- Email: [email protected]
Utility scale wind turbines on a grid-connected island: A feasibility study
ARTICLE IN PRESS
Renewable Energy 33 (2008) 712–719 www.elsevier.com/locate/renene
Utility scale wind turbines on a grid-connected island: A feasibility study Mohit Dua, James F. Manwell, Jon G. McGowan Department of Mechanical and Industrial Engineering, Renewable Energy Research Laboratory (RERL), University of Massachusetts, Amherst, MA 01003, USA Received 6 July 2006; accepted 10 April 2007 Available online 27 June 2007
Abstract This paper analyzes the technical and economic feasibility of installing utility-scale wind turbines on the Fox Islands, located 12 miles from the coast of Maine in the United States. Three locations on the islands, as well as a near offshore site, are analyzed in detail as potential sites for wind turbine installations. As discussed in this work, the logistic problems of transporting and installing wind turbines on the island require innovative solutions. These include locally available amphibious vessels, which can land turbine components at suitable shallow spots on the island, self-erecting towers, which allow use of a smaller crane for installation, and a special turbine foundation suitable for the local ground conditions. In the economic analysis, in addition to standard life-cycle parameters, renewable energy credits (REC) were also included. This work concludes that the installation of sub-megawatt wind turbines on the island is logistically possible and will lead to a reduction in the cost of electricity to the customers. r 2007 Elsevier Ltd. All rights reserved. Keywords: Wind systems; Wind system siting; Grid connected islands; Near offshore wind
1. Introduction/background 1.1. Overview A number of recent studies have been conducted on assessing the feasibility of wind turbine installation for various sites on New England islands. For potential sites, previous work at the University of Massachusetts (UMass) identified that there were more than 3000 islands near the shores of New England, with about 200 inhabited ones having gridconnected or isolated electrical power systems [1]. Feasibility studies at UMass [2,3] have considered the potential implementation of hybrid wind systems on Cuttyhunk Island (MA), grid-connected wind systems on four Boston (MA) harbor, islands, and a preliminary feasibility study of a gridconnected wind system for the Fox Islands (ME). In the United States, very little development of wind energy has actually taken place on New England islands. Cuttyhunk Island, for example, had a 200 kW wind turbine installed by WTG Systems in 1977. It soon ceased operation, however, primarily due to difficulties associated Corresponding author. Tel.: +1 413 545 2756.
E-mail address: [email protected] (J.G. McGowan). 0960-1481/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2007.04.007
with operating the turbine in conjunction with diesel generators (difficulties now largely solved). A MOD-OA 150 kW turbine was also installed in a wind/diesel system on Block Island, RI in the early 1980s [4], but it only operated for a few years. At present, there are no installations of wind turbines on New England islands despite the high wind potential. In Europe, there are numerous examples of wind turbines successfully sited on islands. For example, the Danish Island of Samsø and the German island of Pellworm have utilityscale installations supported by the European Commission. The major share of energy on Samsø comes from eleven, 1 MW wind turbines sited on the island and another ten, 2.3 MW Bonus turbines located off the shore of the island [5,6]. Pellworm presently has 16 wind generators with a total installed capacity of 5.9 MW [7]. It thus appears that both land-based and near offshore wind turbine systems are realistic options for island electric power systems. 1.2. Scope of work Our previous work [2,3] identified the Fox Islands in Maine as a potential site for a grid-connected utility-scale
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
wind turbine power system, and stressed the need for more wind resource data and a more detailed wind turbine siting study. The research summarized in this paper represents a continuation of our previous work, and presents a summary of the most recent wind system siting feasibility study for these islands. 2. Fox islands characteristics 2.1. Geographical location As shown in Fig. 1, the Fox Islands are located 12 miles from the coast of Rockland, Maine. Vinalhaven and North Haven are the two main islands that make up the Fox Islands. Vinalhaven, with approximately 1200 year-round residents has the largest population of the 14 year-round inhabited islands in Maine. Lobster fishing is the main component of its economy. It is approximately 9 miles long and 6 miles wide. North Haven, the second largest island of the group has a year-round population of 350. Since the island has mostly summer homes, the island population increases markedly in the summer. The population of Vinalhaven increases to approximately 6000 in the summer and the population of North Haven increases to 2000. 2.2. Electrical supply system Fox Island Electric Cooperative (FIEC) has been providing electricity for Vinalhaven and North Haven since 1975. Electricity is purchased wholesale on the mainland and delivered to the island via four armored submarine cables. Since the cables are not buried in the ocean floor they have been damaged by ocean currents and fishing activity, and they have required repair numerous times. In case of damage, repair on a cable can take up to 3 days, during which time the spare cable is used. If two of
Fig. 1. Fox Islands map.
