Webinar on Tan Delta Testing

MSc. Andrej Šepčić

RAM
andrej.sepcic@altanova-group.com
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

1. We talked about what is electrical insulation

2. We found out that we can model the

insulation as a Capacitor, and we spoke about
ideal and real insulation scheme

Real Capacitor shown

on PCB
Ideal Real
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

3. We learned that the IR – Resistive current is the

reason why our insulation heats up, and that
heat deteriorates the electrical insulation

P = 𝑉 𝑥 IR 𝑥 cos 𝜑
H
And that it creates
E
a vicious circle (a chain A
reaction) that T
can lead to dramatic Chain reaction resulting
failures in insulation breakdown
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

4. We heard the Dutch legend of Hans Brinker

a boy who saved his village by sticking a
finger in a dike – since a small leak in a dike can
quickly deteriorate the dike and we compared
the dike to electrical insulation

5. We also heard the TV can function normally

in an aquarium filed with mineral oil, but that
we should take care of how much water is in
that oil – and that is what Tan Delta
enables us to do
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

6. And we concluded the first part saying that we

measure the Tan Delta so that we could asses
the state of our electrical insulation – and to
see if there were some geometrical movements
inside our Transformer

7. We said that Tan Delta can also be tested on

any type of primary equipment - Transformers,
but also RM, cables etc…
We test tan delta
to keep the asset running
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

8. We then defined Tan Delta as the ratio of

resistive and capacitive current
𝐼 1
𝑇𝑎𝑛 𝛿 = 𝐼𝑅 = 𝜔𝑅𝐶
𝐶

And we also mentioned that TD = DF Dissipation Factor

9. And we also defined PF – power factor as 

𝐼𝑅
PF = cos  = 𝐼
𝑇𝑂𝑇

We also mentioned that TD and PF in most cases are the Current and voltage vectors
same (up to 10°) and if there is a significant difference, in insulation
then the insulation is probably very much compromised
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

10. Further on we discussed what makes the insulation

deteriorate, so we mentioned:

A) Water (the fact that transformer breaths)

B) Partial Discharge (and other processes that
produce gases)
C) Oxidation processes that create sludge

All this makes insulation more conductive

and less insulating

Service - aged transformer

Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

11. After which we showed two tables of allowed

values for TD

12. We concluded that TD is dependant on

ambient temperature

And we gave an example on temp. compensation

Tables showing allowed TD values

Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

13. And then we said that voltage at which the

test should be done should be not bigger
than 12 kV (since all the polarisation processes
are finished at that voltage) OR the nominal
voltage (if the asset has nominal voltage less
than 12 kV)

14. We also mentioned that test frequency should be

nominal (50 Hz or 60 Hz) or as close as possible to TDX 5000 – Tan Delta test set
that frequency – delta not bigger than 2 Hz
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

15. Then we described in detail the measurement

procedure – which we can sum up as

A) disconnect and deenergise the PT

B) short – circuit the windings (Prim, Sec, Ter)
C) ground the test set and connect the cables
D) set up the measurement in the firmware by
choosing where you connected the generation and
what do you want to measure and by choosing the
voltage and frequency
E) measure (CH and CHL, and maybe also CH + CHL) Short Circuiting the PT windings
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

15. …continuing

F) we hit Emergency button before approaching the PT

and we change the generation from HV winding to
LV winding in order to measure CL (we can also
measure CHL to confirm the consistence as well as
CL + CHL)
G) we now remove short circuits and focus on bushings
in bushings we can measure C1 and C2
Generating on LV side and
measuring from HV side
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

15. …continuing

H) when generating to measure C1 we can generate 12

kV (keep in mind 10 kV is often test voltage of
choice)

I) then we shift generation to measurement point

down below and current measurement to the top of
the bushing – we do not exceed 1 kV
Generating on LV side and
measuring from HV side
Workshop on Tan Delta Testing

Let us repeat what we learned during our first webinar

16. In the end we have our set of results and we need

to evaluate them. In order to evaluate them we can

A) compare them to previous values

B) compare to allowed values
C) compare phases (bushings, cable phases)
D) compare to the same type of an asset working in
similar conditions

17. If we suspect in deterioration we can do sweep Measuring TD on bushing

(voltage or frequency) or continue with maybe
TD monitoring
 Workshop on Tan Delta Testing

And this is where we pick up today

Let us explain better some of the results we

obtain and use the formulas to see how can
they be used to confirm the tan delta

U [V] I [mA] C [nF] TD % PF % Rp [𝐌Ω] Rs [𝒌Ω] ϕ QF Z [𝒌Ω]

