Spatial Analysis of Wave Power Potential in North Carolina by Jeffrey D

Spatial Analysis of Wave Power Potential in North Carolina by Jeffrey D. Bower

Dr. Lincoln Pratson, Advisor

May 2010

Masters project submitted in partial fulfillment of the requirements for the Master of Environmental Management degree in the Nicholas School of the Environment of Duke University 2010

ABSTRACT

North Carolina is actively pursuing electricity generation from various renewable energy technologies. In 2008 the state’s General Assembly commissioned a study by UNC-

Chapel Hill researchers assessing the feasibility of offshore wind power and as a direct result of the study, a project is currently under development in the Pamlico Sound. This analysis is a complementary study evaluating the potential for the use of wave energy conversion devices as another renewable energy technology in North Carolina’s coastal waters.

Using GIS, this study assesses the spatial distribution of wave power potential offshore and identifies ecological and human use conflicts prohibitive of development.

After calculating generation potential and estimating the costs related to wave farm construction, the economic feasibility of wave power is assessed, based on levelized cost of generation (LCOG). The economic analysis employs the same LCOG model used by the UNC offshore wind study team to provide directly comparable results.

The results of the analysis indicate that while there are areas off the coast of North

Carolina where wave power development is technically feasible, it is not nearly cost- competitive with other types of generation, including offshore wind. Wave power technology is still in an early development phase, however, and if anticipated future cost savings and efficiency improvements are realized, wave power may become more economically feasible.

TABLE OF CONTENTS I. Introduction...... 1 A. Wave power resource potential ...... 1 B. Statewide context ...... 2 1. Marine renewable energy in North Carolina, UNC wind study...... 3 II. Project overview...... 4 III. Methodology ...... 6 A. Electricity generation potential...... 6 1. Spatial conflicts to development ...... 6 2. Wave resource potential...... 10 3. WEC device performance ...... 13 B. Wave farm cost...... 19 1. Device costs ...... 20 2. Transmission cost...... 21 3. Operation and maintenance costs ...... 27 C. Overnight cost of plant construction...... 28 D. Levelized cost of generation...... 28 IV. Results...... 31 A. Electricity generation potential...... 31 1. Spatial conflicts to development ...... 31 2. Wave resource potential...... 34 3. WEC device performance ...... 37 B. Overnight cost of plant construction ...... 39 C. Levelized cost of generation...... 41 V. Discussion ...... 43 A. Spatial results ...... 43 B. Economic results...... 44 VI. Recommendations ...... 46

ii LIST OF FIGURES

Figure 1. Total U.S. available incident wave energy flux...... 2 Figure 2. Wave buoys used in interpolation analysis...... 11 Figure 3. AC and DC transmission cost comparison by distance from shore, based on cost estimates from the Cape Wind DEIS ...... 24 Figure 4. Substations identified for grid interconnection...... 26 Figure 5. Ecological/environmental conflict areas identified in UNC offshore wind study...... 32 Figure 6. Human use conflicts identified by UNC offshore wind study, with appropriate buffers ...... 33 Figure 7. Complete exclusion zone and MMS blocks eligible for development ...... 34 Figure 8. Mean wave height values interpolated from buoy data ...... 35 Figure 9. Mean wave period interpolated from buoy data...... 35 Figure 10. Mean wave power calculated with interpolated values...... 36 Figure 11. Estimated mean wave power, in kW/m, from Defne, et al...... 37 Figure 12. Annual energy absorbed by hypothetical wave farm...... 38 Figure 13. Overnight cost of construction, high cost estimates...... 40 Figure 14. Overnight cost of construction, low cost estimates ...... 41 Figure 15. Levelized cost of generation, high capital cost, low efficiency estimates...... 42 Figure 16. Levelized cost of generation, low capital cost, high efficiency estimates...... 42

iii LIST OF TABLES

Table 1. Electric Power Industry Generation by Primary Energy Source, 2008 ...... 2 Table 2. North Carolina REPS requirements...... 3 Table 3. NDBC wave buoy stations used in analysis...... 12 Table 4. EPRI analysis: San Francisco site annual occurence of hours per sea-state...... 15 Table 5. EPRI analysis: Pelamis wave energy conversion absorption performance (kW) in each sea-state (excluding power take off losses) ...... 16 Table 6. EPRI analysis performance assumptions for pilot and commercial phase Pelamis plants...... 17 Table 7. Pelamis WEC device cost components for proposed demonstration and commercial plants...... 20 Table 8. Cape Wind Project transmission cost estimates - AC option...... 23 Table 9. Cape Wind Project transmission cost estimates - DC option...... 23 Table 10. Transmission cost summary, 2008$...... 24 Table 11. Input assumptions for "Cost of Generation Calculator" LCOG model ...... 30 Table 12. Estimated annual electricity generation, demonstration- and commercial-stage plants ...... 38 Table 13. Cost comparison of wave power to other electricity sources...... 45

iv LIST OF ACRONYMS USED

DEIS Draft Environmental Impact Statement GIS Geographic Information System EPRI Electric Power Research Institute LCOG Levelized Cost of Generation MMS Minerals Management Service NDBC National Data Buoy Center NOAA National Oceanic and Atmospheric Administration PCM Power Conversion Module REPS Renewable Energy and Energy Efficiency Portfolio Standard WEC Wave Energy Converter/Conversion

v I. Introduction With increasing concern regarding the environmental consequences of the world’s dependence on carbon-based fuels, the energy industry is constantly exploring alternative, renewable sources of electricity generation. Scientists are investigating ways to harness all types of natural processes to generate clean, sustainable electricity. Over the last few decades, wind and solar power development have been the leading renewable technologies, and these industries have rapidly matured to commercial-scale. In recent years, however, developers have increasingly been looking for ways to harness the powerful and consistent forces of the ocean. Once such strategy is wave power.

This study will examine the feasibility of wave power development in one particular marine environment – the costal waters of North Carolina. The analysis will examine existing marine uses and sensitive ecological areas to identify areas incompatible with wave power development, calculate the electricity generation potential offshore, and evaluate the economic feasibility of development.

A. Wave power resource potential There is a very significant wave resource in the waters of the United States. The most commonly cited estimates show that on an annual basis, approximately 2,100 terawatt-hours (TWh) of wave energy are available offshore (see Figure 1). By comparison, the entire retail electricity market in 2009 totaled approximately 3,500 TWh.1

It is this tremendous potential that prospective wave power developers and investors are looking to harness.

1 Retail sales of electricity to ultimate customers: Total by end-use sector. (2010). Retrieved from http://www.eia.doe.gov/cneaf/electricity/epm/table5_1.html.

1 The development and testing of wave energy conversion (WEC) devices has accelerated in recent years, with the first commercial wave power project in U.S. waters currently under development off the coast of Reedsport, Oregon. Ocean Power Technology is developing a 1.5 megawatt (MW) wave park consisting of 10 buoys which convert the motion of the waves into electricity.2

Figure 1. Total U.S. available incident wave energy flux3

B. Statewide context North Carolina’s current electricity generation portfolio is dominated by coal and nuclear plants (see Table 1). In recent years, however, the state has taken some strong legislative steps to promote the development of renewable sources.

Table 1. Electric Power Industry Generation by Primary Energy Source, 20084 Fuel Source Electricity Generated (MWh) Percentage Share Coal 75,814,787 60.5 Petroleum 320,221 0.3 Natural Gas 4,177,342 3.3 Nuclear 39,776,280 31.8 Hydroelectric 3,033,642 2.4 Other Renewables 1,922,213 1.5 Pumped Storage -121,064 -0.1 Other 315,642 0.3

2 Reedsport OPT wave park. (2010). Retrieved April 20, 2010, from http://www.oceanpowertechnologies.com/reedsport.htm 3 Bedard, R., Hagerman, G., Previsic, M., Siddiqui, O., Thresher, R., & Ram, B. (2005). Offshore wave power feasibility demonstration project: Final summary report. Electric Power Research Institute. 4 North Carolina electricity profile. (2008). Retrieved from http://www.eia.doe.gov/cneaf/electricity/st_profiles/north_carolina.html.

