TEMPUS ENERGY: UNBALANCE IN THREE PHASE GRIDS 1: Introduction Consider a three phase grid as visualized in Figure 1. Suppose the three phase voltages π’π1 (π‘) , π’π2 (π‘) and π’π3 (π‘) are sinusoidal voltages having three times the same magnitude. The phase differences between these voltages equal 120°. In each phase, there is a grid impedance and assume three times the same grid impedance is obtained i.e. π1 = π2 = π3 . In case the load is symmetric, the phase currents ππ1 (π‘) , ππ2 (π‘) and ππ3 (π‘) have the same magnitude and the phase differences equal 120°. This implies the line voltages π’πΏ1πΏ2 (π‘), π’πΏ2πΏ3 (π‘) and π’πΏ3πΏ1 (π‘) also have the same magnitude and the phase differences also equal 120°. When the load of the grid is not symmetric, the phase currents ππ1 (π‘) , ππ2 (π‘) and ππ3 (π‘) do not have the same magnitude and/or do not have phase differences equal to 120° (i.e. the loads have different power factors implying the currents do not have the same phase differences with respect to the phase voltages π’π1 (π‘) , π’π2 (π‘) and π’π3 (π‘)). This implies the three line voltages π’πΏ1πΏ2 (π‘), π’πΏ2πΏ3 (π‘) and π’πΏ3πΏ1 (π‘) do not have the same magnitude and/or do not have phase differences equal to 120°. Figure 1: Three phase power system 2: The use of vectors or complex numbers To simplify calculations, sines are very often respresented by complex numbers (vectors). The sinusoidal voltage π’(π‘) = √2 π π ππ(ππ‘ + π) Μ with magnitude π and phase φ. More precisely, is represented by a complex number π Μ = ππ ππ = π(πππ π + ππ πππ). π In a symmetric three phase power system, there are three phase voltages π’π1 (π‘), π’π2 (π‘) and π’π3 (π‘) having the same amplitude and having phase differences equal to 120°. More precisely, the voltages π’π1 (π‘) = √2 ππ π ππ(ππ‘ + π) 2π π’π2 (π‘) = √2 ππ π ππ (ππ‘ + π − ) 3 2π (π‘) = √2 ππ π ππ (ππ‘ + π + ) {π’π3 3 Μ π1 = ππ π ππ π Μ π2 = ππ π π(π−2π⁄3) {π Μ π3 = ππ π π(π+2π⁄3) π are visualized in Figure 2. Figure 2: Vector diagram of a symmetric three phase power system Figure 3 visualizes a vector diagram for a non symmetric three phase power system. In general, the voltages have different amplitudes and the phases are arbitrary (i.e. there are no phase differences equal to 120°) giving π’π1 (π‘) = √2 ππ1 π ππ(ππ‘ + ππ1 ) {π’π2 (π‘) = √2 ππ2 π ππ(ππ‘ + ππ2 ) π’π3 (π‘) = √2 ππ3 π ππ(ππ‘ + ππ3 ) Μ π1 = ππ1 π πππ1 π Μ π2 = ππ2 π πππ2 {π Μ π3 = ππ3 π πππ3 π Figure 3: Vector diagram of a non symmetric three phase power system 3: Symmetrical components A three phase power system can always be considered as the sum of a direct component (positive sequence), an inverse component (negative sequence) and a homopolar component (zero sequence). The direct component is visualized in Figure 4 and given by π’π1 (π‘) = √2 ππ π ππ(ππ‘ + ππ ) 2π π’π2 (π‘) = √2 ππ π ππ (ππ‘ + ππ − ) 3 2π π’ (π‘) = √2 ππ π ππ (ππ‘ + ππ + ) { π3 3 Μ π1 = ππ π πππ π Μ π2 = ππ π π(ππ−2π⁄3) {π Μ π3 = ππ π π(ππ+2π⁄3) π Figure 4: Direct component The inverse component is visualized in Figure 5 and given by π’π1 (π‘) = √2 ππ π ππ(ππ‘ + ππ ) 2π π’π2 (π‘) = √2 ππ π ππ (ππ‘ + ππ + ) 3 2π (π‘) = √2 ππ π ππ (ππ‘ + ππ − ) {π’π3 3 Μ π1 = ππ π πππ π Μ π2 = ππ π π(ππ+2π⁄3) {π Μ π3 = ππ π π(ππ−2π⁄3) π Figure 5: Inverse component The homopolar component is visualized in Figure 6 and given by π’β1 (π‘) = √2 πβ π