Power quality measurements in a singlehouse

Transcription

1 24th International Conference & Exhibition on Electricity Distribution (CIRED) June 2017 Session 2: Power quality and electromagnetic compatibility Power quality measurements in a singlehouse microgrid Sarah Rönnberg, Math Bollen, Jakob Nömm Department of Energy Sciences, Luleå University of Technology, Skellefteå, Sweden sarah.ronnberg@ltu.se ISSN doi: /oap-cired Abstract: Power quality measurements have been performed during 1 year in a small-scale microgrid in Sweden. The microgrid consists of one residential house that can operate in grid-connected or grid-disconnected (islanded) modes. During the year the microgrid was disconnected from the main grid most of the time. 1 Introduction International council on large electric systems (CIGRE) WG6.22 defines a microgrid as: an electricity distribution system containing loads and distributed energy resources, (such as distributed generators, storage devices, or controllable loads) that can be operated in a controlled, coordinated way either while connected to the main power network or while islanded [1]. The power quality phenomena associated with a microgrid during island operation are the same as for any other grid [2] but the disturbance levels might be different. During island operation, the system impedance for lower frequencies is expected to be higher compared to the grid connected case especially in a small-scale microgrid. Loads that draw a distorted current will thus more likely result in distorted voltage and one major concern is that the harmonic levels within a microgrid will exceed the limits during island operation. The increase in impedance along with a mainly inverter-based generation can also result in resonances being excited over a wide range of frequencies. During island operation voltage and frequency control has to be provided locally within the microgrid. As the microgrid lacks the strength that the main grid can provide this can be challenging and variations in either production or consumption can result in unacceptable variations in frequency and/or voltage magnitude. 2 Residential house This case study is based on measurements in a small-scale microgrid operated at 50 Hz. The microgrid consists of one residential detached house located on the Swedish west coast with a living space of 504 m 2 that is spread over three floors. The house is supplied from a 20 kwp photovoltaic roof top installation and a 2.6 kwp photovoltaic installation on the south facing facade of the building. The entire photovoltaic installation produced about 21,000 kwh in This covers the consumption of about 12,400 kwh for the house and about 3700 kwh to charge electrical vehicles. The surplus of the production is stored in a 144 kwh silicone battery and in an 1100 kwh hydrogen tank (a 5800 kwh hydrogen tank is under construction). The stored energy is then used at times with low solar irradiation. The house is well isolated and is heated with geothermal heating, solar thermal collectors and excess heat from the hydrogen production/conversion. Different heat recovery systems exist to reduce the loss of heat in for instance the ventilation of the house. The house is also equipped with a backup diesel generator. In addition to the aforementioned equipment, the house contains normal household devices like refrigerator, computer and lighting. Efforts have been made to reduce consumption and most equipment is of energy saving type, e.g. all lights in the house are LEDs. The house is intended to operate solely in island mode but has been connected to the main grid at some occasions. 3 Harmonic distortion Devices that draw a distorted current will result in distorted voltage and more so in a weak grid. A concern is that the harmonic levels will exceed the limits when in island mode. During some periods the house was connected to the main grid and a comparison can therefore be made between grid connection and island operation. The absolute value of the voltage total harmonic magnitude for the three phases throughout 1 year of measurements is shown in Fig. 1. The voltage total harmonic levels visibly drops during times when the house is connected to the main grid. The voltage harmonics 3, 5, 7 and 9 are shown in Fig. 2 during a 24 h period running island operation; all harmonics are thus produced within the microgrid. As a comparison, the same harmonics are shown for the duration of one day when the microgrid was connected to the grid (Fig. 3) and for a residential house of similar size located in the north of Sweden (Fig. 4). The magnitudes of the individual harmonics are lower for the grid connected cases and the variations over time are also smaller. The average harmonic spectrum during 1 h is shown in Fig. 5 when the house is connected to the grid and island operated. The load condition during the two occasions are similar, the measurements are taken at the same time of day but the rms current in each phase is slightly lower during island operation. The low order odd harmonics are dominating in both cases. The magnitudes of the voltage odd harmonics 3 11 are higher during island operation with some exceptions (third harmonic in phase B, fifth harmonic in phase C, seventh harmonic in phases B and C). The high-order odd harmonics are only visible during grid operation. Even harmonics up to order 14 are present only during island operation. 4 Unbalanced load One hour during operation of a large single-phase load has been studied and compared between the grid connected situation and island operation. The load is connected single phase and draws an rms current around 35 A which gives an average current unbalance of 88% (grid connected) and 91% (island operation). The 1 min average of the resulting voltage unbalance is shown in Fig. 6. In both cases the voltage unbalance remains below 1.15%. The average voltage unbalance for the duration of this hour is 0.84% 818 This is an open access article published by the IET under the Creative Commons

