- Email: [email protected]
A GIS-based interactive web decision support system for planning wind farms in Tuscany (Italy)
Renewable Energy 36 (2011) 754e763
Contents lists available at ScienceDirect
Renewable Energy journal homepage: www.elsevier.com/locate/renene
A GIS-based interactive web decision support system for planning wind farms in Tuscany (Italy) Riccardo Mari a, Lorenzo Bottai a, Caterina Busillo a, Francesca Calastrini b, Bernardo Gozzini b, Giovanni Gualtieri b, * a b
Laboratory of Monitoring and Environmental Modelling for the sustainable development (LAMMA), Via Madonna del Piano 10, 50019 Sesto Fiorentino, Italy Research National Council-Institute for biometeorology (CNR-IBIMET), Via Caproni 8, 50145 Firenze, Italy
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 April 2010 Accepted 10 July 2010 Available online 2 August 2010
In the framework of regional renewable energy policies, starting from 2008 the Tuscany Regional Authority promoted the “WIND-GIS” project aimed at assessing the large-scale wind potential of Tuscany region, Italy. This goal was achieved by developing an integrated Geographic Information System (GIS) based decision support system (DSS), compliant with Directive 2007/2/EC of European Commission (EC), which was designed to help public operators in the preliminary location of sites eligible for wind harness. To make the system an actually operative tool, it was conceived as a web-oriented interactive system that the public operators may freely access. The DSS was developed by using the MapServer open-source web-GIS application. Furthermore, the “p.mapper” front-end application developed in JavaScript and PHP/Mapscript was used, which enables a user-friendly interface to MapServer to be performed. System’s wind resource data are estimated by the 2-km resolution application over Tuscany of a meteorological model chain through a 4-year period (January 2004eDecember 2007) with a 1-h timestep. Wind estimations at 75 m were taken into account in order to be addressed to large-scale wind turbines according to the Tuscany Energy Plan objectives of 300 MW installed power derived from wind within 2012. Furthermore, to overcome the problems posed by all groups involved with initially opposing positions in the location for new wind farms (e.g., investors vs. environmentalist groups), the DSS also encompasses a number of layers such as landscape, ecological and archaeological constrained areas. This paper presents the description of the DSS, as well as the application results in terms of maps of wind resource and energy yield once a 2-MW wind turbine has been set as a sample. The developed DSS is currently in use by the Tuscany Regional Authority for planning the regional wind energy strategy. Ó 2010 Elsevier Ltd. All rights reserved.
Keywords: Wind resource assessment Wind energy policy GIS Decision support system Interactive web system Tuscany region
1. Introduction The selection of final location for the construction of wind farms must often be negotiated among the groups involved in the planning process, where the different groups may have conflicting interest: for example, investors and power utilities will look for economically more attractive locations, while other agents, such as environmentalist groups, might consider some of these places as unacceptable from an environmental impact standpoint. This conflict of interests can delay, and even block, the realisation of new wind farms. As a matter of fact, wind farm planning is subject to
* Corresponding author. Tel.: þ39 055 4483027; fax: þ39 055 444083. E-mail address: [email protected] (G. Gualtieri). 0960-1481/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2010.07.005
multi-criteria decision-making processes under uncertainty with conflicting technological, economic, environmental and social aspects. Therefore, the selection of suitable geographical locations should be performed by means of a multi-criteria planning tool attempting to consider jointly most of the economic, technical, environmental and social implications of the planning problem [1]. To pursue such a goal, in the present work an integrated Geographic Information System (GIS) based decision support system (DSS) was developed to help public operators in the preliminary step of selecting consensual locations for new wind farms from initially opposing positions and make these processes faster and more effective, obtaining acceptable solutions for all the groups [1]. Successful completion of this preliminary phase is always accompanied by a local impact study involving a detailed field analysis of any landscape or environmental impacts the
R. Mari et al. / Renewable Energy 36 (2011) 754e763 Table 1 Summary of wind energy application features implemented in WIND-GIS.
755
Table 2 List of layers imported in WIND-GIS as sorted by thematic category.
Item
Parameter
Value
Category
Layers
Wind estimations
Processed time period
01/01/2004-31/12/ 2007 1h 35064 99.50% 75 m a.g.l. WRF 10 Km 525 (25 21)
Wind data: CALMET points
Years 2004e2007 Year 2007 Year 2006 Year 2005 Year 2004 Mean Wind Speed (m/s) Full Load Hours (h/y) Annual Energy Production (MWh/y) Archaeological Constraints Landscape Constraint Parks, reserves and natural areas Municipality Borders Province Borders Cartography (1:250,000)
Prognostic model
Diagnostic model
Wind turbine characteristics
Time step Processed data sample Valid data Height Name Spatial resolution No. processed gridded points Name Spatial resolution No. processed gridded points Name Number of blades Cut-in wind speed Cut-off wind speed Rated wind speed Hub height Rotor diameter Swept area Rated power
CALMET 2 Km 12840 (120 107) Typical 2000 KW 3 4 m/s 25 m/s 15 m/s 78 m 80 m 5027 m2 2000 KW
project is likely to generate [2]. The GIS platform is well suited for locating wind farms thanks to its capability to manage and analyse multidisciplinary data, perform “what if” scenarios which can be used to evaluate the effects of different planning policies, model impacts of proposed and operational sites, and suggest modifications to minimise them [3]. The GIS has been applied to other DSSs,
Wind power maps
Exclusion layers
Background
such as in the evaluation of national wind energy classification [4], regional renewable energy potential [5,6], harness of local renewable resources [7], selection of power technology in rural electrification [8], and selection of potential locations for new wind farms [1,3]. It was also used to analyse wind energy scenario policies, both resulting from different adopted constraint criteria [2,9] and addressed to investigate nation-wide long-term incentive programs [10]. To make the current system an actually operative tool, it was conceived as a web-oriented interactive DSS public operators may freely access. This feature can be reckoned as an added value of the developed system and remarked to point out its originality.
Fig. 1. Web layout of the developed “WIND-GIS” interactive wind resource mapping system over Tuscany region (Italy).
756
R. Mari et al. / Renewable Energy 36 (2011) 754e763
Fig. 2. Sample layout of CALMET grid points interactive query and returned alphanumeric information.