713
the four cables are damaged, supply is switched over to single phase. The commercial users on the island are adversely affected by this switchover to single phase. Fig. 2 shows the average monthly load from September 2002 to August 2003. The curve shows two peaks, one during August, due to summer residents and tourists, and the other in January, attributed to the heating load in winter. FIEC paid an average electricity cost of $41.77/MWh during the analysis period (September 2002–August 2003). Including the line losses, cost of operating the utility and maintaining the transmission and distribution, the cost for FIEC customers increased to $80/MWh. In addition, the customers were also charged $100/MWh for the cost of transmission cable, resulting in a cost of $180/MWh. FIEC has recently replaced the cables and upgraded the distribution system. With these upgrades, electricity costs have risen to $220/MWh. 2.3. Wind resource Wind resource monitoring has been carried out at Vinalhaven by the University of Massachusetts since the end of August 2002. The wind-monitoring site is in the middle of the island (441060 1300 N/681520 6000 W) on a granite ledge covered with low trees and bushes. The site is surrounded by trees, which are 6–12 m in height, located approximately 30 m from the monitoring tower. Two sensors each for monitoring wind speed and direction are mounted at 40 m. The data sampling and averaging rates are 2 s and 10 min, respectively. Fig. 3 shows the monthly average wind speeds at the 40 m height (period of measurement: September 2002–August 2003). As expected, the wind speeds are lowest in summer and highest in winter. The average wind speed for the site for the year was 5.4 m/s. By using a log-law extrapolation (surface roughness ¼ 0.5 m), the average wind speed was estimated to be 5.7 m/s at 50 m. A Sonic Detection and Ranging (SODAR) system was also used at the measurement site to measure wind speeds at multiple heights. Fig. 4 shows a plot of wind speed time series at multiple heights for a 10-day period. The readings showed a significant wind shear on the island due to the topography. The SODAR data was not used in this study, however, since recent investigations at the University of Massachusetts indicate that SODAR equipment may underestimate wind speeds at lower elevations when there are trees or other obstacles present [8]. In order to model the wind speeds at different locations on the islands, an estimation approach using the New England wind map from AWS Truewind Solutions [9] was used. Fig. 5 shows the AWS Truewind map that was first used to estimate the average wind speed at the monitoring site. At the 70 m level, hub height of potential wind turbines to be used, the New England wind map estimated an average wind speed of 6.7 m/s. It should be noted that the Truewind map yielded an average wind speed of 5.9 m/s
ARTICLE IN PRESS 714
M. Dua et al. / Renewable Energy 33 (2008) 712–719
Fig. 2. Average monthly electric loads.
Fig. 3. Average monthly wind speeds.
Fig. 4. Wind speeds (SODAR) at multiple heights.
at 50 m, slightly higher than the estimated value using hourly data. We also note that the SODAR data at the measurement site yielded an average wind speed of 7.6 m/s
at 70 m. For this work, for the different island sites (onshore and offshore), a conservative approach of using the hour-by-hour data adjusted to the AWS Truewind map value was used. That is, the hourly monitored wind speed was scaled to the Truewind average to get representative hourly wind speed time series at the selected sites. Finally, to compare the wind speed for the year (September 2002–August 2003) to the long-term wind speed, wind speed data from the C-MANN station on Matinicus Rock were used [10]. The average wind speed for the period of 1984–2001 was found to be 7.91 m/s, whereas average wind speed for the monitoring period was found to be 7.92 m/s. Thus, the average wind speeds during the monitoring period is similar to the long-term average wind speed. Therefore, a scaling factor for the adjustment to the long-term average was not introduced.
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
3. Proposed power system 3.1. Potential wind generation system As previously noted, the customers of FIEC are paying a high charge for electricity (about $220/MWh) due to high transmission charges of $100/MWh. Thus, this utility is
715
considering electricity generation from wind turbines (submegawatt sized) on the island to offset these charges. The proposed wind turbines will be connected to the distribution grid on the island. For each of the proposed sites, one to three wind turbines were considered. Electricity generated using the wind turbines will offset the electricity purchased from the mainland. If electricity in excess of the local demand is generated from the wind turbines, it will be sold to a mainland utility. 3.2. Potential wind turbine sites A number of potential island sites were analyzed with respect to road access, grid connectivity, area availability, site conditions, wind resource, visual impact, land ownership, and noise issue. As shown in Fig. 6, three sites were selected for detailed analysis: Isle au Haut, Round Neck, and Leadbetter Island, with average wind speeds (AWS Truewind map) at 70 m of 6.9, 7.5, and 7.8 m/s, respectively. 3.3. Potential wind turbine choice
Fig. 5. Wind speed map at 70 m height.
As discussed next, the logistics of wind turbine transportation plays a major role in deciding the maximum turbine size that can be transported to the island. Table 1 gives the component details for the Vestas V47–660 kW and V80–1.8 MW wind turbines. The Vestas V52–850 kW, which is an intermediate size between the V47 and V80, is not available for the US market. Likewise, the smallest size supplied by General Electric Wind for the US market was the 1.5 MW series.
Fig. 6. Potential sites for wind turbines.
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
716 Table 1 Component details for wind turbines Wind turbine
V47–660 kW
V80–1.8 MW
Diameter Number of blades Hub height (approx.) Nacelle weight Rotor weight
47 m 3 40–45–50–55 m 20.4 T 7.2 T
80 m 3 60–67–78 m 63 T 35 T
There is no one in the vicinity of the islands who can supply a crane suitable for lifting the 63 ton nacelle, for the V80, to the tower height. The manager of FIEC also pointed out that public acceptance for wind energy on the island would be higher for a smaller turbine. Therefore, in view of these considerations, and transportation and installation issues (discussed next), only the use of a V47 turbine was considered for this study.
