: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

16 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

APPENDICES

Potential Sources of Error/Ensuring Quality Test Results

The following section identifies several areas ofpotential error in insulation testing above 1 kV. Thesefactors may be of less importance in 1 kV testing, butincreased voltages and sensitivities make them criticalfor higher voltage testing.

Test Leads

Beware of instruments with low quality leads whosevoltage rating is less than the test voltages employed. Itis extremely important that the only leakage currentsduring a measurement are those that are developed bythe insulation under test. If the leads themselvesproduce leakage, you may be measuring lead insulationresistance rather than the item under test.

All leads supplied with Megger insulation testers arehigh quality leads, which have been tested to withstandvoltages well above the highest test voltage generatedby the particular instrument. Even then, it is importantto reduce stray leakage by preventing the leads fromcontacting each other, the ground and particularlywater.

Making Measurements above 100 GΩMeasurements up to 100 GΩ can be made without anyspecial precautions, assuming that the leads arereasonably clean and dry. The guard (to be discussedlater) can be used to remove the ef fects of surfaceleakage if necessary. Greater precautions are requiredwhen measuring resistances above 100 G Ω as strayleakage current can spoil the quality of the readingstaken. Be aware of the following:

n Test leads should not be allowed to touch eachother or any other object since this will induceleakage paths.

n Sharp points at the test lead connections should beavoided since this will encourage corona discharge.

n Instrument test jacks should be deep so thatunwanted leakage does not occur between theterminals.

Accuracy Statements

Pay close attention to an insulation tester’ s accuracystatement. Do not accept a mere plus/minus percentagefor digital units. The statement must also includeplus/minus a number of digits, as no digital display canfix its last digit (least significant digit, or l.s.d.) to asingle number. Accuracies specified as “percent ofreading” indicate the same error at all points on thescale.

Analog statements listed as “percent of scale” or “fullscale deflection” (f.s.d.) can be deceptive. Because theaccuracy interval is based on the full-scale length, ittranslates into an increasing percentage error as thereadings rise against a logarithmic scale. In other words,the same number of pointer widths on the expandedlow end of the scale will account for only a fewmegohms, while on the contracted upper end, this willbe hundreds of megohms. Therefore, when meeting adesired or required accuracy spec, don’t stop at thepercentage statement but also examine the terms.

Delivery of Stated Voltage

Voltage regulation is indicated for an insulation testerwith a load graph in the instruction manual showingthe output voltage against resistance load. The loadcurve ensures that, at typical insulation resistancevalues, the insulation tester is delivering full rated testvoltage to the test item. While this may appear to beobvious, it is not necessarily the case unless so stated bythe manufacturer of a given tester. A poorly-regulatedtester may load down under a high-resistance load sothat the insulation of the test item may actually beexperiencing only a fraction of the rated test voltage,which the transformer can output only under maximumconditions. Such instrumentation is not likely to comeprovided with a load curve.

It was this condition that inspectors from specifyingagencies, like UL®, discovered among “testers” that were“jury-rigged” from on-hand transformers and othercomponents at job sites to perform high potential tests.The inadequacies of such systems lead to the highlyspecific language pertaining to output voltage that nowcommonly appears in the standards literature. Meggerinsulation testers conform by delivering and maintainingthe rated test voltage once a minimum load

Page 19: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 17

commensurate with typical insulation values (generally1 to 10 MΩ, depending on model and voltage selection)is applied. Test voltage is typically a few volts aboverated, but should not drop below it, maintaining theintegrity of the test and the repeatability whenperforming scheduled preventive maintenance. Ifexceptionally precise reporting data is mandated, somemodels display the actual test voltage in addition to theselected voltage and this information is included amongthe data provided at the conclusion.

Interference Rejection

Interference is the electrical noise produced at a varietyof frequencies, which can appear in the sample beingtested. It is usually induced currents or voltages fromadjacent equipment and is very common in substations,particularly high voltage substations where powerfrequencies predominate. This electrical noisesuperimposes an ac signal on the dc test current and cancause considerable variations in readings and mayprevent the operator getting a reading at all if it isbeyond the capabilities of your instrument. As anexample, 4 mA of 50/60 Hz noise is fairly typical of theelectrical noise that can be encountered in largesubstations (400+ kV).

Be aware of the capability of the insulation tester beingused to cancel out the effects of this ac noise ef fectively,resulting in the ability to make measurements inincreasingly more difficult conditions. Not all noise islimited to power frequencies, however. Toaccommodate other frequencies some top of the rangeinstruments incorporate further software filters that caneliminate the effects of this noise. It is important thatthe instrument you use is matched to the level ofinterference anticipated.

Rules on Testing and Comparing

Comparison of results in order to determine rates ofdegradation is key to the whole preventive/predictivemaintenance concept. However, it must be emphasizedthat this concept applies to readings taken at discretemaintenance intervals. Even then, strict standardizationof test procedure and conditions is imperative.Comparison of “on-the-spot” readings is a wholedifferent scenario and fraught with potential error.