9994.4 12.855 4.0975 0.228 0.228 342.738 1786.04 89.8692 437.937 782.392


1
𝛿 = 90° − ϕ 𝑇𝑎𝑛 𝛿 =
𝜔𝑅𝑝𝐶
𝛿 = 90° − 89.8692° = 0.1308 °
1
𝑇𝑎𝑛 0.1308° = 0.228 % = = 0.228 %
314 ∙342.738 ∙ 10 6 ∙ 4.0975 ∙ 10 −9
 Workshop on Tan Delta Testing

And this is where we pick up today

…now we calculate PF, Rs and QF

U [V] I [mA] C [nF] TD % PF % Rp [𝐌Ω] Rs [Ω] ϕ QF Z [𝒌Ω]

9994.4 12.855 4.0975 0.228 0.228 342.738 1786.04 89.8692 437.937 782.392

cos 89.8692° PF = 0.228 %

Example of results

𝑇𝑎𝑛 𝛿 0.002283 1
𝑅𝑠= =
314 ∙ 4.0975 ∙ 10 −9
= 1775 𝑄𝐹 = = 438.02
𝜔𝐶 𝑇𝑎𝑛 𝛿
Workshop on Tan Delta Testing

Why do we present the result in percentages

BTW you can have your 𝑇𝑎𝑛 𝛿 result in absolute or in percentages – (0.00228 vs. 0.228 %)
𝐼𝑅
Absolute because 𝑇𝑎𝑛 𝛿 = and it is a non dimensional number (Amps divided with Amps)
𝐼𝐶
And for small angles (and do not forget we talk about values less than 1 degree) 𝑇𝑎𝑛 𝛿 is a small number

δ (degrees) tan δ in absolute tan δ in percentages

0.1 0.00175 0.175
0.25 0.00436 0.436
0.5 0.00872 0.872
0.57 0.00995 0.995

So we simply multiply 𝑇𝑎𝑛 𝛿 in absolute with a 100 and from then on we express the value in percentages
mostly for practical reasons
Workshop on Tan Delta Testing

Typical values of Capacitance in primary equipment

This table is here only as a guideline what to expect in terms of Capacitance when measuring Tan Delta

Capacitance heavily depends on physical size of the asset – that us why bigger assets have bigger capacitance

Keep in mind - this is not written in stone – but it should serve as a guideline

Asset – Primary Equipment Expected Capacitance

Power Transformers Few nf to few tens of nF
Bushings Few hundred pF
Rotative Machines From few hundred nF to few μF
Cables Depends on the type of insulation and
length for example – for 20 kV cable in
XLPE 0.36 𝜇𝐹/km

Table showing aproximate and expected values of Capacitance in primary

equipment
Workshop on Tan Delta Testing

Before we proceed let me show you

…how to set up a firmware in step by step guide using this video

The specimen we are testing in a Capacitor – 4.45 pF and with know

issue in the insulation material

Watch carefully -
 Workshop on Tan Delta Testing

Substation KAZAN
Bay TRAFO

File name TD102

Manufacturer : Model : 1 Serial Number : 1
Session date : 2.11.2018.
Operator : EVGENIJ Instrument : STS 5000 Instrument S/N : 97236

Results

Tangent Delta Test type HV Winding

Test date and time: 2.11.2018. 14:23

Ambient condition
Ambient temperature 25,0 °C
Oil temperature 60,0 °C
Humidity 70,0 %

Capacitance CH Mode GSTg-A+B

Test sequence 1
Nominal capacitance 0,2000 nF
Nominal Tδ 5,0000 m%

Sweep None
Results

Vout [ V] I out [ A] Freq. [Hz] Cp [F] Tδ [% ] QF Loss [W] [W]@10kV QTest [VA] STest [VA] Rp [Ω]
1 9994,4 12,855m 50,0 4,0976n 0,2283 437,9366 0,2934 0,2938 -128,4669 128,4673 342,7383M

Capacitance CHL Mode UST-A

Test sequence 2
Nominal capacitance 0,2000 nF
Nominal Tδ 5,0000 m%

Sweep None
Results

Vout [ V] I out [ A] Freq. [Hz] Cp [F] Tδ [% ] QF Loss [W] [W]@10kV QTest [VA] STest [VA] Rp [Ω]

1 9994,7 10,079m 50,0 3,2127n 0,2518 397,2302 0,2538 0,2541 -100,7282 100,7285 396,7296M

Signature : Approval :

