See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/280559092 Impact of Solar Panels on Power Quality of Distribution Networks and Transformers Article in Canadian Journal of Electrical and Computer Engineering · January 2015 DOI: 10.1109/CJECE.2014.2359111 CITATIONS READS 9 2,450 3 authors: M. A. Awadallah B. Venkatesh Ryerson University Ryerson University 56 PUBLICATIONS 755 CITATIONS 142 PUBLICATIONS 2,288 CITATIONS SEE PROFILE SEE PROFILE Birendra Nath Singh Ryerson University 11 PUBLICATIONS 82 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Hybrid solar-wind-grid power generation system; Modeling, simulation and MPPT” View project Pole-mounted energy storage system for reliability enhancement of local distribution companies View project All content following this page was uploaded by M. A. Awadallah on 29 July 2015. The user has requested enhancement of the downloaded file. CANADIAN JOURNAL OF ELECTRICAL AND COMPUTER ENGINEERING, VOL. 38, NO. 1, WINTER 2015 45 Impact of Solar Panels on Power Quality of Distribution Networks and Transformers Impact de panneaux solaires sur la qualité des réseaux de distribution et de transformateurs de puissance Mohamed A. Awadallah, Bala Venkatesh, Senior Member, IEEE, and Birendra N. Singh Abstract— This paper presents an investigation on the impact of solar panels (SPs) on the power quality of distribution networks and transformers. Both solar farms and residential rooftop SP are modeled with the distribution network according to Canadian Utility data. Total harmonic distortion of voltages and currents on both sides of the distribution transformer are monitored under different operation conditions. A laboratory setup employing a single-phase inverter and three-phase transformer is used to test system performance in the presence of phase unbalance and harmonics. Core and winding temperatures are measured under various loads. Simulation and experimentation results show that the performance of distribution networks and transformers under the impact of SPs is within standard limits. Résumé— Cet article présente une enquête sur l’impact des panneaux solaires (PS) sur la qualité des réseaux de distribution et de transformateurs de puissance. Les fermes solaires ainsi que les toits avec des PS résidentiels sont modélisés avec le réseau de distribution basé sur les données canadiennes de services publics. Les distorsions harmoniques totales des tensions et des courants sur les deux côtés du transformateur de distribution sont contrôlées sous différentes conditions de fonctionnement. Une installation de laboratoire utilisant un onduleur monophasé et un transformateur triphasé est utilisée pour mesurer la performance du système de test avec un déséquilibre de phase et d’harmoniques. Les températures du noyau et d’enroulement sont mesurées avec différentes charges. Les résultats de simulation et expérimentaux montrent que la performance des réseaux de distribution et des transformateurs sous l’impact des PS respecte les standards. Index Terms— Distribution networks, distribution transformers, harmonics, solar panels (SPs). I. I NTRODUCTION OLAR photovoltaic (PV) energy is one of the most rapidly developing renewable sources. Solar cells are made of semiconductor materials which convert light energy of the sun into dc electricity. Therefore, the usage of inverters with solar panels (SP) becomes inevitable before solar power can be used by local loads or transmitted into the grid. SPs are normally installed in distribution networks, rather than the generation or transmission levels of power systems. Both small rooftop SP installations and large solar farms create voltage harmonics and inject current harmonics into the distribution network by the associated inverters. On the other hand, distribution trans- S Manuscript received February 27, 2014; revised May 29, 2014; accepted September 15, 2014. Date of current version March 31, 2015. This work was supported by Hydro One Networks Inc., Toronto, ON, Canada, through the Centre for Urban Energy, Ryerson University, Toronto. M. A. Awadallah was with the University of Zagazig, Zagazig 44516, Egypt. He is now with the Centre for Urban Energy, Ryerson University, Toronto, ON M5B 2K3, Canada (e-mail: [email protected]). B. Venkatesh is with the Centre for Urban Energy, Department of Electrical and Computer Engineering, Ryerson University, Toronto, ON M5B 2K3, Canada (e-mail: [email protected]). B. N. Singh is with Hydro One Networks Inc., Toronto, ON M5G 2P5, Canada (e-mail: [email protected]). Associate Editor managing this paper’s review: Davood Yazdani. Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/CJECE.2014.2359111 formers are subject to second-quadrant operation (i.e., when the active power flow is reversed) under light load conditions when SP operate at or close to full capacity. Although power system harmonics are known to be consequences of nonlinear loads, accurate measurement of voltage and current harmonics is quite tricky . Tracking down harmonic sources is also challenging as well as effective filtering and mitigation techniques . A few publications in the literature have considered the effects of SP on distribution networks. Impact of the SP at the Sydney Olympic Village on accommodating network is addressed in , where voltage and current total harmonic distortion (THD) remain within standard limits even if all SP operate simultaneously. In a weak network supplied by SP, the replacement of incandescent lamps by compact fluorescent lamps—for energy saving— increases voltage THD . The initial THD of 3.14% can reach 10.15%, 22.2%, and 34% if 30%, 60%, and 90% of the lighting load is replaced, respectively. In , the effects of SP on power grids of two small Greek islands are studied and compared with the case of Diesel generator supplies. Although voltage THD with SP operation is higher than the Diesel generator case, yet, it remains under standard limits. With the combination of linear and nonlinear loads on a transformer, an expansion of the standard K-factor evaluates the composite harmonic current . Rad et al.  report 0840-8688 © 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information. 46 CANADIAN JOURNAL OF ELECTRICAL AND COMPUTER ENGINEERING, VOL. 38, NO. 1, WINTER 2015 TABLE I C ASES OF U NBALANCED O PERATION harmonic measurements on six distribution transformers along with other performance indices. Harmonic effect on winding eddy current loss is found much more significant than other stray loss. A laboratory setup is proposed to measure losses of high switching frequency converter transformers, and compare with those computed through finite elements . In , voltage and current harmonics caused by various nonlinear lighting loads are presented, and derating of the distribution transformer is accordingly proposed. A calculation routine of the reduced per unit load, at which distribution transformers maintain full lifetime under current harmonics, is presented in . Harmonics generated by different nonlinear loads are measured in the laboratory, and their impact on transformer losses, temperature rise, and loss of life is studied . The finite element method is used to calculate the hottest spot temperature through field strength solution at different parts of the transformer . The change in the temperature rise because of harmonics is used to estimate the lifetime expectancy and propose a new loading profile to keep lifetime unaffected. This paper introduces a study on the impact of SP and their associated inverters on the distribution network and transformer. A MATLAB/Simulink model is built for SP and distribution network according to Canadian Utility data; both solar farms and residential rooftop SP are considered. Voltage and current THD are monitored under different conditions. In a laboratory setup, a single-phase grid-tied commercially available inverter for solar power applications is used to feed a three-phase transformer connected to the grid. Losses, efficiency, and voltage and current THD are measured as well as core and winding temperatures. Results show that SP do not have significant harmful impact on distribution networks or transformer, as long as they keep low relative rating with respect to the power carrying capacity of the system. II. S IMULATION R ESULTS A. Analysis of Solar Farm A simulation model is built in MATLAB/Simulink for a solar farm on the basis of Canadian Utility data for the system shown in Fig. 1. The transmission system is modeled as 115 kV, 170 MVA, three-phase source. Two 83 MVA, 115/27.6 kV transformers exist within the distribution system shown as one block in Fig. 1. Bus 1 is the point of common coupling (PCC), where the solar farm is connected to the system at this point. The solar farm includes 17 SP of 500 kW each. One SP is separately connected to a 27.6 kV/265 V, 500 kVA transformer, whereas the remaining 16 SPs are connected in pairs to eight transformers of 1 MVA each. The SP module employs maximum power point tracking (MPPT) Fig. 1. Solar farm system model. Fig. 2. SP model. via a dc converter, whereas the associated inverter applies sinusoidal pulse-width modulation (PWM) and has an LCL low-pass filter (Fig. 2). It is still practically accepted to have such ratings by considering that solar farms do not usually operate at rated capacity all the time owing to changes in environmental conditions. SPs typically give rated output at standard test conditions (1000 W/m2 solar irradiation and 25 °C cell temperature) which may not be maintained all the time. On the other hand, from the simulation-work viewpoint, the SP impact on distribution transformers becomes more serious as the SP rating increases. Therefore, simulation results would be more conclusive in the present case. The system shown in Fig. 1 is modeled under balanced and unbalanced conditions with and without capacitors at the PCC when the solar farm operates at rated power. The THD of phase voltages and currents