3.4. Transportation and installation Since the site is located on an island with limited infrastructure there are unique transportation and installation challenges. There is a regular ferry service between Vinalhaven and Rockland. Vinalhaven has a pier with a load-bearing capacity of 36,400 kg. This capacity is suitable for landing the wind turbine components on the pier. However, the pier is located in the most densely populated part of the island. Therefore, for equipment to be transported from the pier to the site, it needs to pass through narrow roads and sharp turns. While the nacelle, turbine rotor, and tower sections can be transported using the pier, alternate transportation arrangements are required for transporting the blades due to the large turning radius of the transportation truck. One such alternative is using an amphibious vessel. A local company in Rockland undertakes transportation to the island using such a vessel. Fig. 7 shows a photograph of the vessel. The vessel, with a length of 29 m and width of 8.5 m can carry up to 100 tons of gross weight. The vessel can be landed at a suitable shallow landing point near the potential site for the turbine installation. The blades can then be transported to the site using the existing roads or new roads may need to be constructed. To allow for installation of a 660 kW wind turbine on the island, with tower sections supplied by the manufacturer, a crane is required for lifting the nacelle (20.4 ton weight) to a hub height of around 60 m. Transportation of such a crane to the island is not feasible. Valmont’s Wind Energy Structure [11] addresses this problem by allowing the use of smaller cranes for wind turbine installation. For installation on site, a 70 ton Linkbelt hydraulic truck crane with extended boom package or a standard 90 ton hydraulic crane is required. The 70 ton Linkbelt hydraulic crane is
Fig. 7. Amphious vessel.
available in Rockland and has been transported to the island using the amphibious vessel. Conventional wind turbine foundations require large quantities of concrete and water. Vinalhaven is a rocky island and the use of conventional foundations will require a large quantity of rock to be removed by blasting. For turbine installation on a rocky site, rock anchor foundations can be used. 3.5. Electrical interconnection Electrical interconnection includes components starting at the wind turbine and leading to the point of interconnection with the distribution system on the island. The main components are generator step-up transformer, overhead cables, meter, and recloser. The pad-mounted generator step transformer will be designed to match the generator voltage (0.69 kV) on the low-voltage side and match the medium-voltage system (7.2 kV) on the other side. Generally, the medium-voltage cables, which comprise the power system between the wind turbines, are underground to minimize the visual impact. For this case, we consider the use of overhead cables because it is difficult to install underground cables in a rocky terrain and the visual aspect of overhead cables are not expected to cause a public acceptance issue. At the interconnection with the distribution line a fault-interrupting device called a ‘‘recloser’’ is used. This is a conventional utility ‘‘circuit breaker’’ that is used to automatically isolate the section of the line connected to the wind turbines in case of an electrical fault. A meter is used at the interconnection to record the wind energy fed into the grid.
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
For one of the sites located on an island (Leadbetter Island), a submarine cable is required to transmit electricity to Vinalhaven. The manager of FIEC noted that for this portion, a part of the old cable (from the undersea cable recently replaced) could be salvaged for this use. 4. Performance analysis: three island sites 4.1. Overview The feasibility of the proposed power systems was evaluated from the resource, logistic, and economic points of view. A detailed evaluation was carried out for installation of 1, 2, or 3 wind turbines at the proposed sites. The wind turbine electrical production for the various sites and different number of turbines was calculated using the hour-by-hour resource data, appropriately scaled for each of the sites. For the analysis, it was assumed that the availability of each wind turbine was 97%, array loss with two wind turbines was 6%, and with three wind turbines was 10%. While calculating the net wind energy sold, 2% transmission losses to the mainland are taken into consideration. A separate program was written to carry out the basic calculations of electricity generation and its export/import. Full details (including economics) of the analysis are given in Ref. [12]. 4.2. Economic analysis A detailed cost study was carried out for each of the sites for installing one to three wind turbines, and the following costs were included:
electrical interconnection, road preparation, site preparation, foundation, tower and Installation, wind turbine, dismantling.
It should be noted that the cost estimates were performed for equipment and installation and did not include any permitting or other predevelopment work. These costs could depend substantially on the methods by which the project would be undertaken. A final cost estimate that would be carried out before the decision to proceed would include these other factors. For a proposed wind energy system to be economically feasible, the market value of the electricity generated must exceed the cost to produce energy from the wind energy system. For this work, the net present value of savings over the life of the project was used to compare the various options. A brief description of the more important factors in the economic analysis follows.
717
4.2.1. Renewable energy credits In April 2002, Massachusetts implemented a Renewable Portfolio Standard (RPS), requiring retail electricity sold in the state to include a minimum percentage of electricity with renewable attributes. With RPS, the environmental attributes of electricity are unbundled from the electricity itself. This is done using Renewable Energy Credits (RECs). Maine also has an RPS, but the inclusion of hydropower in the definition of renewable energy precludes the development of wind energy in the state. Both Maine and Massachusetts are part of the New England Power Pool (NEPOOL), which is a voluntary association of entities engaged in electric power business in New England. NEPOOL has a single regional electric transmission network. Massachusetts allows electric suppliers to achieve compliance using certificates from generation sourced anywhere in NEPOOL or an adjacent power pool, provided that the power flows onto the NEPOOL grid. Therefore, RECs from electricity generated in Maine can be sold to a utility in Massachusetts. At the time of this work, RECs were trading at about $40/MWh [13]. 4.2.2. Electricity selling options For the Fox Islands there are two alternatives for selling the excess electricity generated. The RECs can be directly sold at $35–40/MWh. The other alternative is to sell electricity through a local non-profit organization as ‘green electricity’. We assumed that market forces would ensure that both alternatives yield the same economics. The analysis was therefore carried out for the first alternative with an assumed value of $35/MWh as the selling price of electricity. 4.3. Economic analysis results The economic analysis was carried out using the input parameters summarized in Table 2. The results from this analysis are shown in Table 3. It should be noted that Table 3 lists the results from an analysis that includes the income from RECs but does not include the Renewable Energy Production Incentive. As can be seen from the table, all sites would prove to be profitable with the net present value of savings, over the lifetime of the project ranging from $9,82,000 to $4,399,000 compared with the present case. The calculation of these savings assumes that wind energy offsets electricity bought from the mainland, and for periods when the wind energy is greater than the island load, is exported to the mainland. 5. Performance analysis: offshore wind energy This work, prompted by an interest from FIEC, also considered the use of near offshore wind energy sites for the future. An analysis was carried out using Geographic Information Systems to determine the most favorable locations for offshore installations (a raster analysis with a cell size of 200 200 m was used). The cost of electricity
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
718
generation was evaluated for each cell assuming that cell to be the center of a wind farm (consisting of 10, GE 1.5 MW wind turbines) and conditions (water depth [14], wind speed [9], and distance of cable travel) for the whole farm being the same as the selected cell. Only areas with water depth less than 30 m were considered for the analysis. The cable length to the interconnection point on the island was calculated using a cost function of unity for travel over land and five for travel undersea to account for the differential cost of transmission. A simplified cost model was used for evaluating the levelized cost of energy over a 20-year period using the same interest, discount, and inflation rate as the earlier onshore analysis. Costs were included for wind turbines, support structure material, wind turbine installation, cable to landing point on the island, balance of station, and operation and maintenance. Fig. 8 shows the levelized cost of energy for the offshore installation (values are shown in $/kWh). The best potential is on the southeast side of Vinalhaven with a cost of energy around $65/MWh (excluding RECs). This is a very economic price for offshore wind energy. The reason
for the low cost of energy is the combination of low water depth (less than 20 m), high wind speed (8.5–9.0 m/s), and less distance to the island (0.6 km).
