Transformer Overloading and

Assessment of Loss-of-Life
for Liquid-Filled Transformers

P.K. Sen, PhD, PE, Fellow IEEE

Professor of Engineering
Colorado School of Mines

PSERC Project Webinar, April 19, 2011

 Presentation Outline

• Introduction and Motivations

• Remaining Transformer (Insulation) Life,
Transformer Thermal Model, IEEE Loading
Guides
• Modeling and Economic Evaluation
• Simulation and Results
• Conclusions

2
Transformer Type and Nameplate Rating
• Oil-Filled Transformer
- <500 kVA (distribution, OA)
- 500 kVA - 100 MVA (distribution power) (OA, OA/FA, OA/FA/FA)
- >100 MVA (transmission power) (OA/FA/FA and OA/FA/FOA)
• 65°C average winding temperature rise
(thermally-upgraded paper insulation)

• 55°C average winding temperature rise

(not thermally-upgraded paper, kraft)

• 30/40/50/56 MVA @55o / 65o C TR (OA/FA/FA)

@55°C TR 3
 Issues and Considerations
• Thermal Model
- Design Consideration (Losses & Cooling)
- Temperature (Top & Bottom Oil, Winding Hottest-Spot Temp.)
- Insulation Degradation
- Estimation of Transformer Remaining Life from Its Insulation

• Economic Model
- Total Owning Cost (TOC)
- Revenue Requirement
- Other Models!!
4
Thermal Model Optimization
• Design (Minimum Cost & Optimum Sizing)
•Temperature • New Design
• Insulation • Retrofit Design
• Life - Replace now
- Replace later
- Add a second transf.
- ??

Economical Model
Revenue Requirement Uncertainties & Probability
Approach
• Future load growth projection
•Ambient temp.
•Transformer Failure

5
 Loading Guide Standards

• IEEE Std. C57.91-1995 (USA)

– Clause 7 (Classical Model)
– Annex G (Detailed Model)

• IEC 354-1991 (EUROPE, CANADA and Others)

• IEEE & IEC Comparisons

IEC is somewhat similar to IEEE Clause 7

6
 Transformer Failure Modes
• Long-term Failure Mode


– Mechanical and chemical properties of insulation
deterioration
– Dielectric strength reduction (approx. 10% over
the transformer life)


• Short-term Failure Mode
– Bubble formation (gassing), temporally dielectric
strength reduction (30% impulse strength
reduction @ 180 °C )

7
 Remaining Transformer
(Insulation) Life
• The Tensile Strength (Mechanical) of cellulose
insulation is used as a criteria for loss-of-life
estimation.

• Recently, the Degree of Polymerization (DP)

(Chemical) has also become a popular measurement
of insulation life.

• If a sample of insulation near hot spot location is

made available, the remaining life of the transformer
could be approximately predicted based on specific
criteria.
8
 Remaining Life by Tensile
Strength Method
120 The normalized remaining life
by tensile strength method
100
based on 20% remaining tensile
strength (RTS) as end-of-life criteria,
80
created by least square method
% Tensile Strength

60

 RTS 
40 Remaining Life  1  0.633 ln  
 97.05 
20

0
1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1 -0.2
Remaining Life (per unit)
9
 Remaining Life by DP Method
700

600
The normalized remaining life
by DP method based on remaining
500 DP of 200 as end-of-life criteria,
created by least square method
Degree of Polymerization

400

 RDP 
300
Remaining Life  1  0.881ln  
 622 
200

100

0
1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1 -0.2
Remaining Life (per unit)

10
IEEE Std. C 57.91-1995, Per-Unit Life
 15, 000 
 
• Per-Unit Life 18  THS  273 
L  9.80 x 10 e
 15, 000 
 39.164  
dL THS  273 
• Relative Aging Factor FAA  e 
dt
1
T

