AppNote8

7/29/2019 AppNote8

1/16

ON-LINE BUSHING MONITORING AND

COMPARISON TO OFF-LINE TESTING

Claude Kane | Dynamic Ratings, Inc.

Abstract

In order to understand the aspects of on-line bushing monitoring, it would be good todiscuss what is power factor and capacitance, typical construction of a bushing and

typical off line testing theory. This will allow for comparisons to be made between the

methods and technology.

Condenser Bushing Construction

The key components of a condenser bushing are as shown in Figure 1.

Capacitance LayersC1 C2 C3 C4 C5

If

C1=C2=C3=C4=C5

Then

V1=V2=V3=V4=V5

Foil/Conductive Ink PaperFilled

with Oil

Tap

CenterConductor

Figure 1

The purpose of the capacitive layers allows for the energized center conductor to

penetrate the ground plan. By having equal capacitances, the voltage is stressed

equally across each layer.

7/29/2019 AppNote8

2/16

2

Condenser Bushing Construction (contd)

There are two major capacitances in a bushing - C1 and C2. These are shown in

Figure 2. The C1 capacitance is the capacitance between the center conductor and

the tap. The tap is usually connected to the outer most foil and in some cases to the

second to last foil. The C2 capacitance is the capacitance from the tap to ground.

Typically the tap is grounded, therefore the C2 capacitance is not in the circuit duringnormal operation.

BetweenLast Foil&Flange

Test/Capacitance/Potential/Voltage

C1

C2

Figure 2

What is a Capacitor?

A capacitor is two conductive plates that are separated with a dielectric. In the case of

a bushing, the conductive plates are the foil and the dielectric is the paper and the oil.

All dielectric materials have some sort of loses, expect for a perfect vacuum.

Comparisons of dielectric constants are shown in Table 1.

Material Dielectric Constant

Vacuum 1.0

Air 1.00549Paper 2.0

Oil 2.2Porcelain 7

Water 20

Table 1

DielectricConductive Plates

7/29/2019 AppNote8

3/16

3

Dielectric Loses

Dielectric losses are measured in units of watt loses whereby heat is generated due to

these losses. Losses are created by the following causes.

Natural resistance of the material

Type of the material

Polar molecules such as moisture

Ionization of gases (Partial Discharge)

Losses will vary by the amount of dielectric material. Since bushing are not the same

size and composition, comparison of watt losses between different manufactures,

sizes, etc. is difficult. Therefore the industry uses the term Power Factor to quantify

the condition of the bushing insulation system. As losses increase due to any the

above causes, the Power Factor will also increase.

Power Factor

PowerFactor (PF) is the phase angle relationship between the applied voltage across

a capacitance and the total current through the capacitance.

Power = Voltage (E) x Current (It) x Cosine ().

Watts = E x Ir

Watts = E x It x Cosine()

PF = Cosine() =

If decreases, more resistive current will flow through the insulation and the power

factor will increase. Table 2 shows the angle () and the calculated %PF. It can be

seen, a phase angle difference of only 0.2 degrees (90 89.8) gives a %PF of 0.349,

which is a common value (0.25 to 0.50) for a power factor of a bushing.

%PF (%COS)90 0

89.8 0.349

89.5 1.75

88 3.490

87 5.23

86 8.710

Table 2

E

IrIc

It

7/29/2019 AppNote8

4/16

4

Capacitances

In bushings there are a several capacitors in series. The total capacitance is the sum

of the inverse of each capacitance as shown in formula below.

C1 C9C8C7C6C5C4C3C2

Tap

Flange

When a capacitor layer shorts out, the value of the capacitance will always increase.

For example, using the formula, if a system has three capacitors wired in series and

each capacitors value is 3 pF, the total capacitance will be equal to 1 pF. (Pico farad)

Using the formula, if one of the capacitors is shorted out, the sum is 1.51 pF.

The capacitors in series act as a voltage divider. If a capacitor shorts out the voltageat the tap will increase in proportion. Also, as the voltage varies the leakage will vary.

Therefore, if the voltage increases, there will be an increase in leakage current.

C1C2

7/29/2019 AppNote8

5/16

5

Typical Off-line Test Setup

The basic off-line test circuit is shown below. A high voltage is applied to the test

object and the leakage current is measured.

Test

Object

The test circuit to measure the C1 capacitance of a bushing is as follows:

The high voltage is connected to main conductor of the bushing and the return lead is

connected to the tap. The electronics in the test set can accurately measure leakage

current, the applied voltage, frequency and the phase angle between the applied

voltage and the leakage current. From this data, the watts loss and PF can be easily

calculated as discussed earlier in this document. Typically, 10kV is applied to the mainconductor.