File name : TD102 Serial Number : 1 Print Date : 29.3.2020. 22:55:21

Page 1 of 3
Workshop on Tan Delta Testing

Capacitance between HV and the ground

Good T𝛿 𝑣𝑎𝑙𝑢𝑒

Capacitance between HV and LV

Good T𝛿 𝑣𝑎𝑙𝑢𝑒
Workshop on Tan Delta Testing

Specimen Capacitance

Capacitance between Tertiary and the ground CH 4,0976 nF

Good T𝛿 𝑣𝑎𝑙𝑢𝑒
CHL 3,2127 nF

CHT 3,1983 nF

Sum of capacitances
CH+CHL+CHT 10,5077 nF
Good T𝛿 𝑣𝑎𝑙𝑢𝑒 measured

CH+CHL+CHT 10,5086 nF
summed up Δ = 1,1 𝑝𝐹

Very High Accuracy 0,01 %

Workshop on Tan Delta Testing

Substation ISTOCNAJA
Capacitance & Tanδ
Bay TRAFO 1

File name – SVEL TRAFO 1

Manufacturer : SVEL Model : SVEL 26 Serial Number : 435523
Session date : 24.10.2019.
Operator : EVGENIJ Instrument : STS 5000 Instrument S/N : 21884

Results

Tangent Delta Test type HV Winding

Ambient condition
Ambient temperature 25,0 °C
Oil temperature 60,0 °C
Humidity 70,0 %

Capacitance CHL Mode UST-A

Test sequence 1
Nominal capacitance 0,2000 nF
Nominal Tδ 5,0000 m%

Sweep None
Results

Vout [ V] I out [A] Freq. [Hz] Cp [F] Tδ [% ] QF Loss [W] [W]@10kV QTest [VA] STest [VA] Rp [Ω]
1 10,0k 24,600m 50,0 7,8316n 0,1828 547,0911 0,4504 0,4499 -246,1359 246,1363 224,1598M

Capacitance CH Mode GSTg-A

Test sequence 2
Nominal capacitance 0,2000 nF
Nominal Tδ 5,0000 m%

Sweep None
Results

Vout [ V] I out [A] Freq. [Hz] Cp [F] Tδ [% ] QF Loss [W] [W]@10kV QTest [VA] STest [VA] Rp [Ω]

1 10,0k 25,610m 50,0 8,1546n 0,2185 457,7683 0,5600 0,5595 -256,1964 256,1970 180,0544M

Signature : Approval :

File name : SVEL TRAFO 1Serial Number : 435523 Print Date : 29.3.2020. 21:48:31
Page 1 of 2
Workshop on Tan Delta Testing

CHL is often considered the most important insulation

CHL Good T𝛿 𝑣𝑎𝑙𝑢𝑒

CH Good T𝛿 𝑣𝑎𝑙𝑢𝑒
Workshop on Tan Delta Testing

Specimen Capacitance

Good T𝛿 𝑣𝑎𝑙𝑢𝑒 CH 8,1546 𝑛𝐹

CHL 7,8316 𝑛𝐹

CH + CHL 15,9879
measured

CH + CHL 15,9862
Summed up
Δ = − 1,7 𝑝𝐹

Very High Accuracy

0,01 %
Workshop on Tan Delta Testing
Workshop on Tan Delta Testing

Phase A

Bad T𝛿 𝑣𝑎𝑙𝑢𝑒
Further investigation - Voltage and Frequency sweep
 Workshop on Tan Delta Testing

Voltage sweep is rarely done with busings,

but here we did it to show consistency of Phase A
Observe current
Voltage Sweep that increases with capacitance measurement
voltage

Consistent Cp
proves that
measurement
is correct and stable

T𝛿 stays high
 Workshop on Tan Delta Testing

Freq. Sweep
Bad T𝛿 𝑣𝑎𝑙𝑢𝑒
Phase A
Consistent Cp
proves that
measurement
is correct and stable

T𝛿 is showing
that ominous sign
First big and then
small

Goes to prove
that this bushing
needs to be
replaced
 Workshop on Tan Delta Testing

Freq. Sweep
Bad T𝛿 𝑣𝑎𝑙𝑢𝑒
Phase B
Consistent Cp
proves that
measurement
is correct and stable

T𝛿 is again above
the allowed value
but not as high
as Ph A

Cp is like on
phase A
 Workshop on Tan Delta Testing

Freq. Sweep
Bad T𝛿 𝑣𝑎𝑙𝑢𝑒
Phase C
Consistent Cp
proves that
measurement
is correct and stable

