Wind Resource Development in the Minnesota Coastal Zone (306-02-08)

8/14/2019 Wind Resource Development in the Minnesota Coastal Zone (306-02-08)

1/46

1

Minnesotas Lake Superior Coastal Program

Wind Resource Development in the Minnesota Coastal Zone

Michael Mageau, Brody Sunderland and Stacey Stark

University of Minnesota, DuluthCenter for Sustainable Community Development

Department of Geography328 CINA Hall218-726-6133

August 30, 2008

Project No. 306-02-08Contract No. A92528

This project was funded in part under the Coastal Zone Management Act, by NOAAs Office ofOcean and Coastal Resource Management, in cooperation with Minnesotas Lake SuperiorCoastal Program.


2/46

2

AcknowledgementsWe would like to thank the following individuals for their many contributions to this project.UMD students Nick Salo, Nick Entinger, Janelle Stauff, Nicole Hynum and Melissa Wenker forall their work on the project. We would also like to thank UMDs Linda Klint and Kirsten Lien

for their work on project budgets, invoinces and accounting. As well as the Minnnesota CoastalZones Karla Sundberg and Pat Collins for their advice and guidance over the entire course ofthe project. Finally, we would like to thank our numerous community partners from eachmonitoring site for their support, cooperation, interest and past/current/future wind developmententhusiasm and planning.

This project was funded in part under the Coastal Zone Management Act, by NOAAs Office ofOcean and Coastal Resource Management, in cooperation with Minnesotas Lake SuperiorCoastal Program. We would also like to thank the U of MNs Northeast Region SustainableDevelopment Partnership (NMSDP), Minnesotas NE Region Clean Energy Resource Teams(CERTS), and UMDs College of Liberal Arts for critical matching project funds.

IntroductionWind development is one of the fastest growing sectors in the energy industry today. Under theright conditions it can be a large and sustainable local economic development opportunity.Current wind resource estimates (i.e. MN DOC) for the Lake Superior Coastal Zone indicate thatwe have a poor wind resource not suitable for development. These wind resource estimates arebased on climatological modeling, and are no substitute for site-specific measurements. Weessentially wanted to verify current wind resource estimates with site-specific monitoring, andexplore the economic viability of wind development in our region.

The overall project was designed to achieve three primary objectives: 1. Obtain a minimum ofone years worth of quality wind speed data from eight sites (Duluth, Clover Valley, Silver Bay,Finland, Lutsen, Grand Marais, Hovland and Grand Portage) along the Northshore of LakeSuperior; 2. Use this site-specific wind speed data to create a wind resource map for the entireregion to include several key overlays (topography, roads, transmission, bird migrations); 3. Usethis wind speed data to conduct community-scale wind development economic feasibilitystudies, and determine the direct, indirect and induced (employment) economic impactsassociated with potential future wind development in Southern St. Louis, Lake and CookCounties.

Objective I: Wind Monitoring

II. MethodsA. Site Selection:

Our study began with the process of selecting appropriate sites for monitoring the wind. The keycriteria dictating site selection included the following: 1. We needed an existing structure thatcould be climbed and was at least 100 feet tall; 2. We were looking for monitoring sitesassociated with prominent peaks; 3. We needed year round road access to each site; 4. Weneeded a set of sites that were equally spaced along the Northshore; 5. We needed permission to


3/46

3

have unlimited access to these sites; and finally 6. We needed a good community partner towork with. Fortunately, we were able to meet each these criteria at eight different sites fromDuluth to Grand Portage. Figure 1 illustrates the geographical location of each site. Table 1lists our monitoring sites, and describes several key characteristics of each.

Figure 1. The geographic location of each monitoring site

Table 1. Key characteristics of each monitoring site.

Site Structure Elev (ft) Height (ft)Enger Tower Tower 1146 95Clover Valley Fire Tower 1400 110Silver Bay Weather Stat 620 100Finland Fire Tower 1865 120Lutsen Com Tower 1745 120Grand Marais Com Tower 1722 100Hovland Fire Tower 1774 95Grand Portage Com Tower 1720 30m

50m80m


4/46

4

B. Equipment:

The wind monitoring component of the project required the following equipment: 1.Anemometers for measuring the wind; 2. Dataloggers for recording the wind speed data; 3.Equipment for mounting the anemometers to the existing towers; 4. Software for downloadingthe data; and 5. Excel spreadsheets for data analysis and presentation. We chose a durable and

inexpensive anemometer made by APRS World LLC. Pictures, calibration information and adetailed technical description of their standard anemometer can be found on their website(www.aprsworld.com). We chose a relatively inexpensive, mid-range Madgetech Pulse 110 datalogger, along with necessary cables and data aquisition software. Details for each item can befound on the Madgetech web site (www.madgetech.com). Finally, we used standard one inchsquare steel bars along with square and U bolts for attaching our anemometers to the existingtowers at each site. In addition, some sites required various lengths of 1 inch galvanized steelpoles.

C. Installation:

Installation began with obtaining permission for access to the towers located on each existing

site. Each site was then visited and photographed for use in determining the best method formounting our anemometers. We worked with four different types of towers (Enger Tower,Staircase Fire Tower, Communication Tower, and External Ladder Fire Tower). Each towertype required its own particular installation methodology. At Enger Tower we simply bolted aten foot galvanized steel pole to the railing at the top of the tower, and mounted our anemometerto the top of the steel pole (figures 2 and 3).

Figure 2. Detailed schematics of the Enger Tower mount.
http://www.aprsworld.com/http://www.aprsworld.com/http://www.aprsworld.com/http://www.madgetech.com/http://www.madgetech.com/http://www.madgetech.com/http://www.madgetech.com/http://www.aprsworld.com/


5/46

5

Figure 3. Photo of the Enger Tower mount.

Clover Valley and Finland had conventional staircase fire towers. At these sites we bolted twoshort (1 foot) sections of square steel to the tower window sills to clear the roof overhang, andextended a ten foot section of square steel from these two supports beyond the roofline. We thenattached the anemometer to the top of this extension pole (figures 4 and 5).

Figure 4. Detailed schematics of the conventional staircase fire tower mounts.


6/46

6

Figure 5. Photo of the Clover Valley Fire tower (conventional staircase fire tower) withanemometer extending vertically from the roofline.

Lutsen, Grand Marais and Grand Portage had communication towers. At these sites we bolted aten-foot section of square steel to the towers, and extended the anemometer 6-8 feet from thetowers (figures 6 and 7).

Figure 6. Detailed schematics of the communication tower mounts.


7/46

7

Figure 7. Photo of Lutsen communication tower with anemometer extending laterally.

Finally, Hovland had a unique external ladder fire tower with a hole in the roofline. At this sitewe attached a one-inch diameter, 10 foot long galvanized steel pole to the floor of the fire tower,and to two window sills using straps. The pole extended beyond the roof line, and theanemometer was attached to the top of the pole (figure 8). At each site (with the exception of thetwo stairway fire towers) wire is run down the tower to ground level where it is plugged into thedata logger. The data loggers are stored in a safe, dry container. This allows for easy groundaccess to the data loggers for monthly data collection. In every case we made every attempt tofollow the National Renewable Energy Laboratorys (NREL) guidelines for anemometerinstallation in terms of achieving the appropriate distance from rooflines and communicationtowers.


8/46

8

Figure 8. Photo of Hovland external ladder Fire Tower.

