325

1 INTRODUCTION

Currently specialists are examining possibilities of

using objects on the shore or at sea to determine ship

position during emergency situations or in case of

interference in the functioning of satellite systems. To

that end using DGPS stations and AIS base stations is

being considered. In the authors’ opinion, wind farms

constructed at sea could also serve this purpose,

especially since more and more of such objects are

appearing on coastal areas and are detected by ship

radars at long distances. This research paper includes

the calculations of the probability of detecting a wind

farm via ship radar within the function of distance

from the antenna. Due to the limited size of the

publication only extreme cases are presented on the

graphs: favourable conditions – sea state 3 (wind

speed 8m/s and wave height below 1m) and adverse

conditions – sea state 7 (wind speed 15m/s and wave

height of up to 5 m). The sea states were examined

without precipitation and during low rainfall (with

the intensity of 4mm/h) and intense rainfall (20mm/h).

2 METHOD OF CALCULATING THE RANGE OF

RADAR DETECTION

To conduct the calculations the authors used the

computer programme CARPET (Computer- Aided

Radar Performance Evaluation Tool) [1], which was

developed by the TNO Physics and Electronics

Laboratory (TNO-FEL) in Holland. The programme is

an acknowledged tool for calculating the range of

various radars and used among other things in

designing and trials of marine traffic monitoring

systems. The equation of the range was programmed

as follows and allows analysing the impact of many

factors on the detection capabilities of radiolocation

devices, not only marine ones, thus the assumptions

and initial parameters have to be defined. The

equation looks as follows: [1]

( )

peak t r pc t

t t bs r atm

P GGG F

R LL LL

λσ

(1)

where:

peak

– peak power (W),

– antenna power gain in the transmitter and

receiver,

Radar Observation of Wind Farms in Various Weather

onditions

T. Stupak

Gdynia Maritime University, Gdynia, Poland

S. Świerczyński & M. Wąż

Polish Naval Academy, Gdynia, Poland

ABSTRACT: This article presents a calculation of condition detection wind farm by ship’s radar. The authors

used computer programme CARPET 2 for simulation different propagation condition. Wind farm echoes are

visible in significant distance and can be advantage to ship position mark.

http://www.transnav.eu

the

International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 14

Number 2

June 2020

DOI:

10.12716/1001.14.02.07

326

– transmit wavelength (m),

– radar cross-section (m2),

– wave interference ratio,

– range,

– transmitter/antenna path loss,

– radiation pattern main lobe distance,

– receiver path loss,

atm

– atmospheric attenuation (oxygen, rainfall,

water vapor).

3 RESEARCH CONDITIONS

The research method for this paper was a computer

simulation using the programme CARPET v2.0. The

purpose of the research was to assess the possibility of

detecting a wind farm in various hydro-

meteorological conditions at sea. To make the

calculations, the authors used the parameters of

Raytheon NSC34 radars (Table 1), which can be

deemed typical for the devices currently utilized in

the navy.

Table 1. Parameters of Raytheon NSC34 radar of X band and

S band[6]

_______________________________________________

Band X S

_______________________________________________

Horizontal width of characteristics [°] 1.0 1.5

Vertical width of characteristics [°] 23 23

Gain [db] 29 28

Polarization Horizontal Horizontal

Frequency [MHz] 9410 3050

Pulse length [µs] 0,8 0,8

Pulse frequency [Hz] 1000 1000

Bandwidth [MHz] 4 4

Peak power [kW] 25 30

_______________________________________________

During the research the authors used a wind farm

whose radar cross-section of a single wind farm

equals 300m², whereas the height of one turbine is

60m. Offshore structures with such a surface area and

height are well detected by ship radars. The big

height value of the wind generators’ turbines and

their metal structure render the radar cross-section

huge and thus, multiple indirect echoes and echoes

from side lobes can appear, but also shadow sectors

are generated. [5]

