Patent Application
20110133461
The invention relates to wind-turbine technology and, for example, to methods of operating wind turbines under converter faults, and to related wind turbines.
WO 2004/100102 A1 describes a wind power installation with a generator and converter system divided into four independent branches. The generator and converter branches can be selectively connected by switches (
WO 2006/107548 A1 describes a power converter system adapted to operate in a normal mode and in a fault mode. The converter system has a plurality of parallel converters, referred to as “bridges”. In the event of a fault of a bridge the fault mode is entered and the faulty bridge is bypassed.
According to a first aspect, also referred to as the “reactive-current margin aspect”, a method is provided of operating a wind turbine comprising a generator and an electric-converter system arranged to produce and convert electric power up to a nominal active power. The wind turbine is arranged to operate in at least two different converter modes, referred to as a “fully-functional converter mode” and a “faulty-converter mode”. The electric-converter system comprises a plurality of parallel converters. The converters are dimensioned not only to operate at nominal active current, i.e. electric current corresponding to nominal active power production, but are dimensioned to provide an over-current margin to enable reactive current to be produced on top of the nominal active current in the fully-functional converter mode. The method comprises:
A wind turbine pertaining to the “reactive-current margin aspect” is also provided. It comprises a generator and an electric-converter system arranged to produce and convert electric power up to a nominal active power, and a controller arranged to control the electric-converter system. The controller is arranged, in the fully-functional converter mode, to cause the converters to produce reactive current on top of the nominal active current, for example according to a corresponding set-point. The wind turbine is arranged to operate in at least two different converter modes, the fully-functional converter mode and the faulty-converter mode. The controller is arranged, in response to a fault of one or more of the converters, to change operation from the fully-functional converter mode to the faulty-converter mode. The electric-converter system comprises a plurality of parallel converters. The converters are dimensioned not only to operate at nominal active current, but are dimensioned to provide an over-current margin to enable reactive current to be produced on top of the nominal active current in the fully-functional converter mode. The controller is arranged, in the fully-functional converter mode, to cause the converters to produce reactive current on top of the nominal active current upon request. It is further arranged, in the faulty-converter mode, to cause at least one other converter of the converter system to produce additional active current by using its over-current margin to compensate at least partly for a reduction of active-current production due to the fault of the one of the converters, and to cause the reactive-current production by the at least one other converter to be limited correspondingly.
According to a further aspect, also referred to as the “low-voltage margin aspect”, a method of operating a wind turbine is provided. The wind turbines comprise a generator and an electric-converter system arranged to produce and convert electric power up to a nominal active power to be supplied to an electric grid. The wind turbine is arranged to operate in at least two different converter modes, referred to as a “fully-functional converter mode” and a “faulty-converter mode”. The electric-converter system of the wind turbine comprises a plurality of parallel converters. The converters are dimensioned not only to operate at nominal active current, i.e. electric current corresponding to nominal active power production, but are dimensioned to provide an over-current margin to enable nominal active power to be produced in the event of a lower than nominal grid voltage in the fully-functional converter mode. The method comprises:
A wind turbine pertaining to the “low-voltage margin aspect” is also provided. It comprises a generator and an electric-converter system arranged to produce and convert electric power up to a nominal active power to be supplied to an electric grid, and a controller arranged to control the electric-converter system. The wind turbine is arranged to operate in at least two different converter modes, that is a fully-functional converter mode and a faulty-converter mode. The electric-converter system of the wind turbine comprises a plurality of parallel converters. The converters are dimensioned not only to operate at nominal active current, i.e. electric current corresponding to nominal active power production, but are dimensioned to provide an over-current margin to enable nominal active power to be produced in the event of a lower than nominal grid voltage in the fully-functional converter mode. The controller is arranged, in the fully-functional converter mode, to cause the converters to produce nominal active power and thereby increase the active current beyond nominal active current into the over-current margin in response to a lower-than-nominal grid voltage. The controller is arranged, in response to a fault of one or more of the converters of the wind turbine, to change operation from the fully-functional converter mode to the faulty-converter mode. The controller is arranged to cause, in the faulty-converter mode and at nominal grid voltage, at least one other converter of the converter system of the wind turbine to produce additional active current by using its over-current margin to compensate at least partly for a reduction of active-current production due to the fault of the one of the converters.