ππ(ππ‘ + πβ ) {π’β2 (π‘) = √2 πβ π ππ(ππ‘ + πβ ) π’β3 (π‘) = √2 πβ π ππ(ππ‘ + πβ ) Μ β1 = πβ π ππβ π Μ β2 = πβ π ππβ {π Μ β3 = πβ π ππβ π Figure 6: Homopolar component 4: The use of symmetrical components As already mentioned, an arbitrary non symmetric three phase power system is given by π’π1 (π‘) = √2 ππ1 π ππ(ππ‘ + ππ1 ) {π’π2 (π‘) = √2 ππ2 π ππ(ππ‘ + ππ2 ) π’π3 (π‘) = √2 ππ3 π ππ(ππ‘ + ππ3 ) with arbitrary ππ1 , ππ2 , ππ3 , ππ1 , ππ2 and ππ3 . Starting from ππ1 , ππ2 , ππ3 , ππ1 , ππ2 , ππ3 it is possible to calculate appropriate ππ , ππ , πβ , ππ , ππ and πβ in such a way the sum of the direct, inverse and homopolar voltage sequences gives the original non symmetric power system (voltages). 4.1: Calculation of symmetrical components To calculate the symmetrical components, the voltages π’π1 (π‘), π’π2 (π‘) and π’π3 (π‘) will be represented Μ π1 = ππ1 π πππ1 , π Μ π2 = ππ2 π πππ2 and π Μ π3 = ππ3 π πππ3 . In case ππ1 = by the complex numbers π ππ2 = ππ3 , ππ2 = ππ1 − 2π⁄3 and ππ3 = ππ1 + 2π⁄3, a symmetrical three phase system is obtained which equals Μ π1 = π Μ π1 π Μ π2 = π2 π Μ π1 {π Μ π3 = π π Μ π1 . π Here, π = π π2π⁄3 and notice that 1 + π + π2 = 0 and π3 = 1. Μ π for the first phase, π2 π Μ π for the second phase and π π Μ π for the The direct sequence contains π Μ π for the first phase, π π Μ π for the second phase and third phase. The inverse sequence contains π 2 Μ Μ β for the first phase, π Μ β for the π ππ for the third phase. The homopolar sequence contains π Μ second phase and πβ for the third phase. This implies Μ π1 = π Μ β + π Μ π + π Μ π π Μ π2 = π Μ β + π2 π Μ π + π π Μ π {π Μ π3 = π Μ β + π π Μ π + π2 π Μ π π or equivalently in matrix notation Μ π1 π 1 1 Μ π2 ] = [1 π2 [π 1 π Μ π3 π Μ β 1 π Μ π ] π ] [π 2 Μ π π π Μ π1 , π Μ π2 and π Μ π3 are known, it is possible to calculate the symmetrical In case the voltages π Μ β , π Μ π and π Μ π as components π Μ β π 1 1 1 Μ π ] = [1 π [π 3 Μ π 1 π2 π Μ 1 ππ1 Μ π2 ] π 2 ] [π π π Μ π3 This implies in case ππ1 = ππ2 = ππ3 , ππ2 = ππ1 − 2π⁄3 and ππ3 = ππ1 + 2π⁄3, there is only a Μ β = 0, π Μ π = direct component and there is no inverse nor homopolar component. More precisely, π Μ π = 0. In case ππ1 = ππ2 = ππ3 , ππ2 = ππ1 + 2π⁄3 and ππ3 = ππ1 − 2π⁄3, there ππ1 π πππ1 and π Μ β = is only an inverse component and there is no direct nor homopolar component. More precisely, π Μ π = 0 and π Μ π = ππ1 π πππ1 . The direct and the inverse components are actually symmetrical 0, π three phase systems having another phase sequence (positive or negative). By adding direct, inverse and homopolar voltage sequences, it is not only possible to obtain an arbitrary non symmetric power system of phase voltages. By adding direct, inverse and homopolar sequences, it is also possible to obtain line voltages, phase currents and line currents. 4.2: Exercise Consider a non symmetrical three phase grid having (π = 314 πππ⁄π ) line voltages π’πΏ1πΏ2 (π‘) = √2 240 π ππ(ππ‘) {π’πΏ2πΏ3 (π‘) = √2 210 π ππ(ππ‘ − 2.178) π’πΏ3πΏ1 (π‘) = √2 210 π ππ(ππ‘ + 2.178) Calculate the symmetrical components and study the impact of this non symmetrical voltage on an induction motor (suppose the stator windings are delta connected). Suppose the slip π of the motor equals π π ⁄2 (π π is the slip at the breakdown torque ππ ) and π π = 0.2. Calculate the decrease of the motor torque due to the inverse