A.Data Recovery and MaintenanceApproximately once every one to two months we visited each site to collect data, and performany necessary maintenance. The data was collected in two different ways. The first approachconsists of preprogramming a data logger in the office, and then taking it to a monitoring sitewhere it is switched with the logger currently reading data. The data filled logger is thenreturned to the office, downloaded and reprogrammed to receive data at the next site. In thesecond approach, we bring a computer to the monitoring sites and download the data on site,reprogram the logger and put it back in place.

The data is downloaded onto the computer using the Madgetech software that comes with theMadgetech Pulse 110 data loggers. The desired time interval (10 minute averages arestandard) was selected. The results appear in both graphical and tabular form. This Madgetechdata is then exported into Excel. Once in Excel, we can easily conduct all our data analyses,conversions and graphical presentations as well as compute the desired format for input into theWAsP modeling program.


9/46

9

III. Results

The following graphs depict weekly average wind speed (mph) over the course of the entirestudy for each site. In addition, the average for the entire data set is indicated in the key.Weekly averages were chosen, so that we could include an entire year of data, and stilldemonstrate the typical wind speed variability in the study region. Gaps in the data are clearly

illustrated in the figures, and will be explained in the next section. Significant data loss wasonly a problem with the Finland site. The annual averages range from 14.3 mph at Enger Towerto 16.9 mph at Grand Portage. Included as a project deliverable is an excel spreadsheetcontaining daily average wind speeds for all sites over the entire course of the study (ten-minuteinterval data also available).

Figure 9. Weekly average wind speeds for Enger Tower and Clover Valley over the entirecourse of the study.

Figure 10. Weekly average wind speeds for Silver Bay and Finland over the course of the study.


10/46

10

Figure 11. Weekly average wind speeds for Lutsen and Grand Marais over the entire course ofthe study.

Figure 12. Weekly average wind speeds for Hovland and Grand Portage over the entire courseof the study.


11/46

11

IV. Discussion/ConclusionsAs we suspected, the wind speeds along the Northshore of Lake Superior are far higher than theMN state wind maps (30m) published by the MN Department of Commerce(www.mndoc.state.gov) indicate (Figure 13). According to Figure 13, our study region shouldcontain average wind speeds of 11 mph or less. We found wind speeds from 14-17 mph at

approximately 30 m. Not only do these wind speeds exceed those estimated by the MNDepartment of Commerce, but our windiest sites (16-17 mph) rival those of the well-developedBuffalo Ridge in SE MN, and the recently developed Taconite Ridge on MNs Iron Range justnorth of Virginia, MN.

Figure 13. MN State wind map at 30 m. Estimated by WindLogics and published by the MNDepartment of Commerce in January of 2006 (www.mndoc.state.gov).
http://www.mndoc.state.gov/http://www.mndoc.state.gov/http://www.mndoc.state.gov/http://www.mndoc.state.gov/


12/46

12

Finally, the results indicated in figures 9-12 suggest we had a minor problem with lost data.Data was lost at various sites throughout the study for the following reasons: 1. Data loggerbattery failure; 2. Loosened wiring connections; and 3. Data logger theft. At Enger Tower andClover Valley no data was lost. The initial installation at the Clover Valley site did not occuruntil August. As of this writing we have yet to receive the Data from Northshore mining in

silver Bay. They have been struggling with a consultant throughout the summer to put their datainto a user-friendly format for publication, and to transfer to our project. We anticipate receivingthis data soon.

Finland was our most problematic site in terms of data loss. On two separate occasionspremature battery failure occurred at the Finland site resulting in partial loss of the data betweensampling intervals. In addition, one of our data loggers was stolen from the Finland site resultingin complete loss of data for the sampling period. Finally, in the last sampling interval we found aloose wiring connection between the anemometer and data logger. As the result of this loss ofdata, and the time periods in which the data was lost we believe we significantly underestimatedthe wind speeds at the Finland site.

At the Lutsen site two short time periods of data were lost due to battery failures. No data waslost from the Grand Marais site. The initial installation at this site did not take place until Augustas well. In the final month of reported Grand Marais data the values increase over those ofLutsen for the first time in the entire sampling period. Over this time period (4/15-5/15) werecorded several 10-minute intervals with impossibly high wind speeds (some in excess of 100mph). Since these erroneous readings tended to occur at high wind speed intervals (as comparedwith other sites) we replaced the impossibly high readings (anything over 70 mph) with 40 mph.We felt this was more accurate than simply eliminating these windy intervals. However, webelieve this may have resulted in an over estimate of the Grand Marais wind speeds, and skewedthe power density calculation reported in the next section.

Finally, at the Hovland and Grand Portage sites no significant data was lost. In general, with theexception of Gand Marais, we found the wind speeds tended to increase as we sampled fromDuluth Northeast to Grand Portage. All things considered, we believe the wind resource fromFinland to Grand Portage is worthy of potential wind development. The major limiting factor fordevelopment in the region is transmission capacity.

Objective 2: Wind Resource MappingI. Methods

The process for creating a wind resource map is a complicated one involving many steps andseveral different software applications. The following is a detailed overview of the entireprocess. We began by collecting Digital Elevation Maps/Models (DEMs) of the entire regionfrom the US Geological Survey (www.seamless.usgs.gov). These DEMs are then imported intoArcMap 9.2, and converted to contour maps. The Contour maps were then split into smallersegments corresponding to our 7 monitoring sites to reduce file size and calculation time.Arcv2cad File Converter was then used to convert these contour maps into WAsP 9 readabledxf.files. WAsP has emerged as the industry standard wind development software (RISO website). It is a PC-program that can be used for a wide variety of wind development applications
http://www.seamless.usgs.gov/http://www.seamless.usgs.gov/


13/46

13

including the extrapolation of the wind resource from a particular monitoring station over a largeregion.

Once the contour maps were in WAsP surface roughness values were added/digitized onto theDEMs using WAsPs Map Editor function. Land use maps from

(www.nrri.umn.edu/coastalGIS were used as the template for assigning/digitizing surfaceroughness values. We used four surface roughness categories and chose the following surfaceroughness values: Water (.0001); Swamp (.1); Forest (.9) and City (1.0) (Gipe, 2004, pg. 41).

The 10-miute interval wind speed and direction data was exported from an Excel Spreadsheetinto a Notepad text file. This text file was then imported into WAsP. WAsP then uses theelevation data from the contour maps, the digitized surface roughness values along with the windspeed and direction data to extrapolate the wind resource across the region from a particularmonitoring location.

The regional wind resource maps (one for each monitoring site) generated by WAsP are then

exported into ArcMap 9.2. Once in ArcMap 9.2 the Ascii files are converted into raster files toallow for the re-creation of visual maps and overlaying. The individual maps were piecedtogether to create a single seamless map of the entire study area. This was done using theMosiac Data Management Tool in ArcMap 9.2. This tool allows the user to average the windresource estimate from those areas where the individual resource maps overlapped. Therefore,the final wind resource estimate is calculated from a single monitoring site in areas close to site-specific monitoring stations, and from the average of two monitoring stations in areas betweenmonitoring sites.

Finally, layers (roads, waterways, cities/towns, transmission lines) were added to the windresource map using data from UMDs Geographic Information Systems Laboratory (roads andlakes), MN DNR Deli (state boundary, cities and Lake Superior), Land Management InformationCenter (transmission lines). We did not do a bird migration route overlay due to a lack of currentdata. However, in response to this lack of regional data, the DNR Coastal Zone is currentlyfunding a detailed migratory bird density analysis for the region, and we will produce an overlayusing this data upon conclusion of the study.