According to an another aspect, also referred to as the “low-temperature margin aspect”, a method is provided of operating a wind turbine comprising a generator, an electric-converter system arranged to produce and convert electric power up to a nominal active power. A converter-cooling system is also provided, with a coolant having a coolant temperature. The wind turbine is arranged to operate in at least two different converter modes, referred to as a “fully-functional converter mode” and a “faulty-converter mode”. The electric-converter system comprises a plurality of parallel converters. The converters are dimensioned to operate at nominal active current, i.e. electric current corresponding to nominal active power production, at a predetermined coolant temperature. Less electric power is converted by the converter system in the faulty-converter mode than in the fully-functional converter mode, so that less heat is produced by the converter system in the faulty-converter mode than in the fully-functional converter mode. The claimed method comprises:
A wind turbine pertaining to the low-temperature margin aspect is also provided. The wind turbine comprises a generator, an electric-converter system arranged to produce and convert electric power up to a nominal active power, and a converter-cooling system with a coolant having a coolant temperature. The wind turbine is arranged to operate in at least two different converter modes, the fully-functional converter mode and the faulty-converter mode. The electric-converter system comprises a plurality of parallel converters. The converters are dimensioned to operate at nominal active current, i.e. electric current corresponding to nominal active power production, at a predetermined coolant temperature. The amount of electric power convertible by the converter system is reduced in the faulty-converter mode, so that less heat is produced by the converter system in the faulty-converter mode than in the fully-functional converter mode. The controller is arranged, in response to a fault of one or more of the converters, to change operation from the fully-functional converter mode to the faulty-converter mode. It is also arranged to cause the coolant temperature to be lowered, or to detect a signal indicating a lowered coolant temperature. Moreover, the controller is arranged to cause at least one other converter of the converter system to produce additional active current by using an over-current ability due to the lower coolant temperature to compensate at least partly for a reduction of active-current production due to the fault of the one of the converters.
Other features are inherent in the methods and products disclosed or will become apparent to those skilled in the art from the following description of embodiments and its accompanying drawings.
Embodiments of the present invention are explained, by way of example, and with reference to the accompanying drawings, in which:
a-6c illustrate operation using the reactive-power margin (
a-7b illustrate operation using the reactive-power margin, at the wind-park level;
a-8c illustrate operation using the low-voltage margin, at the wind-park level.
Normally, when a converter system of a wind turbine is built fault-tolerant (which is sometimes also referred to as “redundant”) at least two parallel converters, or “power paths” are provided. In the case of a failure of one of the power paths, the power path concerned can be shut down, and operation can continue through the other path(s); this is referred to as “faulty-converter mode”, or simply “fault operation”. Conventionally, fault operation can only be performed at reduced power of (n−1)/n of the nominal power (where n is the number of parallel power paths). For example, with two parallel power paths the total power would conventionally be reduced to ½ of the nominal power during fault operation.
WO 2004/100102 A1 mentioned at the outset, proposes that the parallel converters be over-dimensioned, e.g. by 20%. While without such over-dimensioning the remaining total power would be reduced to 0.75 of the nominal power (n=4 in WO 2004/100102 A1), over-dimensioning the converters by 20% enables the reduction to be limited to 0.9 of the nominal power. However, over-dimensioning of the converter system only to provide improved fault-operation is relatively costly.
The inventor has recognised that over-current margins provided in converter systems for other functions than fault tolerance can be employed for increasing the active power produced during fault operation by disabling the other function in the case of a fault, i.e. by disabling for example the reactive-power margin and/or the low-voltage margin (these margins will be explained in detail below). This is the concept underlying the reactive-current margin and the low-voltage margin aspects mentioned in the summary section. The inventor has further recognised that an over-current margin can be created if the total active power is reduced below nominal in the case of a failure, and that this can be employed to increase the active power produced during fault operation. This is the concept underlying the low-temperature margin aspect. Essentially, the concept is that one gets more active power in the event of a converter failure at no added expense.
While an electric machine, for example, can usually be overloaded significantly for short time intervals due to its relatively large heat capacity and thermal inertia, semiconductor switches of a converter system can hardly be overloaded at all, not even for relatively short time intervals, due to their little heat capacity and thermal inertia.
Assuming that the reactive-power margin is 15% and the low-voltage margin is 10%, and further assuming a converter system with two power paths, a fault of one of these power paths without use of any of these margins would reduce the power produced to 50% of the nominal power, but with use of these margins 62.5% of the nominal power can be produced. Furthermore, with reduced production, the cooling need from the converter system is also reduced, meaning that the coolant will be cooler, or—in a cooling system with a feedback-controlled coolant temperature—the set-point of the coolant temperature can be lowered, which allows for a yet higher production ending in 65-67.5% of nominal power. This is a significant improvement in energy production in scenarios where fault tolerant operation is active. It also allows for longer operation in faulty condition so that service/repair is less urgent which is particularly useful in offshore wind parks in which accessibility of wind turbines is often limited.
