Fish is a kind of sustainable food for healthy diet, and it plays an important role in fulfilling human's future needs for healthy food. Norway is the world's largest fish exporter, and nearly half of the Norwegian fish come from fish farms. In recent years, Norwegian fish farm industry is developing very fast. Feed barge acts as the control center, and it is very important for operation of fish farm. The purpose of the mooting system of feed barge is to hold the feed barge in the designated position. Wind load is a low frequency force and it is important for mooring and floating structure. In other words, wind load on feed barge and its mooring system must be investigated.

The mooring system of feed barge performs as mass-spring-damping system. When the natural frequency of the system equals to the wind excitation frequency, resonance occurs. The resonance can lead to a larger tension to the mooring line. The natural frequency is related to the mass of the feed barge, the added mass and stiffness of the mooring system. Usually, the wind excitation frequency can be obtained from the power spectral density of the wind spectrum.

In this project, we take three representative Norwegian fish farms Sør Gåsvær, Nordre Skokkeløy, and Sørværet as case study. The three fish farms in exposed sites are marked as Case 1, Case 2 and Case 3. The water depth in Case 1, Case 2 and Case 3 are 100 m, 69 m and 100 m, respectively. Concrete feed barge SeaFarmBase 400T services in Case 1, steel feed barge AB650 services in Case 2, and steel feed barge AC850 services in Case 3. The mooring system of SeaFarmBase 400T in Case 1 is designed as 4 mooring chains with wet weight per length of 49 kg/m. The mooring system of AB650 in Case 2 is designed as 4 mooring chains with wet weight per length of 34.4 kg/m. The mooring system of AC850 in Case 3 is designed as 8 mooring chains with wet weight per length of 28.2 kg/m. The wind data for the three cases from dataset of 1806_Sites_wF_wWind_20022017b_wWaves.mat are long-term wind speed time series data. This data are measured every 6 hours which is not suitable to converse to frequency domain. So Kaimal spectrum, Davenport spectrum and Harris spectrum are used to model the wind in the cases. In Case 1, the mean wind speed is 6.48 m/s and the standard deviation of wind speed is 3.29 m/s. In Case 2, the mean wind speed is 4.44 m/s and the standard deviation of wind speed is 2.22 m/s. In Case 3, the mean wind speed is 6.30 m/s and the standard deviation of wind speed is 3.22 m/s. All of these values are calculated from the dataset of 1806_Sites_wF_wWind_20022017b_wWaves.mat by MATLAB.

The horizontal tension range of the mooring line is from 0.001Tmax to 0.7Tmax. In surge motion, the natural frequency ranges in Case 1, Case 2 and Case 3 are 0.0006-0.0026 Hz, 0.0005-0.0021 Hz and 0.0013-0.0067 Hz, respectively. In sway motion, the natural frequency ranges in Case 1, Case 2 and Case 3 are 0.0004-0.0018 Hz, 0.0003-0.0015 Hz and 0.0006-0.0027 Hz in respectively.

In Case 1 and Case 3, for Kaimal spectrum and Harris spectrum, the wind excitation frequency range is 0.0004-0.006 Hz. But for Davenport spectrum, the wind excitation frequency range is 0.0055-0.09 Hz in Case 1 and Case 3. The wind excitation frequency of Kaimal spectrum and Harris spectrum is 0.0003-0.004 Hz in Case 2. The wind excitation frequency of Davenport spectrum in Case 2 is 0.0039-0.06 Hz.

Based on the analysis of the results, We draw the following conclusions:

1) In sway: There are possibilities of resonance in all the three cases when the wind models are Kaimal and Harris spectra, and almost no possibility of resonance in all the three cases when the wind model is Davenport spectrum.

2) In surge: The resonance in all of the three cases are possible when the wind models are Kaimal and Harris spectra. There is almost no possibility of resonance in Case 1 and Case 2 when the wind model is Dvenport spectrum. The possibility of resonance in Case 3 is high when the wind model is Dvenport spectrum and the horizontal tension range of the mooring line is from 0.5Tmax to 0.7Tmax. There is almost no possibility of resonance in Case 3 when the wind model is Dvenport spectrum and the horizontal tension range of the mooring line is from 0.001Tmax to 0.5Tmax.

If the wind models in the three cases are Kaimal and Harris spectra, the possibilities of resonance in surge and sway are high. When resonance occurs, the vibration amplitude is large. In this situation, the wind excitation can pose a possible threat to the integrity of the feed barge due to the large vibration amplitude in surge and sway.

If the wind model in the three cases is Dvenport spectrum, there is only a possibility of resonance in surge with Case 3. In this situation, the wind excitation will not pose a possible threat to the integrity of the feed barge.