It is tempting to try to back up tests with additionalreadings. You may make some adjustment to the testitem or setup, or someone else may have dif ficultyaccepting the result and wish to verify it. But aninsulation tester is not like a multimeter! High-voltagetesting behaves very much like the HeisenbergUncertainty Principle (you cannot know both the speedand position of an electron) applied to insulation. Thatis to say, the act of measuring affects the item beingmeasured, so that subsequent readings are not beingtaken on precisely the same item.

As has been described, the act of applying an insulationtest polarizes the insulating material. This ef fectivelychanges its electrical configuration and dielectricproperties. Because insulating material is, by design, apoor conductor, it can take considerable time for“relaxation,” or the return to random configuration, tooccur. Immediately upon termination of a test, the itemunder test is not precisely the same piece of equipmentthat it was before the test. An immediate follow-up testwill be affected, sometimes considerably, by the chargeleft from the first test. Which measurement is correct?They both are! They each can be expected to give acorrect measurement of the condition of the insulationat the time of test. Furthermore, industry-standarddischarge procedures are not sufficient for theinstitution of a repeat test. Such procedures are aimedat personnel safety, not qualification of the test item.Residual charges can remain for hours, or even days,that may be below human perception yet enormous toa sensitive meter. Equipment should be left groundedfor several hours, or preferably until the next day ,before additional testing is done. And then, externalfactors, especially temperature, must not be overlooked.

This does not mean that on-the-spot retesting shouldnever be performed. For relative information, it may bequite valuable. But it must be kept in perspective. Donot expect the readings to agree.

Two different operators may also not observe the samedegree of detail with respect to procedure. Temperatureis one factor. If the equipment is turned on, perhaps tocheck performance, then retested, the second test is notnecessarily comparable to the first. Time of test is alsoreadily overlooked. One operator may rigidly time thetest while another merely waits for stabilization of thereading. This can result in measurements being taken atdifferent points on the time-resistance curve (as hasbeen illustrated under the “Spot-Reading” test), andagain the two results will not be comparable.

Page 20: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

18 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

If this seems like excessive attention to detail, considerthe standards agencies. Organizations like UL ® andASTM® do not write procedures that say, in effect,“hook up a meter and take a reading.” Rather , theyspecify every variable, including setup, procedure, andcharacteristics of the test instrument, before results canbe considered in conformance. Standard maintenanceprocedures deserve no less diligence.

Testing Insulation Resistance of Rotating Machinery

In March 2000, the IEEE-SA Standards Board approved arevision of IEEE Std 43-1974 by the Electric MachineryCommittee of the IEEE Power Engineering Society. Thisrevision is IEEE Std 43-2000, the “IEEE RecommendedPractice for Testing Insulation Resistance of RotatingMachinery.” Changes made to the types of insulationused in electric rotating machines have resulted indifferent insulation resistance characteristics, andtherefore, required a substantial revision to the IEEEstandard. According to the IEEE, the standard isintended for:

n Individuals/organizations that manufacture rotatingmachines.

n Individuals/organizations that are responsible forthe acceptance of new rotating machines.

n Individuals/organizations that test and maintainrotating machines.

n Individuals/organizations that operate rotatingmachines.

Megger recommends that anyone involved in testingand/or maintaining rotating machinery review thisstandard in detail. We will provide some of thehighlights.

IEEE Std 43-2000 recommends a procedure formeasuring insulation resistance of armature and fieldwindings in rotating machines rated 1 hp, 750 W orgreater and applies to synchronous machines, inductionmachines, dc machines and synchronous condensers. Itdoes not apply to fractional horsepower machines. Italso recommends the insulation test voltage (based onwinding rating) and minimum acceptable values ofinsulation resistance for ac and dc rotating machinewindings.

The following chart provides guidelines for the dcvoltage to be applied during an insulation resistancetest. Note that voltages up to 10 kV are recommendedfor windings rated at greater than 12 kV.

*Winding Rated Insulation Resistance Test Voltage (V) Direct Voltage (V)

<1000 500

1000-2500 500-1000

2501-5000 1000-2500

5001-12,000 2500-5000

>12,000 5000-10,000

* Rated line-to-line voltage for three-phase acmachines, line-to-ground voltage for single-phasemachines, and rated direct voltage for dc machines orfield windings

The standard recommends that each phase be isolatedand tested separately (if feasible) as this approachallows comparisons to be made between phases. Thetwo phases not being tested should be grounded to thesame ground as the stator core or rotor body . When allphases are tested simultaneously, only the insulation toground is tested. Insulation resistance measurementsshould be made with all external equipment (cables,capacitors, surge arresters, etc.) disconnected andgrounded as these items may influence the resistancereading. A common ground should be used to preventstray losses in the ground circuit that could ef fect thetest results.