on both sides of the transformer are monitored. In balanced case, system voltages of the transmission system are equal to rated value and shifted from each other by 120°. Three cases of voltage unbalance are considered as shown in Table I. In practice, it is unacceptable to have voltages more than 1.1 pu. However, some phase voltage values in Table I are exaggerated to make sure the impact of voltage unbalance on harmonic distortion is insignificant. The THD of phase voltages and currents at both sides of the transformer are given in Table II. Results imply that phase voltage THD is always less than the standard permissible limit of 5%, except for phase-B secondary voltage under unbalanced case #3. However, current THD is mostly under the standard limit, except for phase-A primary current AWADALLAH et al.: IMPACT OF SPs ON POWER QUALITY 47 TABLE II P ERCENTAGE THD OF P HASE V OLTAGES AND C URRENTS U NDER BALANCED AND U NBALANCED C ONDITIONS TABLE III P ERCENTAGE THD OF P HASE V OLTAGES AND C URRENTS U NDER D IFFERENT S HUNT C APACITOR R ATINGS AT PCC under case #3 of unbalanced operation. None of the individual voltage or current harmonic components exceeds the standard limit of 3% under any case. The only exception is the fifth secondary voltage harmonic on phase-B under unbalanced case #2, which is 3.05%. Results show that waveform distortion is more sensitive to unbalance in voltage magnitudes than phases. It should be noted that waveform distortion in cases of unbalance may not be due only to SP. Some phase voltage magnitudes significantly exceed rated value which could cause transformer saturation leading to more distortion. Finally, it could be concluded that voltage and current THD are mostly within acceptable values with SP operation under balanced and unbalanced conditions. The next procedure to remedy the negative impact on voltage and current waveform distortion is to install shunt capacitors at PCC to help filter out harmonics. Capacitors with different VAR ratings are tried; results of new THD are shown in Table III. It can be noticed that there is no significant changes in THD values for capacitor rating up to 8 MVA. At 10 MVA, voltage THD on both sides and secondary current THD are noticeably reduced. Capacitors at such rating are likely to assist low-pass filters of SP by reducing the cutoff frequency. However, for ratings above 14 MVA, all THD increase again, possibly due to transformer saturation as a result of voltage increase at PCC. B. Analysis of Residential Rooftop SP Impact of residential rooftop SP on distribution network and transformer is then considered. The difference between Fig. 3. Residential SP system model. residential SP and solar farms is not only the output power, but also the method of integration with the system. Solar farms are usually located far outside residential areas, and are normally connected to the grid at PCC using long feeders. On the contrary, residential SP are located in urban areas, and are connected to the grid at many points inside the distribution network. Three-phase inverters are commonly used with solar farms, and transformers are mostly employed to raise the voltage up to the distribution level. However, because residential SP are usually less than 10 kW in rating, singlephase inverters are used to connect directly to the 240 V feeders. 48 CANADIAN JOURNAL OF ELECTRICAL AND COMPUTER ENGINEERING, VOL. 38, NO. 1, WINTER 2015 TABLE IV P ERCENTAGE THD OF P HASE V OLTAGES AND C URRENTS U NDER D IFFERENT N UMBERS OF R ESIDENTIAL SP Fig. 4. Experimental setups. (a) Delta. (b) Star. (c) Star grounded. A MATLAB/Simulink model is built for the system (Fig. 3) on the basis of Canadian Utility data. Saturable transformer modules at 27.6 kV/416 V, 100 kVA are used in the model, where a maximum of 10 SP are connected to phase A of each transformer. The model is run for different numbers of SP and THD of voltages and currents are recorded. Results are shown in Table IV. Results indicate that harmonics caused by SP on phase A can intrude into other phases having no SP. With the increase in SP numbers, voltage and current THD boost up, but remain within standard limits. III. E XPERIMENTAL R ESULTS The experimental work aims at testing the SP impact on distribution transformers in a laboratory setup. A three-phase dry-type, /Y, 240/240 V, 10 kVA, 220 °C temperature class transformer is tested using a single-phase grid-tied, 208 V, 3 kW, commercially available inverter for solar power applications. Three different configurations are used for testing as shown in Fig. 4. The primary side connected to the grid could be delta, star, or star grounded, where two terminals of the secondary winding are connected to the inverter. The dc source used to feed the inverter can run in current mode to best represent SP characteristics. The inverter employs MPPT over two channels that could be separately connected to two SP or paralleled together to the same one. The active power flow