Flicker Analysis and Case Studies
Jim Rossman, P.E.
Senior Manager of Power Quality
Gerald Johns, P.E.
Senior Power Utilization Engineer
TVA
August 2008
1
Voltage Drop Creating Visible Light Flicker
Voltage Flicker Is:
Voltage Drop as Seen by
Others
Visible With Incandescent
lights and some CFLs
Complaints Normally From
Residential Customers
Much lower in magnitude
than a fault related
voltage sag – usually
only a few volts on 120V system
Noticeable at low levels
and downright annoying
at higher levels
2
Voltage Flicker Does Not:
Normally does not cause
equipment downtime
Does not damage other
folks equipment
Flicker is Perceived Voltage Drop
Voltage Drop is Graphically Shown Below
Drawing Copied from IEEE 141
Vdrop ≅ IR cos Φ + IX sin Φ
3
Per Unit Approach to Calculating Flicker
Vdrop ≅ IR cos Φ + IX sin Φ
Hand calculations often work best in percent on an MVA base
TVA uses 100-MVA as our base so the following is the same
approach but written on a percent basis:
%Vdrop ≅
MWswing (% R1) + MVARswing (% X 1)
100
Example: Large motor draws 25-MW and 40-MVAR on
startup. With a system Thevenin equivalent
Of Z= 0.988% +j 7.170% (100-MVA base), What is the
expected voltage swing on startup?
Vdrop ≅
4
25MW (0.988%) + 40 MVAR (7.170%)
= 3 .1 %
100
Perceived Voltage Drop in Lights
GE Flicker Curves
Copied From IEEE 141
5
TVA Flicker Curve
6
Flicker Curves
IEEE 141 and TVA
TVA
Curve
In
Red
7
Three Key Flicker Components
Recipe for Complaints:
1. Start with - Varying load/generation equipment
2. Serve from a - High System Impedance
3. Repeat variations until annoying
8
Component 1:
Equipment with Load Swings Leading to
Possible Flicker Issues
• Motors and Compressors
–
–
–
–
•
•
•
•
•
9
Rock Crushers
Saw Mills
Large Motors (Car Shredders, pumping systems)
HVAC systems
Welders
Process ARC Furnaces
Process Induction Furnaces
Wind Generators
Others
Mitigation Systems Reducing Load Swings
Dynamic Var Systems
–
–
–
–
Static Var Compensators
Dstatcom
Intellivar
Thyristor-Controlled Capacitor Systems
Motor Impact Reduction
– Reduced Voltage Starters
– Wound Rotor Motor Resistors
– DC Motor Systems
Sequencing Multiple Welder Systems
Arc Furnace Reactors
10
Component 2:
Impedance Issues Important to Flicker Critical
Point of Common Coupling (PCC) for Flicker Studies
60_MVA
2
AC1
+
1
RL
+
+
PI2
RL
161kVRMSLL /_0
20_MVA
1
2
PI1
3
161/69/13
ZZ_1
+
RL1
69/13.2
.28
7Ohm
Transmission Sub-transmission Distribution Distribution
Level
Circuit Level
Bus Level Circuit Level
My PCC Definition:
Closest Interconnection Location Where Another
Customer Can See The Voltage Drop From The
Disturbing Load
11
Component 2:
Series Impedance to PCC –
Shown in Percent (100-MVA Base)
+
AC2
161kVRMSLL /_0
+
+
RL3
RL
+
+
RL4
+
RL2
RL
PI3
PI4
A
B
C
D
% Series Circuit Impedance
100-MVA base
Possible PCC
Interconnection
Transmission – 161-kV
A
(0.70+j4.82)% – total to A
Sub-transmission – 69-kV
B
(1.08+j12.6)% – total to B
Distribution Bus – 13-kV
C
(1.75+32.6)% – total to C
Distribution Circuit – 13-kV D
(30.47+j138.9)% – total to D
Hopkinsville, KY Facility - Services Went From 13-kV to 161-kV
12
Mitigation Approaches With Impedance
Reductions
Historical Solution to Flicker Problems – Reduce Impedance
and percent voltage drop by Moving PCC upstream:
– General Distribution to Dedicated Distribution Feed
– Dedicated Distribution Feed to Sub-transmission Feed
– Sub-transmission Feed to Transmission Feed
Add Series Capacitors In Line
13
Component 3:
Frequency of Voltage Variation Swings
• Startup Related (once per season, month, week, day)
– Pump Motors
– Process Systems
– Energizing Transformers
• Low Frequency Process cyclic (multiple times per day or
hour)
– Air compressors
– Refrigeration Compressors
• High Frequency Process cyclic (many times per hour of
second)
–
–
–
–
14
Rock crusher cycling
Shredders
Welders
Arc Furnaces
GE Flicker Curves
Copied From IEEE 519
15
Drawbacks of IEEE 141/519 Curve
126
125
290.0
124
287.5
285.0
122
Volts
RMS Volts
123
282.5
121
280.0
120
119
277.5
118
275.0
117
272.5
A Vrms (val)
116
280
300
320
340
seconds
360
16:04:02
03/13/2008
Thursday
380
16:04:04
16:04:06
16:04:08
16:04:10
288
120.01
286
284
282
RMS Volts
RMS volts
120.005
120
280
278
276
119.995
274
272
119.99
270
230
16
230.5
231
231.5
seconds
232
232.5
2530.5
2531
2531.5
2532
2532.5 2533
seconds
2533.5
2534
2534.5
IEEE 1453
• Based on IEC 61000-4-15 (adopts the IEC
standard).