The standard calls out both the insulation resistance testand the polarization index test (PI), and recommendsthat both tests be made (if possible). It indicates thattesting history should be used to track changes. Ifhistory is not available, the standard provides minimumvalues for both tests that can be used to estimate thesuitability of the winding. These are the lowest values atwhich a winding is recommended for an overvoltagetest or for operation.

Page 21: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 19

The recommended minimum values for PI are based onthe thermal class of the insulating materials and applyto all insulating materials regardless of application perIEC 60085-01: 1984. The PI test is not applicable tononinsulated field windings. Be aware that a very highPI (greater than 8) for varnished cambric, shellac mica-folium, or asphaltic stator windings may indicate thatthe insulation has been thermally aged and may be atrisk of failure. Physical inspection can be used to confirmif the insulation is dry and brittle.

The rating of the machine determines whether themotor windings must achieve the minimum value foreither the insulation resistance test or PI test, or mustachieve the minimum for both tests.

Figure 18: Use of the Guard Terminal on a Power Cable

Thermal Class Rating Minimum PI Value

Class A 1.5

Class B 2.0

Class F 2.0

Class H 2.0

The recommended minimum insulation resistance afterone minute at 40o C can be determined from thefollowing chart. The minimum resistance of one phaseof a three-phase armature winding tested with theother two grounded should be approximately twicethat of the entire winding. If each phase is testedseparately (with guard circuits being used on the phasesnot under test), the observed minimum resistanceshould be three times the entire winding.

Minimum InsulationResistance (MΩ) Test Specimen

kV* + 1 For most windings made before about 1970, all field windings,and others not described below.

100 For most dc armature and ac windings built after about 1970(form-wound coils).

5 For most machines with random-wound stator coils and form-wound coils rated below 1 kV.

* kV is the rated machine terminal-to-terminal voltagein rms kV.

Machine Rating Evaluation Criteria

10,000 kVA or less Should have EITHER a value ofthe PI test or a value of theinsulation resistance test (at 40º C) above the minimumrecommended values.

Above 10,000 kVA Should have BOTH a value ofthe PI test or a value of theinsulation resistance test (at 40º C) above the minimumrecommended values.

The Guard Terminal

Some insulation testers have two terminals, others havethree. As these are dc testers, two of the terminals arethe + and -. The third (if present) is a guard. It does nothave to be used and many operators use insulationtesters satisfactorily without ever employing the guard.However, it affords the operator an extra function fordiagnosis of equipment problems. The guard is a shuntcircuit that diverts surface leakage current around themeasurement function. If parallel leakage paths exist, aguard connection will eliminate those from themeasurement, and give a more precise reading of theleakage between the remaining elements.

Page 22: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

20 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

As an example, dirt and moisture on a transformerbushing will promote surface leakage between the +and – connections, thereby bringing down the readingand possibly giving a false impression that the bushingis defective. Connecting the guard to a bare wirewrapped around the bushing will intercept this currentand yield a measurement based predominantly uponleakage through defects in the ceramic.

It is most important not to confuse the guard with aground. Connecting the guard and return lead to thesame element of the test item only shunts the currentthat is supposed to be measured, and thereby short-circuits the measurement function. When selecting atester, consider:

n The goals of testing (basic installation checks don’ tgenerally require a guard).

n The electrical composition of the items to be tested(motors and transformers can be tested for leakagebetween windings, with ground leakageeliminated).

n The possible effects of surface leakage (wire andcable can carry current across the surface, via dirtand moisture, as well as through the insulatingmaterial).

n The degree to which results must be analyzed (are“bad” items merely to be replaced or discarded, orwill it be necessary to localize faults for possiblerepair).

Testers with guards generally cost a bit more than two-terminal models, but in many applications, a two-terminal model won’t be imparting the full spectrum ofinformation that can be accrued by insulation testing.

Something that is often forgotten is the dif ference inthe capabilities of the guard circuit. The guardingcapability of the insulation tester is much moreimportant when measuring leaky insulation than theusually quoted measurement accuracy figure, which maybe 5%. Consider the following example, an extremecase where the surface leakage path is 200 times lessthan the resistance of the insulation.

Here we show an insulator of value 100 M Ω that wewish to measure. It is dirty and contaminated and so ithas a surface leakage path of 500 k Ω. If we apply ourtest voltage from the positive and negative terminalswithout guarding the circuit, 20 times as much currentwill flow through the surface leakage compared withthe current flowing through the insulation we wish tomeasure and we will read a resistance of only 497 k Ω.

If we “guard” the sample, here shown as being guardedsuch that we split the leakage resistance equally oneither side of the guard connection, we will be able toeliminate the effect of the surface leakage to a certainextent. How much we eliminate the ef fect of thesurface leakage is based on the guard circuitry of theinsulation tester used. Depending on the instrumentchosen, this error level can range from less than 1.0% tomore than 80.0%. If you intend to use the guardterminal, investigate the error level before purchasingan instrument.

