Energy Policy ] (]]]]) ]]]]]]

Contents lists available at ScienceDirect

Energy Policy
journal homepage: www.elsevier.com/locate/enpol

On developing a prospecting tool for wind industry and policy decision support$
Charles McKeown a,n, Adesoji Adelaja b,1, Benjamin Calnin c,2
a

The Department of Agricultural Food and Resource Economics (AFRE), Michigan State University, 401 Agriculture Hall, Michigan State University, East Lansing, MI 48823, USA MSU Land Policy Institute, 401 Agriculture Hall, Michigan State University, East Lansing, MI 48823, USA c MSU Land Policy Institute, 305 Manly Miles Building, 1405 South Harrison Road, Michigan State University, East Lansing, MI 48823, USA
b

a r t i c l e in f o
Article history: Received 30 July 2010 Accepted 4 November 2010 Keywords: Decision support Renewable energy Wind energy Resource assessment

abstract
This paper presents the rudiments of a Wind Prospecting Tool designed to inform private and public decision makers involved in wind industry development in reducing transaction costs associated with identifying areas of mutual focus within a state. The multiple layer decision support framework has proven to be valuable to industry, state government and local decision makers. Information on wind resources, land availability, potential land costs, potential NIMBYism concerns and economic development potential were integrated to develop a framework for decision support. The paper also highlights implications for decision support research and the role of higher education in providing anticipatory science to enhance private and public choices in economic development. & 2010 Elsevier Ltd. All rights reserved.

1. Introduction Interest in renewable energy has grown in the United States in recent years. The impetus for this is the growing prospects for rising energy prices, concern about climate change and global warming, and the potential that the US could play a prominent role in the renewable energy industry while enhancing national energy security (US DOE, 2008). Wind energy has received particularly signicant attention. Among the reasons for this are the general abundance of wind resources in coastal states and the great plains; the fact that wind represents a cleaner alternative to coal, natural gas, and nuclear power; the fact that wind does not involve signicant competition with other land-based industries for land; and the perceived existence of an opportunity for the US to drive technological advancements that can result in greater competitive or comparative advantage (OConnell, 2007). In recent years, the Obama Administration has sought to promote wind energy through its investments in research and development, especially via the American Recovery and Reinvestment Act of 2009 (ARRA). The bill included billions of dollars of investment in renewable energy through the Department of Energys community bloc grants,

industry loans and grants, and assistance to the states. This was designed to deepen the US presence in wind energy technology development and deployment. Over time, as fossil fuel based energy has become more expensive, wind energy costs have trended downward (Gross et al., 2003). However, despite this downward trend in cost of development and low operating cost of wind energy, its high initial cost represents a barrier to adoption (Owen, 2006). Other barriers include the limited status of the industry in the US, vis-a-vis its potential; the long duration between the choice of wind energy as an energy source and its development and deployment; the limited incentives in place at the state level to spur wind energy development by increasing consumer, utility company and local adoption (DSIRE, 2009); the limited existence of a state regulatory and management frameworks for managing wind energy expression on the landscape; and NIMBYism3the fact that many homeowners and communities tend to oppose wind turbines for visual, ecological (e.g. bird and bat kills) and other reasons (Alberts, 2005). While wind energy is clean and does not generate the usual environmental pollution typically associated with coal and other fossil fuel based energy systems, the challenges involved in its deployment remain a major barrier to adoption. The search for information and knowledge by policy makers and the industry makes the development of decision support tools essential. For one

$ The project was supported by funding the Hannah Professor Research Endowment at Michigan State University. n Corresponding author. Tel.: + 1 517 355 4702. E-mail addresses: mckeownc@msu.edu (C. McKeown), adelaja@landpolicy.msu.edu (A. Adelaja), calnin@landpolicy.msu.edu (B. Calnin). 1 Tel.: + 1 517 355 4702. 2 Tel.: + 1 517 432 8800x107.

3 Not in My Backyard (NIMBY) is dened here as local residents opposing wind energy developments in close proximity to their residence or place of work or in their community while not necessarily expressing opposition to the same type of development elsewhere.

0301-4215/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.enpol.2010.11.015

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]]

thing, such tools have the potential to reduce barriers to adoption and development. State policies are critical to the adoption of wind energy, as has been shown in states such as Texas, Minnesota, California, and Colorado (DSIRE, 2010). More recently, the province of Ontario Canada has essentially spurred signicant action in wind energy through the adoption of sweeping policy changes that involves a target of scaling back coal generating capacity (Ontario Legislature, 2009). Understanding where wind resources are abundant, where opposition is likely to be lower, where land values are reasonable and where other supportive infrastructure exists could improve state level planning for energy development. For example, in Michigan, the state created a Wind Resource Zoning Board (Michigan Senate, 2008) to identify the best regions where enabling policies and proactive transmission planning and deployment can be targeted. Wind developers also need such tools in order to facilitate their search for locations and their attempt to conduct focused community outreach and education. Perhaps more importantly, developers need to be able to get a quick read not only on resource availability, but also on various environmental, cost and community concern factors. This is why decision support tools such as the Land Policy Institute (LPI) Wind Prospecting Tool have been of value to the industry and government in the state of Michigan (Adelaja et al., 2007). The objective of this paper is to present the framework used in developing the LPI Wind Prospecting Tool. The tool guided the work of the Michigan Wind Resource Zoning Board, state agencies and other decision-makers at the local level. It has also been used by many wind-interested rms in site selection. The state of Michigan, which has been very active in developing policies to spur wind industry development, is our case study. In the balance of this paper, a background is provided on the Michigan wind development environment. This is followed by the description of the methodology used in developing the tool, features of the tool and the public and private sector implications.

