practicing dga 09072010
TRANSCRIPT
-
8/10/2019 Practicing DGA 09072010
1/172
A BPL GLOBAL COMPANY
A BPL GLOBAL COMPANY1
-
8/10/2019 Practicing DGA 09072010
2/172
A BPL GLOBAL COMPANY2
Serveron Corporation 2008. No use without permission.
2
Agenda
Serveron Corporate Background
Transformer Reliability
Dissolved Gas Analysis
DGA Interpretation
Accuracy & Gas levels
On-Line and Lab DGA Case Studies
-
8/10/2019 Practicing DGA 09072010
3/172
A BPL GLOBAL COMPANY3
Serveron Corporation 2007. No use without permission.
3
-
8/10/2019 Practicing DGA 09072010
4/172
A BPL GLOBAL COMPANY4
Serveron Corporation 2007. No use without permission.
4
Hillsboro, USA Headquarters Branch offices in Beijing and Brussels
Incorporated in 2001 EPRI and proprietary GC based
technology Siemens investment and re-
branding agreement-2006
Fully integrated operation Engineering Manufacturing Sales & Marketing Field Service Secure Data Center
Customer Base Over 1000 On-line DGA monitors Over 80 major utilities worldwide Most major transformer
manufacturers
-
8/10/2019 Practicing DGA 09072010
5/172
A BPL GLOBAL COMPANY5
Serveron Corporation 2007. No use without permission.
5
Michel Duval
Dr Michel Duval obtained a B.Sc. in chemical engineering in 1966and a Ph.D. in polymer chemistry in 1970. He has joined IREQ
(Institut de recherche dHydro Qubec) in 1970. Since then, he hasmade significant contributions in 3 main fields of R&D: dissolved gas-in-oil analysis (DGA), electrical insulating materials and lithium-polymer batteries.
In the field of DGA, M. Duval: Is well-known for his Triangle method of DGA interpretation, usedworldwide.
Has developed and promoted the use of gas-in-oil standards in the IECand ASTM standards
Has established the typical levels of gas formation observed in varioustypes of electrical equipment in service, now used as a reference by theindustry
Has been the Convenor of several IEC working groups and CIGRE taskforces and is the principal author of several IEC international standardson DGA (60567, 60599, 61181).
-
8/10/2019 Practicing DGA 09072010
6/172
A BPL GLOBAL COMPANY6
Serveron Corporation 2007. No use without permission.
6
Michel Duval
In the field of electrical insulating oils, M.Duval has researched andpublished in the areas of: Metal passivators
Oil reclamation timing
Unstable oil detection and prevention
Paraffinic content of oils
The characterization of XLPE in HV cables and of HV outdoors insulators
M. Duval holds 16 patents and is the author of more than 75 scientific
and technical papers and books , 5 international standards (IEC,
ASTM), and numerous technical reports and presentations in
conferences. He is a Fellow of IEEE and of the Chemical Institute ofCanada. He may be contacted at [email protected].
-
8/10/2019 Practicing DGA 09072010
7/172
A BPL GLOBAL COMPANY7
Serveron Corporation 2007. No use without permission.
7
-
8/10/2019 Practicing DGA 09072010
8/172
A BPL GLOBAL COMPANY8
Serveron Corporation 2007. No use without permission.8
Transformer Asset Management
Transformers are a critical and costly element in the electrical grid
Unplanned failures at any point in the transformer lifecycle havemajor consequences
DGA condition assessment has been recognized for over 30 years forimproving reliability and lowering transformer asset maintenancecosts
TimeNew End of
Design Life
FailureRate
Bathtub Curve
-
8/10/2019 Practicing DGA 09072010
9/172
A BPL GLOBAL COMPANY9
Serveron Corporation 2007. No use without permission.9
FailuresThey Happen!
-
8/10/2019 Practicing DGA 09072010
10/172
A BPL GLOBAL COMPANY10
Serveron Corporation 2007. No use without permission.10
One Example
True Story: 520 MVA GSU Transformer
$3.5 million replacement cost
$0.5 million environmental cleanup
$1.5 million/day spot market buy
$17 million loss in
eight days!!
-
8/10/2019 Practicing DGA 09072010
11/172
A BPL GLOBAL COMPANY11
Serveron Corporation 2007. No use without permission.11
Increasing Failure Rates
All catastrophic failures in any USA news media
Reported number of failures is increasing
Total % of explosive failures is increasing
0
50
100
150
200
250
79 80 82 84 86 88 90 92 94 96 98 00 02NumberofExplosions
SergiTransformer Explosions in the United States
-
8/10/2019 Practicing DGA 09072010
12/172
A BPL GLOBAL COMPANY12
Serveron Corporation 2007. No use without permission.12
Increasing Failure Rates
William H. Bartley P.E., Hartford Steam Boiler Inspection & Insurance Co.,"Life Cycle Management of Utility Transformer Assets,"
0
1
2
3
4
5
1973 1978 1983 1988 1993 1998 2003 2008 2013
Transforme
r
FailureRate(%)
Projected rates will reach unacceptable levels if nothing is done to
improve the system;
Replacing the fleet is not an alternative
One solution is increased monitoring
Published reports in the news validate an increasing failure rate
-
8/10/2019 Practicing DGA 09072010
13/172
A BPL GLOBAL COMPANY13
Serveron Corporation 2007. No use without permission.13
Factors Leading to Increased Failures
Stress experienced:
Mechanical
Thermal
Electrical
Chemical
Tight budgets:
Cutbacks in O&M
Deferred capital replacement
Stress with age: The average age of US power transformers is>42 yrs, increasing 0.7 yrs./yr.
Age itself is not a cause of failure
Increased energy demand: Transformer peak and average
loading has increased
-
8/10/2019 Practicing DGA 09072010
14/172
A BPL GLOBAL COMPANY14
Serveron Corporation 2007. No use without permission.14
Modes of Stress
MECHANICAL STRESS
Failures due to internal stress, weather,and accidents:
Tank failure Radiator failure Winding buckling Core damage
Insulation damage LTC failures
Lead failures Bushing failures
ELECTRICAL STRESS
Failures initially categorized into areassuch as over-voltage or partial
discharge, typically accompanied bythermal or chemical failure:
Switching surges Lightning
THERMAL STRESS
Failures due to insulation destruction orconductor burn-through as the result of:
Short or long-term overloading Faults
Undersized leads Contact coking
Bad joints Loss of cooling Design issues
CHEMICAL STRESS
Failures due to:
Ingress of water
Ingress of oxygenLoss of insulating oil
Paper degradation due to heat or aging
-
8/10/2019 Practicing DGA 09072010
15/172
A BPL GLOBAL COMPANY15
Serveron Corporation 2007. No use without permission.15
Types of Faults
Other
Lightning
Through FaultsInsulation
Deterioration
Inadequate
Maintenance
Moisture
Loose
Connections
Overloading
Workmanship
Source: Hartford Steam Boiler Inspection and Insurance Co.
-
8/10/2019 Practicing DGA 09072010
16/172
-
8/10/2019 Practicing DGA 09072010
17/172
A BPL GLOBAL COMPANY17
Serveron Corporation 2007. No use without permission.17
-
8/10/2019 Practicing DGA 09072010
18/172
A BPL GLOBAL COMPANY18
Serveron Corporation 2007. No use without permission.18
Monitoring of Gases in Transformers
As insulating material breaks
down due to stress, gases are
formed which dissolve in the
transformer oil
Levels and combinations of the
gases formed are used to detect
incipient faults
For over 50 years, DGA has been
the leading tool to assess
transformer condition
ISSOLVED
AS
NALYSIS
http://members.clipartinc.com/gallery/cgi-bin/imageFolio.cgi?direct=Animations/Mini/Alphabets/Liquid_Letters&img=18http://members.clipartinc.com/gallery/cgi-bin/imageFolio.cgi?direct=Animations/Mini/Alphabets/Liquid_Letters&img=18http://members.clipartinc.com/gallery/cgi-bin/imageFolio.cgi?direct=Animations/Mini/Alphabets/Liquid_Letters&img=18 -
8/10/2019 Practicing DGA 09072010
19/172
A BPL GLOBAL COMPANY19
Serveron Corporation 2007. No use without permission.19
Monitoring of Gases in Transformers
Gas originates from many places:
Mineral insulating oil Conductor paper insulation
Pressboard barriers
Other materials
Gas generation is related to highmaterial temperatures (150C to
1,000C).
Gases are symptoms of:
Poor design or construction Too much electrical stress
Too much thermal stress
Too many short circuits
Overall poor condition
-
8/10/2019 Practicing DGA 09072010
20/172
A BPL GLOBAL COMPANY20
Serveron Corporation 2007. No use without permission.20
Which Gases are Generated?
Insulating Oil
Eight key gases in transformer oil are associated with fault
conditions.