This is a classic example of the need to compare tests ona like to like basis. An unguarded measurement and aguarded measurement yield very different results. Howis an operator to know whether the guard terminal waspreviously used unless the test records record thisseemingly unimportant detail?

Effects of Temperature

Temperature variations can have a significant ef fect oninsulation resistance readings. Resistance dropsmarkedly with an increase in temperature for the samepiece of apparatus. Each type of insulating material hasa different degree of resistance change withtemperature. Temperature correction factor tables havebeen developed for various types of electrical apparatusand can be acquired from the manufacturer. Failingthat, it is recommended that you develop your owncorrection factor tables by recording two resistancevalues for the same piece of equipment at two dif ferenttemperatures. A graph may then be plotted ofresistance (on a logarithmic scale) against temperature

G100 MΩ

250 kΩ

250 k Ω

Figure 19: Guard Terminal Diagram

Page 23: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 21

Effects of Humidity

Humidity (moisture content) has an effect uponinsulation resistance, but it cannot be quantified asneatly as can temperature effect because different typesof insulation will absorb moisture to varying degrees, aswill varying ages and conditions of the same type. Thebest that can be said is that humidity is a factor thatshould not be overlooked when evaluating test results.Unlike temperature, humidity’s effect is not a constantgradient and as long as the temperature remains abovethe dew point, humidity will not appreciably af fectinsulation readings.

Increasing humidity in the surrounding (ambient) air canaffect insulation resistance to varying degrees. Ifequipment operates regularly above the dew-pointtemperature (the temperature at which the moisturevapor in air condenses as a liquid), the test reading willnot be affected much by the humidity. Even if the

(on a linear scale). The graph is a straight line and maybe extrapolated to any temperature so that correctionfactors may be read directly.

In lieu of detailed data, the “rule-of-thumb” is that forevery 10º C increase in temperature, halve theresistance; or for every 10º C decrease in temperature,double the resistance. For example, a 100 G Ω resistanceat 20º C becomes 25 GΩ at 40º C.

Why is temperature correction important? Consider thefollowing example of a motor tested at various times ofthe year at differing temperatures (all within a 15ºband). The temperature adjustments were made usingthe rule-of-thumb correction.

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

Insulation Resistance Graph - XYZ Motor(Not corrected for Temperature)

Figure 20: Insulation Resistance Graph Not Corrected forTemperature

Insulation Resistance Graph - XYZ Motor(corrected for temperature)

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

Figure 21: Insulation Resistance Graph Corrected forTemperature

Insulation Temperature Temp. AdjustedDate Resistance (MΩ) ºF Insulation

Resistance (MΩ)

Jan-90 15,000 68 14,990

Jun-90 9,000 80 14,276

Jan-91 14,500 68 14,490

Jun-91 8,500 82 14,562

Jan-92 14,300 68 14,290

Jun-92 8,700 81 14,341

Jan-93 14,500 68 14,490

Jun-93 8,900 81 14,671

Jan-94 14,200 69 14,748

Jun-94 8,900 80 14,117

Jan-95 13,600 68 13,591

Jun-95 8,900 78 13,071

Jan-96 13,500 66 12,491

Jun-96 7,500 80 11,896

Jan-97 11,300 68 11,292

Jun-97 6,500 80 10,310

Jan-98 8,000 67 7,693

The readings taken create confusion if they are notcorrected for temperature. When plotted, they producea chart that is of limited use in determining a trend.If the same data is corrected for temperature andplotted, the graph begins to provide a valuable pictureof the deterioration of the insulation.

Temperature correction is particularly important whentesting with higher voltages and at higher levels ofsensitivity.

Page 24: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

Protection Against Access to Hazardous Parts (First Digit)

Number Description

0 Non-protected

1 Protected against access with back of hand (50 mm)

2 Protected against access with jointed finger (12 x 80 mm)

3 Protected against access with a tool (2.5 mm)

4, 5, 6 Protected against access with a wire (1.0 mm)

22 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

equipment to be tested is idle, the same is true — solong as its temperature is kept above the dew point(and the insulation surfaces are free of contaminants,such as certain lints and acids or salts, which have theproperty of absorbing moisture).

In electrical equipment, we’re concerned primarily withthe conditions on the exposed surfaces where moisturecondenses and affects the overall resistance of theinsulation. Studies show, however, that dew will form inthe cracks and crevices of insulation before it is visiblyevident on the surface. Dew-point measurements willprovide a clue as to whether such invisible conditionsmight exist, altering the test results.

Humidity effects require greater attention as testvoltages increase because the higher voltages canpromote ionization much more readily than at lowvoltages. As a result, humidity that doesn’ t produce anoticeable effect at 1 kV may produce perplexingly lowreadings at 5 kV. This is not necessarily a problem. Thedifference in response at two dif ferent voltages can beused to detect moisture and tests carried out guardedand unguarded can be used to detect surface moistureor internal moisture.

