Understanding and management of sulfur corrosion from insulating oil

Fabio ScatiggioTERNA S.p.A. – Venice (Italy)

Massimo Pompili University “Sapienza” – Rome (Italy)

Vander TumiattiSea Marconi Technologies – Turin (Italy)

The Second Kuwait Electricity Conference & Exhibition

Generality

The term of corrosive sulfur identifies those forms of sulfur that are corrosive under normal operating thermal conditions of the transformers.

Some of sulfur compounds are beneficent to the oil because they act as retardants or passivants in the oxidation process (“good sulphur”).

2

KUWAIT , MARCH 2011

Generality

From IEC 60296: •Some sulfur compounds, e.g. mercaptans, are very corrosive to metal surfaces, i.e. steel, copper and silver (switchgear contacts) and shall not be present in new oil. This type of corrosive sulfur should be detected following DIN 51353.•Some other sulfur compounds, e.g., dibenzyldisulphide, may result in the deposition of copper sulfide (Cu2S) in paper insulation, reducing its electrical insulation properties. This has resulted in several equipment failures in service.

•IEC 62535 provides the best available assurance that Cu2S deposition in paper will not occur in service as a result of the oil used. This test method, based on work performed by CIGRE WG A2.32, will detect corrosive sulfur compounds in oil. It applies only to oils that do not contain a metal passivator additive (declared or undeclared).

3

KUWAIT , MARCH 2011

Chemistry of Corrosive Sulfur

The low amount sulfur oils are historically appreciated due to the negative properties of sulfur as well.

Sulfur compounds can cause corrosion, accelerate the asphalts production and worsen the dielectric properties of oil

Some of sulfur compounds are beneficent to the oil because they act as retardants or passivants in the oxidation process (“good sulfur”).

4

KUWAIT , MARCH 2011

Chemistry of Corrosive Sulfur

Sulfur in insulating oil = 0.01 - 0.5%, it is present in different

forms, as:

• mercaptans or thiols; in linear, cyclic and aromatic structures

• cyclic and aliphatic sulfides

• thiophens and benzothiophens

• disulfides

• polysulfides and others forms.

5

KUWAIT , MARCH 2011

Chemistry of Corrosive SulfurC C C C SH CCSCC

alchil-mercaptano dialchil-solfuro

C

CC

C

C

SH

ciclopentantiolo

C

CC

C

CC SH

cicloesantiolo

SHC

CC

C

CC

tiofenolo

C

CC

C

S

tiaciclopentano

C

CS

C

CC

tiacicloesano

C

CC

C

S

tiofene

C

CC

C

CC

C

C

S

benzotifene

6

KUWAIT , MARCH 2011

Chemistry of Corrosive Sulfur

Not all the sulfur species present in mineral oil are corrosive in the same way

Source: Lance R. Lewand “The Role of Corrosive Sulfur in Transformers and Transformer Oil” - Proceedings of the Sixty-Ninth Annual International Conference of Doble Clients, Boston, MA, USA, 2002.

7

KUWAIT , MARCH 2011

FORMULA REACTIVITYElemental Sulphur

(free) S very high reactive

Mercaptans (thiols) R-SH very reactive

Sulphides (thioethers) R-S-R1 reactive

Disulphides R-S-S-R stable/reactive

Thiophenes 5 membered ring with S very stable

Chemistry of Corrosive Sulfur

The rate of all chemical reactions is governed by temperature

The rate of sulphide formation reactions seems to approximately double for every 10ºC increase of temperature

8

KUWAIT , MARCH 2011

Total sulfur concentration isn’t related with corrosiveness The market is

requiring “free sulfur” or “low sulfur” oils.

But at the same time it is very interesting to discover that during the last 10-15 years there has been an increase in the presence of corrosive sulfur and related faults

Total sulfur against year of manufacturing

0

1000

2000

3000

4000

5000

6000

7000

19

64

19

66

19

68

19

70

19

72

19

74

19

76

19

78

19

80

19

82

19

84

19

86

19

88

19

90

19

92

19

94

19

96

19

98

20

00

20

02

20

04

20

06

20

08

Year

mg

/kg

90% cum.average value

1995: “miles stone

” year

Historical backgrounds

9

KUWAIT , MARCH 2011

Historical backgroundsElectrical industry lives for more than 70 years with only sporadic cases of transformer failures induced due to the presence of corrosive sulfur in mineral insulating oils.Scientific literature doesn’t reports about “pandemic occurrences”, until the end of the past century, when in Columbia were reported a large number of faults in shunt reactors (2003-2006).Even any official statistic exist at yet, the recurrence rated of transformer failures (tens or may be hundreds) has undergone a very marked increase all around the world:

Brazil ChinaArgentina IndiaItaly USAUnited Kingdom MexicoScandinavia RuandaMiddle East South Africa

10

KUWAIT , MARCH 2011

Test MethodsInternational standard organizations have adopted, since the ’50s, a simple ON/OFF test to detect the corrosiveness of the oils.

