METHOD FOR DETERMINING A WIND TURBINE LAYOUT

Abstract

The invention provides a method for determining a wind turbine layout in a wind power plant comprising a plurality of wind turbines. The method comprises the steps of generating a plurality of random layout candidates fulfilling a set of basic requirements, and then performing a pre screening process on each of the plurality of random layout candidates. Based on the pre-screening process, a subset of layout candidates is selected and detailed optimization is performed on the layout candidates of the selected subset of layout candidates. Based on the detailed optimization, an optimized layout for the wind power plant is selected among the optimized layout candidates of the subset of layout candidates.

Claims

1. A method for determining a wind turbine layout in a wind power plant comprising a plurality of wind turbines, the method comprising: generating a plurality of random layout candidates, each layout candidate fulfilling a set of basic requirements, performing a pre-screening process on each of the plurality of random layout candidates, the pre-screening process including general optimization of the layout candidates, based on the pre-screening process, selecting a subset of layout candidates, performing detailed optimization on the layout candidates of the selected subset of layout candidates, and based on the detailed optimization, selecting an optimized layout for the wind power plant among the optimized layout candidates of the subset of layout candidates.

2. The method of claim 1, wherein the generating a plurality of random layout candidates is performed by defining a probability density function (PDF) proportional to a mean wind speed at the wind power plant.

3. The method of claim 1, wherein the generating a plurality of random layout candidates is performed by means of rejection sampling.

4. The method of claim 1, wherein the basic requirements include boundaries of site, internal exclusion zones, variable minimal spacing between wind turbines according to wind speed and wind direction, wind resources, number of wind turbines to be installed at the site, and/or total nominal power to be generated by the wind power plant.

5. The method of claim 1, wherein the general optimization includes applying a simple physical model involving at least one parameter.

6. The method of claim 1, wherein the pre-screening process includes a comparison of the layout candidates.

7. The method of claim 1, wherein the pre-screening process includes ranking the layout candidates, and wherein the selecting a subset of layout candidates comprises selecting the layout candidates with the best ranking.

8. The method of claim 1, wherein the detailed optimization includes applying a detailed physical model, applying statistical modelling, and/or identifying wind turbines that violate minimal spacing.

9. The method of claim 1, wherein the performing detailed optimization is performed iteratively.

10. The method of claim 1, wherein the performing detailed optimization on the layout candidates of the selected subset of layout candidates includes repositioning of at least one wind turbine until it reaches equilibrium.

11. The method of claim 1, further comprising the performing a final optimization of the selected optimized layout for the wind power plant.

12. The method of claim 1, further comprising the erecting a wind power plant in accordance with the selected optimized layout.

13. The method of claim 1, further comprising the estimating power generated by the wind power plant with the optimized layout.

14. (canceled)

15. A wind power plant, comprising: a plurality of wind turbines arranged at a site according to a layout; the layout determined by: generating a plurality of random layout candidates, each layout candidate fulfilling a set of basic requirements; candidates, the pre-screening process including general optimization of the layout candidates; based on the pre-screening process, selecting a subset of layout candidates, performing detailed optimization on the layout candidates of the selected subset of layout candidates; and based on the detailed optimization, selecting an optimized layout for the wind power plant among the optimized layout candidates of the subset of layout candidates.

16. A method for determining a wind turbine layout in a wind power plant comprising a plurality of wind turbines, the method comprising: generating a plurality of random layout candidates by defining a probability density function (PDF) proportional to a mean wind speed at the wind power plant; wherein each layout candidate fulfils a set of basic requirements including: boundaries of site, internal exclusion zones, variable minimal spacing between wind turbines according to wind speed and wind direction, wind resources, number of wind turbines to be installed at the site, and/or total nominal power to be generated by the wind power plant; performing a pre-screening process on each of the plurality of random layout candidates, the pre-screening process including general optimization of the layout candidates, wherein the general optimization includes applying a simple physical model involving at least one parameter; based on the pre-screening process, selecting a subset of layout candidates, performing detailed optimization on the layout candidates of the selected subset of layout candidates; and based on the detailed optimization, selecting an optimized layout for the wind power plant among the optimized layout candidates of the subset of layout candidates.

17. The method of claim 16, wherein the pre-screening process includes a comparison of the layout candidates.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0045] The invention will now be described in further detail with reference to the accompanying drawings in which

[0046] FIG. 1 shows a flow chart of a method for determining a wind turbine layout in a wind power plant comprising a plurality of wind turbines,

[0047] FIG. 2 illustrates two randomly generated layout candidates, and

[0048] FIG. 3 illustrates optimization process of one randomly generated layout.

DETAILED DESCRIPTION OF THE INVENTION

[0049] FIG. 1 shows a flow chart 100 of a method for determining a wind turbine layout of a wind power plant comprising a plurality of wind turbines. In the first step 101 a plurality of random layout candidates are generated. They may be generated through any standard statistical randomizer which includes certain boundaries as each layout candidate needs to fulfil a set of basic requirements. The set of basic requirements may include boundaries of site, internal exclusion zones, variable minimal spacing between turbines according to wind speed and wind direction, wind resources, number of wind turbines to be installed at the site, and/or total nominal power to be generated by the wind power plant. Random generation of the layouts is performed within a short time and it does not require large computing power.