Other connections of the test leads can be made to test the C 2 capacitance and other

characteristics, but C1 is only discussed since that is the only capacitance that can be

monitored on-line.

7/29/2019 AppNote8

6/16

6

On-line Bushing Monitoring

THEORY

By far, the most common method to monitor bushings is the sum of current method.

Figure 3 shows a basic block diagram of a bushing monitoring system that uses the

sum of currents method. During commissioning the null-meter is balanced to zero.The purpose of the balancing circuit is to take into account the differences in system

voltages and phase fluctuations and bushing characteristics. As a defect develops the

complex conductivity of the bushing insulation changes and the current and its phase

angle in one of the phases also changes. Therefore, the null-meter will no longer be

null.

KEY POINT

THE AMPLITUDE OF THE CHANGE REFLECTS THE SEVERITY OF A PROBLEM AND

THE PHASE ANGLE INDICATES WHICH PHASE IS EXPERIENCING THE CHANGE.

c1

c2

c1

c2

c1

c2

BalancingUnit

SummationUnit Null Meter

Figure 3

Installed bushing sensor

7/29/2019 AppNote8

7/16

7

On-line Bushing Monitoring (contd)

The change in the sum of currents can be approximately represented by the formula

below under the assumption of a single defective phase:

2

0

2

0 )tan(

+

=

CC

I

II

Where: I - Parameter Sum of Currents,tan - Tangent delta change,

0CC - Relative change in bushing capacitance,

0C - Initial Capacitance Reading,

0I - Initial Sum of Current Value.

Ideally, the sum of the three bushing currents should be zero. In reality, not allparameters are equal from each phase. Therefore during commissioning of the

system, the monitor is placed in a Balancing Mode, and the monitor self-adjusts so the

sum of the (3) currents is equal or close to zero.

b.

'

I0A

I

0

CI

0BI

'

AI

AV

'A

I

a.

0=I0

AI

0B

I

0

CI c.

"

I0A

I

0

CI

0B

I

"A

I

AV

"

AI

Figure 4 (a., b., c.)

Figure 4 (a., b., c.) explains the method in vector format. (a.) shows all three currents

from the bushing test taps perfectly balanced and the sum equal to zero. If there is a

change in tangent delta in the phase-A bushing an additional active current will pass

through the A-phase bushing insulation and the new current 'A

I , thus throws the

system out of balance. The consequent imbalance vector is equal to the tangent delta

change and directed along the phase-A voltage vector (b.). A change in capacitance isshown in (c.). This additional current is perpendicular to the A-phase voltage. The

consequent imbalance is now positioned along the vector 0A

I .

7/29/2019 AppNote8

8/16

8

On-line Bushing Monitoring (contd)

The magnitude of the change is an indicator of the problems severity, and the vector

change indicates which bushing is going bad and whether the power factor or

capacitance is changing. The chart and plot below shows a recent example of a

bushing going bad.

A Phase

Power Factor

B Phase

Power Factor

7/29/2019 AppNote8

9/16

9

Calculation of Power Factor and Capacitance

For decades the only method to determine the quality of the bushing insulation was to

perform off-line tests and compare the measured power factor and capacitance to

nameplate values or previous tests. Therefore, this is what maintenance personnel

are used to looking at.

As stated earlier, when performing offline tests all quantities can be measured.Table 3 compares the available data inputs for Off-Line Testing to the data available

for on-line Monitoring.

Parameter Off Line Testing On-LineMonitoring

Applied Voltage XLeakage Current X XPhase angle between Voltageand Leakage current

X

Frequency X X

Table 3

When performing on-line monitoring, the key diagnostic factor is the sum of currents

and the phase angle of the sum. Only estimates of the power factor and capacitance

can be made since all the data required to calculate the absolute PF and capacitance

is not available, as it is for off line tests.

For this reason, on-line bushing monitoring provides relative calculation of PF and

Capacitance. When the system goes out of balance, estimates are made on the

change of PF and/or capacitance. These values are then added/subtracted to baseline

values (nameplate or recent test values) entered into the system. For example if thebaseline PF is 0.35% and the algorithms change calculation show the PF increased by

0.50% the reported PF will be 0.35 + 0.50 = 0.85%.

It must also be noted the sum of currents value is not a Calculated value, but is a

Measured value.

Let us consider a system with three bushings. Total variables required are 4 x 3 = 12.

If a user had a three phase off-line test set, all the data that is required is available.

