Assessment of wind and wave energy potential along the Indian coast

Abstract

The focus is now on sustainable development, which is inevitable without harnessing renewable energy sources. The fundamental element in wind wave generation is the interaction between air and sea which helps in momentum exchange between atmosphere and ocean. The Indian coastline is under a dynamic wave climate with the action of wind. Indian landmass has two tropical basins, the Bay of Bengal and the Arabian Sea, which have tremendous potential to tap renewable energy. The variations in wave climate due to dynamic-wind have to be assessed. Hindcast data obtained from Global Climate Models help us in the long-term analysis of wind and wave climate. In an attempt to explore the renewable energy potential along the Indian coast, a numerical wave model is developed using MIKE 21 SW module to assess the wind and wave climate. A gridded global wind speed dataset from ECMWF called ERA-Interim wind speed data of 38 years (1981 to 2018) is used as input for the numerical model. The dataset and numerical model performance were validated against in-situ measurements. The results showed amongst the locations studied off Goa, Karnataka, Kerala, Tamil Nadu, and Andhra Pradesh had good potential to extract offshore wind energy using offshore wind turbines.

Keywords:

• Wind speed
• long-term analysis
• ERA-Interim
• MIKE21
• wave power
• offshore renewable energy
• climate change

Reviewing Editor:

• Michele Iervolino, Università degli Studi della Campania 'Luigi Vanvitelli’, Italy

SUBJECTS:

• Hydraulic engineering
• Marine and offshore structures
• Water science

1. Introduction

The diminishing supply of fossil fuels and the increase in global energy demand has shifted the focus on environment-friendly renewable energy. The increase in the population has increased energy consumption and intensified greenhouse gas emissions. Tropical regions like India have the potential to extract ocean energy, but there are several challenges to renewable energy extraction (Felix et al., 2019). Sannasiraj and Sundar (2016) identified that the technology of ocean energy is a constraint for exploiting ocean energy. However, there is constant research and development in this field. Waves, tidal, thermal, and offshore wind are potential ocean energy sources. Wind energy is an efficient form of renewable energy, but it requires a detailed analysis to identify a potential location (Chaurasiya et al., 2018). The long-term analysis helps in assessing the wave climate of a particular region (Chaurasiya et al., 2018; Upadhyaya et al., 2020). Analyzing the wind and wave climate not only helps Oceanographers but also helps decision-making agencies for coastal management and assess the prospects of ocean energy. Indian domain experiences wind from the southwest direction during monsoons (June-September) and reversal winds from the northeast direction during winter monsoons (October to January). Wind power forecasting using numerical prediction models on a global to local scale is important for potential future development in the field of wind power generation (Foley et al., 2012).

Offshore Renewable Energy (ORE) as a clean energy source has emerged and options of integrating wave energy converters with offshore wind turbines are given more emphasis as it is economical (Pérez-Collazo et al., 2015). Most countries are investing in renewable energy extracted from offshore wind and ocean waves. ORE generates energy from wind at sea which is unobstructed compared to onshore wind fields, and in addition, ocean energy through waves, tides, or thermal can be extracted. Guedes Soares et al. (2014), in their review on ORE, mention that the past decade has seen the evolution of ORE platforms from shallow water fixed types to floating types. Bagbanci et al. (2012) felt this technology with floating platforms has promising prospects when used for large-scale generation. In order to explore the available marine sources and utilize the ocean space effectively, there has been substantial progress in the development and design of very large floating structures (Praveen et al., 2020). This study focuses on the feasibility of offshore fixed support structures with an offshore wind turbine at shallow water depths. The fixed offshore platforms can be a gravity type, jacket type, suction bucket type, or monopile type structure. The design principles from offshore platforms for oil and gas exploration can be used. The cost of the offshore structure is proportional to the water depth. Hence, the location selected in this study has water depths varying from 45 to 55 m. Additionally, the wave power at selected nearshore locations is also evaluated to assess the feasibility of a combined ORE platform along the Indian coast at shallow depths.

Kumar and Anoop (2015) mention that the wave characteristics can be studied as a time series if the data is captured in a spatiotemporal manner. The scarcity of in-situ measurements along the Indian domain has resulted in hindcast studies based on global modeled and satellite datasets. ERA-Interim is a product of the European Centre for Medium-Range Weather Forecasts (ECMWF), which provides hindcast values of various atmospheric and oceanic parameters. These gridded data are popularly known as Global Climate Models, which can be effectively used for numerical model studies (Umesh et al., 2017). There are limited studies to assess the wind and wave potential along the Indian coast. Hence, In the present study, wind data from 1981 to 2018 (38 years) corresponding to the Indian domain (4˚ S to 30˚ N and 40˚ E to 95˚ E) is analyzed. This wind forcing is also used to simulate the wave climate using the MIKE21 numerical model. The results obtained from this hindcasting study will help to identify the potential location across the Indian coast. In-situ data is used as a reliable measure for validating model results.

