0

JICA Study Team for Enhancing Efficiency of Operating Thermal Power Plants in NTPC-India

Summary of

Generator Assessment
&
GT Assessment

2010

Issued in Feb
Feb.2010
2010
 Contents 1

1. Generator Assessment procedures

2. Generator assessment results of Korba#6,

Rihand#2, Singlauli#4

3. GT Assessment procedures

4. GT assessment results of Korba#6, Rihand#2,

Singlauli#6

Issued in Feb. 2010

 2

Generator Assessment procedures

Issued in Feb. 2010

 Generator Assessment procedures 3

Required specification of stator

Insulator : Mica
Resin: Epoxy
Resin process: VPI (Vacuum Pressure Impregnation)
Target for the assessment
Main specification
Korba #6 588MVA, 16.2kV, Stator: water cooled
Rihand #2 605MVA, 20kV, Stator: water cooled
Singrauli #4 235.3MVA, 15.75kV, Stator: water cooled

Target specification is the same as required specification of stator.

Issued in Feb. 2010

 Assessment procedures 4

Assessment type
Current condition assessment
Judging from the latest test results
Remaining life assessment (RLA)
Judging from lots of historical test data
Generator part for the assessment
Stator coil insulation
Stator coil insulation is most deteriorated.
Sorts of stator coil insulation diagnosis for assessment
1)Insulation resistance test (IR test)
2) Polarization index test (PI test)
3) Tanδ test
4) Step voltage test
5) Partial discharge test (Corona test)

Assessment Is conducted according to Japanese criteria.

Issued in Feb. 2010

 5

Generator assessment results of

Korba#6, Rihand#2, Singlauli#4

Issued in Feb. 2010

 Korba#6 assessment results 6

Test item Test record Evaluation Remark

Current condition assessment
IR test 0.0628930 MΩ Not available
PI test 0.994 Not available = IR (8 min) / IR (1 min)
RLA
PI test – Not available

Korba#6 does not drain and dry stator coil.

So, Korna#6 test data is meaningless for the assessment.

Issued in Feb. 2010

 Rihand#2 assessment results 7

Test item Phase Test record Evaluation Remark/standard

Current condition assessment
IR test 3-phase 1000 MΩ Good JP ≥ 50 MΩ,
NTPC ≥ 21 MΩ
PI test 3-phase 3.75 Good JP, NTPC ≥ 2.0
Tanδ test R 0.74% Good JP: Δtanδ < 2.5%
(ΔTanδ) Y 1.08% Good
B 1.24% Good
Remaining life assessment
PI test 3-phase – Not Fluctuates PI data
available
Tanδ test – – Not No more than one test result
available

Issued in Feb. 2010

 Estimation of Tanδ0 & Tanδmax (Y phase) 8

Tanδ test (Y phase)

3.0%

2.5%
0.0574x
y = 0.0093e
Tanδ(%)

2.0%

1.5%

1.0%

0.5%
0 2 4 6 8 10 12 14
Test voltage (kV)

Issued in Feb. 2010

 Rihand#2 RLA 9
Multiple historical Tanδ data is not available. Therefore
RLA is conducted by multiple historical PI test
PI test historical trend (Whole phases test)
8
PI data has fluctuated recently
7 No deterioration trend
6 Ave.=3.7
5
PI da ta

4
y = -0.1933x + 6.6281
3
Cautio n criteria: Les s than 2.0
2

1 Impossible to conduct RLA

Dang erous criteria: Les s than 1.0
0
0 5 10 15 20
(Jul-1989) Operation year (year)

Issued in Feb. 2010

 Singraulii#4 assessment results 10

Current condition assessment

Phase Test Evaluation Remark/standard
record
IR test R 3100 MΩ Good JP ≥ 50 MΩ, NTPC ≥ 17 MΩ
Y 3500 MΩ Good
B 1900 MΩ Good
PI test R 2.58 Good JP, NTPC ≥ 2.0
Y 2.86 Good
B 2.11 Good
Tanδ test R 0.01% Not available JP: Δtanδ < 2.5%
(ΔTanδ) Y -0.03% Not available The test voltage is low.
B 0.02% Not available
Partial – – Not available The evaluation items differ
discharge test from those in Japan.