DGA detects the level of gases indicative of incipient faults
which may lead to transformer failure.
Acetylene
(C2H2)
Oxygen (O2)
CarbonDioxide (CO2)
CarbonMonoxide (CO)
Ethylene
(C2H4)
Hydrogen
(H2)
Methane
(CH4)
Ethane
(C2H6)
-
8/10/2019 Practicing DGA 09072010
21/172
A BPL GLOBAL COMPANY21
Serveron Corporation 2007. No use without permission.21
The Importance of DGA to Reliability
DGA enables the detection of the presence and severity of faults:
Air leaks
Hot metal faults
Arcing and partial discharges
DGA may help to indirectly detect root cause of faults in:
Windings (short circuits, insulation failure)
Cleats and leads (high contact resistance, loose contacts)
Tanks (ground problems, circulating currents)
Tap changer (resistive contacts, leaks into main tank)
Core (magnetic flux problems)
Oxidation of materials (mostly CO, CO2)
-
8/10/2019 Practicing DGA 09072010
22/172
A BPL GLOBAL COMPANY22
Serveron Corporation 2007. No use without permission.22
-
8/10/2019 Practicing DGA 09072010
23/172
A BPL GLOBAL COMPANY23
Serveron Corporation 2007. No use without permission.23
Traditional DGA Methods
Samples of oil are manually taken
on a set schedule.Samples are sent to Accredited Labs
DGA labs use Gas Chromatography (GC)
-
8/10/2019 Practicing DGA 09072010
24/172
-
8/10/2019 Practicing DGA 09072010
25/172
A BPL GLOBAL COMPANY25
Serveron Corporation 2007. No use without permission.25
Standards and Guidelines GoverningLaboratory DGA
ASTM D3612-2002 Standard Test Method for Analysis of Gases Dissolved inElectrical Insulating Oil by Gas Chromatography
IEC 60567-2005 Oil-filled electrical equipment - Sampling of gases and of oil
for analysis of free and dissolved gases
Some other ASTM standards are available;
D3613-1998 - Standard Practice for Sampling Insulating Liquids for Gas
Analysis and Determination of Water Content
D3305-95 (1999)Standard Practice for Sampling Small Gas Volume in
a Transformer
D2759-2000Standard Practice for Sampling Gas from a Transformer
under Positive Pressure
-
8/10/2019 Practicing DGA 09072010
26/172
A BPL GLOBAL COMPANY26
Serveron Corporation 2007. No use without permission.26
Oil Sampling
Manual Sampling A small volume of oil (30 mL) is collected in a gas-tight syringe,
using a 3-way valve, then transported to the laboratory
ASTM method D3613 details procedures for oil sample handling
On-Line Sampling A small volume of oil is continuously circulated through the
monitor and then returned to the transformer
The oil is sampled and analyzed for gas content by the monitor
On-Line monitors offer a closed-loop repeatable oil sampling
process, with no possibility for contamination
-
8/10/2019 Practicing DGA 09072010
27/172
A BPL GLOBAL COMPANY27
Serveron Corporation 2007. No use without permission.27
Oil Sampling Guidelines
References:
ASTM D923: Standard Practice for Sampling Electrical Insulating Oils for
Gas Analysis
IEC 60475: Method of Sampling Liquid Dielectrics
IEC 60567: Guide for the Sampling of Gases and Oil from Oil-filled
Electrical Equipment for the Analysis of Free Dissolved Gases
Sampling of Dielectric Liquids by Lance Lewand, Doble Client
Conference, 2003
-
8/10/2019 Practicing DGA 09072010
28/172
A BPL GLOBAL COMPANY28
Serveron Corporation 2007. No use without permission.28
Gas Extraction
Dissolved gases are present in transformer oil atconcentrations from
-
8/10/2019 Practicing DGA 09072010
29/172
A BPL GLOBAL COMPANY29
Serveron Corporation 2007. No use without permission.29
Gas Extraction
METHOD A (partial de-gassing method) Introduce the oil sample into a pre-evacuated vessel
The extracted gases are compressed to atmospheric pressure
and the total volume measured
The gases are then analyzed by gas chromatograph
METHOD B
The oil is sparged with a carrier gas on a stripper column
The gases are then flushed from the stripper column into a gas
chromatograph for analysis
-
8/10/2019 Practicing DGA 09072010
30/172
A BPL GLOBAL COMPANY30
Serveron Corporation 2007. No use without permission.30
Gas Extraction
METHOD C (Headspace Method) Introduce an oil sample in a glass vial with a headspace gas
phase of argon above it.
Some of the gases migrate from the oil to the headspace and
equilibrate according to Henrys law.
Henrys Law - At a constant temperature, the amount of a given gas
dissolved in a given type & volume of liquid is directly proportional to the
partial pressure of that gas in equilibrium with that liquid.
At equilibrium, the headspace is transferred to a sample loop thenintroduced into a gas chromatograph
-
8/10/2019 Practicing DGA 09072010
31/172
A BPL GLOBAL COMPANY31
Serveron Corporation 2007. No use without permission.31
Gas Chromatography
Mobile phase = gas
Stationary phase
= solid adsorbent or liquid
Mobile gas phase flows
through the column
under pressure0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
0 50 100 150 200 250
Time (s)
Intensity
1 2 3
Mobile
phase
flow
Stationary phase
3 2 1
Gas Detector
-
8/10/2019 Practicing DGA 09072010
32/172
A BPL GLOBAL COMPANY32
Serveron Corporation 2007. No use without permission.32
-
8/10/2019 Practicing DGA 09072010
33/172
A BPL GLOBAL COMPANY33
Serveron Corporation 2007. No use without permission.33
Gas Solubility Coefficients
Some gases dissolve into the oil more than others
This is known as Solubility In the following graph, Ethane, C2H6dissolves the most and
Hydrogen, H2the least.
Solubility coefficients change over temperature Gases dissolve into oil in different amounts as temperature
changes
Solubility coefficients are required because all Laboratory gasmeasurement systems measure gas-in-gas and these concentrationsare different than the gas-in-oil
Note: Temperature of the Oil during a manual sample is required only for Lab
Moisture-in-Oil calculations
Temperature of the Oil during the gas extraction process is required todetermine the Solubility Coefficient
-
8/10/2019 Practicing DGA 09072010
34/172
A BPL GLOBAL COMPANY34
Serveron Corporation 2007. No use without permission.34
Gas Solubility Coefficients
The triangles and squares in the graph represent ASTM solubilitycoefficients at selected temperatures used in the Laboratory DGA process
-
8/10/2019 Practicing DGA 09072010
35/172
A BPL GLOBAL COMPANY35
Serveron Corporation 2007. No use without permission.35
Gas Solubility Coefficients
Solubility coefficients are applied to the laboratory measured gas-in-gasvalues to derive the true gas-in-oil concentration in transformer oil.
Example
Method C headspace sampling formula (from ASTM D3612)
CL= CG( K + VG/VL) CL= Gas-in-Oil value, ppm C
G= Gas-in-Gas value, ppm
K = Solubility Coefficient VG = Volume of Oil space, 7 mL* VL = Volume of Gas space, 15 mL*
(*)Typical Laboratory values
Example for Hydrogen, H2
CG= 40 ppm K = 0.0558 VG/VL = 7/15 = 0.467mL CL= 40 (0.0558 + 0.467) CL= 40 (0.5225)
CL= 20.9 ppm is the Gas-in-Oil value
-
8/10/2019 Practicing DGA 09072010
36/172
A BPL GLOBAL COMPANY36
Serveron Corporation 2007. No use without permission.36
Manual DGA Process
Gas Chromatograph
Transformer GasTransformer Oil
Extraction Device
Vacuum (A)
"Stripper Column (B)
Headspace (C)
Oil Sampled From Syringe orBottle
Transformer
Oil Sampled into
Syringe or Bottle
Transportation
Manual Injection
Carrier Gas
Supply(8.2 psig)
A
1
09
8
7
65
4
3
2
1
1.0
mL
1.0mL
Vent
Porapack Column(A)
Molecular SievesColumn (B)
ThermalConductivity
Detector
Channel B
Channel A
Each time a sample
is handled, there is achance that error
may be introducedinto the
measurement
Gas Chromatograph
-
8/10/2019 Practicing DGA 09072010
37/172
A BPL GLOBAL COMPANY37
Serveron Corporation 2007. No use without permission.37
Concerns with Laboratory DGA
Sampling:
If not done according to standard procedures by experienced personnel,the samples may lead to erroneous results.
Transport:
If procedures are not followed properly, transportation may introduce
contaminations.
Exposure to external factors may also affect the oil sample:
If low-quality syringes are used, additional O2 & CO2 may be introduced
or key gases (e.g. H2) may be lost.
-
8/10/2019 Practicing DGA 09072010
38/172
A BPL GLOBAL COMPANY38
Serveron Corporation 2007. No use without permission.38
Good sample versus bad sample
It is sometimes very clear to the laboratory performing theanalysis if it was taken improperly.