In the case of a radar operating in X band there is

100% likelihood of detecting a wind farm at a distance

of 5 Nm, however, the distance grows to 10 Nm for a

radar operating in S band. Within this range if we are

approaching a wind farm, we will obtain a better

image in S band than in X band. Nevertheless, at a

distance of less than 1 Nm from the antenna, the S

band radar will start registering a high volume of

noise resulting from signals received in the directions

of the side lobes. The image on the radar operating in

S band will be corrupt, with multiple jamming

signals. An antenna of a radar operating on longer

waves has worse characteristics out of necessity, as

the level of side lobes decreases with the increase of

the antenna’s dimensions calculated in relation to the

wavelength. At small distances from the wind farm

the radar operating in X band is more useful than the

S band radar due to its higher antenna parameters.

The assumed height of the antenna of the utilized

radar was 20m above sea level, which is compatible

with the height of the radar antenna installation on a

small ship. On larger ships the maximum observation

ranges will be bigger, but upon high sea states the

interference from the sea surface will be more intense.

During the research the authors used tables from

the CARPET 2.0 User Manual, in which the scale of

sea states and the Beaufort scale are presented. The

simulations were conducted for the following hydro-

meteorological conditions presented in table 2 and the

results are shown in the consecutive figures. For the

simulation the conditions presented in table 2 were

chosen. The values for each sea state are consistent

with the data of the programme. 80% of the time at

sea the sea state does not exceed 3, in storm

conditions the possibilities of radar detection are

limited, and that is why these sea states were chosen

for analysis. Rainfall decrease the echo signal, thus the

rainfall in the conducted analyses was taken into

consideration.

Table 2. Hydro-meteorological conditions for all the cases

under research

_______________________________________________

No. Wind Sea state

_______________________________________________

1 15w 3

2 29,14w 7

3 58,4w 8

_______________________________________________

The radar cross-section is defined as the surface of

the metal plate which gives the same level of radio-

location signal as in the case of a real object and is

introduced into the range equation as a point gain

parameter, thus it was assumed that it is a square and

its height is located at the point where diagonals

intersect. Because the radar cross-section of the

windfarm is large, it was assumed that it is the same

for both radar bands and equals 300 m2 and is located

at the height of 60 m above sea level.[2]

4 SIMULATION OF WIND FARM DETECTION

Figure 1. The course of object signal and noise for S band

In fig. 1 one can observe the course of radar signals

within the function of the distance from the antenna

for S band upon sea state 8. The course of the signal

reflected from the windfarms’ windmill, interference

reflected from the sea, rain and noise are shown. A

327

signal reflected from an object is bigger than a signal

reflected from waves of approximately 20dB and the

interference disappears at the distance of

approximately 4 NM. A signal reflected from rainfall

of low intensity (4mm/h) is over 40 dB smaller than

that of the echo, so it does not influence the possibility

of echo observation. The echo signal fluctuates at the

distance between 3 and 8 NM, disappears at the

distance of 9 NM, and appears above 10 NM, and

subsequently decreases quickly, because beyond the

radio horizon the signal is reflected by a smaller and

smaller fragment of the windmill, and for the radar

under research the distance to the radar horizon

equals approximately 10 NM.

Figure 2. The distribution of probability of detecting a band

S radar

In figure 2 the authors presented the function of

the probability of detecting wind farms in the same

conditions which are shown in figure 1. The course

emerges as a result of adding the above-mentioned

courses and it is easier to interpret. One can see that

the wind farm is clearly visible on the radar screen

from the smallest distances to 17 NM with one case of

disappearance (up to 50% probability, so in this case it

is visible at every second antenna revolution) at the

distance of 9 NM.

Figure 3. The distribution of probability of detecting a band

X radar

Figure 3 shows the probability of detection for

radar X band in good propagation conditions, with

sea state 3 and no rainfall. The signal fluctuates as

near as 3 NM from the antenna, but it does not pose

any problems up to 5 NM. Over the 5 NM distance

signal loss is detected caused by its interference with

the signal reflected by the sea surface. The echo signal

from the power plant appears and disappears from

the screen. The situation persists until the distance

reaches 20 NM. In this area, it is possible to set own

position based on the power plant echo, but it

requires lengthier radar observation. On the radar

working in band S the wind farm echo is visible

without disappearing, whereas in X band the loss of

echo complicates matters.