Wind energy development may offer a new economic opportunity for Michigan. Renewable energy development has been a boon for some of the progressive nations in Europeresulting, for example, in the creation of tens of thousands of jobs in both Germany and Denmark (EWEA, 2009). Michigan currently relies on coal and nuclear fueled base-load generation units for about 83% of its annual electricity production, with the balance coming from natural gas, hydropower, and other sources. Less than 2% is currently generated from renewable energy (EIA, 2010). All of the fuel feedstock for coal and nuclear are imported from other regions of the US (MPSC, 2007). Annual dollar exports for coal-based electric energy alone were over $1 billion in 2006 (MPSC, 2007). Given the anticipated increase in the prices of fossil fuels, energy security is now a cornerstone of public policy in the state. The largely untapped wind resource in Michigan could give wind energy an edge, provided that some of the unnecessary barriers to wind energy development are reduced. The state is seeking to take critical steps not only to identify those barriers, but to strategically reduce them. Barriers to wind development in Michigan appear to be causing existing rms and investors to work in other states. These barriers create creating developer risk associated with inadequate site specic information about wind potential, contract opportunities, connectivity opportunities and pricing strategy. An operational objective of LPIs Wind Prospecting Tool project was therefore, to increase the chance to reduce the transaction costs associated with deploying wind systems in the state, thus fostering viable market activities where feasible.

3. Specic objectives and rationale The Wind Prospecting Tool was developed to assist state and local policy decision makers and the wind energy industry in understanding where the wind industry development pressure will be expressed the most in Michigan, and to help wind energy developers shorten the search time for viable locations. The framework was designed to be an easily understood, integrated resource that can help:

2. Background on Michigan Despite its unique location and assets, Michigan lags behind many other states in the nation in installed wind generating capacity. According to the National Renewable Energy Laboratory (NREL), while Michigan has the land available and the wind resource to install 59,000 MW (plate capacity4) of potentially developable wind resource capacity onshore when landscape and wind resource are considered (Elliott et al., 2010), of the thirty one states that have installed systems, Michigan ranks twenty sixth in installed capacity but fourteenth in resources. Fewer than 150 MW have been installed in Michigan (AWEA, 2010), in contrast to Texas which has installed a total of 9410 MW of capacity (AWEA, 2010). Michigan is very near the bottom in terms of the ratio of installed to potential wind generating capacity, despite its greater need for energy independence due to its relative isolation as a peninsula state without large resources of coal or other conventional fuel for electricity generation. According to a report by the Renewable Energy Policy Project, the gap in Michigans capacity offers tremendous opportunity for wind energy industry growth (Sterzinger and Syrcek, 2004). Nationwide, states are increasingly pursuing renewable energy sources and many with wind energy potential are pursuing such potential as a matter of policy. According to Michigans 21st Century Energy Plan, approximately 52009200 GWh of additional renewable energy is needed by December 31, 2015 (Lark, 2007).
4 Plate Capacity refers to a manufacturers rated total power output capacity of a wind energy generating system or the sum of these capacities for a wind farm. Actual output varies widely due to wind resource levels and site design.

 Filter out areas of low potential for wind energy development.  Focus efforts of stakeholders on high quality areas, and provide  
critical analysis of policy gaps in those areas and at the state level to enable wind energy development. Target statewide policy and the wind development communitys investment toward those areas that are most conducive for wind energy development. Help communities understand their own wind development potential.

The development of the tool focused on nding and assessing the capacity of the best areas for wind development in Michigan and assessing community by community the potential for local development. It was designed to lter, focus and target information on wind energy development in Michigan by providing information on four categories of receptivity factors5 that the wind energy industry, and government experts indicated were important:

 Geophysical factorsthose factors inherent to the landscape

such as topography, lakes and wetlands.
5 An attempt was made in this study to identify, as much as possible, all major decision factors that could affect business wind site location and government selection of locations to emphasize appropriate industry development enabling policies. All factors considered in this analysis were selected based on the polling if industry and government representatives about relevant factors to consider in building a value added tool.

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]]

Table 1 Data used in the wind index and what it represents. Data Wind speed score Agricultural land contiguity and area Forest land contiguity and area Per acre value of agricultural land Population density: 2000 Population density change: 19902000 Total possible Proxy for Wind density for power generation The number of towers that can be strung together in a reasonably compact setting The number of towers that can be strung together in a reasonably compact setting Land Costs Possible local resistance to wind farm installation Pressure for residential and other types of development Possible index score 350 180 130 130 130 80 1000

 Land/economic factorsthose factors related to land use,  

development pressure and local economic conditions. Environmental concernsthose factors related to the presence or absence of endangered species, rare ecological communities, and highly unique habitats. Local policythose factors that relate to planning and zoning6 and the expression of NIMBYism.