6. Conclusions Based on the work of this study, the Fox Islands appear to have a viable potential for wind energy. Out of the three selected sites, Leadbetter Islands shows the highest potential. Public acceptance and permitting may finally be the deciding factors in selecting the final site. With 14 year-round inhabited islands in Maine, this project could serve as a model for wind energy development in the State of Maine. In addition, future work on wind system siting on these islands should consider the possibility of near offshore sites.
Table 2 Economic parameters used in analysis Economic parameters Down payment Loan interest rate Discount rate General inflation rate Period of loan System life
20% 4.0% 3.00% 2.00% 20 years 20 years
Electricity Electricity Electricity Electricity
$48.3/MWh 2.0% $35.0/MWh
rate buying rate inflation rate selling rate
Incentives RECs Annual escalation
$38.0/MWh 2.0%
Others Operation & maintenance (O&M) Annual escalation Insurance Annual escalation
$12.5/MWh 2.0% $3.5/MWh 20%
Fig. 8. Levelized cost for offshore installation.
Table 3 Results of economic analysis Site
Isle Au haut
No. of wind turbines (V47) Total system cost ($ l000) Electricity produced by wind turbines (MWh) Electricity exported to grid (MWh) Savings ($ l000)
1 955 1578 0.98 982
2 1800 2966 28 1833
Round neck 3 2637 4260 263 2519
1 931 1954 1.02 1486
2 1782 3673 50 2758
Leadbetter Island 3 2633 5275 419 3885
1 1119 2164 1.09 1550
2 1972 4067 66 3054
3 2825 5842 523 4399
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
Acknowledgments This work was sponsored by the US Department of Energy, Region 1. We also wish to acknowledge the help of Dave Folce from the Fox Islands Electric Cooperative.
[6] [7] [8]
References [1] Manwell JF, McGowan JG, Blanco G. Potential for wind energy development on New England Islands. Renewable energy research laboratory report. Amherst: Department of Mechanical and Industrial Engineering, University of Massachusetts/Amherst; 2002. [2] Manwell JF, McGowan JG. Wind/hybrid power system applications for New England islands. Wind Eng 2003;27(2):143–56. [3] Manwell JF, McGowan JG. Development of wind energy systems for New England islands. Renew. Energy 2004;29(10):1707–20. [4] Stiller PH, Scott GW, Shaltens RK. Measured effect of wind generation on the fuel consumption of an isolated diesel power system. IEEE Trans Power Apparatus Syst 1983;PAS-102:1788–92 (ISSN 0018-9510). [5] Towards 100% RES supply on Samsø, Denmark. Three years of experience in a planning period over ten years. Samsø Energy and
[9] [10] [11] [12]
[13] [14]
719
Environment Office. Website: /www.insula.org/islandsonline/ samsoe-2.pdfS, Accessed June, 2006. Offshore wind energy research project portal. Website /http:// www.offshorewinde6.ergy.orgS. Accessed June, 2006. The Pellworm experience. Website: /http://www.insula.org/islandsonline/ pellworm-1.pdfS. Accessed June, 2006. Rogers AL, Manwell JF, Ellis AF. Wind shear over forested areas. In: Proceedings of 2005 wind energy symposium, AIAA, Reno NV, 2005, p. 489–96. AWS Truewind, Website: /http://truewind.teamcamelot.com/ne/S. Accessed June, 2006. The National Data Buoy Center, Website, /http://www.ndbc.noaa.govS. Accessed June, 2006. Valmont Industries. Website: /http://www.valmont.comS. Accessed June, 2006. Dua M. Analysis of siting utility scale wind turbines on Islands off the Coast of New England. MS Project Report, Department of Mechanical and Industrial Engineering, University of Massahusetts, Amherst, 2004. Evolution Markets LLC. Website: /http://www.evomarkets.comS. Accessed June, 2006. Gulf of Maine Coastal Program. Website: /http://gulfofmaine.fws.govS. Accessed June, 2006.
