A41F-0179 Application of Satellite Data to Develop Wind Potential Model: A Case Study of Pakistan Coastal Belt Zeeshan Alam Nayyar1,2, Nayyer Alam Zaigham2 1
Renewable Energy Research Group, Department of Applied Physics, University of Karachi, Pakistan 2 Faculty of Engineering, King Abdulaziz University, Jeddah, Saudi Arabia
[email protected],
[email protected] Abstract Since the independence in 1947, the Pakistan relies on the conventional resources for the generation of electricity. Since the local production of fossil fuel is not sufficient to fulfill the growing need of the country, the major economic burden involves huge import of petroleum products. In such a situation, the renewable energy resources are imperative in view to substantiate the economic burden. Wind energy resource is one of them, which is freely available and environmental friendly in nature. Pakistan is the late starter in the field of wind energy technology mainly because of the unavailability of the baseline wind data. As such, the adequate wind modeling and identification of the potential areas are imperative for the development of wind energy technology in the country. Present research study is carried out, based on the available satellite-collected wind data, to establish the rational wind potential model(s) of lower Indus Plains and Sindh coastal areas of Pakistan. The results of the present study reveals interesting pattern of the wind energy potential demarcating the higher wind energy resource zones and indicating hot spots for the future wind-farm installations. This paper describes the use of available satellite-collected wind data in the demarcation and modeling of wind potential along the lower Indus coastal belt and the methodology could be replicated on other parts of Pakistan and/or other counties.
01 – I n t r o d u c t i o n In comparison to the rapid increasing trend in population, the growth trend of energy generation should have been proportionally maintained to achieve the targets set for a country, but it is not the case in Pakistan. It is because of lack in planning for the energy generation and the development capabilities for rational identification of energy sources during the last sixty years[1]. Since past more than two decades, the Pakistan government’s efforts are on way to speed up the development of energy generation capabilities out of petroleum products, coal reserves and hydel power, in addition to secondary consideration of utilizing wind energy and other renewable sources through public and private sectors as well[2]. In Pakistan, in fact, the preliminary efforts for the exploitation of wind energy were made since early 1960’s by installing wind mills for pumping shallow groundwater on a very limited scale for the development of agricultural activities at few locations in Sindh and Balochistan areas[3,4] by Water & Power Development Authority (WAPDA) and Agriculture Development Bank of Pakistan (ADBP). In general, all the R&D work relevant to wind energy development remained the concern of public-sector organizations till early of this decade. During the last few years, the private sector is also taking keen interest for the development and exploitation of wind power. In spite of all efforts by the public and private sectors, the wind energy generation on commercial scale is almost insignificant except the community-based wind installations by some of the public-sector organizations. In spite of all efforts made during last about 60 years for the development of wind energy technology, the generation of commercial-level wind energy in Pakistan was not developed as it should have been. Although, Pakistan has about 1000 km long coastline[5], which could be utilized for the installation of wind farms, but the installed grid-connected wind energy contribution of Pakistan is still virtually at zero in comparison to some of the neighboring countries of the region like China, India, Iran and Turkey. This is because there is no adequate data collection and wind model generation, and the rational thinking to utilized the easily available energy.
Digital Elevation Model in 16-bit raster format of the study area was acquired from the United State Geological Survey (USGS) for the study area from their global 30 arc-second elevation dataset (GTOPO30) with a grid spacing of 30 arc seconds. Digital data containing the geomorphic and physiographic information covering the study area were also collected from the Digital Chart of the World (DCW) dataset, which is the product of the Environmental Systems Research Institute Inc. (ESRI), U.S.A. The resolution of the DCW data sub-set is 1o x 1o longitude/latitude and contain populated places, railroads, roads, utilities, land cover, aeronautical points and cultural landmarks. The satellite imageries of dust storms and cyclones were collected from time to time. The true-color NASA’s Terra and Aqua satellite Moderate Resolution Imaging Spectroradiometer (MODIS) satellite images of the dust storms have been collected for the period from 2003 to 2006. Similarly, the satellite images of the cyclones have also been collected from National Oceanic and Atmospheric Administration (NOAA) from polar (N4, N5, N6) and geostationary (G8, G10, G12) satellites over the period of 1998 to 2007.
As for the last GIS processing step, the wind power density (WPD) for the lower Indus plains and delta region was calculated utilizing the primary and secondary thematic layers by the formula suggested by Hughes (2000)[17]:
WPD = 0.5 ⋅ k ⋅ ρ ⋅ (wind_speedmean
)3
where
(a)
k is the Weibull shape factor, and ρ is the air density.
(b)
(c)
(d)
Figure 4 Maps showing wind power densities for the lower Indus region for the seasons of (a) Spring, (b) Summer, (c) Autumn, and (d) Winter.