The presence of these factors and their interaction can make a community an excellent place for wind power development or present transaction costs so high as to preclude it entirely. Articulating these factors allows community and state decisionmakers to understand and address policy and educational efforts to reduce or eliminate those transaction costs.

4. Methods Wind site selection decisions are essentially site suitability exercises. Therefore, an appropriate starting point and modeling framework is a site suitability model that weighs various factors (location benets and barriers) relative to each other based on a set of key site selection criteria. This approach is commonly used in ecologic studies to determine the suitability of a particular habitat to an individual species or clades of species (Ahmadi-Nedushan et al., 2006), in land use modeling to predict patterns of development (McDonald and Brown, 1984), and in the corporate and municipal planning process (Malczewski, 2003). This approach, however, has not been widely applied in wind energy industry site selection, Voivontas et al. have developed a decision support tool, however it does not extend beyond the wind resource and landscape issues (Voivontas et al., 2009). Given the specic site characteristics that are necessary for a wind projects development, a suitability index modeling framework that includes economic and policy indicators, was determined to be the appropriate methodology for developing the Wind Prospecting Tool. Denote the desirable attributes of a given potential wind development project site by the matrix X such that each Xij represents the level of the ith attribute at the jth site. The attributes should include all factors relevant to the developers initial site selection process, which our expert panel indicated would include the wind resource capacity, availability of open space, favorability of land values, low opposition to wind development and low potential for long term encroachment of other land uses. Denote the aggregate site suitability of the jth location as Yj such that the n P suitability function takes the form Yj gi Xij . Note that a simple
i1

linear aggregate function is assumed where gi 4 0, for all i.

6 This is an important consideration in Michigan, where local governments (1856 local governments in all) are responsible for planning and zoning and local planning and zoning commissions preside over relatively small jurisdictions and in many cases small populations as well.

Yj is the aggregation coefcient associated with the jth suitability index and Y is the threshold level that the developer is looking for to determine that a project site is worth exploring for making appropriate leasing and site study investments. If Yj Z Y , the developer nds the site attractive. However, if Yj o Y the site is unattractive. As developers and the state seek places to concentrate their efforts, the spatial distribution of Yj becomes important. Of course, Yj varies across locations and is also dependent on the inherent Xi factors. Note that other factors affect site selection, but are considered in the second step in the site selection process (e.g. information on grid access and networking charges) which can only be discovered after a developer zeros in on a potential site and begins the to explore the site specic grid connectivity and other issues. Given the sequential nature of the siting process where initial suitability is determined rst, followed by nal decisions (which can take years), we focused our study on the former. The omission of the availability and proximity of transmission infrastructure is obvious, but our initial focus is initial site selection since that represents the rst step in a complete site assessment. Transmission is obviously a key factor in siting wind energy developments and the lack of transmission (in terms of presence, interconnection options, and congestion) is one of the primary bottlenecks to development in the US (US DOE, 2008). However, our ability to incorporate transmission information into the tool was constrained by homeland security concerns which under the USA Patriot Act (US Congress, 2001), makes the spatially explicit details about transmission infrastructure sensitive information and not for publication. Furthermore, proximity to transmission is not always the most relevant factor in connecting to the grid in the Midwest Systems Operator (MISO) region, where the grid is congested in places and well under capacity in other areas. Each potential interconnection is modeled by both MISO and the transmission company to assess the ability of the grid to absorb the proposed power production at the applicants proposed location. If there are improvements required by the applicant, the cost of those improvements is borne by the applicant. The fact that grid connection studies are carried out in two stages with separate costs for each stage and the uncertainty in the cost of eventual connection and networking makes it virtually impossible to attempt to simultaneously and accurately factor in transmission. Therefore the tool was focused on those early stage decision factors in wind development. Grid connection studies are recognized as problems in wind energy development in Michigan, and this recognition eventually led to the inclusion of provisions in Michigans renewable portfolio standard legislation to help rectify the situation. That outcome is discussed in greater detail later in this paper. To obtain information on relevant X categories, we convened an expert group consisting of wind developers, electricity regulators, transmission companies and local policy makers empanelled to also help weigh the relative importance of each factor (the gs). The process consisted of a series of questions to panelists in March of 2007 to weigh the relative importance of each factor and also

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]]

solicited recommendations regarding the addition or deletion of factors. This provided the basis for understanding the industries decision making process. That panel identied decision factors and determined that the initial decision factors are inherently xed across places and a weighting technique was needed to identify relative weights or gs. Participants were then asked to assign scores to each g ranging from 1 to 1000, bearing in mind the relative values of g and the importance in the decision making process. The gs obtained via this process were the consensus of the panel and there was strong agreement on the nal weights. These weights were then presented to other experts for review and feedback before applying them to the analysis. This process yielded consensus. The 1000 point index was then constructed using the weights provided by the process described above. The nal index and the scoring for each factor are shown in Table 1. From Table 1, it is obvious that the wind speed score is a key factor in site selection, not only for the public sector, but for the private sector as well. This is followed by land availability factors which, jointly, account for 310 possible points. Land acquisition costs come next with 130 possible points; and population density, a proxy for local resistance at 130 possible points. Future development pressure and future+ other pressures are proxied by population density change and this factor has a possible score of eighty points. Various datasets were used in this analysis, including the US Census (US Census Bureau, 2004a, 2004b) for demographic information, USGS National Land Use and Land Cover data (USGS, 2004) for landscape characteristics, the Michigan State Tax Commission(STC, 2007) reports for land valuation, the Michigan Geographic Data Library(CGI, 2009) for Community Mapping, and the fundamental wind resource map used was the National Renewable Energy Laboratory fty meter wind density map for Michigan (AWS Truewind, 2004).

