Comparisons of Rotor Equivalent Wind Speed (REWS) in complex terrain using a 100m mast and several remote sensing devices

Comparisons of Rotor Equivalent Wind Speed (REWS) in complex terrain using a 100m mast and several remote sensing devices Stefanatos N., Foussekis D.,...
Author: Ernest Watts
4 downloads 0 Views 391KB Size
Comparisons of Rotor Equivalent Wind Speed (REWS) in complex terrain using a 100m mast and several remote sensing devices Stefanatos N., Foussekis D., Mouzakis F. Centre for Renewable Energy Sources (CRES), Wind Energy Department, Laboratory for wind turbine testing 19th km Marathonos Av.,GR-190 09, Pikermi, Attiki, GREECE Tel. +302106603300, Fax: +302106603301, e-mail: [email protected]

Summary The present work, aims to investigate the sensitivity of the proposed methodology, the measurement methods, and the sensing devices on the definition of the REWS. This is made by applying different scenarios for the estimation of REWS using data from campaigns including concurrent measurements from remote sensing devices (RSD) and mast based instruments in a complex terrain site. On the complex terrain site of CRES at Lavrion, a 100m mast is operated. Measurement campaigns have been made using four Lidars systems of different types, with concurrent recording of wind speed, wind direction, inclination, and meteorological parameters from mast based sensors on the 100m mast. Data gathered are used to estimate the REWS wind speed following the definition given in the running drafts of the IEC 6140012-1 standard. Various concepts were run, assuming wind turbines with different hub heights and different configurations of measurement layout. A good correlation between measurements from the different RSDs and mast based instruments is usually seen with differences in the resulting REWS being small, but dependent on the flow conditions and measurement configuration. For the cases studied, ddifferences seen between the REWS and hub height as measured by the RSDs are close to 1% for bin-averaged values with Urews being typically lower than Uhub.

1. Introduction In the next generation of standards for power performance measurements, which are currently in their final drafting stages, the use of a Rotor Equivalent Wind Speed (REWS) is proposed as an alternative to the hub height wind speed. On the same time, the use of Remote Sensing Devices is also introduced for measuring the wind speed as an alternative to mast based instruments. Information on the sensitivity of the REWS estimation n the measurement device, the measurement layout and WT main characteristics can be very useful for the correct implementation of the new definition of the power curve reference wind speed.

2. Method 2.1

Site description

CRES Test Station is situated approximately 100km SE of Athens. The facility is divided in two parts: The first and bigger part comprises a commercial Wind Farm with 5 HAWTs. The second part, situated in the north of the Wind Farm, comprises a 100m meteorological mast and two small wind turbines. The elevation of the site is approximately 112m and the sea coast is at 1km east. The landscape could be characterized as complex terrain (rather mild), surrounded by hills with rather gentle slopes, and low vegetation (see Figure 1). A roughness value of 0.15 represents rather well the entire site.

2.2

Mast description

The reference mast of the CRES Test Station is a lattice, guy wired mast with an equilateral

triangular cross-section, the same bottom to the top. Mast height is 100m.

from

Vector A100LK cup anemometers mounted on side booms to the east side of the mast were available at 54m, 76m and 100m AGL. The sensors are supported on the mast, by the aid of telescopic booms providing a 3m separation from mast. The mast influence in the sensors is estimated to be 1.5% following the recommendations of [1]. In the opposite side (Mast’s W-side) and at the same heights as the anemometers, Vector W200P wind vanes were installed using identical booms. All the anemometers were calibrated according to MEASNET procedures Error! Reference source not found., before their installation and right after their replacement.

Figure 1: The 100m reference mast at CRES test station (S view).

2.3

Lidar units

Four different Lidar units were used. • Lidar1: Measurement levels:40m, 54m, 76m, 100m, 120m, 140m, 160m. • Lidar2: Measurement levels:40m, 54m, 76m, 100m, 120m, 140m, 160m. • Lidar3: Meas. levels: 54m, 76m, 100m, 120m. • Lidar 4: Meas. levels: 54m, 76m, 100m.

2.4

Virtual rotor cases

The scope of the work is to investigate the sensitivity of the REWS to the operational characteristics of the measuring device, and the testing layout. Only wind speed measurements from the 100m mast and the Lidars are used. No power performance data are available. Two virtual rotor cases where defined, and the corresponding REWS were evaluated using the available wind speed data. For Case A, where a virtual WT with Hub height H=76m and diameter D=80m is assumed, the rotor spans the height from 36m to 116m AGL, allowing for the REWS to be measured using the mast anemometers and all Lidars. For Case B, a virtual WT with hub height H=100m and diameter D=130m, is assumed, where the rotor spans the height from 35m to 165m AGL. In this case, the height of the mast is not sufficient for REWS estimation directly. For comparison only, the mast readings are used to estimate REWS using shear extrapolation. Lidars1 2 &3 were operated at sufficient ranges for estimating the REWS for Case B. For Lidar4 the maximum measuring height was set at 100m, therefore no results are given for Case B for this Lidar.

