Download SECTION-4-Chapter 10

Chapter 1 – (chapter 10)
Time-Varying and Probabilistic
Considerations: Setting Limits
Thomas Ortmeyer, Wilsun Xu, Yahia Baghzouz
1.1 Introduction
Harmonic limits have been established and widely applied over the last 20 years. The
primary harmonics standards are IEEE 519 and IEC 61000-3-6. IEEE 519 was first published in
1981, was substantially revised in 1992, and is currently undergoing another substantial revision.
IEC 61000-3-6 has undergone a similar cycle of development. The two standards are currently
in widespread use throughout the world, and the steady state limits they impose are not
particularly controversial.
Harmonics are, in theory, steady state quantities.
continually changing, and harmonic levels vary regularly.
Power systems, however, are
Figure 1 shows measured Total
Harmonic Distortion (THD) measured on a 138 kv bus. These levels are generally within the
harmonic limits, except for a 5.3 minute excursion and a later 3.2 minute excursion above 2%.
Figure1: 138kV bus voltage THD as a function of time.
At this time, it is unclear how to deal with these short term excursions above the steady
state harmonic limits. While it is widely recognized that some level of short term excursions
above the limits should be allowable, the area remains under active research, and there is at this
point no consensus on short term harmonic limits.
This chapter provides a perspective on limiting factors on short term harmonic levels.
1.2 Harmonic Definition
Clearly, rapid variations in distortion levels fall into the category of transients rather than
harmonics. As the speed of the variations decrease, at some point the distortion is more clearly
described as being a set of harmonics of varying magnitude rather than a transient phenomena.
IEC 61000-4-3 sets a window of 3 seconds as being the shortest window over which harmonics
should be measured, and IEC 61000-3-2 allows a 50% increase in distortion level for this period,
for certain types of equipment. IEEE 519-1992 does not address this measurement issue, but
does recognize that short term excursions above the limits are to be expected.
1.3 Harmonic Effects
The fundamental principles guiding harmonic standards are to:
a) Avoid system and load damage and disruption due to high harmonic levels
b) Limit harmonic losses to an acceptable level
c) When mitigation is necessary, find an economical and equitable solution
In developing short term harmonic limits, the primary consideration will be the first item,
avoiding damage or disruption of system and load equipment.
Harmonic effects are tied to either voltage magnitudes or current flows. One primary
impact of voltage magnitude on system equipment is the insulation issue—and electrical
overstress, subsequent corona and breakdown are generally high speed phenomena.
Load
disruption is another primary impact of voltage distortion. Load disruption effects are also short
term in nature, with nuisance tripping having been reported to occur regularly as distortion levels
approach 10%. The impact of overcurrent on system equipment, on the other hand, is most
directly tied to I2R heating, and resulting temperature increases, which is a slow phenomena.
Load performance is much more complex—distorted voltage applied to a load causes distorted
current draw.
Heating effects on induction and synchronous machines are a primary concern,
although torque ripple and increased noise also can be a limiting factor. In electronic devices,
high frequency ripple and resulting effects on input filters can raise both heating and insulation
overstress concerns.
On the other hand, certain waveforms can cause power supplies to
prematurely go out of spec on low voltage, for example.
Thermal time constants on power equipment are relatively long, and are an obvious
choice for appropriate time limitations for short term harmonics. Voltage overstress and load
disruption, being fast phenomena, can theoretically provide upper limits on acceptable levels of
short term harmonics. The problem in applying these concepts is that the power systems is an
amalgamation of equipment with widely varying thermal time constants and insulation levels.
1.4 Proposed Limit Methodologies
Figure 2 shows a histogram and cumulative distribution function for a typical harmonic
voltage measurement, in this case, for a week long measurement.
While this data yields
information on the time duration at a given distortion level, it does not contain information on
the duration of individual events at that distortion level. Therefore, data presented in this form
does not give full information for determining acceptability for short term bursts of harmonics.
An alternate method for displaying harmonic data is presented in Figure2, where curves are
presented for both the maximum lengths observed for individual bursts, and the total cumulative
length of bursts during a measurement period. This curve also includes a conceptual limit for
short term harmonic limits.
It seems clear that there is a need for limits on harmonic bursts for individual events. A
statement that harmonic limits should be met 95% of the time is insufficient in that it does not
address upper limits on distortion levels, or the time duration of individual bursts of high
distortion. In order to further explore the impacts of harmonic bursts, it is of interest to study the
impact of these bursts on ac power capacitors.
Figure 2. Drawing of cumulative distribution (Ttotal) and the maximum duration and maximum
duration of individual burst (Tmaximum) for a harmonic measurement. The curve also includes a
conceptual limit for short term harmonic levels.
1.5 Capacitor Aging
Capacitors are known to be sensitive to excessive harmonic levels. Partial discharge and
overtemperature are two primary agents which lead to accelerated aging of power capacitors.
IEC standards call for harmonic measurements taken with approximately 200 millisecond
windows (exactly 10 cycles on 50 hertz or 12 cycles on 60 hertz systems). These measurements
are taken continuously, and the rms value of the harmonic components are then computed over 3
second intervals and over 10 minute intervals. The established total harmonic distortion limits
cannot be exceeded in any 10 minute interval. The 10 minute time frame is in the order of
magnitude of power capacitor thermal time constants, so it is appropriate that steady state limits
be enforced in this time frame, from the capacitor perspective. From a thermal point of view,
shorter term bursts of harmonics could be acceptable.