Using the reactive-power margin and/or the low-voltage margin during fault operation produce at least a fraction of the active power which would normally be produced by the faulty converter means that these margins are not available for reactive-power production and/or active-power production at low grid voltage anymore. However, in some of the embodiments in which the wind turbine with the faulty converter is part of a wind park the remaining turbines of the wind park compensate for the reactive-power deficit and/or increase the voltage “seen” by the wind turbine with the faulty converter in the case of a low grid voltage by producing reactive power. This functionality at the wind-park level is, for example, controlled by a wind-park controller. The wind-park controller is informed by the wind turbines of the wind park about any fault which would require the wind-park controller to direct the other wind turbines to compensate for the reactive-power deficit and/or increase the voltage in the wind park.
In some embodiments the wind turbine comprises a generator and an electric-converter system arranged to produce and convert electric power up to a nominal active power. At relatively low wind speeds, in what is called the partial-load mode, the wind turbine of some embodiments is operated at maximum efficiency, i.e. it attempts to convert as much wind power into electric power as possible. When the wind speed increases, as from a certain wind-speed value (typically between about 10 and 18 m/s) the operation is changed to what is called the full-load mode. In the full-load mode the electric power produced is limited to a predetermined and substantially constant value, even if the wind would allow more power to be produced. This limitation of electric power produced can, e.g. in wind turbines with pitchable blades, be achieved by feathering the blades (i.e. pitching the blades out of the wind). The electric power produced in the full-load mode is what is called the “nominal active power”.
In some embodiments the electric-converter system comprises a plurality of parallel converters, or power paths, and is fault-tolerant or “redundant”. In some embodiments the converters are electrically connected with their inputs and outputs to a common input point and a common output point; thus being “parallel” in a strict sense. In other embodiments, however, the generator's stator winding is subdivided in electrically independent sub-windings, each sub-winding being, e.g. connected to one converter or group of converters. These converters (or groups of converters)—although sharing the current produced by the generator—are not electrically connected both at their inputs and outputs. “Parallel” is used herein in a broader sense of a configuration in which the converters share the current produced by the generator which encompasses, e.g. the subdivided-generator example.
The electricity grid to which electric power is supplied by the wind park and to which the wind turbines of the wind park are connected, e.g. at a point of common coupling (referred to as “PCC”), usually has a nominal voltage. Deviations of the grid voltage from the nominal grid voltage are typically in a range of ±10% around the nominal grid voltage. If the grid voltage is defined to be the voltage at a distant point in the grid (e.g. at the point where a branch line connecting the wind park to the grid meets the grid) the voltage at the PCC is not necessarily equal to the grid voltage owing to a possible voltage drop or voltage increase across the branch, caused by current produced by the wind park and flowing through the impedance of the branch line. However, the voltage at the PCC is linked to the grid voltage at the distant point, so that a nominal voltage can also assigned to the PCC (e.g. defined as the voltage at the PCC when the grid voltage at the distant point is nominal, and the wind park supplies nominal active power but no reactive power at the PCC. The actual voltage at the PCC will approximately follow the ±10% grid-voltage variations at the distant point.
The term “nominal active current” through a circuit element of the converter (such as a power path, or all the parallel power paths together) refers to the active current (opposed to reactive current) flowing through that element in the fully-functional converter mode at nominal voltage at the PCC.
As already mentioned above, in some of the embodiments the wind turbine is arranged to operate in at least two different converter modes, referred to as the “fully-functional converter mode” and the “faulty-converter mode”. In the fully-functional converter mode, in some embodiments, all the parallel converters are operative, at least at nominal active power (below nominal active power, however, some of the converters may be inactive, in some embodiments).
In the event of a failure of one of the power paths, the power path concerned is shut down, and operation continues through the other path(s) in the “faulty-converter mode”. Thus, in response to a fault of a converter, operation is changed from the fully-functional converter mode to the faulty-converter mode. This mode change is, for example, performed by a controller of the converter system, also referred to as the “converter controller”, which may be a part of a wind-turbine controller. In some embodiments the converter monitors the converter system, e.g. by measuring voltages and/or currents in different branches of the converter system, and thereby becomes aware of faults of single converters in the converter system. In other embodiments another entity, e.g. a converter monitoring unit, performs such a monitoring function and notifies the converter controller about the converter fault.
A number of activities are carried out, e.g. by the converter controller, when the change from the fully-functional converter mode to the faulty-converter mode is to be performed:
Further activities are carried out, e.g. by the converter controller and/or the wind-park controller, which are specific to the different over-current margins which may be used to increase the active current flowing through the remaining power paths beyond the nominal active current (reactive-power margin, low-voltage margin, and low-temperature margin). These margin-specific activities will now be described in detail.