Finally, the wind power density model for the lower Indus plains and delta region of Pakistan for the standard heights of 30m & 50m were generated by subtracting the secondary thematic layer of excluded-land from the annual wind power density maps and by combining the results with the layer of the natural barriers impact. The final wind power density model is presented as Figure 3.
(a)
Since the collected wind and meteorological data are from different sources and are of different nature, initially, statistical processes were made to normalized the data and to adjust the error to minimum. The resultant data have been used to develop the GIS based wind potential model. The GIS-based model that have been developed contain thematic layers having resolution of 30 arc-second. The UTM WGS-1984 zone geographical projection system was used to represent data/map arrangement in each thematic layer. The GIS-linked data were digitally processed or interpolated in terms of the thematic layers for the whole area but is only represented for the land areas.
(b)
Figure 5 Satellite images of (a) tropical cyclone 03B from June 25 – 28, 2007 in the Arabian Sea, and (b) dust storm dated March 02, 2003, that affected the lower Indus Delta region of Pakistan (Source: NASA).
The wind potential model for the lower Indus plains and delta region of Pakistan was prepared by utilizing the data processed earlier. In the development of wind model, the main consideration was to incorporate all the features that were lacking in the previously proposed studies/models[6,7,8,9,10] as much as possible. For the present study, the parameters were given priority to develop comprehensive & sustainable wind assessment and wind potential model; and all the data were made normalized for the standard heights of 30m & 50m; roughness length map of the study area was developed to provide a scaling factor for the interpolation/extrapolation of the wind speeds for the different heights; the wind potential model was not only developed by the wind and related metrological data but were also integrated with the natural barriers like dust storms and tropical cyclones originated in and around the vicinity of the study area; land-use, including demographic and urbanization features, were also taken into consideration in the development; Weibull distribution parameters, on the monthly as well as annual basis, were calculated and incorporated. The steps taken are presented in the 3-tier flow chart given in Figure 2.
05 – Conclusion In general, the wind power density model of lower Indus plains and delta region of Pakistan shows the high potential and low potential anomalous areas in relation to the impact of micro-level features and the natural barriers explicitly. The model show the conducive wind energy environments for the commercial-level & community-level wind installations more or less all over the area.
Figure 3 Wind power density model for the lower Indus plains & delta region of Pakistan.
References [1]
Processed Wind & Relevant Data
Land-Use
Transportation
Natural Barrier
Annual
Elevation
Monthly
Seasonal
04 – R e s u l t a n d D i s c u s s i o n Temperature
Seasonal
Annual
Wind Speed
30 m
Monthly
The lower Indus plains and delta region is comprised of Karachi, Thatta, Badin, Mirpur Khas, northwestern tip of Umarkot, western part of Thar, Sanghar, southern part of Nawabshah, Dadu, southern part of Khuzdar and eastern parts of Lasbella districts. It is bounded by the Arabian Sea in the southwest and India in the southeast. Based on the result of present study, two excellent wind power density anomalies have been modeled in this region (Figure 3). A third anomalous zone is partially found in the northwestern corner of the present block. This anomaly seems to extend further northward that needs additional research work in future.
Seasonal
50 m
Annual
The anomaly, along the deltaic coast follows the coastal alignment in northwest-southeast direction. This anomaly covers major portion of eastern part of Karachi and western part of Thatta districts and is lying mainly on the western side of the Indus River. It has characteristic wind power density values from class-2 to class-7. The class-7 wind potential anomaly zone covers the marshy areas of Indus deltaic coastal part. This zone terminates at the location of Jangshahi. Southeastern part of Karachi, Gharo, Mirpur Sakro and Garho lies in the prospective zone. Keti Bandar lies at the moderate zone.
Primary Thematic Layers
Figure 1 Index map of the Study Area.
Present paper describes the details of the adopted research approach for the development of above mentioned GIS linked WPD models for two different heights of 30m and 50m.
Slope & Aspect Roughness
Excluded Land
Shape Factor ‘k’
Air Density
Secondary Thematic Layers
The western larger elongated excellent wind power density anomaly lies on landward side in north-south orientation. The anomaly extends from southern part of Nawabshah upto Run-of Kutch in the north-south alignment and from eastern part of Dadu to the northwestern tip of Umarkot. Hyderabad lies almost in the central part of the anomaly. The excellent wind potential of class-7 is spread over a very large area in comparison to other anomalies found in the study area. Similarly, good and moderate wind power density area is much larger. It is inferred that this anomalous zone represent the main target areas for the commercial-level wind energy exploitation in the Sindh province.
Wind Power Density Model 30 m
02 – Data Collection
50 m
Output Layers (Model)
Figure 2 Flow diagram indicating the GIS processing steps for the development of wind power density model for lower Indus plains and delta region of Pakistan.
