Advanced Machines and Energy Systems
(AMES) Group
ENERGY EFFICIENCY
PROJECTS AT UCT
Department of Electrical Engineering
RMWG Meeting
20 Feb 2008
1
Overview

Vision

Key Group Members

Collaboration

Research Outputs & HR Development

Laboratory Facility

Research Areas

Details of Current Research
RMWG Meeting
20 Feb 2008
2
Vision

To provide feasible technical solutions to relevant industrial problems, whilst
maintaining a high scholarly research content

This is achieved by engaging highly skilled personnel and by applying a
methodical approach to problem solving

To disseminate research findings through technical reports and peerreviewed publications

To develop human resource capacity in electrical machines, drives and
energy systems, which will eventually contribute towards innovation and
poverty alleviation
RMWG Meeting
20 Feb 2008
3
Key Group Members
Prof. Pragasen Pillay
PhD, CEng, FIEEE, FIET
Professor UCT, Clarkson
Electrical Machines, Drives
and Renewable Energy
Dr. Paul Barendse
PhD, MIEEE
Lecturer, UCT
Electric Drives, Fault
Diagnosis
Dr. Azeem Khan
PhD, MIEEE
Senior Lecturer, UCT
PM Machines & Drives,
Wind Energy
Mr. Marubini Manyage
MSc, MIEEE
Research Officer, UCT
Machine Design, Energy
Efficiency
Dr. Ben Sebitosi
PhD, CEng, MIEEE
Snr Research Officer, UCT
Rural Electrification,
Renewables, Energy Policy
Mr. Chris Wozniak
BSc
Technical Officer, UCT
Electrical Machines &
Energy Systems
Post Graduate Students

PhD Students: (5)
•

MSc Students: (8)
•

Mr F. Endrejat, Mr M. Manyage , Mr R. Naidoo, Mr D. Singh, Mr R. Okou
Miss P. Ijumba, Miss K. Masemola, Mr R. Solomon, Mr G. Mwaba, Mr T. Madangombe, Mr H. Mzungu,
Mr S. Sager, Mr P. Kiryowa
BSc(Eng) Final-year Thesis Students: (15)
•
RMWG Meeting
About 15 students supervised per year for final-year thesis project
20 Feb 2008
4
Collaboration – South Africa

Academic Institutions:
• DUT - Deepak Singh, Senior Lecturer – Physics
• DUT - Dr. Poobie Govender, Prof Krishnan Kanny
• University of Pretoria – Mr Raj Naidoo, Senior Lecturer – EE
• University of Stellenbosch – Prof W. van Niekerk, Prof M.J. Kamper, Dr D.
Johnson

Government / Parastatals:
• NRF
• THRIP
• DST – Mr L. Simpsom, Ms T. Mailula, Dr Boni Mehlomakulu
• SANERI – Mr K. Nassiep, Dr M. Bipath, Dr T. Mali

Industry:
• ESKOM – Mr L. Pillay, Mr S. Bisnath, Mr J. Gosling, Dr S. Higgins, Mr R. Koch, Mr
Y. Brijmohan, Dr T.L. Mthombeni,
• SASOL – Mr F. Endrejat, Mr T. Perumal, Mr D. Willemse
• LHM – Mr R. Melaia
RMWG Meeting
20 Feb 2008
5
Collaboration – International

Clarkson University, Potsdam, NY, USA
• SMMA - The Motor and Motion Association
• EMERF Consortium - Electrical Motors Educational and Research Foundation
• NYSERDA - New York State Energy Research and Development Authority

Aalborg University, Denmark (wind energy)

University of Picardie, France (condition monitoring)

Politecnico di Torino, Torino, Italy (motor lamination losses)

Graduate school of Telecom. and IT, Addis Ababa, Ethiopia (Cellphone Keyboard
Customization for Rural Applications)

GTZ, German Aid Agency (environment and rural electrification)

Centre for Research in Energy and Energy Conservation of Makerere University,
Uganda (white LEDs dissemination)
RMWG Meeting
20 Feb 2008
6
Research Outputs & HR Development
2007
2006
2005
2004
Human Resource Development
3xPhD
1xMSc
12xBSc
2xPhD
0xMSc
6xBSc
1xPhD
1xMSc
5XBSc
1xPhD
0xMSc
6XBSc
Journals papers
(Peer-reviewed, international)
6
5
6
11
Conference papers
(Refereed international mainly and some local)
16
12
14
7
Contract Research - Eskom
3
3
1
1
Contract Research - Other Industries
1
1
1
0
RMWG Meeting
20 Feb 2008
7
Laboratory Facility