• Provides specifications for the measurement of
flicker based on IEC 61000-4-15.
• Provides recommended flicker limits on
medium-voltage, high-voltage, and extra-high
voltage systems based on IEC 61000-3-7.
17
Advantages of using IEEE 1453 over
IEEE 141/519 curves
• Measurement that directly represents flicker level in
terms of human perception.
• Provides a way of measuring flicker when voltage
fluctuation is anything but a rectangular change.
• Impact of modulations caused by modern solid-state
converters (interharmonics) on voltage fluctuations
taken into account.
• Can be incorporated into simulation models to provide
future flicker estimates and the effectiveness of various
flicker mitigation options.
18
UIE/IEC Flickermeter
First 4 blocks produce a signal that collectively represent:
• Response of a lamp to a supply voltage variation.
• Perception ability of the human eye
• Memory tendency of the human brain.
Block 5 is a statistical calculation that emulates human irritability
to the flicker level.
Figure Source: J.C. Gomez, M.M. Morcos, “Flicker Measurement and Light Effect”, Power Engineering
Review, Nov. 2002.
19
IEEE 1453 Definitions
• Pst – “A measure (statistical) of short-term perception of
flicker obtained for a ten-minute interval. “
• Plt – “A measure (statistical) of long-term perception of
flicker obtained for a two-hour period. This value is
made up of 12 consecutive Pst values per the following
formula.” - Necessary when duty cycle varies or
multiple loads operating simultaneously.
Plt =
3
1 N
× ∑ Pst 3j
N j=1
where N = The nunber of Pst readings.
N should be based on the duty cycle of the fluctuating
load. If exact duty cycle is unknown, assume N=12 to
represent 2 hours.
20
IEC Flickermeter
• Pst of 1.0 represents a magnitude and
frequency of voltage fluctuation that is
generally considered to be objectionable.
• Measurement based on luminous fluctuation
associated with 60-watt, 60-Hz 120-VAC or 50Hz 240-VAC incandescent lamps.
21
Comparison of IEEE 141 and IEC
Flickermeter Curves
*1 dip = 2 changes
22
IEEE 1453 Recommended Flicker
Levels
However, levels that are not
objectionable may still be
perceivable.
23
Represents the
levels below
which
complaints are
not generally
received.
1453 Statistical Guidelines
• As a general guideline, when designing, Pst
and Plt should not exceed the planning levels
more than 1% of the time (99% probability
level), with a minimum assessment period of
one week.
• IEEE 1453 recommends that Pst and Plt not
exceed 1.0 pu 5% of the time in existing low
voltage and medium voltage systems (95%
probability level.
24
Other Points To Note About Flicker
• People most sensitive to fluctuation frequency of 2 to 10
Hz with flicker visible up to 35 Hz.
• Any change in voltage 6.0 % or greater results in
objectionable flicker, regardless of frequency. [1]
• Lower-wattage incandescent bulbs produce more flicker
for a given change in voltage. At the same rated
wattage, a 230-volt bulb will flicker more than a 120-volt
bulb for a given change in voltage.
• Dimmers can exacerbate the flicker problem because
flicker becomes more perceptible as baseline lumen
levels are reduced.
• Fluorescent lamps typically flicker less for a given
voltage input [2].
[1] R.Dugan, et al., Electrical Power Systems Quality, 2nd Ed., 2002.
[2]T.A. Short, Electric Power Distribution, 2005.
25
EPRI Studies of Compact Fluorescent
Lamp Gain Factors
The gain
factor is defined as the
ratio of relative light
changes to relative
voltage changes.
26
EPRI Studies of Linear T8 Fluorescent
Lamp Gain Factors
27
How About a Break!
Lets Keep It to 15 Minutes – The Best is
Yet to Come!
28
Case #1 Sawmill on Rural Feeder
7.2/12.47 kV distribution source @ mill
Primary Z = 12.94 + j10.45 ohms
9.47 miles of 14.4/24.9 kV
3.76 miles of 7.2/12.47 kV
Flicker caused by load fluctuations as
well as motor starts.
29
Case #1 - Head-Saw Load Up – 9.6% Vdrop @ Mill
Secondary and 6.6 % Vdrop On Primary
Amps
1500
500
-500
-1500
Volts
AI
280
275
270
265
260
255
BI
CI
Significant voltage reduction due to
reactive power draw.
A V rms ( val)
B V rms (val)
C V rms (val)
500
kW
400
300
200
kVAR
100
800
700
600
500
400
300
Real power draw starts to increase
due to increase load torque (large
log), but then declines because of
reduced speed.
TOT P( kW) (val)
TOT Q(kV A R)
14:49:30.0
12/17/2007
Monday
14:49:30.5
14:49:31.0
14:49:31.5
Event #1 – 150-hp head saw abruptly loaded
up – data collected at utility metering point
30
14:49:32.0
14:49:32.5
Reactive power draw increases
substantially due to extreme speed
reduction.
Case #1 – 200 hp Chipper Startup w/softstart 7.5 % instantaneous
Vdrop at Sawmill and app. 5 % corresponding drop at residence
1100
Amps
900
700
500
Volts
300
A Irms (val)
B Irms (val)
C Irms (val)
A Vrms (val)
B Vrms (val)
C Vrms (val)
285
280
275
270
265
260
300
kW
250
200
150
100
TOT P(kW) (val)
800
kVAR
700
600
500
400
300
TOT Q(kVAR)
12:22:50
12/17/2007
Monday
12:22:52
Softstart programmed to start
motor @ 330 % FLA
31
12:22:54
12:22:56
Initial 1-Week Pst Profile
32
Initial Goals
• Reduce flicker Pst on utility system to < 1.0.
• Reduction in voltage sags caused by sawmill
operations on utility system to no more than 2.0
percent.
• Mitigation mechanisms must not cause greater
than 2.0 percent voltage rise on utility system.
33
Examples of Products Providing Fast Reactive
Power Compensation
S & C’s AVC (Adaptive Var
Compensator)
34
ABB’s DynaComp (Dynamic-Var
Compensator
More Examples of Products Providing Fast
Reactive Power Compensation
Square-D
Hybrid-Var
Compensator
Square-D
Active Filter
Eaton-CH
Active Filter
35
Case Study #1 – Examples of Motor
Starts with AutoXFMR starters
290
285
Volts
280
275
270
265
Headsaw
(150hp)
w/67%
tap
260
1000
A Vrms (val)
B Vrms (val)
C Vrms (val)
A Irms (val)
B Irms (val)
C Irms (val)
Amps
750
Edger
(150hp)
w/50 &
67% taps
500
250
12:17:55
12/17/2007
Monday
36
12:18:00
12:18:05
12:18:10
Starter Modifications
Set all taps to 50% on all
autotransformer starters.
Programmed 200hp softstart to
start at 250 % of FLA.
After starter modifications, Max
starting Vdrop on primary ~3%
37
Findings of Additional Recording at
the Sawmill
Dran-View 6.5.00 OEM Site License Tennessee Valley Authority
kW
600
400
200
kVAR
0
TOT P(kW) (val)
800
600
400
200
0
TOT Q(kVAR)
5
4
3
2
1
0
B VPst
12:00
05/12/2008
Monday
00:00
05/13/2008
Tuesday
12:00
00:00
05/14/2008
Wednesday
12:00
Fluctuations were more frequent
and of higher magnitude overall due
to type of wood being sawed.
38
00:00
05/15/2008
Thursday
12:00
00:00
05/16/2008
Friday
Case Study #1 – Sawmill on Rural
Feeder
Worst-case
primary
fluctuation =
8%
39
Calculations - Converting 3.9 miles to
14.4/24.9 kV
New worstcase
primary
fluctuation =
2.8%
40
Pst Calculations and Estimated Pst
with Voltage Conversion
Residence
1.75
1.50
Sawmill
1.25
1.00
0.75
0.50
0.25
B VPst
06:00
05/13/2008
Tuesday
08:00
10:00
12:00
14:00
Created with DranView 6.5.0
41
16:00
18:00
20:00
Estimating Future Pst After Voltage
Conversion
• Max Pst at residence for
day shown = 1.34
• Corresponding Pst at
PstMax @ residence
1.34
=
= 0.77
sawmill secondary
PstCorresponding @ sawmill 1.74
metering point = 1.74
• Typical ratio “every other
interval” ~ 0.7-0.75
ZTo _ XFMR _ Sawmill _ Primary
0.1070
=
= 0.7
• Impedance Ratio =
ZTo _ XFMR _ Sawmill _ Secondary
• By converting last 3.76
miles to 14.4/24.9 kV
0.1535
ZTo _ XFMR _ Sawmill _ Primary
ZTo _ XFMR _ Sawmill _ Secondary
With Impedance Ratio = 0.48 PstMax @
Residence Due to Sawmill Load = 0.84
42
=
0.0448
= 0.48
0.0933
Case Study #2a & #2b – Residential
Air-Conditioning Starting
• Customer complaints at their worst have
resulted in service drop/lateral changeout or
possibly even transformer upsizing
• HVAC technicians often install “hard start” kits
• On average, electric consumers understand
that flicker is simply going to occur when their
A/C starts.
• Members sharing secondary conductors and
sometimes transformer may result in one
member seeing another member’s A/C startup.
43
A/C Starting Case Study #2a
Volts
150
50
-50
-150
Volts
AV
120.0
117.5
Before hard
start installed
115.0
112.5
110.0
Amps
150
A Vrms
50
0
-100
-150
AI
Volts
150
19:55:25.3
07/11/2007
Wednesday
19:55:25.4
19:55:25.5
19:55:25.6
19:55:25.7
50
-50
-150
AV
Volts
120.0
117.5
115.0
112.5
110.0
Amps
150
100
After hard
start installed
0
-50
-100
-150
20:09:57.3
07/11/2007
Wednesday
44
A Vrms (val)
AI
20:09:57.4
20:09:57.5
CI
20:09:57.6
20:09:57.7
Results of Using Hard Start Devices
• Hard start results in increased starting torque
• Compressor accelerates to full speed more
quickly, thereby reducing startup duration
• Magnitude of initial inrush current and voltage
drop still the same as before – capacitor
provides phase shift but no voltage rise
45
Minimum Voltage Magnitude Profile and Corresponding
Pst for A/C Startups Case Study
125.0
Volts
122.5
120.0
117.5
115.0
Flicker could be considered
objectionable although Pst
<1.0
A Vrms (min)
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
A VPst
20:00
06/15/2008
Sunday
46
22:00
02:00
00:00
06/16/2008
Monday
AVrms
AVPst
Min
113.3
0.06869
Max
124.8
0.8114
95%
N/A
0.7575
99%
N/A
0.8114
04:00
06:00
A/C Startup Case Study #2b
• Customer complained about flickering lights, number of
interruptions occurring, and damaged electronics.
• Distributor power-quality monitoring revealed nothing
problematic.
• Distributor changed 50 kVA to 75 kVA transformer and
concerns were not resolved.
• Relationship between customer and distributor was
tense and customer wondered why the transformer had
been changed out if no problem could be detected.
• HVAC technician told customer that the A/C unit needed
a “booster” (hard-start kit).
47
A/C Startup Case #2b Power Quality
Monitoring
A/C unit starting current
300
200
100
0
-100
-200
-300
-400
0
0.02
0.04
0.06
0.08
0.1
T ime (sec)
Switchbounce associated with the mercurybulb thermostat resulted in multiple A/C
starts
48
0.12
0.14
Case Study - Raw Water Pumping
Station – Original Design
• Original study began in 2004
– Design incorporated 1 pump driven by 1250 hp
4,160 V AC induction motor
– Provision for additional pump installation
– 4,160 VAC Electronic softstarter
– 2,125 kVAr 2-stage capacitor bank switched
simultaneously on with motor (electromechanical
contactor-switched)
– 1,800 kVAr switched off just as motor about to
reach full speed
49
Original Design – Circuit
Characteristics
• Fed from 161/69/12.47 sub off of 12.47-kV
system
• App. 3.5 miles with 8 line sections of
distribution line
• 12.47 kV to 4,160 V 2,500 kVA padmount
XFMR
• 0.1 miles from riser pole to XFMR, but
distributor wanted PCC to still be at primary
terminals of XFMR
50
Original Objectives and Constraints
• Determine if original design with both electronic softstart
and simultaneously-switched cap bank:
– Result in less than 2.0 % voltage drop at PCC during
motor start?
• Use softstart manufacturer’s stated inrush current
values for calculations
• Perform calculations by hand and/or using steady-state
type analysis program
51
Original Calculations from 2004-2005
2.0 % limit exceeded but distributor willing to go to 3.0 % with
understanding that pump not started more than once per week
52
2006 Design Changes
• Latest starter mfr/consultant analysis indicated voltage
drop will exceed 4.0 percent at PCC during startup.
• New design called for 2 pumps operating
simultaneously
• Increased runtime/# of startups
• Customer contemplating generator because distributor
does not want to allow more than 3% fluctuation under
any circumstances
53
s c o pe
C a d iz Su b Vo lta g e
Pr ima r y _ C u r r e n t_ C
rms
s c o pe
out
tr _ 8
in
out
s c o pe
tr _ 7
out
i(t)
+
out
p3
in
S
f(u)
Fm6
?v
1
s c p10
s c o pe
Pr im a r y Vo lta g e a tTr a n s fo r me r
s c ope
p4
v(t)
out
tr _ 1 2
ASM 1
?i
?m
PF
4 .1 6 0 k V s c ope
1250hp
Fm 7
SW 1
7 |1 E1 5 |0
1 2 .4 7 /4 .1 6
ia
Se c Vo ltD r o p a tTr a n s fo r me r
s c o pe
AS M
S
+
+
?v
1 6 1 k VR M SL L /_ 0
Tm
Sp e e d
Yg Yg _ n p 2
1
2
AC 1
s c p3
s c op e
rms
in
?v
P
1
Q
2
BU S2
rms
in
tr _ 4
f(u)
s c o pe
Mo to r N o 2 R MSC u r r e n t
Fm4
s c ope
Se c o n d a r y Vo lta tTr a n s fo r m e r
v(t)
s c o pe
s c p8
rms
Fm 2
in
in
tr _ 2
out
tr _ 1
+
scope
sc p4
sc p6
scope
c SW 2
+
C2
276uF
rms
sc p5
scope
+
C1
?i
+
276uF
Mo to r N o 2 Sp e e d
out
rms
c
b
a
BU S2
c SW 1
+
f(u)
s c ope
Fm 3
f(u)
f(u)
1
s c op e
s c p9
p1
Fm 1
out
f(u)
a
Fm
BU S1
1
?
+
1
YgYgD_np1
161/12.47/69
T
s c ope
3
2
3
2
out
tr _ 1 1
D EV4
Pr im a r y Vo ltD r o p a tTr a n s fo r me r
1
out
tr _ 9
rms
?v
1
s c ope
f(u)
Pe r c e n tVd r o p a tC a d iz Su b s ta tio n
rms tr _ 1 0
in
?
s c ope
s c p17
rms
in
s c o pe
tr _ 6
+
•Analysis of across-theline start with and
without simultaneous
cap switching
Pr ima r y _ C u r r e n t_ A
rms
in
YgYgD_np2
161/12.47/69
•EMTP used to simulate
motor start using
dedicated feed from
substation
Fm5
1
p2
Pr ima r y _ C u r r e n t_ B
rms
s c ope
s c p18
v(t)
in
in
out
tr _ 3
c SW 3
?i
+
BU S2
276uF
Fm9
C3
+
?i
f(u)
4 5 .9 8 u F
+
c SW 4
+
C4
4 5 .9 8 u F
?i
+
c SW 6
+
C5
?i
c SW 5
C8
ASM2
Tm_ 1
4 5 .9 8 u F
54
C6
?i
10000
Sp e e d 2
s c ope
Sp e e d
c
ia
S
Mo to r N o 1 R MSC u r r e n t
AS M
+
4 5 .9 8 u F
C7
?i
+
rms
?m
4 .1 6 0 k V
1250hp
in
s c o pe
out
tr _ 5
With Cap Bank Switched - Results
Max 1.4 % drop during start
at substation with caps
55
Without 1,800 kVAr Cap Bank
Switched - Results
Max 1.7 % drop during start
at substation without caps
56
Case Study #4 – Automotive Component
Manufacturer with Welders (Reactive Power)
Estimated 400-kVAr
swings per phase
based on 1-minute
max/avg/min data
57
Case Study #4 – Automotive Component
Manufacturer with Welders (Real Power)
Estimated 200-kW
swings per phase
based on 1-minute
max/avg/min data
58
Case Study #4 – Automotive Component
Manufacturer with Welders (Voltage)
3 to 4.7-volt swings
based on 1-minute
max/avg/min data
59
Case Study #4 – Corresponding Pst on
Bus Feeding Welders (Dirty Feed)
Due to voltage sag
60
Generally, Max Pst ~ 1.5
Case Study #4 – Corresponding Pst on Bus Fed By
Separate Transformer Bank (Clean Feed)
Generally, Max Pst ~ 0.5
61
Vdrop Calculation and Comparison of Welder Bus
Pst to That of Separately Fed Bus
•
•
•
•
•
Max Pst on welding bus = 1.5
Max Pst on separately fed bus = 0.5
Utility-supplied SCC = 4,127 A @ 7.2/12.47 kV
Welding bus XFMR %Z = 6 %
Assume X/R of utility circuit = 7 and X/R of XFMR = 12
7200
× (cos(tan −1 (7)) + j sin(tan −1 (7))) = 0.25 + 1.73 ohms
4127
0.25 + j1.73
On 100 MVA base, Zsource =
= 16 + j111.3 percent
12470 2
108
Transformer Z (on 100 - MVA base) = 20 + j239 percent
0.6 × (16 + 20) + 1.2 × (111.3 + 239)
= 4.4%
100
Separately Fed Bus Pst 0.5
=
= 0 .3
Welding Bus Pst
1.5
Expected %Voltage Drop =
16 + j111.3 111
Common Source Z
=
≈
≈ 0.32
Total Z to Welding Bus
36 + j350
350
62
Case 5 – Wind Generation
63
Buffalo Mountain Wind Farm Generation Flicker Case
TVA originally contacted for three small windmills to be fed from a
local distributor’s 13-kV system. Occasionally during windmill
startup, the distributor received flicker complaints. When the 151.8-MW units went in service, it was decided to serve the new
load at 161-kV.
64
Aggregate Load Swings for 15-1.8-MW Wind
Generators
Voltage
MW
MVAR
65
Case 6 – Large Motor System
Evaluate Possibility of
Serving a 3000-hp Wound Rotor Motor
Metal Shredder
At Two Possible Locations
–
–
66
XXX 69-kV: Z1 = 23.989 + j 52.618%
YYY 161-kV: Z1 = 1.559 + j 9.048%
Characteristics of Shredder
Provided by Vendor
3 times per minute
Shred Load
Cycling
3.1-MVA swings at
78%p.f.
2.4-MW swing
1.9-MVAR swing
67
Hand Calculation Used for Approximation and EMTP
Model Used for More Exact Solution
Site XXX Hand Calculation:
Z1 = 23.99 + j 56.6 % (100-MVA base)
MVA swing = 2.4 + j1.9
%VDrop ≅ 2.4-MW (23.99)/100 + 1.9-MVAR (56.6)/100 = 1.65%
sum1
v(t)
+
+
in
p3
1
Percent_V
scope
f(u)
Use EMTP
Modeling
For More
Exact
Solution
DYg_1
AC1
+
Fm1
DEV6
out
TACS type-66
rms meter
-
+
RL1
1
69kVRMSLL /_0
2
69/4.16
3/7
C4
tmr1
c
Timer
2
C2
1
0.25
1.0
select
?s
Sel1
scope
IArms_Uncorrected
DEV5
out
C3
TACS type-66
rms meter
in
c
IArms_Corrected
DEV3
out
c
in
2
scope
TACS type-66
rms meter
Tm
Teg
Speed
ASM1
SW1
+
500ms| 1E15| 0
i(t)
DEV1
p1
s3
i(t)
s4
Gain2
p2
S
ASM
R
N
MW_MVAR_MW_METER
9.549297
4.16kV
3000hp
+
+
184uF
2.5
C1
+
0.0e+0m
68
Torque
scope
R2
R1
RPM
scope
EMTP Simulation Results
Voltage Drops As Shredder Cycles
YYY – 161-kV
0.3%
XXX 69-kV
1.75%
69
Applying Results to TVA Flicker Curve Shows YYY is
OK and XXX is Not Recommended
TVA
XXX 69-kV – Above Curve – Ruled Out
YYY 161- Below Curve - OK
70
Case 7
My Town Needs a Steel Rolling Mill
What do we need to do to allow us to run
this mill?
Remote Community Fed at 46-Kv – Impedance
to Substation - Z1 = 74.9 +j113.5%
At 26-kV Substation Transformer Secondary
(PCC) - Z1=76.3 +j169.0%
71
Company Choose IntelliVar as Mitigation
System
Not actual
Installation
But Shows
All Major
Components
Transformers
Valves
Filters
Controls
72
IntelliVAR System at 1500-KVAR
Controls Not Working Properly
1.8
1.6
26kV - Pst A
26kV - P Total
from 06/13/2007 to 06/15/2007
from 06/13/2007 to 06/15/2007
PST(95) – 1.47
5000
Min[P Total] (kW)
Avg[P Total] (kW)
Max[P Total] (kW)
4500
1.4
4000
1.2
3500
3000
0.8
2500
Pst A
1.0
2000
0.6
1500
0.4
1000
0.2
0.0
13 Wed
Jun 2007
EPRI/Electrotek
1.8
1.6
500
14 Thu
Time
46kV - Pst A
15 Fri
0
13 Wed
Jun 2007
PQView® EPRI/Electrotek
14 Thu
Time
15 Fri
PQView®
26kV - Q Fund All
from 06/13/2007 to 06/15/2007
from 06/13/2007 to 06/15/2007
PST(95) – 1.17
1.4
4000
Min[Q Fund All] (kvar)
Avg[Q Fund All] (kvar)
Max[Q Fund All] (kvar)
3500
3000
1.2
2500
2000
0.8
1500
Pst A
1.0
1000
0.6
500
0.4
0
0.2
0.0
13 Wed
Jun 2007
EPRI/Electrotek
73
-500
-1000
14 Thu
Time
13 Wed
Jun 2007
PQView® EPRI/Electrotek
15 Fri
14 Thu
Time
15 Fri
PQView®
IntelliVAR Sizing Doubled to 3000-KVAR
Control System Finally Tuned
26kV - P Total
26kV - Pst A
1.8
1.6
from 04/24/2008 to 04/26/2008
from 04/24/2008 to 04/26/2008
5000
26-kV - PST(95) - 1.18
Min[P Total] (kW)
Avg[P Total] (kW)
Max[P Total] (kW)
4500
1.4
4000
1.2
3500
3000
0.8
2500
Pst A
1.0
2000
0.6
1500
0.4
1000
0.2
0.0
24 Thu
Apr 2008
EPRI/Electrotek
500
25 Fri
Time
26 Sat
0
24 Thu
Apr 2008
PQView® EPRI/Electrotek
46kV - Pst A
1.8
1.6
1.4
25 Fri
Time
26 Sat
PQView®
26kV - Q Fund All
from 04/24/2008 to 04/26/2008
from 04/24/2008 to 04/26/2008
1000
46-kV - PST(95) – 0.82
Min[Q Fund All] (kvar)
Avg[Q Fund All] (kvar)
Max[Q Fund All] (kvar)
500
0
1.2
-500
-1000
0.8
-1500
Pst A
1.0
-2000
0.6
-2500
0.4
-3000
0.2
0.0
24 Thu
Apr 2008
EPRI/Electrotek
74
-3500
25 Fri
Time
26 Sat
-4000
24 Thu
Apr 2008
PQView® EPRI/Electrotek
25 Fri
Time
26 Sat
PQView®
Before and After Intellivar Changes
Kosciusko - Attala Steel 26kV Kosciusko - V RMS A
Kosciusko - Attala Steel 26kV Kosciusko - V RMS A
from 06/13/2007 to 06/15/2007
Min[V RMS A] (V)
Avg[V RMS A] (V)
from 04/24/2008 to 04/26/2008
Max[V RMS A] (V)
15400
15200
15200
15000
15000
14800
14800
14600
14600
14400
14400
14200
14200
14000
14000
13800
13800
13600
13 Wed
Jun 2007
EPRI/Electrotek
1.8
26kV - Pst A
15 Fri
24 Thu
Apr 2008
PQView® EPRI/Electrotek
25 Fri
Time
26 Sat
PQView®
from 06/13/2007 to 06/15/2007
from 04/24/2008 to 04/26/2008
1.8
1.4
1.4
1.2
1.2
1.0
1.0
PST(95) - 1.18 – Not Under 1.0
But No Routine Complaints
Pst A
Pst A
1.6
0.8
0.6
0.6
0.4
0.4
0.2
0.2
13 Wed
Jun 2007
EPRI/Electrotek
Max[V RMS A] (V)
26kV - Pst A
1.6
0.0
Avg[V RMS A] (V)
13600
14 Thu
Time
0.8
75
Min[V RMS A] (V)
15400
14 Thu
Time
15 Fri
PQView®
0.0
24 Thu
Apr 2008
EPRI/Electrotek
25 Fri
Time
26 Sat
PQView®
Case 8
My Largest Industry Has Us in a Bad
Situation Due to Flicker, What Are Our
Options?
76
Distributor Not Routinely Serving Any Customer From
46-kV Winding Due to Flicker Complaints
Other Customers
Normally Fed
From 46-kV System
Distributor Removed
Them Due to Flicker
Complaints
YgYgD_np1
2
1
3
161/46/13
+
AC1
161kVRMSLL /_0
+
+
RL1
+
+
RL3
RL2
+
RL4
PI1
T Equivalent
DEV1
PCC
Is Located
Within
Transformer
At T Equivalent
Tie Point
77
DEV2
DEV3
Arc Furnace Arc Furnace Arc Furnace
Site Fed From
Dedicated
13-kV Winding
161/46/13
Transformer
PST Estimate – Traditional Hand Calculation
161-kV MVAsc at Sub – 1558 MVAsc or 3.85% at 60 MVA Base
161-kV:46-kV:13:kV Transformer Nameplate:
Z161-46 = 9.7% at 60-MVA,
Z161-13 = 5.8% at 21-MVA (16.57% at 60-MVA)
Z46-13 = 2.1% at 21-MVA (6.0% at 60-MVA)
Convert to T equivalent
Z161 = ½ (9.7% +16.57%-6.0%) = 10.135% at 60 MVA
Z46 = ½ (9.7% + 6.0% -16.57%) = -0.435% at 60 MVA
Z13 = ½ (16.57%+6.0% -9.7%) = 6.435% at 60-MVA
PCC Includes TVA plus 161-kV portion of winding to 46/13-kV T point
PCC = 3.85% + 10.135% = 13.985% = 429-MVAsc
Estimated AF Swing = 2 x 4-MW = 8-MVA
Swing with 3 AF = 3 3 x 8-MVA = 11.54-MVAsc for 3 units operating
46-kV Short Circuit Voltage Depression (SCVD) = 11.54/429 = 0.0269
161-kV Short Circuit Voltage Depression (SCVD) = 11.54/1558 = 0.0074
46-kV PST Estimate = 0.0269 x KST = 0.0269 x 60 = 1.61
161-kV PST Estimate = 0.0074 x KST = 0.0074 x 60 = 0.44
78
PST Measurements of 46-kV and 161-kV Bus
Confirm Hand Calculations - Target – PST Under 1.0
Moving to 161-kV Accomplishes This Objective
Timeplot
2.00
PST Measurements at 161-kV
1.75
1.50
161-kV PST
(a phase)
161-kV Flicker
0.44 per hand calc.
1.25
1.00
0.75
0.50
0.25
0.00
A FDPst
06/12/2007
2.00
06/14/2007
06/16/2007
46-kV Flicker
1.61 per hand calc.
06/18/2007
06/20/2007
06/22/2007
06/24/2007
06/26/2007
06/28/2007
Event #25 at 06/25/2007 08:10:30.067
AVrms RMS and Waveform
PSTTimeplot
Measurements at 46-kV - PCC
1.75
1.50
1.25
46-kV PST
(a phase)
1.00
0.75
0.50
0.25
0.00
A FDPst
06/12/2007
06/14/2007
06/16/2007
06/18/2007
06/20/2007
Event #1 at 05/23/2007 14:04:48.349
AVrms RMS and Waveform
79
06/22/2007
06/24/2007
06/26/2007
06/28/2007
Option 1 – Add Transformer at Substation
Use Existing 13-kV Feed to Plant
Remove Plant
Load From
Existing 3-winding
Transformer
YgYgD_np1
2
1
3
161/46/13
+
AC1
161kVRMSLL /_0
+
+
+
RL1
+
Other Customers Fed
From 46-kV System
RL3
RL2
+
RL4
PI1
T Equivalent
DEV1
New
PCC
At
161-kV
80
DEV2
DEV3
DYg_1
1
2
Arc Furnace Arc Furnace Arc Furnace
161/13
Dedicated 161:13-kV Existing 13-kV Line
Transformer at Sub
Move PCC Up to 161-kV
Option 1 - Install Dedicated Transformer
Option
1
Move PCC up to 161-kV System at Substation
All cost below in $1000s
Install
Total
1a
New 16-MVA Transformer, oil containment, foundations, bus
$850
$150
$1,000
1b
161-kV SF-6 breaker w/ bay and relays
$250
$150
$400
1c
13-kV Distribution Breaker with relays
$50
$25
$75
1d
Routing Existing 13-kV Wiring to New System
$10
$10
$20
1e
Two-Step - 4.5 (4.0 effective) -MVAR Harmonic Filter
$100
$23
$123
Misc. cost not currently identified - 20% of cost (excluding
transformer)
Grand Total - New Transformer System
81
Equip.
$124
$1,742
Option 2
Build 161-kV Line to Plant Site and Add
Transformer at That Site
Remove Plant
Load From
Existing 3-winding
Transformer
YgYgD_np1
2
1
3
161/46/13
+
AC1
161kVRMSLL /_0
+
+
+
RL1
+
RL3
RL2
PI1
T Equivalent
Other Customers Fed
From 46-kV System
DYg_1
1
2
161/13
DEV1
New
PCC
At
161-kV
82
DEV2
DEV3
+
PI2
New 161-kV Line
To Plant Site
Arc Furnace Arc Furnace Arc Furnace
Dedicated 161:13kV Transformer at
Plant
Move PCC to 161-kV
Option 2 - Install Dedicated 161-13-kv Substation at
Plant Site
Option
2
Move PCC Up to 161-kV System at Plant
All cost below in $1000s
Install
Total
2a
New 16-MVA Transformer, oil containment, foundations, bus
$850
$75
$925
2a
Used 18-MVA Transformer, oil containment, foundations, bus
$350
$75
$425
2b
161-kV SF-6 breaker w/ bay and relays
$250
$75
$325
2c
161-kV Feed to Plant (existing row - big unknown ??)
2d
161-kV Circuit Switcher at Plant with relays
2e
13-kV Distribution Breaker at Plant with relays
2f
Two-Step - 4.5 (4.0 effective) -MVAR Harmonic Filter
Misc. cost not currently identified - 20% of cost (excluding
transformer)
83
Equip.
$400
$100
$25
$125
$50
$25
$75
$100
$23
$123
$210
Grand Total - New Transformer System
$2,183
Grand Total - Used Transformer System
$1,683
Option 3 - Install Quick, Dynamic VAR Correction and
Harmonic Filter System at Plant Site
Add Dynamic Var
Correction at Plant
Site
1x
YgYgD_np1
+
AC1
161kVRMSLL /_0
+
+
RL1
+
+
RL3
RL2
+
SVC_1
2
3x
3
161/46/13
Other
Customers
Fed
From 46kV System
1
2
1
RL4
PI1
T Equivalent
DEV1
No Transmission or
Distribution System
Changes Needed!
84
DEV2
DEV3
Arc Furnace Arc Furnace Arc Furnace
Fix Flicker at Plant 13-kV Bus
Option 3 - Install Quick, Dynamic VAR Correction and
Harmonic Filter System at Plant Site
Option
3
Install ABB Mincomp or S&C Purewave Dstatcom
All cost below in $1000s
Install
Total
$1,000
$100
$1,100
3a
Install +10 /- 2-MVAR Var Compensator
3b
Two Step - 6.6/6.3 (6.0 effective)-MVAR Harmonic Filter
$120
$25
$145
3c
13-kV Switchgear Integrated Into Existing System
$100
$25
$125
Misc. cost not currently identified - 10% of cost
Grand Total
85
Equip.
$125
$1,370
Economic Summary:
Installing Dynamic Var Compensation is Best Choice
Option
#
Power
Factor
Correction
Savings
(1000s)
Electronic
Equip.
Savings
(1000s)
Production
Profit
Increase
(1000s)
Total
Savings
Total
Cost
(1000s)
(1000s)
Option
1
$32.8
$31.6
$0
$64.4
$1,742
27.0
Option
2
$32.8
$31.6
$0
$64.4
$2,183
33.9
Option
3
$32.8
$31.6
$415.0
$479.4
$1,370
2.9
Option 1 – Transformer Isolation at Substation, filter at Plant
Option 2 – Transformer Isolation at Plant, filter at Plant
Option 3 – Install Filter/Dynamic VAR Compensation at Plant
86
Simple
Payback
(years)
Case 9
A Large Steel Mill Has an SVC, Why
is it Creating More Flicker When On
Compared to When its Off?
DEV2
s3
Super_Voltmeter1
PCC
+
AC1
161kVRMSLL /_0
RL2
+
0.285,4.004Ohm
RL1
+
DY_1
1
DD_1
2
1
2
0.804,2.3585Ohm
161/33
33/0.7
DEV1
+
FLICKERMETER_PCC
IEC
FLickermeter
87
10
R1
Arc Furnace
Base Case Without SVC System Operating
2 Days
Phase A Voltage
(Max, Min, Avg)
161PCC - Pst A
from 10/29/2004 to 10/31/2004
1.4
Flicker Summary
PCC Monitor Site
Without SVC/ Filters
PST(95) = 0.93
1.2
Pst A
1.0
0.8
0.6
0.4
0.2
0.0
29 Fri
Oct 2004
EPRI/Electrotek
88
30 Sat
Time
31 Sun
PQView®
Base Case With SVC System Operating
2 Days
Phase A Voltage
(Max, Min, Avg)
161PCC - Pst A
from 10/26/2004 to 10/28/2004
Flicker Summary
PCC Monitor Site
Without SVC/ Filters
PST(95) = 1.22
1.4
1.2
Pst A
1.0
0.8
0.6
0.4
0.2
0.0
26 Tue
Oct 2004
EPRI/Electrotek
89
27 Wed
Time
28 Thu
PQView®
Flicker Reduced by Approximately 24% by
Switching Off SVC/Filters
What is the reason for lower flicker levels
with the SVC out of service?
Without SVC
With SVC
161PCC - Pst A
161PCC - Pst A
from 10/29/2004 to 10/31/2004
from 10/26/2004 to 10/28/2004
1.2
1.2
1.0
1.0
0.8
0.8
Pst A
1.4
Pst A
1.4
0.6
0.6
0.4
0.4
0.2
0.2
0.0
29 Fri
Oct 2004
EPRI/Electrotek
30 Sat
Time
PST (95) – 0.93
90
31 Sun
0.0
26 Tue
Oct 2004
PQView® EPRI/Electrotek
27 Wed
Time
28 Thu
PQView®
PST (95) – 1.22
Base Case With SVC System Operating Weakened TVA System - 2 Days
Complaints Ramped Up During This Period
161PCC - Pst A
from 02/09/2008 to 02/11/2008
2.4
2.2
Flicker Summary
PCC Monitor Site
Without SVC/ Filters
PST(95) = 1.86
2.0
1.8
1.6
Pst A
1.4
1.2
1.0
0.8
0.6
0.4
0.2
9 Sat
Feb 2008
EPRI/Electrotek
91
10 Sun
Time
11 Mon
PQView®
Comparison of Voltage Swings at PCC Both
With and Without SVC Operation
No SVC
SVC
Controls
Overshoot –
One
Possibility
For
Flicker Swing
92
GE Flicker Curves
Copied From IEEE 519
Voltage Swing Possible By TCR Overshoot = 1.1%
93
Case 10
A Large Steel Mill Has an SVC and DC
Electric Arc Furnace, Here is The Way
Things Should Operate!
94
EAF Operation with Great SVC Control Performance
161PCC- Pst A
from 06/03/2008 to 06/04/2008
0.35
0.30
PST
Pst A
0.25
0.20
0.15
0.10
0.05
3 Tue
3AM
6AM
9AM
Jun 2008
EPRI/Electrotek
12PM
Time
3PM
6PM
9PM
4 Wed
161PCC- V RMS A
PQView®
from 06/03/2008 to 06/04/2008
Min[V RMS A] (V)
97000
Avg[V RMS A] (V)
Max[V RMS A] (V)
96500
96000
95500
VOLTAGE
95000
94500
94000
93500
93000
92500
92000
3 Tue
3AM
6AM
9AM
Jun 2008
12PM
Time
3PM
6PM
9PM
4 Wed
161PCC- Q Fund All
EPRI/Electrotek
PQView®
from 06/03/2008 to 06/04/2008
Min[Q Fund All] (kvar)
30000
Avg[Q Fund All] (kvar)
Max[Q Fund All] (kvar)
25000
20000
15000
10000
REACTIVE
POWER
5000
0
-5000
-10000
-15000
-20000
-25000
-30000
Jun 2008
EPRI/Electrotek
95
3 Tue
3AM
6AM
9AM
12PM
Time
3PM
6PM
9PM
4 Wed
PQView®
Questions/Comments
If not, then lets eat!!
96