component in the grid voltage. Solution: Μ πΏ1πΏ2 = 240, π Μ πΏ2πΏ3 = 210 π −π2.178 and π Μ πΏ3πΏ1 = 210 π π2.178, one obtains that π Μ β = 0, π Μ π = Since π Μ π = 20. Using the Kloss equation, the torque in the positive direction equals 220 and π 2 πππ 4 ππ = π π π = 5 πππ π π π + π π due to the direct component (π π = π π ⁄2) (πππ is the breakdown torque due to the direct voltage component). With respect to the inverse rotating magnetic field due to the inverse component in the grid voltage, the slip equals π π = 1 + (1 − π π π π )=2− 2 2 The inverse voltage component gives a breakdown torque 2 ππ 2 20 πππ = ( ) πππ = ( ) πππ ππ 220 implying a torque in the negative direction which equals 2 πππ ππ = π π . π + π π π π π In case π π = 0.2, π π = 0.1 and π π = 1.9 which implies the total torque 20 2 π = ππ − ππ = 0.8 πππ − 0.208 πππ = (0.8 − 0.208 ( ) ) πππ = 0.798 πππ . 220 In the present example, the impact of the inverse voltage component on the torque is limited. In case π π is larger and/or the inverse voltage component is larger, the impact of ππ will be larger. Consider e.g. the situation when π π = 0.5. 5: The relationship between voltages and currents Consider a three phase power system as visualized in Figure 1. The generators generate a symmetric direct sequence π’π1 (π‘) , π’π2 (π‘) and π’π3 (π‘) (with sinusoidal voltages). In case the load is symmetric (with sinusoidal currents), also the currents ππ1 (π‘), ππ2 (π‘) and ππ3 (π‘) only contain a direct component. This implies the line voltages π’πΏ1πΏ2 (π‘), π’πΏ2πΏ3 (π‘) and π’πΏ3πΏ1 (π‘) only contain a direct component and are symmetric. In case the load is non symmetric, the currents ππ1 (π‘), ππ2 (π‘) and ππ3 (π‘) contain a direct component, an inverse component and a homopolar component. The voltage drops across the grid impedances π1 , π2 and π3 will also contain a direct component, an inverse component and a homopolar component. Relying on the voltage law of Kirchoff, the line voltages π’πΏ1πΏ2 (π‘), π’πΏ2πΏ3 (π‘) and π’πΏ3πΏ1 (π‘) will also contain a direct component, an inverse component and a homopolar component. This implies the voltages π’πΏ1πΏ2 (π‘), π’πΏ2πΏ3 (π‘) and π’πΏ3πΏ1 (π‘) will not be symmetric anymore. 6: Consequences of non symmetric grid voltages 6.1: Induction motors Consider an induction motor fed by a non symmetric three phase voltage and this voltage is the sum of a direct component and an inverse component. This implies the stator current of the motor will contain a direct component and an inverse component (assume the direct component is larger than the inverse component). The direct component generates a stator rotating magnetic field. Due to this rotating field, the rotor rotates in the same direction as this rotating field (positive direction). The inverse component generates a second rotating field which rotates in the opposite direction. This second rotating field also affects the rotor. Due to each rotating magnetic field, there is a torque speed characteristic as visualized in Figure 7. The first rotating field (due to the direct component) has a synchronous speed ππ and the second rotating field (due to the inverse component) has a synchronous speed −ππ . The actual speed π of the rotor is somewhat smaller than ππ . Figure 7: Torques due to the direct and inverse rotating fields Due to the direct rotating magnetic