Table 3 Field structure of CALMET points wind data files. Type
Name
Input: site summary
ID X Y Z RO SITE GRID_DIST
Input: wind data
Input: wind turbine
Output: site wind resource
Output: turbineconverted energy
Description
Station ID X-coordinate Y-coordinate Elevation (a.s.l.) Air density Location name Minimum distance from electric grid PERIOD Processed time period PROCESSED_DATA Processed data sample VALID_DATA Valid data sample VALID_DATA_PERC Valid data percentage H_AGL Sensing height (a.g.l.) WIND TURBINE Turbine name/model WT_POWER Rated power WT_H Hub height WT_A Swept area MAX_WS Maximum wind speed MEAN_WS Mean wind speed MEDIAN_WS Median wind speed SCALE_FACTOR Weibull’s scale factor SHAPE_FACTOR Weibull’s shape factor BETZ_ENERGY Betz annual specific energy AF Availability factor CF Capacity factor FLH Full-Load Hours AEP Annual Energy Production PRODUCED_ENERGY Energy produced over the period
Unit e Km Km m Kg/m3 e Km e h h % m e KW m m2 m/s m/s m/s m/s e KWh/ m2 e e h/y MWh/y MWh/ period
At the moment, dynamic and interactive GIS-based wind resource mapping systems (WRMSs) available on the web are numbered, particularly those fully freeware. In particular, one of the most relevant is the “FirstLook” system built-up by 3Tier [11], i.e., a dynamic WRMS providing wind energy resource through a qualitative indicator called “wind rank” which ranges from 0 to 100%. Furthermore, after subscription with fee, numerical annual mean wind speed at hub heights of 20, 50 and 80 m can be derived, too. It is noteworthy at the moment this is the only interactive WRMS covering all over the world. Another remarkable interactive WRMS is the “WindNavigator” system developed by the Associated Weather Services (AWS) Truewind [12], which provides, following purchase of subscription, wind resource maps at hub heights of 30, 60, 80 and 100 m over the countries of Canada, India and United States. Speaking of fully freely available WRMSs, also at a national scale it is to be mentioned the system built-up by the National Renewable Energy Laboratory (NREL) [13], based on the wind energy resource atlas of the United States (1986), as described in [4]. Over whole United States the interactive WRMS performed by the Natural Resources Defense Council (NRDC) in terms of wind power classes (at 10 and 50 m a.g.l.) is available too, also including other renewable energy potential mapping such as solar, biogas and cellulosic biomass [14]. On the other hand, the Canadian territory is supplied with a wind energy atlas performed by the Recherche Prévision Numérique (RPN) [15] at hub heights of 30, 50 and 80 m according to a territory slicing in 65 tiles. A lower scale affects the WRMS developed by the Oklahoma Wind Power Initiative (OWPI), covering the Oklahoma state [16], as well as the one by the Northern Arizona University (NAU), focused
R. Mari et al. / Renewable Energy 36 (2011) 754e763
Fig. 3. Sample layout of wind power maps: mean wind speed (m/s) at 75 m a.g.l. (2004e2007).
Fig. 4. Sample layout of wind power maps: annual energy production (MWh/y) at 75 m a.g.l. (2004e2007).
757
758
R. Mari et al. / Renewable Energy 36 (2011) 754e763
Fig. 5. Sample layout of exclusion and background layers.
Fig. 6. Sample layout of CALMET data query by point.
R. Mari et al. / Renewable Energy 36 (2011) 754e763
759
Fig. 7. Sample layout of CALMET data query by rectangle.
on the Arizona state [17]. Both WRMSs are referred to a hub height of 50 m. The US states of New Jersey and New York are mapped thanks to the Wind Resource Explorer (WRE) developed by AWS Truewind [18], where heights of 30, 50, 70 and 100 m have been considered. As far as Canada is concerned, the Nova Scotia Department of Energy developed the Nova Scotia Wind Atlas (30, 50 and 80 m a.g.l.) [19], whereas the Ontario region is covered by the wind resource atlas by the Canadian Ministry of Natural Resources (10, 30, 50, 80 and 100 m a.g.l.) [20]. In Europe, two websites are noteworthy, i.e., the interactive WRMS built-up by Suisse Eole [21], operative over Switzerland at hub heights of 50, 70 and 100 m, and the one by Action Renewables over Northern Ireland (30, 75 and 100 m on-shore, whilst 50, 75 and 100 m off-shore) [22]. 2. Application features 2.1. Study area Current WRMS has been developed over Tuscany, a region located in central Italy with an area of about 23,000 Km2 and some 3,700,000 inhabitants. Its northern part, where the Firenze metropolitan area is located, also including the cities of Prato and Pistoia, is rather densely populated (311 inhab/Km2 vs. a regional average of 160 inhab/Km2), with about 1,500,000 inhabitants (42% total). Tuscany is characterized by a complex topography and strong land cover variations, surrounded by the Mediterranean Sea westwards and the chain of Apennine Mountains north (up to 2000 m a.s.l.) and eastwards (up to 1600 m). In particular, only 8.4% territory is flat, while 66.5% is made of hills and 25.1% of mountains.
Thereby, all these features suggest it should be a very promising region from a wind energy viewpoint. On the other hand, its unique historical and archaeological heritage, along with landscape one, resulted in a number of constraint criteria to be adopted in terms of particularly wide-ranging constrained areas. 2.2. Wind data estimation In the framework of regional renewable energy policies, starting from 2008 the Tuscany Regional Authority promoted the “WINDGIS” project aimed at assessing the large-scale wind potential of the region. To this end, wind data have been estimated based of the combination of a mesoscale with a microscale meteorological model. As a matter of fact, most of currently web available WRMSs have been developed based on a model combination like that. This is the case, e.g., of WRMSs referred as [12,15,16,18,20]. Thus, over Tuscany wind estimates have been calculated through the application of the coupled Weather Research and Forecasting (WRF) [23] and CALMET [24] models. In particular, the initial 10-km resolution wind fields calculated by the prognostic WRF mesoscale model over 25 21 gridded points have been later downscaled to 2 km by the CALMET diagnostic model, whose computation grid was made by 120 107 points. The use of WRF model for wind resource assessment purposes is currently widely scientifically accepted, e.g., as referred in [25,26]. The 3Tier “FirstLook” WRMS itself [11], on the other hand, is based on a mesoscale 10-year WRF model run. Conversely, CALMET proved to be particularly suitable to work over complex terrain [27], as well as over large areas and long-term simulations [28]. Furthermore, its use to properly downscale prognostic model for
760
R. Mari et al. / Renewable Energy 36 (2011) 754e763
Fig. 8. Sample layout of CALMET data spatial query: search of points falling within the borders of a specific municipality.