N W S E

Index Score
NREL Wind Class Areas Classes 3 - 7
0-1 2-3 4-6 7 - 12 13 - 39 40 - 101 102 - 244
0 30 60 120 Miles 180

Fig. 1. NREL wind classication and scoring by community in Michigan.

5. Index components 5.1. Wind score Class three or better wind, as modeled by NREL, is generally considered to be the threshold for utility scale wind development (Elliot et al., 1991). The wind score is a result of ltering the NREL original 50 m wind density data to produce a map of only class three to seven wind resources in Michigan. The area in each class was scaled and added to produce the nal wind resource score for a community. Fig. 1 provides a visualization of the distribution of wind resources by location in Michigan. In the map, municipalities were scored on the basis of a total of 350 possible points. 5.2. Area of agriculture with wind Agricultural land has proven to be one of the most important land types for the installation of wind turbines. Wind turbine installation on agricultural land allows a farmer to continue farming the land because of the minimal footprint of each tower, and the income generated for the farmer by the leases is typically far greater than the minimal loss in capacity to produce crops where the turbines are installed. The larger the area of agricultural land within a community, the greater the amount of towers that can be installed within that community. In addition, the ability to assemble coalitions of landowners interested in hosting turbines is increased. Fig. 2 provides a visualization of the distribution of agricultural land with adequate wind resources by location in Michigan. Note that the highest score for any location in Michigan is 90, which is half of the total points available for the agricultural land and contiguity score.

N W S E

Index Score
Area of Agricultural Land With Suitable Wind Density at 50m
0-2 3-6 7 - 11 12 - 23 24 - 37 38 - 57 58 - 90
0 30 60 120 Miles 180

Fig. 2. Area of agricultural land by community in Michigan.

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]]

N W S E W

N E S

Index Score
Agricultural Land Contiguity With Suitable Wind Density at 50m
8 - 24 25 - 37 38 - 45 46 - 52 53 - 59 60 - 72 73 - 90
0 30 60 120 Miles 180

Final Areas
Area of Forest Land With Suitable 50m Wind Density
0-1 2-3 4-6 7 - 11 12 - 20 21 - 40 41 - 65
0 30 60 120
Fig. 4. Area of forest land by community in Michigan.

Miles 180

Fig. 3. Agricultural land contiguity with suitable wind by community in Michigan.

5.3. Contiguity of agricultural land with wind The cost of an installation and the ease of interconnection are partially decided by the compactness of the entire wind farm. Some communities have large areas of agricultural lands that are scattered throughout the landscape while others have agricultural lands that are densely packed. The contiguity measure used is derived from the discipline of landscape ecology. Ot is a direct measure of how connected or separated agriculture is in the area in question. Scores are determined within each community using the Fragstats analysis environment (McGarigal et al., 2002). Fig. 3 provides a visualization of the contiguity of agricultural land by location in Michigan. Note that the highest score for any location in Michigan is 90, which is half of the total points available for the agricultural land and contiguity score.

is half of the total points available for the forest land and contiguity score. 5.5. Contiguity of forest land with wind As with agricultural landscapes, the cost of an installation and the ease of interconnection are partially decided by the compactness of the entire wind farm. Some communities have large areas of forest that are scattered throughout the landscape while others have largely connected tracts. Our contiguity measure was also derived from the discipline of landscape ecology as a direct measure of how connected or separated forest land is in the area in question. The scores are also determined within each community using the Fragstats analysis environment. Forest contiguity with wind is important to wind energy development. The reduced attraction of contiguity of forest land is reected in the lower total possible score (as compared to agricultural lands) of 130 points for forested land. Fig. 5 provides a visualization of the contiguity of forest land with adequate wind resources by location in Michigan. Note that the highest score for any location in Michigan is 65, which is half of the total points available for the forest land and contiguity score. 5.6. Open space (agriculture and forest) land value Land value is a fundamental metric in determining lease rates and local taxes, and also serves as an indicator of other development pressures. As the value of open landscapes increases, the cost of wind instillations also increases. Therefore, low land values score high on the index. Agricultural value, as dened by the state tax commission, also includes forest land value. Note that the maximum score for open space land value is 130 points. Fig. 6 provides a visualization of open space land value distribution in Michigan.