field there is a large driving torque in the positive direction. The inverse rotating magnetic field generates a smaller torque in the opposite direction (negative direction). The total torque is the direct torque reduced with the inverse torque. Moreover, due to the large slip of the rotor with respect to the inverse rotating magnetic field additional voltages and currents are induced in the rotor causing a significant additional heat dissipation. This reduces the energy efficiency of the motor. When comparing with ππ , the rotor has a slip π π = ππ − π ππ and when comparing with −ππ , the rotor has a slip π π = −ππ − π . − ππ There are two stator rotating magnetic fields, the first one has the synchronous speed ππ and the second one has the synchronous speed −ππ . This first stator rotating magnetic field generates voltages and currents having a frequency π π 50 π»π§. This generates a rotor rotating magnetic field having a speed π π ππ with respect to the rotor. Since the rotor has a speed (1 − π π ) ππ , the rotor rotating field has an absolute speed ππ . The second stator rotating magnetic field generates voltages and currents having a frequency π π 50 π»π§. This generates a rotor rotating magnetic field having a speed π π (−ππ ) with respect to the rotor. Since the rotor has a speed (1 − π π )(−ππ ) = (1 − π π )ππ , the rotor rotating field has an absolute speed −ππ . This means there are four rotating magnetic fields: - Rotating field 1: stator field with speed ππ Rotating field 2: stator field with speed −ππ Rotating field 3: rotor field with speed ππ Rotating field 4: rotor field with speed −ππ . The interaction of the rotating fields 1 and 3 gives a constant torque. This is the desired useful torque (coming from the direct torque speed characteristic having a positive direction). The interaction of the rotating fields 2 and 4 also gives a constant torque. This is the torque coming from the inverse torque speed characteristic having a negative direction. The interaction between the rotating fields 1 and 4 gives a pulsating torque (which is on average zero). The interaction between the rotating fields 2 and 3 gives a pulsating torque (which is on average zero). Due to these pulsating torques, the torque of the induction motor is no longer constant. 6.2: The neutral conductor Consider a three phase load (wye connected with neutral conductor) as visualized in Figure 8. The impedances π1 , π2 and π3 are different. In general, this implies the currents ππ1 (π‘), ππ2 (π‘) and ππ3 (π‘) have different magnitudes and have phase differences equal to 120°. Figure 8: Non symmetric three phase load Using Kirchoff’s circuit law, ππ (π‘) = ππ1 (π‘) + ππ2 (π‘) + ππ3 (π‘). The currents ππ1 (π‘), ππ2 (π‘) and ππ3 (π‘) are the sum of a direct component, an inverse component and a homopolar component. The direct component contains three currents which have the same magnitude and which have phase differences equal to 120°. The sum of these three currents is zero implying there is no current in the neutral conductor due to the direct current component. The inverse component contains three currents which have the same magnitude and which have phase differences equal to 120°. The sum of these three currents is zero implying there is no current in the neutral conductor due to the inverse current component. The homopolar component