T𝛿 is showing
even bigger values
than before

Cp has similar
values as for
Ph A and Ph B
 Workshop on Tan Delta Testing

Freq. Sweep
Bad T𝛿 𝑣𝑎𝑙𝑢𝑒
Phase N
Consistent Cp
proves that
measurement
is correct and stable

T𝛿 is showing the
biggest values
of all 4 bushigs

Cp has similar
values as for
Ph A, Ph B, Ph C
Workshop on Tan Delta Testing

Let us keep in mind that 𝑇𝑎𝑛 𝛿 should change TD

with frequency Geometry

𝐼𝑅 𝑈ൗ
= 𝑅 = 1
𝐼𝐶 𝑈 𝑗𝜔𝑅𝐶 Moisture
൘1ൗ
𝑗𝜔𝐶
Oil conductivity
While Capacitance – being a phisical reality of a
specimen – should stay the same
Geometry
Moisture
The TD curve shifts to the right as water penetrates 1 mHz 10 mHz 100 mHz 1 Hz 10 Hz 100 Hz

the insulation – that is why when we do frequency Frequency (Hz)

sweep, we tend to see decrease and then a small
increase
Workshop on Tan Delta Testing

Conclusion – All 4 HV bushings were changed

Measurement was repeated and values were
very much within the normal values

Let us repeat – Measurement point (Tap) are usually found on HV bushings

If your bushing has no measurement Tap – then you should use
Hot collar

Also when measuring Tan Delta in bushing – it is a good idea to first

clean thoroughly the bushing to get more accurate results
Workshop on Tan Delta Testing
Workshop on Tan Delta Testing
Workshop on Tan Delta Testing

Good T𝛿 𝑣𝑎𝑙𝑢𝑒

We can see a much smaller value of Tan Delta – 0,013 %

which is quite normal for cables
Also with cables in mind, we measure only one Specimen
Only one Capacitance and Tan Delta – Conductor - GND
Workshop on Tan Delta Testing

We would like to show you a voltage

sweep and how it is configured in
firmware
Workshop on Tan Delta Testing

…. ROTARY MACHINE INSULATION COULD BE SIMULATED USING EQUIVALENT

CAPACITANCES, LIKE ON A 3 WINDING TRANSFORMER

CRS

CRS CST

CR CS CT

CR capacitance between R terminal and ground

CS capacitance between S terminal and ground
CT capacitance between T terminal and ground
38
Workshop on Tan Delta Testing

Insulation system

Electrical model for test purpose:

Two insulation systems:

•Phase to ground insulation
C1, C2, C3
•Phase to phase insulation
C1-2 C2-3 C1-3

Each capacitance models the electric

behavior of each phase.
Workshop on Tan Delta Testing
What about big capacitance

The use of resonance phenomenon is a possible solution to test high

capacitive load

By using a parallel inductor (RCTD5000) a LC resonating load is created.

This permits to minimize the output power required by

TD5000

RCTD5000 is composed by two inductors of 40H

12kV 1A

It is possible to connect one inductor or more

(4 inductors is the maximum allowed) in parallel to the
RCTD unit
capacitive load
Workshop on Tan Delta Testing
LC resonating circuit
insulation
Impedance

VTD

I_Load

By connecting a proper numbers of inductors the maximum

impedance can be set close to the test frequency (50Hz or ftest f
60Hz)
1
𝑓=
𝐿
2𝜋 𝐶
𝑛
This system minimizes the current required by TD5000 and
allows test of capacitance up to 3uF and corresponding tan N number of inductors connected
delta L 40H
C capacitive load
Workshop on Tan Delta Testing
Measurement Setup
Workshop on Tan Delta Testing
Test mode
Two ways to use the device:
• Single Test

• Tip Up/Tip Down Test (Voltage Sweep): perform a series of tests

increasing/decreasing the voltage with step of 20% rated voltage

Tip-Up Test according to IEEE 286 :

tip-up = PFhv - PFlv

On the Tip-Up test, PF can vary between 0.5%

and 1%.
Workshop on Tan Delta Testing

HOW TO EXECUTE THE TEST

1) Ground the motor’s tank

U R
z
y V S
M
x W T
Workshop on Tan Delta Testing

HOW TO EXECUTE THE TEST

2) Disconnect Star connection or star point, if possible,
otherwise perform only the 3phase general test to ground