Figure 4. The probability of detecting X band radar

In figures 4 and 5 the probability of detecting the

power plant during observation of X band in adverse

weather conditions, i.e. the wave height is 5m, the

wind speed over 15m/s and moderate rainfall, is

shown. In figure 4 the probability is shown for a

lower sea state and without rainfall. Up to 2NM, the

power plant is visible (probability between 60 and

80%), and after that, the visibility improves, as the

interference from sea waves diminishes. Over 10 NM

large signal loss can be detected.

Figure 5. The probability of detecting band X radar

Fig. 5 shows the probability function of band X

detection for very adverse weather conditions (sea

328

state 8 and 4mm/h rainfall). High waves and low

rainfall nullify the possibility of observing the wind

farm. The probability for a distance below 1 NM is

over 60%. For distance from the antenna over 2,5 NM

the probability of detection fluctuates between 50 and

30% up to 8 NM, after which the wind farm’s echo

disappears and can only be seen for brief moments.

The maximum range of detecting the power plant in X

band is lower than in the S band by about 20%.

Table 3. Observations for S band and X band

_______________________________________________

Sea S band X band

state Max Losses Rainfall Max Losses Rainfall

range [mm/h] range [mm/h]

[NM] [NM]

_______________________________________________

3 17.1 8.0-9.0 no 13.7 3.0 1

fluctuation

3 17.2 8.0-9.0 4

7 17.1 8.0-8.5 1 12.9 2.0-5.0 1

(60%)

7 17.1 9.0-10.5 4 12.9 2.0-5.0 4

(40%)

7 17.1 9.0-10.5 20 10.2 2.0-5.0 20

(below 20%)

8 17.1 8.0-9.0 1 10.2 1.5-6.0 1

(60%-70%)

6.0-11.0

(do80%)

8 17.2 8.0-9.0 4 0.5 1.0-4.0 4

(50%)

5.0-10.0

(30%-50%)

8 17.2 8.0-9.0 20 0.3 Up to 1.5 20

(50%)

Further

(0%)

_______________________________________________

Based on the results presented in table 3 it must be

stated that in S band the radar observation of wind

farms is possible in all weather conditions and the

detection range is sufficient. In X band, however, the

muffling of the signal by rainfall is high and limits the

range of wind farm detection. Observation during

high sea state in X band is difficult but achievable.

However, if in such sea state rainfall also appears,

even slight, then the possibility of detection is limited

or even impossible. Results like these have been

recorded while research was conducted in the vicinity

of other wind farms. [3] [4].

Comparing the results obtained utilizing the

CARPET software with the wind farm observation

during cruises it must be stated that the farms are

easily visible in X band at distances over 20 NM. It is

probably the result of a simplified model of radar

cross-section for the wind farm. It must be stated that

wind farms are clearly visible from the sea and can be

used to reinforce the position of the ship. No

additional ship equipment is necessary for this

purpose.

5 CONCLUSION

Wind farms, due to their large size, are clearly visible

to the naked eye and to the radar from a significant

distance, but they can also cause a false echo to

appear on the screen and mask other objects and

navigational signage.

The Carpet 2 simulation software, which was

utilized for this research, enables the analysis of radar

signal detection from a significant distance, from

different objects and in different weather conditions.

Utilizing this software allows for a quick analysis of

the influence on the operation of the radar of

meteorological conditions and other parameters, such

as radar cross-section, the height of the antenna. The

Carpet 2 software can be successfully utilized as a

teaching method for naval radiolocation.

As a result of the simulations performed, it can be

stated that the distance from the radar and

meteorological conditions influence the probability of

detection of the wind farm. The wind farm, treated as

an area where wind turbines are situated, is visible on

the radar from a significant distance thanks to its size.