Renewable Energy 33 (2008) 712–719 www.elsevier.com/locate/renene
Utility scale wind turbines on a grid-connected island: A feasibility study Mohit Dua, James F. Manwell, Jon G. McGowan Department of Mechanical and Industrial Engineering, Renewable Energy Research Laboratory (RERL), University of Massachusetts, Amherst, MA 01003, USA Received 6 July 2006; accepted 10 April 2007 Available online 27 June 2007
Abstract This paper analyzes the technical and economic feasibility of installing utility-scale wind turbines on the Fox Islands, located 12 miles from the coast of Maine in the United States. Three locations on the islands, as well as a near offshore site, are analyzed in detail as potential sites for wind turbine installations. As discussed in this work, the logistic problems of transporting and installing wind turbines on the island require innovative solutions. These include locally available amphibious vessels, which can land turbine components at suitable shallow spots on the island, self-erecting towers, which allow use of a smaller crane for installation, and a special turbine foundation suitable for the local ground conditions. In the economic analysis, in addition to standard life-cycle parameters, renewable energy credits (REC) were also included. This work concludes that the installation of sub-megawatt wind turbines on the island is logistically possible and will lead to a reduction in the cost of electricity to the customers. r 2007 Elsevier Ltd. All rights reserved. Keywords: Wind systems; Wind system siting; Grid connected islands; Near offshore wind
1. Introduction/background 1.1. Overview A number of recent studies have been conducted on assessing the feasibility of wind turbine installation for various sites on New England islands. For potential sites, previous work at the University of Massachusetts (UMass) identified that there were more than 3000 islands near the shores of New England, with about 200 inhabited ones having gridconnected or isolated electrical power systems [1]. Feasibility studies at UMass [2,3] have considered the potential implementation of hybrid wind systems on Cuttyhunk Island (MA), grid-connected wind systems on four Boston (MA) harbor, islands, and a preliminary feasibility study of a gridconnected wind system for the Fox Islands (ME). In the United States, very little development of wind energy has actually taken place on New England islands. Cuttyhunk Island, for example, had a 200 kW wind turbine installed by WTG Systems in 1977. It soon ceased operation, however, primarily due to difficulties associated Corresponding author. Tel.: +1 413 545 2756.
E-mail address: [email protected] (J.G. McGowan). 0960-1481/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2007.04.007
with operating the turbine in conjunction with diesel generators (difficulties now largely solved). A MOD-OA 150 kW turbine was also installed in a wind/diesel system on Block Island, RI in the early 1980s [4], but it only operated for a few years. At present, there are no installations of wind turbines on New England islands despite the high wind potential. In Europe, there are numerous examples of wind turbines successfully sited on islands. For example, the Danish Island of Samsø and the German island of Pellworm have utilityscale installations supported by the European Commission. The major share of energy on Samsø comes from eleven, 1 MW wind turbines sited on the island and another ten, 2.3 MW Bonus turbines located off the shore of the island [5,6]. Pellworm presently has 16 wind generators with a total installed capacity of 5.9 MW [7]. It thus appears that both land-based and near offshore wind turbine systems are realistic options for island electric power systems. 1.2. Scope of work Our previous work [2,3] identified the Fox Islands in Maine as a potential site for a grid-connected utility-scale
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
wind turbine power system, and stressed the need for more wind resource data and a more detailed wind turbine siting study. The research summarized in this paper represents a continuation of our previous work, and presents a summary of the most recent wind system siting feasibility study for these islands. 2. Fox islands characteristics 2.1. Geographical location As shown in Fig. 1, the Fox Islands are located 12 miles from the coast of Rockland, Maine. Vinalhaven and North Haven are the two main islands that make up the Fox Islands. Vinalhaven, with approximately 1200 year-round residents has the largest population of the 14 year-round inhabited islands in Maine. Lobster fishing is the main component of its economy. It is approximately 9 miles long and 6 miles wide. North Haven, the second largest island of the group has a year-round population of 350. Since the island has mostly summer homes, the island population increases markedly in the summer. The population of Vinalhaven increases to approximately 6000 in the summer and the population of North Haven increases to 2000. 2.2. Electrical supply system Fox Island Electric Cooperative (FIEC) has been providing electricity for Vinalhaven and North Haven since 1975. Electricity is purchased wholesale on the mainland and delivered to the island via four armored submarine cables. Since the cables are not buried in the ocean floor they have been damaged by ocean currents and fishing activity, and they have required repair numerous times. In case of damage, repair on a cable can take up to 3 days, during which time the spare cable is used. If two of
Fig. 1. Fox Islands map.
