Temperature Rise Tests
Centre for Power Transformer Monitoring, Diagnostics and Life Management (transformerLIFE)
Monash University, Australia
Oleg Roizman
IntellPower, Australia
Valery Davydov
Monash University
Spring 2009 IEEE Transformers Committee Meeting
Miami FL, 21 April 2009

Special Test Transformer Nameplate Data
Year of Manufacture
Rated Power HV/MV for ONAN, kVA
Rated Voltage HV/MV/LV, kV
Rated Current HV/MV
Cooling Types
Amp
Number of Phases
Vector Symbol
Mass Untanking, kg
Mass Each Cooler (excluding oil), kg
2006
468/468
22/4.5/0.415
12.3/60.0
ONAN, ONAF,
OFAF
3
YNyn0yn0
2420
115
Mass Total (including oil), kg
Insulating Oil Each Cooler, l
Insulating Oil Total,
Oil Circulation, l/min litre
6850
43
2650
1200

Temperature Sensors and DAQ System
• 16 Fiber Optic sensors
• 24 Thermocouples, including magnetic and thermal ribbon types
• 9 RTDs, including those of moisture/temperature transmitters
• More than 60 channels of information stored at 1 min interval

Location of Thermocouples and RTDs in Test Transformer
Top Oil Temperatures
Top Core Yoke Temperatures
Top Rings Temp
Top Radiator
Bottom Radiator

Location of Fibre Optic Probes and Core
Thermocouples

Fibre Optic Probe Installation in MV Winding, Phase B

Fibre Optic Probe Installation in HV Winding

Effect of Measuring Instrumentation
• In the following 4 slides, comparisons are made for the two windings of Phase B (OFAF)
– MV (layer type)
– HV (disc type)
• The comparisons are made for results obtained during the Temperature Rise Tests conducted at Monash using two different instrumentation sets for winding resistance measurement

55
60
65
Comparison of Tavr and FO for MV OFAF
Phase B, 100A
1st set of winding resistance measurement instrumentation was used
Top wnd. temp
Mid wnd. temp
50
45
40
Avr wnd. T
35
30
0 1 2
FO12 MV top
3
Bot wnd. T
4 5
FO13 MV middle
6 7 8 t, min
FO14 MV bottom
9 10 11 12
θ wm(i) as measured
13 14 15
θ w(i) as calculated
16

Comparison of Tavr and FO for HV OFAF
Phase B, 100A
1st set of winding resistance measurement instrumentation was used
51
49
47
45
43
41
39
37
35
0
Tawr
2 4
FO7 HV top
Measured average winding temperature
6 8 t, min
10
FO9 HV middle
Calculated average winding temperature
12
FO11 HV bottom
14 16

OFAF (100A) HV Phase B
2 nd set of winding resistance measurement instrumentation was used
90
85
80
75
70
65
60
0:00:00 0:02:53 0:05:46
Twarm
Twarm fitted exp 5mins
0:08:38 0:11:31
Twarm fitted IEC 20mins
FO7
0:14:24 0:17:17
Twarm fitted IEC 15mins
FO1
0:20:10
Twarm fitted IEC 5mins
FO9
0:23:02

OFAF(100 A) MV Phase B
2 nd set of winding resistance measurement instrumentation was used
120
110
100
90
80
70
60
0:00:00 0:02:53 0:05:46 0:08:38 0:11:31 0:14:24
Time since shut down (mins)
0:17:17 0:20:10 0:23:02
Twarm
Twarm fitted exp 5mins
Twarm fitted IEC 20mins
FO13
Twarm fitted IEC 15mins
FO12
Twarm fitted IEC 5mins
FO14
It could be seen that depending on duration of test variations in Twnd (R) is ~10
º
C

84
82
80
78
76
74
72
90
88
86
70
7:12 8:24
Blocked Coolers: 60A
Phase A
Phase C
Blocked Coolers 60 A
Phase B
~4 hours!
9:36 10:48 12:00
FO6
13:12
FO15
14:24
FO16
15:36 16:48 18:00 19:12

72
70
68
0:00
78
76
74
84
82
80
Blocked Coolers: 60 A
2 nd set of winding resistance measurement instrumentation was used
Blocked Cooling
Ph A
Tw = 5.521
* exp(-time/ 4.331) 0.1618
* time + 76.59
Ph C
Ph B
2:24
Twarm A
4:48 7:12
Twarm fitted IEC (A)
9:36
Twarm C
12:00
Twarm fitted IEC ©
14:24
Twarm B
16:48 19:12
Twarm fitted IEC (B)
21:36