wind resource assessment purposes has been proved, e.g., both in combination with MM5 (Fifth-Generation NCAR/Penn State Mesoscale Model) [29], as referred in [30,31], and with WRF [32]. Currently achieved results confirmed that the use of a modelling chain is a valuable tool for generating accurate 3-D meteorological fields for wind resource assessment purposes. In addition, the comparison with observations showed models results were in good agreement with observations and the downscaling process improved the wind speed distribution. The modelling chain has been applied for a 4 calendar-year period (Jan. 2004eDec. 2007) with a 1-h time step. Wind estimations at 75 m were taken into account in order to be addressed to large-scale wind turbines according to the Tuscany Energy Plan, which foresees the installation by 2010 of 300 MW wind power thanks to some 10e15 wind farms. All details of such WRF-CALMET application, which is part of the “WIND-GIS” regional project as well, are broadly described in [33]. A brief summary is reported in Table 1, including the characteristics of the sample wind turbine chosen for energy computation purposes, i.e., a typical 2-MW rated wind generator. 3. System description 3.1. General Wind potential over Tuscany region has been computed according to methods and standards recommended within the Infrastructure for Spatial Information in the European Community (INSPIRE) Directive by the European Community (EC) [34], aimed at the development of a Spatial Data Infrastructure (SDI) to allow a larger spread and utilization of the result by different users and for the subsequent ecological and economic analysis. The current
WRMS has been implemented by using an open-source, free map server implementation tool, i.e., MapServer, developed by the University of Minnesota [35]. The user interface was developed by means of “p.mapper”, an open-source framework that is intended to offer broad functionality and multiple configurations in order to facilitate the setup of MapServer applications based on PHP/MapScript [36]. The WRMS proposed in this work, named “WIND-GIS”, is designed to support the preliminary step of land-planning process aiming at evaluating the possible installation of new wind farms. All needed information was arranged in terms of maps the user may variously overlay in order to make easier the navigation throughout the regional territory. The web working environment of WIND-GIS, as a captured screen snapshot, can be seen in Fig. 1, where the Tuscany study area is also displayed. WIND-GIS may be accessed at: http://geoportale.lamma.rete.toscana.it/eng/windgis. 3.2. Category description WIND-GIS user interface is supplied with fifteen layers, which are grouped according to four thematic categories (Fig. 1), namely: wind data, wind power maps, exclusion layers, and background. In Table 2 the list and details of imported layers as sorted by thematic category are presented. 3.3. Wind data: CALMET points The first category, i.e., “Wind data: CALMET points”, includes point vector layers containing wind data resulting from the aforementioned application of the WRF-CALMET modelling chain (see x 2.2). Wind data point information is managed by the
R. Mari et al. / Renewable Energy 36 (2011) 754e763
761
Fig. 9. Sample layout of CALMET data spatial query: search of points located at a given distance from the existing electric grid.
PostgreSQL relational database by using the PostGIS spatial extension. It consists of a spatial layer containing X and Y coordinates of CALMET grid points joined with the alphanumeric information of CALMET elaboration referred to single years 2004, 2005, 2006, 2007, as well as overall 2004e2007 period. The alphanumeric information of the layers is returned via a dropand drags down menu after the user selects the tool-tip button the mouse on a grid point, as shown and highlighted in Fig. 2. In Table 3 the field structure of CALMET points wind data files is described. Along with (input) site summary, wind data features and turbine characteristics, main (output) wind energy indicators are calculated, such as, e.g., mean wind speed, Weibull’s scale and shape factors, as well as Betz annual specific energy [37]. In addition, turbine-related availability and capacity factors have been calculated, as well as full-load hours (FLH) and annual energy production (AEP) [37]. 3.4. Wind power maps The second category, i.e. “Wind power maps”, consists of three thematic (raster) layers, namely mean wind speed, FLH and AEP. The maps, which are static and pre-calculated, are referred to a hub height of 75 m after processing the full 2004e2007 period. FLH and AEP maps result from the application of the aforementioned 2-MW wind turbine, whose characteristics are reported in Table 1. A proper spatialization of CALMET grid points was performed. As they feature a 2-km spatial resolution, a total of 12,840 grid points was employed for spatialization. As an example, Fig. 3 shows the map of mean wind speed (m/s), whilst Fig. 4 depicts the pattern of AEP (MWh/y).
3.5. Exclusion layers and background The third category, “Exclusion layers”, groups (point and polygon) layers representing constrained areas or areas where a particular attention is required to the planner or decision-maker. These layers are outside the facilities of the current system and reside on the Tuscany Regional Authority server. They are accessible through the “WMS” Web Map Service, which the Tuscany Regional Authority itself made freely available for the entire thematic and technical cartography [38]. The “WMS” technology allows to always obtain the most updated data directly from the producer-maintainer, and thus the user does not have any data stored on his client machine. It is to be noticed that the integration of exclusion layers on worldwide available WRMSs is quite rare, at the moment. As a matter of fact, this is only the case of the NREL WRMS over US [13], the AWS Truewind one over the states of New Jersey and New York [18], the Ontario system [20], and the WRMS over Switzerland [21], which in particular is the only one in Europe. Finally, the system comprises the “Background” category, including (polygon and raster) layers such as municipality and province borders, as well as a 1:250,000 cartography. Also, these layers are accessible through the “WMS” Web Map Service. In Fig. 5 a specific focus is made as an example on exclusion and background layers.
4. Using the system WIND-GIS allows three types of queries of the alphanumeric information linked to the CALMET data points (Table 3) to be made, i.e., by point, rectangle as well as spatial query.
762
R. Mari et al. / Renewable Energy 36 (2011) 754e763
4.1. CALMET data: query by point and clicking the When selecting the query by point button desired CALMET point object of the layer (and thus period) of interest, a window such as the one shown in Fig. 6 will open, where in particular the 2004e2007 full period has been chosen. The system also allows to export the results of a query or a search into a Microsoft (MS) Excel spreadsheet. 4.2. CALMET data: query by rectangle it is possible to draw Using the query by rectangle button a rectangle on the map enabling all encompassed CALMET points to be selected and related window to open. The graphical result of this action should appear as displayed in Fig. 7. Again, the result of this query may be exported into an MS Excel spreadsheet. 4.3. CALMET data: spatial query The system allows the search of the alphanumeric information linked to the specific layer of interest to be performed based on a number of default search criteria. This is achieved through a dropdown menu located above the button bar on top left hand side of the working environment. The following default search criteria have been set:
Municipality borders; Province borders; Distance from electric grid; Years 2004e2007; Year 2007; Year 2006; Year 2005; Year 2004.