Ingress Protection

Somewhere in the fine print of most test equipmentproduct bulletins is an IP rating, a number that gives theoperator vital information. In fact, the IP rating lets theoperator know whether a piece of test equipment issuited for his/her application and test environment.

“IP” stands for “ingress protection.” That is the degreeto which the instrument can withstand invasion byforeign matter. The IP rating system has beenestablished by the IEC (International ElectrotechnicalCommission), in their Standard 529, and is used as aguide to help the operator protect the life of theinstrument. It also can help the operator make a moreinformed purchase decision by ensuring that the pieceof test equipment is designed to work in theenvironment(s) that he/she faces.

The IP rating is comprised of two digits, each signifyinga separate characteristic. The designation indicates howwell the item is sealed against invasion by foreignmatter, both moisture and dust (the higher thenumber(s), the better the degree of protection). Whatwould a typical rating of IP54 tell a buyer about theapplication capabilities of a model? If you want tosound thoroughly knowledgeable, that’s IP five-four, notfifty-four. Each digit relates to a separate rating, not toeach other.

The first digit refers to particulate ingress, reflecting thedegree to which solid objects can penetrate theenclosure. A level of “5” indicates “dust protected” aswell as protected from invasion with a wire down to 1.0mm. There is only one higher category: “dust tight.”The second digit refers to moisture. A rating of “4”means resistance to “splashing water, any direction.”The higher ratings of 5 through 8 indicate “jettingwater” and “temporary” or “continuous” immersion.

So what? Well, suppose an instrument underconsideration was rated only to IP43. What would thattell the operator about its usability? Could it bethoroughly utilized in a quarry or cement plant? Hardly!The particulate rating 4 indicates “objects equal orgreater than 1 mm.” That’s a boulder in comparison toparticles typically produced by industrial processes.Flying dust could put the unit out of commission.

Suppose the unit is rated at IP42. A moisture rating of 2indicates dripping water. Therefore, it would not beresistant to flying spray. Acquiring an instrument for anenvironment that exceeds its IP capabilities likely meansthat the operator will need another very soon. Whatabout a rating of IP40? A moisture rating of 0 meansthat the unit is not protected against any liquid ingress.

The following charts provide a guide to various IPratings and what they mean to the operator:

Page 25: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 23

High Potential Testing

There is no truly singular definition of the “highpotential” test. It is commonly used, but its definition issituational, in the “eye of the beholder” it might besaid. Basically, a high potential test is an electrical stresstest conducted at a voltage two or more times ratedvoltage and sometimes known as a Withstand or ProofTest.

Since the test is conducted at a voltage considerablyhigher than the rated voltage of the equipment beingtested, it is known as an overvoltage test unlike thehigh voltage insulation test, which is generally appliedat a voltage below the rated voltage of the equipment.The act of overvoltage testing creates abnormal stressesin the test sample and can contribute to theacceleration of aging in insulation. Indeed, somestandards require the voltage to be increased until thetest sample breaks down.

If an overvoltage test is to be applied, it is normalpractice to apply an undervoltage PI test beforehand topre-qualify the insulation.

High potential tests may be carried out with ac or dcvoltages, as appropriate. Samples with considerablecapacitance will appear as a short circuit to an ac testrequiring a test set with very large power capabilities toovercome the capacitive charging currents. In situationssuch as this, it is quite normal to apply a dc test with theequivalent peak.

Current (nA) Readings vs. Resistance (M Ω) Readings

Insulation testers measure current and then convert itinto a resistance reading. Why do we do this? W ell,predominantly, it’s tradition. Good insulation produces ahigh reading while poor insulation produces a lowreading. Also, good insulation is predominantly resistive.If we double the test voltage, we double the currentflowing but the resistance remains constant. However,sometimes it is easier to diagnose problems byconsidering the actual currents that flow.

The choice is yours because many modern insulationtesters are capable of presenting their measurements ineither unit.

Burn Capability

Full-function insulation testers above 1 kV often includea “burn” mode. It is a feature that may never be used;yet it does have a viable function within a narrow rangeof application.

Insulation testers will generate high voltages intosignificant resistances. However, if a breakdown occurswithin the insulation, the resistance drops, the currentincreases, and the voltage drops. If left to its owndevices this would cause the breakdown arc toextinguish, the resistance to increase, and the voltage toincrease which in turn causes breakdown and so on. Thiscontinuing cycle does not allow the measurement ofresistance and indeed could open pinholes or enlargeburn tracks. Rather than cause further damage, mostinsulation testers will shut down.

