AND ASSOCIATED METHOD

TECHNICAL FIELD

The invention relates to a converter cell for use in high voltage multilevel converters. An associated high voltage multilevel converter and method are also presented.

BACKGROUND

High Voltage Direct Current (HVDC) is increasing in usage due to a number of benefits compared to AC (Alternating Current) for power transmission. In order to connect a HVDC link or an HVDC grid to an AC grid, conversion needs to occur from DC (Direct Current) to AC or AC to DC. This conversion can for example be performed using voltage source converters (VSC).

In order to reduce harmonic distortion in the output of converters multilevel converters have been introduced, where the output voltages can assume several discrete levels. In particular, converters have been presented where a number of converter cells, each containing a number of switching elements and an energy storage element, are connected in series to form a variable voltage source.

There is always a drive to reduce losses of such converters, while still providing appropriate capabilities for switching control and fault handling.

SUMMARY

It is an object to provide a converter cell with reduced power losses, while controllability is essentially maintained, at least in a normal operating mode.

According to a first aspect, it is presented a converter cell arranged to be used in a high voltage multilevel converter. The converter cell comprises: an energy storage element; and a plurality of switching elements. The plurality of switching elements comprises at least one thyristor and a plurality of transistors; and each one of the at least one thyristor is provided in a position for a switching element in the converter cell where, during normal operation of the converter cell, the at least one thyristors is in a continuous conducting state. By placing the at least one thyristor in positions which are normally in a continuous conducting state, conduction losses are reduced, e.g. compared to a converter cell with only transistors. Moreover, clamping inductors are not required, since the main function of these is to allow arbitrary switching, which is not required for these at least one thyristors.

Normal operation may be operation when the converter cell is not in a failure mode. In normal operation, the converter cell is fully functional and all its switching elements can be controlled by a controller according to a switching schedule.

The at least one thyristors may be respectively provided in every topological position for a switching element in the converter cell where, during normal operation of the converter cell, the switching element is in a continuous conducting state. In other words, all such positions can be filled with thyristors.

Each one of the at least one thyristor may be an Integrated gate-commutated Thyristor, IGCT or a hybrid switch which allows for loss reduction with the above described mode of operation.

Each one of the plurality of transistors may be an insulated gate bipolar transistor (IGBT). Alternatively, each one of the plurality of transistors could be a power Metal-Oxide-Semiconductor Field-Effect-Transistor (MOSFET), power Bipolar Junction Transistor (BJT), Bi-mode Insulated Gate Transistor (BIGT) or any other suitable high power transistor component.

Each one of the at least one thyristor may be arranged to be set in a blocking state when a fault occurs on a DC side of the multilevel converter. The thyristor can be controlled to be in a blocking state, even without a clamping inductor. For example, this can be performed around when the current passes zero. According to a second aspect, it is presented a high voltage multilevel converter comprising a plurality of converter cells according to the first aspect.

According to a third aspect, it is presented a method for controlling a converter cell of a high voltage multilevel converter, the converter cell comprising: an energy storage element; at least one thyristor; and a plurality of transistors; wherein each one of the at least one thyristor is provided in a topological position for a switching element in the converter cell where, during normal operation of the converter cell, the at least one thyristors is in a continuous conducting state. The method comprises the step of: setting at least one thyristor, of the at least one thyristor, in a blocking state when a fault occurs on a DC side of the multilevel converter.

The step of setting at least one thyristor in a blocking state may be delayed until the current through the at least one thyristor is negligible. The step of setting at least one thyristor in a blocking state may be delayed until the current through the at least one thyristor passes through a zero value, i.e. until there is a zero crossing for the current of the thyristor in question.

The method may further comprise the step, prior to the step of setting at least one thyristor in a blocking state, of: detecting a fault in the DC side of the multilevel converter. The step of setting at least one thyristor in a blocking state may then be based on the step of detecting a fault.

It is to be noted that any feature of the first, second and third aspect may be applied to any other of these aspects, where appropriate. Generally, all terms used in the claims are to be interpreted according to their ordinary meaning in the technical field, unless explicitly defined otherwise herein. All references to "a/an/the element, apparatus, component, means, step, etc." are to be interpreted openly as referring to at least one instance of the element, apparatus, component, means, step, etc., unless explicitly stated otherwise. The steps of any method disclosed herein do not have to be performed in the exact order disclosed, unless explicitly stated.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is now described, by way of example, with reference to the accompanying drawings, in which:

Fig l is a schematic diagram illustrating one embodiment of a multilevel converter for converting between DC and AC;

Fig 2 is a schematic diagram show a converter arm of Fig l in some more detail; Figs 3A-B are schematic diagrams showing two embodiments of converter cells e.g. of the converter arm of Fig 2;