1.6 Partial Discharge
Instantaneous voltages above the partial discharge inception voltage causes rapid aging of
the capacitor. This aging is the primary reason for capacitor voltage limits, which are 120% of
rated peak voltage in IEEE Std 18, IEEE Standard for Shunt Power Capacitors. This standard
also limits rms voltage to 110% of rated rms voltage. Therefore, for capacitors which are not
derated and experience fundamental frequency overvoltages, it is important that the harmonics
do not increase the peak instantaneous voltage by more than 10%. Montanari and Fabiani [3]
investigate the effect of voltages above the partial discharge inception limit. They show that
capacitors exhibit rapid aging when the voltage is sufficient to cause partial discharge to occur.
It is clear from this study that harmonic voltages above the partial discharge inception voltage
should be avoided for the 3 second measurement as well as the 10 minute measurement.
A 5% THD limit will not necessarily limit the instantaneous harmonic voltage to less than
10%-- THD is an RMS type measure, while voltage magnitudes add algebraically—experience
shows that it must be assumed that harmonic voltage peaks and the fundamental voltage peak
coincide.
Therefore, the present THD limits on harmonic voltage levels are not directly
compatible with the capacitor voltage standard. In many practical installations, however, the
majority of the harmonic content is centered at one or two harmonic frequencies. Comparison of
THD limits and the capacitor voltage limits shows that capacitor voltage limit is met if the
fundamental voltage is within its limit, the THD limit is maintained, and the harmonics are
limited to a relatively low number of significant frequencies.
IEEE Std. 18 also provides allowance for transient overvoltage. These do require the
capacitor to tolerate a limited exposure to voltages above 120% of the capacitor rated peak
voltage. This transient allowance does not appear to have any relevance to harmonic bursts
measured over a 3 second window. Another complicating factor is that it is not uncommon to
derate capacitors in applications where harmonic levels are expected to be significant. If this
practice is uniformly followed, relatively high distortion levels would not necessarily lead to
overvoltage stress in the capacitor.
1.7 Thermal Heating
Harmonic flows contribute losses in addition to those experienced under purely
sinusoidal conditions. Steady harmonic levels are allowed for in capacitors through rms current
ratings which are somewhat higher than would be experienced by the capacitor with rated
sinusoidal voltage at the rated frequency.
The thermal time constant of a capacitor has been defined as the time it takes for the
capacitor core temperature to reach 65% of its final value in response to a step change in ambient
temperature. Harmonic variations which are slower than the thermal time constant of a capacitor
will cause significant variations in the capacitor temperature. Conversely, harmonic overcurrents
which last less than 10% of the thermal time constant of the capacitor will cause relatively little
temperature rise.
With regard to the IEC time limits, the 10 minute harmonic window can be expected to
be on the order of the thermal time constant of typical power factor correction capacitors.
Significant overtemperature can be expected for overcurrents lasting 10 minutes, therefore it is
appropriate to apply steady state limits in this time frame. Shorter term harmonic overcurrents,
however, can occur without excessive overtemperature. In particular, a doubling of the harmonic
current over a 3 second period should not lead to reduced life when the current rating is
maintained over a 10 minute measurement.
The IEC 3 second and 10 minute limits refer to harmonic voltages, not currents. The
capacitor will experience loss components proportional both to current and to voltage. The low
capacitor impedance at the harmonic frequencies, however, makes capacitor current a primary
concern. It is not possible to specify a general voltage THD limit for a broad range of harmonic
frequencies, and guarantee that the capacitor current will remain within the device rating. These
ratings are, however, can be sufficient in systems with typical harmonic voltage profiles, where
the harmonic content is primarily limited to the lower orders.
Figure 3. Histogram and cumulative distribution of THD levels for a week long measurement.
1.8 Conclusions
The development of limits for short term harmonic levels is clearly a difficult topic.
These limits must provide an economical compromise between operational efficiency and
equipment lifetimes. It is clear that there will be no answer which avoids some level of
compromise.
These limits must involve both magnitude and duration. The correspondence of duration
limits and equipment thermal limits is apparent, and any duration limit should be a small fraction
of power equipment thermal time constants.
The IEC’s 3 second measurement certainly
qualifies in this regard, and the requirement that 10 minute measurements meet harmonic limits
also seems reasonable. There may exist middle ground between the two for an intermediate
limit.
The magnitude of an allowable burst poses more difficulties. In cases where a single
harmonic is dominant (which is not unusual), a single harmonic limited to 3% in the steady state
could be doubled to 6% in short term situations without problem. If, however, there is a THD
level of 3% consisting of two equal harmonics of 2.12%, in the worst case (which again is not
unusual), these could sum and add 4.24% to the peak in the steady state. If this level is allowed
to double in the short term, the peak voltage would increase by 8.5%, which is approaching the
limit imposed by the standard (if the capacitor is operating at it’s rms limit, 10% above nominal
voltage).
This difficulties of applying a short term THD limit on capacitors is apparent. It appears
that a relaxation of short term THD limits should not allow levels to go more than double the
steady state limits, based on the existing levels for general distribution systems. An allowance of
short term variations of 50% above the steady state limits appears to be a conservative level
which should not lead to problems in most cases.
1.9 References
1. IEC 61000-4-7 Electromagnetic compatibility (EMC) - Part 4: Testing and measurement
techniques - Section 7: General guide on harmonics and interharmonics measurements and
instrumentation,
for
power
supply
systems
and
equipment
connected
thereto
2. IEEE Std. 18-2002,, IEEE Standard for Shunt Power Capacitors.
3. G. C. Montanari and D. Fabiani. “The Effect of Non-Sinusoidal Voltage on Intrinsic Aging
of Cable and Capacitor Insulating Materials,” IEEE Transactions on Dielectrics and
Insulation, Vol. 6, No. 6 (Dec., 1999): pp. 798-802.