REWS estimation

The definition described in [1] for the REWS calculation is used. Wind veer is not accounted for. The measurement campaigns examined here were not intended for REWS estimation, therefore the mast layout and the Lidar settings, present some deviations in comparison to the requirements of [1]. Main deviations are -

-

-

Only 3 measurement points are available for the mast and Lidar 4, which is the minimum allowed number of points, while it is recommended to use more than three Positioning of measurement points relative to the virtual rotors is deviating for the mast, with outer anemometers being a little inwards to the 2/3R range. For the higher rotor case, REWS is estimated using the mast as a “Hub height mast” and shear extrapolation, although this is not allowed by the [1]. REWS estimated by the mast using shear extrapolation is presented for comparison only.

To allow direct comparison of results among the different campaigns and devices REWS (Urews) is always presented in non dimensional form divided by the hub height wind speed (Uhub).

3. Results 3.1

One Lidar (Lidar3) is operating next to the mast. The REWS is estimated for two devices (Mast, Lidar3) simultaneously. For this campaign data are available for a south sector (180o to 210o) also and are presented in parallel Campaign C. One Lidar (Lidar4) is operating next to the mast. The REWS is estimated for two devices simultaneously (Mast, Lidar4). Lidar 4 was set to measure up to 100m, therefore the high rotor case (H=100m, D=130m) is not examined. North sector only (350 o to 10o) is run for this case also. The following filtering was applied to all cases run. • 4.0 m/s < Uhub < 25.0 m/s • -0.5 < α < 1.5 where α the shear factor • -15 o < Wind veer< 15o The shear and veer filtering aim to isolate extreme events and have minimal impact on the data base (less than 1%of data rejected for most cases)

3.2

0.25

Campaigns

Three measurement campaigns are examined Campaign A. Two Lidars (Lidar1 & Lidar2) were operating simultaneously next to the mast. The REWS is estimated for three wind speed measuring devises (Mast, Lidar1, Lidar2). To reduce mast interference,only data from a narrow sector to the north (350o to 10o), where the wind speed approaches perpendicular to the anemometer booms, are used. Campaign B.

Overview of flow characteristics

Figures 2, to 5 below present Turbulence intensity at 100m as measured by cup anemometer and, flow inclination at 100m, wind veer across rotor and shear in the lower and upper part of the virtual rotor for the high case, as measured by Lidars. All data presented are bin-averaged per wind speed. The flow characteristics are typical of mild complex terrain sites as the one here.

Turbulence Intensity [ ]

2.5

0.2 0.15 0.1 0.05 0 0

4

8

12 16 Uhub [m/s]

20

24

28

Figure 2. Turbulence intensity at 100m

6

12

4

8

Wind veer [deg ]

Flow inclination [deg ]

8

2 0 -2 -4 0

4

8

12 16 Uhub [m/s]

20

24

4 0

28

-4 0

0.4

3.3 0.2 0 -0.2 -0.4 0

4

8

12 16 U hub [m/s]

20

24

(a) Shear exponent α (upper rotor) [ ]

4

8

12 16 Uhub [m/s]

20

24

Figure 5. Wind veer across rotor (Case B, H=100m, D=130m)

0.4

Presentation of results.

Campaing A Figure 6 presents the ratio of Urews/Uhub versus Uhub and versus shear difference between lower and upper part of rotor as measured by Lidar1 for the high rotor case (H-100m, D-130m) of Campaign A. 10minmean and bin-averaged values are presented. On bin averaged values, the difference between Urews and Uhub is smaller than 1%, while variations up to 2% are typically seen in 10min mean values. Larger scatter in wind speeds below 8m/sec is also observed. As expected, the ratio Urews/Uhub is directly related to the shear difference between upper and lower part of rotor.