contains three currents which have the same magnitude and which have the same phase implying their sum is not zero. Due to the homopolar phase current component, a current is flowing in the neutral conductor. This current in the neutral conductor is three times as large as the homopolar component in the phase current. Indeed, since Μ πΌπβ 1 1 1 Μ ] = [1 π [πΌππ 3 1 π2 Μ π πΌπ Μ 1 πΌπ1 2 ] [πΌ Μ ] , π π2 π πΌπ3 Μ Μ , πΌππ Μ and πΌππΜ . Since πΌπΜ = πΌπ1 Μ + πΌπ2 Μ + πΌπ3 Μ and it is possible to calculate πΌπβ Μ πΌπ1 1 Μ ] = [1 [πΌπ2 1 Μ πΌπ3 1 π2 π Μ 1 πΌπβ Μ ] π ] [πΌππ 2 π πΌππΜ one indeed obtains that Μ . πΌπΜ = 3 πΌπβ In case the homopolar current component is large, this has to be taken into consideration when choosing the neutral conductor. The cross section of the neutral conductor has to be sufficiently large in order to reduce the heat dissipation. Traditionally, the neutral conductor has a smaller cross section than the phase conductors. This remains a good choice when there is no or only a limited non symmetry in the grid. The current in the neutral conductor is zero or small. However, when there is an important non symmetry in the grid, especially when there is an important homopolar component, it can be necessary to take the cross section of the neutral conductor the same as the cross section of the phase conductors. In extreme cases, it can be necessary to give the neutral conductor a larger cross section than the phase conductors. 7: Reducing non symmetries in a grid 7.1: Compensating a non symmetrical load It is important to reduce or (if possible) avoid non symmetries in a three phase grid. Using symmetric three phase loads instead of single phase loads is useful. When using single phase loads, it is important to spread them over the three different phases. In case there is only one non symmetrical load as visualized in Figure 9 which can not be changed, it is possible to compensate the non symmetry by using capacitors and inductors (resistors are not used since resistors consume active power). Figure 9: Non symmetrical load As visualized in Figure 9, the current in phase 1 equals zero. In phase 2 and phase 3, the currents have the same magnitude but there is a phase difference of 180°. These are important non symmetries. Although there is no homopolar current component (there is no neutral conductor), the direct current component and the inverse current component have the same magnitude. Figure 10: Compensation of a non symmetrical load Suppose in Figure 10 the admittance πΜ 23 = 1 πΜ 23 is the useful non symmetric load (e.g. a resistor π as visualized in Figure 9 giving πΜ 23 = 1⁄π ) and the admittances πΜ 12 and πΜ 31 are used to obtain a symmetrical load behavior. In order to avoid heat losses, the admittances πΜ 12 and πΜ 31 are capacitors or inductors (containing no resistors). Suppose the Μ πΏ1πΏ2, π Μ πΏ2πΏ3 and π Μ πΏ3πΏ1 supplied by the grid are symmetrical i.e. they only contain a direct voltages π component. By an appropriate choice of the admittances πΜ 12 and πΜ 31 , the currents supplied by the grid must will only contain a direct component. The grid currents can be written as Μ πΌπ1 πΜ 12 Μ [πΌπ2 ] = [−πΜ 12 Μ 0 πΌπ3 0 πΜ 23 −πΜ 23 −πΜ 31 0 ] Μ π31 Μ πΏ1πΏ2 π πΜ 12 + πΜ 31 Μ πΏ2πΏ3 ] = [ −πΜ 12 [π Μ πΏ3πΏ1 π −πΜ 31 −πΜ 12 πΜ 23 + πΜ 12 −πΜ 23 −πΜ 31 −πΜ 23 ] Μ π31 + πΜ 23 Μ π1 π