Multiple search options may be performed in a row. For example, after chosen the “Years 2004e2007” wind data layer and run the search by municipality borders, the system selects and highlights all CALMET points falling within the borders of the selected municipality, as shown in the sample of Fig. 8. This search may be furtherly refined seeking those municipality points also located at a given distance from the existing (not shown in figure) electric grid, as displayed in the sample of Fig. 9. Accordingly, the tabulated result of the point selection by spatial query is also returned. 5. Conclusions and perspectives In the present work the development of an integrated GISbased DSS is described, which is designed to help public operators in the preliminarily location of sites eligible for wind harness. The DSS has been developed in the framework of the “WIND-GIS” project, promoted by the Tuscany Regional Authority with the aim of assessing the large-scale wind potential of Tuscany region, Italy. The system was conceived as a dynamic and interactive WRMS freely accessible on the web. It was developed by using the MapServer open-source web-GIS application, along with the “p.mapper” front-end application developed in JavaScript and PHP/Mapscript, which enables a user-friendly interface to MapServer to be performed. Systems’s user interface is supplied with fifteen layers, which are grouped according to four thematic categories, i.e., wind resource data, wind power maps, exclusion layers, and background. Wind resource data, referred to a hub height of 75 m, result from the 2-km resolution application over Tuscany of the WRF-CALMET
model chain through a 4 calendar-year period (January 2004 to December 2007) with a 1-h time step. This combination of a mesoscale with a microscale model is in agreement with the methodology approached by most of currently web available WRMSs. On the other hand, the use of WRF and CALMET models for wind resource assessment purposes is currently widely scientifically accepted. Also, currently achieved results confirmed that the use of a modelling chain is a valuable tool for generating accurate 3D meteorological fields. A sample 2-MW wind turbine has been set for energy computation purposes. As a matter of fact, the whole Tuscany territory is covered by a 2-km spaced point grid which all alphanumeric information is associated with. Wind power maps, resulting from processing the full 2004e2007 period and spatializing a total of 12,840 points, include mean wind speed, FLH and AEP. The developed WRMS proved to be a robust and powerful tool to support the preliminary step of land-planning process in the evaluation of possible installation of new wind farms. As a matter of fact, three types of queries are allowed on any wind data grid point, i.e., by point, rectangle and spatial query, as well as the export of related results into an MS Excel spreadsheet. Another system’s key-point is given by all imported exclusion layers, i.e., archaeological and landscape constrained areas, as well as parks, reserves and natural areas. These layers, which are outside the facilities of the current WRMS, reside on the Tuscany Regional Authority server and may be accessed through the “WMS” Web Map Service made available by the Tuscany Regional Authority itself. Summarizing, as compared to the state-of-the art, the “WINDGIS” system is worth noticing for a number of reasons, namely: 1. it is an interactive GIS-based, web freely accessible WRMS in Italy, as well as one of the really few in Europe; 2. in Europe, the Swiss WRMS is supplied by all necessary exclusion layers such as “WIND-GIS”; worldwide, WRMSs including this feature are numbered; 3. in Europe, its wind data resolution (i.e., 2 km) is the highest such as the one of the WRMS over Switzerland, which was built-up through a statistical interpolation of station measurements, actually. A number of system’s possible enhancements could be performed in the future, both in terms of updates (e.g., changes in exclusion layers) and upgrades (e.g., extension of CALMETcomputed wind estimations to a longer period, as well as increase to a finer horizontal resolution). In addition, similarly to other WRMSs available on the web, the assessment of regional wind potential addressed to small wind turbines could be carried out as well, e.g., by considering a hub height of 20 m. Furthermore, the proposed WRMS has been designed to be easily applied to other regions, particularly if considering the Italian territory to be fully covered by WRF-calculated wind estimations currently stored in a suitable meteorological archive. On the other hand, besides monitoring the evolving wind energy production scenario over Tuscany, “WIND-GIS” could also be integrated with information derived from other renewable energy sources. For example, radiation maps could also be derived from WRF model data to help operators in the location of photovoltaic plants.
Acknowledgements This work was supported by the Tuscany Regional Authority in the framework of the project: “WIND-GIS: a project to develop a web service to assess the wind potential of Tuscany region”.
R. Mari et al. / Renewable Energy 36 (2011) 754e763
References [1] Ramirez-Rosado IJ, Garcia-Garrido E, Fernandez-Jimenez A, ZorzanoSantamaria PJ, Monteiro C, Miranda V. Promotion of new wind farms based on a decision support system. Renew Energy 2008;33:558e66. [2] Lejeune P, Feltz C. Development of a decision support system for setting up a wind energy policy across the Walloon region (southern Belgium). Renew Energy 2008;33:2416e22. [3] Baban SMJ, Parry T. Developying and applying a GIS-assisted approach to locating wind farms in the UK. Renew Energy 2001;24:59e71. [4] Heimiller DM, Haymes SR. Geographic Information Systems in Support of Wind Energy Activities at NREL. Preprints of 39th AIAA Aerospace Sciences Meeting. Reno, Nevada, USA; 2001. [5] Voivontas D, Assimacopoulos D, Mourelatos A, Corominas J. Evaluation of renewable energy potential using a GIS decision support system. Renew Energy 1998;13:333e44. [6] Hillring B, Krieg R. Wind energy potential in Southern Sweden e example of planning methodology. Renew Energy 1998;13(4):471e9. [7] Yue C, Yang CG. Decision support system for exploiting local renewable energy sources: a case study of the Chigu area of southwestern Taiwan. Energy Policy 2007;35:383e94. [8] Amador J, Domınguez J. Application of geographical information systems to rural electrification with renewable energy sources. Renew Energy 2005;30:1897e912. [9] Krewitt W, Nitsch J. The potential for electricity generation from on-shore wind energy under constraints of nature conservation: a case study for two regions in Germany. Renew Energy 2003;28:1645e55. [10] Dutra R, Szklo A. Assessing long-term incentive programs for implementing wind power in Brazil using GIS rule-based methods. Renew Energy 2008;33:2507e15. [11] 3Tier. First look world wind map. Available at:; 2010. [12] Associated Weather Services (AWS) Truewind. Wind navigator e Canada, India and United States. Available at: ; 2010. [13] National Renewable Energy Laboratory (NREL). United States annual wind resource potential. Available at: ; 2010. [14] Natural Resources Defense Council (NRDC). Renewable energy for America. Available at: ; 2010. [15] Recherche Prévision Numérique (RPN). The Canadian wind energy atlas. Available at: . [16] Oklahoma Wind Power Initiative (OWPI). Oklahoma wind resource map. Available at: ; 2010. [17] Northern Arizona University (NAU). Arizona interactive wind resource map. Available at: ; 2010. [18] Associated Weather Services (AWS) Truewind. Wind resource explorer e States of New Jersey, and New York. Available at: ; 2010. [19] Nova Scotia Department of Energy. Nova Scotia wind atlas. Available at: ; 2010.