5.4. Area of forest with wind Forest area with wind is important to wind energy development, although possibly less desirable than agricultural land for wind development due to siting concerns. This is reected in the lower total possible score of 130 points for forested land. The NREL map models the effect of land cover, including turbulence and obstructions. Therefore, forested areas of Michigan are shown to have high wind resources, despite land cover related issues. As with agriculture, the more area of forest within a community, the more potential towers a wind developer can concentrate in an area. With high interconnection costs to the grid, it is important to wind developers that there be enough installed power capacity and wind turbines in an area to be able to offset the cost of interconnection into the grid. Fig. 4 provides a visualization of the distribution of forest land with adequate wind resources by location in Michigan. Note that the highest score for any location in Michigan is 65 which

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]]

5.7. Population density: 2000 Population density in this index is used to measure the potential for local resistance in a community to wind development. The higher the population density within a community, the more likely there are going to be individuals with concerns about issues such as view shed impingement, ice throw, icker fusion and bird strikes. While population density is an imperfect measure of the development of a NIMBY reaction, it has been shown to be an indicator in other controversial land uses such as prison siting (Hoyman, 2006), US Federal Emergency Management Administration trailer park siting (Davis and Bali, 2008) and landlls, incinerators and connectional facility siting (Rasmussen, 1992). Note that the maximum score for population density is 130 points. Fig. 7 provides a visualization of population density distribution in Michigan. 5.8. Population density change: 19902000 Population density change measures a communitys potential for other types of development pressures such as residential or commercial development which may be more nancially rewarding than wind leases. This metric is used to capture the willingness on the part of land holders to enter into long term leases vs. the potential payoff from land sale to other types of development. Communities with low population growth are given high index scores; all negative values were given a score of 100. Note that the maximum score for population density change is 130 points. Fig. 8 provides a visualization of population density change throughout Michigan. 5.9. Zoning score One of the key factors in determining the suitability of an area for wind energy development is local zoning laws applicable to

N W S E

Index Score
Forest Land Contiguity With Suitable Wind Density at 50m
6 - 17 18 - 24 25 - 29 30 - 34 35 - 42 43 - 51 52 - 65
0 30 60 120
Fig. 5. Forest land contiguity by community in Michigan.

Miles 180

N W S E W

N E S

Index Score
Value of Open Space
83.3 - 103.9 104.0 - 117.3 117.4 - 121.6 121.7 - 123.9 124.0 - 126.1 126.2 - 128.5 128.6 - 130.0
0 30 60 120
Fig. 6. Open space score by community in Michigan.

Index Score
Population Density: 2000
1 - 56 57 - 86 87 - 106 107 - 118 119 - 125 126 - 128
Miles 180

129 - 130
0 30 60 120

Miles 180

Fig. 7. Population density in 2000 by community in Michigan.

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]]

N W S E W

N E S

Index Score
Population Density Change: 1990 - 2000
22.5 - 42.4 42.5 - 64.3 64.4 - 72.0 72.1 - 75.4 75.5 - 77.7 77.8 - 79.2 79.3 - 80.0
0 30 60 120 Miles 180

Index Score
Zoning Score
(-23) - (-26) (-27) - (-28) (-29) - (-41) (-42) - (-46) (-47) - (-56) (-57) - (-69) (-70) - (-100)
0 30 60 120 Miles 180

Fig. 8. Population density change from 1990 to 2000 by community in Michigan.

Fig. 9. Wind zoning index score by community in Michigan.

wind turbine and energy development within the community. A review of the zoning language in Michigan applicable to wind power development was conducted, and the level of barrier presented by zoning was assessed, ranked and scaled. The nal value was then subtracted from the nal score. Unfortunately, zoning scores have the only potentially negative values as there are no communities that have passed enabling ordinances that reduce barriers for wind development. Positive score are possible. Also communities with no language pertaining to wind were assigned a zoning score of zero. Note that the maximum score for local policy is 130 points which no community actually realizes due to the mixed nature of ordinance restrictions. Fig. 9 provides a visualization of local zoning scores throughout Michigan.

N W S E

5.10. Total index score The scores discussed above were summed in order to produce nal state and community maps. These maps were then intersected with the NREL class three and up areas to clip out the areas without documented wind resources. Fig. 10 shows the index score with the inuence of the zoning subtracted. Though high index scores are most often found in coastal communities, it is important to note that there are a number of inland communities with relatively high index scores. The highest scoring communities in Michigan (index scores of 500 or better) were then selected and aggregated into the top 12 wind utility scale wind development areas in the state. It is important to note that the grid and transmission issues have not been addressed in this framework due to homeland security issues. Some of these areas will possibly be later determined to be impractical due to interconnection and grid capacity issues. Fig. 11 shows the communities that make up each of the top 12 areas for utility scale development. These areas of Michigan

Index Score

Total
With Zoning
59 - 249 250 - 339 340 - 391 392 - 430 431 - 487 488 - 584 585 - 714
0 30 60 120 Miles 180

Fig. 10. Wind index score with adjustments for local policy by community in Michigan.

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]]

represent the areas where state policies have become focused since the development of the wind prospecting tool. It is no surprise that the top 12 areas represented in Fig. 11 are the same areas where wind generation applications have intensied in Michigan. This very fact highlights the signicant role that the Wind Prospecting Tool played in streamlining public and private decision making with respect to wind prospecting and policy focus in Michigan.