713
the four cables are damaged, supply is switched over to single phase. The commercial users on the island are adversely affected by this switchover to single phase. Fig. 2 shows the average monthly load from September 2002 to August 2003. The curve shows two peaks, one during August, due to summer residents and tourists, and the other in January, attributed to the heating load in winter. FIEC paid an average electricity cost of $41.77/MWh during the analysis period (September 2002–August 2003). Including the line losses, cost of operating the utility and maintaining the transmission and distribution, the cost for FIEC customers increased to $80/MWh. In addition, the customers were also charged $100/MWh for the cost of transmission cable, resulting in a cost of $180/MWh. FIEC has recently replaced the cables and upgraded the distribution system. With these upgrades, electricity costs have risen to $220/MWh. 2.3. Wind resource Wind resource monitoring has been carried out at Vinalhaven by the University of Massachusetts since the end of August 2002. The wind-monitoring site is in the middle of the island (441060 1300 N/681520 6000 W) on a granite ledge covered with low trees and bushes. The site is surrounded by trees, which are 6–12 m in height, located approximately 30 m from the monitoring tower. Two sensors each for monitoring wind speed and direction are mounted at 40 m. The data sampling and averaging rates are 2 s and 10 min, respectively. Fig. 3 shows the monthly average wind speeds at the 40 m height (period of measurement: September 2002–August 2003). As expected, the wind speeds are lowest in summer and highest in winter. The average wind speed for the site for the year was 5.4 m/s. By using a log-law extrapolation (surface roughness ¼ 0.5 m), the average wind speed was estimated to be 5.7 m/s at 50 m. A Sonic Detection and Ranging (SODAR) system was also used at the measurement site to measure wind speeds at multiple heights. Fig. 4 shows a plot of wind speed time series at multiple heights for a 10-day period. The readings showed a significant wind shear on the island due to the topography. The SODAR data was not used in this study, however, since recent investigations at the University of Massachusetts indicate that SODAR equipment may underestimate wind speeds at lower elevations when there are trees or other obstacles present [8]. In order to model the wind speeds at different locations on the islands, an estimation approach using the New England wind map from AWS Truewind Solutions [9] was used. Fig. 5 shows the AWS Truewind map that was first used to estimate the average wind speed at the monitoring site. At the 70 m level, hub height of potential wind turbines to be used, the New England wind map estimated an average wind speed of 6.7 m/s. It should be noted that the Truewind map yielded an average wind speed of 5.9 m/s
ARTICLE IN PRESS 714
M. Dua et al. / Renewable Energy 33 (2008) 712–719
Fig. 2. Average monthly electric loads.
Fig. 3. Average monthly wind speeds.
Fig. 4. Wind speeds (SODAR) at multiple heights.
at 50 m, slightly higher than the estimated value using hourly data. We also note that the SODAR data at the measurement site yielded an average wind speed of 7.6 m/s
at 70 m. For this work, for the different island sites (onshore and offshore), a conservative approach of using the hour-by-hour data adjusted to the AWS Truewind map value was used. That is, the hourly monitored wind speed was scaled to the Truewind average to get representative hourly wind speed time series at the selected sites. Finally, to compare the wind speed for the year (September 2002–August 2003) to the long-term wind speed, wind speed data from the C-MANN station on Matinicus Rock were used [10]. The average wind speed for the period of 1984–2001 was found to be 7.91 m/s, whereas average wind speed for the monitoring period was found to be 7.92 m/s. Thus, the average wind speeds during the monitoring period is similar to the long-term average wind speed. Therefore, a scaling factor for the adjustment to the long-term average was not introduced.
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
3. Proposed power system 3.1. Potential wind generation system As previously noted, the customers of FIEC are paying a high charge for electricity (about $220/MWh) due to high transmission charges of $100/MWh. Thus, this utility is
715
considering electricity generation from wind turbines (submegawatt sized) on the island to offset these charges. The proposed wind turbines will be connected to the distribution grid on the island. For each of the proposed sites, one to three wind turbines were considered. Electricity generated using the wind turbines will offset the electricity purchased from the mainland. If electricity in excess of the local demand is generated from the wind turbines, it will be sold to a mainland utility. 3.2. Potential wind turbine sites A number of potential island sites were analyzed with respect to road access, grid connectivity, area availability, site conditions, wind resource, visual impact, land ownership, and noise issue. As shown in Fig. 6, three sites were selected for detailed analysis: Isle au Haut, Round Neck, and Leadbetter Island, with average wind speeds (AWS Truewind map) at 70 m of 6.9, 7.5, and 7.8 m/s, respectively. 3.3. Potential wind turbine choice
Fig. 5. Wind speed map at 70 m height.
As discussed next, the logistics of wind turbine transportation plays a major role in deciding the maximum turbine size that can be transported to the island. Table 1 gives the component details for the Vestas V47–660 kW and V80–1.8 MW wind turbines. The Vestas V52–850 kW, which is an intermediate size between the V47 and V80, is not available for the US market. Likewise, the smallest size supplied by General Electric Wind for the US market was the 1.5 MW series.
Fig. 6. Potential sites for wind turbines.
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
716 Table 1 Component details for wind turbines Wind turbine
V47–660 kW
V80–1.8 MW
Diameter Number of blades Hub height (approx.) Nacelle weight Rotor weight
47 m 3 40–45–50–55 m 20.4 T 7.2 T
80 m 3 60–67–78 m 63 T 35 T
There is no one in the vicinity of the islands who can supply a crane suitable for lifting the 63 ton nacelle, for the V80, to the tower height. The manager of FIEC also pointed out that public acceptance for wind energy on the island would be higher for a smaller turbine. Therefore, in view of these considerations, and transportation and installation issues (discussed next), only the use of a V47 turbine was considered for this study.