Μ π2 ] [π Μ π3 π Μ πΏ1πΏ2 = π Μ π1 − π Μ π2 , π Μ πΏ2πΏ3 = π Μ π2 − π Μ π3 and π Μ πΏ3πΏ1 = π Μ π3 − π Μ π1 . Since with π Μ π1 π 1 1 Μ π2 ] = [1 π2 [π 1 π Μ π3 π Μ πΌπ1 Μ β 1 π 1 1 Μ ] = [1 π 2 Μ π ], [πΌπ2 π ] [π Μ π π2 π 1 π Μ πΌπ3 1 πΌβΜ π ] [πΌπΜ ] π2 πΌπΜ One obtains πΌβΜ 1 1 [πΌπΜ ] = [1 3 1 πΌπΜ giving 1 π π2 πΜ 12 + πΜ 31 1 2 ] [ −π Μ 12 π π −πΜ 31 −πΜ 12 Μ π23 + πΜ 12 −πΜ 23 −πΜ 31 1 −πΜ 23 ] [1 πΜ 31 + πΜ 23 1 1 π2 π 1 π] π2 Μ β π Μ [ππ ] Μ π π πΌβΜ 0 Μ [πΌπ ] = [0 0 πΌπΜ 0 0 πΜ 12 + πΜ 23 + πΜ 31 −(ππΜ 12 + πΜ 23 + π2 πΜ 31 )] −(π2 πΜ 12 + πΜ 23 + π πΜ 31 ) πΜ 12 + πΜ 23 + πΜ 31 Μ β π Μ π ] [π Μ π π Μ π , the In order to avoid an inverse current component πΌπΜ due to the direct voltage component π condition π2 πΜ 12 + πΜ 23 + π πΜ 31 = π2 πΜ 12 + 1 + π πΜ 31 = 0 π Since the real parts of πΜ 12 and πΜ 31 are zero (no resistive part in order to avoid heat losses), there are two unknowns πΌπ(πΜ 12 ) and πΌπ(πΜ 31 ) which satisfy 2π 2π 1 2π 2π (πππ ( ) − ππ ππ ( )) π πΌπ(πΜ 12 ) + + (πππ ( ) + ππ ππ ( )) π πΌπ(πΜ 31 ) = 0 3 3 π 3 3 or equivalently 2π 1 2π π ππ ( ) πΌπ(πΜ 12 ) + − π ππ ( ) πΌπ(πΜ 31 ) = 0 3 π 3 { 2π 2π πππ ( ) πΌπ(πΜ 12 ) + πππ ( ) πΌπ(πΜ 31 ) = 0 3 3 This implies πΌπ(πΜ 12 ) = − 1 1 1 = − π 2 π ππ (2π) π √3 3 πΌπ(πΜ 31 ) = + 1 1 1 = + π 2 π ππ (2π) π √3 3 Therefore, πΜ 12 is an inductor having a value πΏ= π √3 π and πΜ 31 is a capacitor having a value πΆ= 1 π π √3 . Using these inductor πΏ and capacitor πΆ values as visualized in Figure 11, the grid current will only contain a direct component in case the applied grid voltage only contains a direct voltage component. Figure 11: Compensation of a non symmetric load In case the applied grid voltage has another phase sequence (it only contains an inverse voltage component), a direct current component πΌπΜ will be avoided when π πΜ 12 + πΜ 23 + π2 πΜ 31 = 0 Μ π , only an inverse current component πΌπΜ will flow. Verify Due to this inverse voltage component π what values for πΜ 12 and πΜ 31 are needed. 7.2: Avoiding and reducing non symmetries due to photovoltaic inverters Avoiding non symmetries is also important to reduce the heat losses in the grid. Consider a three phase grid without a neutral conductor and suppose a photovoltaic inverter injects power into the grid. There exist single phase inverters and there exist three phase inverters. Suppose the grid has a line voltage π and the inverter injects an apparent power π. When a single phase inverter is used, a current πΌ1 will flow in two phase conductors. When a three phase inverter is used, a current πΌ3 will flow in all three phase conductors. This means π = π πΌ1 = √3 π πΌ3 implying πΌ1 = √3 πΌ3. When π is the resistance of a phase conductor, then the grid losses in case of a single phase inverter equal ππππ π ,1πβ = 2π πΌ12 and the losses in case of a three phase inverter equal ππππ π ,3πβ = 3π πΌ32 = ππππ π ,1πβ . 