Flexible distribution system with the capability of various DC and AC supplies
Two, 250kW DC machines and 4-quadrant drives
• Fed directly from the UCT 11kV ring mains through 11kV/500V transformers
6.6kV, 520kW Alternator (driven by afore-mentioned DC machines)
75kW induction motor with a 75kW drive
Several small and medium DC, AC machines and drives including test benches and
testing equipment
These unique capabilities, allows lab testing of machines that is not capable at some
international institutions
RMWG Meeting
20 Feb 2008
8
Current Research Areas

Energy Efficiency - Machines
• Core loss study
• Electric motors for demand side
management
• MV petrochemical drives

Machine Design Projects
• Small wind generator design
• Low voltage High Current Traction
motor design


Rural Electrification and Alternate
Energy Sources
• Applications of white LEDs
• Optimization of solar water pumping
systems
• Solar water heaters
• Flywheels for energy storage
• Biomass

Power Systems Applications
• Power quality
• Dip classification using wavelets
• Impacts of renewable energy
sources on power systems
Condition Monitoring
• Fault studies and condition
monitoring of induction motors, PM
motors and wind generators
RMWG Meeting
20 Feb 2008
9
Core Losses
Motor Lamination Core Losses

Background:
• Motorized applications are major electricity consumers, in SA
and USA, 64 % and 60 % of total electricity, respectively
• Core losses can be 25 % ~ 30 % of the total losses, even
higher with newer designs, such as SRMs and BDCMs
• Variable speed drives produce harmonics that increase core losses

Research Focus:
• Develop a scientific understanding of lamination core losses
• Develop core loss design equations suitable for motor designs
applications especially in software design packages

Goals:
• Improving motor efficiency by reducing core losses
• Aid motor designers with better models
• Realize energy and dollar savings
• Reducing peak demand levels and delaying the need for new stations

Environmental benefits:
• Reduce C02 emissions by efficient use of electricity
RMWG Meeting
20 Feb 2008
10
Core Losses
Core Loss Predictions

Classical core loss predictions use:
Steinmetz formulae for predicting
area under BH-Loop

Several improvements suggested over
the years


P  Ph  Pe
However current disagreements in
literature on:
• Computing coefficients for formulae
• Structure of coefficients
• Dependence on the operational
parameters, such as frequency and flux
density
Current work in this areas has led to
reliable analytical expressions for
predicting core losses in lamination strips

Formulae validated on Epstein test bench

Plan to integrate formulae into Finite
Element Analysis software - Magsoft
RMWG Meeting
20 Feb 2008
P  kh fBp1.6  ke ( fBp )2
P  kh fBp abB  ke ( fBp )2
P  kh fBp  ke ( fBp )2  kex ( fBp )1.5
P  kh f  Bp  ke ( fBp )2  kex ( fBp )1.5
P  kh fBp
a bBp cBp2
P
kh1
B p
 ke ( fBp )2  kex ( fBp )1.5
f
 ke ( fBp )2
k
k
 h 2  h3
Bp
Bp
11
Core Losses
New Commercial Test Bench System





Fully customized
• Streamlined for high f tests
• The only unit in the US!
Uses a 352-turn frame
Up to 4.0 kHz
Fully automated
• Can run tests faster
Temperature monitoring
RMWG Meeting
20 Feb 2008
12
Machine Design
Traction Motor Design
Pallet truck
Mr. Marbini Manyage
Funding: ESKOM Senior Fellowship, US DOE and NYSERDA
Traction motor

Description and Use
• Design a high efficiency Low Voltage High Current Permanent Magnet
Synchronous Motor for traction applications
• Motor will be used in a 24V battery-operated pallet truck
• Compete with DC and AC induction motors
• Benefits: Long battery lifespan and extended operating cycles

Design Challenges
• Low voltage inverter limit (14.5 VLL AC)
• Cogging and Ripple torque
• No cooling, Maximum temperature (180degC)
• Stator outer diameter < 120mm

Efficiency improvement
• Better core loss prediction using improved core loss formula and new test bench
• Choice of laminations
• Reduce winding resistance
RMWG Meeting
20 Feb 2008
13
Machine Design
Traction Motor Prototyping
RMWG Meeting
20 Feb 2008
14
Machine Design
SMC Axial-flux PM Generator
Dr. Azeem Khan
Funding: 2004 – present NRF Thuthuka; April 2005 – December 2006
ESKOM, US DOE, NYSERDA, Warner Energy