763
[20] Canadian Ministry of Natural Resources. The Ontario wind resource atlas. Available at:; 2010. [21] Suisse Eole. The Swiss wind power data website. Available at: ; 2010. [22] Action Renewables. Wind speed mapping of Northern Ireland. Available at: ; 2010. [23] Skamarock WC, Klemp JB, Dudhia J, Gill DO, Barker DM, Wang W, et al. A description of the Advanced Research WRF version 2. Boulder, Colorado, USA: National Center for Atmospheric Research; 2005. NCAR Technical Note. [24] Scire JS, Robe FR, Fermau ME, Yamartino RJ. A User’s Guide for the CALMET meteorological model (version 5.0). Concord, MA, USA: Earth Tech Inc; 1999. [25] Berge E, Bredesen RE, Mollestad K. Combining WAsP with the WRF meso-scale model. Evaluation of wind resource assessment for three Norwegian wind farm areas. In: Proceedings of European Wind Energy Conference & Exhibition (EWEC). Milano, Italy; 2007. [26] Storm B, Dudhia J, Basu S, Swift A, Giammarco I. Evaluation of the weather research and forecasting model on forecasting low-level jets: Implications for wind energy. Wind Energy 2009;12:81e90. [27] Scire JS, Robe FR. Fine-Scale Application of the CALMET Meteorological Model to a Complex Terrain Site. In: Proceedings of AWMA’s 90th Annual Meeting and Exhibition. Toronto, Canada; 1997. [28] Faghani D, Fozza R, Aït-Driss B. Synchronization of Production Time Series of Geographically Dispersed Wind Farms. In: Proceedings of 7th World Wind Energy Conference 2008 (WWEC2008). Kingston, Canada; 2008. [29] GrellGA Dudhia J, Stauffer DR. A description of the fifth generation Penn StateNCAR mesoscale model (MM5). Boulder, CO, USA: National Center for Atmospheric Research; 1994. NCAR/TN-398þSTR. [30] Yim SHL, Fung JCH, Lau AKH, Kot SC. Developing a high-resolution wind map for a complex terrain with a coupled MM5/CALMET system. J Geophys Res 2007;112:D05106. [31] Prieto M, Navarro J, Copeland J. A comparison of wind field simulations through a coupling of MM5/CALMET in a complex terrain. In: Proceedings of European Wind Energy Conference & Exhibition (EWEC). Milano, Italy; 2007. [32] Radonjic Z, Chambers D, Telenta B, Music S, Janjic Z. Coupling NMM mesoscale weather forecast model with CALMET for wind energy applications. Geophysical Research Astracts 2009;11:1152. [33] Busillo C, Calastrini F, Gozzini B, Gualtieri G, Mari R. WIND-GIS: Progetto per la realizzazione di un servizio web per la valutazione del potenziale eolico della regione Toscana. LaMMA Consortium; 2008 [in Italian]. [34] European Commission. Directive 2007/2/EC of the European Parliament and of the council of 14 march 2007 establishing an infrastructure for spatial information in the European Community (INSPIRE); 2007. [35] University of Minnesota. MapServer documentation. Available at: ; 2010. [36] Burger A. p.mapper e A MapServer PHP/MapScript Framework. Available at: . 2010. [37] Burton T, Scarpe D, Jenkins N, Bossanyi E. Wind energy handbook. Chichester, UK: John Wiley & Sons; 2001. [38] Tuscany Region. GEOSCOPIO_WMS. Available at: ; 2010 [In Italian].
Contents lists available at ScienceDirect
Renewable Energy journal homepage: www.elsevier.com/locate/renene
A GIS-based interactive web decision support system for planning wind farms in Tuscany (Italy) Riccardo Mari a, Lorenzo Bottai a, Caterina Busillo a, Francesca Calastrini b, Bernardo Gozzini b, Giovanni Gualtieri b, * a b
Laboratory of Monitoring and Environmental Modelling for the sustainable development (LAMMA), Via Madonna del Piano 10, 50019 Sesto Fiorentino, Italy Research National Council-Institute for biometeorology (CNR-IBIMET), Via Caproni 8, 50145 Firenze, Italy
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 April 2010 Accepted 10 July 2010 Available online 2 August 2010
In the framework of regional renewable energy policies, starting from 2008 the Tuscany Regional Authority promoted the “WIND-GIS” project aimed at assessing the large-scale wind potential of Tuscany region, Italy. This goal was achieved by developing an integrated Geographic Information System (GIS) based decision support system (DSS), compliant with Directive 2007/2/EC of European Commission (EC), which was designed to help public operators in the preliminary location of sites eligible for wind harness. To make the system an actually operative tool, it was conceived as a web-oriented interactive system that the public operators may freely access. The DSS was developed by using the MapServer open-source web-GIS application. Furthermore, the “p.mapper” front-end application developed in JavaScript and PHP/Mapscript was used, which enables a user-friendly interface to MapServer to be performed. System’s wind resource data are estimated by the 2-km resolution application over Tuscany of a meteorological model chain through a 4-year period (January 2004eDecember 2007) with a 1-h timestep. Wind estimations at 75 m were taken into account in order to be addressed to large-scale wind turbines according to the Tuscany Energy Plan objectives of 300 MW installed power derived from wind within 2012. Furthermore, to overcome the problems posed by all groups involved with initially opposing positions in the location for new wind farms (e.g., investors vs. environmentalist groups), the DSS also encompasses a number of layers such as landscape, ecological and archaeological constrained areas. This paper presents the description of the DSS, as well as the application results in terms of maps of wind resource and energy yield once a 2-MW wind turbine has been set as a sample. The developed DSS is currently in use by the Tuscany Regional Authority for planning the regional wind energy strategy. Ó 2010 Elsevier Ltd. All rights reserved.
Keywords: Wind resource assessment Wind energy policy GIS Decision support system Interactive web system Tuscany region
1. Introduction The selection of final location for the construction of wind farms must often be negotiated among the groups involved in the planning process, where the different groups may have conflicting interest: for example, investors and power utilities will look for economically more attractive locations, while other agents, such as environmentalist groups, might consider some of these places as unacceptable from an environmental impact standpoint. This conflict of interests can delay, and even block, the realisation of new wind farms. As a matter of fact, wind farm planning is subject to
* Corresponding author. Tel.: þ39 055 4483027; fax: þ39 055 444083. E-mail address: [email protected] (G. Gualtieri). 0960-1481/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2010.07.005
multi-criteria decision-making processes under uncertainty with conflicting technological, economic, environmental and social aspects. Therefore, the selection of suitable geographical locations should be performed by means of a multi-criteria planning tool attempting to consider jointly most of the economic, technical, environmental and social implications of the planning problem [1]. To pursue such a goal, in the present work an integrated Geographic Information System (GIS) based decision support system (DSS) was developed to help public operators in the preliminary step of selecting consensual locations for new wind farms from initially opposing positions and make these processes faster and more effective, obtaining acceptable solutions for all the groups [1]. Successful completion of this preliminary phase is always accompanied by a local impact study involving a detailed field analysis of any landscape or environmental impacts the
R. Mari et al. / Renewable Energy 36 (2011) 754e763 Table 1 Summary of wind energy application features implemented in WIND-GIS.
755
Table 2 List of layers imported in WIND-GIS as sorted by thematic category.