The presence of free water or foreign objects such as
insects, pipe sealant, tape or putty are strong
indicators that the drain valve was not adequatelyflushed out prior to sampling.
Materials used for collecting samples;
Galvanic fittings (zinc coated) used in the drain valveassembly such as the drain plug, can create a galvanic
reaction with water, and cause very high levels of
hydrogen to be produced.
G d l b d l
-
8/10/2019 Practicing DGA 09072010
39/172
A BPL GLOBAL COMPANY39
Serveron Corporation 2007. No use without permission.39
Good sample versus bad sample
Sample ports should have Brass Fittings or Stainless
Steel
Tubing - Tygon tubing or other compatible tubing is necessary for
sample collecting
Tubing should only be used once and then discarded as the wallsof the tubing have a memory (can hold gases, water and other
chemicals compounds in the wall of the tubing)
Incompatible tubing such as natural rubber or PVC will
contaminate a sample with unwanted materials.
-
8/10/2019 Practicing DGA 09072010
40/172
A BPL GLOBAL COMPANY40
Serveron Corporation 2007. No use without permission.40
Laboratory handling is a concernthe sample may be contaminated
in the process of being analyzed
Gas Chromatography analysis in
the laboratory is another source
of errors if not done by
experienced personnel and not
following QA/QC procedures.
The accuracy required for laboratory analysis is +/- 15 % on gas
concentrations, but several laboratories are much less accurate than
that.
Concerns with Laboratory DGA
Diff A L b DGA R lt
-
8/10/2019 Practicing DGA 09072010
41/172
A BPL GLOBAL COMPANY41
Serveron Corporation 2008. No use without permission.41
Differences Among Lab DGA Results
Independent studies have shown lab-to-lab DGA differences can belarge, with some results necessarily not accurate
This study consisted of Sixteen (16) samples from the sameTransformer - All samples were collected by the same individual usingthe same process.
All samples were sent to 3 different
laboratorys simultaneously.
Possible causes of differences are; Even with the same sampler
and sample process, contaminationcould have occurred
Transportation may have affectedthe samples
Laboratory accuracies are +15%on average; some are much worse
Methane
0
500
1000
1500
2000
2500
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Lab A Lab B Lab C
Hydrogen
0
500
1000
1500
2000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Lab A Lab B Lab C
-
8/10/2019 Practicing DGA 09072010
42/172
A BPL GLOBAL COMPANY42
Serveron Corporation 2007. No use without permission.42
Questions
What is the current DGA process at your utility? How are samples taken?
What is the frequency of samples?
What method (A, B or C) is used in your lab?
Do you have your own lab or rely on third party labs?
How often is the Lab Equipment Calibrated?
Do you send a known sample to the Lab?
-
8/10/2019 Practicing DGA 09072010
43/172
A BPL GLOBAL COMPANY43
Serveron Corporation 2007. No use without permission.43
-
8/10/2019 Practicing DGA 09072010
44/172
A BPL GLOBAL COMPANY44
Serveron Corporation 2007. No use without permission.44
Standards and Guidelines Governing
-
8/10/2019 Practicing DGA 09072010
45/172
A BPL GLOBAL COMPANY45
Serveron Corporation 2007. No use without permission.45
Standards and Guidelines Governingthe Interpretation of DGA
IEEE Std. C57.104.1991 IEEE Guide for the Interpretationof Gases Generated in Oil Immersed Transformers
IEEE PC57.104 Draft 11d IEEE Guidenot approved
IEC 60599-1999 Mineral Oil Impregnated Electrical
Equipment in Service: Guide to the Interpretation of
Dissolved and Free Gas Analysis.
IEC 60599-1999, Amendment 1, 04/2007
-
8/10/2019 Practicing DGA 09072010
46/172
A BPL GLOBAL COMPANY46
Serveron Corporation 2007. No use without permission.46
-
8/10/2019 Practicing DGA 09072010
47/172
A BPL GLOBAL COMPANY47
Serveron Corporation 2007. No use without permission.47
Gas Sources
Gases in oil always result from the decomposition ofelectrical insulation materials (oil or paper), as a result of
faults or chemical reactions in the equipment
For example: Oil is a molecule of hydrocarbons, i.e., containing hydrogen and
carbon atoms, linked by chemical bonds (C-H, C-C)
-
8/10/2019 Practicing DGA 09072010
48/172
A BPL GLOBAL COMPANY48
Serveron Corporation 2007. No use without permission.48
Gas Formation
Some of these bonds may break and form H*,
CH3*, CH2* and CH* radicals.
-
8/10/2019 Practicing DGA 09072010
49/172
A BPL GLOBAL COMPANY49
Serveron Corporation 2007. No use without permission.49
Gas Formation
All these radicals then recombine to form the fault gasesobserved in oil:
Gas Formation
-
8/10/2019 Practicing DGA 09072010
50/172
A BPL GLOBAL COMPANY50
Serveron Corporation 2007. No use without permission.50
Gas Formation
In addition to these gases, the decomposition of paper
produces CO2, CO and H2O, because of the presence of
oxygen atoms in the molecule of cellulose:
-
8/10/2019 Practicing DGA 09072010
51/172
A BPL GLOBAL COMPANY51
Serveron Corporation 2007. No use without permission.51
Hydrogen H2
Methane CH4
Ethane C2H6
Ethylene C2H4Acetylene C2H2
Carbon monoxide CO
Carbon dioxide CO2
Oxygen O2
Nitrogen N2
The Main Gases Analyzed by DGA
-
8/10/2019 Practicing DGA 09072010
52/172
A BPL GLOBAL COMPANY52
Serveron Corporation 2007. No use without permission.52
Fault Gas Formation
Most of the time, all gases are present in DGA results,
however, some are formed in larger or smaller quantities
depending on the energy content of the fault
Example; Low energy faults such as Corona PartialDischarges in gas bubbles, or low temperature hot spots,
will form mainly Hydrogen, H2and Methane, CH4
-
8/10/2019 Practicing DGA 09072010
53/172
A BPL GLOBAL COMPANY53
Serveron Corporation 2007. No use without permission.53
Fault Gas Formation
Faults of higher temperatures are necessary to form large quantities
of Ethylene, C2H4
Finally, it takes faults with a very high energy content, such as in
electrical arcs, to form large amounts of Acetylene, C2H2
By looking at the relative proportion of gases in the DGA results it is
possible to identify the type of fault occurring in a transformer in
service
-
8/10/2019 Practicing DGA 09072010
54/172
A BPL GLOBAL COMPANY54
Serveron Corporation 2007. No use without permission.54
Gas Formation Patterns
Are related only to the materials used and faults involved
Are the same in all equipment where these materials are
used Sealed or air-breathing power transformers
Reactors Instrument transformers
LTCs
Etc.
-
8/10/2019 Practicing DGA 09072010
55/172
A BPL GLOBAL COMPANY55
Serveron Corporation 2007. No use without permission.
-
8/10/2019 Practicing DGA 09072010
56/172
A BPL GLOBAL COMPANY56
Serveron Corporation 2007. No use without permission.
-
8/10/2019 Practicing DGA 09072010
57/172
A BPL GLOBAL COMPANY57
Serveron Corporation 2007. No use without permission.
-
8/10/2019 Practicing DGA 09072010
58/172
A BPL GLOBAL COMPANY58 Serveron Corporation 2007. No use without permission.
-
8/10/2019 Practicing DGA 09072010
59/172
A BPL GLOBAL COMPANY59 Serveron Corporation 2007. No use without permission.
-
8/10/2019 Practicing DGA 09072010
60/172
A BPL GLOBAL COMPANY60 Serveron Corporation 2007. No use without permission.
-
8/10/2019 Practicing DGA 09072010
61/172
A BPL GLOBAL COMPANY61 Serveron Corporation 2007. No use without permission.
-
8/10/2019 Practicing DGA 09072010
62/172
A BPL GLOBAL COMPANY62 Serveron Corporation 2007. No use without permission.
Fault Gas Formations
-
8/10/2019 Practicing DGA 09072010
63/172
A BPL GLOBAL COMPANY63
Serveron Corporation 2007. No use without permission.
63
(Ref. S. D. Myers 1996)
-
8/10/2019 Practicing DGA 09072010
64/172
A BPL GLOBAL COMPANY64
Serveron Corporation 2007. No use without permission.
64
-
8/10/2019 Practicing DGA 09072010
65/172
A BPL GLOBAL COMPANY65
Serveron Corporation 2007. No use without permission.
65
-
8/10/2019 Practicing DGA 09072010
66/172
A BPL GLOBAL COMPANY66
Serveron Corporation 2007. No use without permission.