1.1. Study domain

India has a coastline of about 7500 km passing through nine states. The Indian Ocean, Southern Hemisphere, and Arctic regions have a higher probability of extreme waves and wind (Dobrynin et al., 2012). The Arabian Sea region is exposed to high winds and wave activities during monsoons (June to September) and is otherwise calm. South-West is the predominant wind direction during monsoons, and it shifts to North-East during winter monsoons from October to January (Vikas et al., 2015). The wave climate is severe during the southwest and northeast monsoons (Sandhya et al., 2018). The changes in wave climate will affect the operations of offshore platforms as it increases the risks of damage (Reistad, 2001; Ruggiero et al., 2010). There are many potential areas for development where the resources can be harnessed along the Indian coast (Figure 1).

Figure 1. Typical mean Significant wave height variation along the Indian domain with location details for the study (−4° to 30° N 40° to 95°E).

1.2. Oceanography data products

Location-specific long-term measured data of ocean parameters are sparse, which has resulted in fewer regional studies (Hithin et al., 2015). Compared to the Atlantic and Pacific region, the Indian Ocean region has limited in-situ observations (Patra and Bhaskaran, 2016). Oceanography products are mostly mathematical equations modeled based on the ocean and atmospheric processes. Extensive computation facilities are required for studies of this magnitude which will be taken up by government metrological agencies supported by other bureaus or agencies. The use of wind data in ocean modeling is a common practice. The accuracy of the dataset plays a vital role in wave simulations obtained from the numerical model (Cucco et al., 2019).

The European Centre for Medium-Range Weather Forecasts (ECMWF) provides medium-range forecasts and aims for accurate prediction of climate data. Among the various products offered by ECMWF, ERA-Interim is one such reanalyzed dataset that spans the entire twentieth century. Dee et al. (2011) mention ERA-interim as a gridded data product, where the data assimilation technique is used. Through ERA-Interim data, details of zonal and meridional wind data for a particular latitude and longitude can be obtained at a definite time interval. The wind speed data (m/s) is available from January 1979 and is updated in real-time with a delay of two months.

Assessing the historical variability of wind speeds based on the ERA-Interim dataset is one of the objectives of this study. Studies of this kind are relevant as the changes in wind patterns due to climate change alters the wave climate (Dobrynin et al., 2012). In turn, this will alter the wave loads that the marine structure will be subjected to.

2. Methods

Numerical models provide quick and realistic simulations based on the accuracy of data imputed. Global climate models contain downscaled data of variables at larger spatial grids. These global datasets are highly regarded in the scientific community as they provide reasonably good estimates closer to the in-situ data (Umesh et al., 2017). In this study, a numerical wave model was developed using MIKE 21 SW module to simulate the hindcasted wave climate. The spatiotemporal wind forcing provided ERA-Interim is used to assess the nearshore wind and wave energy potential along the Indian coast.

2.1. Wind speed dataset

The daily wind speeds (m/s) at 10 m above sea surface in the eastward direction (U₁₀) and northward direction (V₁₀) is downloaded from the ECMWF website. The wind data with a grid size of 0.5° x 0.5° for the Indian domain (Figure 2) from 1981 to 2018 is used as an input to the numerical wave model. In-situ data for the year 2013, corresponding to 84.00°E, 13.50°N (BD11) installed offshore of Tamil Nadu coast, is used to validate the wind speed values.

Figure 2. The resultant mean wind speed (m/s) variation along the Indian domain considered for the study (−4° to 30° N 40° to 95°E).

2.2. Data processing

Processing of this huge data of wind speed is performed using the FERRET tool by National Oceanic and Atmospheric Administration (NOAA) which works on the Ubuntu platform. FERRET is a versatile tool that can analyze, process, and visualize climate data. The downloaded data is imported to FERRET, which is followed by operations like re-gridding, computing resultant wind speed, and monthly mean wind speed, followed by exporting the data into the required format.

2.3. Numerical modeling

MIKE21 SW, a spectral wind-wave model, was developed by the Danish Hydraulic Institute (DHI). MIKE 21 SW is a third-generation spectral wind-wave model based on unstructured meshes (DHI, 2015). MIKE21 SW simulates wave parameters spatially by solving energy and mass balance equations. (Sørensen et al., 2005). (1) $$\frac{\partial N}{\partial t}+\nabla .\left(\mathrm{ṽ}N\right)=\frac{S}{\mathrm{\sigma }}$$(1)

Where, N (x̅, σ, θ, t) is the density function, with t = time, x̅ is (x,y) in Cartesian coordinate, S = S_in+S_nl+S_ds+S_bot+S_surf with S_in being momentum transfer due wind, S_nl corresponds to energy transfer due to non-linear energy function, S_ds is energy dissipation due to white capping, S_bot is energy dissipation bottom friction, and S_surf is energy dissipation depth induced wave breaking parameter, ṽ= (c_x, c_y, c_σ, c_θ) is wave group propagation velocity in four-dimensional phase space of x̅, σ, and θ with a differential operator ∇.