Item
Parameter
Value
Category
Layers
Wind estimations
Processed time period
01/01/2004-31/12/ 2007 1h 35064 99.50% 75 m a.g.l. WRF 10 Km 525 (25 21)
Wind data: CALMET points
Years 2004e2007 Year 2007 Year 2006 Year 2005 Year 2004 Mean Wind Speed (m/s) Full Load Hours (h/y) Annual Energy Production (MWh/y) Archaeological Constraints Landscape Constraint Parks, reserves and natural areas Municipality Borders Province Borders Cartography (1:250,000)
Prognostic model
Diagnostic model
Wind turbine characteristics
Time step Processed data sample Valid data Height Name Spatial resolution No. processed gridded points Name Spatial resolution No. processed gridded points Name Number of blades Cut-in wind speed Cut-off wind speed Rated wind speed Hub height Rotor diameter Swept area Rated power
CALMET 2 Km 12840 (120 107) Typical 2000 KW 3 4 m/s 25 m/s 15 m/s 78 m 80 m 5027 m2 2000 KW
project is likely to generate [2]. The GIS platform is well suited for locating wind farms thanks to its capability to manage and analyse multidisciplinary data, perform “what if” scenarios which can be used to evaluate the effects of different planning policies, model impacts of proposed and operational sites, and suggest modifications to minimise them [3]. The GIS has been applied to other DSSs,
Wind power maps
Exclusion layers
Background
such as in the evaluation of national wind energy classification [4], regional renewable energy potential [5,6], harness of local renewable resources [7], selection of power technology in rural electrification [8], and selection of potential locations for new wind farms [1,3]. It was also used to analyse wind energy scenario policies, both resulting from different adopted constraint criteria [2,9] and addressed to investigate nation-wide long-term incentive programs [10]. To make the current system an actually operative tool, it was conceived as a web-oriented interactive DSS public operators may freely access. This feature can be reckoned as an added value of the developed system and remarked to point out its originality.
Fig. 1. Web layout of the developed “WIND-GIS” interactive wind resource mapping system over Tuscany region (Italy).
756
R. Mari et al. / Renewable Energy 36 (2011) 754e763
Fig. 2. Sample layout of CALMET grid points interactive query and returned alphanumeric information.
Table 3 Field structure of CALMET points wind data files. Type
Name
Input: site summary
ID X Y Z RO SITE GRID_DIST
Input: wind data
Input: wind turbine
Output: site wind resource
Output: turbineconverted energy
Description
Station ID X-coordinate Y-coordinate Elevation (a.s.l.) Air density Location name Minimum distance from electric grid PERIOD Processed time period PROCESSED_DATA Processed data sample VALID_DATA Valid data sample VALID_DATA_PERC Valid data percentage H_AGL Sensing height (a.g.l.) WIND TURBINE Turbine name/model WT_POWER Rated power WT_H Hub height WT_A Swept area MAX_WS Maximum wind speed MEAN_WS Mean wind speed MEDIAN_WS Median wind speed SCALE_FACTOR Weibull’s scale factor SHAPE_FACTOR Weibull’s shape factor BETZ_ENERGY Betz annual specific energy AF Availability factor CF Capacity factor FLH Full-Load Hours AEP Annual Energy Production PRODUCED_ENERGY Energy produced over the period
Unit e Km Km m Kg/m3 e Km e h h % m e KW m m2 m/s m/s m/s m/s e KWh/ m2 e e h/y MWh/y MWh/ period
At the moment, dynamic and interactive GIS-based wind resource mapping systems (WRMSs) available on the web are numbered, particularly those fully freeware. In particular, one of the most relevant is the “FirstLook” system built-up by 3Tier [11], i.e., a dynamic WRMS providing wind energy resource through a qualitative indicator called “wind rank” which ranges from 0 to 100%. Furthermore, after subscription with fee, numerical annual mean wind speed at hub heights of 20, 50 and 80 m can be derived, too. It is noteworthy at the moment this is the only interactive WRMS covering all over the world. Another remarkable interactive WRMS is the “WindNavigator” system developed by the Associated Weather Services (AWS) Truewind [12], which provides, following purchase of subscription, wind resource maps at hub heights of 30, 60, 80 and 100 m over the countries of Canada, India and United States. Speaking of fully freely available WRMSs, also at a national scale it is to be mentioned the system built-up by the National Renewable Energy Laboratory (NREL) [13], based on the wind energy resource atlas of the United States (1986), as described in [4]. Over whole United States the interactive WRMS performed by the Natural Resources Defense Council (NRDC) in terms of wind power classes (at 10 and 50 m a.g.l.) is available too, also including other renewable energy potential mapping such as solar, biogas and cellulosic biomass [14]. On the other hand, the Canadian territory is supplied with a wind energy atlas performed by the Recherche Prévision Numérique (RPN) [15] at hub heights of 30, 50 and 80 m according to a territory slicing in 65 tiles. A lower scale affects the WRMS developed by the Oklahoma Wind Power Initiative (OWPI), covering the Oklahoma state [16], as well as the one by the Northern Arizona University (NAU), focused
R. Mari et al. / Renewable Energy 36 (2011) 754e763
Fig. 3. Sample layout of wind power maps: mean wind speed (m/s) at 75 m a.g.l. (2004e2007).
Fig. 4. Sample layout of wind power maps: annual energy production (MWh/y) at 75 m a.g.l. (2004e2007).
757
758
R. Mari et al. / Renewable Energy 36 (2011) 754e763
Fig. 5. Sample layout of exclusion and background layers.
Fig. 6. Sample layout of CALMET data query by point.
R. Mari et al. / Renewable Energy 36 (2011) 754e763
759
Fig. 7. Sample layout of CALMET data query by rectangle.
on the Arizona state [17]. Both WRMSs are referred to a hub height of 50 m. The US states of New Jersey and New York are mapped thanks to the Wind Resource Explorer (WRE) developed by AWS Truewind [18], where heights of 30, 50, 70 and 100 m have been considered. As far as Canada is concerned, the Nova Scotia Department of Energy developed the Nova Scotia Wind Atlas (30, 50 and 80 m a.g.l.) [19], whereas the Ontario region is covered by the wind resource atlas by the Canadian Ministry of Natural Resources (10, 30, 50, 80 and 100 m a.g.l.) [20]. In Europe, two websites are noteworthy, i.e., the interactive WRMS built-up by Suisse Eole [21], operative over Switzerland at hub heights of 50, 70 and 100 m, and the one by Action Renewables over Northern Ireland (30, 75 and 100 m on-shore, whilst 50, 75 and 100 m off-shore) [22]. 2. Application features 2.1. Study area Current WRMS has been developed over Tuscany, a region located in central Italy with an area of about 23,000 Km2 and some 3,700,000 inhabitants. Its northern part, where the Firenze metropolitan area is located, also including the cities of Prato and Pistoia, is rather densely populated (311 inhab/Km2 vs. a regional average of 160 inhab/Km2), with about 1,500,000 inhabitants (42% total). Tuscany is characterized by a complex topography and strong land cover variations, surrounded by the Mediterranean Sea westwards and the chain of Apennine Mountains north (up to 2000 m a.s.l.) and eastwards (up to 1600 m). In particular, only 8.4% territory is flat, while 66.5% is made of hills and 25.1% of mountains.