Fig 4 is a schematic diagram illustrating two full bridge cells connected in series;

Fig 5 is a schematic diagram illustrating a structure with a full bridge cell and a half bridge cell connected in series;

Fig 6 is a schematic diagram illustrating a structure similar to that of Fig 5, being an amended structure of a full bridge cell and a half bridge cell connected in series;

Fig 7 is a schematic diagram illustrating a structure similar but slightly different to that of Fig 5;

Figs 8A-B are schematic diagrams illustrating how the switching elements of Figs 3A-B and Figs 4-7 can be implemented; and

Figs 9A-B are flow charts illustrating methods which can be executed in embodiments presented herein. DETAILED DESCRIPTION

The invention will now be described more fully hereinafter with reference to the accompanying drawings, in which certain embodiments of the invention are shown. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided by way of example so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. Like numbers refer to like elements throughout the description. Fig l is a schematic diagram illustrating one embodiment of a multilevel converter 8 for converting between DC and AC. The DC connection comprises a positive terminal DC ⁺ and a negative terminal DC" and can be an HVDC connection. The AC connection in this embodiment is a three phase connection comprising three AC connections AC _a , ACb and AC _C and can be connected e.g. to an AC grid. While the multilevel converter 8 is here shown with three phases, the multilevel converter 8 can equally well have one, two, four or more phases.

Since there are three phases here, there are three phase legs 7a-c. The three phase legs 7a-c are connected in parallel between terminals DC ⁺ , DC- of the DC connection. In this embodiment, a first phase leg 7a comprises a first converter arm la, a first inductor 9a, a second inductor, 9b and a second converter arm lb connected serially between the terminals DC ⁺ , DC" of the DC connection. Analogously, a second phase leg 7b comprises a third converter arm lc, a third inductor 9c, a fourth inductor, 9d and a fourth converter arm id connected serially between the terminals DC ⁺ , DC" of the DC connection, and a third phase leg 7c comprises a fifth converter arm le, a fifth inductor 9e, a sixth inductor, 9f and a sixth converter arm if connected serially between the terminals DC ⁺ , DC" of the DC connection. The AC terminals AC _a , ACb and AC _C are provided between the inductors 9a-b of the respective phase legs 7a-c. Optionally, only one inductor is provided in each phase leg 7a-c. A main controller 6 is connected to all converter arms la-f and sends control signals to control the operation of the converter arms la-f. In this way, the main controller 6 controls the operation of the converter arms for conversion from AC to DC and/or from DC to AC. The multilevel converter 8 can be unidirectional in either direction between AC and DC or bidirectional. The multilevel converter 8 in this embodiment is a voltage source converter.

While the multilevel converter 8 is here shown with two converter arms for each phase leg, each phase could comprise any suitable number (l, 2, 3, etc.) of serially connected converter arms, controlled by the main controller 6. Fig 2 is a schematic diagram show a converter arm 1 of Fig 1 in some more detail. The converter arm 1 can be any one of the converter arms la-f shown in Fig 1. The converter arm 1 comprises any suitable number of converter cells 10, where each converter cell 10 is controlled by the main controller 6. As explained in more detail below, each converter cell 10 comprises one or more switching elements, each of which can be individually controlled by the main controller 6, e.g. to achieve a desired switching pattern for conversion between AC and DC of a converter device comprising the converter arm 1.

Figs 3A-B are schematic diagrams showing two embodiments of converter cells 10 e.g. of the converter arm 1 of Fig 2. Fig 3A illustrates a converter cell 10 implementing a half bridge structure. The converter cell 10 here comprises a leg of two serially connected active components in the form of switching elements Si, S2. The leg of two serially connected switching elements Si, S2 is connected in parallel with an energy storage element 16. The energy storage element 16 can e.g. be implemented using a capacitor, a super capacitor, an inductor, a battery, etc. The output voltage synthesized by the converter cell can thus either be zero or the voltage of the energy storage component 16, denoted Vc. The output voltage of the converter cell 10 is controlled by control signals from the main controller to the switching elements S1-S2, optionally via one or more gate units. Fig 3B illustrates a converter cell 10 implementing a full bridge structure. The converter cell 10 here comprises four switching elements S1-S4. An energy storage component 16 is also provided in parallel across a first leg of two switching elements S1-S2 and a second leg of two switching elements S3-S4. Compared to the half bridge of Fig 3A, the full bridge structure allows the synthesis of an output voltage capable of assuming both signs, whereby the voltage of the converter cell can either be zero, the voltage of the energy storage component 16, i.e. Vc, or a reversed voltage of the energy storage component 16, i.e. -Vc. The output voltage of the converter cell 10 is controlled by control signals from the main controller to the switching elements S1-S4, optionally via one or more gate units.