The fraction of reactive power produced is commonly expressed by what is called the “power factor”. The power factor PF is defined as the ratio of active power P to apparent power S. The square of the apparent power S is the sum of the squares of the active power P and the reactive power Q, S2=P2+Q2. Therefore, the power factor is the absolute value of the cosine of the phase angle φ, PF=|cos φ|. Reactive power flowing along a transmission line having an impedance causes a voltage drop or a voltage increase across the impedance (depending on the sign of the impedance and the sign of φ, i.e. whether the reactive power is inductive or capacitive). As a consequence, reactive-power production can be used in voltage control. By producing a suitable amount of reactive power (with the suitable sign of φ) a wind park can increase or decrease the voltage at the PCC or a more distant point in the grid. Wind parks are therefore usually required to be able to provide reactive power on demand by the grid operator, with a value of the power factor and sign of the phase-angle which can be freely chosen by the grid operator within certain limits. The demand can be represented, for example, in relative terms by a required power-factor or phase angle, or in absolute terms by a required reactive power or reactive current demand signal. Without any limiting intent all these alternatives are commonly referred to herein as “power-factor demand signal”. In some embodiments, for example, the power-factor demand signal is provided by the grid-operator to the wind-park controller. In other embodiments the wind-park controller itself measures the voltage, e.g. at the PCC and produces a power-factor demand signal on its own, without any prescription by the grid provider. In both cases the wind-park controller transforms the global wind-park-related demand signal into wind-turbine-individual demand signals for each wind turbine. The wind-turbine-individual demand signals may be the same for all the wind turbines of the wind park, or may be different for the individual wind turbines.
In order to cope with such reactive-power requirements, in some embodiments the converter system is dimensioned not only to operate at nominal active current, but is specially dimensioned to produce reactive current to be produced on top of the nominal active current in the fully-functional converter mode. To this end, a reactive-power over-current margin is provided (where “over-current” refers to “beyond nominal active current”) to enable the converters to bear the additional reactive current. For example, in order be able to produce electricity with any demanded power factor in the range 0.85≦PF≦1 (with either sign of φ) the reactive-power over-current margin is dimensioned to be about 1 minus the lower power-factor limit, i.e. about 15%. The converter controller is arranged, in the fully-functional converter mode, to cause the converters dimensioned in this way to produce reactive current on top of the nominal active current. In particular, the required phase shift φ between current and voltage is obtained by changing the on/off timing of the converters' semiconductor switches.
When the mode change from the fully-functional converter mode to the faulty-converter mode is made, in some embodiments the controller causes the reactive-power over-current margin of the remaining converter(s) to be partially or completely used for active-power production in order to compensate partly for the active-power deficit caused by the converter failure. As a result, the reactive-power over-current margin is not (or not completely) available any more for reactive current production. Thus, reactive-current production is reduced or completely stopped by the controller, as required by the increased active-power production (reducing and stopping reactive-current production are collectively referred to as “limiting reactive-current production”). In some embodiments limiting the reactive-current production is performed autonomously by the wind-turbine's controller converter. That is to say, in some embodiments the converter controller, when using the reactive-power over-current for additional active-power production in the faulty-converter, does not comply with any power-factor demand signaled by the wind-park controller.
This non-compliance with the power-factor demand is resolved, in some embodiments, by other turbines of the wind park compensating for the deficit of reactive-current production by the wind turbine operating in the faulty-converter mode. To this end the wind-park controller—which is in some embodiments notified about the operation in the faulty-converter mode of the wind turbine—directs one or more of the remaining wind turbines of the wind park to increase production of reactive current to compensate for the reduced reactive-current production by the wind turbine which is operating in the faulty-converter mode. Thus, the collective coordinated operation of a plurality of wind turbines in the wind park enables the active-power production to be maximised in the event of a converter fault, provided the remaining wind turbines were not already being required to produce maximum reactive-current production before the fault.
As indicated above, the voltage in an electricity utility grid is not constant but is typically allowed to vary, for example, in a range of ±10% around the nominal grid voltage. Such variations are, for example, due to the fact that consumption of electric power takes places in a manner which is not controlled or pre-determined by the grid operator. Owing to the increasing penetration of wind-power conversion, and to the rather unpredictable characteristics of the wind, the production side also increasingly contributes to such voltage variations. Grid operators often try not to exhaust the allowed voltage-variation range over longer periods of time, i.e. bring back a low voltage of say, 90% of the nominal voltage, to nominal voltage after a few seconds or minutes.
Besides such voltage variations in the allowed range there are grid faults with more significant voltage drops, e.g. short-circuits which cause the grid voltage to go down to e.g. 15% of the nominal voltage in the vicinity of the short-circuit location. Such faults are normally cleared within a second. The present description mainly deals with the voltage variations in the allowed range rather than such grid faults.