U R
z
y V S
M
x W T
Workshop on Tan Delta Testing

HOW TO EXECUTE THE TEST

3) Connect test set to the motor

46
Workshop on Tan Delta Testing

HOW TO EXECUTE THE TEST

4) With STS family you could perform at least 7 different tests,

Execute the following tests using STS – Power transformer – 3 winding menu

N^ C Capacitance Description STS 12kV STS ref. Method

connect to
1 R between R terminal and ground R CH GSTg A + B
2 S between S terminal and ground S CL GSTg A + B
3 T between T terminal and ground T CT GSTg A + B
4 RS between R and S terminals R CHL UST A
5 ST between S and T terminals S CLT UST B
6 TR Between T and R terminals R CHT UST B
7 RST between R, S, T terminals & ground Phases Cgeneric GSTg A+B

47
Workshop on Tan Delta Testing

HOW TO EXECUTE THE TEST: DETAILS

EXAMPLE OF CONNECTIONS FOR TEST 1, 4, 6

Capacitance and tan delta between:

- R terminal and Ground (CH – GSTg A+B)
- R & S terminal (CHL – UST A)
U R
- R & T terminal (CHT – UST B) z
y V S
M
x W T

48
Workshop on Tan Delta Testing

HOW TO EXECUTE THE TEST: DETAILS

EXAMPLE OF CONNECTIONS FOR TEST 2, 5, 4

Capacitance and tan delta between:

- S terminal and Ground (CL – GSTg
A+B)
- S & T terminal (CLT – UST B) U R
z
- R & S terminal (CHL – UST A)
y V S
M
x W T

49
Workshop on Tan Delta Testing

HOW TO EXECUTE THE TEST: DETAILS

EXAMPLE OF CONNECTIONS FOR TEST 3, 4, 5

Capacitance and tan delta between:

- T terminal and Ground (CT – GSTg A+B)
- T & R terminal (CHT – UST A)
- S & T terminal (CLT – UST B)
U R
z
y V S
M
x W T

50
Workshop on Tan Delta Testing

HOW TO EXECUTE THE TEST: DETAILS

EXAMPLE OF CONNECTIONS FOR TEST 7

Capacitance and tan delta between:

- R, S,T terminal and Ground
(Cgeneric – GSTg A+B)
U R
z
(*) y V S (*)
M
x W T

NOTE:
The 2 short circuit connection (*) must be performed before starting the test

51
Workshop on Tan Delta Testing

RESULTS

TAN DELTA Insulation type

0,5% Modern epoxy resin & polyester impregnated
3–5 % Asphaltic mica windings
TAN DELTA Values should be constant over time

If TAN DELTA increase, it is indication that

- insulation overheating is occurring
-Winding is becoming contaminated

-If TAN DELTA increased > 1% by initial value → there is a significat

deterioration

52
Workshop on Tan Delta Testing

RESULTS VALUES OVER TIME

Capacitance and tan delta tested simultaneously

C TAN DELTA Indication

Decrease Increase General thermal deterioration
Increase Increase Winding contaminated
or absorbed moisture

TAN DELTA trend shows the average condition of the insulation,

since it measure the total dielectric loss in a winding

NOTE: the above “decrease” and “increase” are referred to the references
values (manufacturers, etc).

53
 Workshop on Tan Delta Testing

Later we realised that Motor and

STS/TD were not grounded in
the same point

Bad measurement. Totally unexpected values of Tan Delta and Capacitance

Workshop on Tan Delta Testing

Finally we managed to get good

and sensible result

After which we turned to

Tip Up and Tip Down test
 Workshop on Tan Delta Testing

Tip Up and Tip Down test

is done at
20%, 40%, 60%, 80% and 100%
Nominal Voltage
and then viceversa

Also we did test with freq.

different than nominal

Since we are talking about

testing of Rotary Machine – Tan
Delta values can go up to 3% or
even 5% depending on the
insulating material

Since this was freshly refurbished

motor, we got good results
Workshop on Tan Delta Testing

• Location: Europe
• After the installation of a HV transformer, the
utility started immediately to see a critical
level of equivalent gas (using 2 gas DGA unit)
• TECHIMP global monitoring system also
confirmed that Interface PDs (PD activities
between two different insulation materials)
were happening in every phases
• Utility is looking for the TECHIMP suggestion
corresponding the situation
Workshop on Tan Delta Testing

Sensor placing for TD

and PD measurment
Workshop on Tan Delta Testing

Sensor placing for TD

and PD measurement
Workshop on Tan Delta Testing

Partial Discharges (PD)

Tap Adapter
• Tap Adapters installed in the bushings capacitive taps
• Both Tank PD and Bushing PD can be detected
• Trend over time can be evaluated
Dissolved Gas Analysis (DGA ) DGA
• Monitoring dissolved H2 + CO and Moisture
Leakage current in the bushing (Tan-) & Capacitance
• Tap Adapters installed in the bushings capacitive taps
• Measures tan- and Capacity of Bushings TD-Guard
Workshop on Tan Delta Testing

