            Department of Telecommunications,Kerala State requested NIT Calicut to conduct a harmonic study at   the telephone   Exchanges   located in the Calicut area in order to derive a better understanding of the harmonic distortions and the level of neutral currents at those Exchanges and to arrive at recommendations on remedial measures if needed.
            Subsequent discussions with Executive Engineer (E),TED,Calicut resulted in identifying two Exchanges - called Exchange-I and Exchange-II in this article - for detailed harmonic study.
            The Department of Electrical Engineering,NIT Calicut assigned the work to me. I visited the two Exchanges on 9-12-1999 and 10-12-1999 and carried out detailed harmonic measurements at various points in the Electrical System at the two Exchanges.I also discussed the various harmonic related issues with the team of Telecom Engineers who witnessed the measurements.
            This article describes the measured harmonic levels,analyses the data and reports the findings.

2. Executive Summary

The Power Quality Study conducted at Exchange-I and Exchange-II results in the following recommendations/observations.

IEEE 519 Standards specify limits for the individual harmonic content and THD for currents drawn by non-linear loads from a Point of Common Coupling. The Exchange-I goes beyond these limits marginally.
The harmonic levels at the Exchange-I are not at levels detrimental to the system performance.
There are no statutory limits to harmonic currents drawn from mains at present.
There is no need for a Harmonic Filter installation at the Exchange-I at present.
The harmonic currents drawn at the Exchange-II contain even harmonics and particularly second harmonic. This is due to insufficient smoothing inductance in the d.c side of three phase half-controlled converters in the Power Plant.
The harmonic content in the Exchange-II is much above the IEEE limits due to its even harmonic content. Harmonic Filtering is not the solution to this.
There is no need to install a Harmonic Filter at Exchange-II.
Exchange-II will have to replace the Power Plant with modern units as and when the Exchange is required to meet statutory harmonic standards.
The Power Factor at Exchange-II is at 0.89 even without employing any Power Factor Correction Capacitor and is satisfactory at that level.
No Zig-zag transformer or Zero Sequence Filter is needed at the Exchanges since there is only little triplen harmonic currents in the Neutral at both Exchanges and the 50Hz current level in the Neutral is well within the allowed limits in both the Exchanges.

3. Harmonic Distortion in Voltages and Currents at Exchange-I

The Exchange-I receives Grid Supply at 11kV level and steps it down to 415V level using two 500kVA transformers. Both transformers are used in the normal condition; but they are not paralleled. There are two 250kVA DG Sets to provide standby supply. Only one is run at a time and the other is maintained as standby. The LT supply is regulated at the transformer output by means of 500kVA a.c voltage stabilisers which have 300-460V input range and maintain 415V output within ± 1%. The transformers have 4.3% impedance. A 70kVAr-capacitor installation is used to improve the power factor at this Exchange. The active power demand of the Exchange-I varies between 100kW to 160kW over the day.

The load consists of air conditioning plants, lighting, UPS and PC loads, Lifts, Pumps and the Float Charger-Rectifier units called the Plant. The consumption is dominated by the consumption in A/C units and in the Plant. There are 11 Package AC Plants of 10T rating and 10 Window A/C Units of 1.5T rating. The Window A/C units constitute the major single phase load. The UPS & PC load and the Plant load are the distorting loads in the system. The Plant uses phase controlled Thyristor Converters to provide the charging current to the 48V battery banks as well as the load current from the battery node.

This Exchange was monitored on 9-12-1999 to study the harmonic distortion contents in voltages and currents during different times of the day. The measurements were performed at the secondary side of Transformer-2 after switching all the loads to this transformer. Measurements were repeated with and without the power factor correction capacitor. The entire Exchange load was put on one of the 250kVA DG Set and harmonic measurements were carried out at the output of the Generator. Also the harmonic distortion in the Plant Feeder was studied separately in detail. At all stages the corresponding Neutral currents were observed for magnitude as well as harmonic content.

Summary of Measurements at Exchange-I

The Measurements are summarised in the Table below.