6. Development scenario results Perhaps as important as the understanding of high suitability areas is the understanding of the potential economic impacts of future development. This is particularity so in Michigan where the state is in dire need of new job opportunities and where wind industry development is at the forefront of state strategies to promote such opportunities. To provide relevant information on

potential employment and income opportunities, the top 12 areas in the state were further examined to determine the possible number of towers they can accommodate as well as estimating power outputs, lease values, maintenance and upkeep jobs as well as construction job creation. Wind turbines are generally spaced no closer than ve times their rotor diameter (Patel, 2006). Using this guideline, 450 m spacing was determined to be a reasonably conservative estimate of tower density as it represents a 90 m rotor diameter. The largest turbines commissioned for installation in Michigan have an 80 m rotor diameter. The power possible was calculated by assuming a 1.65 MW turbine at 28% capacity factor (Lark, 2007). Job creation was estimated by employing the Jobs and Economic Impact Model (JEDI) tool developed by the National Renewable Energy Laboratory (NREL, 2007). To generate reasonable estimates of impacts, several scenarios were calculated using 5%, 10%, 15% and 20% of the wind resource area. These levels were chosen as reasonable scenarios of development potential and they provide a basis for comparison and contrast of the impacts of wind development. The results of this analysis at the 15% and 20% levels are shown in the Tables 2 and 3.

7. Community tool The community level wind tool component includes information on each Michigan communitys potential for wind development. This analysis involved not only the top 12 areas, but all areas in Michigan with adequate wind resources. Communities were provided customized printable reports that include information on the communitys LPI wind index score and the potential number of towers, jobs, and revenue the community can receive with wind development. The tool component also included zoning laws applicable to wind energy development in each community, allowing residents and policy makers within the community to explore whether or not local zoning can impact on wind development. Fig. 12 represents a screenshot from the larger tool provided on the LPI website for Michigan communities to access. The area of the state represented in the map is the thumb area in the Eastern part of the lower peninsula of Michigan.

N W E S

Final Areas
Top 12
Eastern UP Keweenaw Leelanau Mason County Presque Isle Straits Thumb 1 Thumb 2 Thumb 3 Traverse Bay UP Ribbon UP Ribbon 2
0 30 60 120 Miles 180

8. Additional value added information available through the tool map server Beyond the factors already addressed above, there is a host of environmental and landscape issues that affect wind power siting and development. With the help of project partners, this information was included in the tool as well. These factors include areas of critical habitat for threatened and endangered species obtained from the Michigan Natural features Inventory (MNFI, 2007), conservation and

Fig. 11. The Top 12 contiguous areas for wind development in Michigan.

Table 2 Economic and power projections if 15% of the available land area is used in the top 12 locations in Michigan. Area name Turbines possible Power production possible (MW) 84.7 118.0 91.1 32.7 11.2 67.7 126.2 147.6 126.2 69.3 49.7 19.5 Potential land lease revenue $366,588 $510,714 $394,434 $141,372 $48,348 $293,148 $546,210 $638,928 $546,516 $300,186 $215,118 $84,456 Potential maintenance and upkeep jobs 24 34 26 9 3 19 36 42 36 20 14 6 Potential construction jobs 372 518 400 143 49 297 554 648 555 305 218 86

Eastern IP Keweenaw Leelanau Mason County Presque Isle Straits Thumb 1 Thumb 2 Thumb 3 Traverse Bay UP Ribbon UP Ribbon 2

183 255 197 71 24 147 273 319 273 150 108 42

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]]

Table 3 Economic and power projections if 20% of the available land area is used in the top 12 locations in Michigan. Area name Turbines possible Power production possible 553.5 771.1 595.5 213.4 73.0 442.6 824.7 964.7 825.1 453.2 324.8 127.5 Potential land lease value $479,200 $667,600 $515,600 $184,800 $63,200 $383,200 $714,000 $835,200 $714,400 $392,400 $281,200 $110,400 Potential maintenance and upkeep jobs 32 44 34 12 4 25 47 55 47 26 19 7 Potential construction jobs 486 677 523 188 64 389 725 848 725 398 285 112

Eastern IP Keweenaw Leelanau Mason County Presque Isle Straits Thumb 1 Thumb 2 Thumb 3 Traverse Bay UP Ribbon UP Ribbon 2

240 334 258 92 32 192 357 418 357 196 141 55

Fig. 12. Screenshot of the community wind tool web interface for Michigan.

recreation land holdings obtained from the Nature Conservancy (TNC, 2007), wetlands obtained from the US National Wetlands Inventory (USGS, 2006), inland lakes and rivers obtained from the Michigan Center from Geographic Information (CGI, 2009) and steep slopes obtained from the US Geological Survey (USGS, 2005). These areas were not subtracted from the total area available for development. They serve as indicators that, as part of comprehensive site assessment, may be some areas of concern. For example, a conservation easement for agriculture may or may not eliminate the possibility of wind turbine instillation. The location of the easements on the map indicates that this issue must be investigated. Similarly, the presence of an endangered species may eliminate an area or simply require special construction considerations. With future funding, these issues can be examined in more detail. This also applies to the modeling of migratory yways and avian habitat to minimize bird strike potential. An example of potential exclusionary areas in Manistee County Michigan is shown in Fig. 13.

