Loading...
HomeMy WebLinkAboutPVSC_Wilber-et-al_Performance_Cost_Solar_Photovoltaic_Technology_Alaska_ACEP 2015 Catching the Midnight Sun: Performance and Cost of Solar Photovoltaic Technology in Alaska Michelle Wilber, Erin Whitney, Christopher Pike and Jeremy Johnston Alaska Center for Energy and Power, University of Alaska Fairbanks, P.O. Box 755910, Fairbanks, AK 99775-5910, USA Abstract — This paper is an update to the original 2017 review of solar photovoltaic (PV) technology cost and performance in Alaska by E. Whitney and C. Pike in the Journal of Renewable and Sustainable Energy (volume 9, pp. 061704:1- 9). Solar PV technology remains a nascent but promising energy option in remote Alaska microgrids, which can serve as examples for isolated electrical grids worldwide. Capacity factors for community-scale solar PV installations in Alaska range from 7% to 15%, which can still be economically viable given the high cost of diesel typically used in rural powerhouses. With remote locations and challenging conditions, total installed costs can be significantly higher than those in the rest of the United States, although trends indicate that costs per kilowatt-hour (kWh) are lower for larger system sizes and are decreasing with time as seen elsewhere. Index Terms — Alaska, arctic, microgrids, photovoltaic, solar I. INTRODUCTION Significant volatility in fuel prices in the past decade and concerns over energy security have thrust isolated grids such as those found in rural communities, island states, and remote military installations using conventional fossil fuel power generation into an energy crisis. Many of these remote locations are turning to renewable energy to reduce fuel consumption and costs and to ensure a more independent and reliable energy source. At the same time, vulnerability of larger grids to disruption from natural disasters and other sources has led to an interest in creating ‘islandable’ microgrids incorporating renewables to improve resiliency. Over 200 remote communities in Alaska are largely dependent on diesel generators and arguably have the highest electric rates in the nation due to the logistics of importing fuel. These stand-alone village microgrids typically serve 300–450 people with average loads of ~200 kilowatts (kW) [1]. As such, these microgrids provide ideal settings to validate variable generation and load control strategies with broader applications to high-penetration renewable islanded systems globally. The renewable energy contribution in many small Alaska communities is already much higher proportionally than what utilities in larger grids nationally would even consider, although larger grids are on a trajectory that will require doing so in future operations. One of these renewable energy sources is solar photovoltaic (PV) power [2]. Although Alaska’s high latitude creates large fluctuations in sunlight throughout the year, computer simulations show the solar PV potential in Alaska to be on par with or greater than that of Germany, one of the largest solar PV power markets in the world [3]. Furthermore, Alaska’s cold temperatures increase system voltage, reduce electrical resistance, and yield higher-than-rated outputs associated with reflected light and albedo effects [4]-[5]. These factors, combined with declining module prices, are making solar PV technology more economical. Solar PV arrays have been installed in all areas of the state from the southwest to the Arctic, and low sun angles and long daylight hours represent opportunities to mount panels vertically on walls as well as on the east and west sides of buildings. To illustrate the solar resource in Alaska, Figure 1 shows the expected average daily solar radiation levels for varying surface angles for different cities in Alaska relative to Seattle and Phoenix in the continental United States [6]. Anchorage (61°N) is located in southcentral Alaska, Fairbanks (65°N) is in the interior, and Kotzebue (67°N) is in the far north. Figure 1c shows the amount of solar radiation that horizontal collectors would be expected to receive throughout the year, with the highest radiation levels in May and June. In the graph showing the collectors tilted at 15° steeper than the latitude angle (Figure 1a) and tilted vertically (Figure 1b), a strong improvement in springtime performance is seen. Figures 1a and 1b show that Fairbanks and Kotzebue receive almost 5 kWh/m2/day in March and almost 6 kWh/m2/day in April. This early season performance improvement is attributed to more direct radiation from the low sun angles and high levels of reflected radiation from the snow-covered ground. Significant space-heating demands also coincide with this springtime arctic solar resource. While solar PV power is not a viable year-round resource for Alaskan communities, it can be of use for seasonal applications and paired with other energy sources for winter energy demands. Fig. 1. Expected average daily solar radiation levels for varying surface angles by city in Alaska, compared with Seattle and Phoenix. For 1a, the collectors are angled at 15° steeper than the respective latitudes of Seattle (48°N), Anchorage (61°N), Fairbanks (65°N), Kotzebue (67°N), and Phoenix (33°N). Figure 1 uses data from the National Renewable Energy Laboratory (NREL), collected between 1961 and 1990 and based on averaged values of radiation. It does not account for weather patterns and cloud cover. In the NREL model, the albedo of snow was taken into account to calculate reflected radiation [7]. Surface albedo was adjusted depending on the presence of snow cover. If there was snow on the ground, the surface albedo was set to 0.6 (albedo for snow ranges from ~0.35 for old snow to 0.95 for dry new snow). If no snow was indicated, the surface albedo was set to 0.2, a nominal value for green vegetation and some soil types [7]. II. METHODS Cost and production estimates are vital in the economic analysis of a proposed solar project. An initial analysis of solar installations in Alaska was compiled in 2017, which included data before the fall of 2016 [8]. At that time, 17 installations had enough data to calculate a capacity factor, and 18 installed or proposed systems had enough data to estimate installed costs. Large variability in costs per installed kW was seen in the 2017 compilation [8], and insufficient data were available to see trends in cost with time. An additional three years of installation and production data were available to add to the 2017 analysis. At least four additional community-scale solar PV systems have been installed in isolated microgrids in Alaska since the fall of 2016, and many more systems (and capacity) have been installed on the larger ‘railbelt’ electric grid which services the area along Alaska’s road and rail system connecting the interior and southcentral portions of the state. An additional 13 systems now have enough data to calculate a capacity factor, and costs are available for four new systems and for Solarize Anchorage, a neighborhood solar installation program in Anchorage, Alaska. Much of the data is not public and was obtained by consultation with installers, community development staff, and others with knowledge of the projects. Many of the installed systems are currently monitored, with data available through online portals. Some cost breakdowns by component (hardware, labor, shipping, etc.) are possible, but often these are estimations given a lump sum bid by installers. III. ANALYSIS AND DISCUSSION A. Total Installed Costs As reported previously [8], costs for solar PV projects installed in Alaska before fall of 2016 ranged from $3.19/W to $13.33/W with all systems smaller than 24 kW, although the low end of this cost range involved volunteer labor and/or shipping deals. Adding on to that original analysis, a proposed (but uninstalled) 50 kW project had an anticipated cost of about $4.00/Watt. Additional costs are now available for four community and utility scale PV systems installed in Alaska in 2018. Costs are also available, as a function of size, from a residential solar participation program (Solarize Anchorage) in Anchorage, Alaska for 2018 and 2019. Costs of the 2018 and 2019 projects range from $1.25 to $4.60/Watt. All of these costs, translated to 2018 dollars and plotted as a function of the year of installation, show a clear trend of falling installation costs with time (Figure 2). Total installed costs as reported here are the sum of labor, parts and materials, and shipping. In some cases, as reported in the 2017 analysis, lower than average installation costs were realized by special circumstances in one or more of those components. Most labor costs were not included in a 138 kWDC project in Willow, Alaska as the owners installed the project themselves. For a number of other installations, figures are based on verbal estimates from batched purchases and are not public record. The inconsistency of information is indicative of the nascent solar PV industry in Alaska. Total installed costs in dollars per watt plotted as a function of installation size for the 2018 and 2019 installations show a trend toward lower costs with larger installation sizes, as seen in Fig. 3. For the remote installations, costs ranged from $2.20 to $4.60/Watt. All of these are for larger (> 45 kW) systems. Installations on the less-remote road system in Alaska range from $1.25 to $3.50/Watt, with higher costs for smaller systems. Fig. 2. Total installed costs ($/W) as a function of installation year show a trend towards lower costs with time in Alaska. The yellow and red lines show the trend for U.S. residential and commercial installed costs, respectively, according to NREL [9]. All costs are in 2018 dollars. Prices in Alaska are approaching but still generally higher than prices in the contiguous United States (Figure 2). This is unsurprising considering the high transportation and labor costs of remote Alaska. It appears that the large variability and high premiums in Alaska relative to contiguous U.S. prices are falling, which is expected with a maturing market. Fig. 3. Total installed costs ($/W) as a function of installation size (kW) for 2018 and 2019 installations in Alaska show a trend towards lower costs with larger installation sizes. All costs are in 2018 dollars. B. Cost Breakdown Cost breakdowns for solar installations in rural Alaska (Table 1) are difficult to obtain. Often the contractor bids on a job as a lump sum, and separating labor from equipment and materials is difficult to do accurately. Systems in Ambler, Kobuk, Shungnak, Noorvik, Noatak, Deering, Kotzebue, Selawik, Kiana, Buckland, and Kivalina (shaded in green in Table 1) were installed by Bering Straits Development Company through coordination with the Northwest Arctic Borough and were bid as a group. Costs for these systems were difficult to separate from the main lump sum bid. Based on input from Rob Bensin, former Energy Efficiency and Renewable Energy Division Manager at Bering Straits Development Company (personal communication), costs were separated using 30% for logistics, 15% for labor, and the remainder for racking, hardware, and materials. Systems in Eagle and Kaltag were installed by the utilities using funding from the Renewable Energy Fund. Systems in Galena and Fort Yukon were installed with assistance from the Tanana Chiefs Conference. The 18 kW installation in Fort Yukon was accomplished with volunteer labor and a shipping deal. A 50 kW Galena system was only bid, and not installed. Per price quote, “Heavy equipment to be provided for trenching/anchors/material handling.” Shipping is estimated here at $30,000 per Dave Pelunis-Messier, Rural Energy Coordinator for the Tanana Chiefs Conference, based on other similar systems in the Interior. The Willow installation had additional estimated costs not included in the other categories of: land preparation: $12,000, fencing and security: $5,000, electrical and interconnection: $22,000, equipment rental: $10,000. The Hughes installation was not completely installed as of spring 2019, and thus a cost per watt and total cost are not given in Table 1. Of note, inverter installation, electrical work, and interconnection are ongoing in Hughes. Therefore, hardware cost in Table 1 does not include inverters, and electrical labor and interconnection are also not included in these cost breakdowns. Other costs incurred to date for this installation, but not included in the categories of Table 1 are: land preparation: $5,000 plus in-kind equipment use, and $1335 for pile installation equipment rental. C. Operation and Maintenance Costs Costs for operation and maintenance (O&M) of a solar PV system include scheduled maintenance and cleaning, unscheduled maintenance, and inverter replacement reserves [10]. According to NREL, costs dropped from $54/kW/yr in 2010 to $22/kW/yr in 2018 for residential-sized systems (less than 10kW) in the