from the inverter to the grid through the transformer Fig. 5. Voltage THD under different configurations. (a) Primary. (b) Secondary. is controlled via the dc input. Voltage and current THD are measured across the transformer as well as active and reactive powers. Core and winding temperatures are also measured by two independent thermocouples. It should be noted that such experimental setup represents what exists in reality when the load profile on the distribution network is very low and SP are connected to one phase at the secondary side of the distribution transformer. It also represents the case of only one house in a neighborhood having a rooftop SP connected to the distribution transformer through a single-phase inverter. Nevertheless, three-phase energization of the transformer is essential for the balance of the flux in the core. As inverter rated power is 3 kW, active power is varied between 500 W and 2500 W in steps of 500 W; measurements AWADALLAH et al.: IMPACT OF SPs ON POWER QUALITY Fig. 6. Current THD under different configurations. (a) Primary. (b) Secondary. are taken at each loading point. Owing to the large thermal time constant of the transformer, the highest load of 2500 W is set for about 4 h to reach steady-state temperature. Then, active power is reduced through the inverter dc input, and every load power is set for about 40 min. Both core and winding temperatures are recorded every 10 min. The voltage and current THD values are plotted against active power as shown in Figs. 5 and 6 at the grid and inverter sides of the transformer. The voltage THD is always within the permissible 5% limit. Current THD has a high value close to 30% at light load; it decreases with loading approaching the standard limit. Transformer losses and efficiency are plotted for different winding configurations as shown in Fig. 7. It should be noted that losses are still much less than rated, as transformer is lightly loaded. It is evident from the plots that efficiencies are almost comparable for different configurations, while the delta connection yields the lowest losses. Temperature variations are shown in Figs. 8–10 for delta, star, and star grounded connections, respectively. From such temperature plots, the heating thermal time constant for core and winding average out to 104.575 and 67.214 min, respectively. By considering that the temperature class of the transformer is 220 °C, it is evident that both core and winding temperatures are well below the rated limit at all loads. It is true even in cases the current THD becomes close to 30%. It is believed that current harmonics add to the winding loss, and hence to the winding temperature rise. However, because the transformer is lightly 49 Fig. 7. Transformer losses and efficiency under different configurations. (a) Losses. (b) Efficiency. Fig. 8. Core and winding temperature under delta connection. loaded, overall winding temperature is still within very safe limits. Although such high current distortion does not have serious effect on the distribution transformer, its impact on other network components could be critical. In addition, it is believed that the reversal of active power flow, which results in the second quadrant operation, has insignificant impact on the transformer performance. The temperature plots also show that both winding and core temperatures are consistently increasing exponentially as far as the load is kept constant. However, as 50 CANADIAN JOURNAL OF ELECTRICAL AND COMPUTER ENGINEERING, VOL. 38, NO. 1, WINTER 2015 Such voltage is zero under all load levels as the primary voltage is more influenced by the balanced grid voltage. In addition, the current flowing from the star point to ground, in case of star-grounded connection, is given in Table V, as well. IV. C ONCLUSION Fig. 9. Core and winding temperature under star connection. This paper presents a study on the impact of SPs on the power quality of distribution networks and transformers. Solar farms and rooftop residential SPs are independently simulated when incorporated into distribution systems. Results show that voltage and current are mostly within permissible limits under different conditions. An optimum value of capacitors connected at the PCC can help the system reduce voltage and current THD. Results also show that voltage and current distortion increases as the number of SP inverters connected to the system increases. An experimental setup is built in the laboratory to test a three-phase dry-type transformer when fed by a singlephase grid-tied inverter, and connected to the grid. Voltage and current THD, losses and efficiency of the transformer, and core and winding temperatures are all measured under different SP output powers. Results show that all THD values are within standard limits, losses and efficiency increase with loading, and temperature is always well below the rated load value. A PPENDIX Fig. 10. Core and winding temperature under star-grounded connection. TABLE