For on-line monitoring the following conditions apply:

The Line Voltage at the bushing terminals is assumed to be constant on all

three phases.

The phase angles between the Phase Voltages are constant.

In additional to the leakage current, the phase angles between Phase A and B

and A and C are also measured and frequency is measured.

7/29/2019 AppNote8

10/16

10

Calculation of Power Factor and Capacitance (contd)

In actual applications, the voltage is not constant and the phase angles between

phases are not always exactly 1200 apart. If all changes remained symmetrical, then

the system would stay in balance and the PF and capacitance calculations would be

fairly accurate. Since this is not the case, small variations in the sum of currents will

occur.As a bushing deteriorates, the variations have less of an impact on the calculations.

As can be seen in the chart below of the sum of currents, at lower values the variations

are a lot larger than those when the sum of currents is higher. In fact, as the sum of

current increases the accuracy of the PF and capacitance calculations will improve.

As stated earlier, a change in voltage at the tap (because the amplitude of the leakage

current will change) will indicate a change in capacitance and a change in phase anglewill indicate a change in PF.

The actual calculation of PF and capacitance is quite complex with extensive

averaging, filtering and proprietary algorithms.

7/29/2019 AppNote8

11/16

11

Other items to consider when comparing Offline Testing to On-line Bushing

Monitoring Values

Many bushing defects are temperature and voltage dependent. When testing off-line,

the bushing is at ambient conditions and only 10 kV is applied. With on-line

monitoring, higher voltages are applied and elevated temperatures are present.

The chart and plot below shows the affect of top oil temperature (red) to that of thesum of currents (blue). Top oil temperature is used as a relative temperature

measurement for the temperature of the bushing. Approximately 60% of the bushing

temperature comes from the oil. . As can be seen, there is significant temperature

correlation.

Changes in

B Phase

7/29/2019 AppNote8

12/16

12

Other items to consider when comparing Offline Testing to On-line Bushing

Monitoring Values (contd)

The following chart shows the PF calculations for the above chart and plot.

Diagnostics show the B phase bushing is deteriorating.

Table 4 shows the offline tests of the three high voltage bushings discussed above.

One can observe the changing values of the B Phase PF, with the calculated PF for all

three bushing. As can be seen, the offline power factor test on B Phase is out of

specification.

Table 4

The plots below show that significant partial discharge was occurring in the B phase

bushing.

7/29/2019 AppNote8

13/16

13

Additional Information

The following graph shows both the temperature and voltage dependency of a

bushing.

7/29/2019 AppNote8

14/16

14

Case Study

To better illustrate the benefits of on-line bushing monitoring and the points above, letus look at some data collected on some 230 kV bushings.

This particular transformer has had two different bushings going bad separated byabout 8 months. The trend of the sum of currents is shown below. From observation,one can see the sum of currents is varying quite a bit. This is normal. The sum ofcurrents for July 2010 slightly exceeded 1.0%. Default alarm points were 4% (Yellow)and 6% (Red). Even though alarm points were not reached the customer had aplanned outage, and they replaced the bushing at that time.

If we zoom in on the data collected during the July 2010 timeframe and review the

correlation with top oil temperature, we see a lagging temperature dependency. The

cause of the lagging is related to the time constant in heating up the oil in the bushing.

Correlation with Top Oil Temperature is a key diagnostic.

KEY POINT

WHEN MONITORING BUSHINGS, CORRELATION WITH TOP OIL TEMPERATURE

IS A KEY DIAGNOSTIC.

July 2010 March 2011

7/29/2019 AppNote8

15/16

15

Case Study (contd)

Review of the Polar Chart below identifies the two faulty bushings. C Phase in July

2010 and B phase in March 2011.

Trend of the Relative Bushing Power Factor calculations

July 2010

March 2011

July 2010 March 2011

7/29/2019 AppNote8

16/16

16

Summary

This paper reviewed the concepts of:

Bushing Construction

Capacitance

Dielectrics

Power Factor (PF)

Off line bushing testing theory

Comparisons of offline testing to on-line monitoring were made and why absolute

calculations of the power factor and capacitances cannot be made with traditional

bushing monitoring systems.

Sum of Currents and balancing method

Correlation of the affects of oil temperature and voltage and bushing failure

This paper documented a case study whereby; On-line monitoring was used to avoidbushing failures, which could have led to unplanned service outages and theassociated costs.

Conclusion

When performing on-line monitoring of bushings, the sum of currents and the phaseangle are the key diagnostics parameters. A paradigm shift must be made from

focusing on the capacitance and power factor readings to that of the sum of currents.

References