Grid-voltage variations are also “seen” by the wind turbines of a wind park, e.g. at the PCC of the wind turbines, even if the park is connected to the grid by a branch line across which there may be some voltage drop or increase due to the current flowing from the wind park to the grid. Thus, there is a nominal voltage at the PCC (e.g. defined as the voltage at the PCC when the grid voltage at the distant point is nominal, and the wind park supplies nominal active power but no reactive power at the PCC). The actual voltage at the PCC will approximately follow the grid-voltage variations. For simplicity, we assume herein that the voltage at the PCC equals the voltage at the terminals of the wind turbine in question which will be a good approximation if the distance between the wind turbine and the PCC is relatively small; However, if there is a significant difference between the voltage at the PCC and that at the terminals the latter shall be used instead.
Active power is the product of voltage and current. To provide nominal active power Pnom at nominal voltage Vnom the converter system produces a nominal current Inom: Pnom=Vnom·Inom. However, if the voltage decreases by a certain factor lv (for “low voltage”); e.g. 0.9, the current produced by the converter system has to be increased correspondingly (by the inverse of the factor lv−1, e.g. by 1/0.9) to maintain nominal-power production: Pnom=lv·Vnom·lv−1·Inom.
Therefore, in order to maintain nominal power in such under-voltages situations, in some embodiments the converter system is dimensioned not only to operate at nominal active current, but is dimensioned to provide an over-current margin to enable nominal active power to be produced in the event of a lower-than-nominal grid voltage in the fully-functional converter mode, i.e. to increase the current beyond the nominal current to compensate for the reduced voltage, e.g. increase the current to lv−1·Inom. Thus, in some embodiments the width of the low-voltage over-current margin is lv−1. The wind-turbine controller is arranged, in the fully-functional converter mode, to cause the converters of the wind turbine to produce nominal active power and thereby, in periods in which the voltage at the PCC is below nominal voltage, to increase the active current beyond nominal active current into the low-voltage over-current margin. This increase may be obtained by increasing the on-times of the converters' semiconductor switches.
When the mode change from the fully-functional converter mode to the faulty-converter mode is made, in some embodiments the low-voltage over-current margin is used in order to compensate partly for the active-power deficit caused by the converter failure. The controller causes the low-voltage over-current margin of the remaining converter(s) to be partially or completely used for active-power even if the grid voltage is at nominal grid voltage (or above nominal voltage, which is included in the term “at nominal grid voltage”). Thus, in the fault converter mode and at nominal grid voltage, the controller causes the remaining converter(s) to produce active current beyond nominal, as allowed by the low-voltage over-current margin. The remaining converter(s) therefore produce(s) more active power than its(their) nominal active power.
If the grid-voltage is low, e.g. at the lower limit of the allowed voltage range the remaining converter continues to produce active current beyond nominal, but since the voltage is low the converter only produces nominal active current; i.e. the low-voltage over-current margin allows no compensation of the deficit caused by the converter fault if the voltage is at the lower limit of the allowed voltage range. This is acceptable as the duration of periods with low voltage is normally short.
In some embodiments in which the wind turbine is one of a plurality of wind turbines of a wind park, the other, remaining turbines of the wind park may even enable the wind-turbine which is operating in the faulty converter mode to sustain its production beyond nominal active power in the case of low voltage in the (distant) grid. As was explained above, the voltage at the PCC can be raised by reactive-current production, as reactive current (with a suitable sign of φ) flowing through the branch line from the PCC to the distant grid causes a voltage across the branch line's impedance. Thus, in some embodiments the wind-park controller—after having been notified that the wind turbine in question operates in the faulty-converter mode—directs, in response to a lower-than-nominal voltage at the PCC, one or more of the remaining wind turbines of the wind park to produce power with a power factor suitable to increase the voltage at the PCC to the nominal voltage. Although the grid voltage at the distant point in the grid is lower than nominal the wind turbine operating in the faulty converter mode will then “see” nominal power, and its remaining converters can use their low-voltage over-current margin to produce active power beyond their nominal active power.
As indicated above, the sensitivity of typical semiconductor switches of a converter system is partially related to their thermal sensitivity, in the sense that any over-current may cause temperature rise which may immediately damage or even destroy the semiconductor switch concerned owing to its small heat capacity and small thermal inertia,. However, the amount of current that will actually damage or destroy a semiconductor switch depends heavily on the coolant temperature to which the semiconductor switch is subjected. A lower coolant temperature enables a semiconductor switch to withstand higher current.
In some embodiments, in the faulty-converter mode the remaining converters do not use a reactive-current margin or a low-voltage margin to compensate for the deficit caused by the failure, which reduces the total active-power production significantly. For example, in an embodiment with two parallel converters a fault of one converter will reduce the total active-power convertible by the converter system to 50% of the total active power in the fully functional mode. But even in embodiments with compensation based on a use of the reactive-power over-current margin and/or the low-voltage over-current margin, the total active-power will be reduced, e.g. in the two-converter example to about 62.5% of the total active power in the fully functional mode.