The wind and the other relevant data and information were acquired from many sources including NASA, WUI, USGS, NOAA, ESRI, PMD, etc. for the estimation and generation of wind potential model of the study area. The 10-year monthly average wind and other meteorological parameters including average, minimum, maximum and three-hourly intervals averaged wind speed for 50m and 10m; frequency distribution of wind speed for 50m; wind direction and three-hourly intervals wind direction at 50m; average air temperature and three hourly intervals averaged air temperature at 10m; average daily temperature range; etc. were acquired from Surface Meteorology and Solar Energy (SSE) Section of National Aeronautics and Space Administration (NASA) on a 1o x 1o grid system for whole working area.
The wind and related data, processed earlier, were used to create the primary thematic layers. These layers served as the input(s) for the spatial analysis to calculate the meso-scale wind power density. The primary thematic layers include salient cities layer; transportation network (roads, rail tracks & airports)layer; elevation layer; land use (distorted surface, inundated area, wetland, city area, cropland/grassland, forest, water bodies, barren land, game reserve, national park, sanctuary & wildlife) layer; natural barrier impact (dust storm & cyclone) layer; temperature layer on seasonal and annual bases; wind speed spatially interpolated layers on monthly, seasonally and annual bases for 50m & 30m.
The wind and meteorological data including average wind speed, average wind direction, and average ambient temperatures for Karachi, Badin and Hyderabad were collected from Pakistan Meteorological Department (PMD) archive for the period of January 1942 to December 2002 for different anemometer heights.
To generate the wind speed layer(s), the spatial interpolations were implemented by using the regularized spline formula, as suggested by Franke (1982) and Mitas & Mitasova (1988)[15, 16]:
The wind and related meteorological data were also collected from the Weather Underground Inc. (WUI) in addition to the PMD and NASA datasets for Karachi, Badin and Hyderabad airports. The data is for the period of January 2004 to December 2005 and include 30-minute/6-hourly mean & maximum wind speed at 10m, and mean, maximum & minimum temperature at 10m.
υ(x,y) = ξ(x,y) + Σ(j=1,n)[λk R(rj)]
Beside the wind and other general meteorological data, certain other parameters are also considered very essential to develop the wind potential model(s) of a particular region. In this connection, a
By combining the primary thematic layers and the processed data mathematically, secondary thematic layers were generated. The secondary layers include air density layer on monthly and annual bases; slope and aspect ratio layers; roughness length layer; excluded land layer; and Weibull shape factor k layer.
03 – D a t a P r o c e s s i n g
Cities
Considering the technical shortcomings of the studies conducted earlier for the assessment and exploitation of wind energy potential in the country, the present study was undertaken for the development of GIS-linked wind potential model for the southeastern part of Pakistan comprising of lower Indus River plain and its delta region covering about 90,000 square kilometers (Figure 1).
For the interpolation, the value of r is set equal to 0.00833 (i.e. 30 arc-second), τ equal to 0.1 and N equal to 12.
where
ξ(x,y) = α1 + α2x + α3y, R(r) = 1/2π {r2/4 [ln (r/2τ) + ς - 1] + τ2 [Ko (r/τ) + ς + ln(r/2τ)]},
n = number of data points, j = coefficients found by the solution of a system of linear equations, rj = distance from the point (x,y) to the jth point, τ2 = parameter influencing the character of surface interpolation, Ko = modified Bessel function, ς = constant equal to 0.577215, αi = coefficients found by sol. of a system of linear eqs.
Based on the seasonal variation behavior of the wind densities (Figure 4), it is observed that this anomalous zone is sustainable more or less all over the year except during the winter season, when the wind power density relatively decreases to class-4. On the other hand, the eastern anomaly of wind power density is not sustainable over the year i.e. wind power density decreases to class-1 (background value) during winter period. It is inferred that the best wind potential in this coastal part may be available during the monsoon period.