•
•
•

C
Workshop on Tan Delta Testing

• Thanks to separation a precise PD Trend plotting is possible

• No amplitude increase over time

qMAX95% = PD amplitude
Nw = PD repetition rate
Workshop on Tan Delta Testing

• DGA H2 increased constantly by a rate of 60 – 70 ppm/month

• DGA H2 reached a value of 1250 ppm
• DGA CO value (not plotted) was always below 350 ppm

Hydrogen increase
confirmed due to the
three interface PD
activities inside the
Transformer!
Workshop on Tan Delta Testing

• No significant change in Tan- value of the Bushings

• No significant change in Capacitance value of the Bushings

Capacitance

Tan-
Workshop on Tan Delta Testing

• Three PD activities inside the Transformer, one for each phase

• Such PD activities are the cause of H2 increase
• High Frequency content → PD in the upper part of the Transformer
• PD identified as an Interface phenomenon between different dielectric
materials
• Stable Trend over time would suggest that no solid insulation systems
are concerned

Conclusion: PD possibly generated at three interfaces air/oil in the three

domes just below the bushings due to not perfect oil filling of the
Transformer
Workshop on Tan Delta Testing
Utility actions:
• Transformer oil was drained and degassed
• Transformer was filled up again slowly with hot oil in order to avoid any
kind of empty regions.
• Transformer was put on-service again with the monitoring installed

Result:
• Transformer PD Free!
• No further H2 increase
• Elimination of interfaces air/oil
Workshop on Tan Delta Testing
Workshop on Tan Delta Testing

UST and GST modes

There are 3 modes of testing:

UST – Ungrounded Specimen Test (Floating mesurement system)

GST – Grounded Specimen Test (Grounded measurement system)
GSTg – Grounded Specimen Test Guard (Grounded measurement system with some currents
not measured or „guarded“ – one can also say – removed – from the ammeter measurent)

Specimen = Sample (by that inteding the sample of our insulation system). We could also
freely say „Capacitor“ instead of Specimen since we are measuring the speciments (parts of
overall insulating system) of Capacitors (insulation)

Generally It is good to have as much “speciments“ as possible, since it gives us better

resolution when performing a test
Workshop on Tan Delta Testing

Examples of Speciments:
Ungrounded Specimen -
none of the sides touch the
ground

HV (H) LV (L)
CHL Equivalent Scheme
H L
R

r y
CH CL
b
B
Y

Grounded Specimens
- one side is touching
the ground
Workshop on Tan Delta Testing

Examples of Speciments:

UST
UST

HV (H) LV (L) T CHT

Equivalent Scheme

CHL CLT
H T
R
r y r y L
1 1 2 2
b

b
1

2
B
Y

CH CL CT

GST GST GST

Workshop on Tan Delta Testing

UST and GST modes

In order to have a better – resolution – it is always a good idea to test the smallest
possible sample (thus having a better chance to see the influence of a small
imperfection) – actually we can „zoom in“on separate segments of insulation (CH, CHL,
CL…)

That way we can more easiliy discern if the problem lies in (for 2 Winding transformer):

1. HV to GND insulation or
2. LV to GND insulation or
3. HV – LV insulation
4. Inside the Bushing (C1)
5. Between Bushing Tap and GND (C2)
 Workshop on Tan Delta Testing

Herein lies the first reason why

Tan Delta test is better than Polarisation CHL
Indeks – better „resolution”. HV (H) LV (L)

While with PI you mostly get one

number that describes the state of the
whole insulation, with Tan Delta you get
a more detailed insight per segment –
In order to specify the state of insulation
of each segment of insulation
CH CL
Example 1:

TD (CHL): 0.75 % TD (C1): 0.55 %

TD (CL): 0.56 % TD (C2): 0.62 %
TD (CH): 0.98 %

In this case what is obvious is that CH – insulation HV – GND is the

weakest. That in turn presents a bigger issues than if TD would be
0.98 % for CL, because HV side (having higher voltage) presents a
bigger risk for discharge than LV (because it is lower voltage)
 Workshop on Tan Delta Testing

….going back to modes of testing – or – how do we get such a nice „resolution”

One could say that the choice of method (UST, GST….) is kind of a switch that
is placed before the Ammeter/Wattmeter and that chooses which currents flows
Into the said Ammeter/Wattmeter:

All this because

In todays modern test sets the current we
We use the measurement measure is actually
Matrix reflects the
SWITCH capacitance we
mesure