2 Fig. 1 One hour average of the total harmonic distortion measured during 1 year. Different colours indicate the three phases Fig. 4 One minute average of individual harmonics (from top to bottom) 3, 5, 7 and 9 for one phase during one day for a detached house in the north of Sweden Fig. 2 One minute average of individual harmonics (from top to bottom) 3, 5, 7 and 9 for one phase during one day Fig. 5 Individual voltage harmonics for the three phases during grid connection (blue bars) and island operation (red bars) at similar load condition Fig. 3 One minute average of individual harmonics (from top to bottom) 3, 5, 7 and 9 for one phase during one day when the house is connected to the grid for the grid connected case and slightly higher at 0.91% during island operation, both values well below the permitted 2% limit. Fig. 6 Voltage unbalance during one hour with unbalanced load 5 Supraharmonics An increase of waveform distortion in the frequency range between 2 and 150 khz (supraharmonics) can be expected due to the inverter-based generation and use of modern energy saving devices as both types of equipment are associated with supraharmonic emission. The photovlotaic (PV) inverters switching frequency is commonly located somewhere between 3 and 20 khz [3], emission from household devices, including electrical vehicles can be found throughout the supraharmonic frequency range [4]. The potential risk of resonances within a microgrid at or near the switching frequency of the inverters during island operation is also mentioned in [5]. This is an open access article published by the IET under the Creative Commons 819

3 Fig. 7 Measured supraharmonics at an instance with high supraharmonic content during grid connection Fig. 9 Measured supraharmonics at an instance with high supraharmonic content during island operation. Note the difference in vertical scale compared to Fig. 7 The presence of supraharmonics in the studied microgrid is shown in Figs In Fig. 7, the discrete fourier transform (DFT) of a 200 ms snapshot is shown during grid connection. There are some components present around 10 khz with a peak magnitude just over 1.3 V and some components below 5 khz. The levels of supraharmonics increase during island operation. As for the grid connected mode the supraharmonic emission is not always present and no clear trend of when they appear can be seen. The voltage waveforms shown in Fig. 8 (the results from the DFT are shown in Fig. 9) are taken during similar conditions as the measurement shown in Fig. 7. There are more components present and the magnitudes are also higher during island operation. The components with the highest magnitude can again be found around 10 khz with a peak value of 4.8 V (2.2% of the fundamental component at 50 Hz). These were measurements taken during daytime but it is worth mentioning that these components are not always present and no conclusion can be drawn about their origin. The sampling frequency of the instrument used sets the limit at 25 khz, no conclusion can therefore be drawn with regard to the upper supraharmonic range. Fig. 10 Frequency variation during three days in island operation 6 Frequency variations One concern with running a microgrid in island operation is the frequency variations that can occur. The 1 h average of the frequency during three days is shown in Fig. 10. The increase in frequency occurs during the hours when the sun is up and the PV systems are likely injecting active power. The low demand in relation to the available generation results in this increase. Looking at the frequency variation during 1 year the instances when the microgrid was connected to the grid can be identified but also times when there is likely no or low-pv production (see Fig. 11). Fig. 8 Voltage waveform at an instance with high supraharmonic content during island operation Fig. 11 Frequency variation during 1 year 820 This is an open access article published by the IET under the Creative Commons