However, if you want to find the location of thebreakdown this may be extremely inconvenient. For thisreason some instruments offer an operator selectable“burn” mode; the automatic shutdown is overriddenand a low current arc is maintained. It must beunderstood, however, that the instrument’s short circuitlimitation is still in effect. The tester will not provide a“dead” short. The function enables the operator to

Protection Against Ingress of Solid Foreign Objects (First Digit)

Number Description

0 Non-protected

1 Objects equal or greater than 50 mm

2 Objects equal or greater than 12.5 mm

3 Objects equal or greater than 2.5 mm

4 Objects equal or greater than 1 mm

5 Dust protected

6 Dust tight

Protection against Ingress of Liquids (Second Digit)

Number Description

0 Non-protected

1 Water dripping vertically

2 Water dripping, enclosure tilted up to 15°

3 Spraying water, up to 60° angle from vertical

4 Splashing water, any direction

5 Jetting water, any direction

6 Powerful jetting water, any direction

7 Temporary immersion in water

8 Continuous immersion in water

Page 26: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

24 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

localize or identify the fault by looking for a spark orwisp of smoke or perhaps by use of an ionizationdetector. Pinholes in windings can be identified, coveredwith insulating varnish, and the equipment returned toservice. In cable maintenance, a high potential testerwith much higher currents than insulation testers is usedto “break down” a high-resistance fault, converting it toan “open” that is much more easily recognized by arcreflection techniques.

Drying Out Electrical Equipment

Electricity and water do not form a happy partnershipand so it is often necessary to “dry out” insulation. Thismay be done to remove surface moisture or perhaps todrive moisture from the internals of the insulation.Indeed some pieces of equipment have in-built heatercoils which can be used for this purpose. However ,several other methods are also available for dryingelectrical equipment.

The most satisfactory solution to the problem involvesplacing the windings in an oven with suitabletemperature control and proper air circulation. Banks ofinfrared lamps may be used when this is not possible, ora suitable housing may be built around the machine,using steam coils or electric resistance type units for asource of heat. Openings must be provided for the freecirculation of air as otherwise the expulsion of moisturewould simply result in an increasing humidity inside thedrying chamber. Blowers may be used to increase the airmovement.

Vacuum drying has been also effectively used toexpedite the return of equipment to service, but thismethod requires extra precautions and should only beundertaken by experienced personnel.

Another method often used is to circulate low-voltagecurrent through the windings. This method should notbe used, however, until the insulation resistance hasreached a value of at least 100 M Ω. The current shouldbe limited to only a fraction of nameplate amperes, anda careful check must be maintained on maximumtemperatures on the insulated parts. Maximum dryingtemperatures on windings should not exceed 194º F (90º C) as measured by a thermometer. This will preventnot only the rapid thermal deterioration of theinsulation but damage from the high vapor pressuresthat would be obtained if steam were produced.

If drying is required, records help determine when theinsulation is moisture free. As an example of theimportance of past readings, consider a motor that’ sbeen flooded. After a cleanup, a spot reading with theMegger tester shows 15 MΩ. If past records showed theinsulation resistance to run from 10 to 20 M Ω, themotor would be in good shape. If, on the other hand,past records showed the normal resistance values to runfrom 100 to 150 MΩ, the operator would know thatmoisture was still present in the motor windings.

During drying operations, when insulation resistancevalues are used as an indicator of the suitability ofwindings for service or for application of test potential,the drying must be continued for a suf ficient time tomake sure that the values are reliable. Often theresistance curve will take one or more sharp dips beforeleveling off or continuing to increase in a positivedirection. This is due to moisture working out of thewindings. When the machine is completely dried out,further work is required to remove any remaining dust.This may be done through the use of dry compressed airat pressure not exceeding 40 psi.

Figure 22 shows a typical drying-out curve for a dcmotor armature, which shows how insulation resistancechanges. During the first part of the run, the resistancedecreases because of the higher temperature. Then itrises at a constant temperature as drying proceeds.Finally, it rises to a high value, as room temperature (20° C) is reached.

There is a significant caveat when testing wet insulationwith an insulation tester; wet equipment is susceptibleto voltage breakdown. If windings have absorbed a lotof moisture even low voltages can puncture insulation.Therefore, the operator should be very careful beforeapplying high voltages. More advanced Meggerinsulation testers allow the test voltage to be set atanything from a low of 25 volts to a high of 5000 voltsin 25-volt increments.

Page 27: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 25

Test Item Discharge

Perhaps at school you were taught to discharge acapacitor and then to store the capacitor with theterminals shorted together. Did you ever wonder why,since you have discharged the capacitor and maybe hadchecked that there was no voltage across the terminals,you needed to short the terminals?

The reason is the dielectric absorption current. If theterminals are left unshorted, the energy stored bydielectric absorption will slowly release with negativecharge migrating to one terminal and positive charge tothe positive terminal. Over a period of time this chargecan build up to a dangerous level, as high as theoriginal test voltage, and with a considerable amount ofenergy to back it up. This energy can kill.

At the end of an insulation test the test sample closelyresembles a charged capacitor; there remains aconsiderable amount of energy stored within theinsulation dielectric.