II. ResultsFigure 14 illustrates the average annual wind speed for the entire study region (100 meterresolution) as calculated by WAsP. Also included electronically as a deliverable is the actualArcMap 9.2 wind resource map. ArcMap users can use the map to manipulate layers, zoomin/out on particular areas of interest, and to pull actual wind speed estimates from any point onthe regional map. Any region on the map colored orange or red would likely contain aneconomically viable wind development project given adequate transmission capacity. This windresource map is the first of its kind at this resolution in our region, and will likely serve as thestarting point for any wind development project in NE Minnesota. Again, the average annualwind speeds measured and estimated using WAsP are far greater than the 10-12 mph windspeeds indicated in Figure 13.
http://www.nrri.umn.edu/coastalGIShttp://www.nrri.umn.edu/coastalGISwereusedasthetemplateforassigning/digitizingsurfaceroughnessvalues.Weusedfoursurfaceroughnesscategoriesandchosethefollowingsurfaceroughnessvalues:Water(.0001http://www.nrri.umn.edu/coastalGISwereusedasthetemplateforassigning/digitizingsurfaceroughnessvalues.Weusedfoursurfaceroughnesscategoriesandchosethefollowingsurfaceroughnessvalues:Water(.0001http://www.nrri.umn.edu/coastalGISwereusedasthetemplateforassigning/digitizingsurfaceroughnessvalues.Weusedfoursurfaceroughnesscategoriesandchosethefollowingsurfaceroughnessvalues:Water(.0001http://www.nrri.umn.edu/coastalGISwereusedasthetemplateforassigning/digitizingsurfaceroughnessvalues.Weusedfoursurfaceroughnesscategoriesandchosethefollowingsurfaceroughnessvalues:Water(.0001http://www.nrri.umn.edu/coastalGISwereusedasthetemplateforassigning/digitizingsurfaceroughnessvalues.Weusedfoursurfaceroughnesscategoriesandchosethefollowingsurfaceroughnessvalues:Water(.0001http://www.nrri.umn.edu/coastalGISwereusedasthetemplateforassigning/digitizingsurfaceroughnessvalues.Weusedfoursurfaceroughnesscategoriesandchosethefollowingsurfaceroughnessvalues:Water(.0001http://www.nrri.umn.edu/coastalGISwereusedasthetemplateforassigning/digitizingsurfaceroughnessvalues.Weusedfoursurfaceroughnesscategoriesandchosethefollowingsurfaceroughnessvalues:Water(.0001http://www.nrri.umn.edu/coastalGIS


14/46

14

Figure 14. Wind resource map of entire study region at 100 meter resolution and 30 meterheight. Data from the CSCD, Seamless USGS, LMIC and the MN DNR.

III. Discussion/ConclusionsOverall the WAsP calculations appear to have accurately estimated the wind resource across theregion, but there are some peculiarities worthy of explanation. In some cases, there is a sharpcontrast along the junction between two individual resource maps (i.e. between Finland andLutsen) despite the averaging using the Mosiac Data Management Tool. This is the result of

significantly different wind speed measurements at neighboring monitoring stations incombination with the resolution of our color scale. Increasing the resolution fixes this boundaryproblem, but makes the overall map less detailed, hence less useful.

Another issue that is more difficult to explain is the relatively high values estimated in theFinland region of the wind resource map. The average wind speed measured at the Finland


15/46

15

Monitoring site was relatively high, but less than Lutsen, Hovland and Grand Portage. We haveno reason to believe the lack of data from the Finland site has resulted in an overestimate of itswind resource. In fact, we believe given the period of data that was lost, we are more likely tohave underestimated the Finland wind resource. Therefore, given the moderate wind resourcemeasured at the Finland site, one would not expect the regional wind resource estimated from the

Finland monitoring site to be so high relative to the others. We believe this is because WAsP hasa tendency to assign the highest wind resource values to the prominent peaks located close toLake Superior. Most of our monitoring sites occupy these peaks, so WAsP is extrapolatingdown from them as you head inland. However, the Finland monitoring site is further inland.Therefore, WAsP is actually extrapolating up from the measured value to assign high values tothe prominent peaks closer to Lake Superior, and pulling up the inland values as well. The sameis true, but to a lesser extent in the Clover Valley region of the map.

Objective 3: Economic Feasibility

The wind monitoring data was used to estimate the economic feasibility of wind resourcedevelopment at each monitoring site. In addition, an economic I/O model was used to estimatethe county level economic development potential of wind resource development.

A.Economic FeasibilityI. Methods

There are several different common approaches to conducting this sort of economic feasibilityanalysis. We took a unique approach using a STELLA computer simulation model because of itsvalue as a community educational tool, and its ability to incorporate the uncertainty associatedwith key system parameters. We began by exporting our 10-minute interval wind speed data inexcel. These 10-minute averages were then exported from excel into a text file, and then inputedinto the WAsP wind modelling software package. WAsP then bins these 10-minute intervalaverage wind speeds into 1 m/s wind speed bins, and sums the number of 10-minute intervals(graphed as relative frequency) in each wind speed bin. The result is the statistical distributionillustrated in figure 15. WAsP then calculates the wiebel-k constant based on the shape of thisdistribution. The more skewed the distribution to the right (higher wind speed bins) the smallerthe weibel-k constant, and the greater the power density. Finally, WAsP calculates the powerdensity (w/m2) for each bin and sums them to arrive at the total power density found in the windfrom the particular monitoring site. This total power density value was then entered in theSTELLA model. This process was repeated for each monitoring site.


16/46

16

Figure 15. The wind speed distribution, k-value and power density from the Lutsen monitoringsite.

The STELLA model developed to estimate economic feasibility is illustrated in figure 16. Themathematical equations governing the relationships between the key parameters can be seen inthe working model included as a project deliverable. The model was designed to calculate andcompare the annual revenues associated with a given wind development project and theassociated annual costs. To calculate annual revenues the power density (PD) value (w/m2)computed by WAsP is adjusted for the difference between the anemometer monitoring heightand future turbine hub height as well as the wind shear exponent associated with the monitoringsites surface roughness (adj PD). The adjusted PD is then multiplied by the proposed windturbines swept area which is calculated from its turbine diameter along with the proposednumber of turbines. The result is the total turbine output in watts. This output is then convertedto kilowatt hours (kwh) per year by dividing by 1000 and multiplying by the number of hours ina year and the turbines rated efficiency. The turbines annual output (kwh/year) is then multipliedby the electricity price (cents/kwh) to arrive at annual revenue. In addition, because the price perkw paid by the community partner is in many cases equal to the price they can expect from theelectricity their turbine generates the annual revenue value can be expected to increase by anannual percentage as the price of electricity increases. The model multiplies annual revenue byan annual electricity rate inflation value to achieve this effect.


17/46

17

installation costs

PD

adj PD

wind shear exp

hub height

anemometer height

turbine diameter

swept area

turbine outputKwh annual output

turbine eff iciency

hours per y ear built in

rate inf lation

total wind sys tem cost

~ turbine capacity kw

annual cost

interest rate

loan amount

cents per kwh

equipment lifet ime

annual project costs

annual rev enue

cost per kw

Figure 16. Diagram of the STELLA model used to calculate and compare annual wind systemcosts and revenues for community economic feasibility.