Background:
• Axial-flux PM generator design has highest torque density
• However, slotting of AFPM stator core is problematic:
 Difficult to machine tape-wound stator core
 Magnetic properties of core affect by machining process
SMC Axial-flux PM generator with single rotor, double stator:
• Uses Soft Magnetic Composite (SMC) material
• Easy to manufacture - Slotted cores are pressed
• Shorter flux paths, high torque density, high efficiency
Previous work showed need for composite (SMC + steel)
stator core structure:
• Steel in magnetic circuit increases effective
permeance of circuit, thus reducing effect
of lower SMC permeability
• Steel in circuit also reduces SMC required,
thus reducing effect of higher SMC core loss
RMWG Meeting
20 Feb 2008
15
Machine Design
Construction of the prototype




SMC teeth:- Machining of SMC cores by end-mill:
• Cost effective, easily accessible process
• Avoids high cost of pressing SMC parts for prototyping
• Good dimensional tolerances
Pre-machined SMC core:
• Core in pressed and heat-treated state
• Good insulation between iron regions through bulk of material
• Microscopic image of SMC surface shows this clearly:
Machined SMC core:
• Machining action results in elongation / smear of iron regions on machined surfaces
• Degradation of insulation between iron regions
• Increased conductivity and hence eddy current losses on / near machined surfaces
Acid treatment process introduced to eliminate smeared iron on machined surfaces
• Phosphoric acid solution used to etch smeared iron
• Acid reacts with iron only and not with insulation epoxy
• Etched parts ultrasonically cleaned in methanol to prevent corrosion
• Lower surface conductivity
Pre-machined
RMWG Meeting
Machined
20 Feb 2008
Acid treated
16
Machine Design
Stator cores tested

Two identical machined SMC cores prototyped:
• 1st case : machined SMC core with untreated teeth
• 2nd case: machined SMC core with acid treated teeth
Untreated core
RMWG Meeting
Acid treated core
20 Feb 2008
17
Motor Efficiency
Effect of Armature Rewinding on
Induction Motor Efficiency
Mr. Heskin Mzungu
Funding: 2007 – ESKOM



Background
• Motorized loads such as induction motors account for 60% the load in industry.
• With more than 5 billion Rands spent on over 2 years on motor repair, the impact
of motor repair on efficiency in South Africa is unknown.
Research Focus
• The comparison of the different procedures followed in international standards
such as the IEEE 112, IEC 60034, JEC 37 and others
• The impact of the process of armature rewind on induction motor efficiency
Objectives
• Construction of state of the art test rigs to produce very accurate efficiency
values for the different motors
• Rigorous testing of motors before and after rewinding
RMWG Meeting
20 Feb 2008
18
Motor Efficiency
Test Rigs


Two test rigs commissioned in Electric Machines lab at UCT
Range of motors to be tested and rewound: 3kW, 7.5kW, 11kW, 15kW, 22
kW, 37.5kW, 45kW, 55kW
Inline
Torque
Transducer
2:3 Pulley
System
Dynanometer
250kW
Dynamometer
Tested AC
Motor
15kW tested
motor
Inline Torque
Transducer
Torque transducer
amplifier
15kW test rig to test from 3-15kW
RMWG Meeting
55kW test rig to test from 22-55kW
20 Feb 2008
19
Motor Efficiency
Test Rigs

Speed vs Load
Accuracy and repeatability of
tests are very important. A 15kW
motor is used to validate this.
1500.00
1490.00
Speed (rpm)
1480.00
Efficiency vs Load for 15kW motor
94
1470.00
1460.00
1450.00
1440.00
1430.00
92
1420.00
0
20
40
60
80
100
120
140
160
Loading (%)
Losses vs Torque^2
88
800
700
86
Stray Load Losses (W)
Efficiencies
90
84
82
600
500
400
300
200
80
0
20
40
60
80
100
120
140
Loading (%)
160
100
0
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
Torque Squared (N.m ^2)
RMWG Meeting
20 Feb 2008
20
Motor Efficiency
Comparison of Efficiency Standards

Efficiency tests performed as per
SANS IEC 34-2, JEC 2137, IEEE
112 and CSA 390 standards and
compared with Pout/Pin and catalogue
IEEE
IEC 61972
IEEE 112
SANS IEC 34-2
CSA-390
Catalogue
JEC
SANS IEC 34-2
Pout/Pin
89
Discrepancy between the efficiencies
are due to the treatment of stray
losses, temperature (Pout/Pin) and
manufacturers quoting calculated
catalogue efficiencies rather than
tested values
CSA-390
IEC 61972
90
Efficiency (%)