Issued in Feb. 2010

 Singrauliu#4 current condition assessment 11
Results of Tanδ
0.93
0.92 ●: R phase
●: Y phase
0.91 ●: B phase
0.9
Tanδ(%)

0.89
0.88
y = 0.8791e 0.0028x
0.87
0.86
0.85
y = 0.8567e -3E-05x
0.84
0 1 2 3 4 5 6
Test voltage (kV)
Technically, the higher test voltage is, the higher tanδdata become.
But Y-phase data declines. So, these data are not reliable.

Impossible to conduct Current condition by Tanδ test

Issued in Feb. 2010

 Singraulii#4 assessment results 12

RLA
Test Phase Operation Year
years
PI test R 32 years 2015 Up to the caution level *1
39 years 2022 Up to the dangerous level *1
Y 29 years 2012 Up to the caution level *1
36 years 2019 Up to the dangerous level *1
B – Not available The PI data does not show a
deteriorating trend.
Tanδ All – Not available The test voltage is low.
*1: PI data is easily changeable due to stator dryness level

Issued in Feb. 2010

 Singrauli #4 RLA 13
RLA by PI test (R phase)
PI historical trend (R-phase)
7
y = -0.1371x + 6.3445
6

5 Approximation formula by
Excel function
4
PI

2
Caution criteria: Less than 2.0
1
Dangerous criteria: Less than 1.0
0
0 10 20 30 40
(1983) Operation year (year)

Although PI data fluctuate due to stator coil dryness,

the deterioration trend is shown.

Issued in Feb. 2010

 Recommendations 14
1. Conducting insulation diagnosis without cooling water (Korba)
It is recommended that Korba#6 conduct the insulation
diagnosis without cooling water so that Korba#6 can grasp the
current condition, the deterioration trend.
2. Conduct test by appropriate test voltage (Singrauli)
When Tanδ test is conducted, it is required to conduct it by
appropriate test voltage. Because it mainly focuses on partial
discharge in voids of the insulation. The partial discharge does
not occur at low test voltage.

<General>
1. Continue to conduct the insulation diagnosis
NTPC should conduct the insulation diagnosis periodically in
the future so that NTPC can grasp the deterioration trend.
2. Get reliable and accurate test data
Some test data is not consistent with technical theory. Check
the test data and get reliable and accurate test data .
Issued in Feb. 2010
 Status of conducting insulation diagnosis test 15

Plant IR PI S-V Tanδ PD Remarks

Korba#6 △*1 △*1 − − −
Singrauli#4 ○ ○ − ○*2 ○*3
Rihand#2 ○ ○ − ○ −
Japan ○ ○ (○)*4 ○ ○
S-V: Step-voltage test PD:Partial discharge test
*1: Test with cooling water is meaningless.
*2: Max test voltage is low.
*3: Measurement data is different from Japanese one
*4: Some OEMs do, the others do not in Japan.
Insulation diagnosis is totally evaluated by all the tests in
Japan.
Issued in Feb. 2010
 16

GT Assessment procedures

Issued in Feb. 2010

 Required specification & Target GT 17

Required specification of GT
Oil immersed transformer
Insulation paper : craft paper
Sealed type
No absorbent in oil
Target for the assessment
Unit Main specification
Korba#6 Single phase 200 MVA × 3 units, OFAF , Sealed type
Rihand#2 Single phase 201.7 MVA × 3 units, OFWF, Sealed type
Singrauli#6 Single phase 200 MVA × 1 unit, OFWF, Sealed type

Target specification is the same as required specification.

Issued in Feb. 2010

 Assessment method 18

Assessment type
Current condition assessment
Judging from the latest test data
Remaining life assessment (RLA)
Judging from lots of historical test data
GT part for the RLA
Coil insulation paper
Because Coil insulation paper is most deteriorated.
Sorts of GT assessment
1)Insulation resistance test (IR test)
2) Dissolved gas analysis (DGA)
3) Furfural analysis
4) CO+CO2 analysis

Assessment Is conducted according to Japanese criteria.