It is a non-moveable construction, clearly visible and

one that can function as navigational assistance.

The calculations made in this research show the

probability of detection of the wind farm by a ship

radar in diminishing weather conditions, here the

only way for safe cruising is the radar, and the

distance at which the wind farm is detectable can

have a significant influence over the performed

navigation.

The research was done for a radar working in the

bands X and S. Based on the analyzed scenarios, it is

clear that there is a significant difference in detection

of the wind farm between those two bands. For X

band the probability of detection of the farm below

the distance of 4NM is around 80%. In this range, the

possibility of detection of the wind farm is high

regardless of the weather conditions.

The larger the distance between the wind farm and

the radar the more fluctuations occur, meaning that

the image appears and disappears.

In the range between 4NM and 8NM the

possibility of detection falls below 50%, and in the

range of up to 20NM there are areas of high

probability of detection alongside areas where the

echo disappears.

For the radar working in S band usually only a few

occurrences of lack of detection of the wind farm

appear. Comparing the analyzed scenarios, it can be

observed that the lack of visibility of the wind farms

is the highest in mild meteorological conditions and

falls alongside the diminishing of the weather

conditions. The conclusion is that radars provide

better visibility and detection in diminishing weather

conditions.

The probability of detection for this band was

around 75 to 90% with the sea state at 8. In these

research conditions, the object is better detectable in S

band.

The great height of the wind generator turbines

and their metal construction make for a wide radar

cross-section, that is why the occurrence of indirect

multiple echoes is possible from side lobes as well as

generating shadow sectors.

In the case of the radar that works in the X band, a

large possibility of detecting a wind farm at the

329

distance of 5 NM exists, however, this distance is

prolonged to 10 NM for a radar working in S band. In

that range one would get a clearer image on the

approach to the wind farm using S band rather than X

band. However, if indirect distance is smaller than

1NM, the S band radar will register a high volume of

interference caused by the signals received on the side

lobes’ direction. The image on the radar working in S

band will be unclear, and interfering signals will be

plentiful. The antenna of the radar that works in a

longer wavelength by definition has worse

characteristics, as the level of side lobes diminishes

concurrently with the antenna's growth in size as

calculated in relation to the wavelength. At small

distances from the wind farm the radar working in X

band is more useful than S band radar due to the

antenna's higher parameters. The antenna of the radar

used is situated 20m above sea level.

REFERENCES

[1] Albert G. Huizing and Arne Theil. CARPET Version 2.0

(Computer Aided Radar Performance Evaluation Tool).

Manual Version 1.2. Developed at the TNO Physics and

Electronics Laboratory by Albert G. Huizing and Arne

Theil.

[2] David Rugger, Alan Pieramico, Terry Koontz. Appendix

M Report of the Effect on Radar Performance of the

Proposed Cape Wind Project and Advance Copy of

USCG Findings and Mitigation. USCG Order #HSCG24-

08-F-16A248 Cape Wind Radar Study. TSC Technology

Service Corporation, Trumbull 2010.

[3] L. S. Rashid, A.K. Brown. Impact Modelling of Wind

Farms on Marine Navigational Radar. MACS

Engineering Research Group School of Electrical &

Electronic Engineering University of Manchester,

UK,2010. From:http://www.supergen-

wind.org.uk/Phase1/docs/ Rashid,Brown-Supergen

Oxford2007.pdf

[4] Roy Baker. Wind Farm Effects on Marine Radar. Marine

& Risk Consultants Ltd. MARICO House, Bramshaw

Southampton SO43 7JB. From: www.marico.co.uk

[5] Tadeusz Stupak, Ryszard Wawruch. Problemy Instalacji

Farm Elektrowni Wiatrowych na Morzu. Opracowanie

dla Urzędu Morskiego w Gdyni, na prawach rękopisu,

2009.

[6] http://www.raytheon-radary.az.pl