There is an important “rule of thumb” on charging anddischarging items under test. This rule suggests that theoperator discharge the item under test for five times aslong as it was tested. If the operator performs a 10-minute PI test, he/she should allow the unit to dischargefor 50 minutes.

A good quality instrument will automatically dischargethe test sample as soon as a test is completed orinterrupted. Some lower quality instruments have aseparate discharge selection knob, or switch, which addsa step to a test. If this step is forgotten, the test itemcan be deadly for the next person who handles it.

Megger insulation testers also detect the voltage acrossthe test sample during the discharge phase and willshow this voltage until it has fallen to a safe level. Atthis point, the item is safe to handle.

However, all we have discharged at this point is thestored capacitive charge. As explained at the start ofthis booklet, any capacitance is charged relativelyquickly at the start of a test. Similarly , the capacitivecharge is discharged relatively quickly at the end of atest. But the dielectric absorption current takes muchlonger to go in and also takes much longer to come out.

Thus while the sample is immediately safe to handle, ifthe terminals are not shorted they will gradually acquirecharge and become dangerous once again. So, unlessthe equipment is going back into service, ensure thatthe terminals are shorted and grounded.

Figure 22: Typical Drying-out Resistance Graph

Page 28: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

26 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

Charging Time for Large Equipment

One question we are often asked is, “How long will ittake to charge a particular piece of equipment?” Theanswer is, “We don’t know!”

Why not? Well, the answer depends on the actualconfiguration of the particular piece of equipmentconcerned. For example, the Megger S1-5010 specifies acharging rate of “less than 5 seconds per microfaradwith 2 mA short circuit current” and “2.5 seconds permicrofarad with 5 mA short circuit current.” Thus, if youknow the capacitance of the test sample you can workout the charging time; it doesn’ t matter if it is a motor,a cable or just a slab of insulating material.

Motor Driven Insulation Testers

Another question we are frequently asked is “Whathappened to the old wooden box motor driveninsulation testers?” Some people seemed to think thatthey set the standard for insulation testing and still do.

These motor driven wooden boxes, with an externalmotor were produced between 1910 and 1972 and usedthe original Evershed patented “Cross Coils Ohmmeter.”This was a large heavy movement that, as the namesuggests, had two coils set at an angle to each other .This was the first “true ohmmeter.” The construction ofthe movement had benefits and drawbacks.

The main benefit was, because of the weight of themovement, it had considerable inertia and was,therefore, quite insensitive to interference or transientevents. This resulted in a very smooth motion.

Unfortunately, the sheer weight of the movement madeit fairly delicate and so the instruments needed to behandled with care. Furthermore, the instrumentsneeded to be leveled before use and was, therefore,supplied with a spirit level on the scale and adjustablefeet. The movements were also fairly insensitive withmaximum resistance capabilities that could be measuredin high megohms or low gigohms.

Alternative power sources were developed. The oldgenerator was big and heavy as anyone who has triedto hand crank one of these old instruments will attest;you certainly wouldn’t want to do a PI test while handcranking, but if you lacked a mains supply there was noalternative.

Technology advances meant that “electronicmovements” could be used which were more ruggedand more accurate. New low-voltage generators weredeveloped which made hand cranking much easier andthen ultimately battery technology enabled purebattery power to be employed. This resulted in the longterm, very stable power supplies that we see today .

The use of electronics has resulted in lighter, morerugged, more accurate instruments that respond morequickly. They can provide more information, whichresults in us seeing transient events that were previouslytotally hidden by the relative instability of the powersupply and the inertia of the movement.

Which is better? The decision is yours.

Page 29: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV 27

MEGGER 5-KV INSULATION TESTERS

MIT510 and MIT520

n Line supply or battery operated

n Measures to 15 TΩ

n Digital/analog backlit display

n Automatic IR, DAR, PI, SV and DD tests

The MIT510 and MIT520 provide a high level of qualitytesting along with prime safety, convenience andportability. MIT510 offers selectable test voltages at 2 50,500, 1000, 2500, and 5000 V. The MIT520increases the functionalityby permitting test voltageto be set in any 10 Vincrement from 50 V to 1 kV, then in 25 V steps all the way to 5 kV.

Test results are displayed on our patented electronicdigital/analog arc, therebyaffording the precision of a digital reading. For fastercharging and testing of large inductive loads theMIT510 and 520 feature a 3 mA short circuit current.

MEGGER INSULATION TESTERS

Megger offers a wide selection of 5 and 10-kV testers tofit your needs. Megger insulation testers give theelectrician, maintenance engineer and repair technicianflexibility and versatility to meet the demands of highvoltage applications.

Model BM11

n Large, direct reading analog display

n Guard terminal

n Supplied with internal charger and battery check

This model is ideal for careful observation of insulationperformance during anextended test. Withrechargeable powersupply, the unit is bothportable and rugged. Fourtest voltages, up to 5000 V,permit measurement to100,000 MΩ. There is awarning of externalvoltage, while stored testcharge is safelydischarged, and monitoredon a voltmeter scale to1000 V ac or dc.

