METHOD FOR EVALUATING EXPECTED PERFORMANCE OF A WIND FARM

Abstract

Evaluating expected performance of a wind farm during a voltage event of a power grid, using a simulation tool. The tool initiates a simulated voltage event, at time, t.sub.0; initiates simulation of a subsequent response to the voltage event by the farm; retrieves information regarding an initial voltage state of a model of the farm, at t.sub.0, and predicts a final voltage state of the model of the wind farm, based on the retrieved information regarding the initial voltage state. At t.sub.0, a simulated output current, I, of the farm is lowered, based on the retrieved information and the predicted final voltage state, and a simulated output voltage, V, of the model of the farm, is monitored. The simulated output current, I, of the farm is subsequently adjusted, based on V, while monitoring V, I, and/or simulated output power of the model. Expected performance of the farm is then evaluated.

Claims

1. A method of evaluating expected performance of a wind farm during a voltage event of a power grid, using a simulation tool, the wind farm comprising a plurality of wind turbines connected to the power grid, and the simulation tool comprising a model of the wind farm and a model of the power grid, the method comprising: causing a simulated voltage event in the model of the power grid, at a time, t.sub.0; by the simulation tool: initiating the simulated voltage event, at time, t.sub.0; initiating simulation of a subsequent response to the voltage event by the wind farm, the response by the wind farm including transfer from an initial voltage state towards a final voltage state of the model of the wind farm; retrieving information regarding the initial voltage state of the model of the wind farm, at time, t.sub.0; and predicting the final voltage state of the model of the wind farm, based on the retrieved information regarding the initial voltage state; at time, t.sub.0, lowering a simulated output current, I, of the wind farm, based on the retrieved information regarding the initial voltage state and the predicted final voltage state; monitoring a simulated output voltage, V, of the model of the wind farm, in response to the simulated voltage event; subsequently adjusting the simulated output current, I, of the wind farm, based on the monitored simulated output voltage, V, while monitoring simulated output voltage, V, simulated output current, I, and/or simulated output power of the wind farm model; and evaluating expected performance of the wind farm, based on the simulated output voltage, V, simulated output current, I, and/or simulated output power.

2. The method of claim 1, wherein predicting the final voltage state is performed prior to time, t.sub.0.

3. The method of claim 1, wherein monitoring a simulated output voltage, V, of the model of the wind farm, in response to the simulated voltage event, comprises monitoring a time derivative of the simulated output voltage, dV/dt, and/or a total change in simulated output voltage, ΔV, from the initial voltage state to the final voltage state.

4. The method of claim 1, wherein lowering a simulated output current, I, of the wind farm comprises lowering the simulated output current, I, to a predefined level in the case that the initial voltage state is a fault ride through state, and maintaining the simulated output current, I, in the case that the initial voltage state is not a fault ride through state.

5. The method of claim 1, wherein subsequently adjusting the simulated output current, I, of the wind farm comprises iteratively predicting the final voltage state, based on the monitored simulated output voltage, V, and adjusting the simulated output current, I, iteratively, based on the predicted final voltage state.

6. The method of claim 1, wherein the simulated voltage event is a low voltage ride through (LVRT) event.

7. The method of claim 1, wherein the simulated output current is a simulated reactive output current, I.sub.q, and wherein the simulated output power is a simulated reactive output power, Q.

8. The method of claim 1, wherein the model of the wind farm comprises a plurality of wind turbine models, each wind turbine model corresponding to one of the wind turbines of the wind farm, and wherein the step of lowering a simulated output current, I, of the wind farm comprises lowering a simulated output current, I.sub.WT, of at least some of the wind turbine models.

9. A computer readable medium storing a simulation tool comprising a model of a power grid and a model of a wind farm, the wind farm comprising a plurality of wind turbines connected to the power grid, wherein the simulation tool, when executed by one or more processors, is configured to perform an operation of evaluating expected performance of the wind farm during a voltage event of the power grid; the operation, comprising: causing a simulated voltage event in the model of the power grid, at a time, t.sub.0; by the simulation tool: initiating the simulated voltage event, at time, t.sub.0; initiating simulation of a subsequent response to the voltage event by the wind farm, the response by the wind farm including transfer from an initial voltage state towards a final voltage state of the model of the wind farm; retrieving information regarding the initial voltage state of the model of the wind farm, at time, t.sub.0; and predicting the final voltage state of the model of the wind farm, based on the retrieved information regarding the initial voltage state; at time, t.sub.0, lowering a simulated output current, I, of the wind farm, based on the retrieved information regarding the initial voltage state and the predicted final voltage state; monitoring a simulated output voltage, V, of the model of the wind farm, in response to the simulated voltage event; subsequently adjusting the simulated output current, I, of the wind farm, based on the monitored simulated output voltage, V, while monitoring simulated output voltage, V, simulated output current, I, and/or simulated output power of the wind farm model; and evaluating expected performance of the wind farm, based on the simulated output voltage, V, simulated output current, I, and/or simulated output power.