9. Summary and conclusions The LPIs Wind Prospecting Tool represents an innovative integration of key industry and policy considerations in developing

a decision support framework for wind industry development. While the analysis might appear simple and straightforward, it highlights the importance of decision support tools to guide the choice of synergistic opportunities for industry and government to work together. While this tool was developed specically for Michigan, it has been of interest to companies and policy makers in many other states. It has also been the subject of keen interest by state ofcials in Michigan and beyond who are exploring how university research can help in public policy, especially in areas where industry and government goals are not necessarily in conict. Zeroing in on the sweet spots where opportunities exist, while avoiding prescriptive value judgments in policy research is an area of important need with respect to university research, especially during these times when the private and public sectors are seeking new opportunities to streamline public and private choices that are meaningful in economic development. From the analysis above, it is obvious that the range of considerations involved in informing a public policy challenge is vast, if not outright complicated. Unfortunately, however, opportunities for constructive dialogue that will lead to valuable frameworks for moving forward are limited, especially when the issue of concern involves multiple companies, multiple consumer interests and concerns, and sometimes overlapping and conicting public

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

10

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]]

Fig. 13. Map showing one county with potential exclusionary areas for ecological reasons.

policy objectives. While this paper stops short of presenting details about how Michigan leveraged this tool to create winwin solutions, the value of the tool has been signicant. For example the Michigan Wind Resources Zoning Board, which was mandated by the Michigans Clean Renewable and Energy Efciency Act of 2007, contracted with the LPI to use the tool framework to help the state delineate areas of priority for wind energy development and transmission planning as specied in that legislation. To date, though two contracts, the LPI provided such service to the state of Michigan (PSC, LPI, 2009). These decision support activities played a critical role in the assessment of wind in Michigan, in the analysis of the transmission grids ability to integrate wind energy, and eventually in the adoption of a cooperative model for designing and funding upgrades to transmission infrastructure to allow wind energy a faster path to grid integration. The tool has also been leveraged by several international, national, and state based companies in the process of prospecting for wind development projects in Michigan. Decision support tools are valuable only when forward looking choices are made to anticipate and provide, ahead of time, information and analysis that can add value. Such information is typically characterized by the integration of multiple analysis techniques and disciplines, including spatial analysis, economic analysis, policy analysis and the polling of policy makers and industry decision makers. This raises a challenge to researches that are neither close enough to government or industry to understand the needs of either. Obviously, synthesis is important in generating effective tools that add value. Perhaps more importantly, a multidisciplinary approach is required, which can often be a challenge in the setting of higher education. Other studies of a similar nature by the authors include the Michigan Offshore Wind Resource Assessment, Michigans Browneld and Renewable Energy Resource

Assessment, and LPIs analysis of Renewable Portfolio Standards and their inherent features in the development of renewable energy. All three led to signicant policy changes and industry coalescence around catalytic industry development efforts in renewable energy. The analyses presented above also involved signicant outreach, including the creation of a web based tool as the primary delivery mechanism. This is made possible by the integration of research, targeted outreach and technology. This raises the question of whether or not the private sector should develop this type of decision support. It probably should if the tool is valuable only to the private sector. However, on potentially contentious issues such as wind development, the benets of these types of tools are largely to the public sector. This in turn raises the question of whether or not adequate resources will be provided ahead of time by the public sector, especially in an environment where tools development must involve the anticipation of public policy needs in order for such tools to be valuable within the time frame in which public choices are made.

References
Adelaja, S., et al., 2007. The Land Policy Institute Wind Prospecting Tool Prototype. Land Policy Institute, Michigan State University, East Lansing, Michigan. Ahmadi-Nedushan, B., St-Hilaire, A., Berube, M., Robichaud, E., Thiemonge, N., Bobee, B., 2006. A review of statistical methods for the evaluation of aquatic habitat suitability for instream ow assessment. River Research and Applications 22, 503523. doi:10.1002/rra.918. Alberts, D.J., 2005. Delphi Inquires Into Wind Siting Issues, Report to the Michigan Energy Ofce, 2005, Lawrence Technical University, MI. American Wind Energy Association, 2010. Year End 2009 Market Report. American Wind Energy Association, Washington, DC.

