HomeMy WebLinkAboutKotzebue Wind Farm Expansion Project Power Quality Investigation Report - May 1999 - REF Grant 21954271
u
POWER QUALITY INVESTIGATIONS
FOR
MULTIPLE WIND TURBINES WITH INDUCTION GENERATORS
Prepared for:
KOTZEBUE ELECTRIC ASSOCIATION
Kotzebue, Alaska
In cooperation with:
The National Rural Electric Cooperative Association
Cooperative Research Network
By:
Malcolm A. Lodge. P., Eng.
Island Technologies Incorporated
Charlottetown, P.E.I., Canada
Jie Han, P., Eng.
Maritime Electric Co Ltd.
Charlottetown, P.E.I. Canada
In Association with:
Craig Thompson, P., Eng.
Thompson Engineering Co., Inc.
Anchorage, Alaska
May 25, 1999
Ll
C
POWER QUALITY INVESTIGATIONS
FOR
MULTIPLE WIND TURBINES WITH INDUCTION GENERATORS
TABLE OF CONTENTS
1.0 INTRODUCTION.......................................................1
2.0 INTERCONNECTION ISSUES .............. I ............... I .... I ....... 13
3.0 POWER QUALITY ...................................................... 3
4.0 KEA SYSTEM..........................................................5
5.0 STUDY METHOD ...................................................... 8
6.0 MODELLING.........................................................10
1. Assumptions.....................................................10
2. Parameter Preparation ............................................. 11
3. Computer Modelling .............................................. 11
4. Wind Turbine Model .............................................. 11
5, Load Flow Cases.................................................12
6. Load Flow Results and Conclusions .................................. 12
7. Voltage Profile ................................................... 14
7.0 RESULTS OF INVESTIGATIONS ........................................ 17
1. Analytical Summary .............................................. 17
8.0 MEASUREMENT PROGRAM ........................................... 18
1. Measurement Procedure ............................................ 18
2. Measurement Results .............................................. 19
9.0 CONCLUSION........................................................20
FIGURES5 - 14 ..............................................................21
Appendix 1, Load Calculation for Feeder #4...................................... 32
Appendix 2, Impedance Calculation for Feeder #4................................. 36
Appendix 3, Wind Turbine Generator Model ..................................... 38
w POWER QUALITY INVESTIGATIONS
FOR
MULTIPLE WIND TURBINES WITH INDUCTION GENERATORS
1 1.0 INTRODUCTION
KEA) is a rural electric
Kotzebue Electric Association, ( cooperative located in the city of
Kotzebue Alaska. KEA provides electric energy to about 3,000 residents and owns and operates
a diesel electric power plant and electric distribution system with about 1000 customers.
Kotzebue is located just north of the Arctic Circle on the tip of a peninsula in Kotzebue Sound, a
large bay along the coast of the Bering Strait. The plant has six diesel engine -generators with a
combined capacity of 11,000 kW. Annual peak load and minimum load on the plant are about
4,000 kW and 1800 kW respectively. The plant is manually operated and dispatched.
All fuel and most other supplies and equipment necessary for operation of the diesel plant
are imported by ocean going barge from the south. The remote location of the community and
the Arctic climate makes transportation of fuel expensive. Kotzebue Sound is frozen for nine
months and typically only open to shipping from July to September. The result of these
conditions is a relatively high cost to produce electricity as compared to that for other larger
communities in the south who benefit from connection to hydro electric generation and larger
electric grids with lower production costs. To enable residents and consumers in Kotzebue and
other comparable communities to afford reliable electricity, the State of Alaska has for several
years funded a Power Cost Equalization (PCE) program since 1983 which provides financial
assistance to qualifying utilities to offset high operating costs. Since the cost of this program
increases as the communities grow and the State of Alaska would like to reduce the burden of the
PCE program on their budget, they have offered to assist, promote and sponsor the development
of wind energy as an alternative and a supplement to diesel generated electricity. They have been
joined in this effort by the US Government though the Department of Energy (DOE), the
National Renewable Energy Laboratory (NREL) and The National Rural Electric Cooperative
Association (NRECA). Wind energy was recognized during early investigations of renewable
1
F',
0
1
I
. I
energy options for rural Alaska as the most appropriate renewable energy alternative for coastal
areas at high latitudes and where small hydro power is not practical.