Reduced total active power production also means that the power loss (which equals the heat produced) in the converter system will be reduced approximately proportionally, since the power loss is approximately proportional to the total current switched by the converter system. In some embodiments a cooling system is provided for transferring the heat produced by the converter system to the environment. The cooling power of the cooling system increases with increasing coolant temperature. “Nominal coolant temperature” can he defined to be the coolant temperature needed at the upper design limit of ambient temperature (e.g. 40° C.) to be able to transfer the heat produced by the generator system at nominal active-power production to the environment. As the cooling power of the cooling system increases with increasing coolant temperature, in some embodiments the reduction of the total active power produced in the faulty-converter mode enables the coolant temperature to be lowered below the nominal coolant temperature.
A lower coolant temperature, in turn, enables the current switched by the remaining converters to be increased, thereby providing an additional “low-temperature over-current margin”.
In some embodiments the coolant temperature is feedback-controlled. The lowering of the coolant temperature can there be achieved by lowering the temperature set-point for the controller which controls the cooling power. In other embodiments in which the coolant temperature is not feedback controlled the coolant temperature adjusts itself to an equilibrium temperature value between the heat flowing into the cooling system and the heat flowing out of it. The reduction of the total active power produced shifts this equilibrium to a lower temperature so that, in these embodiments, the coolant temperature falls by itself, caused by the power reduction.
Accordingly, in some embodiments the converter controller is arranged, in response to a fault of one or more of the converters, to change operation from the fully-functional converter mode to the faulty-converter mode. It is further arranged, in embodiments with feedback-controlled coolant temperature to cause the coolant temperature to be lowered upon the mode change, e.g. by having a coolant controller change the temperature set-point. In some embodiments (with or without feedback control of the coolant temperature) the converter controller is arranged to detect a signal indicative of the lower coolant temperature. The converter controller is also arranged, once it has detected that the coolant temperature is lower, to cause one or more of the remaining converters to produce additional active current by using the low-temperature over-current margin.
Some of the embodiments only use the reactive-current margin, or the low-voltage margin, or low-temperature margin, to increase the power produced during fault operation. Other embodiments use two or three of these margins to increase the power produced during fault operation. These may be, for example the reactive-current margin and the low-voltage margin; the reactive-current margin and the low-temperature margin; the low-voltage margin and the low-temperature margin; or the reactive-current margin, the low-voltage margin, and the low-temperature margin. In some embodiments an additional margin is provided to cope with imbalances of load-sharing between the parallel power paths. During fault operation, however, the number of parallel power paths is reduced, or is even zero (meaning that only power path is left, with no power path parallel to it). This enables the load-sharing margin to be completely or partially used to increase the power produced during fault operation. Moreover, over-dimensioning as in WO 2004/100102 A1 can be used on top of this.
It is noted that the different margins have no individuality. In a converter system dimensioned to provide a plurality of over-current margins (e.g. a reactive-power margin and a low-voltage margin), if the active current is raised beyond nominal active current in the event of a failure one therefore cannot directly assign the margin then used specifically to a particular over-current margin (e.g. specifically to the a reactive-power margin, or the low-voltage margin). However, if at the same time reactive-power production is (partially or completely) disabled in order to make the margin designed at for reactive-power production available for active over-current it is justified to say that the active current is increased “into the reactive-power margin”. Likewise, if at the same time the ability to increase the active current inversely to the voltage at the wind turbine's terminals is disabled and, optionally, other wind turbines of the wind park prevent under-voltage from occuring, in order to make the margin designed at for low-voltage production available for active over-current to compensate for active-power reduction to a converter failure, it is justified to say that the active current is increased “into the low-voltage margin”.
In some embodiments, the controllers (the converter controller, a cooling controller, a wind-power-conversion controller, the wind-turbine controller, and/or the wind-park controller) comprise digital micro-processors. In those embodiments, for example, the definition that a controller is “arranged” to perform a certain activity may mean that the controller is programmed such that when the program is executed on the controller's micro-processor the activity is carried out.
The mention of different controllers does not imply that they are based on separate hardware or software. For example, the converter, cooling, and wind-power-conversion controllers can be functional parts of the wind-turbine controller's hard- and software, without any structural separation.