Issued in Feb. 2010

 DGA Judgment criteria in Japan 19
Judgment criteria in Japanese criteria (ppm)
Criteria Caution level 1 Caution level 2 Abnormal level
Item
TCG ≥ 500
H2 ≥ 400
CH4 ≥ 100
C2H6 ≥ 150
C2H4 ≥ 10 ≥ 10 & TCG ≥ 500 ≥ 100 & TCG ≥ 700 or
≥ 100 & the increasing ratio
of TCG ≥ 70 per month
C2H2 ≥ 0.5 ≥ 0.5 ≥5
CO ≥ 300
*1: Caution level-1
The level that transformer is out of normal condition although it is not judged to be abnormal
and dangerous condition
*2: Caution level-2
The level that transformer becomes abnormal condition gradually
*3: Abnormal level
The level that transformer is clearly abnormal condition. (aggravating further from Caution
level-2)
ΔTCG means the gap of TCG between the test and the previous test
Issued in Feb. 2010
 CO+CO2 Analysis Judgment Criteria 20
7. Japanese assessment criteria
Level Criteria Remarks
(CO+CO2 volume)／(Insulation paper weight)
Caution 0.2 Ave. DP=
level mL/g 450 level(*1)

(mL/g)
Dangerous 2.0 Ave. DP=

(CO+CO2 volume)/(Insulation paper weight)

level mL/g 450 level(*2)

*1： Transformer with DP of 450, with

the lowest CO+CO2 volume per
insulation paper weight
*2： Transformer with DP of 450, with
the highest CO+CO2 volume per
insulation paper weight

Assessment is conducted
according to the criteria.

Ave. DP ofIssued
insulation paper2010
in Feb.
 Furfural Analysis Judgment Criteria 21
7. Japanese assessment criteria
(Furfural weight)／(Insulation paper weight)
Level Criteria Remarks

(mg/g)
Caution 0.0015 Ave. DP=
mg/g 450 level(*1)

(Furfural weight)/(Insulation paper weight) 

level

Dangerous 0.015 Ave. DP=

level mg/g 450 level(*2)

*1： Transformer with DP of 450, with

the lowest furfural weight per
insulation paper weight
*2： Transformer with DP of 450, with
the highest furfural weight per
insulation paper weight

Assessment is conducted
according to the criteria.
Ave. DP of insulation paper

Issued in Feb. 2010

 22

GT assessment results of
Korba#6, Rihand#2, Singlauli#6R

Issued in Feb. 2010

 Korba#6R DGA Result 23
H2 CH4 C2H4 C2H6 C2H2 C3S CO CO2 TCG TCG increase
Date Remarks
ppm ppm ppm ppm ppm ppm ppm ppm ppm ppm/month
Caution level-1 ≧400 ≧100 ≧10 ≧150 ≧0.5 ≧300 ≧500
C2H4≧10+TCG≧500 or
Caution level-2 ≧10 ≧0.5 ≧500
C2H2≧0.5
C2H4≧100+TCG≧700 or
Abnormal level ≧100 ≧700 ≧70
ΔTCG≧70ppm/month
Apr-1990 387 3 0 2 51 0 34 47 477 Abnormal level *1
Nov-1990 11 8 0 0 0 0 82 819 101 -53.7
Feb-1995 80 17 0 30 0 10 15 1723 152 1.0
Sep-1995 75 16 0 22 0 23 10 1855 146 -0.9
Jun-2000 85 56 50 58 0 28 381 2853 658 8.8 Caution level-2
Oct-2000 84 54 52 84 0 12 398 3606 684 6.0 Caution level-2
Jul-2005 20 60 22 66 0 28 450 3810 646 -0.7 Caution level-2
Oct-2005 16 70 20 70 0 30 440 3890 646 0.0 Caution level-2
Jan-2006 42 60 16 67 0 25 420 3710 630 -5.5 Caution level-2
Apr-2006 Recently within criteria
32 62 2 118 0 60 262 4135 536 -26.9 Caution level-1
Jul-2006 23 15 1 50 0 10 130 1906 229 -99.0
Oct-2006
Apr-2007
So, Good condition
29
16
24
30
1
2
26
38
0
0
8
5
128
91
1903
1597
216
182
-4.9
-5.2
Jul-2007 12 42 0 45 0 10 75 2062 184 0.9
Oct-2007 102 24 0 31 0 9 34 1394 200 4.9
Dec-2007 91 35 0 51 0 12 43 2111 232 17.8
Mar-2008 70 30 0 29 0 8 40 1922 177 -17.6
Jun-2008 71 2 2 8 0 0 127 1865 210 10.8
Aug-2008 60 20 0 25 0 7 25 1229 137 -32.7
Jul-2009 21 81 1 76 0 31 111 1869 321 16.8
Dec-2009 18 61 2 58 0 21 109 1916 269 -9.8
Mar-2010 25 32 0 33 0 2 92 1790 184 -26.8
Jun-2010 21 49 3 91 0 4 112 1896 280 30.0
Issued in Feb. 2010
 Estimation of CO+CO2 generation speed (R phase)
(Korba#6) 24