11

KUWAIT , MARCH 2011

Method DescriptionASTM D130 – IP 154 –

ISO 2160 COPPER STRIP AT 100°C

ASTM D1275/A - ISO 5662 - NBR 1505

COPPER STRIP AT 140°C FOR 19 HOURS

DIN 51353 SILVER STRIP AT 100°C FOR 18 HOURS

TERNA (Extended ASTM D1275)

COPPER STRIP AT 140°C FOR 48 OR 72 HOURS

DOBLE - ASTM D1275/B – NBR 1505:05

COPPER STRIP AT 150°C FOR 48 HOURS

CCD (Covered Conductor

Discoloration Test) IEC62535

COPPER AND PAPER STRIP AT 150°C FOR 72 HOURS

Test Methods

12

KUWAIT , MARCH 2011

DBDSOn 2006 Terna, Sea Marconi and University of Missouri with a cooperative research discovered the sulfur compound responsible of oil insulating corrosiveness, as: dibenzyl-disulphide (DBDS).

SS

R. Maina, F. Scatiggio, S. Kapila, V. Tumiatti, M. Tumiatti and M. Pompili, “Dibenzyl disulfide (DBDS) as corrosive sulfur contaminant in used and unused mineral insulating oils”, CIGRE Web Site, http://www.cigrea2.org/Site/Publications/download/DBDS_paper_ColorVersionID44VER52.pdf.

13

KUWAIT , MARCH 2011

DBDSUsing an innovative analytical technique based on merging gas- chromatography with atomic emission (GC-AED), capable to detect any single organic sulfur compound.

14

KUWAIT , MARCH 2011

DBDS

AED reveals the constant presence of a sulfur peak towering above a sea of uniformly sized peaks of sulfur compounds residual from deep hydro-treating

AED reveals the constant presence of a sulfur peak towering above a sea of uniformly sized peaks of sulfur compounds residual from deep hydro-treating

Nynas Nytro 10GBN ((with DBDS)

Petrobras AV 58 (DBDS free)Nynas Nytro 11GBX-US oil (DBDS

free)

15

KUWAIT , MARCH 2011

Bitumen

LMOHMOBS

Lubricants

CRUDE OIL

Topping Cracking

DieselKeroseneGasoline

Gas combustibile

Gas (H2)GasolineKerosene

DieselFuel oil

Furfural treatment

Other process

Waxeddistillate

Aromatics + Furfural

FUEL cut

Dewaxing

Paraffine & others

Hydrogenation

Spindle OIL (insulating oil base)

DBDS and Refining Process: Native or Additive?

16

KUWAIT , MARCH 2011

Hydrogenation process is able to remove the very large majority of organic sulfur compounds (OSC). Why not DBDS?

HydrogenationLP Hydrogenation

(4/10 atm)

MP Hydrogenation(40/80 atm)

HP Hydrogenation(> 180 atm)

Naphtenic base + Lubricants

ADVANTAGES:

1. High yield of production

2. Low pour point

3. Low PCA

4. Better impact on environment (low waste)

Oxidation stability ?

DBDS and Refining Process: Native or Additive?

17

KUWAIT , MARCH 2011

DBDS and Refining Process

18

KUWAIT , MARCH 2011

With the kindly permission of ErgonWith the kindly permission of Ergon

min0 5 10 15 20 25

Counts

0

10

20

30

40

50

AED3 B, Sulfur 181 (07-03-19\07-0211.D)

18.

075

- D

BD

S

17.

183

16.

191

min0 5 10 15 20 25

Counts

100

200

300

400

500

AED3 A, Carbon 179 (07-03-19\07-0211.D)

13.

462

1.0

80

min0 5 10 15 20 25

Counts

-2

-1

0

1

2

3

AED3 B, Sulfur 181 (07-03-19\07-0520.D)

16.

194

17.

189

min0 5 10 15 20 25

Counts

0

100

200

300

400

500

600

AED3 A, Carbon 179 (07-03-19\07-0520.D)

16.

374

15.

333

17.

182

12.

865

- D

BP

C 1

3.46

3

min0 5 10 15 20 25

Counts

0

1

2

3

4

5

6

AED3 B, Sulfur 181 (07-03-19\07-0521.D)

16.

192

17.

194

12.

892

min0 5 10 15 20 25

Counts

0

200

400

600

800

1000

1200

1400

1600

AED3 A, Carbon 179 (07-03-19\07-0521.D)

12.

892

- D

BP

C

17.

628

16.

685

15.

327

18.

523

19.

359

13.

460

min0 5 10 15 20 25

Counts

-1.5

-1

-0.5

0

0.5

AED3 B, Sulfur 181 (07-03-19\07-0523.D)

min0 5 10 15 20 25

Counts

0

100

200

300

400

500

AED3 A, Carbon 179 (07-03-19\07-0523.D)

12.

868

- D

BP

C 1

3.46

4

min0 5 10 15 20 25

Counts

-2

0

2

4

6

8

10

12

AED3 B, Sulfur 181 (07-03-19\07-0522.D)

16.

201

17.

196

17.

386

18.

268

16.

581

min0 5 10 15 20 25

Counts

0

200

400

600

800

AED3 A, Carbon 179 (07-03-19\07-0522.D)

15.

341

16.

386

17.

186

15.

232

18.

142

12.

865

- D

BP

C 1

3.46

2

18.

863

min0 5 10 15 20 25

Counts

-2

-1

0

1

2

3

AED3 B, Sulfur 181 (07-03-19\07-0524.D)

16.

188

17.

184

min0 5 10 15 20 25

Counts

0

100

200

300

400

500

AED3 A, Carbon 179 (07-03-19\07-0524.D)min0 5 10 15 20 25

Counts

-2

-1

0

1

2

3

AED3 B, Sulfur 181 (07-03-19\07-0519.D)

16.

199

17.

192

min0 5 10 15 20 25

Counts

0

100

200

300

400

500

600

AED3 A, Carbon 179 (07-03-19\07-0519.D)

16.

379

15.

334

13.

466

NYNAS OILS ON MARKET AFTER DBDS DISCOVERY

YES

NO

NO

NO

NO

NO

NO

Best seller Nynas oil before DBDS Discovery

DBDS typically ~150 ppmDBDS typically ~150 ppm

KUWAIT , MARCH 2011

19

SAME OIL AFTER DBDS DISCOVERY

YES

NO

Best seller Shell oil before DBDS Discovery

DBDS typically ~150 ppmDBDS typically ~150 ppm

KUWAIT , MARCH 2011

20

Paper turns contaminated by Cu2S

Contaminated Copper

Copper as well

~ 6% (w) Cu in paperDBDS effects

21

KUWAIT , MARCH 2011

Magnificatio

n 2000 x

DBDS effects

22

KUWAIT , MARCH 2011

All the metallic

surfaces of naked

copper result covered

by an adherent layer of black dust of Cu2S

Inside the transformer

23

Unpredictability

24

KUWAIT , MARCH 2011

Test on equipmentFactor

yLaboratory

Sub-Station

Response

Paper Volumetric Resistivity X Yes

Paper Dissipation Factor (tang d) X Yes

PF% (Winding – 100 : 1000V) X Yes

PF% (Reactor)X

10-70kV

X1-10kV

No

SFRA (Sweep Frequency Response Analyses) X X No

FDS (Frequency Domain Spectroscopy) X X No

RVM (Return Voltage Measurements) X X No

Partial Discharges (476kV) X No

Unpredictability

25

KUWAIT , MARCH 2011

Test on equipment Laboratory Sub - Station Response

DGA X X No

Water X X No

Chemical- Physical (NN, IFT, tan δ, BDV, etc.)

X No

2-FAL and derivative X No

Corrosive sulfur (DIN 51353, silver, 100°C, 18 h)

X No

Corrosive sulfur (ASTM D1275-A, copper, 140°C, 19 h)

X No/ Yes

Corrosive sulfur (TERNA, copper, 140°C, 72 h)

X No/ Yes

Corrosive sulfur (CIGRE CCD, copper and paper, 150°C, 72 h)

X Yes

DBDS (Dibenzyl-Disulphide) X Yes

DGA is unable to tackle the problem

Unpredictability

POST FAILURE

POST FAILURE

26

Proposed by Terna (*) :

1. Build-up of sulfides on the conductor surface until the Cu2S

film becomes mechanically unstable;

2. Transfer of debris and particles to the adjacent paper tapes;

3. Migration/diffusion of fine particles of Cu2S across the turn

insulation;

4. Increase of local thermal losses due tan δ increase;

5. Breakdown at power frequency and rated voltage stress.

Breakdown mechanism

A

dPT T tan2CVP

c

l

I

Itan

(*) F. Scatiggio, V. Tumiatti, Maina, M. Tumiatti, M. Pompili and R. Bartnikas “Corrosive Sulfur Induced Failures in Oil-Filled Electrical Power Transformers and Shunt Reactors” – IEEE Transaction on Power Delivery – Vol. 24, N° 3, July 2009

27

KUWAIT , MARCH 2011

Breakdown mechanism

Copper sulfide (Cu2S) is an electric semi-conductor

28

KUWAIT , MARCH 2011

tan Resistivity (m)

Insulating windings paper, without visible copper sulfide contamination

0.003 5 1012

Insulating windings paper, with low copper sulfide contamination

0.005 5 1010

Insulating windings paper, with high copper sulfide contamination

> 1 5 104

Mitigation TechniquesEQUIPMENT Winding replacement (high economic impact) Cooling improvement (only some cases).

OIL Oil passivation with Irgamet 39 (worldwide used, but 38% of breakdown in

Terna experience) Oil retrofilling (no faults) Chemical – Active Depolarization (DBDS elimination ).

29

KUWAIT , MARCH 2011

Retrofilling: Oil change

30

Operational history of a retro-filled shunt reactor 500 kV

Time (months) ActionCorrosiveness

Test (ASTM D1275-B code)

DBDS (mg/Kg)

0 Start-up 11 (4b) 16512 11 (4b) 13114 11 (4b) 11015 Oil change 1 (1a) 524 1 (1a) 632 Still in service 1 (1a) 8

Corrosiveness test code is based on ASTM D 1275-B, levels 1-12 correspond to tarnish levels 1a to 4c, where values higher than 9 are rated as corrosive

KUWAIT , MARCH 2011

Passivation

31

Source list of metal deactivators:

• Irgamet® 30 Ciba Triazole derivative (liquid pure reagent)• Irgamet® 39 Ciba Tolutriazole derivative (liquid pure reagent)• Irgamet ® 42 Ciba (water-soluble tolutriazole derivative)• Irgamet® BTZ Ciba Benzotriazole• Irgamet® TTZ Ciba Tolutrizole• Cobratec® TT100 Tolyltriazole (solid pure reagent)• Nynas AB - Nypass (pre-blend of 10% passivator and a transformer oil base stock)• Shell Diala Concentrate P (10% concentrate)• DSI Sulphur Inhibitor – liquid concentrate mixture• Ethanox® 4705 Albemarle N,N-disalicylidene-1,2-diaminopropane• Cuvan® 303 RT Vanderbilt N,N-bis(2-ehtylhexyl)-ar-methyl-1H-benzotriazole-1-methanamine• Cuvan 484 RT Vanderbilt 2,5-Dimercapto-1,3,4-thiadiazole derivative• Cuvan 826 RT Vanderbilt 2,5-Dimercapto-1,3,4-thiadiazole derivative• NACAP® RT Vanderbilt Sodium 2-mercaptobenzothiazole, 50% active• ROKON® RT Vanderbilt 2-Mercaptobenzothiazole• Vanchem® NATD RT Vanderbilt Disodium, 2,5-dimercaptothiadiazole, 30% active• Vanlube 601 RT Vanderbilt Heterocyclic sulfur–nitrogen compound• Vanlube 601E RT Vanderbilt Heterocyclic sulfur• Vanlube 704 RT Vanderbilt Proprietary blend

KUWAIT , MARCH 2011

Passivation

32

Cu2S Cu2S Cu2S Cu2S Cu2S Cu2S

DBDS DBDS DBDS DBDS

DBDS

DBDS

TTA and derivatives react to copper surface

Manufacturers recommendation usage is 100 mg/Kg

TTA TTA TTA TTA TTA TTA TTA TTA TTA

DBDS

DBDSDBDS

DBDS

DBDS

DBDS

KUWAIT , MARCH 2011

Passivation

Time (months)

ActionCorrosiveness test

(ASTM D1275-B code)DBDS

(mg/Kg)

Irgamet 39

(mg/Kg)0 Start-up 11 (4b) 153 < 1

12 11 (4b) 116 < 115 11 (4b) 108 < 116 Passivation 2 (1b) 106 11224 2 (1b) 109 11126 Failure 3 (2a) 106 110

Operational history of a passivated 500 kV shunt reactor, failed 10 months after passivation

Corrosiveness test code is based on ASTM D 1275-B, levels 1-12 correspond to tarnish levels 1a to 4c, where values higher than 9 are rated as corrosive

33

KUWAIT , MARCH 2011

Passivation & Side Effects: Stray gassing

34

Operational history of a passivated 380 kV power transformer, with stray gassing generation

Corrosiveness test code is based on ASTM D 1275-B, levels 1-12 correspond to tarnish levels 1a to 4c, where values higher than 9 are rated as corrosive

Time (months) Action Corrosivity (ASTM D1275-B) DBDS (mg/Kg) Irgamet 39 (mg/Kg)

H2 (μ//l)

0 Start-up 11 (4b) unknown < 1 1112 12 (4c) unknown < 1 1315 11 (4b) 127 < 1 1218 11 (4b) 121 < 1 1419 Passivation 5 (2c) 120 131 16520 5 (2c) 121 100 17523 6 (2d) 115 77 18426 7 (2e) 113 51 20332 Still in service 7 (2e) 109 43 182

KUWAIT , MARCH 2011

Passivation: silver corrosion

35

230/130 kV 150 MVA Transmission TR:• Age: 5 years• Filled in 2005 with Italian oil, later discovered as

corrosive (DBDS: 150 ppm)• Passivated two years later in 2007 • No Irgamet 39 depletion over the years• Failed in February 2010• Inspected in May 2010, by manhole

KUWAIT , MARCH 2011

ElementWeight

%

Copper 2.134

Silver 87.945

Sulphur 9.918

Mechanism:

• Pre-selector OLTC’s copper parts are not corroded

• Pre-selector OLTC’s silver parts are completely corroded

• Build-up of Ag2S on the conductor surface until the film becomes mechanically unstable

• Transportation of Ag2S particles

• Breakdown at power frequency and rated voltage stress?S

EM

/ED

X

36

Passivation: silver corrosion

KUWAIT , MARCH 2011

Passivation

37

Corrosiveness persistence

Irgamet 39 consumption

Stray gassing

H2 CO CO2 CxHy

% 8 39 33 16 16 4

Long time survey (complied over 3 years) on more than 200 reactors & transformers

Most widely mitigation action usedFound to be effective in the large majority of casesSome cases of failure, where critical conditions for instability did exist. Some evidences come from:

• Brazil, 9 shunt reactors failed in a range of 1 ÷ 22 months later

• Brazil, 1 GSU (hydraulic) failed after passivation • USA, 1 GSU (thermal) failed after 1 year from

passivation

KUWAIT , MARCH 2011

Industrial application: - more than 200 transformers - more than 2.000 tons of oil treated

Principles of the process: - The process is performed on-site - The process is performed in closed-circuit, without emptying the unit even partially - The process can be operated on loaded transformers, up to 500 kV of rated voltage

- The oil is firstly forced through a solid chemical reagent that convert DBDS and other corrosive compounds in a more polar by-product - A second step of solid reagents blocks the corrosive compounds previously converted - The oil is submitted to degassing and de-humidification before reverting back in the transformer

- The process can be done without service interruption (on-load)

Depolarization

38

KUWAIT , MARCH 2011

Real case of Selective Depolarization – survey of oil’s parameters after the treatment

Parameter Units ValuePower MVA 15

Rated voltage kV 15Oil type - Nynas Nytrafo 11

Year of installation - 2000Cooling mode - ONAFSegregation - Breathing

0123456789101112

020406080

100120140160180200

11/2

003

11/2

004

11/2

005

11/2

006

11/2

007

11/2

008

11/2

009

DBDS

Corrosivity (ASTM D 1275 B)

Dep

olar

izati

on p

roce

ss

DBD

S(m

g/kg

)

Corr

osiv

enes

s (A

STM

D 1

275

B)

Evolution of

corrosion parameter

s in 4 years

Depolarization

39

KUWAIT , MARCH 2011

Comparison

  PASSIVATION OIL CHANGEDEPOLARIZATIO

N

SIMPLICITY High Low/Medium Low

TIME CONSUMING Low Low/Medium Medium/High

ON LOAD No No Yes

EFFICIENCY Low Medium/High High

OIL’S PROPERTIES IMPROVEMENT

No Yes Yes

LONG TERM PERFORMANCE

Low High High

ENVIRONMENTAL Unknown Low High

COST Very Low Medium/High Medium/High

40

DBDS is like the rust: it never sleeps!

KUWAIT , MARCH 2011

Conclusion

In according with CIGRE A2-32 (Brochure 308) recommendations

KUWAIT , MARCH 2011

41

Conclusion

42

In according with CIGRE A2-32 (Brochure 308) recommendations

KUWAIT , MARCH 2011

THANKS FOR THE ATTENTION

fabio.scatiggio@terna.itfabio.scatiggio@terna.it

KUWAIT , MARCH 2011