The annual costs associated with the wind generation system chosen above are simultaneouslycalculated. The turbine diameter is linked to turbine capacity in kw by a graphical function thatrelates these two values. Total wind system cost is then calculated by multiplying the turbinecapacity in kw by an average cost per kw of cpacity and the number of turbines. This totalsystem cost embodies the complete cost of the turbines and their towers. Added to this cost to

get the total loan amount is the total installation costs (consulting fees, site prep, installation,legal, permitting, interconnect, MISO etc). These installation costs typically varyconsiderably, and we attempted to estimate these costs as a percentage of total wind system costsfor each site given their unique characteristics. The cost data used in our model comes from asurvey of many installed projects (Gipe 2004), and Windustries Community Wind Toolbox(www.windustry.org). In addition, the annual operation costs are calculated based on the totalsystem cost and added to the annual payment on a loan for the total annual wind system cost.The term of the annual loan is equal to or less than the expected equipment lifetime. The annualcost for the wind system is then a function of the total system cost, the interest rate and the termof the loan. Finally the annual revenues are compared with the annual costs to determineeconomic feasibility given the sites wind resource.

The problem with conducting economic feasibility studies for wind development in this manneris that just about every variable mentioned above changes from site to site and over the course ofthe typical 20 year lifetime of any given project. This is the reason for using STELLA. It allowsthe user to adjust all the key model parameters (interest rates, system costs, surface roughnessetc) to determine their individual effects on overall economic feasibility. If this type of modelexperimentation is conducted in front of a community interested in the potential for local wind
http://www.windustry.org/http://www.windustry.org/


18/46

18

development it becomes not only a very useful planning tool, but a great educational tool as well.The typical output from the STELLA model and all the key adjustable variables are shown infigures 17 and 18.

10:45 AM Tue, Jul 22, 2008

Untitled

Page 1

1.00 5.75 10.50 15.25 20.00

Years

1:

1:

1:

2:

2:

2:

200000

400000

600000

1: annual revenue 2: annual cost

1

1

1

1

2 2 2 2

Figure 17. Graphical output comparing annual costs to annual revenues using data from theGrand Portage Site, and a positive value (2%) for utility rate inflation.


19/46

19

60.400

10.000 150.000

?U

hub height

30.000

10.000 150.000

anemometer height

0.23500

0.10000 0.40000

?U

turbine eff iciency

0.06200

0.02000 0.22000

U

cents per kwh

75.250

1.000 100.000

?U

turbine diameter

0.21000

0.10000 0.40000

?

wind shear exp

0.06040

0.01000 0.10000

U

interest rate

0.0200

0.0000 1.0000

U

rate inf lation

0.25000

0.10000 0.30000

?U

installation costs

1750.00

1500.00 2000.00

?U

cost per k

Figure 18. The key model parameters (variables) included in our community specificsimulations, and their values for the Grand Portage Simulation.

II. Results

Table 2 indicates the average annual wind speed (m/s), weibel-k constant, resulting powerdensity (w/m2) and the estimated installation costs (as a percentage of total wind system costs)for each monitoring site. Windustrys Community Wind Toolbox (www.windustry.org)estimates installation costs are typically 10-30% of total project costs. Installation costs dependprimarily on available infrastructure at the site (roads and foundation) and the distance toavailable transmission lines or point of consistent electricity use (i.e. Grand Portage Casino). Weassigned installation cost percentages for each site based on these considerations. These valuesare reported in table 2 and used in the site-specific simulation runs described in the next section.


20/46

20

Table 2. Average annual wind speed (m/s), weibel-k constant, resulting power density (w/m2)and the estimated installation costs (as a percentage of total wind system costs) for eachmonitoring site.

Site avg (m/s) weibel-k PD (w/m2) install cost (%)Enger 6.31 1.77 335 14Clover Valley 6.58 2.29 295 21Silver BayFinland 6.86 2.3 333 18Lutsen 7.03 1.94 419 16Grand Marais 6.14 1.32 470 20Hovland 6.99 1.99 401 28Grand Portage 7.44 2.03 475 26

Table 3 indicates the annual revenues, annual costs and the resulting net revenues for each siteassuming no utility inflation rate. Several variables were held constant for each simulation todirectly compare the net revenues from each site based on their unique wind climate and site-specific estimated installation costs. The variables held constant included: 1. 1.5 MW windturbine with a 76 m blade diameter; 2. Hub height twice as high as the anemometer height; 3.Wind Shear Exponent of .21; 4. 20 year equipment lifetime; 5. Loan interest rate of 6%; 6.Electricity cost of 6.2 cents per kwh; 7. Cost per kw of installed capacity of $1,750; and 8.Turbine efficiency of 23.5%. These typical cost estimates (model parameters) were obtainedfrom (Gipe, 2004), various wind turbine manufacturers, community partner electricity bills andWindustrys Community Wind Toolbox (www.windustry.org). The wind sheer exponent wascalculated using data from anemometers installed at two different heights at the Lutsen andGrand Portage sites. In addition to the exemplary static results reported in this section, the entiredynamic STELLA model is included electronically as a project deliverable.

Table 3 Annual revenues, annual costs and the resulting net revenues for each site assuming noutility inflation rate.

Site Annual Revenues Annual Costs Net RevenuesEnger Tower $295,364 $308,864 -$13,499Clover Valley $260,096 $327,829 -$67,732Silver BayFinland $293,600 $319,701 -$26,000Lutsen $369,425 $314,282 $55,143Grand Marais $414,391 $325,120 $89,271Hovland $353,555 $346,794 $6,760Grand Portage $418,800 $341,376 $77,424

IV. Discussion/Conclusions


21/46

21

The results of the economic feasibility calculations largely parallel the average annual windspeed and resulting power density from each monitoring site. This was expected given thepower embodied in the wind is a cubic function of the average annual wind speed. Therefore a 2mph increase in average wind speed results in 8x the electricity generated from a given windturbine. Our simple comparative analysis suggested the region Northeast of Finland has a strong

enough wind resource to make even a small (1MW) wind development project economicallyfeasible. Project economics improve with larger scale development.

Our electricity production estimates (revenues) are the most accurate available given the windspeed data gathered during this study and regional electricity pricing. However, our comparativecost estimates are more general. As previously described in the report, they are based on typicalvalues given the geology, transmission capacity, road availability and sources of nearbyelectricity consumption. More detailed site-specific cost estimates are possible, but beyond thescope of this project.

The data gathered over the course of this study, and its integration using our modeling methods

not only indicate the possibility for economically feasible wind development projects in NEMinnesota, but they can be used to facilitate the community decision making process. As theresult of this study, each community in our study region now has a solid estimate of their windresource, and the resulting project economics. The STELLA Model can now be used to facilitatemore detailed site-specific project exploration and the corresponding community decisionmaking process regarding future wind development.

B. County-wide Economic Impacts (see appendix A for detailed description ofmethods,results and discussion)

References

Gipe, Paul. 2004. Wind Power:Renwable Energy for Home Farm and Business. Chelsea GreenPublishing Company. White River Junction, Vermont.

Minnesota Department of Commerce. 2006. MN Wind Resource Map. Atwww.state.MN.gov.

Windustrys Community Wind Toolbox. 2008. Atwww.windustry.org.