Returning now to
Each wind turbine 2, 3 has a rotor 4 with pitchable rotor blades 5. The rotor 4 drives a generator 6. In order to enable variable rotor speed, the electric current produced by the generator 6 is converted by a converter system 7 to current adapted to a substantially fixed grid frequency (e.g. 50 Hz or 60 Hz; a typical tolerance of the grid frequency is ±1%). The converter system 7 enables current to be produced with an arbitrary phase, as desired, relative to the grid voltage, thereby enabling variable reactive power to be produced. Each wind turbine 2, 3 has a wind-turbine controller 8. One of its tasks is to control the converter system 7 and to command it to produce electricity with a certain frequency, phase and amplitude. Those functions of the wind-turbine controller 8 which relate to control of the converter system 7 are also referred to as “converter controller”.
Each wind turbine 2, 3 is connected with its terminals to a wind-park internal grid 9 which leads to a point of common coupling (PCC) 10. At the PCC 10 the electricity produced by the wind park 1 is fed into a branch line 11 which connects the wind park 1 at a distant connection point 11 to an electric utility grid 13. The impedance of the branch line 11 is illustrated by an impedance symbol at 11a. The voltage at the terminals of the wind turbines 2, 3 is assumed herein to be approximately equal to the voltage at the PCC 10, that is to say an approximation is made in which the impedance of the internal-grid lines is neglected.
A wind park controller 14 is arranged to communicate bi-directionally with the wind-turbine controllers 8 via a communication network 15. The wind-park controller 14 is also arranged to measure the voltages at the PCC 10 and the distant connection point 12 and the electricity produced (e.g. current and phase angle) at the PCC 10 by sensors 16a, b. The wind-park controller is also able to receive signals representing requirements by the grid operator at input line 17, such a demand indicating a power-factor value to be produced by the wind park 1.
The converter system 7 includes a plurality of parallel converters 18, also referred to as “power paths”. In the implementation example shown each converter 18 includes a generator-side sub-converter 18a and a grid-side sub-converter 18b which are arranged “back to back” by an intermediate DC link 19. The DC link 19 is equipped with an energy-storing element 20, e.g. a capacitor. The generator-side input to the converter 18 is denoted by “21”, and its generator-side output is denoted by “22”.
In the implementation examples described below the smallest element of the converter system is the converter 18. Therefore, a failure of one of the sub-converters 18a, 18b (
In an implementation example illustrated in
A cooling system 30 is provided to dissipate the heat produced by the converters 18A, 18B of the converter system 23. It includes, in an implementation example, a liquid coolant circulating in a closed cooling circuit 31 which includes a heat exchanger 32. The heat exchanger 32 enables heat to be transferred from the coolant to the environment, e.g. to the atmosphere or to sea water. The cooling power can be regulated, for example, by controlling the flow of air, or (sea) water, through the heat exchanger 32.
The wind-turbine controller 8 monitors the operation of the converters 18A, 18B, causes mode changes from the fully-functional converter mode to the faulty-converter mode and vice versa, and actuates semiconductor switches of the converters 18A, 18B, and therefore acts as a converter controller. In this implementation example it is also responsible for controlling the temperature of the coolant in the cooling system 29.
a illustrates operation in the fully-functional converter mode. All the switches 24a, 24b, 26a, 26b (if present) are closed. Current is equally shared between the controllers 18. Each converter 18 operates with nominal active current. In the implementation example shown with two parallel converters 18, each of the converters 18A, 18B carries 50% of the total active power produced, which is the nominal active power of the wind turbine 2.
On the other hand,
In order for the low-temperature margin to become available, in some implementation examples when the change from the fully-functional converter mode to the faulty-converter mode is made, the controller 8 lowers the coolant-temperature set-point and, optionally, verifies that the required lower coolant temperature is reached before it causes the current through the converter 18B to be increased into the low-temperature margin. In other implementation examples without set-point-controlled coolant temperature, the controller 8 only verifies that the required lower coolant temperature is reached before it causes the current through the converter 18B to be increased into the low-temperature margin.
In the faulty-converter mode, however, the faulty converter (18A in
A mode change from the fully-functional converter mode to the faulty-converter mode is performed when the controller 8 observes a converter fault and vice versa when the converter fault is cleared. During the mode change the switches 24a, 26a are opened, the coolant temperature is lowered, current production is increased into the margins and the blade pitch is adapted to the reduced power production, and vice versa, under the directive of the controller 8.
In all of the
Before the mode change both power converters 18A, 18B are operative. The total active power produced by the wind turbine 2 is 1 pu. Each converter 18A, 18B carries 50% of the total active power (in the implementation example with two parallel converters). After the mode change the faulty-converter 18A is stopped, and all the power produced now only comes from the remaining converter 18B (in the implementation example with only two parallel converters shown). However, the active power produced by the remaining converter 18B is now increased by more than ⅓, e.g. from 0.5 pu to about 0.675 pu, by using the power-factor (PF) margin, the low-voltage (lv) margin, and the low-temperature (lT) margin. For example, the widths of the PF margin, lv margin, lT margin may be about 15%, 10%, 10%, respectively, of the nominal active power of the single converter 18B. To express these percentages as power units of the total active power of the wind turbine 2, the PF margin amounts to about 0.075 pu, the lv margin to about 0.05 pu and the lT margin to about 0.05 pu, in this implementation example.
a illustrates the use of the power-factor margin. Three parameters are shown for both operation modes, in an exemplary manner: the power factor demanded by the wind-park controller 14 (dashed line), the power factor actually produced by the converter 18B (solid line), and the active current produced by the converter 18B (solid line, in units pu corresponding to a nominal power for that converter), before and after the mode change.