10. The computer readable medium of claim 9, wherein predicting the final voltage state is performed prior to time, t.sub.0.

11. The computer readable medium of claim 9, wherein monitoring a simulated output voltage, V, of the model of the wind farm, in response to the simulated voltage event, comprises monitoring a time derivative of the simulated output voltage, dV/dt, and/or a total change in simulated output voltage, ΔV, from the initial voltage state to the final voltage state.

12. The computer readable medium of claim 9, wherein lowering a simulated output current, I, of the wind farm comprises lowering the simulated output current, I, to a predefined level in the case that the initial voltage state is a fault ride through state, and maintaining the simulated output current, I, in the case that the initial voltage state is not a fault ride through state.

13. The computer readable medium of claim 9, wherein subsequently adjusting the simulated output current, I, of the wind farm comprises iteratively predicting the final voltage state, based on the monitored simulated output voltage, V, and adjusting the simulated output current, I, iteratively, based on the predicted final voltage state.

14. The computer readable medium of claim 9, wherein the simulated voltage event is a low voltage ride through (LVRT) event.

15. The computer readable medium of claim 9, wherein the simulated output current is a simulated reactive output current, I.sub.q, and wherein the simulated output power is a simulated reactive output power, Q.

16. The computer readable medium of claim 9, wherein the model of the wind farm comprises a plurality of wind turbine models, each wind turbine model corresponding to one of the wind turbines of the wind farm, and wherein the step of lowering a simulated output current, I, of the wind farm comprises lowering a simulated output current, I.sub.WT, of at least some of the wind turbine models.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The invention will now be described in further detail with reference to the accompanying drawings in which

[0052] FIG. 1 illustrates a wind farm connected to a power grid,

[0053] FIG. 2 is a flow chart illustrating a method according to an embodiment of the invention,

[0054] FIG. 3 illustrates simulated reactive power, simulated output voltage, and simulated output reactive current during a simulated voltage event, as a function of time, obtained by means of a prior art method, and

[0055] FIG. 4 illustrates simulated reactive power, simulated output voltage, and simulated output reactive current during a simulated voltage event, as a function of time, obtained by means of a method according to an embodiment of the invention.

DETAILED DESCRIPTION OF THE DRAWINGS

[0056] FIG. 1 illustrates a wind farm 1 comprising a plurality of wind turbines 2, three of which are shown. The wind turbines 2 are connected to a power grid 3 via a point of common coupling 4. Accordingly, the wind turbines 2 convert energy from the wind into electrical energy, which is supplied to the power grid 3, via the point of common coupling 4.

[0057] The wind farm 1 further comprises a power plant controller 5 which is communicatively connected to the wind turbines 2 via communication connection 6. Thus, the power plant controller 5 may receive operational data from the wind turbines 2, and the power plant controller 5 may dispatch control commands, e.g. in the form of setpoint values, to the wind turbines 2.

[0058] Before the wind farm 1 is connected to the power grid 3 for the first time, it may be required to demonstrate that the wind farm 1 will react appropriately to voltage events, such as low voltage ride through (LVRT) or high voltage ride through (HVRT) events of the power grid 3. To this end the performance of the wind farm 1 during such events may be evaluated by means of a method according to an embodiment of the invention, and using a simulation tool comprising a model of the wind farm 1 and a model of the power grid 3.

[0059] FIG. 2 is a flow chart illustrating a method according to an embodiment of the invention. The process is started at step 7. At step 8 a simulated voltage event in a model of a power grid forming part of a simulation tool is planned to take place at a future time, to. The voltage event could, e.g., be a low voltage ride through (LVRT) event.

[0060] At step 9, the time, t.sub.0, has been reached, and the simulated voltage event is initiated by the simulation tool.