• Equivalent Aging FEQA   FAA dt

T0
FEQA x T
• % Loss-of-Life = × 100
Normal Insulation Life

• Normal Insulation Life

11
• Hottest-Spot Temperature (THS)
 THERMAL MODEL

Hottest-Spot Temperature (T )
HS

Hottest-spot temperature = ambient temperature (TA)

+ top oil rise over ambient (DT TO)
+ hot spot rise over top oil (DTG)
12
Thermal Model (IEEE Clause 7)
Steady-State Equations

THS  TA  DTTO  DTG

n
 K R 1
2
DTTO  DTTOR  
 R 1 

DTG  DTGR  K   2 m

K = Per-Unit Loading (Actual Load/Rated (max) Load)

R = Loss Ratio (Load Loss @ Rated Load/No-Load Loss)
Exponent „n‟and „m‟depends on “Cooling”
13
Thermal Model (IEEE Clause 7)
Transient Equations

Oil dDTTO
 TO  DTTO  DTTO ,u
dt
n
 K R 1
2
DTTO ,u  DTTOR   
 R 1 

Winding dDTG
G  DTG  DTG ,u
dt
DTG ,u  DTGR  ( Ku2 ) m
14
Modified Transient Equations

dDTTO
 TO  DTTO  DTTO ,u  DTA
dt

1 1
DTTO ( s)   DTTO ,u ( s)   DTA ( s)
1   TO s 1   TO s

15
Thermal Block Diagram
(IEEE Clause 7)

16
Aging Rate and % Loss-of-Life
 15, 000 15, 000 
  
383  273
Aging Rate  e  
THS

• Loading cycle, Complex load growth

• Ambient temperature variation Thermal

Model
 15, 000 15, 000 
  
383  273

Dt *100 
T
e HS

%loss of life 
Normal Life
Aging rate,
%Loss-of-life

17
 IEEE “Clause 7” & “Annex G”
Comparisons

• “Clause 7” thermal model: Simple, Requires minimum

information, uses top-oil temperature

• “Annex G” thermal model: More complex, Requires

more information, Focus on duct oil temperature change,
uses bottom-oil temperature

18
 “Clause 7” and “Annex G”
Data Requirements
IEEE Classical Model (Clause 7)
1. Top oil temperature rise at rated load
2. Hot spot temperature rise over top oil at
rated load
3. Loss ratio at rated load
4. Winding time constant
5. Oil time constant or
6. Type of cooling system
IEEE Detailed Model (Annex G)
1 Top oil temperature rise at rated load
2 Hot spot temperature rise at rated load
3 Average winding temperature rise at rated
load
4 Bottom oil temperature rise at rated load
5 Losses data from test report
6. Weight of Core & coil
7. Weight of tank & fittings
8. Gallons of fluid
9. Type of cooling system
10. Type of cooling fluid
11. Type of winding material
12. Winding time constant
13. Location of hot spot
14. Per unit eddy current losses at hot spot
location
19
IEEE Annex G & Clause 7 Thermal
Model (Step Load Response)
T e m p e r a tu r e Pr o f ile Plo ts ( IEEE A n n e x G T h e r m a l M o d e l)
120 5
Ho t s p o t

4 .5 T o p o il
L o a d in g
100
4

3 .5
80

Loading (Per unit)

3

60 2 .5

40
1 .5

1
20

0 .5

0 0
0 2 4 6 8 10 12
T im e ( h r )

20
 Economic Evaluation
(Total Owning Cost)

Buying a New Transformer

Simplified Practice,
TOC = IPP + A * NLL + B * LL

Owner Specifies: Manufacturer Specifies:

• Size (MVA) • Initial purchase price (IPP)
• A & B Factors • Losses (Winding & Core)
21
 Economic Evaluation
(Revenue Requirement)

• Daily load curves

Revenue Requirement
• Daily ambient temp. variations
• Overloading & end-of-life
Applicable for • Accurate energy loss
Investor Owned Utilities
(IOU) • Transformer replacement
• Refined & flexible method
• Detailed economic analysis

22
 ECONOMIC
THERMAL MODEL
MODEL

OPTIMIZATION
• NEW DESIGN
• RETROFIT APPLICATIONS

23
 Optimization
(New Procurement)
Transformer Data

Daily load curve, Complex load growth

Ambient temperature variation Thermal Model

Life Expectancy
Economic data
- discount rate, tax, Economic Model
- energy cost, etc.