Appendix A County-wide Economic Impacts
http://www.state.mn.gov/http://www.state.mn.gov/http://www.state.mn.gov/http://www.windustry.org/http://www.windustry.org/http://www.windustry.org/http://www.windustry.org/http://www.state.mn.gov/


22/46

22

The Economic Impact of Adding Wind EnergyIn Communities on Minnesotas

North Shore of Lake Superior, 2007

April 30, 2008

For the

Minnesota Department of Natural Resources

and the University of Minnesota Duluth

Center for Sustainable Community Development


23/46

i

Prepared by:

UMD Labovitz School of Business and EconomicsBureau of Business and Economic Research

James A. Skurla, Acting DirectorJean Jacobson, Senior EditorSyed Jafri, Undergraduate Research AssistantTaha Kasim, Undergraduate Research AssistantVickie Almquist-Minko, Executive Administrative SpecialistBureau of Business and Economic Research19 School of Business and EconomicsUniversity of Minnesota DuluthDuluth, MN 55812-2496 (218) 726-8614

Start date:

July 2007

Contact:

Michael T. MageauUMD Environmental Studies329 CinaH1123 University DrDuluth, MN [email protected]


24/46

ii

Table of Contents

Executive Summary ...................................................................................................................................... iii

PROJECT DESCRIPTION .................................................................................................................................. 6

Deliverables............................................................................................................................................... 6

Study Area ................................................................................................................................................. 6

METHODOLOGY ............................................................................................................................................ 7

INPUTS TO THE MODEL ................................................................................................................................. 9

Assumptions: ............................................................................................................................................. 9

IMPACT FINDINGS ....................................................................................................................................... 11

1) The Three-County Region ................................................................................................................... 11

2) St. Louis County, Minnesota ............................................................................................................... 14

3) Lake County, Minnesota ..................................................................................................................... 16

4) Cook County, Minnesota .................................................................................................................... 18REFERENCES ................................................................................................................................................ 20

APPENDIX .................................................................................................................................................... 21


25/46

iii

Executive Summary

These summary tables present the total economic impact from wind-generated energy as tenpercent and twenty percent of total energy supply in the combined three-county region, as well asCook, Lake and St. Louis Counties. Impacts are projected for 2010, 2015, and 2020, usingpopulation and energy consumption to estimate inputs for these projections. Sources for all tablesare IMPLAN and the UMD Labovitz Schools Bureau of Business and Economic Research. Thecalculations in these models use 2008 dollars for reports, without adjustments for inflation.Readers should also be mindful of the diminishing reliability of modeling an economy asprojections proceed into the future; therefore the values projected for 2010 will warrant mostscrutiny.

Summary: Economic Impact of Wind-Generated Energy as

10% and 20% of Total Energy Supply, in Cook, Lake andSt. Louis County, Minnesota, Estimated for Years 2010,

2015, and 2020 in 2008 dollars10% of Totalsupply 2010 2015 2020

Output $24,334,673 $24,519,171 $24,727,311

Employment 79 79 79

Value Added $18,345,355 $18,484,443 $18,641,354

Tax $4,971,737 $5,009,430 $5,051,95520% of Totalsupply

Output $48,669,346 $49,038,342 $49,454,622

Employment 156 157 158

Value Added $36,690,710 $36,968,886 $37,282,708

Tax $9,943,474 $10,018,860 $10,103,910

Summary: Economic Impact of Wind-Generated Energy as10% and 20% of Total Energy Supply, in St. Louis County,

Minnesota, Estimated for Years 2010, 2015, and 2020 in2008 dollars

10% of Total

supply 2010 2015 2020Output $22,404,173 $22,530,897 $22,684,776

Employment 72 72 72

Value Added $16,894,205 $16,989,765 $17,105,800


26/46

iv

Tax $4,571,894 $4,597,754 $4,629,15520% of Totalsupply

Output $44,808,346 $45,061,794 $45,369,552

Employment 143 144 145Value Added $33,788,410 $33,979,530 $34,211,600

Tax $9,143,788 $9,195,508 $9,258,310

Summary: Economic Impact of Wind-Generated Energyas 10% and 20% of Total Energy Supply, in Lake County,


10% of Total supply 2010 2015 2020

Output $1,118,438 $1,146,691 $1,168,124

Employment 3 3 3

Value Added $883,253 $905,565 $922,491

Tax $244,207 $250,376 $255,056

20% of Total supply

Output $2,236,875 $2,293,382 $2,336,249

Employment 7 7 7

Value Added $1,766,505 $1,811,130 $1,844,983

Tax $488,414 $500,752 $510,112

Summary: Economic Impact of Wind-Generated Energy as10% and 20% of Total Energy Supply, in Cook County,


10% of total supply 2010 2015 2020

Output $527,269 $548,142 $572,706

Employment 2 2 2

Value Added $420,776 $437,433 $457,037

Tax $107,076 $111,315 $116,304

20% of total supply

Output $1,054,537 $1,096,284 $1,145,413

Employment 3 3 3

Value Added $841,552 $874,867 $914,073Tax $214,153 $222,630 $232,607

Because projected demographic growth and therefore energy demand is expected to be small,employment remains almost constant in this sector. Output measures increase, however, and


27/46

v

reflect available support for wind energy. The scale of these impacts is in proportion to the threecounties, reflecting both the sizable St. Louis County and the much smaller Lake and CookCounty demographics. Readers are reminded that only residential (not commercial andindustrial) consumption is modeled here. The contribution by wind energy to total energy

consumption is modeled as ten and twenty percent of total consumptionthese should beconsidered conservative estimates of possible conversion from other sources of energy andgrowth in demand. The Bureau of Business and Economic Research believes that the outputfrom these estimated wind energy contributions would be sufficient to support a wind energyproject in these counties. Finally, it is important to note that energy consumption in the countiesof the study area could be dramatically increased by a group of industrial and commercialprojects currently being proposed for the Iron Range of Northeastern Minnesota.1

1 For instance The Duluth News Tribune reported an estimate for energy use from the MesabiNugget project : The process is energy intensive. Mesabi Nugget expects to buy 25 megawatts

of electricity from Minnesota Power when it begins production, and that consumption couldgrow to 75 megawatts if the plant is built out to 1.5 million metric tons. See Mesabi Nuggetplant is on target by Peter Pass, in the Duluth News Tribune - 05/11/2008.


28/46

6

The Economic Impact of Adding Wind EnergyIn Communities on Minnesotas

North Shore of Lake Superior, 2007

PROJECT DESCRIPTION

This project is a sub-contract in a larger proposal funded by the Minnesota Department ofNatural Resources through the Coastal Zone Program, and University of MinnesotaNortheast Region Sustainable Development Partnership, as well as a CLA TechnicalGrant. The UMD Labovitz School of Business and Economics research bureau wasasked to deliver estimates of economic impact numbers for conversion wind energy.This project contributes to a feasibility study for several communities consideringinvestment in wind power on the North Shore of Lake Superior. Estimates are based onoperations in MW hours of electricity from wind power, projected from data gathered in

2007 and 2008 by Michael T. Mageau and the UMD Center for Sustainable CommunityDevelopment. The economic impact of construction is not provided here.

Deliverables

Economic impact tables for all measures and all effects in the IMPLAN model for a studyregion comprised of three counties in Northeast Minnesota.

Study Area

Figure 1. St. Louis, Lake, and Cook Counties, in Northeastern Minnesota


29/46

7

METHODOLOGY

There are two components to the IMPLAN system, the software and databases. Thedatabases provide all information to create regional IMPLAN models. The software

performs the calculations and provides an interface for the user to make final demandchanges. Comprehensive and detailed data coverage of the IMPLAN study areas bycounty, and the ability to incorporate user-supplied data at each stage of the modelbuilding process, provides a high degree of flexibility both in terms of geographiccoverage and model formulation, in this case definition of the region study areas, and thedefinition of specific models for construction and operations.

Data

IMPLAN data files use federal government data sources including: US Bureau of Economic Analysis Benchmark I/O Accounts of the US US Bureau of Economic Analysis Output Estimates US Bureau of Economic Analysis REIS Program US Bureau of Labor Statistics County Employment and Wages (CEW) Program US Bureau of Labor Statistics Consumer Expenditure Survey US Census Bureau County Business Patterns US Census Bureau Decennial Census and Population Surveys US Census Bureau Economic Censuses and Surveys US Department of Agriculture Crop and Livestock Statistics

IMPLAN data files consist of the following components: employment, industry output,value added, institutional demands, national structural matrices and inter-institutional

transfers.Impacts for models in these analyses use the most recent IMPLAN data available whichis for the year 2006. Economic impacts are made up of direct, indirect, and inducedimpacts. The following cautions are suggested assumptions for accepting the impactmodel:

IMPLAN input-output is a production based model Local or export based purchases that represent transfers from other potential local

purchases are not counted. The numbers (from U.S. Department of Commerce secondary data) treat both full

and part time individuals as being employed.

Assumptions need to be made concerning the nature of the local economy beforeimpacts can be interpreted.

The IMPLAN model was constructed for the year 2006 (most recent dataavailable).


30/46

8

Model

The impact measures and effects for the following tables use these definitions:

Measures

Gross Output represents the value of local production required to sustainactivities.

Value Added is a measure of the impacting industrys contribution to thelocal community; it includes wages, rents, interest and profits.

Employment estimates are in terms of jobs, not in terms of full-timeequivalent employees. Hence, these may be temporary, part-time or short-term jobs.

Effects

Direct Initial new spending in the study area resulting from the project

Indirect The additional inter-industry spending from the direct impact

Induced The impact of additional household expenditure resulting fromthe direct and indirect impact.

The input-output modeling tool IMPLAN can provide a general estimate of the possibleeconomic impact of . . .

Figure 2. Basic information from the IMPLAN software model for the Counties of theNorth Shore.


31/46


32/46

10

consumption was deducted from total consumption to arrive at an approximate figure fortotal residential consumption.

Estimation of 10% and 20% Residential Consumption fromPopulation Projections for Three Counties, 2010, 2015, and 2020

MWh/person xpopulation 2010 2015 2020

Three counties 2,559,214 2,578,611 2,600,507

St. Louis 2,356,319 2,369,647 2,385,831

Lake 136,612 140,063 142,681

Cook 66,283 68,901 71,995

Total consumption (in2008 dollars) 2010 2015 2020

Three counties $179,144,980 $180,503,190 $182,035,490

St. Louis $164,942,330 $165,875,290 $167,008,170Lake $9,562,840 $9,804,410 $9,987,670

Cook $4,639,810 $4,823,490 $5,039,650

10% of total consumption(in 2008 dollars) 2010 2015 2020

Three counties $17,914,498 $18,050,319 $18,203,549

St. Louis $16,494,233 $16,587,529 $16,700,817

Lake $956,284 $980,441 $998,767

Cook $463,981 $482,349 $503,965

20% of total consumption(in 2008 dollars) 2010 2015 2020

Three counties $35,828,996 $36,100,638 $36,407,098

St. Louis $32,988,466 $33,175,058 $33,401,634

Lake $1,912,568 $1,960,882 $1,997,534

Cook $927,962 $964,698 $1,007,930Source: MN State Demographer; UMD Labovitz School BBER

Reported as an average cost in 2008 for the study area, the approximate cost of a KWh isseven cents. (See for instance Minnesota Power total production divided by dollars.

Mageau et al. also corroborated these calculations and recommend $.07/KWh.) Based onthis assumption BBER estimated projections of direct inputs of wind-powered supply, in2008 dollars. The above table shows the value of 10% and 20% consumption of totalconsumption for the three counties.


33/46

11

IMPACT FINDINGS

Tables showing the economic impacts of wind energy as 10% and 20% of the totalsupply of energy in the three-county region are presented below. The combined countiesand the individual county detail are modeled. The impact is measured in terms of the

value of local production, as well as in terms of wages and jobs. The impacts areprojected for the years 2010, 2015, and 2020, based on estimated demand from projectedpopulation and energy consumption, as noted in the previous section.

1) The Three-County Region

10% of total energy:

Summary of Economic Impacts of 10% Wind-GeneratedEnergy in St. Louis, Lake, and Cook Counties of

Minnesota, in the Year 2010, 2015 and 2020 in 2008 dollars

2010 2015 2020

Output $24,334,673 $24,519,171 $24,727,311

Employment 79 79 79

Value Added $18,345,355 $18,484,443 $18,641,354

Tax $4,971,737 $5,009,430 $5,051,955

Again, these measures are defined as follows: Gross Output represents the value of localproduction required to sustain activities. Value Added is a measure of the impactingindustrys contribution to the local community; it includes wages, rents, interest and

profits. Employment estimates are in terms of jobs, not in terms of full-time equivalentemployees. Hence, these may be temporary, part-time or even short-term jobs. Taximpacts are reported here as the total of State, Federal and Corporation taxes.

The economic impact, as it moves through the economy, is modeled as the followingeffects:The direct effect is initial new spending in the study area resulting from the project. Theindirect effect is the additional inter-industry spending from the direct impact. TheInduced effect is the impact of additional household expenditure resulting from the directand indirect impact.


34/46

12

Economic Impact of 10% Wind-Generated Energy in St. Louis, Lake, and CookCounties of Minnesota, in the Year 2010 in 2008 dollars

2010 Direct Indirect Induced Total Multiplier

Output $18,797,148 $2,651,665 $2,885,860 $24,519,171 1.3

Employment 37 12 30 79 2.1

Value Added $15,280,056 $1,338,426 $1,726,873 $18,345,355 1.2

Tax $4,971,737

In the detail tables we show the economic multipliers from the IMPLAN model. Theseratios can be interpreted as follows: The output multiplier: for every dollar of output inthe reference industry, how many dollars are generated in the region. The employmentmultiplier: for every employee in the reference industry, how many employees aregenerated in the region? The value added multiplier: for every dollar of value addedearned from the operations of the reference industry, how many dollars of value addedare generated in the region.

There are actually several layers that can be analyzed through multipliers; the most basicuses a multiplier that takes only the interaction between local industries into account.Interaction between industries is measured as purchases and sales from and to oneanother so that further production can take place.



Output $18,939,662 $2,671,769 $2,907,740 $24,519,171 1.3Employment 37 12 30 79 2.1ValueAdded $15,395,904 $1,348,573 $1,739,966 $18,484,443 1.2

Tax $5,009,430



Output $19,100,440 $2,694,448 $2,932,424 $24,727,311 1.3

Employment 37 12 30 79 2.1Value Added $15,526,599 $1,360,020 $1,754,736 $18,641,354 1.2

Tax $5,051,955

The following tables for further percentages in the three-county region and the individualthree counties can be interpreted in the same way for impact measures (output,


35/46

13

employment, value added, and taxes) and effects (direct, indirect, induced, and totaleffects).

20% of total energy:Summary of Economic Impacts of 20% Wind-

Generated Energy in St. Louis, Lake, and CookCounties of Minnesota, in the Year 2010, 2015 and 2020in 2008 dollars

2010 2015 2020

Output $48,669,346 $49,038,342 $49,454,622


Value Added $36,690,710 $36,968,886 $37,282,708

Tax $9,943,474 $10,018,860 $10,103,910

Economic Impact of 20% Wind-Generated Energy in St. Louis, Lake, and Cook

Counties of Minnesota, in the Year 2010 in 2008 dollars2010 Direct Indirect Induced Total Multiplier

Output $37,594,296 $5,303,330 $5,771,720 $48,669,346 1.3

Employment 74 23 59 156 2.1

Value Added $30,560,112 $2,676,852 $3,453,746 $36,690,710 1.2

Tax $9,943,474



Output $37,879,324 $5,343,538 $5,815,480 $49,038,342 1.3

Employment 74 23 60 157 2.1

Value Added $30,791,808 $2,697,146 $3,479,932 $36,968,886 1.2

Tax $10,018,860



Output $38,200,880 $5,388,896 $5,864,848 $49,454,622 1.3

Employment 74 24 60 158 2.1

Value Added $31,053,198 $2,720,040 $3,509,472 $37,282,708 1.2Tax $10,103,910


36/46

14

2) St. Louis County, Minnesota

10% of total energy:Summary of Economic Impact of 10% Wind-

Generated Energy in St. Louis County of Minnesota,

in the Year 2010, 2015 and 2020 in 2008 dollars2010 2015 2020

Output $22,404,173 $22,530,897 $22,684,776

Employment 72 72 72Value

Added $16,894,205 $16,989,765 $17,105,800

Tax $4,571,894 $4,597,754 $4,629,155

Economic Impact of 10% Wind-Generated Energy in St. Louis County ofMinnesota, in the Year 2010 in 2008 dollars


Output $17,306,906 $2,455,815 $2,641,450 $22,404,173 1.3

Employment 34 11 27 72 2.1

Value Added $14,068,649 $1,241,991 $1,583,566 $16,894,205 1.2

Tax $4,571,894



Output $17,404,800 $2,469,706 $2,656,391 $22,530,897 1.3

Employment 34 11 27 72 2.1

Value Added $14,148,225 $1,249,016 $1,592,523 16,989,765 1.2

Tax $4,597,754



Output $17,523,668 $2,486,574 $2,674,533 $22,684,776 1.3

Employment 34 11 27 72 2.1

Value Added $14,244,854 $1,257,547 $1,603,400 $17,105,800 1.2

Tax $4,629,155


37/46

15


Summary of Economic Impact of 20% Wind-Generated Energy in St. Louis County of Minnesota,

in the Year 2010, 2015 and 2020 in 2008 dollars2010 2015 2020Output $44,808,346 $45,061,794 $45,369,552


ValueAdded

$33,788,410 $33,979,530 $34,211,600

Tax $9,143,788 $9,195,508 $9,258,310



Output $34,613,812 $4,911,630 $5,282,900 $44,808,346 1.3Employment 67 22 54 143 2.1

Value Added $28,137,298 $2,483,982 $3,167,132 $33,788,410 1.2

Tax $9,143,788



Output $34,809,600 $4,939,412 $5,312,782 $45,061,794 1.3

Employment 68 22 54 144 2.1

Value Added $28,296,450 $2,498,032 $3,185,046 $33,979,530 1.2

Tax $9,195,508



Output $35,047,336 $4,973,148 $5,349,066 $45,369,552 1.3

Employment 68 22 55 145 2.1

Value Added $28,489,708 $2,515,094 $3,206,800 $34,211,600 1.2

Tax $9,258,310


38/46

16

3) Lake County, Minnesota


Economic Impact of 10% Wind-Generated Energy inLake County of Minnesota, in the Year 2010, 2015 and

2020 in 2008 dollars

2010 2015 2020

Output $1,118,438 $1,146,691 $1,168,124


Added $883,253 $905,565 $922,491

Tax $244,207 $250,376 $255,056

Economic Impact of 10% Wind-Generated Energy in Lake County of Minnesota,in the Year 2010 in 2008 dollars


Output $1,003,400 $26,128 $88,910 $1,118,438 1.1

Employment 2 0 1 3 1.5

Value Added $815,656 $14,095 $53,502 $883,253 1.1

Tax $244,207



Output $1,028,748 $26,788 $91,156 $1,146,691 1.1


Value Added $836,261 $14,451 $54,853 $905,565 1.1

Tax $250,376



Output $1,047,976 $27,288 $92,859 $1,168,124 1.1

Employment 2 0 1 3 1.5Value Added $851,892 $14,721 $55,879 $922,491 1.1

Tax $255,056


39/46

17


Economic Impact of 20% Wind-Generated Energy in

Lake County of Minnesota, in the Year 2010, 2015 and2020 in 2008 dollars

2010 2015 2020

Output $2,236,875 $2,293,382 $2,336,249


Added $1,766,505 $1,811,130 $1,844,983

Tax $488,414 $500,752 $510,112


2010 Direct Indirect Induced Total MultiplierOutput $2,006,801 $52,255 $177,819 $2,236,875 1.1


Value Added $1,631,313 $28,189 $107,003 $1,766,505 1.1

Tax $488,414



Output $2,057,496 $53,576 $182,312 $2,293,382 1.1

Employment 4 1 2 7 1.8Value Added $1,672,522 $28,902 $109,706 $1,811,130 1.1

Tax $500,752



Output $2,095,953 $54,577 $185,719 $2,336,249 1.1


Value Added $1,703,784 $29,442 $111,757 $1,844,983 1.1

Tax $510,112


40/46

18

4) Cook County, Minnesota


Economic Impact of 10% Wind-Generated Energyin Cook County of Minnesota, in the Year 2010,

2015 and 2020 in 2008 dollars

2010 2015 2020

Output $527,269 $548,142 $572,706


Added $420,776 $437,433 $457,037

Tax $107,076 $111,315 $116,304

Economic Impact of 10% Wind-Generated Energy in Cook County ofMinnesota, in the Year 2010 in 2008 dollars


Output $486,841 $8,285 $32,142 $527,269 1.1

Employment 1 0 1 2 2.0ValueAdded $395,750 $4,290 $20,736 $420,776 1.1

Tax $107,076

Economic Impact of 10% Wind-Generated Energy in Cook County of

Minnesota, in the Year 2015 in 2008 dollars2015 Direct Indirect Induced Total Multiplier

Output $506,114 $8,613 $33,415 $548,142 1.1


Tax $111,315



Output $528,795 $8,999 $34,912 $572,706 1.1


Tax $116,304


41/46

19


Economic Impact of 20% Wind-Generated Energy

in Cook County of Minnesota, in the Year 2010,2015 and 2020 in 2008 dollars

2010 2015 2020

Output $1,054,537 $1,096,284 $1,145,413


Added $841,552 $874,867 $914,073

Tax $214,153 $222,630 $232,607



Output $973,683 $16,569 $64,285 $1,054,537 1.1


Tax $214,153



Output $1,012,229 $17,225 $66,830 $1,096,284 1.1Employment 2 0 1 3 1.5ValueAdded $822,833 $8,920 $43,114 $874,867 1.1

Tax $222,630



Output $1,057,591 $17,997 $69,825 $1,145,413 1.1


ValueAdded $859,707 $9,320 $45,046 $914,073 1.1

Tax $232,607


42/46

20

REFERENCES

Minnesota IMPLAN Group, Inc., IMPLAN System (data and software), 1725 TowerDrive West, Suite 140, Stillwater, MN 55082www.implan.com

Minnesota Department of Commerce, A Reference guide to Minnesota electric andnatural-gas utilities 1965-2002.