U.S. [9]. Improved inverter reliability was one driver of the decrease in O&M. In Alaska, most grid-tied PV systems have been installed for less than 7 years, generally well below the lifetime of major components. TABLE I COST DATA FOR SELECTED SOLAR INSTALLATIONS IN ALASKA Installed Cost by Major Components Location System Size (kW) Installation Date (month/year) Hardware ($) Support Structure Labor/Travel ($) Shipping ($) Cost/Watt ($) Total Cost ($) Installed systems. Costs were based on percentages of estimated total system cost Ambler 8.4 3/2013 41,250 included in hardware 11,250 22,500.00 8.93 75,000 Ambler IRA 2.2 3/2013 13,750 included in hardware 3,750 7,500.00 11.36 25,000 Kobuk 7.4 3/2013 41,250 included in hardware 11,250 22,500.00 10.14 75,000 Shungnak 7.5 10/2013 41,250 included in hardware 11,250 22,500.00 10.00 75,000 Noorvik 12 10/2013 41,250 included in hardware 11,250 22,500.00 6.25 75,000 Noatak 11.3 11/2013 41,250 included in hardware 11,250 22,500.00 6.64 75,000 Deering 11.1 11/2013 41,250 included in hardware 11,250 22,500.00 6.76 75,000 Kotzebue- 1 10.5 10/2014 45,650 included in hardware 12,450 24,900.00 7.90 83,000 Kotzebue- 2 10.5 11/2014 45,650 included in hardware 12,450 24,900.00 7.90 83,000 Selawik 9.7 11/2014 45,650 included in hardware 12,450 24,900.00 8.56 83,000 Kiana 10.5 8/2015 45,650 included in hardware 12,450 24,900.00 7.90 83,000 Buckland 10.5 2015 45,650 included in hardware 12,450 24,900.00 7.90 83,000 Kivalina 10.5 2015 45,650 included in hardware 12,450 24,900.00 7.90 83,000 Installed systems with detailed cost records Location System Size (kW) Installation Date (month/year) Hardware Cost ($) Support Structure Cost ($) Labor/Travel Cost ($) Shipping Cost ($) Cost/Watt ($) Total Cost ($) Eagle 24 7/2015 115,552 included in hardware 94,632 10.88 261,000 Kaltag 9.6 2012 78,657 included in hardware 15,946 6,465.00 13.33 128,000 Galena 6.7 11/2012 14,400 2000 5,000 City covered shipping cost 3.19 21,400 Fort Yukon 18 7/2015 45,000 Included in Hardware 20,000 5,000.00 3.89 70,000 Galena 50 Dec 2015 Estimate Only Lumped together in bid 30,000.00 4.07 203,613 Buckland 45 Fall 2018 Breakdown not available 42,000.00 4.60 207,000 Willow 138 Summer 2018 91,000 30,000 10,000 Included in other costs 1.25 172,500 Hughes 122 2018-2019 71,412 40,920 51,429 32,195 The Cold Climate Housing Research Center (CCHRC) has some of the oldest grid-tied solar installations in Alaska; it maintains three pole-mounted PV systems on two-axis tracking systems with a total installed size of 8 kW. A relay has needed replacement, but otherwise very little maintenance has been required. According to staff at the CCHRC, 4 hours of maintenance are devoted to the systems per year (2 hours twice each year). Assuming $60/hour, yearly maintenance costs equal $30/kW/yr, without taking into account inverter replacement. The trackers are locked at a fixed angle of 80° azimuth facing due south between November and February, when solar insolation is at a minimum and temperatures are coldest; they are set to track the rest of the year. One aspect of O&M in Alaska that deserves special mention is that of snow clearing. A study by students at the University of Alaska Fairbanks simulated the cost and benefit of clearing snow from a hypothetical 1 MW solar installation that faced south at a panel angle of 70°. The study demonstrated that the cost savings from increased generation of electricity due to snow having been cleared from the panels did not justify the cost of labor to perform the task of clearing snow. This study was performed in Fairbanks, where winds are light and extended cold temperatures cause snow in the fall and winter to stay on the ground into springtime. The results would likely be the same, if not more exaggerated, in Western Alaska, where high winds blow and mid-winter warm-ups melt snow from roofs [11]. As stated in the 2017 review [8], discussions with a number of individuals involved in the solar industry in Alaska yielded O&M costs of approximately $100 per installed kW of solar PV power on the high side. This is still considered a reasonable upper limit, given that many of the solar PV arrays installed around the state have still not needed any maintenance since installation. Yearly bolt-tightening is recommended and occasionally needed. Given