Currents (A) R         Y       B     N P.F R          Y        B kW kVA kVAr THD in Voltages %
R       Y       B THD in Currents%
R         Y        B

I 298 276 338 53 0.80 0.78 0.77 159 204 127 ~0 ~0 ~0 9.4 8.8 7.5

II 329 310 387 59 0.81 0.77 0.79 182 233 142 2.7 2.1 2.6 13.4 10.9 10

III 274 257 273 27 0.73 0.73 0.71 140 193 134 5.2 5.2 5.1 6.7 7.4 10.4

IV 106 111 110 6 0.72 0.75 0.72 56 77 52 4.2 3.7 3.4 28.4 26.4 25.5

Case I - Transformer Secondary, Average Load, Capacitor OFF, Measurement on one cable of a 3-run system-Data shown after multiplication to account for cable paralleling-Cable Balance was bad and hence data is somewhat in error esp. in the power factor.

Case II - Transformer Secondary, Peak Load, Capacitor ON, Other comments as in Case I.

Case III - DG Set Output Lines, Typical Load, and Measurement by Clamping all the Cables of multi-run Cable together and hence data accurate.

Case IV - At Plant Feeder, Measurement by clamping all the Parallel Cables together.

The above measurements reveal that about 1/3^rd of total consumption takes place in the Plant and this also happens to be the dominant distorting load in the system. The three-run cable that connects the LT side of the transformer to the Main SwitchBoard was not balanced. The layout of the runs was such that the Clamp CT of the Harmonic Analyser Equipment could not get all the three cables in its jaw. Hence the measurement was done on one of the three cables and a multiplication factor of 3.4 was used to arrive at the total current, power etc. in each phase. The factor should have been 3 ideally; but 3.4 had to be used in order to adjust for the cable unbalance. Under these conditions the measured value will be in error by as much as 10% especially in power factor measurement.

The system power factor is about 0.73 while on DG Set. The system reactive power will be about 150kVAr under peak conditions. The 70kVAr capacitor at the bus will improve the power factor to about 0.85.This is thought to be sufficient. However if the power factor has to be 0.95 a total of 100kVAr capacitor will be needed and hence an additional 30kVAr has to be installed.

Detailed Harmonic Analysis at Exchange-I

Detailed waveform observation and harmonic data were stored in the Harmonic Analyser Equipment memory and was analyzed later using the software available along with the Equipment.

Fig.1 shows the waveform printouts for the condition in Case I. The Phase Voltages and line currents at the 500kVA transformer secondary and the neutral current at the secondary are shown. The values correspond to one of the cables in a three-run system. Notice that the voltage waveforms are reasonably pure while the current waveforms are perceptibly distorted. The neutral current (total) shows predominantly 50Hz current with the peak in its half cycles explained by presence of single-phase UPS loads in the system. The neutral current level is small and is about 15% of line current typically and is explained by the unbalances in lighting loads, single phase Window A/C loads and UPS loads . No equipment exclusively aimed at improving the current balance or eliminating third harmonic in neutral needs to be installed at this exchange. There is only very little third harmonic in neutral and that does not justify any effort to remove it.

Fig.2 shows the harmonic contents in the line currents as % of fundamental current. The fundamental currents were 297A, 275A and 338A in R, Y and B phases respectively. Higher order triplen harmonics are absent. 5^th and 7^th harmonics dominate in general.

Fig.3 shows some waveforms for Case II and Fig.4 shows the harmonic contents in the line currents as % of fundamental current for this Case. The fundamental currents were 330A, 312A and 384A in R, Y and B phases respectively. 13^th (and 11^th to a certain extent) harmonic in currents seem to have become dominant in this Case. This indicates a partial harmonic resonance of the power factor correction Capacitor with the transformer leakage impedance. The presence of 11^th and 13^th harmonics can be seen as wiggles in the current waveforms in Fig.3. However, though there is partial amplification of 11^th and 13^thharmonics in currents it is not high enough to demand urgent installation of harmonic filters.

Fig.5 shows the line and neutral currents at the Plant input point and show the well-known square wave currents of a Thyristorised converter with continuous conduction at the d.c side. Fig.6 shows the harmonic contents in these currents.

Fig.7 shows the phase voltages and line currents at the output of DG Set when the entire system load was on it. Notice that the distortion in voltages is perceptible now. However the distortion in currents have come down.

Fig.8 shows the distortion content in % of fundamental current.

Does Exchange-I Need Harmonic Filtering?

The Table below shows the internationally accepted IEEE STD 519 Current Distortion Limits for Non-Linear Loads. These limits have to be observed by every power consuming equipment taking more than 50W or so in the European countries. Also, United States and other North American countries have imposed similar restrictions on customers who employ non-linear loads.