Thereby, all these features suggest it should be a very promising region from a wind energy viewpoint. On the other hand, its unique historical and archaeological heritage, along with landscape one, resulted in a number of constraint criteria to be adopted in terms of particularly wide-ranging constrained areas. 2.2. Wind data estimation In the framework of regional renewable energy policies, starting from 2008 the Tuscany Regional Authority promoted the “WINDGIS” project aimed at assessing the large-scale wind potential of the region. To this end, wind data have been estimated based of the combination of a mesoscale with a microscale meteorological model. As a matter of fact, most of currently web available WRMSs have been developed based on a model combination like that. This is the case, e.g., of WRMSs referred as [12,15,16,18,20]. Thus, over Tuscany wind estimates have been calculated through the application of the coupled Weather Research and Forecasting (WRF) [23] and CALMET [24] models. In particular, the initial 10-km resolution wind fields calculated by the prognostic WRF mesoscale model over 25 21 gridded points have been later downscaled to 2 km by the CALMET diagnostic model, whose computation grid was made by 120 107 points. The use of WRF model for wind resource assessment purposes is currently widely scientifically accepted, e.g., as referred in [25,26]. The 3Tier “FirstLook” WRMS itself [11], on the other hand, is based on a mesoscale 10-year WRF model run. Conversely, CALMET proved to be particularly suitable to work over complex terrain [27], as well as over large areas and long-term simulations [28]. Furthermore, its use to properly downscale prognostic model for
760
R. Mari et al. / Renewable Energy 36 (2011) 754e763
Fig. 8. Sample layout of CALMET data spatial query: search of points falling within the borders of a specific municipality.
wind resource assessment purposes has been proved, e.g., both in combination with MM5 (Fifth-Generation NCAR/Penn State Mesoscale Model) [29], as referred in [30,31], and with WRF [32]. Currently achieved results confirmed that the use of a modelling chain is a valuable tool for generating accurate 3-D meteorological fields for wind resource assessment purposes. In addition, the comparison with observations showed models results were in good agreement with observations and the downscaling process improved the wind speed distribution. The modelling chain has been applied for a 4 calendar-year period (Jan. 2004eDec. 2007) with a 1-h time step. Wind estimations at 75 m were taken into account in order to be addressed to large-scale wind turbines according to the Tuscany Energy Plan, which foresees the installation by 2010 of 300 MW wind power thanks to some 10e15 wind farms. All details of such WRF-CALMET application, which is part of the “WIND-GIS” regional project as well, are broadly described in [33]. A brief summary is reported in Table 1, including the characteristics of the sample wind turbine chosen for energy computation purposes, i.e., a typical 2-MW rated wind generator. 3. System description 3.1. General Wind potential over Tuscany region has been computed according to methods and standards recommended within the Infrastructure for Spatial Information in the European Community (INSPIRE) Directive by the European Community (EC) [34], aimed at the development of a Spatial Data Infrastructure (SDI) to allow a larger spread and utilization of the result by different users and for the subsequent ecological and economic analysis. The current
WRMS has been implemented by using an open-source, free map server implementation tool, i.e., MapServer, developed by the University of Minnesota [35]. The user interface was developed by means of “p.mapper”, an open-source framework that is intended to offer broad functionality and multiple configurations in order to facilitate the setup of MapServer applications based on PHP/MapScript [36]. The WRMS proposed in this work, named “WIND-GIS”, is designed to support the preliminary step of land-planning process aiming at evaluating the possible installation of new wind farms. All needed information was arranged in terms of maps the user may variously overlay in order to make easier the navigation throughout the regional territory. The web working environment of WIND-GIS, as a captured screen snapshot, can be seen in Fig. 1, where the Tuscany study area is also displayed. WIND-GIS may be accessed at: http://geoportale.lamma.rete.toscana.it/eng/windgis. 3.2. Category description WIND-GIS user interface is supplied with fifteen layers, which are grouped according to four thematic categories (Fig. 1), namely: wind data, wind power maps, exclusion layers, and background. In Table 2 the list and details of imported layers as sorted by thematic category are presented. 3.3. Wind data: CALMET points The first category, i.e., “Wind data: CALMET points”, includes point vector layers containing wind data resulting from the aforementioned application of the WRF-CALMET modelling chain (see x 2.2). Wind data point information is managed by the
R. Mari et al. / Renewable Energy 36 (2011) 754e763
761
Fig. 9. Sample layout of CALMET data spatial query: search of points located at a given distance from the existing electric grid.
PostgreSQL relational database by using the PostGIS spatial extension. It consists of a spatial layer containing X and Y coordinates of CALMET grid points joined with the alphanumeric information of CALMET elaboration referred to single years 2004, 2005, 2006, 2007, as well as overall 2004e2007 period. The alphanumeric information of the layers is returned via a dropand drags down menu after the user selects the tool-tip button the mouse on a grid point, as shown and highlighted in Fig. 2. In Table 3 the field structure of CALMET points wind data files is described. Along with (input) site summary, wind data features and turbine characteristics, main (output) wind energy indicators are calculated, such as, e.g., mean wind speed, Weibull’s scale and shape factors, as well as Betz annual specific energy [37]. In addition, turbine-related availability and capacity factors have been calculated, as well as full-load hours (FLH) and annual energy production (AEP) [37]. 3.4. Wind power maps The second category, i.e. “Wind power maps”, consists of three thematic (raster) layers, namely mean wind speed, FLH and AEP. The maps, which are static and pre-calculated, are referred to a hub height of 75 m after processing the full 2004e2007 period. FLH and AEP maps result from the application of the aforementioned 2-MW wind turbine, whose characteristics are reported in Table 1. A proper spatialization of CALMET grid points was performed. As they feature a 2-km spatial resolution, a total of 12,840 grid points was employed for spatialization. As an example, Fig. 3 shows the map of mean wind speed (m/s), whilst Fig. 4 depicts the pattern of AEP (MWh/y).
3.5. Exclusion layers and background The third category, “Exclusion layers”, groups (point and polygon) layers representing constrained areas or areas where a particular attention is required to the planner or decision-maker. These layers are outside the facilities of the current system and reside on the Tuscany Regional Authority server. They are accessible through the “WMS” Web Map Service, which the Tuscany Regional Authority itself made freely available for the entire thematic and technical cartography [38]. The “WMS” technology allows to always obtain the most updated data directly from the producer-maintainer, and thus the user does not have any data stored on his client machine. It is to be noticed that the integration of exclusion layers on worldwide available WRMSs is quite rare, at the moment. As a matter of fact, this is only the case of the NREL WRMS over US [13], the AWS Truewind one over the states of New Jersey and New York [18], the Ontario system [20], and the WRMS over Switzerland [21], which in particular is the only one in Europe. Finally, the system comprises the “Background” category, including (polygon and raster) layers such as municipality and province borders, as well as a 1:250,000 cartography. Also, these layers are accessible through the “WMS” Web Map Service. In Fig. 5 a specific focus is made as an example on exclusion and background layers.
4. Using the system WIND-GIS allows three types of queries of the alphanumeric information linked to the CALMET data points (Table 3) to be made, i.e., by point, rectangle as well as spatial query.