all of the documents reviewed to date, for PV systems less than 20 kW in Alaska, O&M ranges from ~$25/kW/yr on the road system or in hub communities to ~$100/kW/yr in more remote areas. Operation and maintenance costs are not completely dependent on system size; they are also a function of the level of local expertise available for repairs, the cost of travel to and from the site, occasional cleaning and inspection, unscheduled warranty work, and inverter replacement reserves. D. Expected Life Most installers assume a system life of 25 years, although it is useful to consider expected lifetimes of individual components. Panels are typically warrantied for 10 years on materials and 25 years for power output, and inverters can be warrantied from 10–20 years. Only one early inverter failure has been noted in a remote Alaskan installation. Tracking systems are commonly reported to have failed in Alaska - likely due to factory installation of inappropriate (“warm- weather”) grease. Panels are sometimes damaged from thrown rocks. E. Capacity Factors Capacity factor is a function of weather, system design, system installation location, angle, and azimuth. It is a unitless ratio of the average power generated, divided by the rated peak power. Note that many of the systems installed in the Northwest Arctic Borough were installed in a semicircular fashion, with the goal of a broad production curve rather than maximum power production at midday. More systems are installed around the state than the ones reported here; however, insufficient data were available to obtain capacity factor information on the systems not listed. Production data available in the 2017 analysis showed that capacity factors ranged from 6% to 15% [8]. Additional production data from these and other installations are now available to update this analysis for more locations in Alaska (Table 2). Including newer production data in the analysis, capacity factors are found to range from 7% to 15%, with the highest values at the one site with tracking. Predicted capacity factors are calculated using NREL’s PVWatts modeling program. Orientation and tilt were estimated from photos, installer’s information, or assumed to be south at 30° tilt if no other information was available. Production data from these installations indicate that capacity factors are close to those modeled by resources such as NREL’s PVWatts, although PVWatts tends to over-predict yearly production at installations in Alaska. This discrepancy points to the need for refinements in local insolation, albedo, snow cover, and other data. Underperformance compared to PVWatts modeling could also be due to incorrect assumptions in the model, shading, panel damage, clipping from inverters sized lower than panel output, or other factors. F. Levelized Cost per kWh The simple levelized cost of renewable energy (cents/kWh) was calculated at 34.4 cents/kWh based on the following inputs into the National Renewable Energy Laboratory (NREL) levelized cost of electricity (LCOE) calculator detailed below. This figure is roughly half of that calculated in the 2017 review [8]. Period: 25 years Discount Rate: 3% Capital Cost (average): $3,000/kW Capacity Factor (average): 9% Fixed O&M Cost: $100/kW/yr Variable O&M Cost: none Heat Rate: none Fuel Cost: none TABLE II ANNUAL AVERAGE PRODUCTION AND CAPACITY FACTORS FOR SELECTED SOLAR INSTALLATIONS IN ALASKA Station Location System Size (kW) Measured Annual Average Production (kWh) Capacity Factor PVWatts predicted Capacity factor Ratio of Actual to PVWatts Capacity Factor Ambler 8.4 5351 7% 9% 78% Arctic Village Clinic 5 3722 8% 11% 78% Bethel Yuut 10 12058 14% 11% 129% Deering 11.1 9671 10% 10% 100% Fairbanks Tracking1 2.64 3555 15% 14% 112% Fairbanks Fixed 2.64 3344 14% 11% 130% Fairbanks Tracking2 2.72 3516 15% 14% 107% Manley Hot Springs 6 4080 8% 9% 86% Noatak 11.3 8921 9% 10% 91% Nome 16.8 13654 9% 10% 91% Nenana Teen Center 4.4 4161 11% 11% 97% Ruby Health Clinic 5.5 4239 9% 9% 97% Ruby Washateria 5.4 4223 9% 9% 99% Kobuk 7.38 5618 9% 9% 93% Shungnak 7.49 5049 8% 8% 93% Noorvik 12 9674 9% 9% 101% Kotzebue Bailing 10.53 8819 10% 10% 96% Kotzebue 10.53 8684 9% 10% 95% Kiana 10.53 8089 9% 11% 82% Buckland 10.53 7317 8% 9% 87% Galena 10.35 6474 7% 10% 69% Fort Yukon 18 16423 10% 12% 88% Koyukuk