Dangerous level =2.0

1.2

1.0
Average RLA level = 1.1
CO+CO2 (mL/g)

0.8
Approximation formula of CO+CO2
0.6 generation speed (ml/g-year)
0.4
y = 0.0283x
0.2
Caution level =0.2
0.0
0 5 10 15 20
(1989 Jan.) Operation year (Years)

The generation speed of CO+CO2 per insulation paper

weight is estimated by approximation formula function of
Excel soft and the current condition & the RLA are
estimated.

Issued in Feb. 2010

 DP estimation by CO+CO2 analysis(Korba#6) 25
(CO+CO2 volume) / (Insulation paper weight) (mL/g)

R phase: 0.4400
Y phase: 0.5194
B phase: 0.6074

R phase: 335
Y phase: 318 R phase: 652
B phase: 295 Y phase: 622
B phase: 595

Ave. DP of insulation paper

Issued in Feb. 2010

 Furfural generating trend (Korba#6) 26

0.010
Dangerous level=0.015
0.009
0.008
R phase Average RLA level=0.00825
0.007
Furfural (mg/g)

Y phase
0.006
B phase
0.005
0.004
0.003
0.002
0.001 Caution Level=0.0015
0.000
0 5 10 15 20 25
(Jan.1989) Operation years (years)

Because 2010 data is little bit larger than 2009 data, furfural
generating speed increases.
Issued in Feb. 2010
 DP estimation by furfural analysis (Korba#6) 27

DP estimation by
(mg/g)

2010 June data.

(Furfural weight)/(Insulation paper weight) 

R phase: 0.00463
Y phase: 0.00541
B phase: 0.00386

R phase: 335 R phase: 675

Y phase: 318 Y phase: 645
B phase: 352 B phase: 710

Ave. DP of insulation paper

Issued in Feb. 2010
 Korba#6 assessment results 28
Current condition assessment
Test Phase Test record Evaluatio Remark/standard
n
IR test R 3500 MΩ Good
Y 3500 MΩ Good JP ≥ 80 MΩ
B 3250 MΩ Good
DGA R DGA data Good
Y DGA data Good
B DGA data Good
CO+CO2 R 0.4305 mL/g Caution Caution ≥ 0.2 mL/g
analysis Y 0.5082 mL/g Caution Danger ≥ 2.0 mL/g
B 0.5943 mL/g Caution
Furfural R 0.00463 mg/g Caution Caution ≥ 0.0015 mg/g
analysis Y 0.00541 mg/g Caution Danger ≥ 0.015 mg/g
B 0.00386 mg/g Caution
Issued in Feb. 2010
 Korba#6 assessment results 29

RLA
RLA Operation Year
years
CO+CO2 R 53.7 Sep.-2042 Up to the average service
analysis Y 45.5 Jul.-2034 life*1
B 38.9 Dec.-2027
Furfural analysis R 38.2 Mar-2027 Up to the average service
Y 32.8 Oct-2021 life*2
B 45.9 Nov-2034

*1:The average service life point was set to 1.1 mL/g, which is the average
between the caution level (0.2 mL/g) and the danger level (2.0 mL/g).
*2:The average service life point was set to 0.00825 mg/g, which is the
average between the caution level (0.0015 mg/g) and the danger level
(0.015 mg/g).