90
85
80
75
70
65
60
55
50
45
0:00
Blocked Coolers HV Phase B 60A hottest measured temp mid measured temp
2:24 4:48
Winding temp by resistance bottom measured temp
16:48 19:12 7:12 9:36 12:00
Time since shutdown (mins)
14:24
Twarm FO7 FO9 FO11 (TV4+TV5)/2
21:36

ONAN(60A) HV phase B
2 nd set of winding resistance measurement instrumentation was used
70
65
60
55
50
45
0:00:00
Twarm
0:02:53
Twarm fitted IEC 15mins
0:05:46 0:08:38
Time since shutdown (mins)
0:11:31
Twarm fitted IEC 5mins Twarm fitted exp 5mins
0:14:24
FO9 FO3
0:17:17
FO11

ONAN(60A) MV phase B
2 nd set of winding resistance measurement instrumentation was used
75
60
55
50
45
70
65
40
0:00:00
Twarm
0:02:53
Twarm fitted IEC 15mins
0:05:46 0:08:38
Time since shutdown (mins)
0:11:31
Twarm fitted IEC 5mins Twarm fitted exp 5mins
0:14:24 0:17:17
FO12 FO13 FO14

Comparison of FO Data for Phases A, B & C
• In the following slide a comparison is made for Phases A, B and C for the HV (disc type) winding for the OFAF cooling mode
• The comparison is made for the measurements obtained during the Temperature Rise Tests conducted at Monash for the FO probes installed in the 2 nd top disc of Phases A, B & C
• The differences in the FO measurement results were observed due to the following reasons:
– Phase A is the most remote phase from the oil inlet pipe; the velocity of oil flow through the winding ducts of Phase A is the lowest
– Phase C is the closest phase to the oil inlet pipe; the velocity of oil flow through the winding ducts of Phase C is the highest
– Phase B is in the middle between the inlet pipe and Phase A

FO Sensors Data for Phases A, B and C
OFAF (100A)
95
90
85
80
75
70
65
60
55
8:24
Phase B
Phase C
9:36
Cooling curves
10:48 12:00
Phase A
13:12
Time, h:mm
14:24 15:36
FO6 FO15
16:48
FO16
18:00

75
70
65
60
0:00
Cooling Curves for A, B and C Phases,
OFAF (100A), HV
2 nd set of winding resistance measurement instrumentation was used
90
Phase B Phase C Phase A
85
80
2:24 4:48 7:12 9:36 12:00
Time since shutdown, min
14:24 16:48 19:12 21:36

Effect of first valid time point for MV in OFAF
80
75
70
90
85
65
60
0:00:00
1:45 min
4:00 min
0:02:53
Twarm
0:05:46 0:08:38 0:11:31 0:14:24
Time since shutdown (mins)
Twarm fitted IEC 20mins (from 1m45s)
0:17:17 0:20:10
Twarm fitted IEC 20mins (from 4m)
0:23:02

Factors Affecting Winding Temperature Rise
• Winding resistance measuring equipment
• Ambient temperature determination
• Inadequate calculation of the average oil temperature (leads to wrong g factor)
• Accuracy of Load Loss measurement
• Assumed total loss as sum of NL + LL
• Effect of the Core temperature dynamics
• Cold resistance measurement errors
• Not reaching steady state before shutdown

Factors Affecting Winding Temperature Rise
(cont’d)
• Connection circuit (two windings at a time)
• Time interval – first and last data point resistance measurement
• Fitting curve method
• Ambient oil temperature consideration

Conclusions
• Depending on winding time constant taking first resistance measurement at 4 min may be too long wait and could lead to significant error in determination of winding temperature at shutdown
• 10 min cooling curve period could be well justified for small and medium distribution transformers, but does not seem to be adequate for large power transformers, where 20 min should be considered as more appropriate
• 15 sec acquisition rate was found to be easily achievable with the modern acquisition systems and is recommended, especially for a winding with a short time constant

Conclusions (Cont’d)
• Considerations should be given to the following recommendations when FO sensors are used: a.
Number and locations of FO temperature probes should be determined on the basis of analysis of heat and mass transfer with assistance of numerical methods such as FEM and CFD; b.
Hot-spot temperature should be continuously measured by FO sensors installed in each winding of each phase and verified by calculations in accordance with the latest relevant standards and/or more detailed inhouse thermo-hydraulic models; c.
Average winding temperature rise by resistance should be measured only in the winding of the phase with the highest hot-spot temperature found in b) unless the difference between that temperature and the average of all phases exceeds agreed value (e.g. 3
º
C).