0.2 1.04

0 -0.2 -0.4 0

4

8

12 16 U hub [m/s]

20

(b) Figure 4. Wind shear for lower (a) and upper (b) rotor part (Case B, H=100m, D=130m)

24

U rews_Lidar / UHub_Lidar

Shear exponent α (lower rotor) [ ]

Figure 3. Flow inclination at 100m

1.03 1.02 1.01 1 0.99 0.98 0.97 0.96 0

4

8

12

16

UHub_Lidar [m/s]

20

24

1.04 1.03

1.04

1.02

U rews / UHub

Urews_Lidar / Uhub_Lidar

1.06

1.02 1 0.98

1.01 1 0.99 0.98 0.97 0.96

0.96 -0.4

-0.2

0

0.2

0.95

0.4

Shear exponent diff. αup- αlow [-]

0

Figure 6. Campaign A, H=100m, D=130m, Lidar1 Ratio Urews/Uhub versus Uhub and shear difference between lower and upper rotor Figure 7 presents Urews/Uhub versus Uhub for all devices used simultaneously in campaign A (Mast, Lidar1 & Lidar2) for both high and low virtual rotor cases. Very good agreement in the estimation of Urews is seen between the two Lidars with differences being smaller than 0.5%. For all devices (Lidar1, 2 & mast) the bin–averaged values for Urews/Uhub are within 1% from unity, with the exception of low wind speeds range for the mast, for the low rotor case. 1.04 1.03

4

8

12

16

20

24

Uhub [m/s]

(b) Figure 7. Campaign A, Urews/Uhub versus Uhub , all devices (a) H=76m, D=80m (b) H=100m, D=130m Campaign B Figure 8 presents results for Campaign B, where one Lidar operates (Lidar3), but two sectors (north & south) are investigated. Again differences between Urews and Uhub are close to 1%, for the complete wind speed range.

1.01 1

1.04

0.99

1.02

1.03

U rews / UHub

U rews / UHub

1.02

0.98 0.97 0.96 0.95 0

4

8

12

16

Uhub [m/s]

20

24

1.01 1 0.99 0.98 0.97 0.96 0

4

8

12

16

Uhub [m/s] (a)

(a)

20

24

1.04 1.03

U rews / UHub

1.02 1.01 1 0.99 0.98 0.97 0.96 0

4

8

12

16

20

24

Uhub [m/s]

(b) Figure 8. Campaign A, Urews/Uhub versus Uhub , all devices (a) H=76m, D=80m (b)H=100m, D=130m

28

1.04

1.04

1.03 1.02

U rews / U Hub

Campaign C Results from campaign C are given in Figure 9 (Lidar4, low case only)

1.03

1 0.99 0.98 0.97

1.02

U rews / U Hub

1.01

0.96

1.01

0

1

4

8

12

16

20

24

20

24

Uhub [m/s]

0.99 0.98 0.97 0.96 0

4

8

12

16

20

(a)

Uhub [m/s]

1.04 1.03

Figure 9. Campaign C, Urews/Uhub versus Uhub , Lidar4 & mast H=76m, D=80m

U rews / UHub

1.02 1.01 1 0.99 0.98 0.97

An overview of Urews/Uhub ratio for all cases studied for the low and high virtual rotor is given in Figure 10.

0.96 0

4

8

12

16

Uhub [m/s]

(b) Figure 10. Urews/Uhub versus Uhub , all Lidars (a) H=76m, D=80m (b) H=100m, D=130m

4. Discussion Data from simultaneous measurements with remote sensing devices and a 100m are used to evaluate Rotor equivalent wind speed for virtual wind turbines rotors. For the Low rotor cases, where Urews can be directly estimated from the mast data, differences in the non dimensionalised ratio Urews/Uhub between the mast based and the Lidar measurements are around 1% for most cases. For Lidars operating in parallel (Campaign A, Lidar1 & Lidar2) the differences on the Urews/Uhub are around 0.5% on the bin averaged values, however this applies to the non-dimensionalised ratio, and does not account for absolute differences between the different Lidars that may be due to individual calibration or operating characteristics. For all Lidars, differences seen between the REWS and hub height as binned averaged values are close to 1%, while 10-min values show a scatter of 2 to 3% typically. Shear difference between the lower and upper part of the rotor directly influences the variation in Urews. For the cases studied, Urews tends to be smaller than Uhub, following the

trend where shear closer to the ground (lower rotor part) is usually more intense than higher up (upper part of rotor). It must be noted that wind speed binaveraged values presented here, come from a linear averaging of the corresponding values. However, the response of a wind turbine rotor to wind speed is not linear, and varies across the WT operating range. Therefore, the effect of the differences seen between Urews and Uhub on the power performance and the corresponding energy production may not follow linearly the ratio variations. Further Urews/Uhub investigations, including power performance measurements should also be evaluated, to better describe the sensitivity to the parameters mentioned here.

5. References [1]. IEC 61400-12-1 ( Draft CDV / 22.7.15) “Wind Turbines-Part. 12-1: Power performance. measurements of electricity producing w/t”