Model S1-5010

n Flexible power supply options

n PC control and display option

n Heavy duty, line/battery operation

A premium insulation tester for advanced insulationresistance diagnostic testing of all types of electricalequipment. The S1-5010 performs automatic tests,enabling consistent operation without operator input,plus it stores results and can be operated from a PC.Portable construction allows theinstrument to be used in avariety of applications.It is particularly suitablefor analyzing insulationof large machines, HVgenerators, and cables.

S1-552 and S1-554

n 5 mA short circuit current

n Extra noise rejection capability up to 4 mA (S1-554)

n 15 TΩ measurement range

n Automatic DAR, PI, SV and DD tests

The new S1-552 and S1-554 includes all the capability of the MIT510 and 520. The units of fer variabletest voltages of 50 V to 1 kV in 10 V steps and 1 kV to 5 kV in 25 V steps.

The S1-552 and S1-554 drives 5 mA of short circuitcurrent, allowing for the fastest charging andtesting of large inductive loads such as generatorsand transformers.

Model S1-552 features a 2 mA noise rejectionwhile the S1-554 provides an industry best of 4 mA @ 200 V and above.

Page 30: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

28 A GUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE 1 KV

MIT1020

The Megger MIT1020 10-kV insulation resistance testeris also designed specifically to assist you with the testingand maintenance of high voltage electrical equipment.The unit features anexpanded test voltagerange from 50 V to 10kV, in 10 V incrementsto 1 kV and 25 Vincrements above 1 kV.

For faster charging andtesting of largeinductive loads theMIT1020 features a 3mA short circuit current.

Models MJ15 and BM15

n Pass/fail overlays for rapid go/no go testing

n Insulation resistance to 20 GΩ

n Voltage range to 600 V indicates auto discharge

The BM15 and MJ15 are compact 5-kV insulation testersthat are simple to use and provide a quick, accuratereading of insulation resistance. Both instruments of ferfour test voltages (500 V, 1 kV, 2.5 kV, 5 kV), analogscales, and measurement sensitivity to 20 G Ω.

The BM15 is powered by 8 “AA” or rechargeablealkaline batteries while the MJ15 includes a hand-crankgenerator in addition to battery power.

MJ15BM15

MEGGER 5-kV INSULATION TESTERS Continued MEGGER 10-kV INSULATION TESTERS

Our 10-kV testers fully conform to the requirements ofIEEE Standard 43-2000, the “IEEE Recommended Practicefor Testing Insulation Resistance of Rotating Machinery.”

n Line supply or battery operated

n Digital/analog backlit display

n Variable test voltage from 50 V to 10 kV

n Automatic IR, PI, DAR, SV and DD tests

n Measures to 35 TΩ

S1-1052 and S1-1054

The new S1-1052 and S1-1054 are also designed fortesting the insulation of high-voltage electricequipment. The units carry all the features of theMIT1020 with the addition of a short circuit current of 5 mA — that’s 5 times faster to raise the capacitor to the desired test voltage! Large loads like generators and transformers can be tested quickly.

Model S1-1052 features a 2 mA noise rejection while the S1-1054 provides an industry best of 4 mA @ 200 Vand above.

Designed in accordance with the requirements ofIEEE43-2000, these units, like the MIT1020, are ideal formeasuring the insulation resistance of armature andfield windings in rotating machines rated 1hp (750 W)or greater. The standard applies to synchronous,induction and dc machines as well as synchronouscondensers.

For more information and full specifications, visit ourwebsite www.megger.com.

And there’s more! Megger, the originator of theinsulation tester, also offers a full line of 1-kVinsulation testers to fit your budget and applications. Log on www.megger.com and see for yourself!

Page 31: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE
Page 32: US $9.95 AGuide To Diagnostic Insulation Testing Above · PDF fileAGuide To Diagnostic Insulation Testing ... Polarization Index Test ... AGUIDE TO DIAGNOSTIC INSULATION TESTING ABOVE

MEG-562/3M/5.2007

UK Archcliffe Road, DoverCT17 9EN England T (0) 1 304 502101 F (0) 1 304 207342

UNITED STATES 4271 Bronze WayDallas, TX 75237-1088 USAT 1 800 723 2861 T 1 214 333 3201 F 1 214 331 7399

OTHER TECHNICAL SALES OFFICESNorristown USA, Toronto CANADA,Mumbai INDIA, Trappes FRANCE, Sydney AUSTRALIA, Chonburi THAILAND and The Kingdom ofBAHRAIN.

ISO STATEMENT

Registered to ISO 9001:2000 Reg no. Q 09250

Registered to ISO 14001 Reg no. EMS 61597

5KV_UG_en_V03www.megger.comMegger is a registered trademark

michael ryan
MR STAMP 2