66
Fault Types
Partial discharges of the corona-type (PD)
Typical examples: Discharges in gas bubbles or voids trapped in paper A result of poor drying or poor oil-impregnation
Discharges of low energy (D1)
Typical examples: Partial discharges of the sparking-type Inducing carbonized punctures in paper Low-energy arcing, inducing surface tracking of paper and carbon
particles in oil
Ref. IEC 60599-1999
-
8/10/2019 Practicing DGA 09072010
67/172
A BPL GLOBAL COMPANY67
Serveron Corporation 2007. No use without permission.
67
Fault Types
Discharges of high energy (D2)
Typical Examples High Energy Arcing Flashovers Short Circuit with power follow through
These result in; Extensive damage to paper Large formation of carbon particles in oil Metal Fusion Tripping of the equipment or gas alarms
Ref. IEC 60599-1999
F l T
-
8/10/2019 Practicing DGA 09072010
68/172
A BPL GLOBAL COMPANY68
Serveron Corporation 2007. No use without permission.
68
Fault Types
Thermal faults of temperatures 200C) Black or carbonized (>300C)
Ref. IEC 60599-1999
F lt T
-
8/10/2019 Practicing DGA 09072010
69/172
A BPL GLOBAL COMPANY69
Serveron Corporation 2007. No use without permission.
69
Fault Types
Thermal faults of temperatures between 300 and700C (T2)
Typical Examples: Defective contacts
Defective welds Circulating currents
Evidenced by: Carbonization of paper
Formation of carbon particles
Ref. IEC 60599-1999
F lt T
-
8/10/2019 Practicing DGA 09072010
70/172
A BPL GLOBAL COMPANY70
Serveron Corporation 2007. No use without permission.
70
Fault Types
Thermal faults of temperatures >700C (T3)
Typical Examples: Large circulating currents in tank and core Short circuits in laminations
Evidenced by: extensive formation of carbon particles in oil metal coloration (800C) or metal fusion (>1000C)
Ref. IEC 60599-1999
-
8/10/2019 Practicing DGA 09072010
71/172
A BPL GLOBAL COMPANY71
Serveron Corporation 2007. No use without permission.
71
Diagnostic Tools for DGA
-
8/10/2019 Practicing DGA 09072010
72/172
A BPL GLOBAL COMPANY72
Serveron Corporation 2007. No use without permission.
72
Diagnostic Tools for DGA
Tool
Reference Standard
IEEE C57.
104-1991
IEEE PC57.
104 D11d
IEC 60599-
1999
Dornenburg Ratios
TDCG Procedure
Key Gas Procedure
TCG Procedure
Rogers Ratios
IEC Gas Ratios
Duval Triangle
CO2/CO Ratio
O2/N2 Ratio
C2H2/H2 Ratio
Diagnostic Tools for DGA
-
8/10/2019 Practicing DGA 09072010
73/172
A BPL GLOBAL COMPANY73
Serveron Corporation 2007. No use without permission.
73
Diagnostic Tools for DGAProcedures vs. Methods
Procedures are guidelines on what steps are necessary to
continue to examine the potential for a fault or problem
condition.
Methods are results that provide condition assessments
of potential faults directly
K G P d
-
8/10/2019 Practicing DGA 09072010
74/172
A BPL GLOBAL COMPANY74
Serveron Corporation 2007. No use without permission.
74
Key Gas Procedure
Complements the TDCG procedure in order to identify
faults
Focuses on the main (or key) gas formed
Limitations:
High tendency to return inconclusive results
Key Gas Procedure
-
8/10/2019 Practicing DGA 09072010
75/172
A BPL GLOBAL COMPANY
y
75
Serveron Corporation 2007. No use without permission.
Key Gas Examples
-
8/10/2019 Practicing DGA 09072010
76/172
A BPL GLOBAL COMPANY76
Serveron Corporation 2007. No use without permission.
76
0
20
40
60
80
100
CO H2 CH4 C2H6 C2H4 C2H2
Thermal Oil and Cellulose Fault
0
20
40
60
80
100
CO H2 CH4 C2H6 C2H4 C2H2
Thermal Oil Fault
0
20
40
60
80
100
CO H2 CH4 C2H6 C2H4 C2H2
High Energy Arcing
0
20
40
60
80
100
CO H2 CH4 C2H6 C2H4 C2H2
Low Energy Partial Discharge
Rogers Ratio Method
-
8/10/2019 Practicing DGA 09072010
77/172
A BPL GLOBAL COMPANY77
Serveron Corporation 2007. No use without permission.
77
Rogers Ratio Method
IEC Gas Ratio Method
-
8/10/2019 Practicing DGA 09072010
78/172
A BPL GLOBAL COMPANY78
Serveron Corporation 2007. No use without permission.
78
IEC Gas Ratio Method
Summary of the Ratio Methods
-
8/10/2019 Practicing DGA 09072010
79/172
A BPL GLOBAL COMPANY79
Serveron Corporation 2007. No use without permission.
79
Summary of the Ratio Methods
The Basic, Rogers and Dornenburg methods use thesame 3 basic gas ratios: (CH4/H2, C2H2/C2H4and
C2H6/C2H4).
Depending on the values of these gas ratios, codes or
zones are defined for each type of fault
One drawback of these ratio methods is that no diagnosis
can be given in a significant number of cases, Dead
Zones - fall outside defined zones.
3-D IEC Gas & Rogers Ratio Methods
-
8/10/2019 Practicing DGA 09072010
80/172
A BPL GLOBAL COMPANY80
Serveron Corporation 2007. No use without permission.
80
g
-
8/10/2019 Practicing DGA 09072010
81/172
A BPL GLOBAL COMPANY81
Serveron Corporation 2007. No use without permission.
81
Triangle Method
-
8/10/2019 Practicing DGA 09072010
82/172
A BPL GLOBAL COMPANY82
Serveron Corporation 2007. No use without permission.
82
Triangle Method
The Triangle, developed empirically in the early 1970s,
and is used by the IEC.
Based upon the 3 gases (Methane, CH4, Ethylene, C2H4
and Acetylene, C2H2) corresponding to the increasing
energy levels of gas formation. One advantage of this method is that it always provides a
diagnosis, with a low percentage of wrong diagnoses.
There are no indeterminate diagnostics using the Triangle
method.
Triangle Method
-
8/10/2019 Practicing DGA 09072010
83/172
A BPL GLOBAL COMPANY83
Serveron Corporation 2007. No use without permission.
83
Triangle Method
The triangle method plots the relative % of the 3 gases on
each side of the triangle, from 0% to 100%.
The 6 main zones of faults are indicated in the triangle,
plus a DT zone (mixture of thermal and electrical faults)
Approximately 200+ inspected cases in service were used
to develop the Triangle
Triangle Method
-
8/10/2019 Practicing DGA 09072010
84/172
A BPL GLOBAL COMPANY84
Serveron Corporation 2007. No use without permission.
84
Triangle Method
-
8/10/2019 Practicing DGA 09072010
85/172
A BPL GLOBAL COMPANY85
Serveron Corporation 2007. No use without permission.
85
Triangle Method
How have fault zones been defined in the Triangle?
Answer:
They are based on a large number of cases of faulty
transformers in service which have been inspected visually.
The root cause of the failure was determined and matched to
the DGA data.
The Triangle was tested with all these cases and correctly
identifies the zone that matches the root cause of failure at a
very high percentage.
Triangle Method
-
8/10/2019 Practicing DGA 09072010
86/172
A BPL GLOBAL COMPANY86
Serveron Corporation 2007. No use without permission.
86
Cases of faults PD and D1
Tracking; Sparking; Small Arcing
Triangle Method
-
8/10/2019 Practicing DGA 09072010
87/172
A BPL GLOBAL COMPANY87
Serveron Corporation 2007. No use without permission.
87
Cases of faults D2
Triangle Method
-
8/10/2019 Practicing DGA 09072010
88/172
A BPL GLOBAL COMPANY88
Serveron Corporation 2007. No use without permission.
88
Circulating Currents; Laminations; Bad Contacts
Cases of thermal faults in oil only
Triangle Method
-
8/10/2019 Practicing DGA 09072010
89/172
A BPL GLOBAL COMPANY89
Serveron Corporation 2007. No use without permission.
89
Brownish Paper; Carbonized Paper; Not Mentioned
Cases of thermal faults in paper
Triangle Method FAQs
-
8/10/2019 Practicing DGA 09072010
90/172
A BPL GLOBAL COMPANY90
Serveron Corporation 2007. No use without permission.
90
Triangle Method FAQ s
How are corona PDs, which form a lot of Hydrogen, H2, identified in
the Triangle without using this gas?
Answer: In such faults, Methane, CH4 is indeed formed in
smaller amounts than Hydrogen, H2 (typically 10 to 20 times
less), but which can still be measured easily by DGA
0
20
40
60
80
100
CO H2 CH4 C2H6 C2H4 C2H2
Low Energy Partial Discharge
Triangle Method FAQs
-
8/10/2019 Practicing DGA 09072010
91/172
A BPL GLOBAL COMPANY91
Serveron Corporation 2007. No use without permission.