Maximum Harmonic Current Distribution in % of Fundamental
Harmonic Order ( Odd Harmonics)
I_sc/I_L	h<11	11£ h <17	17 £ h <23	23 £ h <35	35 £ h	TDD
< 20 20-50 50-100 100-1000 >1000	4.0 7.0 10.0 12.0 15.0	2.0 3.5 4.5 5.5 7.0	1.5 2.5 4.0 5.0 6.0	0.6 1.0 1.5 2.0 2.5	0.3 0.5 0.7 1.0 1.4	5.0 8.0 12.0 15.0 20.0
Even Harmonics are Limited to 25% of the Odd Harmonics Limits Above.
Where I_sc = Maximum Short Circuit Current at Point of Common Coupling , I_L = Average Maximum Monthly Demand Load Current , TDD = Total Demand Distortion , Harmonic Current Distortion in % of maximum Demand Load Current ( 15 or 30 min Demand Current)

            The Point of Common Coupling in the case of Exchange-I is the transformer secondary and the net leakage impedance at that point will be about 5% on 500kVA basis (accounting for the Stabiliser also). And the maximum demand could be about 250kVA.This gives a figure of 40 for the ratio I_sc/I_L at this point of coupling. Thus the harmonic limits are 7% for components upto 11^th, 3.5% for components upto 17^th and 8% for THD. Fig.4 shows that under peak load conditions, R and Y Phase currents exceed limits in 5^th and 13^th harmonics and the THD of currents in all the three phases (13.4%, 10.9% and 10%) exceed the stated limit of 8%. However, there is no statutory restriction/limits on the harmonic currents that can be drawn by a non-linear load in this State as yet.
            Some form of restriction on the harmonic currents that can be drawn by a customer from a Point of Common Coupling is expected to become a statutory requirement in the coming years and the above referred IEEE limits will be followed more or less. The harmonic content in currents at Exchange-I goes beyond the limits slightly. But notice that strict control of Maximum Demand will easily put Exchange-I in the next category of I_sc/I_Lratio i.e. at 50-100 level and the corresponding harmonic individual and THD limits are well above the current maximum harmonic contents and THD recorded in the system. The system has to maintain its MD below 200kVA for this. If it comes to that it can be done by additional installation of about 40kVAr of Capacitor and installing an MD Monitor/Controller which will switch off some non-critical loads whenever the MD is likely to cross the set limit. In any case any such harmonic restrictions are yet to appear.
            The level of harmonic distortion in voltages at various points is well below 5% when the system is on KSEB and slightly above 5% when the system is on DG Set and does not need any correction in either situation. The harmonic content in currents at around 13-14% max does not need any immediate correction since the over heating in distribution equipment due to this level of distortion is not significantly harmful and is negligible considering the available capacity in the transformers and cables.
            Hence, the Exchange-I does not need any Harmonic Filter since there are no statutory limits to be met at present and the existing distortion level will cause significant overheating anywhere in the system. The Exchange-I does not need any Zig-zag transformer (or Zero Sequence Filter) in the neutral since there is only little third harmonic current in the neutral and the fundamental component present in the neutral is not high enough to warrant load balancing by Zig-zag transformer. The 70kVAr Capacitor may be raised to 100kVAr if power factor correction to 0.95 is needed. This will not cause any more harmonic resonance than what is already present. In fact the present partial resonance may be alleviated with 100kVAr unit since the resonance frequency will shift to around 10^th harmonic and there is little harmonic current at that frequency in the system.

4. Harmonic Distortion in Voltages and Currents at Exchange-II, Calicut.

The Exchange-II receives KSEB Supply at 11kV level and steps it down to 415V level using two 400kVA transformers. Both transformers are used in the normal condition; but they are not paralleled. There are two 500kVA DG Sets and one 200kVA DG Set to provide standby supply. Only one is run at a time and the others are maintained as standby. The LT supply is regulated at the transformer output by means of 400kVA a.c voltage stabilisers which have 300-460V input range and maintain 415V output within ± 0.5%. The transformers have 4.57% impedance. There is no capacitor installation at this Exchange. The active power demand of the Exchange-II varies between 100kW to 160kW over the day.