4 Fig. 12 P st values, different colours indicate the three phases Fig. 14 Voltage unbalance during the transition from grid connection to island operation be linked to any event inside the microgrid and are likely originating in the grid. After islanding the average unbalance decreases but short pulses with an increased level are seen. These pulses correspond to a single phase load drawing a pulsating current. These pulses of current are present also before the microgrid is taken in island operation but without affecting the levels of voltage unbalance in any significant way (see Fig. 14). 9 Short-circuit capacity Fig. 13 Transition from grid connection to island operation, voltage (blue curve) and current (orange curve) for one phase According to EN the variation of the 50 Hz system frequency in an island operated grid should stay within ±15% ( Hz) at all time. 7 Flicker P st values for the three phases are shown in Fig. 12. It can be seen that as the microgrid goes from being island operated to grid connection the P st values decrease in one of the phases (yellow), increase in one and (red) and somewhat remain fluctuating around the same values in the third. The P st value remains at acceptable values throughout the measured period. 8 Transition to island operation The microgrid has, as shown previously, been connected and disconnected from the main grid at some instances. All transitions to and from island operation has been planned. Fig. 13 shows the total harmonic current and voltage at the instance when the microgrid is disconnected from the main grid. An increase in voltage harmonic distortion is clearly visible, however the level of current harmonic distortion somewhat decreases. The voltage harmonic distortion increases from 4 V (1.7%) to almost 13 V (5.6%) as the microgrid is disconnected from the main grid. Even after the increase of voltage harmonic magnitude the levels are well below the permitted limit at 8% set by IEEE 519 and EN Looking at the voltage unbalance at the instance of transition from grid connection to island operation the moment of disconnection is seen with a drop in voltage unbalance. The background voltage unbalance is higher during grid connection. The variations cannot The short-circuit capacity is a regular subject of discussion with island operation of individual customers or smaller parts of the distribution grid. There is concern that the local production units, typically with power-electronics interface, are not able to provide sufficient short-circuit current to guarantee a safe and selective fault clearing. It is however important to emphasis that in a microgrid with power-electronic-based production the short-circuit capacity (source impedance) can be significantly different for a normal load current, for an inrush current, and for a short circuit or earth fault. The kind of measurements as presented in this paper can also be used to obtain information on this. For the installation presented here, there are no indications yet that the fault current would be insufficient for a safe and selective fault clearing. 10 Conclusions Measurements have been performed during 1 year in a small-scale microgrid run in both island operation and grid connected. Several power quality indices have been analysed and in most cases a deterioration of the power quality index is seen as the microgrid is islanded: increase in harmonic voltage; increase in supraharmonic voltage; large increase in frequency variation; slight increase in voltage unbalance. No violation of the limits found in IEEE 519 or EN have been found. No general conclusion can however be drawn from the measurements presented here but for the fact that the differences are not obvious and mere studies and long term measurements in island operated microgrids are needed. This is an open access article published by the IET under the Creative Commons 821

5 11 References 1 CIGRÉ WG C6.22: Microgrids 1: engineering, economics, & experience. TB Zavoda, F., Bollen, M., Rönnberg, S., et al.: CIGRE/CIRED JWG C4. 24, Power quality and EMC issues associated with future electricity networks status report. CIGRE Sessions, JWG C4/C6.29: Power quality aspects of solar power. Technical Brochure, Cigré Rönnberg, S.: Emission and interaction from domestic installations in the low voltage electricity network, up to 150 khz. PhD dissertation, Luleå University of Technology, Laaksonen, H., Kauhaniemi, K.: Voltage and current THD in microgrid with different DG unit and load configurations. CIRED th Int. Conf. and Exhibition on Electricity Distribution, Prague, Czech Republic, 2009, pp This is an open access article published by the IET under the Creative Commons