Long-term power quality measurements in medium voltage networks

Abstract

A power quality survey has been performed in several Swedish medium voltage (MV) substations and a study of the long-term power quality levels under normal operation has been done. The measurements have been performed during several years and the study includes seasonal variations as well as weekly and daily variations. This paper will focus on harmonic voltage distortion (THD and individual harmonics). An analysis has been done looking into metrological data to see any possible correlation with variations in harmonic distortion. Conclusions show that there are some significant differences over the years and between days for all sites. Long term measurements are needed to gain information on which levels one can expect, measuring for only a few weeks or month can give an over or underestimation of the actual levels. Results also shows that the magnitudes of harmonic distortion are well below the permitted level at all time at the selected sites. INTRODUCTION Power quality is defined by IEC as the “characteristics of electric current, voltage and frequency at a given point in an electric system, evaluated against a set of reference technical parameters”. Utilities are required to deliver electrical power that are of good quality. What is considered as good quality is, in Sweden, defined in EIFS 2013:1 [1] where a number of indices are listed, all required to be kept within a certain limit, e.g. the voltage unbalance shall not exceed 2% and the total harmonic distortion (THD) shall not exceed 8%. Power quality measurements are normally performed by the grid owner at several places in the grid to ensure that the levels are kept within the limits. These measurements are however rarely published and detailed knowledge on long and short-term (years to hours) variations in the medium voltage grids are somewhat missing. In this study, a few sites were selected for analyses: \uf0a7 Site 1, 30 kV. Substation close to several wind parks. Data over eight consecutive years. \uf0a7 Site 2, 10 kV and 30 kV. Substation close to domestic and industry load. Data over six consecutive years. \uf0a7 Site 3, 30 kV. Substation close to wind parks, domestic and industry load. Data over five consecutive years. RESULTS Variation over years There is no significant difference between the three phases in any of the measurements presented here; only one phase is shown throughout the paper. Unless otherwise stated, one-hour average values has been used. The cumulative distribution function (CDF) of the voltage total harmonic distortion (THD), each measured year at the three sites has been calculated and shown in Figure 1. It can be seen that there are variation during individual years and between years. The variation between years is greatest at Site 1 where 2011 has the lowest values and 2016 the highest. Different percentiles have been calculated, including the entire duration of the measurements, these results are presented in Table 1. The highest 95% value of the THD is found at Site 3 whereas the highest 50% value is seen at Site 2 (only considering the 30 kV level). The variation over the years can partly be explained by changes in the network topology that has occurred over the years. Table 1 Percentiles of the entire measured duration for the three sites, values in % of FND 50% 75% 95% Site 1 THD 0.57 0.71 0.96 Site 1 H5 0.21 0.32 0.53 Site 1 H7 0.37 0.49 0.67 Site 2, 10 kV THD 0.86 1.02 1.23 Site 2, 10 kV H5 0.53 0.73 0.99 Site 2, 10 kV H7 0.56 0.66 0.81 Site 2, 30 kV THD 0.69 0.82 1.05 Site 2, 30 kV H5 0.42 0.59 0.89 Site 2, 30 kV H7 0.46 0.53 0.64 Site 3 THD 0.63 0.92 1.39 Site 3 H5 0.54 0.84 1.33 Site 3 H7 0.27 0.33 0.44 25th International Conference on Electricity Distribution Madrid, 3-6 June 2019 Paper n°652 CIRED 2019 2/5 Figure 1 CDF of voltage THD, 8 consecutive years, Site 1 (top), 6 consecutive years, Site 2, 30 kV (middle) and 5 consecutive years, Site 3 (bottom) Seasonal variations Some seasonal variations of the harmonic magnitudes are expected, the load will somewhat change during the year as will the climate conditions, e.g. the temperature variations has, during the considered time period, varied between -35°C and +35°C at the locations in question. In [3] the seasonal variations of power quality in German low voltage networks were studied and variations were found in a considerable number of sites, these variations would also impact the medium voltage levels. The 95-percentile value for each month during the measured period is calculated and the result is shown in Figure 2, Figure 3 and Figure 4. A trend of slightly higher values during the summer months (June to September) is seen at Site 1 and 2 even though it cannot be concluded that this period always has the highest values during a year. In 2012 (Site 1), for instance, the magnitude of the 5th harmonic is 0.32% in July and 0.47% in December. For Site 3 it can be concluded that the summer period has higher values compared to the winter months, this trend is visible during all five years. Figure 2 95 percentile of magnitude harmonic 5 per month during the period 2011-2017, Site 1 Figure 3 95 percentile of magnitude harmonic 5 per month during the period 2011-2017, Site 2 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 5t h H ar m on ic (% FN D ) 2011 2012 2013 2014 2015 2016 2017 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 0.2 0.4 0.6 0.8 1 1.2 1.4 5t h H ar m on ic (% FN D ) 2013 2014 2015 2016 2017 25th International Conference on Electricity Distribution Madrid, 3-6 June 2019 Paper n°652 CIRED 2019 3/5 Figure 4 95 percentile of magnitude harmonic 5 per month during the period 2014-2018, Site 3. The average value of the 5th harmonic each week is shown in Figure 5 for 2017 and for 2014 in Figure 6 at all three sites. Similar trends are seen here as in the monthly 95% value. Site 3 shows seasonal variations with higher values during summer (mid-May until mid-August) for all 5 years. Values greater than the yearly average at this site are all found during summer. Site 2 has the highest values during summer (2017) and fall (2014). Weeks with a weekly average greater than the yearly average are found throughout the year but somewhat more frequently in summer 2017 and fall 2014. For 2017, the overall variations in Site 2 are significantly smaller compared to Site 3. During 2014 there are six weeks in the fall that have significantly higher values compared to the rest of the year at Site 2, these are the same weeks that show the highest values at Site 1 as well. The average weekly value at Site 1 varies but these variations cannot be linked to a seasonal change in the same way as for Site 3. For this site, the weeks with highest values are found in early spring (April 2014) and summer (June 2014) and fall (mid-September until November 2014 and 2017). Values above the yearly average are randomly spread throughout the year. Figure 5 Average value of 5th harmonic calculated per week during 2017 at each site. Figure 6 Average value of 5th harmonic calculated per week during 2014 at each site. Correlation with climate conditions Hourly temperature and wind speed has been downloaded from Sweden s Meteorological and Hydrological Institute (SMHI) and from the website renewable ninja [2] for the considered locations and years. The correlation between temperature and average THD, per hour, at Site 1 is shown in Figure 7. No clear conclusion can be drawn from this. There are somewhat higher magnitudes at temperatures above 0°C but in most cases similar values can be found also at temperatures below zero. Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 0 0.5 1 1.5 2 2.5 3 5t h H ar m on ic (% FN D ) 2014 2015 2016 2017 2018 2017 Site 1 0 20 40 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Av er ag e m ag ni tu de 5 th h ar m on ic (% FN D ) 2017 Site 2 0 20 40 Week 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 2017 Site 3 0 20 40 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1