Current coming to Input B Current coming from the Ground

Current coming to Input A

Workshop on Tan Delta Testing

And an example with transformer that has Primary, Secondary and Tertiary winding

SWITCH
Current coming from the Ground
Current coming to Input A

Current coming to Input B

Workshop on Tan Delta Testing

UST and GST modes

If we have 2 current measurement inputs (input A and input B) plus the
current measured through the ground we get 7 combinations overall

Here are the 7 combinations (X) marks what input is taken into consideration:
UST-A UST-B UST-A+B GST GSTg-A GSTg-B GSTg-A+B
Input A x x x x
Input B x x x x
Ground x x x x
By definition UST will never measure the current through the Ground
By definition GST will always measure the current through the Ground
Workshop on Tan Delta Testing

UST and GST modes

UST-A UST-B UST-A+B GST GSTg-A GSTg-B GSTg-A+B
Input A x x x x
Input B x x x x
Ground x x x x
So we could say also that when While for GST type, we have to clarify actually
mentioning the UST type, we have to WHAT CURRENT (CAPACITANCE) WE DO NOT
clarify actually MEASURE (from what value we guard the result):
WHAT CURRENT (CAPACITANCE) WE DO
MEASURE: GST – we measure all currents (GND, A and B)
GSTg – A – we do not measure A input
UST – A – we measure A input GSTg – B – we do not measure B input
UST – B – we measure B input GSTg – A+B – we do not measure neither A nor B
UST – A+B – we measure A and B inputs input
Workshop on Tan Delta Testing

For 2 winding PT these are the rules

Capactitance/TD to measure Method Generation on/Measurement on

CH (HV - GND) GSTg - A HV Winding / LV Winding

CL (LV – GND) GSTg - A LV Winding / HV Winding
CHL (HV – LV) UST - A Both options are accetable
CH + CHL GST HV Winding / LV Winding
CL + CHL GST LV Winding / HV Winding
C1 UST - A Top of bushing / Tap
C2 Tap / Top of bushing

Table correlating Capacitances/Measurement Methods/Placement of generation and

measurement probes
Workshop on Tan Delta Testing – Practical Example 1 – UST-A
UST-A Ungrounded Specimen Test - Input A measured – I am measuring CHL
• Both side of the Capacitor are not Grounded
• HV Generation of Test Set is connected to the HV part of the Power Transformer
• Current measured is current running through cable going to Input A

Connection Scheme HV generation of test set connected to HV side of Trafo

Measurement Input GND

Measurement Input A
 Workshop on Tan Delta Testing – Practical Example 1 – UST-A

Here you would say where have you connected the

voltage output of the testing device
In the SW it would look like this:
Here you would choose what
capacitance you would want to measure

This is determined autmatically by the test set depending on what you want to measure
 Workshop on Tan Delta Testing – Practical Example 1 – UST-A

This is what would be happening inside the test set

The fact that current going
in the ground is not fed to
the Ammeter but that
ammeter is bypassed like here CHL
means that by definition it is
UST

This is the current H L

(and thus also the capacitance)
that we do not measure

AC

CH CL
Ammeter
Cmeasured = CHL This is the current
(and thus also the capacitance)
Input A
Input B that we measure (CHL)

There is nothing connected

to Input B, so it can’t be
UST-B mode
So in modern TD measurment devices, all this is done autmatically. One simply
chooses where the HV generation of test set is connected and what one wants to
measure. The mode is chosen autmatically by test set
Workshop on Tan Delta Testing – Practical Example – GSTg-A
GSTg-A Grounded Specimen Test - Input A not measured – I am measuring
CH
• One side of the Capacitor is not Grounded
• HV Generation of Test Set is connected to the HV part of the Power Transformer
• Current measured is current through the ground

Connection Scheme HV generation of test set connected to HV side of Trafo

Measurement Input GND

Measurement Input A
 Workshop on Tan Delta Testing

Here you would say where have you connected the

voltage output of the testing device
In the SW it would look like this:
Here you would choose what
capacitance you would want to measure

This is determined autmatically by the test set depending on what you want to measure
 Workshop on Tan Delta Testing
This is what would be happening inside the test set
CHL

H L
Cmeasured = CH This is the current that we do not
measure because we have „guard A”
In this case, since we are which means „do not measure A”
AC
measuring in GST mode,
by definition it means that CH
we are measuring the Ground current CL
This also is an answers
Ammeter to the question why we
Input GND can measure CH, CHL
but also CH + CHL without
changing connections
Input B