88
87
86
85
84
JEC
83
15
25
35
45
55
65
75
85
95
105
Loading (%)
90
Standard
motor
88
Efficiency (%)
86
84
82
80

78
76
High Eff
motor
Two 3kW motors tested (standard
and high efficiency motors)
74
72
70
10
20
30
40
50
60
70
80
90
100
Loading (%)
RMWG Meeting
20 Feb 2008
21
Motor Efficiency
Comparison of Efficiency Standards


Standard
Method of SLL
IEEE 112
Indirect Method*
SANS IEC 60034-2
Assumed Values
JEC 2137
Ignore SLL
CSA 390
Indirect Method
Stray Load losses are losses that are the most difficult to measure. This is
due to there non-linearity and numerous causes.
The different standards employ different ways of calculating them. This is
where the biggest difference is in the standards
*
Pstray  (Pelec  Pmech )  (PFr ,W  PFe  Protor  Pstator )
RMWG Meeting
20 Feb 2008
22
Motor Efficiency


Investigation on the effects of
temperature on efficiency during
testing
Losses increase with an increase in
temperature. Efficiency therefore
is affected
100%
115.00
Thermocouple 1
Thermocouple 2
Stray Load Loss (W)
Effects of Temperature
490
440
390
340
290
240
105.00
0
75%
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
Torque Sqaured
Temperature (C)
95.00

SLL variation of up to 1%
85.00
50 %
75.00
25%
65.00
55.00
0.00
1000.00
2000.00
3000.00
4000.00
5000.00
6000.00
7000.00
8000.00
Time (sec)
RMWG Meeting
20 Feb 2008
23
Motor Efficiency
Effects of Temperature



Efficiency maximum variation of up to 0.5%
This is significant when trying to investigate the effects of rewinding
It has been reported in literature that motor can loss or gain up a 1%
in efficiency
91
Efficiency (%)
90
89
Series1
88
Series2
87
86
85
0
20
40
60
80
100
120
140
160
Loading (%)
RMWG Meeting
20 Feb 2008
24
Motor Efficiency
Armature rewinding


Impact on motor efficiency (and/or losses)
can occur at any point during the rewind
process
South African Rewind and Refurbishment
standards such SANS 1804 and 10242 have
a standard procedure
Motor loss
Winding
Reducing conductor cross-sectional area
Changing number of turns
Using wrong winding configuration
Machining rotor
Oven Burn
Out or
Mechanical
Stripping
RMWG Meeting
Winding data and
Winding Stripping
All parts are
cleaned
Electrical tests
are done on Rotor
Core Testing
Paint and
Insulate Core
Varnishing by
impregnating
and baking
Rotor is
balanced
Bearings are
replaced if needed
Stator rewound
and insulated
All electrical connections
done and reassembly of motor
Rotor losses Altering rotor bars and end rings
(I2R losses) Changing cage design
Change in air gap symmetry and air gap
unconventional behavior
Bent motor shaft or damaged end
shields
Mechanical
Overhang is
cut
Bearings and cooling fan not replaced to
Friction and
original conditions
Windage
Incorrect bearing fits
losses
Incorrect bearing preload
Stray load
losses
Visual inspection for fault
Affected by
 Overheating core steel during stripping
Damaging core insulation during winding
Stator core
removal
losses
Excessive abrasion and grinding during
core cleaning
Stator
Winding
Losses (I2R
losses)
Visual inspection
for fault
Final Testing
20 Feb 2008
25
Motor Efficiency
Comparison of 3kW Std and HE Motor

Two induction motors were compared:
• Standard motor: 3kW, 4-pole
• High efficiency: 3kW, 4-pole

Objectives:
• Efficiency differences between motors assessed
•
Operating performance differences between motors, when subjected to the same
load assessed
•
To assess effectiveness of
retrofitting standard motors
with high efficiency motors
RMWG Meeting
20 Feb 2008
3kW HE Motor with Dynamometer
26
Motor Efficiency
Methodology

IEC 60034-2-1 standard used to assess efficiency both motors

Equivalent circuit parameters determined for both motors:
• No-load test
• Locked Rotor test

Operating performance differences assessed using equivalent circuits and
superimposing pump load curve
RMWG Meeting
20 Feb 2008
27
Motor Efficiency
IEC60034 Test Results

IEC 60034-2-1 standard used to assess efficiency both motors:
• 3kW Standard and High Efficiency
motors tested
• Full temperature compensation for losses
as per standard
Stator Copper losses vs load
400
350
Stator Cu losses [W]
300
Indirect Method Efficiency vs Load
90
Std-Motor
Best-fit
88
H-Eff-Motor
250
200
150
Best-fit
100
86
50
0
20
84
Efficiency [%]
Std-Motor
Cubic-fit
H-Eff-Motor
Cubic-fit
40
60
80
100
Load [%]
120
140
160
Rotor Copper losses vs load
82
700
600
80
Rotor Cu losses [W]
500
78
76
400
300
200
74
20
40
60
80
100
120
140
Load [%]
Efficiency vs Load comparison
RMWG Meeting
20 Feb 2008
160
Std-Motor
Cubic-fit
H-Eff-Motor
Cubic-fit
100
0
20
40
60
80
100
Load [%]
120
140
160
28
Motor Efficiency
Equivalent Circuit Parameters

Equivalent circuit parameters determined for both motors:
• No-load test
• Locked Rotor test
• IEEE recommended equivalent circuit used:
IEEE recommended equivalent circuit

Equivalent circuit parameters:
•
Standard motor
RMWG Meeting
High Efficiency motor
20 Feb 2008
29
Motor Efficiency
Operating Performance

Equivalent circuits used with Matlab program to predict performance of motors

Centrifugal pump load curve superimposed on characteristics of both motors

Comparison with experimental results
RMWG Meeting
20 Feb 2008
30
Motor Efficiency
Operating Performance: Torque vs Speed


Torque vs speed curves of both motors with centrifugal pump load curve superimposed
Experimental results show good correlation
IM Torque vs speed characteristic
60

High Eff Calc
Std Calc
High Eff Exp
50
Std Exp
Centrifugal load
High Efficiency motor:
• Slip=4.85%
• Speed=1427.3rpm
40
Torque [Nm]

Standard motor:
• Slip=6.1%
• Speed=1408.5rpm
30
20
10
0
0
500
1000
1500
n [rpm]
RMWG Meeting
20 Feb 2008
31
Motor Efficiency
Operating Performance: Current vs Speed


Current vs speed curves of both motors
Experimental results show good correlationIM per-phase stator current vs speed characteristic
30

Standard motor:
• Slip=6.1%
• Speed=1408.5rpm
• I1_load=6.28A
• pf=0.78 lagging
High Eff Calc
Std Calc
High Eff Exp
25
Std Exp

High Efficiency motor:
• Slip=4.85%
• Speed=1427.3rpm
• Current=6.1A
• pf=0.8 lagging
Current [A]
20
15
10
5
0
0
500
1000
1500
n [rpm]
RMWG Meeting
20 Feb 2008
32
Motor Efficiency
Operating Performance: Efficiency


Efficiency vs speed curves of both motors
Experimental results show good correlation
IM Efficiency vs speed characteristic
100

Standard motor:
• Slip=6.1%
• Speed=1408.5rpm
• Eff_load=83.1%
High Efficiency motor:
• Slip=1%
• Speed=1427.3rpm
• Efficiency=87.8%
High Eff Calc.
90
Std. Calc.
High Eff. Exp
Std. Exp
80
70
60
Efficiency [%]

50
40
30
20
10
0
1000
1050
1100
1150
1200
1250
1300
1350
1400
1450
1500
n [rpm]
RMWG Meeting
20 Feb 2008
33
Motor Efficiency
Operating Performance: Energy

Energy saving based on reduction in input power drawn by High Eff motor:
• Same centrifugal pump load
• 3kW Std motor replaced with 3kW High Eff

Standard motor input power:
• Pin = 3.39kW

High Efficiency motor input power:
• Pin = 3.38kW

Reduction in input power drawn by High Eff motor:
• ∆Pin = 10.64W !!!
• ∆Pin = 0.3% !!!
RMWG Meeting
20 Feb 2008
34
For further enquiries, please contact:
Name:
Number:
E-mail:
Dr Azeem Khan
Dr Ben Sebitosi
Prof Pragasen Pillay
021-650-5956
021-650-5253
Azeem.Khan@uct.ac.za
Adoniya.Sebitosi@uct.ac.za
pillayp@clarkson.edu
RMWG Meeting
20 Feb 2008
35