The wind and wave climate has been simulated for 38 years. The daily wind speed ERA-Interim data from 1980 to 2018 is used as an input for the numerical model. The bathymetry is derived from C-MAP (Figure 3). The wind and wave parameters like significant wave height (H_s), wave period (T), wave power (P), wind speed, and wind direction were extracted at nine locations across 38 years.

Figure 3. Bathymetry details used as input file (.mesh) for MIKE 21.

2.4. Validation

A short-duration validation of historical data is performed against buoy measurements. The validation is against buoy measurement recorded by BD11 (Figure 4) deployed offshore of Tamil Nadu coast (Latitude 13.50°N, Longitude 84.00°E). The daily wind speed data measured for the year 2013 is compared with ERA-Interim wind speeds. The buoy measured wind speed at 3 m from sea surface level, whereas ERA-Interim wind speeds were at 10 m height. Hence the required conversion is performed by the commonly used Hellmann exponential law (Suvire 2011), (2) $$\frac{\mathit{V}}{{\mathit{V}}_{\mathit{o}}}=\frac{{\mathit{H}}^{\mathrm{\alpha }}}{{\mathit{H}}_{\mathit{o}}}$$(2)

Figure 4. Daily wind speed variation at BD11 along with corresponding ERA-Interim values for the year 2013.

Where,

V is the wind speed at height H, and Vo is the wind speed at height H_o = 10 m

Frictional factor α = 0.1 (for Lakes, ocean, and smooth, hard ground)

Daily wind speed variation at BD11 along with corresponding ERA-Interim had an R-value of 0.84 and an RMSE of 1.59 m/s for the year 2013. ERA-Interim modeled wind speed performance is good, which is reflected in the scatter plot (Figure 5(a)). A similar comparison was performed in the previous study on a bouy data AD02 located off Goa coast for the year 2011. The measured daily wind speed shows a good Correlation with ERA-Interim dataset values with a Correlation coefficient (R) of 0.93 and RMSE of 1.29 m/s for the year 2011 (Upadhyaya et al., 2021). Hence, the gridded ERA-Interim dataset can be considered a reliable dataset as the wind speed variation for the larger Indian domain is captured effectively.

Figure 5. Scatter plots of daily mean wind speed values (m/s) at BD11 and Simulated significant wave heights (H_s) at OB03.

The numerical model output has been quantitatively validated against an in-situ measurement recorded by buoy (OB03) off Mangaluru coast (Latitude 12.50°N, Longitude 72.01°E) for the year 2005. The simulation period is set depending on the buoy data availability for validation; hence start and end period of the simulation is accordingly set. The offshore location OB03 had an R-value of 0.66 and an RMSE of 1.42 m for the year 2005. The scatter plot (Figure 5(b)) indicates that MIKE simulated significant wave heights are on a higher side when compared with in-situ measurements of wave heights. The variation of significant wave heights is fairly captured, and the MIKE numerical model majorly overestimated the wave heights in both locations. The model output at the buoy location is found satisfactory considering the coarser-resolution and wider domain datasets being used in this study.

3. Results and discussion

The wind and wave climate is simulated for 38 years hindcast based on ERA-Interim wind speed data in the MIKE21 numerical model. The mean values of wave parameters in terms of significant wave height, peak wave period, and wave power are extracted at nine locations. The wave energy potential (Figure 6) is evaluated based on the mean wave power (kW/m) obtained from the equation below, (3) $$\text{Wave Power},P_{\mathit{\text{energy}}}=\rho \mathit{\text{gE}}{C}_g$$(3)

Figure 6. The mean wave power distribution in the study domain.

Where E is energy density, C_g is the group celerity of waves, ρ is the density of water, and g is the acceleration of gravity.

The mean wind speed and the distance of the selected locations from the coast, water depth, and MIKE21 numerical models hindcasting results at the nine locations are extracted and tabulated in Table 1. The plot composer tool of MIKE zero is used to obtain the time-series plot at each location with information on significant wave height (in m), wind speed (m/s), and wave direction. The typical variation of significant wave height (in m), wind speed (m/s), and wave direction in the nine nearshore locations are shown in Figures 7 and 8.

Figure 7. Variation of wind and wave climate across Gujarat, Maharashtra, Goa, Karnataka, and Kerala for the year 2000.

Figure 8. Variation of wind and wave climate across Andhra Pradesh, Odisha, and West Bengal for the year 2000.

Table 1. Details of the locations considered in the study with simulated results.

Location no	Lat. (ºN)	Long. (ºE)	Location (off the coast)	Mean Significant wave height (m)	Mean Peak wave period (s)	Mean Wave power (kW/m)	Mean wind speed (m/s)	Water depth/Distance from coast (m/km)
1	22.094	68.701	Gujarat	0.86	7.56	7.57	2.87	50/20
2	17.903	72.446	Maharashtra	0. 91	7.67	8.70	2.07	50/40
3	15.375	73.454	Goa	1.28	7.01	13.67	5.90	51/25
4	12.904	74.423	Karnataka	1.30	7.07	12.27	6.12	47/30
5	9.501	74.944	Kerala	1.52	7.22	14.93	6.14	56/80
6	12.446	80.377	Tamil Nadu	0.83	5.11	3.06	5.75	56/20
7	16.294	81.582	Andhra Pradesh	0.73	5.70	2.97	4.62	51/5
8	19.732	86.329	Odisha	0.54	6.18	1.87	1.39	54/12
9	20.921	88.427	West Bengal	0.49	6.30	1.66	1.33	54/60

From the hindcast analysis based on daily wind speed for 38 years, the Gujarat location has fair wind potential as the location majorly experiences wind speed above 3 m/s. The hindcasting study showed a maximum wind speed of 25.92 m/s. The wind directions over this region consistently ranged between 45° and 225°. The second location off the Maharashtra coast experiences higher winds only during monsoons which are above the cut-in speed of 4 m/s for wind turbine operation in a cost-effective manner (Boudia and Santos, 2019). The inconsistent wind direction also adds to the unlikely wind potential across seasons.

The remaining three southern coastal locations on the west coast have similar wind and wave climates with mean wind speeds close to 6 m/s across seasons. Based on the 38-year hindcasting data, maximum wind speeds experienced are 24.46 m/s, 29.80 m/s, and 18.66 m/s off Goa, Karnataka, and Kerala coasts, respectively. The higher wind speeds also reflect the wave climate dominated by wave heights above 1 m.

The east coast has two southern coasts off Tamil Nadu and Andhra Pradesh which have a mean wind speed above 4.5 m/s. The time series shows peaks at regular intervals, which is indicative of better wind energy potential. The other two locations such as off Odisha and West Bengal, experience a relatively calm wind climate with wind speeds below 3 m/s. The east coast of India is prone to extreme winds during cyclones which is more frequent when compared to the west coast of India, increasing the risk to devices during the extreme events at the two sites.

Analysis of wave climate across the nine locations studied indicates that the wind and wave climate is severe in the monsoon months of June to September (JJAS). The regions with yearly mean wave power of 15 kW/m have the potential to generate wave energy (Nelson & Starcher, 2015). The mean wave power at all the sites is less than 15 kW/m which is deficient in harnessing wave energy effectively at competitive prices (Chybowski and Kuźniewski, 2015).

However, Korde and Ringwood (2016) mention that there are efforts to make cost-effective wave energy extraction with low energies. The power output expected from WECs rated below 1000 kW is not significant in the Indian ocean domain (Rusu and Onea, 2017). Therefore, considering all these aspects, at present, extraction of wave energy might not be economically viable at the selected nearshore water depths. Wave energy extraction might be practical in offshore locations whose potential can be assessed by numerical model studies.

Overall, the locations of the southern states such as off Goa, Karnataka, Kerala, Tamil Nadu, and Andhra Pradesh have fair wind energy potential, which can be effectively harnessed through offshore wind turbines. As the nearshore locations assessed have water depth below 55 m, fixed offshore wind turbines like gravity type, jacket type, suction bucket type, or monopile type structure can be the options to choose from.

4. Conclusions

The study discusses the wind and wave potential at nine locations along the Indian coast. Based on the hindcasted wind and wave climate obtained from MIKE 21 driven by ERA-Interim, the following conclusions can be drawn.

Southern coastal states off Goa, Karnataka, Kerala, Tamil Nadu, and Andhra Pradesh have favourable offshore wind speeds (>4 m/s) to operate offshore wind turbines. With water depths below 55 m, the effective utilization of renewable energy is always a possibility. However, with mean wave power being on a lower side (<15kW/m) wave energy converter might not produce the required performance.

Author contributions

Dr. Sandesh Upadhyaya K: Collected and processed the data, performed numerical modeling, data analysis and drafted the original manuscript; Prof. Subba Rao: contributed to the conceptualization of the problem statement, supervision the work and review of the manuscript; Dr. Manu: supervised the work, suggested the resources and edited the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Authors agree to make data and materials supporting the results or analyses presented in their paper available upon reasonable request.

Additional information

Notes on contributors

Sandesh Upadhyaya

Sandesh Upadhyaya K, Ph.D. is a faculty in the Department of Civil Engineering, MIT Manipal with interests in Coastal Engineering, Numerical Analysis and Concrete Technology.