762
R. Mari et al. / Renewable Energy 36 (2011) 754e763
4.1. CALMET data: query by point and clicking the When selecting the query by point button desired CALMET point object of the layer (and thus period) of interest, a window such as the one shown in Fig. 6 will open, where in particular the 2004e2007 full period has been chosen. The system also allows to export the results of a query or a search into a Microsoft (MS) Excel spreadsheet. 4.2. CALMET data: query by rectangle it is possible to draw Using the query by rectangle button a rectangle on the map enabling all encompassed CALMET points to be selected and related window to open. The graphical result of this action should appear as displayed in Fig. 7. Again, the result of this query may be exported into an MS Excel spreadsheet. 4.3. CALMET data: spatial query The system allows the search of the alphanumeric information linked to the specific layer of interest to be performed based on a number of default search criteria. This is achieved through a dropdown menu located above the button bar on top left hand side of the working environment. The following default search criteria have been set:
Municipality borders; Province borders; Distance from electric grid; Years 2004e2007; Year 2007; Year 2006; Year 2005; Year 2004.
Multiple search options may be performed in a row. For example, after chosen the “Years 2004e2007” wind data layer and run the search by municipality borders, the system selects and highlights all CALMET points falling within the borders of the selected municipality, as shown in the sample of Fig. 8. This search may be furtherly refined seeking those municipality points also located at a given distance from the existing (not shown in figure) electric grid, as displayed in the sample of Fig. 9. Accordingly, the tabulated result of the point selection by spatial query is also returned. 5. Conclusions and perspectives In the present work the development of an integrated GISbased DSS is described, which is designed to help public operators in the preliminarily location of sites eligible for wind harness. The DSS has been developed in the framework of the “WIND-GIS” project, promoted by the Tuscany Regional Authority with the aim of assessing the large-scale wind potential of Tuscany region, Italy. The system was conceived as a dynamic and interactive WRMS freely accessible on the web. It was developed by using the MapServer open-source web-GIS application, along with the “p.mapper” front-end application developed in JavaScript and PHP/Mapscript, which enables a user-friendly interface to MapServer to be performed. Systems’s user interface is supplied with fifteen layers, which are grouped according to four thematic categories, i.e., wind resource data, wind power maps, exclusion layers, and background. Wind resource data, referred to a hub height of 75 m, result from the 2-km resolution application over Tuscany of the WRF-CALMET
model chain through a 4 calendar-year period (January 2004 to December 2007) with a 1-h time step. This combination of a mesoscale with a microscale model is in agreement with the methodology approached by most of currently web available WRMSs. On the other hand, the use of WRF and CALMET models for wind resource assessment purposes is currently widely scientifically accepted. Also, currently achieved results confirmed that the use of a modelling chain is a valuable tool for generating accurate 3D meteorological fields. A sample 2-MW wind turbine has been set for energy computation purposes. As a matter of fact, the whole Tuscany territory is covered by a 2-km spaced point grid which all alphanumeric information is associated with. Wind power maps, resulting from processing the full 2004e2007 period and spatializing a total of 12,840 points, include mean wind speed, FLH and AEP. The developed WRMS proved to be a robust and powerful tool to support the preliminary step of land-planning process in the evaluation of possible installation of new wind farms. As a matter of fact, three types of queries are allowed on any wind data grid point, i.e., by point, rectangle and spatial query, as well as the export of related results into an MS Excel spreadsheet. Another system’s key-point is given by all imported exclusion layers, i.e., archaeological and landscape constrained areas, as well as parks, reserves and natural areas. These layers, which are outside the facilities of the current WRMS, reside on the Tuscany Regional Authority server and may be accessed through the “WMS” Web Map Service made available by the Tuscany Regional Authority itself. Summarizing, as compared to the state-of-the art, the “WINDGIS” system is worth noticing for a number of reasons, namely: 1. it is an interactive GIS-based, web freely accessible WRMS in Italy, as well as one of the really few in Europe; 2. in Europe, the Swiss WRMS is supplied by all necessary exclusion layers such as “WIND-GIS”; worldwide, WRMSs including this feature are numbered; 3. in Europe, its wind data resolution (i.e., 2 km) is the highest such as the one of the WRMS over Switzerland, which was built-up through a statistical interpolation of station measurements, actually. A number of system’s possible enhancements could be performed in the future, both in terms of updates (e.g., changes in exclusion layers) and upgrades (e.g., extension of CALMETcomputed wind estimations to a longer period, as well as increase to a finer horizontal resolution). In addition, similarly to other WRMSs available on the web, the assessment of regional wind potential addressed to small wind turbines could be carried out as well, e.g., by considering a hub height of 20 m. Furthermore, the proposed WRMS has been designed to be easily applied to other regions, particularly if considering the Italian territory to be fully covered by WRF-calculated wind estimations currently stored in a suitable meteorological archive. On the other hand, besides monitoring the evolving wind energy production scenario over Tuscany, “WIND-GIS” could also be integrated with information derived from other renewable energy sources. For example, radiation maps could also be derived from WRF model data to help operators in the location of photovoltaic plants.
Acknowledgements This work was supported by the Tuscany Regional Authority in the framework of the project: “WIND-GIS: a project to develop a web service to assess the wind potential of Tuscany region”.
R. Mari et al. / Renewable Energy 36 (2011) 754e763
References [1] Ramirez-Rosado IJ, Garcia-Garrido E, Fernandez-Jimenez A, ZorzanoSantamaria PJ, Monteiro C, Miranda V. Promotion of new wind farms based on a decision support system. Renew Energy 2008;33:558e66. [2] Lejeune P, Feltz C. Development of a decision support system for setting up a wind energy policy across the Walloon region (southern Belgium). Renew Energy 2008;33:2416e22. [3] Baban SMJ, Parry T. Developying and applying a GIS-assisted approach to locating wind farms in the UK. Renew Energy 2001;24:59e71. [4] Heimiller DM, Haymes SR. Geographic Information Systems in Support of Wind Energy Activities at NREL. Preprints of 39th AIAA Aerospace Sciences Meeting. Reno, Nevada, USA; 2001. [5] Voivontas D, Assimacopoulos D, Mourelatos A, Corominas J. Evaluation of renewable energy potential using a GIS decision support system. Renew Energy 1998;13:333e44. [6] Hillring B, Krieg R. Wind energy potential in Southern Sweden e example of planning methodology. Renew Energy 1998;13(4):471e9. [7] Yue C, Yang CG. Decision support system for exploiting local renewable energy sources: a case study of the Chigu area of southwestern Taiwan. Energy Policy 2007;35:383e94. [8] Amador J, Domınguez J. Application of geographical information systems to rural electrification with renewable energy sources. Renew Energy 2005;30:1897e912. [9] Krewitt W, Nitsch J. The potential for electricity generation from on-shore wind energy under constraints of nature conservation: a case study for two regions in Germany. Renew Energy 2003;28:1645e55. [10] Dutra R, Szklo A. Assessing long-term incentive programs for implementing wind power in Brazil using GIS rule-based methods. Renew Energy 2008;33:2507e15. [11] 3Tier. First look world wind map. Available at:
763
[20] Canadian Ministry of Natural Resources. The Ontario wind resource atlas. Available at: