Energy return on investment (EROI), economic feasibility and carbon intensity of a hypothetical Lake Ontario wind farm
With an average annual growth rate of over 20% in recent years, wind energy is now the fastest growing energy source in the U.S. To date, although notable projects such as Cape Wind in Nantucket Sound and Bluewater Wind off Long Island are proposed, wind power projects in the U.S. have been entirely limited to onshore locations. Recent market trends toward offshore wind in Europe, the world leader in wind energy development, suggest that future applications of offshore wind will be utilized in the U.S. as well. Offshore wind power is currently more expensive to generate than from comparably windy sites on land. However, turbine technology and construction processes are improving such that the economics of offshore wind are now more encouraging. As additional onshore locations are utilized and the mitigation of aesthetic impacts continues to be important to project viability, offshore locations will become increasingly practical and profitable.
As in other states, wind energy capacity has increased dramatically in New York State (NYS) in recent years. As of October, 2006, NYS was ranked 15th in the nation with an installed capacity of 280 MW. While over 100 MW are planned near-term, no concrete plan exists to tap the offshore winds on Lake Ontario, a major region of wind resource potential in New York State (see Figure 1). With average wind speeds greater than 7 m/s, Lake Ontario poses significant opportunity for future wind energy development in NYS.
With the idea of reducing costs and optimizing project feasibility, an appropriate location for a Lake Ontario wind farm is determined to be offshore from the Ginna nuclear power plant in Ontario, NY. Assuming some amount of existing electrical infrastructure (transformers, high voltage transmission lines) may be utilized, Ginna is considered to be an optimal location for the farm where cost efficiencies can be realized that might not be elsewhere. With the Ginna location in mind, analyzing the economic feasibility of a Lake Ontario wind farm is a central goal of this analysis. Additionally, net energy analysis is applied to provide an accurate account of the energy return on investment (EROI) and carbon dioxide emissions expected from such a plant.
There are many appropriate onshore locations yet to be utilized for wind power generation in NYS, and the majority of development in the coming years will take place in these areas. However, given the state’s wind resource distribution and continuing trends toward offshore wind in European markets, tremendous potential exists for utility-scale wind development to move offshore in NYS in the near future. As offshore wind power is still more costly to produce, it is crucial to find a strategic offshore location where enhanced cost competitiveness and suitable average wind speeds can be realized.
Studies indicate that electrical interconnection constitutes as much as 17% of the total installed cost of offshore wind farms. Reducing this cost by utilizing existing transformers and high-voltage transmission lines can significantly improve the payoff from offshore wind energy. Construction cost savings of this type are at the heart of the location chosen for a hypothetical Lake Ontario wind farm in this analysis. Located adjacent to the Ginna nuclear facility in Ontario, NY, the farm could assumedly utilize existing infrastructure to deliver electricity to the power grid.
Regional winds at and offshore from the Ginna site are predominantly from the west/southwest. From the modeled NYS wind resource map (Figures 1 and 2), an average annual wind speed of 7.5 m/s is predicted at 70 m height offshore from Ginna. Wind speeds greater than 7 m/s are generally considered suitable for utility-scale wind power generation and are similar to those found at locations of other utility-scale projects in NYS. Of course, greater wind speeds can produce more electricity and thus increase the cost competitiveness of an offshore wind farm. This would favor moving the farm north into areas of greater wind speeds. However, water depths increase from about 12 m to 30 m one to two miles offshore from Ginna. As construction costs become more expensive in deeper water, an appropriate balance between utilizing greater wind speeds and minimizing construction costs must be achieved.
Using early offshore wind farms in Europe as a guide, a capacity of 75 MW is chosen for the Lake Ontario farm. This is large enough to contribute significantly to NYS’s renewable generation capacity while not consuming an extensive portion of the lake. To capitalize on greater wind speeds further offshore and reduce the effect of land-based obstacles on wind speed in times when the wind is blowing from the south/southwest, a distance from shore of 1/2 mile was chosen. The farm would continue north for about another 1.65 miles, with a turbine spacing of four rotor diameters (RD) in the north-south direction and nine or ten turbines total in the row (depending on turbine choice, see next section). Turbine construction would be required in water depths from about 12 m to 30 m. Turbine spacing in the east-west direction would be 8 RD (~1.55 miles) in order to help minimize wake effect losses, as prevailing regional winds are from the west. Rows would consist of five turbines in this direction. The shape of the farm would closely resemble a square with an area of about 2.5 square miles (Figure 4).
Two existing utility-scale wind projects in NYS were used as references in selecting the types of turbines for this analysis. The Maple Ridge and Fenner wind farms—both in areas of similar wind resource potential as the Lake Ontario site—utilize the Vestas V82 1.65 MW and General Electric (GE) 1.5 MW turbines, respectively. Since both turbines are produced by industry leaders, have proven successful in areas of NYS with moderate wind speeds, and have a multi-megawatt capacity suitable for offshore use, they were considered appropriate candidates for a wind farm. Additionally, both turbines incorporate reactive power control technology that makes them more favorable to grid managers.
The GE farm concept consists of 50 turbines in a spacing of ten by five, as described in the farm layout section. The Vestas farm concept consists of 45 turbines (74.25 MW capacity) in a spacing of nine by five. Since the Vestas turbine has a rotor diameter that is five meters larger than the GE turbine, the two farm layouts have a similar area, even though fewer Vestas turbines are required. For both turbines, a monopile type foundation is assumed. This consists of a single steel tube-like foundation that is driven into the seabed beneath the turbine.
Analysis and Results
|Table 1: Per turbine energy production values assumed for GE and Vestas turbines.|
| ||GE 1.5 MW sl||Vestas V82 1.65 MW|
|Annual gross production (kWh)||5,000,000||5,926,140|
|Gross capacity factor||38%||41%|
|Annual net production (with 10% losses)||4,500,000||5,333,526|
|Net capacity factor||34%||36.9%|
Information in the manufacturer-supplied turbine brochures is used to calculate the expected energy output from the GE 1.5 MW and Vestas V82 1.65 MW wind turbines (see Further Reading). The power curves for both turbines assume a standard Rayleigh distribution of wind speeds. A Rayleigh distribution is also assumed for the wind resource potential offshore from the Ginna facility, with an average annual wind speed of 7.5 m/s (Figures 1 and 2). Using these parameters, the GE turbine is expected to produce roughly 5 million kWh/year With a stated gross capacity factor of 41% at an average annual wind speed of 7.5 m/s, the Vestas turbine is expected to produce about 5.9 million kWh/year. For both turbines, a 10% loss is assumed from their baseline power curve expectations: 5% due to unavailability and transformer losses and another 5% due to wake effect losses from the wind farm arrangement (Table 1).
Two life cycle analyses (LCA) of offshore wind turbines are used as references to approximate the energy required to manufacture, transport, construct, operate and maintain, and decommission the GE and Vestas turbines proposed for the Lake Ontario wind farm. “The Energy Balance of Modern Wind Turbines,” published in 1997 in Wind Power Note, a Danish wind journal, contains a LCA for a hypothetical 1.5 MW offshore wind turbine. “Life Cycle Assessment for Onshore and Offshore Sited Wind Farms,” produced by Elsam Engineering, contains a LCA for existing Vestas 2 MW offshore turbines in operation at the Horns Rev wind farm in Denmark.
Table 2 provides a technical description of the turbines assessed in both LCA studies, a technical description of the GE and Vestas turbines, as well as the assumptions used to calculate the energy balance for both turbine options. In both LCA studies, a project lifetime of 20 years is assumed. As a Lake Ontario farm is predicted to operate for 30 years, a recalculated 30-year energy demand is applied to both LCA studies. Since operation and maintenance (O&M) is the only significant variable energy cost during a turbine’s lifetime, the recalculated 30-year energy demand is performed by calculating a per-year O&M energy cost and extending it another 10 years. The new 30-year energy demand is then normalized on a per MW basis to help correct for the difference between the 1.5 MW turbine assessed in the Wind Power Note study and the 2 MW turbine assessed in the Elsam study. (It should also be noted that energy costs for transportation and decommissioning are not included in the Wind Power Note study. A review of LCA literature for wind turbines suggests these life stages account for only a small portion of the lifetime energy cost: roughly about 2% each. Thus, the energy demand value used in this analysis is 4% larger than originally published in Wind Power Note.)
Although the energy demands used from the Wind Power Note and Elsam studies are normalized on a per-MW basis, it is not assumed nor suggested that the energy requirement for a wind turbine is linearly proportional to its capacity rating. No evidence suggests that this is the case; rather, it can be assumed that larger turbines will require proportionally less energy during their life cycles than smaller turbines. This is a likely result of economies of scale energy cost savings. However, each of the four turbines analyzed (in the Wind Power Note, Elsam and this analysis) are within a half-megawatt of capacity rating. Thus, the margin of error will be much smaller than if an analysis of a 100 kW turbine was performed using this method. Additionally, the goal of this analysis is to provide a reasonable approximation of the energy requirement and energy return on investment (EROI) expected for turbines used in a Lake Ontario wind farm. As the farm remains hypothetical, a close approximation is the extent of what can sensibly be achieved, and certainly well suited for this initial analysis.
Table 3 provides a breakdown of the predicted energy requirements of the GE 1.5 MW and Vestas 1.65 MW turbines using the values from the Wind Power Note and Elsam studies. In relation to their expected lifetime power production, an EROI of between 28.3 and 36.7 is calculated for the GE turbine and between 30.5 and 39.6 for the Vestas turbine. The average energy requirement is about 4.22 million kWh for the GE turbine and 4.64 million kWh for the Vestas turbine. The energy payback period for both turbines is less than one year.
In 2003, the most recent year state electricity profiles were compiled by the Energy Information Agency, NYS generated over 137 million MWh of electricity. The breakdown of generation by energy source was 17.1% coal, 14% petroleum, 20.5% natural gas, 29.6% nuclear, 17.6% hydroelectric, and 1.9% renewables. These sources combined to produce 56.9 million metric tonnes of carbon dioxide (CO2) in NYS in 2003 alone, representing an average rate of 413 g CO2/kWh. A meta-analysis of the CO2 intensities of wind turbines by Lenzen and Munksgaard (2002) determined an average life cycle CO2 requirement of 26 g CO2/kWh. This corresponds to a reduction of 387 g CO2/kWh compared to the average energy supply and subsequent CO2 emissions in NYS.
With 50 turbines producing about 4.5 million kWh each per year, the GE Lake Ontario wind farm would generate 225 GWh of electricity each year, and 6,750 GWh over a 30-year lifetime. This amount of wind-generated electricity would reduce annual CO2 emissions in NYS by about 87 thousand metric tonnes, and lifetime emissions by over 2.6 million metric tonnes. With 45 turbines producing about 5.3 million kWh each per year, the Vestas Lake Ontario wind farm would generate about 240 GWh of electricity each year, and 7,200 GWh over its lifetime. This would reduce CO2 emissions in NYS by about 92 thousand metric tonnes annually, and almost 2.8 million metric tonnes over a 30-year period. While these numbers seem significant, it is important to remember that NYS currently emits about 57 million metric tons of CO2 each year from electricity production alone. An additional 75 MW of wind-generated power capacity in NYS offers significant reductions in greenhouse gas emissions compared to the same capacity added under the current average energy supply in the state. However, much larger-scale implementation of cleaner electricity-producing technologies is necessary to seriously combat future emissions of CO2 in the state.
As specific turbine prices and installation costs are generally reserved for developers and contractors, a general figure is used to create a reference case economic scenario. The Danish Wind Power Association, who represent leaders in offshore wind development, estimates an average installed cost for offshore wind turbines of US $1.7 million/MW. This includes turbine construction, electrical interconnection, foundation costs, and other installation fees. As the exact savings that might be realized by utilizing Ginna infrastructure is unknown, a value of $1.5 million/MW is assumed for this analysis. This value is in the lower end of the range of capital costs for existing offshore wind farms in Europe.
Table 4 provides a list of assumptions used in the economic analysis for the Lake Ontario wind farm. For both the GE and Vestas turbine cases, the net present value of the installed capital cost, interest, sales tax and lifetime operation and maintenance is close to $200 million. Electricity from the farm would be sold on the NYS wholesale competitive electricity market at a price of about 5.8 ¢/kWh. Based on the expected generation from the GE and Vestas turbines, this corresponds to a gross annual revenue of about $13 million for the GE farm and $13.9 million for the Vestas farm.
Influential in the economic merit of the Lake Ontario farm is the inclusion or exclusion of the 1.8 ¢/kWh federal Production Tax Credit (PTC) during the first ten years of operation. If included, the PTC amounts to over $4 million in tax credits each year for both the GE and Vestas turbine cases. For the GE turbine case, the PTC determines a rate of return of either 19.9% or 42.8% and levelized production cost of 4.7 ¢/kWh or 5.3 ¢/kWh. For the Vestas turbine case, the rate of return is between 29.2% and 53.9% and the levelized cost of production is between 4.5 ¢/kWh and 5.1 ¢/kWh, depending on the PTC. Thus, in order to optimize the profitability of such a project, it is necessary to begin construction in a period when it is sure to qualify for the PTC.
Figures 5 and 6 show the net payments and revenue from the two Lake Ontario wind farm cases. For the GE turbine case, a payback period of between 17 and 24 years is calculated, with a net profit of between $38.6 and $83.1 million over the 30-year lifetime. Again, this range is dependent on the PTC. For the Vestas turbine case, the payback time is between 15 and 22 years, with a net profit of between $56 and $103.6 million.
While this analysis provides a reasonable approximation of the energy, CO2 and economic assessment of a Lake Ontario wind farm, its accuracy can be confirmed and refined by additional information. Better knowledge of the Ginna offshore site itself, including measured wind speeds (through anemometers), detailed water depths and the degree to which Ginna infrastructure may be utilized will add concrete viability to the energy and economic analyses.
Though both turbine choices produce positive results, the Vestas turbine appears to be the better option in all areas of analysis. This is likely due to the additional generating capacity it has over the GE turbine and an economies of scale benefit. Thus, actual turbine choice for a Lake Ontario wind farm should consider the benefits of larger turbines with bigger rotor diameters, better ability to function in moderate wind speeds, and more power capacity. Other turbines may prove to perform better than the Vestas 1.65 MW is shown in this analysis, including larger GE units currently in development.
- Cleveland, Cutler, Ida Kubiszewski and Peter K. Endres. 2006. Energy return on investment (EROI) for wind energy. Encyclopedia of Earth. Eds. Peter Saundry. (Washington, D.C.: Environmental Information Coalition, National Council for Science and the Environment).
- The Energy Balance of Modern Wind Turbines. Wind Power Note (16). Editor Soren Krohn, December 1997.
- Life Cycle Assessment for Onshore and Offshore Sited Wind Farms. Elsam Engineering, translated by Vestas Wind Systems A/S, Alsvej 21, 8900 Randers, Denmark. October 2004.
- M. Lenzen and J. Munksgaard. Energy and CO2 life-cycle analyses of wind turbines-review and applications. Renewable Energy, Volume 26, Number 3, July 2002, pp. 339-362.
- Offshore Wind Farm Layout Optimization (OWFLO) Project: An Introduction. University of Massachusetts-Amherst, Mechanical and Industrial Engineering Department. Engineering Laboratory Building, Box 32210, Amherst, MA 01003.
- GE 1.5 MW Series Wind Turbine Brochure.
- Vestas V82 1.65 MW Wind Turbine Brochure.