In the example shown, the power factor demanded by the wind-park controller 14, e.g. based on a corresponding demand from the grid operator is constant; i.e. it is unaffected by the converter fault. The demand is here, by way of example, at about PF=0.925. The active current produced by the converter 18B is nominal (1 pu) before the mode change, because it is assumed in this example that all the power converters 18 (including 18B) operate at their nominal current before the mode change. The power factor actually produced before the mode change complies with the demand, i.e. is at about 0.925. The reactive-current component corresponding to this power factor is produced by using the reactive-power margin which allows reactive current to be produced on top of the nominal active current. This is actually the intended normal use of the reactive-power margin.
After the mode change, the reactive-power margin is used to increase the active-current production by the converter 18B beyond its nominal active current instead of producing reactive current. To this end, the controller 8 does not comply with the power-factor demand from the wind-park controller 14 anymore and correspondingly causes the power factor of the current produced to be 1. This can be achieved, for example, by inhibiting the demand signal to the controller 8. The reactive-power margin is now completely, or at least partially, used for active-current production beyond nominal. For example, as shown in
b illustrates the use of the low-voltage margin. Two parameters are shown for both operation modes, in an exemplary manner: the voltage at the terminals of the wind turbine 2, and the active current produced by the wind turbine 2.
In the example shown, the voltage at the wind turbine's terminals is normally at nominal voltage (=1 pu), but there are two voltage drops—one before and another after the mode change—during which the voltage is lowered to 90% of the nominal voltage (=0.9 pu).
Before the mode change the active current produced by the converter 18B is normally nominal (=1 pu). However, during the low-voltage interval, it is increased into the low-voltage margin so that the voltage drop is compensated for and the active power produced (=the product of voltage and current) is nominal active power. In the example shown, the active current produced by the power converter 18B is about 1.1 pu. This is the intended normal use of the low-voltage margin.
After the mode change the active current produced by the converter 18B is constantly kept at an increased value by making use of the low-voltage margin (e.g. at 1.1 pu in
c illustrates the use of the low-temperature margin. Two parameters are shown for both operation modes, in an exemplary manner: the temperature of the coolant, and the active current produced by the converter 18B.
Before the mode change the temperature of the coolant is at the nominal temperature Tnom. Therefore, the low-temperature margin is not available at this stage. Consequently, the active current produced by the converter 18B is nominal (=1 pu).
After the mode change, the temperature of the coolant is lowered, which is possible due to the reduction of the total current switched by the converter system 23 in the faulty-converter mode. The lower coolant temperature establishes the low-temperature margin which is then used to increase the active power production by the converter 18B beyond nominal, For example, it is to increase by 10% of the nominal active current, i.e. to 1.1 pu.
a and 7b relate to operation with use of the reactive-power margin at the wind-park level.
On the other hand,
a corresponds to the voltage diagram of
b illustrates the power factor produced by another wind turbine 3 of the wind park 1 (solid line) and the power factor demanded from it by the wind-park controller 14 (dashed-line). As in
c illustrates the voltage at the terminals of the faulty wind turbine 2 and the active current produced by it, before and after the mode change. The faulty wind turbine's behavior corresponds to what is shown in
However, after the mode change the wind-turbine controller 14 instructs the remaining wind turbine(s) 3 to counteract the voltage drop at the distant connection point 12 by producing reactive power. This is represented by the power-factor demand signal of
The reactive power injected by the other turbine(s) during the second voltage drop into the branch line 11 causes increase of the voltage across the branch line 11 at the distant connection point 12. Thereby, the voltage in the wind-park-internal grid 9 is kept approximately constant. As a result of this intervention of the wind-park controller 14 and the remaining wind turbine(s) 3 to support voltage, the faulty wind turbine 2 does not “see” the second voltage drop at the distant connection point 12, and is therefore able to maintain nominal active-power production during the voltage drop after the change, in contrast to the situation illustrated in
Although certain methods and products constructed in accordance with the teachings of the invention have been described herein, the scope of coverage of this patent is not limited thereto. On the contrary, this patent covers all embodiments of the teachings of the invention fairly falling within the scope of the appended claims either literally or under the doctrine of equivalents.