Input A Current coming

Even thou it seems that current
to Input A is bypassed
on input A is measured,
Ammeter from being measured
because there is cable This is what happens internally Input GND by internal ammeter
connected to input A, internally
this current is not taken into
Input B
consideration because of „guard A” Input A
Workshop on Tan Delta Testing – Practical Example – GSTg-A
GSTg-A Grounded Specimen Test - Input A not measured – I am measuring CL
• One side of the Capacitor is not Grounded
• HV Generation of Test Set is connected to the LV part of the Power Transformer
• Current measured is current through the ground

Connection Scheme
HV generation of test set connected to LV side of Trafo
Measurement Input A

Measurement Input GND

 Workshop on Tan Delta Testing – Practical Example – GSTg-A

Here you would say where have you connected the

voltage output of the testing device
In the SW it would look like this:
Here you would choose what
capacitance you would want to measure

This is determined autmatically by the test set depending on what you want to measure
 Workshop on Tan Delta Testing – Practical Example – GSTg-A
This is what would be happening inside the test set
This is the current that we do not
measure because we have „guard A”
which means „do not measure A” CHL
Cmeasured = CL So actually by changing
the place where we
H L generate – from HV to LV
AC
side of Transformer, we
change from measuring
CH CH to measurement of CL
CL
Ammeter
Input GND Also CL cannot be measured
Input B if generation is on the HV winding
of Transformer.

Input A
Even thou it seems that current
on input A is measured, Current coming
Ammeter
because there is cable This is what happens internally Input GND To Input A is bypassed
connected to input A, internally from being measured
this current is not taken into by internal ammeter
Input B
consideration because of „guard A” Input A
 Workshop on Tan Delta Testing – Custom mesurement

GSTg - A+B - Grounded Specimen Test with Guard A and B

• Let us see what will we actually measure when we apply GST generating on HV winding
• Only current through the ground is measured

Let us play and see what C Measured

happens if we choose a CHL
=
custom mode - (CHL series CL ) ll CH
GSTg – A+B H L
AC

CH CL
No measurment on Input A or Input A
Input B is done since nothing is Input B
connected to it (and even if we
would connect something, it
woud not be measured)
 Workshop on Tan Delta Testing

In case you were wondering why do we use Input B

For autmatic measurement on a transformer that has primary, secodary and terriary winding.
Also in some rare cases if Input A does not work properly (either not calibrated or is burned) – we can use Input B

CHT

AC
CHL CLT
H
L T

Input B Input A CH CL CT

GND

C & Tan delta on a Three winding Transformer.

Workshop on Tan Delta Testing
Why is TD better than Polarisation Index (DC method)

1. Tan Delta results more repetable (consistant) than PI (DC) measurements

2. Tan Delta does not leave a lot of free electrons in the insulation (which DC test does) –
which can later trigger Partial Discharge activity so it is considered less destructive than DC tests

3. Tan Delta is a faster test

4. Giving us more similar conditions (AC) than PI (DC) test since we are using the said equipment with AC
voltage

5. Time Independent (Result of DC testing changes if the test duration changes)

Workshop on Tan Delta Testing
Why are TD and PD two sides of the same story

1. PD are „micro sparks” that partially bridge the insulation – TD is more like overall statistical value of the
insulation’s conductivity

2. Tan Delta shows us the oveall state of the insulation while PD is showing us if we have some issues
happening right now – at this moment

3. Seeing if something is happening right now – we can corelate certain events with the behaviour of
insulation (switching or OLTC, overvoltages, near by faults…)

4. Continous monitoring is the best solution for maintaining properly transformers

5. PD is better seen on bushings (that are very prone to deterioration) while TD sees well whole PT
Workshop on Tan Delta Testing
Negative Tan Delta

Even thou this topic itself deserves a small webinar itself we will try to simplify and list few main reason
why tan delta has negative sign.

First of all negative sign means that angle ϕ is bigger than 90 degrees

1. Connections are not done properly – check the conections

2. Test set and Test Object are not grounded in the same point (certain currents not being measured properly)
3. Special aditives could have been put in the oil (rarely)
4. How „negative” is the TD value ? Very much or it is just slightly over 90 degrees ?

If it is just slightly over 90 degrees – it might be due to measurement inaccuracy

Workshop on Tan Delta Testing

Thank you for your valuable attention

we hope you learned something
☺
MSc. Andrej Šepčić
RAM https://www.altanova-group.com/en/regional-isa-offices/isa-regional-offices
andrej.sepcic@altanova-group.com https://www.altanova-group.com/en/local-distributors/local-distributors