KEA expressed an interest in becoming active in the testing and evaluation of wind
energy in their system and were selected as the key location for the development of a relatively
large, multiple wind turbine project intended to lead the way for wider spread use of wind energy
in rural Alaska. The proposed scale of the project was to install up to 2,000 kW of wind power
by about year 2000.
In 1997, KEA obtained three AOC 15150 production prototype wind turbines
manufactured by Atlantic Orient Corporation (AOC) of Norwich Vermont. This equipment was
selected because of its physical size and rating and certain other design features. Although rated
at 50 kW the wind turbine will produce average power levels of 66 kW or more in cold dense air
conditions such as found in Arctic climates. AOC is the only US manufacturer of wind turbines
of this type and size. KEA has future plans to increase the capacity of their wind generation to
500 kW in 1998 by adding seven AOC 15/50 wind turbines and then later to between 1,000 and
2,000 kW.
Many utilities consider wind turbines as non -conventional means of generating electricity
due to their dependence on the vagaries of the wind for prime mover power and the use of simple
induction generators for producing electricity. For these reasons they are often concerned that
the reliability and quality of power in their power system may be adversely affected by wind
generation. This concern increases as the ratio of the capacity of wind generation to conventional
generation increases. For this reason KEA and their sponsors have chosen to investigate power
quality issues associated with the existing project and future increases in capacity on the KEA
system.
The investigation is intended to predict and measure the power quality effect or impact of
the first three wind turbines and to predict the effect of wind turbines added later. The
investigation was not to consider how the plant is operated and dispatched or the effect the
1
2
1
1
A
operation of additional wind turbines will have on the operation and control of the diesel
generators. This will become an issue as wind capacity increases and will require further
investigation and possible revision of current plant operations methods.
2.0 INTERCONNECTION ISSUES
An electric utility's concerns, when considering a new or alternative power generation
sources on their system, are generally to insure that the new source will not cause or contribute to
a reduction of standards for safety, reliability, quality or economics of their operations. To this
end utilities who have experienced a demand for the interconnection of such sources have
developed regulations and codes of practice to be followed. In the area of power quality the
requirements are generally that the system must be shown by prediction and/or by demonstration
not to adversely effect or contribute to a reduction in the quality of power on the system or to
adversely effect performance, operation, safety or economics of operations.
3.0 POWER QUALITY
The quality of power on a power system is described with a few basic parameters. The
most often used are related to variations from the desired values of average and instantaneous
voltage, frequency, waveform harmonic content (distortion) and power factor. The degree to
which these contribute to power quality concerns are determined by their effect on utility
operation and equipment and on their customers and their equipment.
Voltage variation from the desired level can occur both as a more of less steady variation
which is generally termed as an under or over voltage condition or as a short term transient
variation. The former is usually the result of voltage drops in transmission and distribution lines
and equipment caused by load variations. The latter is usually the result of an event such as the
starting of a large motor or some other sudden placement of a relatively large load on the system
to which the generation and distribution system is unable to respond. Variations in voltage level,
dependent upon their magnitude, duration and frequency of occurance may cause such
1
3
I
0
I
E
annoyances as dimming or flickering of lights and poor performance of certain sensitive
equipment. At one time variations in voltage were a common source of complaint by the owners
of early television receivers which did not have internally regulated power supplies. Modern
electronic appliances are more tolerant of voltage fluctuations than their predecessors. Extreme
and persistent variations of 20% or more can cause equipment damage and overheating of
motors.
Variations of frequency in a utility system are unusual and typically occur for relatively
short periods. The length of the period is usually determined by the response time of the speed
governor and control on the prime mover or engine. Results of frequency variations are the
variation in speed of motors and inaccuracy of clocks and other devices dependent on a fixed and
accurate utility frequency. As long as utility control systems and customer equipment are not
adversely effected, frequency variations are seldom noticed.
Harmonic distortion in a power system results from the saturation of magnetic devices
and components in generators, transformers, motors and fluorescent lighting fixture ballasts. It
also results from the use of certain types of power supplies, electric welders, lamp dimmers and
electronic power converters and inverters. Excessive harmonic distortion can cause undesired
heating in motors and transformers, over voltage in unregulated power supplies and interference
with telephone and other communication systems.
Power factor in a utility system is a measure of the effectiveness of the generation and
distribution system to supply the load or power demand with the least amount of current
necessary. Loads such as low quality fluorescent lighting ballasts and lightly (under) loaded
induction motors contribute to a poor power factor in a power system. Low power factor requires
the generators and conductors in the system to carry more current than should be necessary and
can result in undesirable voltage variations between various points in the system. Low power
factor also contributes to higher distribution system losses.