The load consists of air conditioning plants, lighting, UPS and PC loads, Lifts, Pumps and two Float Charger-Rectifier units called the Power Plant and SMPS. The consumption is dominated by the consumption in A/C units and in the Power Plant & SMPS. There are 12 Package AC Plants of 7T rating and 38 Window/Split A/C Units of 1.5T rating. The Window/Split A/C units constitute the major single phase load. The UPS & PC load and the Plant & SMPS loads are the distorting loads in the system.

The Power Plant uses half-controlled Thyristor Converters to provide the charging current to the 48V battery banks as well as the load current from the battery node. The SMPS uses Active Power Factor Corrected Switched Mode Power Supply Modules to carry out the same function. The Power Plant has 3 units of 430V/50Hz input 50V, 400A output rating and two of them were functioning at the time of measurement and were delivering 350A and 300A to the 50V battery bus. The SMPS has installed capacity of 8 modules each capable of delivering 200A into 50V-battery node and five modules were functioning at around 160A each at the time of measurement.

This Exchange was monitored on 9-12-1999 and 10-12-1999 to study the harmonic distortion contents in voltages and currents during different times of the day. The measurements were performed at the secondary side of Transformer-1 that was supplying the distorting load. The entire Exchange load was put on the 500kVA DG Set-2 and harmonic measurements were carried out at the output of the Generator. Also the harmonic distortion in the Power Plant Feeder was studied separately in detail due to even harmonic content observed in the currents. The SMPS Feeder currents were separately monitored to study the efficacy of such units in improving the power factor and in reducing the current distortion. At all stages the corresponding Neutral currents were observed for magnitude as well as harmonic content.

The HT Side voltages and total HT Side currents were captured in an oscilloscope in order to measure the power factor angle and to calculate the HT Side power factor. This became necessary due to a prominent disparity between the power factor measured by the TOD Meter at the HT side and the power factor metered by Harmonic Analysis Equipment at the LT Side. Oscilloscope measurements at the HT Side resolved the issue in favor of the Harmonic Analyser measurement.

Summary of Measurements at Exchange-II

The Measurements are summarised in the Table below.

Currents (A)
R Y B N
P.F
R Y B
kW kVA kVAr THD in Voltages %
R Y B
THD in Currents %
R Y B

I 182 171 191 26 0.98 0.98 0.96 123 126 27 4.2 3.3 4.4 13.2 14.3 11.4

II 308 294 303 17 0.87 0.86 0.86 190 220 112 2.4 2.7 2.5 6.3 7.9 7.4

III 37 38 39 1.5 0.79 0.81 0.82 21.3 26.4 15.6 3.3 2.0 2.0 67.8 68.5 61.0

IV 25 26 27 1.5 0.99 0.99 0.99 18 18.3 2.4 3.2 2.1 2.0 5.4 12.5 12.4

Case I - Transformer-1 Secondary, Typical Load, Supplies all the distorting loads. But Transformer-2 was supplying the A/C Plants and some other loads too and it is not included.

Case II - DG Set Output Lines, Entire System Load, and Measurement by Clamping one out of 4 parallel cables.

Case III - Power Plant Feeder Load

Case IV - SMPS Load

The system recorded a power factor of 0.86 while on DG Set. The same value was metered at the HT Side by an independent measurement described later. Thus the system power factor is 0.86 and is satisfactory at that level. There is no immediate need for installation of power factor correction capacitors at this Exchange.

Detailed Waveform and Harmonic Analysis at Exchange-II.

Fig.9 shows the voltage and current waveforms for Case I and the corresponding harmonic content table containing the values of dominant harmonics. The current waveforms show prominent presence of even harmonics. These even harmonic components come from the Power Plant load.

Fig.10 shows the voltage and current waveforms for Case-II when the entire Exchange load is on a 500kVA DG Set. Here also the most dominant harmonic in current is the second harmonic.

IEEE harmonic limits for even harmonics are at 25% of corresponding odd harmonic limits. Thus the harmonic content in the currents at Exchange-II are above the limits allowed by IEEE 519 standard.

Fig.11 shows the waveforms of R, Y and B currents drawn by the Power Plant load. These currents are seen to be asymmetric about time axis and contain very high amounts of 100Hz component (about 60% of 50Hz component). See Fig.12 for the harmonic content information of these currents. This Power Plant uses three phase half controlled rectifier (using three thyristors and three diodes plus one output free wheeling diode) with d.c side smoothing inductance. This is an old design strategy and was aimed at reducing thyristor count at a time when the cost of thyristor was high.