Fig 4 is a schematic diagram illustrating two full bridge cells loa-b connected in series. Each one of the two full bridge loa-b cells has the structure of the full bridge cell of Fig 3B. The first full bridge cell 10a thus comprises four switching elements S1-S4 and an energy storage element 16a. The second full bridge cell comprises its own four switching elements S5-S8 and an energy storage element 16b.

Table 1 below shows the switching patterns for the two full bridge cells loa-b of Fig 4, as controlled by the main controller for various desired output voltages Vout. A '0' denotes a blocking state and a '1' denotes a conducting state for the switching element in question.

Si S2 S3 S ₄ S5 S6 S7 S8 Vout Iarm

Normal operation

1 0 1 0 1 0 0 1 Vc

Iarm<o

1 0 1 0 0 1 0 1 0

OR

0 1 1 0 1 0 0 1 2VC

Iarm>o

0 1 1 0 0 1 0 1 Vc

Fault case (All switching elements are blocked) 0 0 0 0 0 0 0 0 2VC Iarm<o

0 0 0 0 0 0 0 0 -2VC Iarm>o

Table ι: Switching schedule for two serially connected full bridge cells

The upper part of Table l shows the switching pattern during normal operation, i.e. when the converter cells loa-b are not in a failure mode. In the normal mode, the main controller controls switching elements of the two full bridge cells loa-b to achieve a desired combined voltage Vout. It is to be noted that this topology supports a current Iarm in both directions, i.e. Iarm < o or Iarm > o.

Observing the pattern for normal operation, it can be seen that one switching element S3 of the first full bridge cell 10a and one switching element S8 of the second full bridge cell 10b are always on. Hence these two switching elements S3, S8 (marked with bold lines and hereinafter called normally on switching elements) are in a continuous conducting state, regardless of what output voltage the main controller desires, during normal operation.

Embodiments presented herein exploit this pattern by providing either or both of the normally on switching elements S3, S8 using components which have a lower resistance when in a conducting state, even if this is at some expense of controllability when the normally on switching elements S3, S8 need to be set in a blocking state. For example, the normally on switching elements S3, S8 can be provided using suitable controllable thyristors, such as IGCTs (Integrated Gate-Commutated Thyristors). The remaining switching elements are provided using elements which have low switching losses and are fully controllable, e.g. transistors such as IGBTs (Insulated Gate Bipolar Transistors) or alternatively, power Metal-Oxide-Semiconductor Field-Effect-Transistors (MOSFETs), power Bipolar Junction Transistors (BJTs), or Bi-mode Insulated Gate Transistors (BIGTs). This structure provides a significantly reduced conduction loss compared to using only transistors. Observing now the pattern in the lower half of Table 1, for a fault case e.g. a fault on the DC side, all switches are blocking. Depending on the direction of the current Iarm, the resulting voltage of the two converter cells loa-b is either 2Vc or -2Vc. It is to be noted that also the normally on switching elements S3, S8 are set to be blocking in the fault case. Hence, in the transition from normal operation to the fault case, the normally on switching element S3, S8 need to transition from a conducting state to a blocking state. In the case that the normally on switching elements S3, S8 are IGCTs or similar, a clamping inductor is typically provided in the prior art to achieve suitable controllability. However, in this application, the switch from conducting to blocking state is not as time critical as in a situation where the IGCTs are used for normal operation switching. Hence, the IGCTs can be controlled to go to a blocking state when the current passes zero (zero crossing), i.e. when the current through the IGCT is negligible. In this way, there is no need for clamping inductors, which significantly reduces cost and complexity.

Using IGCTs for the normally on switching elements has several other effects. The IGCTs are explosion proof for a surge fault on the DC link, which provides a more stable short circuit failure mode compared to if IGBTs or similar are used. Moreover, the gate unit for the IGCT requires only a very low power since the only switching occurs to (or potentially from) the failure mode. Also, very efficient and low cost cooling can be used since IGCTs are double-side cooled.

While the structure of Fig 4 has been described with switching elements S3 and S8 being the normally on switching elements, the control strategy can be modified (using the same structure) using a different switching pattern.

Consequently, any other one of the switches in each one of the full bridge cells loa-b can be configured to be a normally on switching element. In order to achieve the beneficial effects, one or more of the normally on switching elements is replaced with a controllable thyristor, which thus depends on the switching pattern used for the converter cell. Fig 5 is a schematic diagram illustrating a structure with a full bridge cell 10a and a half bridge cell lob connected in series. The full bridge cell 10a comprises four switching elements S1-S4 and the half bridge cell comprises two switching elements S5-S6. The full bridge cell 10a has the structure of the full bridge converter cell of Fig 3B and the half bridge cell 10b has the structure of the half bridge converter cell of Fig 3A.