[0050] Once the plurality of layout candidates is generated in the first step 101, a pre-screening process 102 is performed on each of the plurality of random layout candidates. The pre-screening process 102 includes general optimization of the layout candidates 103. In the pre-screening process 102 the layouts may be evaluated and further compared. General optimization 103 may, e.g., relate to decreasing the number of layout candidates as during pre-screening it may be revealed that some of the layout candidates are non-workable solutions which are then disregarded during the general optimization step 103.

[0051] Based on the pre-screening process 102, a subset of layout candidates is selected in the further step 104. If the layout candidates are compared in the pre-screening process 102, the best candidates may be selected for the subset. The selected layout candidates may be the most promising candidates for further optimization and which may not require large number of iterative optimization in order to achieve an optimal layout.

[0052] The step of detailed optimization 105 is further performed on each layout candidate of the selected subset of layout candidates. The detail optimization 105 may encounter for a plurality of physical conditions. In this step 105 at least one wind turbine may need to be repositioned within the layout. Optimization 105 of the layout candidates from the subset aims to generate a layout with the best possible performances. For instance, the detailed optimization 105 may improve the layouts in terms of maximum power generation, and/or for maximum profit for power in a given market, it may increase wind power plant's utilization rate, etc. This step 105 does not require large computing power and time as the detailed optimization 105 is performed only on a reduced number of layout candidates which already have decent characteristics, e.g., in terms of maximum power generation, etc.

[0053] Based on the detailed optimization 105, an optimized layout for the wind power plant is selected 106 among the optimized layout candidates of the subset of layout candidates. The selected layout candidate may have the best characteristic(s) among the all layout candidates from the subset, i.e., the selected layout may be the best in terms of minimum loads, maximum profit for power in a given market, the best in terms of cost, etc.

[0054] FIG. 2 illustrates three randomly generated layout candidates, 201-203. The layout candidates 201-203 may be generated by a randomizer. As an input, the randomizer may have wind resources map illustrated by a large number of contour lines 204 and a simplified wind rose 208, boundaries 205 defining a site, one exclusion zone 206, and a number of wind turbines available. The wind resources map 204 represents the energy carried by the wind and which may be based on at least one year of wind measurements in a region of the site. The wind rose 208 shows the frequency of winds blowing from particular directions over a specified period indicating the wind blowing in north east direction. In the initial step, a large number of layout candidates are generated and here only three of them are shown. Each layout candidate comprises 40 wind turbines 207 which are distributed in the best wind resources locations. Based on the wind rose 208, wind resources 204, boundaries 205 and the exclusion zone 206, the randomizer generates a plurality of layout candidates, three of which are shown in FIG. 2(a)-(c). Each of the layout candidates 201-203 have different spacing between the turbines 207, defined by the randomizer. In the pre-screening process, the plurality of layout candidates are compared and/or ranked based on their characteristics. The layout candidate 202 shown in FIG. 2(b) has the best expected production among the three shown layouts 201-203 and may be selected for further optimization. It should be noted that the layout candidates 201-203 are only an illustrative example. In a real case, the number of wind turbines may be larger, the wind rose may be more complex and the site may include more than one exclusion zone.

[0055] FIG. 3 illustrates optimization process of one randomly generated layout candidate 300. As in the previous FIG. 2, the layout candidate 300 is generated by the randomizer, based on the wind resources 204, boundaries of the site 205, one exclusion zone 206, 40 available wind turbines 207, and the simplified wind rose 208. The randomizer generates the layout candidate 300 shown in FIG. 3(a) in such a way that the predefined number of wind turbines 207 is distributed in areas with the most favourable wind, within the boundaries of the site 205 and outside the exclusion zone 206. Beside this particular layout candidate 300, a plurality of other random layout candidates are generates. Assuming that the layout candidate 300 is selected in the pre-screening, it will further be optimized. Further optimization may involve a number of iterative steps repositioning one by one wind turbine 207, or a group of wind turbines 207. Here, only two iterations are represented in FIGS. 3(b) and (c). From the FIG. 3(a) it can be seen that the wind turbines 207 are placed relatively close to each other and there is a possibility that the wind turbines 207 may influence each other's operation, i.e. wake effects may not be taken into account. FIG. 3(b) shows the layout candidate 300 after 100 iterations during which the majority of wind turbines 207 are repositioned and placed further apart to at least take wake effects into account. The final layout candidate 300 shown in FIG. 3(c) is a result of a further detailed optimization of the layout candidate 300 shown in FIG. 3(b). The number of the wind turbines 207 is maintained at 40 as the size of the site 205 is sufficient for 40 wind turbines 207. In another example, the number of the wind turbines may be changed. The optimization process of the layout candidate 300 is performed by optimizing total power generated by the wind power plant comprising 40 wind turbines 207. By comparing FIG. 3(a) and FIG. 3(c) it can be noticed that a general trend is that the wind turbines 207 are pushed away from each other and towards the site boundaries.