TOC

24
 Optimization
(Retrofit Design)
Existing system Historical load and
ambient temp. data Replace with new
transformer now

Estimate
Remaining Life Present load Overload
Load growth existing transf.
and replace with


Load shape new one later
Ambient temp.
Economic data Add a 2nd
Remaining etc. transformer
Life Expectancy
25
 Complete Method Block Diagram
Probability Tree Structure Path Levelized Probability
Revenue
Requirement
1 $134,000 0.0016
Probability
Distribution Probability

Load
81 $101,000 0.0016

Energy and Demand Charge

Escalation Rate Equivalent Levelized
Economic Data Revenue Requirement (ERR)

Ambient Temperature and

Load Profiles Cost of Losses Utility’s
Historical Transformer Thermal Financial
Data Estimated Remaining Life Model Cost of Transformer
Model
Cost of Failure

New Transformer End-of-Life

Sizing Criteria Debt Ratio
Equity Return
Transformer Data Thermal Data, Electrical Losses Borrowed Money Rate
Size, Losses, Cost Random Failure Model Tax Rate
Depreciation Method
Book Life

26
 SIMULATION
CASE STUDIES
AND
RESULT

• Case #1 - New Design

• Case #2 - Retrofit Applications

27
 Case #1: New Design
Example Data

• Buy new transformer for a new project

• Initial peak load = 20 MVA
• Load growth in first 4 years (Probability)
- High growth = 2.00%
- Moderate growth = 1.75%
- Low growth = 1.00%
• Load growth after the 4th year is 2.00%

28
 Case #1: Hot-Spot & Top-Oil
Temperature Plots
Hot Spot and Top Oil Temperature

Hot Spot
Top Oil
160

150

140

130

120

110

100

90

80

70

60

50

40

2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030
Year
29
Case #2: Retrofit Design
Example Data
• Existing transformer rating = 18 MVA
• In service for 25 years
• Estimated remaining life = 0.25 pu.
• Initial peak load = 20 MVA
• Load growth in first 4 year (Probability)
- High growth = 2.00%
- Moderate growth = 1.75%
- Low growth = 1.00%
• Load growth after 4th year is 2.00%
• Replace with new unit now or continue overload
..and replace later ? 30
Windows Based Computer Program

• Graphic interface, Load, Save, Print, and Copy

• Transient Loading Analysis (Daily load)

• Life Cycle Study (with load growth from table)

• Optimize Transformer Size

• Delay Replacement or Replace now Analysis

31
Screen Shot (1), Main Menu

32
Screen Shot (2), Annex G Input

33
Screen Shot (3), Load Profile Input

34
Screen Shot (4), Load Growth Input

35
Screen Shot (5), Temperatures Plot

36
Screen Shot (6), Life Cycle Charts

37
Screen Shot (8), Economic Data Input

38
 Conclusions

• Developed an Optimization Model for the

“Best Utilization” of a transformer:
– Transformer replacement strategy (retrofit
design)
– Optimum transformer sizing (new design)
• Computer program is written to facilitate
long and complex calculations
• Windows based computer program
• Asset management tools
39
 Future Work
• New programming is required for faster
optimization computation

• The different depreciation method (economic

model) shall be explored

• Loss-of-life and hottest-spot calculation shall

be updated when more real data is made
available

• Other financial model shall be explored

• Sensitivity analysis on Loss Ratio (R) 40

 Now publicly available
on the PSERC Website
PSERC Project Report:
Transformer Overloading and
Assessment of Loss-of-Life
for Liquid-Filled Transformers
Project Leader: P.K. Sen
Professor of Engineering
Colorado School of Mines
Golden, CO 80401
psen@mines.edu, (303) 384-2020