91
Triangle Method FAQ s
In the Triangle method, why not use Hydrogen, H2 rather
than Methane, CH4 to represent low energy faults? Answer: Because CH4 provides better overall diagnoses for all types of
faults
A possible explanation
Is that H2 diffuses muchmore rapidly than
hydrocarbon gases from
transformer oil.
This will affect gas ratiosusing H2 but not those
using hydrocarbon gases.
-
8/10/2019 Practicing DGA 09072010
92/172
Triangle Method Example
-
8/10/2019 Practicing DGA 09072010
93/172
A BPL GLOBAL COMPANY93
Serveron Corporation 2007. No use without permission.
93
Triangle Method Example
Using the Triangle Method
-
8/10/2019 Practicing DGA 09072010
94/172
A BPL GLOBAL COMPANY94
Serveron Corporation 2007. No use without permission.
94
Using the Triangle Method
The calculation of triangular coordinates can easily be done manually,
or with the help of a small algorithm or software
Errors are often made when developing such an algorithm, so check it
first with the free algorithm available. ([email protected])
For those familiar with computer graphics, it is also possible to
develop a software displaying the point and the fault zones
graphically in the triangle
Software from vendors is available for that purpose
An example of the software follows, Courtesy of Serveron
Using the Triangle Method
mailto:[email protected]:[email protected] -
8/10/2019 Practicing DGA 09072010
95/172
A BPL GLOBAL COMPANY95
Serveron Corporation 2007. No use without permission.
95
Using the Triangle Method
Note: This is the same data as shown using Rogers Ratio example
Using the Triangle Method
-
8/10/2019 Practicing DGA 09072010
96/172
A BPL GLOBAL COMPANY96
Serveron Corporation 2007. No use without permission.
96
Using the Triangle Method
The Triangle, being a graphical method, allows you to
easily follow the evolution of faults with time
The previous example shows evolution from a thermal
fault to a potentially much more severe fault such as D1
- Fault Severity -
-
8/10/2019 Practicing DGA 09072010
97/172
A BPL GLOBAL COMPANY97
Serveron Corporation 2007. No use without permission.
97
Fault Severity
The most severe faults: Faults D2 in paper and in oil (high-energy arcing) Faults T2-T3 in paper (>300C) faults D1 in paper (tracking, arcing) faults T3 in oil (>700C)
The less severe faults: Faults PD/ D1 in oil (sparking) Faults T1 in paper (
-
8/10/2019 Practicing DGA 09072010
98/172
A BPL GLOBAL COMPANY98
Serveron Corporation 2007. No use without permission.
98
Fault Method Comparisons
% CorrectDiagnoses
% UnresolvedDiagnoses
% WrongDiagnoses
IEEE Key Gas
Method 42 0 58
IEEE RogersRatios 62 33 5
Doernenburg
Ratios 71 26 3
IEC Basic GasRatios 77 15 8
IEC Duval
Triangle 96 0 4
% CorrectDiagnoses
% UnresolvedDiagnoses
% WrongDiagnoses
IEEE Key Gas
Method 42 0 58
RogersRatio 62 33 5
Doernenburg
Ratios 71 26 3
IEC GasRatio 77 15 8
IEC Duval
Triangle 96 0 4
CO2/CO Ratio
-
8/10/2019 Practicing DGA 09072010
99/172
A BPL GLOBAL COMPANY99
Serveron Corporation 2007. No use without permission.
99
CO2/CO Ratio CO2/CO ratio used to detect paper related faults
If 200C
If >10, it indicates a fault of temperature T
-
8/10/2019 Practicing DGA 09072010
100/172
A BPL GLOBAL COMPANY100
Serveron Corporation 2007. No use without permission.
100
O2/N2: a decrease of this ratio indicatesexcessive heating
Acetylene/ Hydrogen, C2H
2/H
2; A ratio >3 in the main tank indicates contamination by the LTC
compartment
Gassing not related to faults in service
-
8/10/2019 Practicing DGA 09072010
101/172
A BPL GLOBAL COMPANY101
Serveron Corporation 2007. No use without permission.
101
g
Catalytic reactions on metal surfaces: Formation of H2only
Stray gassing of oil:
The unexpected gassing of oil at relatively lowtemperatures (80 to 200C):
Gassing of the T1 or T2 type
-
8/10/2019 Practicing DGA 09072010
102/172
A BPL GLOBAL COMPANY102
Serveron Corporation 2007. No use without permission.
102
Any questions about what was just presented?
-
8/10/2019 Practicing DGA 09072010
103/172
A BPL GLOBAL COMPANY103
Serveron Corporation 2007. No use without permission.
103
-
8/10/2019 Practicing DGA 09072010
104/172
A BPL GLOBAL COMPANY104
Serveron Corporation 2007. No use without permission.
104
Accuracy and Gas Levels
-
8/10/2019 Practicing DGA 09072010
105/172
A BPL GLOBAL COMPANY105
Serveron Corporation 2007. No use without permission.
105
y
IEC & CIGRE recommendation
DGA diagnosis should be attempted only if gas concentrations or
gassing rates are high enough to be considered significant
Low gas levels may be due to contamination or aging of
insulation, not necessarily to an actual fault.
Limitations to Accurate Diagnostics
First limit - LabAccuracy & Gas Concentrations
Second limit - Typical Values & Diagnosis
Accuracy and Gas Levels
-
8/10/2019 Practicing DGA 09072010
106/172
A BPL GLOBAL COMPANY106
Serveron Corporation 2007. No use without permission.
106
First limit: LAB ACCURACY & Gas Concentrations
Accuracy is defined here as the deviation from a prepared known
value
The accuracy of the average CIGRE /IEC lab (including in theUS) is ~15% at medium (routine) gas concentrations (> 10ppm for hydrocarbons)
Lab Accuracy & Gas Concentrations
-
8/10/2019 Practicing DGA 09072010
107/172
A BPL GLOBAL COMPANY107
Serveron Corporation 2007. No use without permission.
107
(medium gas concentrations of >10 ppm)
Lab Accuracy & Gas Concentrations(low gas concentrations of
-
8/10/2019 Practicing DGA 09072010
108/172
A BPL GLOBAL COMPANY108
Serveron Corporation 2007. No use without permission.
108
The accuracy of the average laboratory decreases to ~ 30% at 6ppm, and 100% near the lab detection limit (2 ppm)
Lab Accuracy & Diagnostics
-
8/10/2019 Practicing DGA 09072010
109/172
A BPL GLOBAL COMPANY109
Serveron Corporation 2007. No use without permission.
109
Effect of lab accuracies of 15% (in RED) and 30% (in BLUE) on DGAdiagnosis uncertainty
When an area of uncertainty crosses several fault zones, a reliable diagnosiscannot be given
This is particularly true for lab accuracies > 30%
Lab Accuracy & Diagnostics
-
8/10/2019 Practicing DGA 09072010
110/172
A BPL GLOBAL COMPANY110
Serveron Corporation 2007. No use without permission.
110
Diagnosis uncertainties corresponding to lab inaccuracies of15, 30, 50
and 75 % (at 10, 6, 4 and 3 ppm for the average lab):
This applies not only to the triangle but to all diagnosis methods
Lab Accuracy & Diagnostics
-
8/10/2019 Practicing DGA 09072010
111/172
A BPL GLOBAL COMPANY111
Serveron Corporation 2007. No use without permission.
111
The actual accuracy of your laboratory can be obtained byusing gas-in-oil standards, which are available
commercially for instance from Morgan Schaffer
(Truenorth)
Below 15% of lab accuracy, a calculation of diagnosis
uncertainty should be done, and commercial software are
available for that purpose, for instance from Delta-X
Research (TOA4)
Typical Values and Diagnosis
-
8/10/2019 Practicing DGA 09072010
112/172
A BPL GLOBAL COMPANY112
Serveron Corporation 2007. No use without permission.
112
Second limit: TYPICAL VALUES
Typical Values are not gas Limit Values
Typical Values are not the same for all Transformer Systems (may
change slightly from system to another).
Typical Values correspond to a given percentile of the populationof DGA results
Typical Values should be calculated for a specific Transformer
System (Utility specific)
Typical Values observed nationwide or worldwide may be used,
but with some caution.
Typical Values and Diagnosis
-
8/10/2019 Practicing DGA 09072010
113/172
A BPL GLOBAL COMPANY113
Serveron Corporation 2007. No use without permission.
113
Second limit: TYPICAL VALUES
Diagnosis only attempted when gas concentrations levels and
rates of gas increase is above Typical Values.
When gas levels are below Typical Values, gas concentrations
may be too close to detection limits.
Below Typical Values, the IEC recommends to use normal
sampling frequency (monthly, semi-annual, etc.,..) and not to
attempt a diagnosis.