Nayyar Z. A., 2009, Investigation & development of GIS-linked aerodynamic wind potential models of the Sindh and Balochistan coastal areas of Pakistan, Ph.D. Dissertation (unpublished), University of Karachi, Pakistan, 334p. [2] Zaigham N. A. and Nayyar Z. A.,2005,Prospects of Renewable Energy Sources in Pakistan: In Renewable-Energy Technologies and Sustainable Development, Khan H. A., Qurashi M. M., Hussain T., Hayee I. (eds.), COMSATS’ series 4, 65-86. [3] Shah M. A., 1991, Wind Power-an answer for socio economic uplift of rural areas: In Wind Energy for Rural Areas, Irene de Jong and Frans Van Hule (eds.), 89-102. [4] Futehally M. A., 1991, Wind Energy and OFIT: In Wind Energy for Rural Areas, Irene de Jong and Frans Van Hule (eds.), 59-63. [5] Nayyar Z. A., 2004, Identification of the Wind Channels for the Prospective Wind-Energy Generation Sites in Pakistan by using Satellite Imageries: In Special Digital Publication of ISNET on Satellite Technology Applications in Communications and Remote Sensing, 62-69. [6] Nasir S. M., 1993, Estimation of wind energy potential in Pakistan: Ph.D. Dissertation (unpublished), University of Balochistan, Pakistan, 251p. [7] Khan N. A., 2001, Wind Mapping of Pakistan: Report submitted to Pakistan Council for Appropriate Technology, Pakistan, 18p. [8] Elliott D. L., 2007, Wind resource assessment and mapping for Afghanistan and Pakistan: Report submitted to NREL-USA, 18p. [9] Empower, 2010, Welcome to Empower [Online], Available: http://www.mft.govt.nz. [10] Hodgetts R. D., 2003, Assessment of the wind measurements Undertaken in Pasni Pakistan: Document # 2814/BR/01, UNOPS, 34p. [11] Franke R., 1982, Smooth interpolation of scattered data by local thin plate splines: Comp. & Maths. With Appls., Vol . 8, 237-281. [12] Mitas L. and Mitasova H., 1988, General variational approach to the interpolation problem: Comput. Math. Applic., Vol . 16, 983-992. [13] Hughes T., 2000, Oklahoma wind power tutorial series - lesson III: Environmental Verification & Analysis Center, University of Oklahoma, 4p. [14] NASA, 2008, Tropical Cyclone 03B – Natural Hazards [Online], Available: http://earthobservatory.nasa.gov/NaturalHazards/view.php?id=18567. [15] NASA, 2006, Visible Earth – Dust over the Arabian Sea [Online], Available: http://visibleearth.nasa.gov/view_rec.php?id=5110.
Considering the natural barriers as demarcating from the satellite imageries of NASA and NOAA, it is observed that this area is subjected to heavy impacts of the frequent tropical cyclones (Figure 5a)[14]. The impact of cyclones appear more on the anomaly observed in the delta region as compared to the eastern anomaly observed inside the land area within the lower Indus basin. Moreover, the western anomaly is more affected by the finer sticky particles of the dust storms from the Arabian Sea (Figure 5b)[15] as compare to the eastern anomaly. Whereas, the areas having eastern wind density anomaly are expected to be affected by the dust storms from the Thar desert region in addition to dust storms from the Arabian Sea.
- - - - - - - - - - A c k n o w l e d g m e n t
- - - - - - - - - -
The authors acknowledge many thanks to the King Abdulaziz University, Saudi Arabia for extending full financial support and cooperation for presentation of the research work. The authors are also greatly thankful to National Aeronautics and Space Administration (NASA), Computerized Data Processing Center of Pakistan Metrological Department (CDPC-PMD), Weather Underground Inc. (WUI), US Geological Survey (USGS), National Oceanic and Atmospheric Administration (NOAA), Environmental Systems Research Institute (ESRI) for providing the required data necessary for this research work. This paper is extracted from the Ph.D. dissertation of the principle author.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - S h o r t
B i o g r a p h y
Zeeshan A. Nayyar received his Ph.D. degree in GIS-linked Wind Potential Model of Pakistan from the department of Applied Physics, University of Karachi in 2009. Presently, Dr. Nayyar is working on GIS modeling and data integration for a research project of faculty of Engineering, King Abdulaziz University, Saudi Arabia in the capacity of Assistant Professor. Concurrently, he is also holding Assistant Professorship at the University of Karachi, Pakistan. He was the principal investigator for a research studies “Investigation and development of GIS-linked aerodynamic-wind potential models of the Sindh and Balochistan coastal areas of Pakistan” during 20052009, and “Determination of Wind Profile of the Sindh Coastal Area for the Development of Wind Energy Technology” during 20032004. He has a visiting scientist status in the Abdus Salam International Center for Theoretical Physics (ICTP), Italy and also attended research programs in Italy in 2001 and 2003. Furthermore, he server as a technical expert to Power Department of Sindh Government for Wind-Solar Renewable Technologies during 2004-2006. Dr. Nayyar was a invited speaker to deliver a lecture on ‘Identification of the Wind Channels for the Prospective Wind-Energy Generation Sites in Pakistan by using Satellite Imageries’ at Iran Space Agency, Tehran in 2004. He has published and presented more than 15 research papers in international & national journals/conferences. He is also a co-author of book chapter entitled ‘Prospects of Renewable Energy Sources in Pakistan’, In Renewable-Energy Technologies and Sustainable Development, Khan H. A., Qurashi M. M., Hussain T., Hayee I. (eds.), COMSATS’ series 4, 2005, 65-86. He is member of many international organizations including Global Wind Energy Council.