2 In 2007, then-Governor Mike Easley signed Session Law 2007-397, establishing a renewable energy and energy efficiency portfolio standard (REPS). A first of its kind in the southeastern U.S., the REPS requires that utilities in North Carolina derive a certain percentage of their electricity from renewable sources. The requirements begin at 3% in

2012 and escalate to 12.5% in 2020 (see Table 2).

Table 2. North Carolina REPS requirements5 Requirements for Requirements for Year Electric Cooperatives and Investor-Owned Utilities Municipal Utilities 2012-2014 3% 3% 2015-2017 6% 6% 2018-2020 10% 10% after 2020 12.5% 10%

The effect of the REPS bill has been the rapid exploration of several different renewable energy solutions for the state. Several biomass and small solar projects have been developed, and the General Assembly is actively debating a wind farm permitting bill which would allow commercial-scale wind facility development.

1. Marine renewable energy in North Carolina, UNC wind study Part of the statewide effort to develop renewable electricity sources has included the investigation of marine renewable energy, specifically offshore wind power. In the summer of 2008 the North Carolina General Assembly asked researchers at the University of North Carolina-Chapel Hill to perform a feasibility study on the development of wind turbines in the Pamlico and Albemarle Sounds and in the state’s offshore waters. The resulting study was completed in June 2009 and was titled “Coastal Wind: Energy for North

Carolina’s Future.”6

5 A citizen's guide: The North Carolina renewable energy and energy efficiency portfolio standard. (2009). North Carolina Sustainable Energy Association. 6 Coastal wind: Energy for North Carolina's future. (2009). University of North Carolina, Chapel Hill.

3 This comprehensive study included the first detailed assessment of the wind resource potential in North Carolina’s waters. Using meteorological data and advanced wind modeling techniques, researchers determined the areas of highest resource and estimated the electricity generation potential in those areas. In addition, the researchers performed extensive spatial analysis to determine areas of competing use and environmental sensitivity. Combining these analyses, the study identified target locations for offshore wind development, areas of high wind resource and low human and environmental impact.

UNC’s offshore wind study concluded that North Carolina’s state waters feature several areas suitable for offshore wind development. As a direct result of the study, Duke

Energy has partnered with UNC to develop an offshore wind demonstration project to consist of one to three turbines in North Carolina’s Pamlico Sound.7

II. Project overview Considering the quick and favorable response to UNC’s offshore wind study, it is the goal of this analysis to provide a complimentary analysis evaluating the potential for wave power development. This study will offer wave power as an additional marine renewable energy technology for consideration.

Specifically, this analysis has two primary goals:

• Assess the technical feasibility of wave power development based on wave resource potential and spatial conflicts • Determine the economic feasibility of wave power development on overnight cost of plant construction and levelized cost of generation.

7 Bonner, L. (2009, September 26). Pamlico Sound towers to test wind energy. News & Observer.

4 In order to aid in the evaluation of the technical and economic feasibility of wave power of the coast of North Carolina, a geographic information system (GIS) was used for several parts of the assessment. The spatial analysis component has three main goals:

• Identify spatial conflicts: There exist current marine uses and ecologically sensitive areas that would preclude wave power development. The spatial analysis will use a multi-criteria analysis to identify and exclude such areas. This portion of the analysis will be largely adopted from the extensive analysis performed by UNC in the offshore wind feasibility study.

• Determine electricity generation potential: A key goal of the spatial analysis component is to calculate the wave power potential for all waters offshore North

Carolina. This data will be used in concert with published WEC device performance data to spatially locate areas of high electricity generation potential.

• Calculate spatially variable cost components: The construction of a wave farm project features several cost components that are constant regardless of the location of the farm. The cost of transmission, however, varies with distance from the grid. This portion of the analysis is designed to calculate this variable cost component.

This study will use a hypothetical wave power plant as a basis for analysis. The plant will consist of 96 Pelamis wave devices. The Pelamis device was chosen because it was the first device used in a commercial-scale project, in Portugal in 2005. Additionally, the Pelamis device was identified as the most commercial-ready device by the Electric

Power Research Institute (EPRI) in its 2005 feasibility study of wave power in the U.S.8

8 Bedard, Hagerman, Previsic, Siddiqui, Thresher, & Ram.

5 Pelamis devices are situated in an array with 150 meter spacing between the devices and can be developed with three rows of the devices.9 Based on these parameters, it was determined that 96 devices could be moored within one Minerals Management

Service (MMS) lease block. MMS blocks are the standard unit for offshore activities because any development requires the leasing of these blocks.

Each Pelamis wave device has a rated capacity of 750 kilowatts (kW). Therefore, the hypothetical 96 device wave farm used in this analysis has a rated capacity of 72 MW.

More details of wave device performance are described below in the methodology section.

Using this hypothetical 72 MW wave farm as a basis for analysis, this study will evaluate the electricity generation potential of such a plant as well as its economic feasibility using the methodology outlined below.

III. Methodology A. Electricity generation potential The first goal of this analysis stated above is to determine the potential for and technical feasibility of electricity generation by WEC devices off the coast of North Carolina.

First, spatial conflicts that would preclude wave power development are identified.

Second, the wave resource potential is calculated, representing the inherent power in the waves. Finally, electricity generation potential is calculated based on data on the performance of the Pelamis WEC devices.

1. Spatial conflicts to development As the development of offshore energy generation facilities is a relatively new concept in the U.S., special care must be taken in order to assess any conflicts that might

9 Previsic, M., Bedard, R., & Hagerman, G. (2004). Offshore wave energy conversion devices. Electric Power Research Institute.

6 arise due to the siting of the project. Specifically, there are two main categories of spatial conflicts: Human use and environmental/ecological.

As part of its assessment of offshore wave power, the UNC researchers completed a comprehensive spatial assessment of current human uses of the marine environment off the coast of North Carolina as well as any ecologically sensitive areas not compatible with offshore wind power development. The assessment drew upon experts in many natural science fields with specialized knowledge of North Carolina’s coastal ecosystems.

Together, the research team developed criteria that would preclude offshore wind development.

Due to the extensive and comprehensive nature of their assessment, this analysis will adopt the general structure of the UNC spatial feasibility study as well as many of the specific criteria identified therein. Exceptions will be made for differences between wave and wind power devices.

a) Ecological and environmental conflicts The waters off North Carolina’s coast are rich and productive, but they are also ecologically sensitive. Any energy development projects need to consider the impact on the fish, turtles, marine mammals, and benthic invertebrates. In their evaluation of offshore wind power feasibility, the UNC researchers developed a complete list of environmental criteria that would conflict with offshore energy facilities. This was based on extensive research and the guidance of experts on the North Carolina coastal environment, such as the North Carolina Division of Marine Fisheries. Since the study evaluated wind power, a major component of the analysis included impacts on coastal birds and bats. That is not a major concern for wave power development, so this analysis

7 will consider only impacts on the aquatic environment. The UNC study identified the following areas that would not be compatible with wind power development:

• Primary and secondary nurseries • Strategic habitat areas (SHAs) • Artificial oyster reef sanctuaries • Oyster cultch planting areas • Shell bottom • Live bottom • Wreck habitats • Submerged aquatic vegetation (SAV)

The UNC team processed these potential conflicts and classified areas as “High,”

“Medium,” or “Low” conflict zones and created a composite GIS layer for each classification.

For this analysis of wave power feasibility, the UNC results were fully adopted. The data layers were acquired from the authors of the UNC report, and the data layer representing the areas of “High” fish habitat conflict was used as an exclusion zone precluding wave power development.

b) Human use conflicts The UNC offshore wind study also evaluated the impact of offshore energy facilities on competing human uses. The North Carolina coast is very active with many commercial, recreational, and industrial uses and since ocean waters are typically not highly regulated, and the process of delineating areas and excluding certain activities is a challenge that all types of offshore energy development will face. Marine activity is largely unrecorded, so spatial data regarding types and levels of use is sparse. Certain human activities offshore

8 can be mapped, and the following data layers were identified by the UNC researchers as those that would exclude wave farm development:

• Ferry routes • Major shipping lines • Major navigation channels • Intracoastal Waterway • Dredge material disposal sites • Shipwrecks • Cape Hatteras National Seashore • Monitor National Marine Sanctuary Conspicuously absent from these areas are major commercial and recreations fishing grounds. Again, data on the specific location of heavily fished areas is sparse, but the UNC researchers were able to identify some areas of high conflict based on interviews with industry experts and individual fishermen. The areas identified were incorporated in the fish conflict layer mentioned above.

GIS data layers for the criteria listed above were acquired from publicly available sources and appropriate buffers were applied to each, as determined by the UNC researchers. For example, major shipping and navigation routes were assigned a 1 kilometer buffer where energy facilities can not be developed. A full list of these data layers and their sources is included in the attached Appendix A.

Once the areas of spatial conflict were identified and appropriately buffered, the data layers were combined into a general exclusion zone layer. Any MMS lease block that overlapped with the exclusion zone was determined to be unsuitable for development and was eliminated from further analysis.

9 2. Wave resource potential A key aspect of the feasibility of wave power development is a suitable wave resource. WEC devices operate by converting the inherent power of waves into electricity to be delivered to the grid. Device performance is measured in kilowatts and is typically represented in a power matrix such as Table 5 on page 16. The amount of electricity generated is dependent on two characteristics of the sea-state: Significant wave height and peak wave period.

Significant wave height is measured in meters and is approximately equal the average of the highest one-third of waves over a specified time period. Peak (or dominant) wave period is measured in seconds and is defined as “the reciprocal of the frequency, fp

(peak frequency), for which spectral wave energy density is a maximum.”10

Since significant wave height and peak wave period are the key indicators of WEC device performance, it is necessary to estimate these measurements for all areas under consideration. Therefore, in order to create an estimation of electricity generation potential for all areas offshore North Carolina, GIS layers for each wave measurement were created based on data gathered from offshore wave buoys. The National Ocean and

Atmospheric Administration (NOAA) operates the National Data Buoy Center, a publicly- available internet database of ocean buoy data. The database is constantly updated with data on climate and ocean conditions from hundreds of buoys in U.S. territorial waters.

For this analysis, 15 buoys from the NDBC were chosen from those offshore Virginia,

North Carolina, and South Carolina. These buoys were chosen because they contained

10 Nondirectional and directional wave data analysis procedures. (1996). Retrieved from http://www.ndbc.noaa.gov/wavemeas.pdf.

10 archived data of wave height and period, and because of their location in or near the chosen study area. Figure 2 shows the locations of the buoys selected.

Figure 2. Wave buoys used in interpolation analysis.

Historical data is available for the operational life of each buoy through the NDBC website. Data was downloaded for each buoy for all available years 1998 or later. These data were processed, removing any data entries for which the values were 0 or 99 (null values). Once the data were processed, mean values for significant wave height and dominant wave period were calculated. These mean values are displayed in Table 3.

Table 3. NDBC wave buoy stations used in analysis Mean Significant Mean Dominant Station Data Station Name Wave Height Wave Period ID Available (meters) (seconds) 41001 150 NM East of Cape Hatteras 1998-2008 2.06 8.06 41002 S HATTERAS – 250 NM East of 1998-2008 1.88 8.10 Charleston, SC 41004 EDISTO – 41 NM Southeast of 1998-2009 1.32 7.30 Charleston, SC 41013 Frying Pan Shoals, NC Buoy 2003-2009 1.34 7.60 41025 Diamond Shoals 2003-2009 1.52 7.79 41036 Onslow Bay Outer, NC 2006-2009 1.33 7.32 41048 W Bermuda 2007-2009 1.90 8.73 41110 Masonboro Inlet, NC 2008-2009 0.96 7.81 44014 Virginia Beach - 64 NM East of 1998-2009 1.44 7.84 Virginia Beach, VA 44099 Cape Henry, VA 2008-2009 1.01 8.22 44100 Duck FRF 26m, NC (430) 2008-2009 1.20 8.47 DSLN7 Diamond Shls Lt., NC 1998-2003 1.31 8.38 DUCN7 Duck Pier, NC 1998-2008 0.84 9.01 FPSN7 Frying Pan Shoals, NC 1998-2008 1.34 7.65 SMBS1 Springmaid pier, SC 2007-2009 0.38 8.45

Using the buoy locations as base points and the mean values for each wave measurement, values were interpolated for each wave metric. The ArcGIS Trend interpolation technique was used, using a 2nd order polynomial, linear regression method with a resolution of 50 meters. This technique was chosen because it most accurately captured the sea-state characteristics offshore North Carolina. The products of this operation were two raster grids: interpolated significant wave height values and interpolated dominant wave period values.

A layer of polygons representing MMS lease blocks was overlaid on top of these rasters. Using the Zonal Statistics as Table tool, the wave height and period values were summarized for each MMS lease block. This operation yielded a table with an entry for each MMS lease block and the average of the significant wave height values for all raster grid cells within that MMS block, as well as a similar value for dominant wave period.

12 These values will be used directly in the calculation of electricity generation potential by the WEC devices (described below). However, in order to compare the results of the interpolation with those published elsewhere, the wave height and period data layers were also used to calculate incident wave power. Wave power is measured in kilowatts per meter (kW/m) and was calculated using the following equation:

J = 0.42 x (Hs)2 x Tp where J is incident wave power (kW/m) Hs is significant wave height (meters) and Tp is dominant wave period (seconds).11

3. WEC device performance A central component of determining the feasibility of wave power is to assess how the devices will perform offshore North Carolina. Electricity generation from wave power, like many renewable energy sources, is highly variable and depends on conditions of nature as much as the devices themselves.

To assess the performance of the Pelamis WEC devices in the modeled sea-states, this analysis will use a modified version of the methodology used in the EPRI study of wave power feasibility, which is outlined below.

a) EPRI performance methodology The wave energy feasibility study completed by EPRI assesses the production potential of a Pelamis wave farm at a specific development site off the cost of San Francisco.

While this is a sit-specific analysis, a major objective of the EPRI study was to develop a thorough methodology for feasibility assessment. In addition to the San Francisco site evaluation, the EPRI research team produced several supporting documents detailing their

11 Hagerman, G., & Bedard, R. (2003). E2I EPRI specification: Guidelines for preliminary estimation of power production by offshore wave energy conversion devices. Electric Power Research Institute.

13 methods. Subsequent to the project’s completion in 2005, these documents have been frequently cited by other wave power feasibility studies and the methodology is often adopted for other site analyses.12,13

The EPRI methodology begins with a characterization of the sea-state at the potential site. This characterization is derived from wave buoy data and describes the sea- state as a combination of significant wave height and dominant wave period. Table 4 shows the scatter diagram of annual occurrence of hours per sea-state at the San Francisco site. For example, there are 126 hours in which the wave period, Tp, is between 7.5 and 6.5 seconds and the wave height, Hs, is between 1.75 and 2.25 meters.

12 Dunnett, D., & Wallace, J. S. (2009). Electricity generation from wave power in Canada. Renewable Energy, 34(1), 179-195. 13 Dalton, G. J., Alcorn, R., & Lewis, T. (2010). Case study feasibility analysis of the Pelamis wave energy convertor in Ireland, Portugal and North America. Renewable Energy, 35(2), 443-455.

14 Table 4. EPRI analysis: San Francisco site annual occurence of hours per sea-state14 Upper Tp: 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 14.5 17.5 20.5 Lower Tp: 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 11.5 13.5 16.5 19.5 Hs and Tp bin Tp (sec) boundaries Lower Upper Hs Hs Hs (m) 3 4 5 6 7 8 9 10 11 12 14 17 20 9.75 10.25 10 0 0 0 0 0 0 0 0 0 0 0 0 0 9.25 9.75 9.5 0 0 0 0 0 0 0 0 0 0 0 0 0 8.75 9.25 9 0 0 0 0 0 0 0 0 0 0 0 0 0 8.25 8.75 8.5 0 0 0 0 0 0 0 0 0 0 0 0 0 7.75 8.25 8 0 0 0 0 0 0 0 0 0 0 0 0 0 7.25 7.75 7.5 0 0 0 0 0 0 0 0 0 0 0 0 0 6.75 7.25 7 0 0 0 0 0 0 0 0 0 0 0 1 0 6.25 6.75 6.5 0 0 0 0 0 0 0 0 0 0 0 2 1 5.75 6.25 6 0 0 0 0 0 0 0 0 0 0 1 2 1 5.25 5.75 5.5 0 0 0 0 0 0 1 1 0 1 4 6 1 4.75 5.25 5 0 0 0 0 0 1 1 2 2 2 8 13 3 4.25 4.75 4.5 0 0 0 0 0 3 2 3 4 6 14 19 4 3.75 4.25 4 0 0 0 0 1 6 6 5 7 13 38 32 8 3.25 3.75 3.5 0 0 0 0 5 21 16 17 18 38 85 53 12 2.75 3.25 3 0 0 0 3 13 62 39 36 47 97 161 76 23 2.25 2.75 2.5 0 0 0 12 47 139 82 82 110 200 253 105 38 1.75 2.25 2 0 0 4 41 126 272 165 168 226 325 302 132 51 1.25 1.75 1.5 0 3 21 127 212 367 263 292 301 338 308 195 52 0.75 1.25 1 2 18 35 97 117 255 224 210 213 246 387 264 37 0.25 0.75 0.5 2 4 3 7 11 37 26 25 22 37 62 20 1 0 0.25 0.125 0 0 0 0 0 0 0 0 0 0 0 0 0

Pelamis Wave Power provided EPRI with the performance data for the WEC device shown in Table 5. This summarizes the power absorbed, in kW, in each sea-state excluding power takeoff losses, the device’s inherent inefficiencies in converting wave power to electricity. The maximum value is the device capacity, 750 kW. Using the values from the example above, the Pelamis device absorbs 219 kW of energy at a sea-state characterized by Tp=7 and Hs=2.

14 Previsic, M., Bedard, R., Hagerman, G., & Siddiqui, O. (2004). System level design, performance and costs for San Francisco California Pelamis offshore wave power plant. Electric Power Research Institute.

15 Table 5. EPRI analysis: Pelamis wave energy conversion absorption performance (kW) in each sea-state (excluding power take off losses) Tp (sec) 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 20 10 750 750 750 750 750 750 750 750 750 750 750 750 711 750 750 738 734 9.5 750 750 750 750 750 750 750 750 750 750 750 750 691 750 710 694 662 9 750 750 750 750 750 750 750 750 750 750 750 750 670 746 668 650 592 8.5 750 750 750 750 750 750 750 750 750 750 750 750 650 699 626 606 551 8 750 750 750 750 750 750 750 750 750 750 750 750 630 653 584 562 509 7.5 750 750 750 750 750 750 750 750 750 750 750 748 610 607 542 518 467 7 750 750 750 750 750 750 750 750 750 750 750 692 566 560 500 474 425 6.5 750 750 750 750 750 750 750 750 750 750 723 592 617 513 458 430 384 6 597 630 663 684 750 750 750 750 750 750 616 633 525 476 396 386 329 5.5 428 497 566 612 750 750 750 750 750 635 642 532 482 400 399 341 322 5 259 364 469 539 750 750 750 750 644 641 531 482 399 394 330 308 274

Hs (m) 4.5 94 233 371 467 735 744 738 634 626 520 473 390 382 319 299 250 208 4 105 216 326 394 632 616 583 585 494 454 374 361 339 283 236 197 153 3.5 0 86 211 326 484 577 568 502 421 394 330 312 260 216 196 164 140 3 0 91 180 246 402 424 417 369 343 331 275 229 208 173 144 120 93 2.5 0 7 93 171 279 342 351 320 274 230 210 174 145 120 100 84 65 2 0 0 66 109 199 219 225 205 195 162 135 112 93 77 64 54 41 1.5 0 0 26 62 112 141 143 129 110 91 76 63 52 43 36 30 23 1 0 0 11 27 50 62 64 57 49 41 34 28 23 0 0 0 0 0.5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.125 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

To derive the device performance at the San Francisco site, the EPRI study simply

multiplied the corresponding cells together to summarize annual device performance. For

example, using the values above, there are 126 hours per year when the sea-state at the San

Francisco site shows Tp=7 and Hs=2. At that same sea-state, the Pelamis device generates

219 kW of electricity. Thus, for that particular sea-state at the San Francisco site, the

Pelamis device will generate 126 hours x 219 kW = 27,594 kWh of electricity. This

operation is executed for each cell in the scatter plot. The final value is the sum of all these

values and is described as the annual energy absorbed by the Pelamis device, in this case

1,229 MWh/year.

The EPRI study then applies certain factors to this value in order to determine the

amount of electricity that would be delivered to the grid. The device availability factor is

16 the percentage of the time that the device is functioning. The device would not be functioning during maintenance or mechanical failure. For the pilot plant, the estimated availability factor is 85%. EPRI estimates that this will increase to 95% in the commercial plant after plant operators have gained more experience and the technology has matured.

The power conversion efficiency is the percentage of the wave energy absorbed that is converted to electricity. The EPRI researchers estimated that the commercial plant would exhibit a 10% improvement over the 80% conversion efficiency of the demonstration plant with the use of more customized components in the power conversion modules. Finally, the authors of the EPRI study determined that by changing the mooring configuration originally utilized by the Pelamis devices, the overall plant could show an absorption performance improvement of 37%. These factors are summarized in Table 6.

Table 6. EPRI analysis performance assumptions for pilot and commercial phase Pelamis plants Pilot Plant Commercial Plant Device availability 85% 95% Power conversion efficiency 80% 88% Efficiency gained from mooring reconfiguration -- 37%

b) EPRI methodology modifications EPRI’s methodology requires very specific sea-state data for the potential development site. Since the goal of this analysis is wave power estimation over a large area, rather than an evaluation of a specific site, the same matrix technique could not be used. This analysis uses mean wave height and period values generalized for each MMS block instead of hours of annual occurrence for each sea-state used by EPRI. Therefore this analysis will alter the EPRI methodology to use mean values rather than the annual occurrence matrix.

17 Using the rasters of wave height and period along with the shapefile of MMS lease blocks, the ArcGIS Zonal Statistics tool was used to assign average values to each block.

The resulting table included an entry for each block with average values of all cells contained within the block for both wave measurements. Average performance values for each lease block were determined using these values with the Pelamis performance matrix

(see Table 5). The average performance value, in kilowatts, was then extrapolated over the course of the year to determine the annual energy absorbed, using the following equation:

Kilowatts x 24 (hours/day) x 365 (days/year). The result of this operation is the annual energy absorbed by one Pelamis device for each MMS lease block, according to its mean sea-state. To model the total energy absorbed by the hypothetical wave farm, these values were multiplied by 96, the number of devices in the farm.

To translate the annual energy absorbed by the wave farm to actual electricity delivered to the grid, efficiency and availability factors were applied. As previously mentioned, the EPRI study modeled two wave farms – a demonstration project and a full- scale commercial plant. As part of modifications the researchers made when modeling the full-scale plant, they assumed certain improvements in efficiency and availability (See

Table 6). The values for annual energy absorbed were reduced by these factors to create the values for annual energy produced. This final value represents the total amount of electricity, in kilowatt-hours, that is delivered to the grid.

This analysis uses the same set of assumptions for modeling the hypothetical wave farm. The early development stage project is characterized by lower efficiency and higher cost, and the later stage project has improvements to both efficiency and cost. The cost components will be described in the following section.

18 The overall result of the device performance assessment, therefore, is two separate estimates for the annual electricity generated and delivered to the grid by the wave farm, a demonstration-stage estimate and a commercial-stage estimate.

B. Wave farm cost In addition to the amount of electricity generated, a key determinant of the feasibility of wave power off the coast of North Carolina is the development and operation cost. The cost of development of a wave farm is comprised of several important cost components. Making a generalized estimation of these costs requires making several assumptions. This section outlines the costs considered for this analysis and details any assumptions made.

In the wave power feasibility analysis performed by EPRI, several cost components were considered for the proposed Pelamis wave farm: Onshore transmission and grid interconnection, subsea transmission cables, Pelamis WEC power conversion modules

(PCMs), Pelamis WEC structural sections, device mooring, installation facilities, installation, construction management and commissioning.15

Since the EPRI analysis was designed for a specific development site, rather than a generalized assessment of development feasibility such as this analysis, many of the cost components were site-specific and acquired through personal communication with local vendors and service providers. For this analysis, the plant development cost components will be limited to the following:

• Pelamis WEC devices (PCMs, structural sections, mooring)

• Subsea transmission

15 Ibid.

19 These two cost components represent the majority of the cost of plant development.

In fact, in the EPRI analyses of costs of a demonstration and commercial plant, these cost components represented 78% and 88% of the total initial capital costs, respectively.16

In addition to the initial capital costs of wave farm development, operation and maintenance (O&M) costs will be considered for the economic feasibility analysis.

1. Device costs There exist in the literature a wide variety of cost estimates for wave power plants and Pelamis plants in particular.17,18,19 The EPRI feasibility study of a Pelamis wave power plant contains the most thorough and comprehensive assessment of plant costs and is often relied upon in other analyses. Therefore, figures from that study will be used in this analysis. In the EPRI analysis, researchers investigated two possible plants: An early stage demonstration plant, and a full-scale commercial plant. The two plants differed in many respects, including capital cost estimates. Table 7 below summarizes the device component costs estimated for each plant.

Table 7. Pelamis WEC device cost components for proposed demonstration and commercial plants Component Demonstration Plant Commercial Plant Power Conversion Modules $1,565,000 $624,000 Manufactured Steel/Concrete Sections20 $850,000 $245,000 Mooring $243,000 $117,000 Total $2,658,000 $986,000

These cost estimates are significantly different. Therefore this analysis will use both estimates to present results under high and low capital cost cases.

16 Ibid. 17 Dunnett, & Wallace. 18 Dalton, G. J., Alcorn, R., & Lewis, T. 19 Previsic, Bedard, Hagerman, & Siddiqui. System level design, performance and costs for San Francisco California Pelamis offshore wave power plant. 20 The Pelamis WEC device is designed to have cylindrical steel structures between the PCMs. In the EPRI assessment, one of the cost savings envisioned by researchers for a commercial-scale plant derived from the adoption of concrete as the main structural material, rather than steel.

20 2. Transmission cost The second major capital cost is transmission. Electricity generated at any offshore power plant must be brought to shore through underwater transmission lines to be interconnected with the larger grid. This can be a considerable portion of capital expenditures for a plant. The capital cost of subsea transmission connection depends largely on two factors: cost of materials and installation, and distance to grid interconnection

a) Cost of materials and installation With the increased attention on potential offshore electricity generation, particularly offshore wind, there have been many recent publications on the cost of underwater transmission.21,22 Many of these are based on experiences in Western Europe, where offshore wind turbines are well under development. There are fewer accurate estimates for underwater transmission in U.S. waters because so far there are no marine electricity generation installations and all subsea transmission cable examples are based on distributing electricity from the mainland to islands.

The most fully developed offshore electricity project in the U.S. is the Cape Wind project, featuring 130 turbines rated at 3.6 MW each, for an installed capacity of 468 MW.

As part of the project’s draft environmental impact statement (DEIS) completed by the U.S.

Army Corps of Engineers, the project developer provided details on estimated costs of underwater transmission. These estimates were provided as Appendix 3-C of the DEIS.23

21 de AlegrÌa, I. M., MartÌn, J. L., Kortabarria, I., Andreu, J., & EreÒo, P. I. (2009). Transmission alternatives for offshore electrical power. Renewable and Sustainable Energy Reviews, 13(5), 1027-1038. 22 Li, G. (2000). Feasibility of large scale offshore wind power for Hong Kong - a preliminary study. Renewable Energy, 21(3-4), 387-402. 23 Cape Wind energy project draft environmental impact statement. (2004). U.S. Army Corps of Engineers.

21 Since the cost estimates included in the DEIS are the most comprehensive for a U.S. offshore energy project, this analysis will use the figures cited in the report as estimated costs of connecting a wave farm to the transmission grid. The Cape Wind DEIS provided cost estimates for two transmission types: Alternating current (AC) and direct current

(DC).

AC and DC transmission systems have differing material requirements and costs associated with them. The primary difference is that AC transmission requires an underwater substation to which all electricity generating devices connect. The substation then “steps up” the voltage for transmission to shore. A DC transmission system does not specifically require a substation, but instead requires AC/DC converters. Electricity is generated as AC and must be converted to DC for transmission through the underwater line. Once it reaches shore, it must be converted back to AC in order to be connected to the grid.

Several studies have examined the advantages and disadvantages of each type of transmission.24,25 AC systems are generally best suited for shorter distances and lower voltages. They are also typically the least expensive option under these conditions and are simpler to interconnect with the grid. The disadvantage of AC systems is that transmission capacity depends on the distance of the line, as transmission resistance increases the farther the generation is from the interconnection point. DC systems, on the other hand, have essentially no line losses. Despite higher capital costs - especially for the AC/DC converters – DC systems are thought to be the preferred method for large projects or those

24 Wright, S. D., Rogers, A. L., Manwell, J. F., & Ellis, A. (2002). Transmission options for offshore wind farms in the United States. American Wind Energy Association. 25 Green, J., Bowen, A., Fingersh, L. J., & Wan, Y. (2007). Electrical collection and transmission systems for offshore wind power. National Renewable Energy Laboratory.

22 sited far from shore. Most analyses provide estimates of the “break even” distance (where

DC becomes the more economical option) between 50 – 100 km from shore.

The Cape Wind DEIS provided estimates for both AC and DC systems for comparison. The values pertinent to this analysis are summarized in Table 8 and Table 9 below. The values are presented in the DEIS in 2003$ and line costs are listed as dollars per mile.

Table 8. Cape Wind Project transmission cost estimates - AC option Unit Unit Material Total Cost Item Installation Cost ($2003) Cost 115 kV AC $2,500,000 $1,200,000 $3,700,000 Submarine per mile per mile per mile cable Offshore $12,000,000 $12,000,000 Substation per station Electrical Equipment

Table 9. Cape Wind Project transmission cost estimates - DC option Unit Unit Material Total Cost Item Installation Cost ($2003) Cost 150 kV DC $750,000 $375,000 $1,125,000 Submarine per mile per mile per mile cable AC/DC $62,000,000 $124,000,000 Converter per station Stations

For consistency with other values in the current analysis, the values were converted to 2008$ and to dollars per kilometer. Costs in 2003$ were converted to 2008$ using a conversion factor of 0.877, derived from Consumer Price Index data.26 The figures were then rounded. The final costs used for this analysis are summarized in Table 10 below.

26 Sahir, R. (2009). Inflation conversion factors for years 1774 to estimated 2019. Oregon State University. Retrieved from: http://oregonstate.edu/cla/polisci/download-conversion-factors

23 Table 10. Transmission cost summary, 2008$ Item AC Costs DC Costs 115 kV AC Submarine $2,700,000 - cable (per km) Transmission Line 150 kV DC Submarine - $820,000 cable (per km) AC: Offshore Substation $14,000,000 - Other Transmission Electrical Equipment Equipment DC: AC/DC Converter - $145,000,000 Stations (2 required)

Because of the low cost of AC substations (as compared to AC/DC converters), it is the lowest cost system for areas near shore. But as distance increases, the costs of the systems converge. The convergence point is at approximately 70 km. For areas farther than 70 km from shore, DC is estimated to be the lower cost option (see Figure 3 below).

Figure 3. AC and DC transmission cost comparison by distance from shore, based on cost estimates from the Cape Wind DEIS

b) Distance to grid interconnection Based on the cost of transmission materials and installation per kilometer outlined in the previous section, it is clear that the distance of an offshore energy project is a large factor in overall capital costs. This analysis takes this fact into consideration by determining the distance of offshore areas to suitable interconnection sites onshore.

While there are transmission and distribution lines along the coast of North

Carolina, not every area is suitable for interconnection. Interconnection typically requires

24 a substation, construction of which would be cost prohibitive for a relatively small offshore energy project. An ideal interconnection spot would feature an existing substation and excess transmission capacity on the connected lines. In its analysis of offshore wind feasibility, the UNC researchers identified three suitable areas for interconnection near the coast.27 This analysis will adopt the conclusions of the UNC report and utilize the same three interconnection points.

The northern-most substation identified by the UNC study is on the Outer Banks near Kitty Hawk. This is in the Dominion service territory. The substation’s proximity to the shore makes it the optimal interconnection location for any project of the northern coastal region.

The UNC researchers also identified several substations near the shore in the central region of North Carolina’s coast near the Pamlico Sound. These substations were presented in a 2008 draft report of the North Carolina Transmission Planning

Collaborative.28 The locations were submitted by Progress Energy as part of a feasibility study of interconnection options for hypothetical wind energy project in the coastal region.

The UNC study identified four substations in this region for its analysis. For the purposes of this analysis, one substation in Morehead City was chosen for modeling potential transmission distance.

Finally, in the southern coastal region, the UNC researchers identified a suitable substation at Carolina Beach. This substation is south of Wilmington and is owned by

Progress Energy.

27 Coastal wind: Energy for North Carolina's future. 28 Report on the NCTPC 2008-2018 collaborative transmission plan [DRAFT]. (2008). North Carolina Transmission Planning Collaborative.

25 The three suitable interconnection locations are displayed in Figure 4 below.

Figure 4. Substations identified for grid interconnection.

Analyzing the potential transmission cost for any offshore wave farm requires data on distance from these substations. To accomplish this, GIS was used. The three substations were represented as an ArcGIS point file and a Euclidean distance raster was generated using these points. Using the shapefile of MMS lease blocks, the average distance from the substations was calculated for each block. A minimum average distance was used, meaning that each lease block was assigned the distance to the nearest substation. This assumes each substation is equally suitable and no priority was assigned to any.

Using the distances calculated for the lease blocks, the cost of transmission was calculated for a wave farm constructed in each block. This calculation used the cost figures outlined in the previous section and used the least cost transmission system (i.e. AC or DC).

As explained above, lease blocks within 70 km of shore were assigned AC transmission costs and those farther than 70 km were assigned DC transmission costs.

26 3. Operation and maintenance costs The operation and maintenance costs account for all activities related to the regular upkeep of the wave farm facilities and are represented on an annual basis. These costs are important to the levelized cost of generation calculations as part of the economic feasibility analysis.

As with Pelamis WEC device costs, this analysis uses the estimates from the EPRI feasibility study.29 Unlike the estimates for device costs, EPRI researchers did not make two different O&M cost estimates. Estimates were only completed for the full-scale 213- device commercial plant. For O&M labor on a plant this size, researchers concluded that it would require a 21-person maintenance crew with an annual cost of $2,584,000. Since the wave farm being modeled in this analysis (96 devices) is approximately half the size of the plant assessed in the EPRI study, annual O&M labor costs were halved, i.e. $1,292,000 per year.

Operation and maintenance costs also include device replacement parts. Because to date there have been no long-term test or commercial projects, accurate estimates for device durability and replacement needs could not be made. Instead, the EPRI analysis allocated a flat rate of 2% of the initial plant capital cost for replacement parts.

This analysis uses the same 2% estimate, so total yearly O&M costs consist of $1.3 million plus 2% of the initial plant capital costs. Thus, O&M costs differ for the high and low cost estimates.

29 Previsic, Bedard, Hagerman, & Siddiqui. System level design, performance and costs for San Francisco California Pelamis offshore wave power plant.

27 C. Overnight cost of plant construction A common metric for the economic comparison of different electricity generation types is the overnight cost of construction. This is typically expressed as cost per megawatt of installed capacity. It is essentially the sum total of all initial capital expenditures divided by the plant capacity.

For this analysis, the cost components that comprise the overnight capital cost are plant costs and transmission costs (and not O&M costs). Plant costs, as described above, are constant for construction at any location offshore. Transmission costs, however, vary by distance from interconnection. Therefore, the overnight cost of construction is variable for different locations offshore, depending on this distance.

Since this analysis uses high and low cost estimates, two overnight construction cost estimates were calculated for a hypothetical wave farm constructed in each MMS lease block. These are equal to the transmission cost for the block plus the high or low plant cost estimate, divided by the plant capacity (72 MW).

D. Levelized cost of generation A second economic metric used in this analysis is levelized cost of generation

(LCOG). This is another common measure of economic performance of electricity generation plants and is essentially the average cost to generate 1 megawatt-hour of electricity.

In the UNC study of offshore wind, researchers performed a preliminary economic analysis of a hypothetical project and reported LCOG results.30 To calculate LCOG, the study employed a public-domain model developed by the California Energy Commission in

30 Coastal wind: Energy for North Carolina's future.

28 2008.31 The model, named “Cost of Generation Calculator,” was developed so the commission could compare electricity sources under its Renewable Energy Transmission

Initiative (RETI). The Cost of Generation Calculator is an excel-based spreadsheet that reports expenditures and generation over the lifetime of a project (defined by the user) and reports the project’s LCOG.

For this analysis, the Cost of Generation Calculator model was chosen so that LCOG results could be directly compared to the results reported by the UNC offshore wind study.

The LCOG model requires several inputs specific to the project being modeled.

These assumptions are listed in Table 11. Items in bold are specific to the MMS block characteristics under analysis. The rest of the input values were replicated from the values used in the UNC wind analysis. Again, this consistency is intended to allow direct comparison of results.

31 A revised version of the model was created in 2009. The 2008 version was used for this analysis for consistency with the UNC offshore wind analysis. Model spreadsheet is available on the California Energy Commission’s website at: http://www.energy.ca.gov/reti/documents/index.html

29 Table 11. Input assumptions for "Cost of Generation Calculator" LCOG model Technology Assumptions Project Capacity (MW) 72 Capital Cost ($/kW) Variable Fixed O&M ($/kW) Variable Fixed O&M Escalation 1.5% Variable O&M ($/MWh) 0 Variable O&M Escalation 1.5% Fuel Cost ($/MBtu) 0 Fuel Cost Escalation 2.5% Heat Rate (Btu/kWh) 0 Capacity Factor Variable Financial/Economic Assumptions Debt Percentage 50% Debt Rate 6% Debt Term (years) 20 Economic Life (years) 20 Depreciation Term (years) 5 Percent Depreciated 100% Cost of Generation Escalation 0% Tax Rate 40% Cost of Equity 11% Discount Rate 8.5% Incentives PTC ($/MWh) $21 PTC Escalation 1.5% PTC Term (years) 10 ITC 0%

With the required inputs, the model reports annual values for generation, operating expenses, debt service, taxes, and incentives. From these values the LCOG is calculated. An example output of the model is attached as Appendix B.

Once the specific cost figures were calculated for an offshore wave power plant in each of the MMS lease blocks, this LCOG model was run for each block using high and low cost estimates. The two LCOG values were recorded for each block, allowing a spatial representation of economic feasibility.

30 IV. Results Based on the methodology criteria outlined above, electricity generation potential was calculated and economic feasibility was assessed. The results of each stage of the analysis are presented here.

A. Electricity generation potential The assessment of electricity generation potential includes the evaluation of spatial conflicts, wave power resource potential, and WEC device performance. The combination of these elements produces an estimation of the electricity generated by the hypothetical wave power plant in the various MMS lease blocks that do not overlap with conflict zones.

The following sections present the specific results of each element.

1. Spatial conflicts to development Based on the methodology and data used in the UNC offshore wind study, the ecological and human use conflicts identified were generally close to shore. Farther offshore there are many areas available for wave power development that do not overlap with the As described in the next section the wave resource in these areas is very low, so the exclusion zone is not particularly detrimental to the prospect of wave power development.

Figure 5 below displays the ecological and environmental conflicts with their appropriate buffers. This data layer was acquired directly from the authors of the UNC offshore wind feasibility study and represents best estimates of areas of highest conflict with sensitive marine life habitat and fishing areas. The majority of areas of high conflict are near shore.

31 Figure 5. Ecological/environmental conflict areas identified in UNC offshore wind study

Figure 6 displays the human use conflicts which preclude wave power development.

These criteria were identified in the UNC offshore wind feasibility study. In the map, the

“Major waterways” layer includes ferry routes, shipping and navigation lanes, and the

Intracoastal Waterway. The blue buffer layer represents the exclusion buffers defined by the UNC study (details are included in Appendix A). As with the ecological conflicts, the majority of conflict areas are near shore, with the exception of the shipping and navigation lanes.

32 Figure 6. Human use conflicts identified by UNC offshore wind study, with appropriate buffers

The ecological and human use conflicts were combined into one exclusion zone layer. The result is displayed in Figure 7. All MMS lease blocks which overlap any portion of the conflict areas were excluded from further analysis, and the remaining eligible blocks are also displayed on the map.

33 Figure 7. Complete exclusion zone and MMS blocks eligible for development

2. Wave resource potential Using the average values derived from the wave buoys, significant wave height and dominant wave period were interpolated. Figure 8 displays the results of the wave height interpolation. Buoys are labeled by the value used in the interpolation process. Figure 9 displays the same data for dominant wave period.

34 Figure 8. Mean wave height values interpolated from buoy data

Figure 9. Mean wave period interpolated from buoy data.

35 These wave measurements were used to calculate mean incident wave power using the equation reference in the methodology section. The results are displayed below in

Figure 10. For comparison, Figure 11 is a map of estimated wave power published by

Defne, et al.32 The values from the interpolation method very closely track the values from

Defne, et al., which used a spectral wave density method of calculating wave power.

Figure 10. Mean wave power calculated with interpolated values

32 Defne, Z., Haas, K. A., & Fritz, H. M. (2009). Wave power potential along the Atlantic coast of the southeastern USA. Renewable Energy, 34(10), 2197.

36 Figure 11. Estimated mean wave power, in kW/m, from Defne, et al.

3. WEC device performance The final stage of this portion of the analysis is the electricity generation potential of

Pelamis WEC devices off the coast of North Carolina. As described in the methodology, an estimated value of annual energy absorbed by one device was calculated for each MMS block using the interpolated mean wave height and period values. This value was then multiplied by 96 (number of devices) to estimate the annual energy absorbed by the hypothetical plant. The results are displayed in Figure 12.

37 Figure 12. Annual energy absorbed by hypothetical wave farm

The efficiency and availability factors defined in the EPRI study were then applied to these values for annual energy absorbed. Two estimates were derived from these factors using the lower-efficiency parameters of the demonstration-stage devices and the higher- efficiency commercial-stage performance parameters. Estimates for annual electricity generation were calculated using these factors and the results are displayed below in Table

12.

Table 12. Estimated annual electricity generation, demonstration- and commercial-stage plants Initial Estimated Annual Electricity Generated Annual Electricity Generated Annual Energy Absorbed Demonstration-stage Commercial-stage (GWh/year) (GWh/year) (GWh/year) 3,836 2,608 4,393 4,757 3,235 5,448 4,911 3,339 5,624 8,594 5,844 9,843 10,820 7,358 12,392 16,805 11,427 19,247

38 The values for the two classes of plants are drastically different. In fact, the annual generation values for the commercial-stage plant are higher than the initial estimates for energy absorbed by the devices in the wave farm. This is due to the 37% increase in device performance that the EPRI study estimates can be achieved with a mooring reconfiguration. These values demonstrate the impact of the efficiency improvements envisioned by the authors of the EPRI feasibility study.

B. Overnight cost of plant construction Overnight cost of construction consists of wave farm device costs and transmission costs. The EPRI cost estimates for a demonstration- and commercial-stage device were detailed in the methodology section. The final values of $2,658,000 and $986,000 per device are reported in 2003$. These values were converted to 2008$ to create per-device estimates of approximately $3,031,000 and $1,124,000. Multiplying these values yielded by 96 (devices in the hypothetical farm) yielded values of $290,976,000 for the demonstration-stage farm and $107,904,000 for the commercial stage wave farm. These values were constant for wave farms built at any MMS lease block.

Transmission costs calculated for each MMS block were added to the device costs to create estimates for block specific total overnight construction cost. Since the transmission costs vary with distance from interconnection, the overnight capital costs predictably increase farther offshore. Separate values were calculated using the high and low capital cost estimates and the results are displayed in Figure 13 and Figure 14 below.

As with the electricity generation values discussed in the previous section, the difference between the overnight construction cost values of the demonstration and commercial-stage plants exhibits the impact of the EPRI study’s estimations regarding

39 future cost savings. Under the high cost estimates of the demonstration plant, the overnight capital cost for a 96 device wave farm is as much as two and a half times the overnight capital cost of the same plant under the low cost estimates of the commercial- stage plant.

Figure 13. Overnight cost of construction, high cost estimates

40 Figure 14. Overnight cost of construction, low cost estimates

C. Levelized cost of generation Using the “Cost of Generation Calculator” with the wave farm cost and performance data, a LCOG value was calculated for each MMS lease block using the two sets of wave farm parameters (demonstration- and commercial-stage). The results of these calculations are shown in Figure 15 and Figure 16 below.

As with the overnight capital cost maps, the pattern is the same, but the values in each map are radically different.

41 Figure 15. Levelized cost of generation, high capital cost, low efficiency estimates

Figure 16. Levelized cost of generation, low capital cost, high efficiency estimates

42 V. Discussion A. Spatial results Based on the results of the spatial analysis, specifically the areas of high conflict

(Figure 7) and the spatial distribution of electricity generation potential (Figure 12), wave power appears to be technically feasible off the coast of North Carolina. The results of the spatial conflict assessment show that there would be serious constraints to any offshore energy development in the areas near the coast, as these areas are characterized by high ecological sensitivity and high levels of human use. Conversely, the areas of greatest electricity generation potential are far offshore, and do not overlap with spatial conflicts to any significant degree. This conclusion, however, is based on the data available. It is likely that so many of the ecological conflicts identified by the UNC study are near the shore because these areas have been most thoroughly studied and mapped. Any proposed offshore energy facility would require a full environmental impact statement to assess potential damages.

As mentioned, the best wave resource exists far off the coast. In addition to posing transmission challenges, there are also many technical challenges to developing a wave power project far offshore, particularly in the deep waters off the continental shelf. This analysis did not impose a maximum depth restriction on lease blocks assessed, but it is likely that areas off the shelf – where quickly drop from 100 meters to several thousand meters – would be beyond the limits of current mooring technology. Also, in areas far off the coast, installation costs would be greater. This analysis did not include installation costs because in the EPRI analysis the cost of installation was only a very small fraction of

43 the total plant cost. This would likely change if construction was taking place hundreds of kilometers from shore.

B. Economic results The results of the economic analysis illuminate several important issues. First, transmission cost can be a very large portion of a project’s cost. In this analysis, the percentage of the total overnight capital cost attributable to transmission ranged between

10% and over 60%. As the market for subsea transmission grows, it is highly likely that improved technologies will develop and these costs will decrease.

The maps of LCOG prices (Figure 15 and Figure 16) highlight the important tradeoffs between near-shore and far-offshore development. Clearly, the cost of developing a plant increases with distance from shore due to the cost of transmission. But at the same time, the wave resource generally improves with distance from shore. This explains why there is no consistent trend of LCOG prices. If the gains in device performance outweigh the extra costs of development, LCOG can decrease farther offshore.

Overall, the values for both overnight construction cost and for LCOG are relatively high. Table 13 below compares the results of the analysis with cost estimates for conventional generation sources. It also includes an LCOG estimate for offshore wind in

North Carolina from the UNC feasibility study. The wave power values in red are derived from the higher capital costs, lower efficiency estimates and the values in blue are the result of lower capital costs and higher efficiency.

44 Table 13. Cost comparison of wave power to other electricity sources.33 Overnight Construction Cost of Generation Plant Type $Million/MW $/MWh Coal: Pulverized $2.5 $63.10 Coal: IGCC $3.4 $82.99 Nuclear $3.9 $83.22 Natural Gas Combined Cycle $1.2 $61.77 Wind $2.1 $80.74 Geothermal $3.2 $59.23 Solar Thermal $3.4 $100.32 Solar Photovoltaic $6.6 $255.41 NC Offshore Wind (UNC)34 -- $101 NC Wave Power $4.5-$10.5 $431-2,032 $2.0-$7.9 $137-819

From this table it is clear that, in general, costs for wave power far exceed costs for traditional generation. The ranges in values are very large, due to the cost variability for a project built at different distances from the coast. At the low end of the estimates, costs are within the range of conventional generation, so it is conceivable that wave power could be economically feasible in North Carolina. However, as previously discussed, these estimates have a considerable amount of inherent uncertainty since a commercial scale wave farm has not yet been constructed.

Cost estimates for offshore wind power, on the other hand, have a much higher degree of certainty due to the maturity of the technology and high levels of development in

Western Europe. This, combined with the generally lower cost of technology, suggests that for marine renewable energy in North Carolina, offshore wind is almost certainly the better strategy.

33 All values except offshore wind and wave power from: Kaplan, S. (2008). Power Plants: Characteristics and Costs. Congressional Research Service. 34 Coastal wind: Energy for North Carolina's future.

45 VI. Recommendations There are several steps that the state of North Carolina could take to improve the development of marine renewable energy generally, and wave power specifically. The first would be to develop a strategy to improve the collection of data on the marine environment. The interpolated values (wave height and period) in this study would have improved accuracy with more buoys offshore. This would increase confidence in any performance calculations and reduce uncertainty and risk.

North Carolina should also monitor the results of wave power demonstration projects worldwide, and particularly U.S. projects, such as the one currently under development in Oregon. As noted by the authors of the UNC study, North Carolina fell behind other states in offshore wind development. As a result of the quick response to the conclusions of the UNC study, North Carolina is quickly catching up to other states with the development of the test project in the Pamlico Sound. By monitoring the progress of wave power demonstration sites, North Carolina would be able to assess whether or not to pursue a demonstration project and would not fall behind other states in another renewable energy technology.

Whether through experimentation with wave power or the continued efforts on offshore wind development, North Carolina should continue to develop its offshore infrastructure. By developing an offshore transmission plan that includes a strategy for onshore interconnection, the state would encourage marine renewable energy developers to site their projects off the coast of North Carolina. Offshore transmission development would encourage mixed use or hybrid projects. Such projects feature multiple offshore

46 technologies (such as wind, wave, or tidal current) sharing infrastructure, such as transmission.

By continuing the exploration of marine electricity generation spurred by the conclusions of the UNC wind study, North Carolina can develop itself as a renewable energy leader, not just in the southeast, but for the entire country.

47 Appendix A:

Data Layer Date Data Source Ferry routes n/a NC DOT data layer acquired from UNC wind study authors Major shipping lanes 2009 -Vanderbilt Engineering Center for Transportation Operations and Research -Research and Innovative Technology Administration’s Bureau of Transportation Statistics (RITA/BTS) Major navigation channels 2009 -Vanderbilt Engineering Center for Transportation Operations and Research -Research and Innovative Technology Administration’s Bureau of Transportation Statistics (RITA/BTS) Intracoastal waterway 2009 -Vanderbilt Engineering Center for Transportation Operations and Research -Research and Innovative Technology Administration’s Bureau of Transportation Statistics (RITA/BTS) Dredge material disposal 1998 NOAA Coastal Services Center sites Shipwrecks 2007 Point file created from data available on website: http://www.nc-wreckdiving.com Cape Hatteras National 2005 NOAA and U.S. Department of the Interior (DOI) Seashore Monitor National Marine 2005 NOAA and U.S. Department of the Interior (DOI) Sanctuary

Note: With the exception of the ferry routes and shipwrecks layers, all data files were downloaded from the Florida Fish and Wildlife Conservation Commission’s Fish and Wildlife Research Institute website at: http://ocean.floridamarine.org/efh_coral/ims/Description_Layers.htm

48 Appendix B: Example output of LCOG model

49 Bibliography

Bedard, R., Hagerman, G., Previsic, M., Siddiqui, O., Thresher, R., & Ram, B. (2005). Offshore wave power feasibility demonstration project: Final summary report. Electric Power Research Institute. Bonner, L. (2009, September 26). Pamlico Sound towers to test wind energy. News & Observer. Cape Wind energy project draft environmental impact statement. (2004). U.S. Army Corps of Engineers. A citizen's guide: The North Carolina renewable energy and energy efficiency portfolio standard. (2009). North Carolina Sustainable Energy Association. Coastal wind: Energy for North Carolina's future. (2009). University of North Carolina, Chapel Hill. Dalton, G. J., Alcorn, R., & Lewis, T. (2010). Case study feasibility analysis of the Pelamis wave energy convertor in Ireland, Portugal and North America. [doi: DOI: 10.1016/j.renene.2009.07.003]. Renewable Energy, 35(2), 443-455. de AlegrÌa, I. M., MartÌn, J. L., Kortabarria, I., Andreu, J., & EreÒo, P. I. (2009). Transmission alternatives for offshore electrical power. Renewable and Sustainable Energy Reviews, 13(5), 1027-1038. Defne, Z., Haas, K. A., & Fritz, H. M. (2009). Wave power potential along the Atlantic coast of the southeastern USA. Renewable Energy, 34(10), 2197. Dunnett, D., & Wallace, J. S. (2009). Electricity generation from wave power in Canada. Renewable Energy, 34(1), 179-195. Green, J., Bowen, A., Fingersh, L. J., & Wan, Y. (2007). Electrical collection and transmission systems for offshore wind power. National Renewable Energy Laboratory. Hagerman, G., & Bedard, R. (2003). E2I EPRI specification: Guidelines for preliminary estimation of power production by offshore wave energy conversion devices. Electric Power Research Institute. Kaplan, S. (2008). Power Plants: Characteristics and Costs. Congressional Research Service. Li, G. (2000). Feasibility of large scale offshore wind power for Hong Kong - a preliminary study. Renewable Energy, 21(3-4), 387-402. Nondirectional and directional wave data analysis procedures. (1996). Retrieved from http://www.ndbc.noaa.gov/wavemeas.pdf. North Carolina electricity profile. (2008). Retrieved from http://www.eia.doe.gov/cneaf/electricity/st_profiles/north_carolina.html. Previsic, M., Bedard, R., & Hagerman, G. (2004). Offshore wave energy conversion devices. Electric Power Research Institute. Previsic, M., Bedard, R., Hagerman, G., & Siddiqui, O. (2004). System level design, performance and costs for San Francisco California Pelamis offshore wave power plant. Electric Power Research Institute. Reedsport OPT wave park. (2010). Retrieved April 20, 2010, from http://www.oceanpowertechnologies.com/reedsport.htm Report on the NCTPC 2008-2018 collaborative transmission plan [DRAFT]. (2008). North Carolina Transmission Planning Collaborative. Retail sales of electricity to ultimate customers: Total by end-use sector. (2010). Retrieved from http://www.eia.doe.gov/cneaf/electricity/epm/table5_1.html.

1 Sahir, R. (2009). Inflation conversion factors for years 1774 to estimated 2019. Oregon State University. Wright, S. D., Rogers, A. L., Manwell, J. F., & Ellis, A. (2002). Transmission options for offshore wind farms in the United States. AWEA.