Above Typical Values, the IEC recommends to use increasedsampling frequency (e.g., monthly or weekly) and a DGA
diagnosis may be attempted.
Typical Values and Diagnosis- 90% Typical Values -
-
8/10/2019 Practicing DGA 09072010
114/172
A BPL GLOBAL COMPANY114
Serveron Corporation 2007. No use without permission.
114
90% Typical Values
90% Typical Values are represented;
Typical Gas Concentration (TGC) levels
Typical Rate of Gas Increase (TRGI) levels.
TGC and TRGI values are the calculated values from a specific
population of transformers in service.
Reference - CIGRE Brochure # 296, Recent developments in DGA
Interpretation, Joint Task Force D1.01/A2.11, June 2006 (available
from [email protected])
Typical Values and Diagnosis- 90% TGC Values -
-
8/10/2019 Practicing DGA 09072010
115/172
A BPL GLOBAL COMPANY115
Serveron Corporation 2007. No use without permission.
115
C2H2 H2 CH4 C2H4 C2H6 CO CO2
All transformers 50-
150
30-
130
60-
280
20-
90
400-
600
3800-
14000
No OLTC 2-20
Communicating
OLTC
60-280
90% TGC Values -
Typical Gas Concentration (TGC) levels observed by CIGRE inpower transformers, in ppm:
NoOLTCrefers to a transformer not equipped withan OLTC or has a LTC not communicatingwith or
leaking to the Main Tank
Reference; Table 3, pg 5; CIGRE Brochure #296
The data shown here represents 25 Electrical Networks worldwide,including more than 20,000 Transformers
CommunicatingOLTCmeans that some oil and/or gas communication
is possible between the OLTC compartment and the Main Tank of a
Transformer.
Typical Values and Diagnosis- 90% TGC Values -
-
8/10/2019 Practicing DGA 09072010
116/172
A BPL GLOBAL COMPANY116
Serveron Corporation 2007. No use without permission.
116
C2H2 H2 CH4 C2H4 C2H6 CO CO2
All transformers 35-
132
10-
120
32-
146
5-
90
260-
1060
1700-
10,000
No OLTC 0-4
Communicating
OLTC21 - 37
90% TGC Values
Typical Rate of Gas Increase(TRGI) levels observed by CIGRE inpower transformers, in ppm/yr
NoOLTCrefers to a transformer not equipped withan OLTC or has a LTC not communicatingwith or
leaking to the Main Tank
Reference; Table 3, pg 5; CIGRE Brochure #296
The data shown here represents 25 Electrical Networks worldwide, includingmore than 20,000 Transformers
CommunicatingOLTCmeans that some oil and/or gas communication
is possible between the OLTC compartment and the Main Tank of a
Transformer.
Typical Values and Diagnosis- 90% TGC & TRGI Values -
-
8/10/2019 Practicing DGA 09072010
117/172
A BPL GLOBAL COMPANY117
Serveron Corporation 2007. No use without permission.
117
90% TGC & TRGI Values
The large ranges of TGC & TRGI values are an indication of the
differences in operation practices of individual Utilities.
Values in the preceding tables are from Core-Type XfmrShell type
values are likely to be higher.
Each individual Utility should preferably calculate its own specificTGC & TRGI values.
To calculate the TGC & TRGI values, separate out the different types
or sizes of transformers to be evaluated into specific groups or
classifications.
Each transformer group and classification will have its own TGC &
TRGI ranking or 90% values
Typical Values and Diagnosis- 90% Typical Values -
-
8/10/2019 Practicing DGA 09072010
118/172
A BPL GLOBAL COMPANY118
Serveron Corporation 2007. No use without permission.
118
90% typical concentration value
1 10 100 1000 ppm
% 100
90
50
Cumulative curves of DGA results (concentrations orgassing rates)
Typical Values and Diagnosis- 90% Typical Values -
-
8/10/2019 Practicing DGA 09072010
119/172
A BPL GLOBAL COMPANY119
Serveron Corporation 2007. No use without permission.
119
yp
Calculate TGC Concentration Values - Example
Hydrogen, H2 Methane, CH4 Ethylene, C2H4 Ethane, C2H6 CarbonMonoxide, CO
CarbonDioxide, CO2
Acetylene
0 0 0 0 0 177.3 0
0 0 0 0 0 258.6 0
0 0 0 0 21.5 487.4 0
1.1 0 0 0 26.3 494.9 0
1.7 0 0 0 29.1 662.3 0
2.3 0 0 0 52.9 1206 0
7.4 0 0 0 56 1263.6 0
7.6 1.5 0 0 121.5 1289.2 0
16.3 6.2 0 1.8 135.1 1748.4 0
18.8 9.4 2.9 2.5 143 2812.4 0
22.8 10.2 3.3 15.9 189.4 3369.8 0
23.6 19.9 5.4 39.5 222.1 4604.5 0
24 30.5 7 49 239.4 4639 0
30.6 49.5 9.9 76.2 282.3 6820.6 0
51.0 52.6 11.7 76.4 296.8 7244.1 0
53.8 63 18.4 76.7 306.9 8381.3 0
90% TGC 87.4 76.4 22.8 118.3 342.1 8426.1 2.1136.5 186.8 55.8 541.2 582.3 9457.3 3.2
138.6 1867.8 5930.9 1078.2 1760.1 19661.7 5.7
90% TGC
(CIGRE)
50 - 150 30 - 130 60 - 280 20 - 90 400 - 600 3800 - 14000 2 - 20
Typical Values and Diagnosis- 90% Typical Values -
-
8/10/2019 Practicing DGA 09072010
120/172
A BPL GLOBAL COMPANY120
Serveron Corporation 2007. No use without permission.
120
Calculate TRGI Gassing Rates Values - Example
Hydrogen, H2 Carbon Monoxide, CO Carbon Dioxide, CO2 Acetylene, C2H21st 2nd 1st - 2nd 1st 2nd 1st - 2nd 1st 2nd 1st - 2nd 1st 2nd 1st - 2nd
0 0 0 24.8 26 1.2 0 0 0 0 0 0
22.3 22.6 0.3 47.5 51.5 4 0 0 0 0 0 0
6.7 5.8 0.9 262.2 248.9 13.3 0 0 0 0 0 0
81.7 82.8 1.1 4358.5 4319 39.5 133.9 135.1 1.2 0 0 0
30.1 28.5 1.6 123.7 168 44.3 30.3 32 1.7 0 0 0
28.1 30.1 2 2476.5 2563.5 87 22.8 20.9 1.9 0 0 0
0 2.1 2.1 134.1 223.8 89.7 145.5 142.9 2.6 0 0 0
0 2.4 2.4 138.1 229.3 91.2 291.8 294.9 3.1 0 0 0
3.7 0 3.7 31.1 161.3 130.2 303.4 328 24.6 0 0 0
0 4.8 4.8 2.9 177.6 174.7 1578.9 1615 36.1 0 0 0
11.5 16.3 4.8 316 494.8 178.8 517 573.8 56.8 0 0 0
24 32.4 8.4 9751.9 10095 343.1 304 367 63 2.2 2.3 0.1
15.4 2.8 12.6 607 1011 404 1160.2 1254.7 94.5 2.4 3.1 0.7
4.5 23.2 18.7 872.7 1281.4 408.7 152.8 298.8 146 7.3 6.6 0.7
28.6 50.2 21.6 17666.6 17120.6 546 6379.4 6627.5 248.1 1 2.2 1.2
144.3 172 27.7 4382 5069 687 326.2 668.2 342 1.5 0 1.5
48.3 11.7 37 6338.5 7110.4 772 489.2 1524.9 10367 2.2 0 2.2
140.2 91.8 48.4 6202.9 7296.9 1094 2324.6 5435.9 3111.3 6.4 1.4 5
186.2 134.1 52.1 6642.1 8622.1 1980 34 3426 3392 1.8 10.4 8.6
Typical Values and Diagnosis- 90% Typical Values -
-
8/10/2019 Practicing DGA 09072010
121/172
A BPL GLOBAL COMPANY121
Serveron Corporation 2007. No use without permission.
121
yp
Influences on Typical Values:
Transformer Age Values are significantly higher in young equipment (suggesting some
unstable chemical bonds in new oil and paper)
Values are a bit higher in very old equipment
Transformer Type Values are higher in shell-type and shunt reactors (operating at
higher temperatures), lower in instrument transformers
Typical values are very similar in air-breathing and in sealed or
nitrogen blanketed equipment, contrary to a common belief in the
US
Transformer Oil Volume Values are not affected by oil volume (suggesting that larger faults
are formed in larger transformers)
90% Typical Concentration valuesobserved in the US
-
8/10/2019 Practicing DGA 09072010
122/172
A BPL GLOBAL COMPANY
122
H2 CH4 C2H4 C2H6 C2H2 CO CO2 TDCG
Condition 1 IEEE-2008 100 120 50 65 2 350 2500 720
Weidmann 96 88 57 79 3 613 5990 1058
APS 80 45 70 30 2 950 1250
GE 80 50 73 28 2 950 1183
IEC/CIGRE 100 80 170 55 3 500 8900 908
Typical Values and Diagnosis- 90% Typical Values -
-
8/10/2019 Practicing DGA 09072010
123/172
A BPL GLOBAL COMPANY
123
Serveron Corporation 2007. No use without permission.