2 When a single phase photovoltaic inverter is used, the heat losses in the grid are doubled in comparison with a three phase inverter. A similar situation occurs when another renewable and/or distributed energy source generates power or when the three phase grid is loaded with a single phase load (in parallel with a line voltage). When using three identical photovoltaic panels with identical single phase inverters, it is also possible to obtain a symmetrical situation (in case the panels produce the same electrical power). Figure 12 visualizes a three phase grid (having a line voltage of 230 π) where three single phase inverters inject their power. The currents πΌπ , πΌπ and πΌπ have a magnitude which equals √3 times the magnitude of the inverter currents πΌπ−π , πΌπ−π and πΌπ−π . Figuur 12: Single phase inverters injecting power in a three phase grid For instance the Sunny Tower of SMA Solar Technology allows to mount a number of single phase inverters (Sunny Boy inverters) which allows them to inject their power symmetrically into a three phase grid. In case the grid has a line voltage of 230 π, the single phase inverters are placed in parallel with the line voltage. A Sunny Tower with six Sunny Mini Central 8000TL inverters (each having a nominal AC power of 8 ππ) is able to inject a total power of 48 ππ into the grid. Figuur 13: The use of the Sunny Tower When a three phase grid with a line voltage of 400 π is used, then a neutral conductor is needed to inject power using a 230 π single phase inverter. The inverter is mounted between a phase conductor and the neutral conductor. A symmetrical injection of the power is obtained by injecting the same power using a first inverter between πΏ1 and π, a second inverter between πΏ2 and π and finally a third inverter between πΏ3 and π. Figure 14: Single phase inverter between phase conductor and neutral conductor When a three phase grid has a symmetrical load, no current is flowing in the neutral conductor which implies this neutral conductor often has a smaller cross section than the phase conductors. In order to limit the heat dissipation in the grid conductors, a symmetrical power injection is really useful when the neutral conductor has a smaller cross section. Indeed, when the three single phase inverters inject their power using πΏ1 − π, πΏ2 − π and πΏ3 − π as already mentioned and they all inject the same power, than they inject currents with the same magnitude which have phase differences of 120°. This implies no current is flowing in the neutral conductor and this avoids heat losses in this neutral conductor. Notice Figure 14 not only visualizes the single phase inverter. Between the photovoltaic panels and the inverter, an EMC/EMI filter has been mounted. This filter reduces the high frequent currents in the photovoltaic cells due to the inverter. This avoids an accelerated aging of the photovoltaic cells. There is also an EMC/EMI filter mounted between the inverter and the grid. This filter limits the emission of high frequent disturbances into the grid in order to comply with national and/or international EMC normalizations (e.g. see http://www.schaffner.com). The principle that it is useful to spread single phase inverters of photovoltaic installations over the phases is a very general principle. This principle is valid for all kind of distributed resources like the Hydro Boy (an inverter fed by a fuel cell) and the Windy Boy (which injects the power of a micro wind turbine) of SMA Solar Technologies. References J.D. Glover, M.S. Sarma, Power System: Analysis and Design, Brooks/Cole, Pacific Groove, California USA, 2002. J. Peuteman, Lokale productie: enkelfasig of driefasig?, Vlaams Elektro Innovatiecentrum VEI, Technologiewacht, 2010. J. Willems, Elektrische netten, Wetenschappelijke uitgeverij en boekhandel, Gent, 1978.