[0061] At step 10, also at time, t.sub.0, the simulation tool retrieves information regarding an initial voltage state of the model of the wind farm. The retrieved information comprises information regarding whether or not the initial voltage state is a fault ride through (FRT) state. Furthermore, a final voltage state of the model of the wind farm is predicted, based on the retrieved information.

[0062] At step 11 it is investigated whether or not the initial voltage state of the model of the wind farm is a fault ride through (FRT) state. If this is the case, then it is likely that the planned simulated voltage event will cause the simulated output voltage, V, to increase abruptly, leading to a corresponding decrease in the simulated output current, I. Therefore, in this case, the process is forwarded to step 12, where the simulated output current, I, is lowered, already at time, t.sub.0, or immediately thereafter, in order to prevent that the output current, I, is too high, due to the duration of the simulation steps, thereby avoiding overshoots or spikes in the simulated output current, I, and the simulated output power.

[0063] At step 13, the simulated output voltage, V, is monitored, and the simulated output current, I, is adjusted in accordance therewith. Thus, it is continuously evaluated whether or not the voltage state of the model of the wind farm is transferring towards the predicted final voltage state. If not, the simulated output current, I, is adjusted accordingly, and thereby the actual behaviour of the real wind farm is followed accurately by the simulation.

[0064] In the case that step 11 reveals that the initial voltage state of the model of the wind farm is not a fault ride through (FRT) state, then the process is forwarded directly to step 13, i.e. the simulated output voltage, V, is monitored and the simulated output current, I, is adjusted as described above, but without the initial lowering of the simulated output current, I.

[0065] At step 14, the simulated output voltage, V, the simulated output current, I, and/or the simulated output power is/are monitored, in order to monitor the simulated behaviour of the wind farm in response to the voltage event.

[0066] Finally, at step 15, the performance of the wind farm, in response to the voltage event, is evaluated based on the monitored simulated output voltage, V, simulated output current, I, and/or simulated output power.

[0067] FIG. 3 illustrates simulated reactive power, simulated output voltage, and simulated output reactive current during a simulated voltage event, as a function of time, obtained by means of a prior art method.

[0068] A simulated voltage event in a power grid is initiated at approximately time=5.8 s, in which the simulated output voltage is decreased abruptly, and is maintained at a lower voltage level. Following the abrupt decrease in simulated output voltage, the simulated reactive output current increases abruptly, and is maintained at a higher current level, in order to support the power grid.

[0069] At approximately time=6.0 s, another simulated voltage event occurs, in which the simulated output voltage is increased abruptly to restore the initial voltage level. Following this, the simulated reactive output current decreases abruptly and restores the initial current level, since the increased reactive output current is no longer required in order to support the power grid.

[0070] However, since the simulation is performed in small time steps, rather than in a continuous manner, a time period corresponding to a simulation step may elapse from the simulated output voltage is abruptly decreased or increased, and until this is detected and a corresponding abrupt increase or decrease in the simulated reactive output current takes place. This results in spikes in the simulated reactive output current, and consequently in the simulated reactive power. In particular, it can be seen that a very large spike occurs in the simulated reactive power when the initial voltage and current levels are restored, approximately at time=6.0 s. These spikes are not ‘real’ in the sense that they would not occur when a real wind farm reacts to a similar voltage event in the power grid, where the control is continuous rather than stepwise. Accordingly, the spikes represent undesired discrepancies between the behaviour of the simulation model of the wind farm and the behaviour of the real wind farm.

[0071] FIG. 4 illustrates simulated reactive power, simulated output voltage, and simulated output reactive current during a simulated voltage event, as a function of time, obtained by means of a method according to an embodiment of the invention.

[0072] A voltage event is simulated, essentially in the manner described above with reference to FIG. 3. However, in the situation illustrated in FIG. 4, the abrupt changes in the simulated reactive output current are performed substantially simultaneously with the abrupt changes in the simulated output voltage. This can be done because it is known in advance that a voltage event will occur, and it is possible to predict, or provide a qualified guess on, a final voltage state of the wind farm, following the voltage event. As a consequence, the delay in the response by the simulated reactive output current is considerably reduced. It can be seen from FIG. 4, that this significantly reduces the spikes in the simulated reactive output current and the simulated reactive power. In particular, the large spike which occurred approximately at time=6.0 s in the simulated reactive power of FIG. 3 is essentially eliminated in FIG. 4. Accordingly, the simulation reflects the behaviour of the real wind farm in a much more accurate manner.