3.4. Transportation and installation Since the site is located on an island with limited infrastructure there are unique transportation and installation challenges. There is a regular ferry service between Vinalhaven and Rockland. Vinalhaven has a pier with a load-bearing capacity of 36,400 kg. This capacity is suitable for landing the wind turbine components on the pier. However, the pier is located in the most densely populated part of the island. Therefore, for equipment to be transported from the pier to the site, it needs to pass through narrow roads and sharp turns. While the nacelle, turbine rotor, and tower sections can be transported using the pier, alternate transportation arrangements are required for transporting the blades due to the large turning radius of the transportation truck. One such alternative is using an amphibious vessel. A local company in Rockland undertakes transportation to the island using such a vessel. Fig. 7 shows a photograph of the vessel. The vessel, with a length of 29 m and width of 8.5 m can carry up to 100 tons of gross weight. The vessel can be landed at a suitable shallow landing point near the potential site for the turbine installation. The blades can then be transported to the site using the existing roads or new roads may need to be constructed. To allow for installation of a 660 kW wind turbine on the island, with tower sections supplied by the manufacturer, a crane is required for lifting the nacelle (20.4 ton weight) to a hub height of around 60 m. Transportation of such a crane to the island is not feasible. Valmont’s Wind Energy Structure [11] addresses this problem by allowing the use of smaller cranes for wind turbine installation. For installation on site, a 70 ton Linkbelt hydraulic truck crane with extended boom package or a standard 90 ton hydraulic crane is required. The 70 ton Linkbelt hydraulic crane is
Fig. 7. Amphious vessel.
available in Rockland and has been transported to the island using the amphibious vessel. Conventional wind turbine foundations require large quantities of concrete and water. Vinalhaven is a rocky island and the use of conventional foundations will require a large quantity of rock to be removed by blasting. For turbine installation on a rocky site, rock anchor foundations can be used. 3.5. Electrical interconnection Electrical interconnection includes components starting at the wind turbine and leading to the point of interconnection with the distribution system on the island. The main components are generator step-up transformer, overhead cables, meter, and recloser. The pad-mounted generator step transformer will be designed to match the generator voltage (0.69 kV) on the low-voltage side and match the medium-voltage system (7.2 kV) on the other side. Generally, the medium-voltage cables, which comprise the power system between the wind turbines, are underground to minimize the visual impact. For this case, we consider the use of overhead cables because it is difficult to install underground cables in a rocky terrain and the visual aspect of overhead cables are not expected to cause a public acceptance issue. At the interconnection with the distribution line a fault-interrupting device called a ‘‘recloser’’ is used. This is a conventional utility ‘‘circuit breaker’’ that is used to automatically isolate the section of the line connected to the wind turbines in case of an electrical fault. A meter is used at the interconnection to record the wind energy fed into the grid.
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
For one of the sites located on an island (Leadbetter Island), a submarine cable is required to transmit electricity to Vinalhaven. The manager of FIEC noted that for this portion, a part of the old cable (from the undersea cable recently replaced) could be salvaged for this use. 4. Performance analysis: three island sites 4.1. Overview The feasibility of the proposed power systems was evaluated from the resource, logistic, and economic points of view. A detailed evaluation was carried out for installation of 1, 2, or 3 wind turbines at the proposed sites. The wind turbine electrical production for the various sites and different number of turbines was calculated using the hour-by-hour resource data, appropriately scaled for each of the sites. For the analysis, it was assumed that the availability of each wind turbine was 97%, array loss with two wind turbines was 6%, and with three wind turbines was 10%. While calculating the net wind energy sold, 2% transmission losses to the mainland are taken into consideration. A separate program was written to carry out the basic calculations of electricity generation and its export/import. Full details (including economics) of the analysis are given in Ref. [12]. 4.2. Economic analysis A detailed cost study was carried out for each of the sites for installing one to three wind turbines, and the following costs were included:
electrical interconnection, road preparation, site preparation, foundation, tower and Installation, wind turbine, dismantling.
It should be noted that the cost estimates were performed for equipment and installation and did not include any permitting or other predevelopment work. These costs could depend substantially on the methods by which the project would be undertaken. A final cost estimate that would be carried out before the decision to proceed would include these other factors. For a proposed wind energy system to be economically feasible, the market value of the electricity generated must exceed the cost to produce energy from the wind energy system. For this work, the net present value of savings over the life of the project was used to compare the various options. A brief description of the more important factors in the economic analysis follows.
717
4.2.1. Renewable energy credits In April 2002, Massachusetts implemented a Renewable Portfolio Standard (RPS), requiring retail electricity sold in the state to include a minimum percentage of electricity with renewable attributes. With RPS, the environmental attributes of electricity are unbundled from the electricity itself. This is done using Renewable Energy Credits (RECs). Maine also has an RPS, but the inclusion of hydropower in the definition of renewable energy precludes the development of wind energy in the state. Both Maine and Massachusetts are part of the New England Power Pool (NEPOOL), which is a voluntary association of entities engaged in electric power business in New England. NEPOOL has a single regional electric transmission network. Massachusetts allows electric suppliers to achieve compliance using certificates from generation sourced anywhere in NEPOOL or an adjacent power pool, provided that the power flows onto the NEPOOL grid. Therefore, RECs from electricity generated in Maine can be sold to a utility in Massachusetts. At the time of this work, RECs were trading at about $40/MWh [13]. 4.2.2. Electricity selling options For the Fox Islands there are two alternatives for selling the excess electricity generated. The RECs can be directly sold at $35–40/MWh. The other alternative is to sell electricity through a local non-profit organization as ‘green electricity’. We assumed that market forces would ensure that both alternatives yield the same economics. The analysis was therefore carried out for the first alternative with an assumed value of $35/MWh as the selling price of electricity. 4.3. Economic analysis results The economic analysis was carried out using the input parameters summarized in Table 2. The results from this analysis are shown in Table 3. It should be noted that Table 3 lists the results from an analysis that includes the income from RECs but does not include the Renewable Energy Production Incentive. As can be seen from the table, all sites would prove to be profitable with the net present value of savings, over the lifetime of the project ranging from $9,82,000 to $4,399,000 compared with the present case. The calculation of these savings assumes that wind energy offsets electricity bought from the mainland, and for periods when the wind energy is greater than the island load, is exported to the mainland. 5. Performance analysis: offshore wind energy This work, prompted by an interest from FIEC, also considered the use of near offshore wind energy sites for the future. An analysis was carried out using Geographic Information Systems to determine the most favorable locations for offshore installations (a raster analysis with a cell size of 200 200 m was used). The cost of electricity
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
718
generation was evaluated for each cell assuming that cell to be the center of a wind farm (consisting of 10, GE 1.5 MW wind turbines) and conditions (water depth [14], wind speed [9], and distance of cable travel) for the whole farm being the same as the selected cell. Only areas with water depth less than 30 m were considered for the analysis. The cable length to the interconnection point on the island was calculated using a cost function of unity for travel over land and five for travel undersea to account for the differential cost of transmission. A simplified cost model was used for evaluating the levelized cost of energy over a 20-year period using the same interest, discount, and inflation rate as the earlier onshore analysis. Costs were included for wind turbines, support structure material, wind turbine installation, cable to landing point on the island, balance of station, and operation and maintenance. Fig. 8 shows the levelized cost of energy for the offshore installation (values are shown in $/kWh). The best potential is on the southeast side of Vinalhaven with a cost of energy around $65/MWh (excluding RECs). This is a very economic price for offshore wind energy. The reason
for the low cost of energy is the combination of low water depth (less than 20 m), high wind speed (8.5–9.0 m/s), and less distance to the island (0.6 km).