S
�I
IThe power quality in an electric utility system and the extent to which it will be effected
by changes and additions of generation (or load) are dependent on the characteristics of the utility
generators, generator controls and distribution system. The latter includes the distribution system
conductors, transformers and characteristics of the consumer load on the system.
Since the KEA wind project will be implemented in stages the effect on power quality,
whether perceptible or significant, is expected to be incremental.
4.0 KEA SYSTEM
The KEA power plant arrangement and single line diagram is illustrated in Figure 1. The
plant has been recently renovated. The name plate capacity of 11, 000 kW compared to the
community peak load of4,000 kW provides a relatively high reserve capacity ratio. This is
common in remote Arctic communities such as Kotzebue where the plant and distribution system
is not connected to an electric grid having other generation resources. In such systems a
breakdown of the largest engine may require several weeks to obtain a replacement or to make
Iextensive repairs.
The KEA distribution system has four principal feeders which service districts in the
community. The wind turbines are located at the extreme end of feeder No. 4. as shown if Figure
2. This is a radial feeder supplying the airport complex and a few other small customers. The
feeder was constructed originally to supply a Iarge military defence radar station which has since
been modernized and reduced in size. The nearest customer to the wind turbine site is a
commercial radio station transmitter located about 1/4 mile away. The section numbers on the
distribution line are for identification of the line characteristics in the utility distribution system
data.
r
9 7.0 RESULTS OF INVESTIGATIONS
0 1. Analytical Summary
0 The investigations carried out were comprised of a series of calculations or estimates of the
of the power quality effects of starting, operating and stopping the wind turbines. The effect of most
interest is the variation in voltage which occurs in the utility system. In the case of the KEA wind
turbines, the condition of the voltage at points along the distribution line between the wind turbine
site and the plant bus is of most interest. To determine this condition the impedance of each section
of the distribution line and the load on each section along the line are used in a calculation to
determine the voltage at each point. The result of this calculation is a profile of the voltage regulation
along the line which can be used to identify where there may be problems. Problem areas are
indicated by either an elevation or depression of the voltage above or below the desired level. A
range of +1- 5% is often used as the acceptable deviation from the desired nominal voltage.
The voltage regulation for the extreme cases of full output operation of 1, 3, 6, 12 and 24
wind turbines on the feeder 44 is shown in Figure 3. When compared to the base case of no wind
turbines in operation, it can be seen that the calculated voltage deviation at any point along the line
is very slight and amounts to less than 2.0 Volts on a 120 Volt base. This result indicates that even
with a considerable VAR requirement for the wind turbines, the KEA distribution system voltage
is well regulated. This condition is a measure of the quality of the line. Examination of the profiles
for each case indicates that the voltage at points near the plant tends to fall slightly with increasing
wind turbine penetration while the voltage at the wind site tends to rise. Comparison of Figures 3
and 4 illustrates that the voltage regulation improves as the wind turbine output increases. The power
introduced by the wind turbines can be seen to compensate somewhat for the increased negative or
lagging VARS contributed by the wind turbines.
It should be realized that the wind turbines do not always run at full output and that VAR
requirements become less in magnitude but higher proportionately as the wind power decreases in
low winds. This will tend to cause a reduction in voltage at the extreme end of the distribution
17
system as the capacity of wind turbine generation increases. The results for these calculations
indicate that the worst case is a reduction of voltage of only 2.2 volts or about 2%. Table 3 lists the
voltage regulation values for the base case of no wind power and for the wind turbine groups at
100%, 0% and 40% load.
It should also be realized that the voltage profile in a system to which wind turbine are
connected is also influenced by the load on the line contributed by the other customers. Also these
calculations assume an infinite bus condition at the plant. This means that load on the plant feeder
being analysed does not cause the plant bus voltage to change. This is a simplification that is usually
done otherwise the response characteristics of the engine governors and voltage regulators must be
considered. This would increase the complexity of the calculations considerably.