Table 2 below shows the switching patterns for the converter cells loa-b of Fig 5, as controlled by the main controller for various desired output voltages Vout. A '0' denotes a blocking state and a '1' denotes a conducting state for the switching element in question.

Table 2: Switching schedule for a full bridge cell serially

connected with a half bridge cell

In this embodiment, there is one normally on switching element S3, which is implemented using a controllable thyristor such as an IGCT. The other switching elements S1-S2 and S4-S6 are implemented using transistors, such as IGBTs. An equivalent control scheme can be achieved using a different switching schedule, resulting in a different switching element being the normally on switching element.

The same effects are achieved here as explained above with reference to Fig 4. Fig 6 is a schematic diagram illustrating a structure similar to that of Fig 5, essentially with a full bridge cell 10a and a half bridge cell 10b connected in series. Here, however, the fourth switching element S4 has been replaced with a diode D4, and there is an additional connecting line between the converter cells loa-b with another diode D7.

Table 4 below shows the switching patterns for the converter cells loa-b of Fig 7, as controlled by the main controller for various desired output voltages Vout. A '0' denotes a blocking state and a '1' denotes a conducting state for the switching element in question.

Table 4: Switching schedule for a full bridge cell serially connected with a half bridge cell comprising a diode

In this embodiment, there is one normally on switching element S3, which is implemented using a controllable thyristor such as an IGCT. The other switching elements S1-S2 and S5-S6 are implemented using transistors, such as IGBTs. The same effects are achieved in this embodiment as explained above with reference to Fig 4. Fig 7 is a schematic diagram illustrating a structure similar but slightly different to that of Fig 5, comprising a full bridge cell 10a and a half bridge cell 10b connected in series. Here, however, the fourth switching element S4 has been replaced with a diode D4. Table 4 below shows the switching patterns for the converter cells loa-b of Fig 7, as controlled by the main controller for various desired output voltages Vout. A '0' denotes a blocking state and a '1' denotes a conducting state for the switching element in question.

Table 3: Switching schedule for a modified structure of a full bridge cell serially connected with a half bridge cell

In this embodiment, there is one normally on switching element S3, which is implemented using a controllable thyristor such as an IGCT. The other switching elements S1-S2 and S5-S6 are implemented using transistors, such as IGBTs. The same effects are achieved in this embodiment as explained above with reference to Fig 4.

Figs 8A-B are schematic diagrams illustrating how the switching elements of Figs 3A-B and Figs 4-7 can be implemented. In Fig 8A, an embodiment is shown where the switching element S is implemented using a transistor 25 such as an IGBT, controlled via a control connection C. This type of switching element is typically used for situations when the switching element is not a normally on switching element. An antiparallel diode can be provided in addition or as part of the transistor 25.

In Fig 8B, an embodiment is shown where the switching element S is implemented using a controllable thyristor such as an IGCT 26, controlled via a control connection C. This type of switching element is used for situations when the switching element is a normally on switching element.

Figs 9A-B are flow charts illustrating methods which can be executed in embodiments presented herein. The methods are used to control one or more converter cells of a high voltage multilevel converter as explained above. First, the method of Fig 9A will be described.

In a set thyristor in blocking state step 30, at least one thyristor being a normally on switching element, is set in a blocking state when a fault occurs on a DC side of the multilevel converter. This may optionally be delayed until a current through the at least one thyristor is negligible, e.g. at a zero crossing for the current through the thyristor. Negligible can e.g. be defined as having an absolute value less than a predetermined threshold. This predetermined threshold is selected to indicate a negligible current. In one example, the threshold itself is zero. In this way, clamping inductors are not required for controlling the thyristor to be in a blocking (or conducting) state.

Fig 9B is a flow chart illustrating an embodiment of another method. In this method, there is a detect fault step 28 prior to the set thyristor in blocking state step 30. In the detect fault step 28, a fault is detected in the DC side of the multilevel converter. The detection can be implemented using a detection circuit (hardware), using software instructions executed in a processor, or a combination of both hardware and software. This detection can then cause the set thyristor in blocking state step 30 to be executed. In other words, the set thyristor in blocking state 30 step may then be based on the detect fault step 28.

While the embodiments presented herein show a finite set of converter cell structures, the structure of one or more normally on switching elements being implemented using a controllable thyristor such as an IGCT can be applied to any suitable converter cell structure where a switching element is in a continuous conducting state during normal operation.

The invention has mainly been described above with reference to a few embodiments. However, as is readily appreciated by a person skilled in the art, other embodiments than the ones disclosed above are equally possible within the scope of the invention, as defined by the appended patent claims.