V A SSESSMENT OF E XPERIMENTAL S YSTEM U NBALANCE load is reduced, winding temperature decreases exponentially. Whereas, core temperature maintains the same rising behavior because the supply voltage and frequency are unchanged. To assess the unbalance of transformer operation, the standard deviations of line current magnitudes at the primary (grid) side are computed for different configurations in Table V. It is clear from such table that the star connection gives less unbalance than both delta and star-grounded, which are comparable. It is also obvious that unbalance in line currents increases with loading for all configurations. The unbalance is also assessed by measuring the neutral point voltage with respect to ground in case of star connection. AWADALLAH et al.: IMPACT OF SPs ON POWER QUALITY R EFERENCES  A. Cataliotti and V. Cosentino, “A new measurement method for the detection of harmonic sources in power systems based on the approach of the IEEE Std. 1459–2000,” IEEE Trans. Power Del., vol. 25, no. 1, pp. 332–340, Jan. 2010.  S. K. Jain and S. N. Singh, “Harmonics estimation in emerging power system: Key issues and challenges,” Electric Power Syst. Res., vol. 81, no. 9, pp. 1754–1766, Sep. 2011.  E. Vasanasong and E. D. Spooner, “The effect of net harmonic currents produced by numbers of the Sydney Olympic Village’s PV systems on the power quality of local electrical network,” in Proc. Int. Conf. Power Syst. Technol., Perth, Australia, Dec. 2000, pp. 1001–1006.  P. N. Korovesis, G. A. Vokas, I. F. Gonos, and F. V. Topalis, “Influence of large-scale installation of energy saving lamps on the line voltage distortion of a weak network supplied by photovoltaic station,” IEEE Trans. Power Del., vol. 19, no. 4, pp. 1787–1793, Oct. 2004.  G. A. Vokas and A. V. Machias, “Harmonic voltages and currents on two Greek islands with photovoltaic stations: Study and field measurements,” IEEE Trans. Energy Convers., vol. 10, no. 2, pp. 302–306, Jun. 1995.  G. W. Massey, “Estimation methods for power system harmonic effects on power distribution transformers,” IEEE Trans. Ind. Appl., vol. 30, no. 2, pp. 485–489, Apr. 1994.  M. S. Rad, M. Kazerooni, M. J. Ghorbany, and H. Mokhtari, “Analysis of the grid harmonics and their impacts on distribution transformers,” in Proc. IEEE Power Energy Conf. Illinois, Champaign, IL, USA, Feb. 2012, pp. 1–5.  Y. Han and Y.-F. Liu, “A practical transformer core loss measurement scheme for high-frequency power converter,” IEEE Trans. Ind. Electron., vol. 55, no. 2, pp. 941–948, Feb. 2008.  M. Bagheri, M. S. Naderi, T. Blackburn, and T. B. Phung, “Transformer efficiency and de-rating evaluation with non-sinusoidal loads,” in Proc. IEEE Int. Conf. Power Syst. Technol., Auckland, New Zealand, Oct./Nov. 2012, pp. 1–6.  D. M. Said and K. M. Nor, “Simulation of the impact of harmonics on distribution transformers,” in Proc. IEEE 2nd Int. Conf. Power Energy, Johor Baharu, Malaysia, Dec. 2008, pp. 335–340.  D. M. Said, K. M. Nor, and M. S. Majid, “Analysis of distribution transformer losses and life expectancy using measured harmonic data,” in Proc. 14th Int. Conf. Harmon. Quality Power, Bergamo, Italy, Sep. 2010, pp. 1–6.  S. Taheri, H. Taheri, I. Fofana, H. Hemmatjou, and A. Gholami, “Effect of power system harmonics on transformer loading capability and hot spot temperature,” in Proc. 25th IEEE Can. Conf. Elect. Comput. Eng., Montreal, QC, Canada, Apr./May 2012, pp. 1–4. View publication stats 51 Mohamed A. Awadallah was born in Zagazig, Egypt, in 1971. He received the B.S. (Hons.) and M.S. degrees from Zagazig University, Zagazig, in 1993 and 1997, respectively, and the Ph.D. degree from Kansas State University, Manhattan, KS, USA, in 2004, all in electrical engineering. He is currently a Visiting Research Fellow with the Centre for Urban Energy, Ryerson University, Toronto, ON, Canada. His current research interests include motor drives, smart grids, and renewable energy. Dr. Awadallah is a member of the Eta Kappa Nu, Tau Beta Pi, and Phi Kappa Phi. Bala Venkatesh (SM’08) received the Ph.D. degree from Anna University, Chennai, India, in 2000. He is currently a Professor and the Academic Director of the Centre for Urban Energy, Ryerson University, Toronto, ON, Canada. He is also a Registered Professional Engineer in the provinces of Ontario and New Brunswick, Canada. His current research interests include power system analysis and optimization. Birendra N. Singh received the M.Eng. degree from the Memorial University of Newfoundland, St John’s, NL, Canada. He taught Electrical Engineering Courses with Ryerson University, Toronto, ON, Canada. He is currently the Manager of the Technology Development with Hydro One Networks Inc., Toronto. He has over 30 years of diversified experience in the electric utility industry with Newfoundland and Labrador Hydro, St John’s, Toronto Hydro, Toronto, and Hydro One, Toronto. He is also a Registered Professional Engineer in the province of Ontario, Canada.