WO2012103889A2

WO2012103889A2 - Potential compensation for electric inverter

Info

Publication number: WO2012103889A2
Application number: PCT/DK2012/000009
Authority: WO
Inventors: Uffe Borup; Frerk HASSE
Original assignee: Danfoss Solar Inverters A/S
Priority date: 2011-01-31
Filing date: 2012-01-30
Publication date: 2012-08-09
Also published as: WO2012103889A3

Abstract

The invention relates to an electric frequency changing device (2, 3) for changing the frequency of an electric power supply (6, 8). The electric frequency changing (2, 3) device comprises at least one fast-acting electric potential changing means (11) for changing the electric potential (U) of at least one part of the electric frequency changing device (2, 3) relative to ground potential.

Description

Potential compensation for electric inverter

The invention relates to an electric frequency changing device for changing the frequency of an electric power supply, in particular for increasing the frequency of an electric power supply. Furthermore, the invention relates to a photovoltaic power plant. Even further, the invention relates to a method of controlling a photovoltaic power plant, comprising at least one frequency changing device for changing the frequency of an electric power supply.

Increasing concern about environmental problems (in particular with respect to the emission of carbon dioxide) and the associated increase public funding opportunities, the number of photovoltaic power plants is on the increase. In photovoltaic power plants, electric power is directly produced from sunlight, using photovoltaic cells. The size of such photovoltaic power plants varies from relatively small arrangements (such as arrangements that are placed on the roof of a living house for (partially) supplying the house with electrical energy) to large-size complexes with a generation electric power capacity of several megawatts or even more.

Photovoltaic cells can only produce direct currents. However, presently used electrical consumers are usually designed to use alternating current (typically 230 V/50 Hz or 110 V/60 Hz). Therefore, it is necessary to alter the direct electric current that is produced by the photovoltaic cells into an alternating electric current. Another advantage of alternating currents is that they can be easily transformed to high voltages, using transformers. High voltages, however, prove to be essential for transporting electric power over long distances without excessive losses. It is therefore a particular advantage for large-size photovoltaic power plants to be able to supply alternating currents of appropriate voltages and frequencies. According to the state of the art, devices known as inverters are used for generating alternating current (AC) from direct current (DC). Such devices generate an alternating signal by switching the output repeatedly between positive and a negative DC bus lines in a way that forms the required output to the load. A frequently used method of doing this is to use a technique known as pulse width modulation (PWM), where a constant frequency square wave switching pattern is generated with a duty cycle which varies in (for example) a sinusoidal manner, yielding an averaged signal which corresponds to the desired signal to the load.

Two commonly used switching device topologies are the H-bridge design and the NPC-design (where the NPC stands for "Neutral Point Clamped"). In particular, those two topologies are used when so-called "transformerless inverters" are used.

In the H-bridge design, each output line is separately connected to the positive and the negative DC buses by a switching device. A particular disadvantage of the H-bridge design is that a voltage with respect to ground which has half of the output voltage of the generated alternating current and about the same frequency of the generated alternating current will be superimposed on the input direct current potential. This effect occurs despite the fact that the direct current voltage component on the input side of the inverter remains essentially constant. In the single-phase NPC design, one output line is clamped between the two DC buses, and the second output line is switched between the two DC buses and the neutral bus. Although the NPC design is more complex, and hence more expensive, it has the advantage that only a comparatively small AC component with respect to ground is superimposed on the input DC. This AC component has usually twice the frequency of the output frequency of the generated alternating current. This feature is sometimes referred to as "quiet rails". Nevertheless, there is still an AC component on the DC input side that can cause problems. The above examples serve to demonstrate that two different inverter topologies lead to AC components relative to ground at the DC inputs. It should be said that the presence of such AC components is not restricted to these two inverter topologies, but is present in other inverter topologies as well. The two examples above serve to illustrate that such AC components may have different sizes and different frequencies which are dependent upon the topology of the inverter.

One of the major disadvantages of an AC component on the input side of the electric inverter (i.e. of an alternating current part on the conducting parts of the relatively large scaled photovoltaic module arrays) is that it may lead to a capacitive leakage current. This current is generated between the photovoltaic modules, having a very large area, and the ground. Further, this capacitive leakage current is very hard to deal with, since the capacity between the photovoltaic modules and the ground can vary widely and - even worse - its variations are usually highly unpredictable, since the capacitance varies significantly with temperature or humidity (i.e. the presence of rain, mist, fog or snow). Even if the AC component can be predicted, for example from

modelling, it would not be possible to predict the resulting leakage current. There is therefore a great advantage in avoiding such AC components on the DC side of the inverters. The problem could be solved by using transformers that galvanically isolate the (DC) input side from the (AC) output side. However, such galvanically isolating inverters are more expensive, weigh more and are less efficient as compared to inverters of the design without such isolation, these being typically referred to as "transformerless" inverters.

The presence of leakage currents also influence the use of residual current devices (RCDs) in power plants such as those described here. RCDs are devices that disconnect a circuit whenever it detects that the electric current is not balanced between a pair of conductors. Such an imbalance is sometimes caused by current leakage through the body of a person who touches a live part of the circuit and who is also grounded. In some situations a lethal shock can result, and RCDs are designed to disconnect fast enough to prevent harm. In a power plant protected by an RCD, the power plant will be disconnected from the grid if a residual current exceeds the triggering level of the RCD. Leakage currents produced by AC components relative to ground as described above may cause residual currents large enough for this to happen. Even if the trigger level of the RCD is not reached, the dynamic voltage components, and thus leakage currents, could put the life of a photovoltaic module installer at risk if he or she touches the photovoltaic module. Additionally, uncertainty lies within the type of RCD since its bandwidth to trigger is not well defined - meaning that the harmonic content of some topologies may not trigger some RCDs even if above the typical 30 mA trigger level.

Another problem inherent in the use of switching, such as the techniques used in inverters, is the production of high-frequency noise as current flows are abruptly switched. Such high-frequency noise can lead to problems with electromagnetic compatibility (EMC), since the noise can be radiated by the switching circuits and cause disturbance to sensitive electrical equipment in the neighbourhood.

It is therefore an object of the invention to suggest an electric frequency changing device for changing the frequency of an electric power supply that is improved over presently known electric frequency changing devices. It is another object of the invention to suggest a method of controlling a photovoltaic power plant, comprising at least one frequency changing device for changing the frequency of an electric power supply, that is improved over presently known methods.

The invention achieves these objects.

It is suggested to design an electric frequency changing device for changing the frequency of an electric power supply in a way that it comprises at least one fast-acting electric potential changing means for changing the electric potential of at least one part of the electric frequency changing device relative to ground potential. In particular, said electric frequency changing device can be designed in a way that it can be used for increasing the frequency of an electric power supply. In principle, a potential changing means as such for changing the electric potential of at least one part of an electric and/or electronic device (including at least one part of an electric frequency changing device) relative to ground potential is known. For example, a normal grounding contact can be used as an electric potential changing means. More elaborately, the use of voltage sources for setting the electric potential of certain parts of an

electric/electronic device to a defined potential with respect to ground potential is also known. However, a simple grounding contact will of course fix the respective part to a fixed static level, namely to ground potential. Even if the electric voltage sources are used for setting the respective part of an electric device to a certain level, so far only quasi-static potential changing means have been used. For example, if a battery has been used for changing the potential versus ground potential, the potential will of course change when the battery is emptied. However, this occurs regularly on a timescale of weeks, months or even years. Further, sometimes some kind of a calibration has already been employed for the potential changing means. However, usually such a

calibration has been either only used for an initial setup of the arrangement (and/or after maintenance of the respective arrangement) or has been used for compensating for environmental changes, for example to adapt the potential according to the varying solar input during the day. However, such changes also occur on a quasi-static timescale, i.e. at a timescale of minutes, quarters of an hour, hours, a couple of hours or even days, weeks and months. Presently, however, it is suggested to vary the electric potential changing means on a (in comparison) very fast timescale - and hence to design the electric potential changing means as a fast-acting electric potential means. In particular, the timescale of the changes of the electric potential should be at least

approximately of the order of a frequency that can occur with at least one of the alternating power supply frequencies involved with the electric frequency changing device, or, for example, of transient frequencies that occur as a side- effect of the control methods used in the electric frequency changing device. This statement can include the meaning that the sampling

frequency/operational frequency of the fast-acting electric potential changing means can be (essentially) equivalent to at least one of the above-mentioned frequencies. However, this statement can also be understood in a way that the sampling frequency/operational frequency of the fast-acting electric potential changing means is chosen way that it can compensate "electric noise" of at least one of the above-mentioned frequencies. Usually, for effectively compensating "electric noise", a higher sampling frequency/operational frequency is needed. Typically, a factor of 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, 14, 15 and/or 20 is needed for this. The aforementioned alternating current frequencies that are involved with the electric frequency changing device can be, in particular, an alternating current on the output side of the electric frequency changing device and/or the frequency of an alternating electric current on the input side of the electric frequency changing device. By the term "approximately in the order", a factor of 1 , 2, 5, 10, 20, 50 and/or 100 (in both "directions" and/or with respect to the input and/or output terminals) can be encompassed. Although the speed of the fast-acting electric potential changing means can be slower, as compared to the lowest frequency of one, several or all of the electric frequencies involved (in particular on the input side and/or on the output side), it is of course preferred if the fast-acting electric potential changing means is at least as fast or preferably faster as compared to such a frequency (such frequencies). In particular, if the fast-acting electric potential changing means is faster (presumably with some additional "sampling factor") as compared to such a frequency (such frequencies), even harmonics (at least the lower harmonics) can be dealt with by the fast-acting electric potential changing means.

Quantitatively, the fast-acting electric potential changing means can operate at 25 Hz, 30 Hz, 50 Hz, 60 Hz, 100 Hz, 120 Hz, 150 Hz, 180 Hz (or even higher harmonics of the "main" frequency 50 Hz or 60 Hz, respectively) and/or 300 Hz, 600 Hz, 1 kHz, 2 kHz, 5 kHz, 10 kHz, 50 kHz or even higher harmonics of the signals generated as a side-effect of one or more control methods used by the electric frequency changing device. Although it is in principle possible that the electric frequency changing device is used for changing the frequency of a first alternating power supply into a second alternating power supply with another frequency, it is preferred if the electric frequency changing device is designed in a way that it can be used for converting DC power into AC power. Additionally or alternatively it is preferred if said electric frequency changing device comprises at least one inverter device, preferably at least one transformerless inverter device. Although it is in principle possible to design the electric frequency changing device in other ways as well, inverter devices are very efficient and comparatively cheap. In particular, transformerless inverter devices are usually particularly efficient and comparatively cheap for performing a frequency change. The generation of AC power from DC power is a very common task, since electric batteries and photovoltaic cells (just to name two different devices) are only able to produce DC power, while AC power is required by standard consumers and/or by transformers for transforming the voltage of the electricity produced.

Furthermore, it is preferred if the electric frequency changing device is designed in a way that it can be used at least in part and/or at least at times to convert and/or to generate at least at times a periodically changing electric power supply, in particular with a square wave electric current, a sawtooth wave-like electric current and/or a sinusoidal-like electric current and/or if the electric frequency changing device is arranged in a way that it can be used at least in part and/or at least at times to convert and/or to generate an alternating electric current with an essentially constant frequency. Such devices are particularly advantageous when it comes to transforming DC power from a battery and/or from a photovoltaic module into a standard network current (with 50 Hz or 60 Hz, for example). However, the device could also be used for changing the frequency of a 50 Hz power network to 16 2/3 Hz, for example (which is the frequency of the electric supply of railway systems in Germany, Austria and Switzerland) or vice versa.

Even more, it is preferred to design the electric frequency changing device in a way that at least one of said fast-acting electric potential changing means is designed and arranged in a way that it can compensate at least in part and/or at least at times frequency-induced effects of at least one alternating electric current generated by said electric frequency changing device. As an example, if an inverter of the H-bridge topology is used, the fast-acting electric potential changing means should be able to operate at 50 Hz (60 Hz), if electric power of 50 Hz (60 Hz) is generated. Likewise, if an NPC-type typology is used, the fast- acting electric potential changing means should be able to operate at 100 Hz (120 Hz) if electric power with the frequency of 50 Hz (60 Hz) is generated by the electric frequency changing device. Again, the expression "operate" can refer to the numbers given, as well as to higher numbers, taking a certain "sampling rate factor" into account. The indicated frequencies of the output powers are the typical standard frequencies of electric supply networks worldwide. Of course, although the "basic" frequencies to be compensated by the fast-acting electric potential changing means have been mentioned, it is also possible that the fast-acting electric potential changing means is designed to be able to compensate the first few higher harmonics of such frequencies.

It is furthermore suggested to design the electric frequency changing device in a way that at least one of said fast-acting electric potential changing means is designed and arranged in a way that at least a part of the electric frequency changing device is set to a certain potential relative to ground potential, preferably to a certain, essentially constant potential relative to ground potential. Although it is possible that this certain potential can vary with time (for example in a way that the "normally occurring" amount of variation is

decreased, where "normally occurring" means the amount of variation that would occur without the fast-acting electric potential changing means and/or if said fast-acting electric potential changing means is switched off), it is preferred if the resulting potential is kept at an essentially constant level. This way, electric losses can usually be avoided best. Furthermore, the certain potential can be essentially zero (so that the respective parts of the electric frequency changing device would be essentially set to ground potential). Additionally and/or alternatively a certain offset from ground potential can be envisaged. If a certain offset from ground potential is envisaged, it is normally possible that degradation effects with thin-film photovoltaic cells and/or with rear contacted photovoltaic cells can be reduced or even essentially avoided. It has been observed that such photovoltaic cells are prone to damage, if they are operated in part above/below ground potential. The at least one part that is set to a certain potential relative to ground potential can be essentially any part of the electric frequency changing device. In particular, it can be an external connection of said electric frequency changing device, in particular on an input side and/or on an output side of the electric frequency changing device. In particular, it can be a DC-part and/or an AC-part of the electric frequency changing device. It should be mentioned that it is possible that the part of the electric frequency changing device that is set to a certain potential relative to ground and the part, where the fast-acting electric potential changing means is (initially/directly) connected to, may differ. A further preferred embodiment of the electric frequency changing device can be achieved if said device is designed and arranged in a way that it can be used for changing the frequency of power generated by at least one

photovoltaic cell, preferably from a plurality of photovoltaic cells and/or from at least one photovoltaic module, preferably from a plurality of photovoltaic modules. This way, it is easy to build a very advantageous (part of a) photovoltaic power plant. It has to be noted that in such power plants, very large areas are covered by the photovoltaic cells/photovoltaic modules.

Therefore, a use of the presently suggested electric frequency changing means is very effective in reducing capacitive currents and hence in improving the efficiency of the resulting arrangement.

Furthermore it is suggested to design the electric frequency changing device in a way that at least one electric frequency changing device, preferably at least one inverter device, comprises a plurality of electric interfaces and/or in a way that said one electric frequency changing device comprises a plurality of inverter devices. This way, it is particularly simple to scale up the overall dimension of the resulting arrangement (for example of a photovoltaic power plant). It has to be noted that usually electric frequency changing devices, photovoltaic cell modules and/or inverter devices can only be scaled up to a certain size for economic and/or technical reasons. With the suggested preferred design, it is easy to reduce the cost and increase the simplicity of such scaling up tasks, since it is often sufficient to use only a single fast-acting electric potential changing means.

Furthermore, it is suggested to design the electric frequency changing device with at least one transferring device to an external electric grid, preferably with at least one galvanically isolating transferring device, more preferably with at least one transformer device. This way it is possible to use the electric frequency changing device (and the devices connected to it) for a local arrangement, but also as a part of a wide area electrical supply system. In particular, if the transferring device is galvanically isolating (for example, if it comprises a transformer) the action of the fast-acting potential changing means can be performed essentially independent of the wide area network.

According to an embodiment of the electric frequency changing device, at least one of said fast-acting electric potential changing means is directly connected to at least one electric main conductor device, in particular to at least one main AC electric conductor device. Such a conductor device can be a wire, where a major electric current flows through. In particular, it can be a wire, where DC current of (part of) a photovoltaic module/photovoltaic power plant and/or AC current of (part of) a photovoltaic module/photovoltaic power plant is routed through. This can be a wire (or several wires), where an electric phase is present (quite often referred to as a "hot wire") and/or a neutral wire. With the notion "main" conductor device is usually meant that the respective conductor device is not (only) used for controlling purposes, but instead for transmitting a significant amount of electric power. Preferably, the electric frequency changing device is equipped with at least one potential measuring means. This way it is possible to measure the actual offset of the respective potential with regard to ground potential. Therefore, it is easy to provide the fast-acting electric potential changing means with an input signal, in particular with a regulating signal. This way, the electric potential can be better controlled. In particular, a fast-acting electric potential changing means and a potential measuring means can be arranged essentially "in parallel". Another preferred embodiment of the electric frequency changing device can be achieved if at least one voltage measuring device, at least one current measuring device, and/or at least one interface device is provided. The at least one interface device is preferably connected to at least one of said fast-acting electric potential changing means and/or to at least one of said inverter devices. The aforementioned measuring devices have proven to be particularly useful when it comes to generating a signal for regulating the fast-acting electric potential changing means. Using an interface device to at least one fast-acting electric potential changing means (in particular to an inverter device) the information gained by such devices (or other means as well) can be easily transmitted to the respective device(s). The information can be transmitted using a bus system, for example. Especially in the latter case, it is particularly easy to even use a readout signal from the fast-acting electric potential changing means for some kind of a feedback. Furthermore, it is particularly simple to use external signals (for example by a control terminal or the like) to get some control over the whole system. It is even possible (although usually not preferred) that a regulating signal is generated, using the actual switching information of the switching device. In particular, a model of the particular inverter topology in use can be utilised to calculate an appropriate regulating signal (which might provide at least a partial compensation of the AC

components relative to ground appearing on the DC input) from the actual switching behaviour of the switching device. Such an appropriate regulating signal may be a signal indicating the phase of the switching pattern being used in the frequency changing device. A preferred embodiment can be achieved if such a switching model is combined (refined) with simultaneous measurements of voltages and/or currents.

Preferably, at least one controller device, even more preferably at least one electronic controller device, is provided for the electric frequency changing device. This way, it is possible to easily create very advantageous and precise controlling signals for the at least one fast-acting electric potential changing means. Furthermore, a feedback output can be generated by said at least one controller device, as well.

Furthermore, a photovoltaic power plant is suggested that comprises at least one electric frequency changing device of the aforementioned type. This will result in a particularly advantageous photovoltaic power plant. In particular, losses due to leakage currents can be largely avoided. Furthermore, the photovoltaic power plant will also show the same effects and advantages as previously described, at least by analogy.

Furthermore, it is suggested to perform a method of controlling a photovoltaic power plant that is comprising at least one frequency changing device for changing the frequency of an electric power supply in a way that the electric potential of at least parts of the photovoltaic power plant relative to ground potential, in particular of at least parts of the electric frequency changing device relative to ground potential, is altered at least in part and/or at least at times by at least one fast-acting electric potential changing means. In particular, it is possible to design the at least one frequency changing device according to the previously suggested design of an electric frequency changing device. When performing the suggested method, the features, effects and advantages, already described in the context of the electric frequency changing device can be achieved as well, at least by analogy. Furthermore, the method can be modified in the already described sense as well, at least by analogy.

In particular, it is possible to perform the method in a way that the alteration of the electric potential of at least a part of the photovoltaic power plant is performed in a way that the respective part is set to an essentially constant potential relative to ground potential. This way, it is possible to minimise (or even to completely avoid) electric losses due to capacitive electric currents. Furthermore, the dangers involved with a (comparatively large) AC component to ground of the electric voltage on the photovoltaic cell-side can be reduced or even completely avoided.

The present invention and its advantages will become more apparent, when looking at the following description of possible embodiments of the invention, which will be described with reference to the accompanying figures, which show:

Fig. 1 : a possible embodiment of a solar power plant in a schematic view; Fig. 2: a simplified circuitry of a first embodiment of a solar power plant

comprising an H-bridge design in a schematic view;

Fig. 3: possible voltages at the input connectors of the H-bridge inverter of Fig.

2;

Fig. 4: a simplified circuitry of a second embodiment of a solar power plant comprising an NPC type inverter in a schematic view;

Fig. 5: the voltage at the input connectors of the NPC inverter of Fig. 4.

In Fig. 1 , a possible embodiment of a solar power plant 1 is shown in a schematic drawing. The embodiment is simplified, in particular with respect to larger sized solar power plants, and mainly given for illustrative purposes.

The solar power plant 1 comprises a plurality of solar collectors 4 (for example an arrangement of several solar modules; not shown). Each of the solar collectors 4 comprises a plurality of solar cells 5, being arranged on the surface of the respective solar collector 4. The arrangement, size, number, design and shape of the solar cells 5 on each of the solar collectors 4 is given for illustrative purposes and can differ widely. The same applies to the number, arrangement, size, design and shape of the solar collectors 4 themselves. Each of the solar collectors 4 is mounted on a fixture 7. Preferably, the fixture 7 comprises some moving mechanism, so that the solar collectors 4 can be moved to follow the sun during the day. Each solar collector 4 is connected to an inverter 2 via a DC connection 6. In the current embodiment, the inverters 2 are of a design, commonly referred to as a transformerless inverter design, more particularly they are of the so-called H-bridge transformerless inverter design. However, different designs are possible as well, in particular a transformerless inverter of the NPC design (see Figs. 4, 5) is possible.

The inverters 2 transform the direct current electric output of the solar collectors 4 into an alternating current. In the current embodiment, the alternating current output is single phase. In Fig. 1 , two conductors are shown. Of course, it would be possible as well to provide a third conductor for providing a protective earth, for example. Also, it would be possible to use a polyphase electric current, in particular, a three phase alternating electric current, a four phase alternating electric current, a five phase alternating electric current and so on. The alternating currents, produced by each one of the transformerless inverters 2 are sent to an AC-bus 8, connecting the inverters 2 and the transformer 9. The transformer 9 will increase the voltage on the AC bus 8 to the level of the electric long-distance network 10. Furthermore, the transformer 9 is galvanically isolating the solar power plant 1 from the electric long-distance network 10. Therefore, the parts of the solar power plant 1 , in particular the AC-bus 8, can be set to an electric potential with respect to ground potential, essentially independently from the setting of the electric long-distance network 10.

In the presently shown embodiment of Fig. 1 , one of the two conductors of the AC-bus 8 is connected to ground via a fast-acting DC source 11. Furthermore, the same wire of the AC-bus 8 is connected to a voltage sensor 12 that is measuring the potential of the respective conductor versus ground potential. Alternatively or additionally, another (second) voltage sensor 13 (of the same and/or of a different design) could be employed that is connected to one of the conductors of one of the DC connection 6. Of course, this second voltage sensor 13 could be alternatively connected to another DC connector 6 as well. Further, instead of voltage sensors 12, 13, current sensors could be used at different points in the circuit, for example in the DC line 6 or one of the AC bus lines 8. Furthermore, data buses 14 are provided for connecting the voltage sensors 12 and 13, the inverters 2 and the controllable DC source 11 to a controller unit 15. Of course, the detailed layout of the data buses 14 could be different, as well.

It is even possible that no voltage and/or current sensors 12, 13 are employed at all. Instead, information on the present switching behaviour of the switches 22 of the inverters 2, 3 (for example a data bit indicating phase) could be used for generating an input signal for the fast-acting DC source 11. In particular, the controller unit 15 can be programmed with an appropriate model of the inverter for generating such a correcting signal. Such a control method is commonly known as "open loop", as opposed to the "closed loop" method utilising sensors. Of course, usually such a simple open loop corrective action will only be of a limited quality. However, even with such a limited quality, a huge improvement over existing photovoltaic power plants 1 can be achieved.

Furthermore, a combination of such a model and measured data (in particular by voltage and/or current sensors) can be used, usually yielding improved results. Since the inverters 2 employed in the solar power plant 1 of Fig. 1 , are of the transformerless type, an offset or superimposed voltage produced by a fast- acting DC source 11 will not only change the potential to ground of the AC-bus 8, but also of the DC connections 6 and finally on the solar collectors 4, as well. To further clarify the problem, let us first assume that the DC source 11 forms a direct connection between the respective wire of the AC-bus 8 to ground while keeping in mind that the transformerless inverters 2 are of an H-bridge type. This would result in a significant AC component on the DC connections 6 of the solar power plant 1 (as illustrated in Fig. 3a). Likewise the (large) surfaces of the solar collectors 4 will also show a strong AC component. The large surfaces involved with the solar collectors 4 will lead to a non-negligible capacitive leakage current to ground. This is illustrated in Fig. 1 by the pseudo-capacitors 16, 17, where a first type of pseudo-capacitors 16 shows the capacitance of the upper surface of the solar collectors 4 to ground and a second type of pseudo- capacitors 17 shows the capacitance of the lower surface of the solar collectors 4 to ground. The pseudo-capacitors 16, 17 are not physically present as separate components; however, the effect of the surfaces of the solar collectors 4 will be similar to the situation where the solar collectors 4 are replaced by the pseudo-capacitors 16, 17.

This capacitive leakage current may cause the problems described above, including the unwanted triggering of RCDs.

A particular problem of the surface capacitance of the solar collectors 4, symbolised by the pseudo-capacitors 16, 17, is that the capacity is strongly variable and non-predictable. Specifically, the capacitance not only depends on the overall surface area, but also on environmental conditions like temperature, air humidity, ground humidity, precipitation (rain, fog, mist, snow) and - even worse - also on condensation or the like. Therefore, it is very hard (if possible at all) to compensate the effects of the varying capacitance. Even if the harmonic and transient content of the AC component to ground of the DC connections 6 is predictable through, for example, modelling of the inverter topology, it would still not be possible to predict the resulting leakage current since the capacitances 16, 17 would not be known in advance.

The embodiment illustrated in Fig 1 allows for control of the leakage current in the solar power plant 1 even if the inverter topology and/or the significant capacitances 16, 17 are liable to produce such a leakage current. Such control is enabled by compensating for the AC components to ground seen at the DC connections 6 actively by implementing the controllable DC source 11 somewhere on the AC side 8 or the DC side 6 of the inverter 2. The

controllable DC source 11 could be placed, as shown in Fig 1 , between on of the AC output lines 8 and ground, or, in another embodiment (not shown) between neutral (or any phase) and ground in a three-phase system or in yet another embodiment (not shown) between ground and one of the lines on the DC input side 6 of any of the inverters 2. Sensing of the AC component could be made at an appropriate position, as for example using the voltage

measuring devices 12, 13 shown in Fig 1 , and these measurements are then used by the controller 15 which tries to minimise the AC components by driving the fast-acting DC source in an appropriate manner. This could be by supplying a voltage of the required magnitude, but 180° out of phase with the AC component to be minimised. This will have the effect that the leakage current is also minimised. Another possibility is to measure and compensate the leakage current directly, by means of current measuring devices appropriately placed. Such current measuring devices (not shown) could be placed so as to measure currents in one or more of the DC input lines 6, or the AC bus lines 8.

The leakage current can be minimised by a single controllable DC source 11 since any AC or DC offset produced by the source relative to ground will have a direct influence on not only the AC bus lines 8 of the system shown in Fig 1 , but also the DC input side 6. This is a direct result of the inverters 2 being of a transformerless type. Thus voltage fluctuations on the DC input side 6 (of all the inverters 2) can be controlled by offsets applied to the AC bus lines 8. Applying a controllable DC source 11 to one of the DC input side lines 6 can be utilised in a similar way to control the AC components to ground appearing at all the other DC inputs 6.

The controllable DC source 11 could, in a further embodiment (not shown) be located inside the housing of the transformer 9 or in a suitable location inside a substation, where it could be connected between the star point of the power plant side of the transformer and earth.

In Fig. 2, the situation of the solar power plant 1 is shown in another view, that focuses on the circuitry of the H-bridge transformerless inverter 2 (while simplifying the solar collector part 18 and the AC output part 19). In particular, in Fig. 2, the basic design of an H-bridge type transformerless inverter 2 is depicted. In the H-bridge design, each output line 8 is separately connected via inductors 23 to the positive and the negative DC buses 21 using a switching device 22 (which are in this case electronic switches such as FETs or IGBTs). If the fast-acting DC source 11 is simply a direct electric connection to ground potential, the situation shown in Fig. 3a would occur. In Fig. 3a (and likewise in Fig. 3b) the evolving time t is shown on the abscissa 25, while the voltage of the two conductors of the DC connection 6 (i.e. the potential at the two input connections 21 versus ground potential) is shown on the ordinate 26. As can be seen from Fig. 3a, a substantial AC component 27 is superimposed on the DC part 28 on the DC connection 6. Usually, when the H-bridge design is employed, the frequency of the AC part 27 is the same as the frequency of the generated AC current (i.e. the frequency on the AC-bus 8). The peak-to-peak voltage of the superimposed AC component 27 is usually about half of the AC voltage on the AC-bus 8 (i.e. of the AC voltage generated) in this type of inverter topology. Now, the AC voltage on the AC-bus 8 (at least on one wire of the AC-bus 8) is measured by a voltage sensor 12 (as illustrated in Fig. 1). Alternatively, different sensors could be employed as well. Using this information, the controller unit 15 will generate a correcting signal with which the fast-acting DC source 11 will be regulated. In particular, the fast-acting DC source 11 will be driven in a way that on the input side 21 of the H-bridge type transformerless inverter essentially only the DC component 28 of the input voltage will remain. Thus, the situation in Fig. 3b will be achieved. Of course, by adding some additional correcting signal, the absolute "height" of the voltage lines in Fig. 3b can be adapted as well. For example, it is possible to use such an offset that all voltages are either above or below ground potential. Using such an offset, degradation effects or the like can be avoided (in particular if thin-film

photovoltaic modules are used) and/or charging effects can be avoided (in particular if back-contacted solar cells are used). Figs. 4 and 5 correspond to Fig. 2 and 3, respectively. However, this time a transformerless inverter 3 of the NPC design (NPC for Neutral Point Clamped) is used. In analogy to Fig. 2, details of the basic design of an NPC type transformerless inverter 3 are shown in the figure. In the NPC design, one output line 24 is clamped between the two DC buses by connection to the neutral point 32. The remaining output line 24 is switched between the two DC buses 21 and the neutral point 32 by the switches 22. Figure 4 also shows a booster circuit 29, comprising an input capacitor, and inductor a switch and a diode 30. The circuit can be used to increase the voltage across the output capacitors 31 above the input voltage appearing at the inputs 27 and produced by the solar collector 18. Such a booster circuit is not necessarily part of an inverter with an NPC topology, but it is often used in photovoltaic power plant applications. An advantage of NPC transformerless inverters 3 is that they do not show the very strong input AC components relative to ground that occur in connection with H-bridge type inverters. This configuration is usually referred to a "quiet rail" inverter.

Nevertheless, there is still some residual AC component to ground 27 that is superimposed on the DC component 28 of the voltage at the input connectors 21. The frequency of the AC component 27 is generally two times the frequency of the AC voltage generated (i.e. on the AC-bus 8). Once again, the situation, shown in Fig. 5a, would be observable if the DC source 11 is constant and/or simply a direct connection to ground.

If, however, input data, received by the data bus 14 and/or by the voltage sensors 12 and/or 13 (or other data as well) are used for generating a corrective signal via the fast-acting DC source 11 , the remaining AC component 27 ripples can be essentially compensated or reduced. This situation is shown in Fig. 5b. Once again, by adding a certain constant signal to the fast-acting DC source 11 , the "absolute height" of the potential lines in Fig. 5b could be changed as well.

Claims

C l a i m s

1. Electric frequency changing device (2, 3) for changing the frequency of an electric power supply (6, 8), in particular for increasing the frequency of an electric power supply (6, 8), characterised by at least one fast-acting electric potential changing means (11) for changing the electric potential (U) of at least one part of the electric frequency changing device (2, 3) relative to ground potential.

2. Electric frequency changing device (2, 3) according to claim 1 ,

characterised in that said electric frequency changing device (2, 3) is at least in part designed for converting DC power (6) into AC power (8) and/or in that said electric frequency changing device (2, 3) comprises at least one inverter device (2, 3), preferably at least one transformerless inverter device (2, 3).

3. Electric frequency changing device (2, 3) according to any of the preceding claims, characterised in that at least one of said fast-acting electric potential changing means (11) is designed and arranged in a way that it can compensate at least in part and/or at least at times frequency-induced effects of at least one alternating electric current (8) generated by said electric frequency changing device (2, 3).

4. Electric frequency changing device (2, 3) according to any of the preceding claims, characterised in that at least one of said fast-acting electric potential changing means (11) is designed and arranged in a way that at least a part of the electric frequency changing device (2, 3) is set to a certain potential (U) relative to ground potential, preferably to a certain, essentially constant potential (U) relative to ground potential.

5. Electric frequency changing device (2, 3) according to any of the preceding claims, characterised in that it is designed and arranged in a way that it can be used for changing the frequency of power generated by at least one photovoltaic cell (5), preferably from a plurality of photovoltaic cells (5) and/or from at least one photovoltaic module (4), preferably from a plurality of photovoltaic modules (4).

6. Electric frequency changing device (2, 3) according to any of the preceding claims, in particular according to any of claims 2 to 5, characterised in that said electric frequency changing device (2, 3), preferably at least one inverter device (2, 3), comprises a plurality of electric input interfaces (6, 21) and/or characterised by a plurality of inverter devices (2, 3).

7. Electric frequency changing device (2, 3) according to any of the preceding claims, characterised by at least one transferring device (9) to an external electric grid (10), preferably at least one galvanically isolating transferring device (9), more preferably at least one transformer device (9).

8. Electric frequency changing device (2, 3) according to any of the preceding claims, characterised in that at least one of said fast-acting electric potential changing means (11) is directly connected to at least one electric main conductor device (6, 8), in particular to at least one main AC electric conductor device (8) .

9. Electric frequency changing device (2, 3) according to any of the preceding claims, characterised by at least one potential measuring (12, 13) means.

10. Electric frequency changing device (2, 3) according to any of the preceding claims, preferably according to claim 9, characterised by at least one voltage measuring device (12, 13), at least one current measuring device and/or at least one interface device (14), wherein preferably at least one interface device (14) is connected to at least one of said fast-acting electric potential changing means (11) and/or to at least one of said inverter devices (2, 3).

11. Electric frequency changing device (2, 3) according to any of the preceding claims, characterised by at least one controller device (15), preferably at least one electronic controller device (15).

12. Photovoltaic power plant (1), characterised by at least one electric frequency changing device (2, 3) according to any of claims 1 to 11.

13. Method of controlling a photovoltaic power plant (1), comprising at least one frequency changing device (2, 3) for changing the frequency of an electric power supply (6, 8), in particular at least one frequency changing device (2, 3) according to any of claims 1 to 11 , characterised in that the electric potential (U) of at least parts of the photovoltaic power plant (1) relative to ground potential, in particular of at least parts of the electric frequency changing device (2, 3) relative to ground potential, is altered at least in part and/or at least at times by at least one fast-acting electric potential changing means (11).

14. Method according to claim 13, characterised in that the alteration of the electric potential (U) of at least a part of the photovoltaic power plant (1) is performed in a way that the respective part is set to an essentially constant potential (U) relative to ground potential.