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015

C. McKeown et al. / Energy Policy ] (]]]]) ]]]]]] AWS Truewind, 2004. Wind Speed of Michigan at 70 meters. /http://www. michigan.gov/documents/dleg/MI_FinalSPD70m_09_09_04_238941_7.pdfS (accessed October 1, 2008). Database of State Incentives for Renewables and Energy Efciency, 2010, DSIRE. Available at: /www.dsire.orgS (accessed March 25, 2010). Davis, B.C., Bali, V.A., 2008. Examining the role of race, NIMBY, and local politics in FEMA trailer park placement. Social Science Quarterly 89, 11751194. doi:10.1111/ j.1540-6237.2008.00563.x. Elliot, D.L., Wendell, L.L., Gower, G.L, 1991. An Assessment of the Available Windy Land Area and Wind Energy Potential in the Contiguous United States. National Renewable Energy Laboratory, Golden Colorado. Elliott, D.L., Schwartz, Marc, Haymes, Steve, Heimiller, Donna, Scott, George, Flowers, Larry, Brower, Michael, Hale, Erik, Phelps, Bryon, 2010. 80 m and 100 m Wind Resource Map for the United States. National Renewable Energy Laboratory, Golden Colorado AWS Truewind, Albany NY. European Wind Energy Association (EWEA), 2009. Wind at Work: Wind Energy and Job Creation in the EU. European Wind Energy Association, Brussels, Belgium. Robert, Gross, Leach, Matthew, Bauen, Ausilio, 2003. Progress in renewable energy. Environment International 29 (1). doi:10.1016/S0160-4120(02)00130-7 ISSN 0160-4120. Hoyman, M., Weinberg, M., 2006. The process of policy innovation: prison sitings in rural North Carolina. Policy Studies Journal 34, 95112. doi:10.1111/ j.1541-0072.2006.00147.x. Lark, P., 2007. Michigans 21st Century Energy Plan, Michigan Public Service Commission Report, Lansing Michigan. Malczewski, J., 2003. GIS-based land-use suitability analysis: a critical overview. Progress in Planning 62 (1), 365. doi:10.1016/j.progress.2003.09.002 July 2004, ISSN 0305-9006. McDonald, G.T., Brown, A.L., 1984. The land suitability approach to strategic landuse planning in urban fringe areas. Landscape Planning 11 (2), 125150. doi:10.1016/0304-3924(84)90035-2 July, ISSN 0304-3924. McGarigal, K., Cushman, S.A., Neel, M.C., Ene, E. 2002. FRAGSTATS: Spatial Pattern Analysis Program for Categorical Maps. Computer software program produced by the authors at the University of Massachusetts, Amherst. Accessed at: /http://www.umass.edu/landeco/research/fragstats/fragstats.htmlS. Michigan Center for Geographic Information (CGI), 2009. Michigan Geographic Framework: State of Michigan. Version 9b. Michigan Center for Geographic Information, Lansing, MI. Michigan Senate, 2008. The Clean, Renewable, and Efcient Energy Act (Public Act 295 of 2008). Michigan Legislature, Lansing Michigan. Michigan State Tax Commission (STC), 2007. Ad Valorem Property Tax Reports. Michigan Department of Treasury, Lansing MI. Michigan Public Service Commission, (MPSC), 2007. Fuel Mix Disclosure Data. Lansing Michigan. Accessed at: /http://www.dleg.state.mi.us/mpsc/electric/ download/fuelmix.pdfS.

11

OConnell, R., Pletka, R., 2007. Percent Wind Energy Penetration in the United States: A Technical Analysis of the Energy Resource. Black and Veatch, Overland Park, KS. Ontario Legislature, 2009. Green Energy Act, Ontario Legislature, Bill 150, Toronto Canada. Owen, A., 2006. Renewable energy: externality costs as market barriers. Energy Policy 34(5), Hong Kong Editorial Board meeting presentations, March 2006, ISSN 0301-4215, 10.1016/j.enpol.2005.11.017. Patel, M., 2006. Wind and Solar Power Systems, Design, Analysis and Operation. CRC Press, Boca Raton, Florida. Public Sector Consultants, Land Policy Institute, 2009. Final Report of the Michigan Wind Energy Resource Zone Board, Lansing Michigan. Available at: /http:// www.dleg.state.mi.us/mpsc/renewables/windboard/werzb_nal_report.pdfS. Rasmussen, Thomas H., 1992. Not in my backyard: the politics of siting prisons, landlls, and incinerators. State & Local Government Review 24 (3), 128134 Autumn. Sterzinger, G., Syrcek, M., 2004. Wind Turbine Development: Location of Manufacturing Activity, Renewable Energy Policy Project, Washington, DC. The Nature Conservancy (TNC), 2007. The Conservation and Recreation Lands Database. /http://www.nature.org/wherewework/northamerica/states/michigan/science/ art22873.htmlS (accessed June 15th, 2007). US Congress, 2001. Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism (USA Patriot) Act of 2001. P.L. 107-56. US Census Bureau, 2004a, US Census of Population and Housing, Washington, DC, 2004. US Census Bureau, 2004b, US Population Projections, Washington, DC, 2004. US Department of Energy, Energy Efciency and Renewable Energy, 2008. 20% Wind Energy by 2030: Increasing Wind Energys Contribution to US Electricity Supply. US Department of Energy, Washington, DC. US Geological Survey (USGS), 2004. National Land Use and Land Cover Dataset, US Geological Survey, Washington, DC. US Geological Survey (USGS), 2006. The National Wetlands Inventory, US Geological Survey, Washington, DC. US Geological Survey (USGS), 2005. The National Elevation Dataset, US Geological Survey, Washington, DC. US Energy Information Administration (EIA), 2010. State Electricity Proles, Washington DC. Accessed at: /http://www.eia.doe.gov/cneaf/electricity/ st_proles/sep2008.pdfS. Voivontas, D., Assimacopoulos, A., Mourelatos, J.Corominas, 2009. Evaluation of renewable energy potential using a GIS decision support system. Renewable Energy 13 (3), 333344. doi:10.1016/S0960-1481(98)00006-8 March 1998, ISSN 0960-1481.

Please cite this article as: McKeown, C., et al., On developing a prospecting tool for wind industry and policy decision support. Energy Policy (2010), doi:10.1016/j.enpol.2010.11.015