123
yp
When DGA results in service reach typical values:
A diagnosis may be attempted to identify the fault (if
lab accuracy is good enough)
The equipment should not be considered at risk
However, it should be monitored more frequently by
DGA
-
8/10/2019 Practicing DGA 09072010
124/172
A BPL GLOBAL COMPANY
124
Serveron Corporation 2007. No use without permission.
124
What are your current typical (normal) values for the
different gases?
-
8/10/2019 Practicing DGA 09072010
125/172
A BPL GLOBAL COMPANY
125
Serveron Corporation 2007. No use without permission.
125
Gas Levels above Typical ValuesRisk of Failure in Service
-
8/10/2019 Practicing DGA 09072010
126/172
A BPL GLOBAL COMPANY
126
Serveron Corporation 2007. No use without permission.
126
Probability of Failure in Service (PFS); Probability of having a failure related event in service, such as any
tripping, fault gas alarm, fire, etc. event
PFS has been defined as;
PFS = Number of DGA analyses followed by a failure
related event/Total number of Analyses@ each gas
concentration value
Gas Levels above Typical ValuesRisk of Failure in Service
-
8/10/2019 Practicing DGA 09072010
127/172
A BPL GLOBAL COMPANY
127
Serveron Corporation 2007. No use without permission.
127
90 98 99 Norm, in %
ppm
PFS, in %
100 300 400
Probability of having a failure-related event (PFS, %) vs. the
concentration of C2H2 in ppm at HQ
Pre -FailureValue
200
Gas Levels above Typical ValuesPre-Failure Values
-
8/10/2019 Practicing DGA 09072010
128/172
A BPL GLOBAL COMPANY
128
Serveron Corporation 2007. No use without permission.
128
The PFS remains almost constant below and above the
90% typical value, until it reaches an inflexion point on thecurve (Pre-Failure Value)
DGA monitoring should be done more and more
frequently as gas concentrations increase from typical toPre-Failure Value
Time to reach the Pre-Failure Value is unknown, could be
Hours, Days, Weeks or Months.
The IEC/CIGRE Approach(Gas Concentrations)
-
8/10/2019 Practicing DGA 09072010
129/172
A BPL GLOBAL COMPANY
129
Probability of having a failure-related event in service (PFS)
in %, vs. the concentration of all gases in ppm. T = 90%
typical value; P = pre-failure value
The IEC/CIGRE Approach(rates of Gas Increase)
-
8/10/2019 Practicing DGA 09072010
130/172
A BPL GLOBAL COMPANY
130
Probability of having a failure-related event in service
(PFS) in %, vs. the rate of increase of all gases in ppm/ yr.
Gas Levels above Typical ValuesPre-Failure Values
-
8/10/2019 Practicing DGA 09072010
131/172
A BPL GLOBAL COMPANY
131
Serveron Corporation 2007. No use without permission.
131
Pre-Failure Value concentrations were found by CIGRE to
be surprisingly close on different networks:
This suggests that failure occurs when a critical amount of
insulation is destroyed
H2
CH4
C2H
4C
2H
6C
2H
2CO
240-
1320
270-
460
700-
990
750-
1800
310-
600
984-
3000
(in ppm)
Gas Levels above Typical ValuesPre-Failure Gassing Rates
-
8/10/2019 Practicing DGA 09072010
132/172
A BPL GLOBAL COMPANY
132
Serveron Corporation 2007. No use without permission.
132
Pre-failure rates of gas increase in power transformers at
CIGRE, in ppm/day
C2H2 H2 CH4 C2H4 C2H6 CO CO2
0.5 3 5 5 11 NS NS
Note: This is an unpublished chart
Gas Levels above Typical ValuesCaution and Warning Values
-
8/10/2019 Practicing DGA 09072010
133/172
A BPL GLOBAL COMPANY
133
Serveron Corporation 2007. No use without permission.
133
In-between typical and pre-failure values, specific caution and
warning values can be defined, depending on the tolerance to risk ofthe maintenance personnel, and on the maintenance budget available
For example, higher values may be used when the maintenance
budget is low, and lower values in the case of strategic equipment
CIGRE caution and warning values are significantly different from
those proposed by IEEE and are available upon request(from
Sample Intervals & Concentration Limits
-
8/10/2019 Practicing DGA 09072010
134/172
A BPL GLOBAL COMPANY
134
Sampling intervals and gas concentration limits in ppm
calculated for an average US power transformer
-
8/10/2019 Practicing DGA 09072010
135/172
Numerical values of sampling intervals and
rate of gas increase limits in ppm/ year
-
8/10/2019 Practicing DGA 09072010
136/172
A BPL GLOBAL COMPANY
136
calculated for an average CIGRE/ IEC power transformer
Gassing rate H2 CH4 C2H4 C2H6 C2H2 CO CO2 TDCG Sampling intervals
Nuclear Transmission
Typical 83 65 89 47 2 660 5850 946 6 months Yearly
Level 2 151 141 180 132 6 1408 12472 2049 Monthly
179 175 218 176 7 1737 15382 2539 Monthly
Level 3 247 266 319 308 13 2607 23067 3841 Weekly
280 313 369 382 17 3054 27012 4513 Weekly
Level 4 475 620 686 951 42 5940 52492 8896 Daily
509 679 745 1074 47 6491 57351 9738 Daily
Pre-failure 1095 1825 1825 4015 182 17000 150000 26000 Hourly Hourly
Sampling Intervals based on Combined Gas
Rate and Gas Concentration Levels of
-
8/10/2019 Practicing DGA 09072010
137/172
A BPL GLOBAL COMPANY
137
Individual Gases(IEC/CIGRE approach)
Sampling Intervals based on CombinedGas Rate and Concentration Levels
Rate
Level #
Conc.
Level #
Daily Weekly Monthly Quarterly Yearly
4 4 X
4 3 X
4 2 X
4 1 X
3 4 X
3 3 X
3 2 X
3 1 X
2 4 X
2 3 X
2 2 X
2 1 X
1 4 X
1 3 X
1 2 X
1 1 X
TDCG Procedure
-
8/10/2019 Practicing DGA 09072010
138/172
A BPL GLOBAL COMPANY
138
Serveron Corporation 2007. No use without permission.
138
Focuses on Total Dissolved Combustible Gas (TDCG) content
TDCG, is the sum of all combustible gases, (H2+CH4+C2H2+
C2H4+C2H6+CO) dissolved in the oil
The table provides guidelines concerning gas levels in
transformers and suggests actions in terms of samplingintervals and operating procedures
IEEE Guide for the Interpretation of GasesIEEE Std C57.104-1991
-
8/10/2019 Practicing DGA 09072010
139/172
A BPL GLOBAL COMPANY
139
IEEE Guide for the Interpretation of GasesIEEE Std C57.104-1991
-
8/10/2019 Practicing DGA 09072010
140/172
A BPL GLOBAL COMPANY
140
-
8/10/2019 Practicing DGA 09072010
141/172
A BPL GLOBAL COMPANY
141
Serveron Corporation 2007. No use without permission.
141
What are your current caution/ warning/ alarm values for
the different gases?
Triangle 2 for LTCsof the Conventional Oil-type
(UZB, UBB, URS, UTT, URT, 550, 394, TLH, TC, LR, LRT, etc)
-
8/10/2019 Practicing DGA 09072010
142/172
A BPL GLOBAL COMPANY
142
: Normal operation; : Severe coking; : Light coking;:
Heating; : Strong arcing D2; : Arcing D1
Triangle 2 for LTCsof the ConventionalVacuum Type (LRT, UVT)
-
8/10/2019 Practicing DGA 09072010
143/172
A BPL GLOBAL COMPANY
143
: no inspection made
Triangle 3 for Non-Mineral oils
-
8/10/2019 Practicing DGA 09072010
144/172
A BPL GLOBAL COMPANY
144
FR3 Mineral oil
Triangle 4 for Low-Temperature Faults inTransformers
-
8/10/2019 Practicing DGA 09072010
145/172
A BPL GLOBAL COMPANY145
: Corona partial discharges; : Stray gassing at
120C; ; Stray gassing at 200C; ; : Hot spots with
carbonization of paper; ;: Overheating (T < 250C).
Triangle 5 for Low-Temperature Faults inTransformers
-
8/10/2019 Practicing DGA 09072010
146/172
A BPL GLOBAL COMPANY146
: Corona partial discharges; : Stray gassing at
120C; ; Stray gassing at 200C; ; : Hot spots with
carbonization of paper; ;: Overheating (T < 250C).
Triangle 6 for Low-Temperature Faults in FR3
Oils
-
8/10/2019 Practicing DGA 09072010
147/172
A BPL GLOBAL COMPANY147
stray gassing at T < 150 C;;
stray gassing at T > 200C;;
gas formation in Alliant transformers
-
8/10/2019 Practicing DGA 09072010
148/172
-
8/10/2019 Practicing DGA 09072010
149/172
A BPL GLOBAL COMPANY149
Serveron Corporation 2007. No use without permission.
149
How to proceed with DGA results coming from thelaboratory ?
-
8/10/2019 Practicing DGA 09072010
150/172
A BPL GLOBAL COMPANY150
Serveron Corporation 2007. No use without permission.
150
1. Examine DGA laboratory values for errors or
inconsistencies.
2. Attempt a fault diagnosis when possible
3. Evaluate the severity of the fault
4. Decide on appropriate actions on the equipment
#1 - Examine DGA laboratory values
(
-
8/10/2019 Practicing DGA 09072010
151/172
A BPL GLOBAL COMPANY151
Serveron Corporation 2007. No use without permission.
151
Eliminate zero values (replace them by the detection
limits of the laboratory, e.g., 2 or 1 ppm)
Compare with previous values in ppm on the same
transformer (DGA history)
Check for inconsistencies which might indicate
contamination during sampling or a laboratory analytical
error. Consider suspect values with caution or discard
them
#2 - Attempt a fault diagnosis
-
8/10/2019 Practicing DGA 09072010
152/172
A BPL GLOBAL COMPANY152
Serveron Corporation 2007. No use without permission.
152
Compare with in-house typical or normal values of
concentration and gassing rate, or compare with publishedvalues (e.g., from IEEE and CIGRE)
If measured DGA values are above typical/ normal values, a
fault diagnosis may be attempted
If measured values are below routine concentration values (10ppm for hydrocarbons), a fault diagnosis should be attempted
only after calculating the uncertainty on the diagnosis, basedon the accuracy of the laboratory at these low concentrationlevels
#3a - Evaluate the severity of the fault
-
8/10/2019 Practicing DGA 09072010
153/172
A BPL GLOBAL COMPANY153
Serveron Corporation 2007. No use without permission.
153
The severity of the fault will depend on:
The rate of gas formation
The concentration of gases
How far measured values are from typical values, and how closethey are to pre-failure values
The severity of the fault will also depend on:
The nature of the fault (electrical or thermal)
The location of the fault (paper or oil only)
#3b - Evaluate the severity of the fault
-
8/10/2019 Practicing DGA 09072010
154/172
A BPL GLOBAL COMPANY154
Serveron Corporation 2007. No use without permission.
154
Gas formation not related to faults (e.g., catalytic reactions of oil withmetals or stray gassing of oil) is of no particular concern
The CO2/CO ratio (incremented) can be used to determine paperinvolvement in the fault
The C2H2/ H2ratio can be used to detect contamination of the maintank by an LTC leak
To fully assess the severity of a fault, it may be useful at this stage toseek the advice of experienced DGA experts
-
8/10/2019 Practicing DGA 09072010
155/172
-
8/10/2019 Practicing DGA 09072010
156/172
A BPL GLOBAL COMPANY156
Serveron Corporation 2007. No use without permission.
156
Any questions about what was just presented?
-
8/10/2019 Practicing DGA 09072010
157/172
A BPL GLOBAL COMPANY157
Serveron Corporation 2007. No use without permission.
157
Benefits of On-Line DGA
Detects both gradual and sudden trends in all gases
-
8/10/2019 Practicing DGA 09072010
158/172
A BPL GLOBAL COMPANY158
Serveron Corporation 2007. No use without permission.
158
Detects both gradual and sudden trends in all gases
Correlates gassing events with external events such as transformer
load, oil temperature, LTC changes, etc.
Provides historic trail for delayed analysis of gassing events
May provide diagnosis on-line
Advantages of On-Line DGA
Oil i l d t ti ll i l d
-
8/10/2019 Practicing DGA 09072010
159/172
A BPL GLOBAL COMPANY159
Serveron Corporation 2007. No use without permission.
159
Oil is sampled automatically via closed-
loop GC:
No risk of human intervention
Repeatable sampling technique
No atmospheric exposure
Data is collected up to hourly:
Resulting in faster, more accuratedetermination of trends
All 8 fault gases + moisture are monitored
and correlated with oil temperature and
load
On-Line DGA
-
8/10/2019 Practicing DGA 09072010
160/172
A BPL GLOBAL COMPANY160
Serveron Corporation 2007. No use without permission.
160
Serveron Transformer Monitor
On-line DGA
-
8/10/2019 Practicing DGA 09072010
161/172
A BPL GLOBAL COMPANY161
Serveron Corporation 2007. No use without permission.
161
On-line gas monitors
Are best suited for measuring rates of gas increase (trends)
Will detect faults between regular oil samplings
May now also provide on-line diagnosis
-
8/10/2019 Practicing DGA 09072010
162/172
A BPL GLOBAL COMPANY162
Serveron Corporation 2007. No use without permission.
162
Serveron Story #11
-
8/10/2019 Practicing DGA 09072010
163/172
A BPL GLOBAL COMPANY163
Serveron Corporation 2007. No use without permission.
163
Three-phase 700 MVA 400 kV GSU transformer
Hints: - Increasing Methane (CH4), Ethane (C2H6), Ethylene (C2H4)
- Sudden increase in Acetylene (C2H2),- Transformer refurbished and re-commissioned 10/27/03
Serveron Story #11 Analysis:
-
8/10/2019 Practicing DGA 09072010
164/172
A BPL GLOBAL COMPANY164
Serveron Corporation 2007. No use without permission.
164
1. The Duval Triangle shows problem evolving from a
T3 Thermal Fault to D1 Discharge of low energy
2. Intermittent grounding was provided by the fastening bolt causing a transient potential
rise and subsequent discharges occurring between the corona ring and the main tank
ground point.
3. An on-site repair was performed and the transformer
was returned to service.
The Duval Triangle is a DGA tool included in the IEC 60599 Gas Guide.HV lead corona ring
shield connection(note burn mark on corona ring
material & eroded bolt)
Serveron Story #12
-
8/10/2019 Practicing DGA 09072010
165/172
A BPL GLOBAL COMPANY165
Serveron Corporation 2007. No use without permission.
165
3-phase, 1100 MVA, 345 kV GSU transformer
Serveron Story #12 Analysis:
-
8/10/2019 Practicing DGA 09072010
166/172
A BPL GLOBAL COMPANY166
Serveron Corporation 2007. No use without permission.
166
The Duval Triangle is a DGA tool included in the IEC
60599 Gas Guide.
Rogers Ratios are included in IEEE 57.104 Gas
Guide (similar to Basic Gas Ratios in IEC-60599)
1. Both the Duval Triangle and Rogers Ratio analysis shows the fault condition is in T2
indicating a thermal problem getting worse in the range of 300
C to 700
C
2. Combustible gas levels were rising very quickly, exceeding preset rate of change limits.
Transformer load reduction began approximately 32 hours after levels began to change
and was fully de-energized within approximately 52 hours
3. Root cause of the problem was insulation design issues around HV and LV leads
Serveron Story #12
-
8/10/2019 Practicing DGA 09072010
167/172
A BPL GLOBAL COMPANY167
Serveron Corporation 2007. No use without permission.
Serveron Story #12
-
8/10/2019 Practicing DGA 09072010
168/172
A BPL GLOBAL COMPANY168
Serveron Corporation 2007. No use without permission.
How to proceed with DGA results coming from the on-line gas monitor
No need to check for errors or inconsistencies in DGA
-
8/10/2019 Practicing DGA 09072010
169/172
A BPL GLOBAL COMPANY169
Serveron Corporation 2007. No use without permission.
169
No need to check for errors or inconsistencies in DGA
values Diagnosis may be available on-line using the main
diagnosis methods
What should still be evaluated is the reliability of the
diagnosis, depending on gas level, and the severity of thefault as in the case of manual DGA
Also, decide on appropriate actions on the equipment asin the case of manual DGA, once the diagnosis is
confirmed
Your Laboratory DGA
E i ti f DGA d t f l t d t f
-
8/10/2019 Practicing DGA 09072010
170/172
A BPL GLOBAL COMPANY170
Serveron Corporation 2007. No use without permission.
170
Examination of your DGA data from selected transformers
to get a diagnosis
The TM series monitors
-
8/10/2019 Practicing DGA 09072010
171/172
A BPL GLOBAL COMPANY
Transformer Mounted
-
8/10/2019 Practicing DGA 09072010
172/172