Issued in Feb. 2010

 Rihand#2 assessment results 30
Current condition assessment
Phase Test record Evaluation Remark/standard
IR test R 707 MΩ Good JP ≥ 80 MΩ
Y 979 MΩ Good
B 835 MΩ Good
DGA R DGA data Good
Y DGA data Good
B DGA data Good
CO+CO2 R 0.95 mL/g Caution Caution ≥ 0.2 mL/g
analysis Y 0.90 mL/g Caution Danger ≥ 2.0 mL/g
B 0.97 mL/g Caution
Furfural R 0.16328 mg/g Danger Caution ≥ 0.0015 mg/g
analysis Y 0.07161 mg/g Danger Danger ≥ 0.015 mg/g
B 0.07161 mg/g Danger The measurement data is
not reliable.
Issued in Feb. 2010
 Rihand#2 assessment results 31
RLA
Test Phase Operation Year
years
CO+CO2 R 23.9 May-2013 Up to the average service
analysis Y 25.1 Jul.-2014 life*1
B 23.4 Nov.-2012
Furfural R – Not available Up to the average service
analysis Y – Not available life*2
B – Not available The measurement data is not
reliable.

*1:The average service life point was set to 1.1 mL/g, which is the average
between the caution level (0.2 mL/g) and the danger level (2.0 mL/g).
*2:The average service life point was set to 0.00825 mg/g, which is the
average between the caution level (0.0015 mg/g) and the danger level
(0.015 mg/g).

Issued in Feb. 2010

 Singraulii#6R assessment results 32
Current condition assessment
Test record Evaluation Remark/standard

IR test HV-E 250 MΩ Good JP ≥ 80 MΩ

LV-E 150 MΩ Good
HV-LV 300 MΩ Good
DGA DGA data Good
CO+CO2 analysis 1.38 mL/g Caution JP Caution ≥ 0.2 mL/g
JP Danger ≥ 2.0 mL/g
Furfural analysis 0.0283 mg/g Danger JP Caution ≥ 0.0015 mg/g
JP Danger ≥ 0.015 mg/g
The measured values are
not reliable.

Issued in Feb. 2010

 Singraulii#6R assessment results 33
RLA
Operation Year
years
CO+CO2 18.0 Jun-2005 Up to the average service life*1
analysis 32.7 Sep-2019 Up to the danger level
Furfural – Not available Over danger level
analysis The measured values are not reliable.

*1:The average service life point was set to 1.1 mL/g, which is the average
between the caution level (0.2 mL/g) and the danger level (2.0 mL/g).
*2:The average service life point was set to 0.00825 mg/g, which is the
average between the caution level (0.0015 mg/g) and the danger level
(0.015 mg/g).

Issued in Feb. 2010

 Recommendations for GT assessment 34
1. Conduct DP measurement (Singrauli)
To know actual GT condition, DP measurement of insulation
paper is effective. Conduct DP test when Singrauli GT is
replaced in order to confirm the correlation between CO/CO2 and
DP.
<General>
1. Continue to conduct the tests
NTPC should conduct the tests periodically in the future so that
NTPC can grasp the deterioration trend.
2. Get reliable and accurate test data
Some test data is not consistent with technical theory. Check the
test data and get reliable and accurate test data .
3. When NTPC purchases transformer, NTPC should get the
initial DP and the insulation paper weight from the manufacturer
in order to conduct CO+CO2/furfural analysis in the future.
4. Accurate oil treatment records are required for CO+CO2 analysis.
5. Accurate oil change records are required for Furfural analysis.

Issued in Feb. 2010

 35

Thank you!

Issued in Feb. 2010