6. Conclusions Based on the work of this study, the Fox Islands appear to have a viable potential for wind energy. Out of the three selected sites, Leadbetter Islands shows the highest potential. Public acceptance and permitting may finally be the deciding factors in selecting the final site. With 14 year-round inhabited islands in Maine, this project could serve as a model for wind energy development in the State of Maine. In addition, future work on wind system siting on these islands should consider the possibility of near offshore sites.
Table 2 Economic parameters used in analysis Economic parameters Down payment Loan interest rate Discount rate General inflation rate Period of loan System life
20% 4.0% 3.00% 2.00% 20 years 20 years
Electricity Electricity Electricity Electricity
$48.3/MWh 2.0% $35.0/MWh
rate buying rate inflation rate selling rate
Incentives RECs Annual escalation
$38.0/MWh 2.0%
Others Operation & maintenance (O&M) Annual escalation Insurance Annual escalation
$12.5/MWh 2.0% $3.5/MWh 20%
Fig. 8. Levelized cost for offshore installation.
Table 3 Results of economic analysis Site
Isle Au haut
No. of wind turbines (V47) Total system cost ($ l000) Electricity produced by wind turbines (MWh) Electricity exported to grid (MWh) Savings ($ l000)
1 955 1578 0.98 982
2 1800 2966 28 1833
Round neck 3 2637 4260 263 2519
1 931 1954 1.02 1486
2 1782 3673 50 2758
Leadbetter Island 3 2633 5275 419 3885
1 1119 2164 1.09 1550
2 1972 4067 66 3054
3 2825 5842 523 4399
ARTICLE IN PRESS M. Dua et al. / Renewable Energy 33 (2008) 712–719
Acknowledgments This work was sponsored by the US Department of Energy, Region 1. We also wish to acknowledge the help of Dave Folce from the Fox Islands Electric Cooperative.
[6] [7] [8]
References [1] Manwell JF, McGowan JG, Blanco G. Potential for wind energy development on New England Islands. Renewable energy research laboratory report. Amherst: Department of Mechanical and Industrial Engineering, University of Massachusetts/Amherst; 2002. [2] Manwell JF, McGowan JG. Wind/hybrid power system applications for New England islands. Wind Eng 2003;27(2):143–56. [3] Manwell JF, McGowan JG. Development of wind energy systems for New England islands. Renew. Energy 2004;29(10):1707–20. [4] Stiller PH, Scott GW, Shaltens RK. Measured effect of wind generation on the fuel consumption of an isolated diesel power system. IEEE Trans Power Apparatus Syst 1983;PAS-102:1788–92 (ISSN 0018-9510). [5] Towards 100% RES supply on Samsø, Denmark. Three years of experience in a planning period over ten years. Samsø Energy and
[9] [10] [11] [12]
[13] [14]
719
Environment Office. Website: /www.insula.org/islandsonline/ samsoe-2.pdfS, Accessed June, 2006. Offshore wind energy research project portal. Website /http:// www.offshorewinde6.ergy.orgS. Accessed June, 2006. The Pellworm experience. Website: /http://www.insula.org/islandsonline/ pellworm-1.pdfS. Accessed June, 2006. Rogers AL, Manwell JF, Ellis AF. Wind shear over forested areas. In: Proceedings of 2005 wind energy symposium, AIAA, Reno NV, 2005, p. 489–96. AWS Truewind, Website: /http://truewind.teamcamelot.com/ne/S. Accessed June, 2006. The National Data Buoy Center, Website, /http://www.ndbc.noaa.govS. Accessed June, 2006. Valmont Industries. Website: /http://www.valmont.comS. Accessed June, 2006. Dua M. Analysis of siting utility scale wind turbines on Islands off the Coast of New England. MS Project Report, Department of Mechanical and Industrial Engineering, University of Massahusetts, Amherst, 2004. Evolution Markets LLC. Website: /http://www.evomarkets.comS. Accessed June, 2006. Gulf of Maine Coastal Program. Website: /http://gulfofmaine.fws.govS. Accessed June, 2006.