Minnesota Department of Commerce, Minnesota Utility Data Book, A reference guide

to natural gas and electric utilities for the years 1965 through 2005, athttp://www.state.mn.us/.

Minnesota State Demographers Population Projections, athttp://www.demography.state.mn.us/
http://www.implan.com/http://www.implan.com/http://www.implan.com/http://www.state.mn.us/http://www.state.mn.us/http://www.state.mn.us/http://www.state.mn.us/http://www.implan.com/


43/46

21

APPENDIX

The combined and compressed tables below provide opportunities to compare the 10%and 20% of total supply estimates of economic impact, for all years and countiesmodeled. Again, the calculations in these models use 2006 dollars for reports, without

adjustments for inflation. Readers should also be mindful of the diminishing reliability ofmodeling an economy as projections proceed into the future; therefore the valuesprojected for 2010 will warrant most scrutiny.

Economic Impact of Wind-Generated Energy as 10% and 20% of Total EnergySupply, in Cook County, Minnesota, Estimated for Years 2010, 2015, and 2020 in

10% of total supply

2010 Direct Indirect Induced Total Multi lierOutput $486,841 $8,285 $32,142 $527,269 1.1


Tax $107,0762015

Output $506,114 $8,613 $33,415 $548,142 1.1Em lo ment 1 0 1 2 2.0Value Added $411,416 $4,460 $21,557 $437,433 1.1

Tax $111,3152020

Output $528,795 $8,999 $34,912 $572,706 1.1Em lo ment 1 0 1 2 2.0Value Added $429,854 $4,660 $22,523 $457,037 1.1

Tax $116,30420% of total supply

2010 Direct Indirect Induced Total MultiplierOutput $973,683 $16,569 $64,285 $1,054,537 1.1


Tax $214,1532015

Output $1,012,229 $17,225 $66,830 $1,096,284 1.1Employment 2 0 1 3 1.5Value Added $822,833 $8,920 $43,114 $874,876 1.1

Tax $222,630

2020 Output $1,057,591 $17,997 $69,825 $1,145,413 1.1Em lo ment 2 0 1 3 1.5Value Added $859,707 $9,320 $45,046 $914,073 1.1

Tax $232,607Economic Impact of Wind-Generated Energy as 10% and 20% of Total EnergySu l in Lake Cook Count Minnesota Estimated for Years 2010 2015 and

10% of total supply


44/46

22

2010 Direct Indirect Induced Total MultiplierOut ut $1,003,400 $26,128 $88,910 $1,118,438 1.1


Tax $244,207

2015 Out ut $1,028,748 $26,788 $91,156 $1,146,691 1.1Employment 2 0 1 3 1.5Value Added $836,261 $14,451 $54,853 $905,565 1.1

Tax $250,3762020

Out ut $1,047,976 $27,288 $92,859 $1,168,124 1.1Employment 2 0 1 3 1.5Value Added $851,892 $14,721 $55,879 $922,491 1.1

Tax $255,05620% of total su l

2010 Direct Indirect Induced Total MultiplierOutput $2,006,801 $52,255 $177,819 $2,236,875 1.1

Em lo ment 4 1 2 7 1.8Value Added $1,631,313 $28,189 $107,003 $1,766,505 1.1

Tax $488,4142015

Out ut $2,057,496 $53,576 $142,614 $1,772,928 1.1Employment 4 1 2 7 1.8Value Added $1,672,522 $28,902 $109,706 $1,811,130 1.1

Tax $500,7522020

Out ut $2,095,953 $54,577 $185,719 $2,336,249 1.1Employment 4 1 2 7 1.8Value Added $1,703,784 $29,442 $111,757 $1,844,983 1.1

Tax $510,112


45/46

23

Economic Impact of Wind-Generated Energy as 10% and 20% of Total EnergySupply, in St. Louis County, Minnesota, Estimated for Years 2010, 2015, and

2020 in 2008 dollars

10% of total supply

2010 Direct Indirect Induced Total MultiplierOutput $17,306,906 $2,455,815 $2,641,450 $22,404,173 1.3

Employment 34 11 27 72 2.1

Value Added $14,148,225 $1,241,991 $1,583,566 $16,894,205 1.2

Tax $4,571,894

2015

Output $17,404,800 $2,469,706 $2,656,391 $22,530,897 1.3

Employment 34 11 27 72 2.1

Value Added $14,068,649 $1,249,016 $1,592,523 $16,989,765 1.2

Tax $4,597,754

2020Output $17,523,668 $2,486,574 $2,674,533 $22,684,776 1.3

Employment 34 11 27 72 2.1

Value Added $14,244,854 $1,257,547 $1,603,400 $17,105,800 1.2

Tax $4,629,155

20% of total supply


Output $34,,613,812 $4,911,630 $5,282,900 $44,808,346 1.3

Employment 67 22 54 143 2.1

Value Added $28,137,298 $2,483,982 $3,167,132 $33,788,410 1.2

Tax $9,143,7882015

Output $34,809,600 $4,939,412 $5,312,782 $45,061,794 1.3

Employment 68 22 54 144 2.1

Value Added $28,296,450 $2,498,032 $3,185,046 $33,979,530 1.2

Tax $9,195,508

2020

Output $35,047,336 $4,973,148 $5,349,066 $45,369,552 1.3

Employment 68 22 55 145 2.1

Value Added $28,489,708 $2,515,094 $3,206,800 $34,211,600 1.2

Tax $9,258,310


46/46

Economic Impact of Wind-Generated Energy as 10% and 20% of Total EnergySupply, in Cook, Lake and St. Louis County, Minnesota, Estimated for Years

2010, 2015, and 2020 in 2008 dollars

10% of total supply

2010 Direct Indirect Induced Total MultiplierOutput $18,797,148 $2,651,665 $2,885,860 $24,334,673 1.3

Employment 37 12 30 79 2.1

Value Added $15,208,056 $1,338,426 $1,726,876 $18,345,355 1.2

Tax $4,971,737

2015

Output $18,939,662 $2,671,769 $2,907,740 $24,519,171 1.3

Employment 37 12 30 79 2.1

Value Added $15,395,904 $1,348,573 $1,739,966 $18,484,443 1.2

Tax $5,009,430

2020Output $19,100,440 $2,694,448 $2,932,424 $24,727,311 1.3

Employment 37 12 30 79 2.1

Value Added $15,526,599 $1,360,020 $1,754,736 $18,641,354 1.2

Tax $5,051,955

20% of total supply


Output $37,594,296 $5,303,330 $5,771,720 $48,669,346 1.3

Employment 74 23 59 156 2.1

Value Added $30,560,112 $2,676,852 $3,453,746 $36,690,710 1.2

Tax $9,943,4742015

Output $37,879,324 $5,343,538 $5,815,480 $49,038,342 1.3

Employment 74 23 60 157 2.1

Value Added $30,791,808 $2,697,146 $3,479,932 $36,968,886 1.2

Tax $10,018,860

2020

Output $38,200,880 $5,388,896 $5,864,848 $49,454,622 1.3

Employment 74 24 60 158 2.1

Value Added $31,053,198 $2,720,040 $3,509,472 $37,282,708 1.2

Tax $10,103,910

Documents

Wind Resource Development in the Minnesota Coastal Zone (306-02-08)