8.0 MEASUREMENT PROGRAM
l . Measurement Procedure
Measurements to compare with the analysis were made using an Metrosonics model pa-7
Power Analyser/Recorder with a pq-2 Enhanced Power Quality Option and expanded solid state
memory. The equipment uses three voltage and four current channels to make measurements. When
connected to utility power circuits the pa-7/pq-2 enables the measurement of KW, KVAR, KVA,
Power Factor, Waveform Harmonics and also enables the capture of waveforms at up to 128 samples
per cycle. This enables the actual observation of in -cycle events which result from switching
transients. An event trigger facility enables the capture and recording of abnormal events. A pre -
trigger memory enables observation of the cycle immediately preceding a triggering event. This
makes it possible to actually see the complete triggering condition.
Measurements were made at the wind site to determine the effect of operation of a wind
turbine at the end of the distribution system. The measurements consisted of voltage and current
transients and steady state values for normal and extreme operation of the wind turbines.
I ._
18
4
4 Normal operation of a wind turbine was performed by allowing the wind turbine to execute
a normal start and then to force a normal stop and braking cycle. During a normal start the wind
turbine is allowed to rotate as caused by the wind until it reaches operating speed at which time it
connects automatically to the grid. Because the wind turbine is rotating at operating speed there is
no inertial energy required to accelerate the mass of the turbine. This reduces the duration of inrush
current very significantly and requires only that the generator excitation current be supplied. A
normal stop is simulated by simply turning the wind turbine control to OFF following which all the
normal braking functions are exercised.
Extreme operation of the wind turbine was simulated by causing an across the line start of
the wind turbine followed by an emergency stop. A rotor `jog' function is provided within the
control system for this test. During a rotor jog the brakes are released and the wind turbine is
motored up to speed. This causes about an initial seven times rated inrush current which reduces to
about 15% rated after a few seconds. An emergency stop is caused by causing a full interruption of
the mains power to the wind turbine. In this case all brakes are applied instantly. An emergency stop
does not in fact alter the transient expected when the wind turbine comes off line.
The response of the distribution system to these events is shown in Figures 5 - 14 following.
2. Measurement Results
Figures 5 and Figure 6 illustrate the RMS voltage and current transients for the first several
seconds following an across the line or jog start. Figure 7 and Figure 8 illustrate the corresponding
instantaneous voltage and current during the first few cycles after starting. It can be seen from Figure
6 that the current immediately rises to 550 Amperes for about five seconds and then falls to a much
lower value as the wind turbine comes up to near synchronous speed. Figure 5 illustrates the Iine to
neutral voltage for the same period. The voltage immediately drops from the nominal value of about
285 Volts to 265 Volts and then recovers as the wind turbine increases in speed. This drop is about
7% and should not cause concern unless problems occur with control systems which will not tolerate
such supply voltage swings. This voltage depression could be much greater at the end of a smaller
19
a
M
feeder. It is to be realized also that this measurement is for one wind turbine only. It would not be
possible to jog start several wind turbine simultaneously. In fact the wind turbine controls must be
arranged so that only one wind turbine can be started at a time in this fashion. The short ( 3 ms )
burst is the result of contact bounce. Figure 8 illustrates the commencement of the current inrush and
the beginning of the 5 second or so transient for across the line starting. These waveforms are normal
for an induction motor start.
Figure 9 and Figure 10 illustrate the RMS voltage and current during the first few seconds
of a normal start or connection respectfully. It can be seen that the current inrush is only for few
cycles or so. It is seen however that the initial value is still several times the rated current. This
current includes the current necessary for excitation of the fixed power factor correction capacitors
which are wired across the generator terminals. Figure I I and Figure 12 illustrate the instantaneous
voltage and current transients during the same normal start up. Figure 12 is a good illustration of
how quickly the current transient decays.
Figure 13 and Figure 14 illustrate the instantaneous voltage and current transients during a
normal shut down. It can be seen that the voltage is hardly affected and that the current decays more
or less instantly. The corresponding waveforms for an emergency stop are identical.
9.0 CONCLUSION
The interpretation of these results is that operation of the up to twenty four AOC 15/50 wind
turbines on the KEA line will not contribute to an unacceptable power quality deterioration. It must
be realized however that because of the high instantaneous current demand caused by either a jog
start or a normal start, only one wind turbine should be normally started or jog started at a time.
Although power quality may not be adversely affected, the addition of larger numbers of wind
turbines will require that consideration be given to changes in control and operation of the diesel
generators. To commence learning the extent of the changes necessary it would be prudent to
commence monitoring and testing diesel plant operations as the number of wind turbines on the
system is incrementally increased.
►.1