Fig.11 Current Waveforms (R,Y,B) in the Power Plant Feeder

But the current in even such a half-controlled converter will have half wave symmetry and will contain only odd harmonics if the d.c side smoothing inductance were high enough. A similar converter was simulated using a circuit simulator to study the effect of d.c side inductance value on the harmonic content of input current. Based on the simulation study (which is not reproduced in this report) it is concluded that the Power Plant at Exchange-II draws large even harmonic current due to insufficient smoothing inductance in the d.c side and consequent discontinuous current operation of the Converter.

Harmonic standards are more stringent on even harmonics. And it is difficult (if not impossible) to filter a harmonic component like second harmonic by passive filtering. Passive Filtering works from only 5^th harmonic practically. Hence Exchange-II will have to consider replacing the Power Plant with modern units like SMPS Charger or Fully Controlled Thyristor Converter as the only method to comply with the harmonic requirements if and when Utility Authorities impose such requirements. No Harmonic Filters have to be installed at present because the dominant harmonic is second harmonic and it is not easily filtered.

Fig.13 shows the R, Y and B currents drawn by the SMPS units and Fig.14 show the harmonic content information of the same currents. The SMPS does a good job of maintaining power factor at close to unity. However the THD in currents drawn in Y and B phases is excessive considering what is practically achievable in SMPS Technology. Notably these two phases draw even harmonics too.

Fig.13 Current Waveforms (R,Y,B) in the SMPS Feeder

Does Exchange-II Need Harmonic Filtering?

The Exchange has excessive even harmonic content in current at the Point of Common Coupling compared to what the IEEE 519 Standard allows. Especially the second harmonic. However there is no stringent harmonic standard to be conformed to at present. Secondly 6 to 8% THD in currents will not cause significant over heating anywhere in the system. Moreover filtering the second harmonic currents by Passive Harmonic Filter is going to be a very costly affair. The system power factor is 0.86 at present and there is no urgent need to add a power factor correction capacitor unless the firm wants to improve the p.f still further for some reason or other. Thus installing a harmonic filter is going to be very costly and ineffective unless it is tuned to 100Hz and its contribution to 50Hz reactive power is really not needed much (since the p.f. is already good enough).

Hence Exchange-II does not need a Harmonic Filter at present. The Exchange will have to replace the existing Power Plant with modern equivalents as and when it is called upon to meet statutory specifications on the harmonic currents drawn. No such statutory specifications are in force at present.

HT Side Measurements at Exchange-II

It was observed during the Power Quality Audit at this Exchange that the TOD Meter was registering a power factor of 0.5 only. The Harmonic Analyser used by the consultant recorded 0.98 on one transformer and 0.86 when the entire exchange load was on one of the 500kVA DG Set. There was no reason for the low p.f recorded by the TOD Meter. However the TOD Meter is new and hence many more measurements at individual phase level as well as three phase level were carried out using different instruments to confirm the value arrived at by the consultant's Harmonic Analyser.
A Tektronix 710A Digital Storage Oscilloscope was hooked on to the HT Side PT Output and CT Output. A Hall effect DC/AC current probe was used to conduct the CT Secondary current into the Oscilloscope. The waveforms were stored in the Oscilloscope and the phase angle between voltage and current was measured using cursor measurement functions in the Oscilloscope. The HT side current waveforms (representing the entire load of the Exchange) in the three phases are shown in Fig.15 and shows negligible distortion as expected. The power factor angle measured by this technique turned out to be 26 deg and the corresponding power factor is 0.89. The value obtained by monitoring the DG Set was 0.86,but this could have been slightly in error due to unbalances in the 4 cables in the 4-run cable of DG Set.

Fig.15 Total HT Side Currents (R,Y,B) at CT Secondary

Hence it is concluded that the system power factor is around 0.89 and that the TOD Meter is in error due to some reason and needs to be looked into.

5. Equipment Used in the Harmonic Study

SIM 50 Energy/Harmonic Analyser from DUCATI Instruments
WASED Software for SIM 50
ITT Make DC/AC Clamp Energy Audit Meter
Tektronix 710A Digital Storage Oscilloscope and Hall Effect Current Probe
MicroSim Design Lab 8.0 EDA Package