City Building 5.16 3943 9% 11% 82% Selawik 9.72 8490 10% 11% 93% Kaltag 9.6 8827 10% 11% 99% Bethel 4.9 4601 11% 11% 97% Anchorage - Bear Valley1 4.06 3209 9% 8% 108% Anchorage - Bear Valley2 7.78 6560 10% 9% 111% Kivalina 10.5 6800 7% 10% 72% Homer 4.13 3678 10% 11% 95% Naknek 80.08 52125 7% 11% 69% Ouzinke 7 5344 9% 11% 81% Considering the capacity factors for installations in Alaska, the LCOE ranges from 20.6–44.2 cents/kWh over a capacity factor range of 7–15%, all other variables remaining constant. Similarly, the LCOE ranges from 21.7–45.9 cents/kWh over a capital cost range of $1,250–$4,600/kW, all other variables remaining constant at the values listed above. The capital cost range is assumed equal to the total installed cost range seen in the 2018-2019 data, since solar PV costs are predominantly capital costs. Capital costs are also the factor that has changed the most since the 2017 review. This LCOE does not take in to account tax credits and benefits or financing costs. G. Storage Systems With the continuing decrease in the cost of storage, solar PV with storage is beginning to garner interest and be installed in Alaska. Solar PV installations with large instantaneous penetrations in remote microgrids can use storage to maintain grid stability. Installations in Buckland, Deering and Hughes are all planning to eventually include battery storage on their grids as well. Storage costs are not included in any of the data in the current analysis. IV. SUMMARY Solar PV technology is a nascent but promising energy option in remote Alaskan microgrids and serves as an example for isolated electrical grids worldwide. This study examines community-scale solar PV installations in Alaska. Total installed costs show a trend toward lower values with larger installation sizes and with time, although prices in Alaska are still generally higher than in the rest of the United States. Capacity factors range from 7% to 15%. However, it should be noted that some installation configurations, particularly in the northwestern part of the state, were installed with the goal of a broad production curve rather than maximum power production. ACKNOWLEDGEMENTS Many people have contributed information and insight to this study. For their review and comments, we wish to thank Rob Bensin, former Energy Efficiency and Renewable Energy Division Manager at Bering Straits Development Company, now with ACEP; Ingemar Mathiasson, Energy Manager for the Northwest Arctic Borough; Paul Schwabe of the National Renewable Energy Laboratory; Dave Messier, Rural Energy Coordinator for the Tanana Chiefs Conference; Bob Deering of the United State Forest Service; Bruno Grunau of the Cold Climate Housing Research Center; Alan Mitchell of Analysis North; Brian Hirsch of Deerstone Consulting; and Dave Lockard, Sam Tappen, and Neil McMahon of the Alaska Energy Authority. REFERENCES [1] AEA (Alaska Energy Authority), Renewable Energy Atlas of Alaska, AEA, 2011. [2] P. Schwabe, “Solar energy prospecting in remote Alaska: An economic comparison of electricity generation costs between solar photovoltaics and diesel fuel expenditures,” National Renewable Energy Laboratory Report, No. 65834, 2016. [3] H. Wirth, “Recent Facts about Photovoltaics in Germany,” Fraunhofer ISE, 2015. [4] J. Nelson, The Physics of Solar Cells, Imperial College, London, 2003. [5] M. Brennan, A. Abrahamse, R. Andrews, and J. Pearce, “Effects of spectral albedo on solar photovoltaic devices,” Solar Energy Materials and Solar Cells, vol. 124, pp. 111–116, 2014. [6] Alaska Center for Energy and Power, “An Investigation of Solar Thermal Technology in Arctic Environments: A Project by Kotzebue Electric Association,” Denali Commission – Emerging Energy Technology Grant report, 2012. [7] National Renewable Energy Laboratory, “Solar Radiation Data Manual for Flat-Plat and Concentrating Collectors.” Retrieved from http://rredc.nrel.gov/solar/pubs/redbook/HTML/appendix.html# -calcsolrad. [8] E. Whitney and C. Pike, “An Alaska Case Study: Solar PV Technology in Remote Microgrids,” Journal of Renewable and Sustainable Energy, vol. 9, pp. 061704:1-9, 2017 [9] R. Fu, D. Feldman, and R. Margolis, “U.S. Solar Photovoltaic System Cost Benchmark: Q1 2018.” Golden, CO: National Renewable Energy Laboratory Report, No. 72399, 2018. [10] N. Enbar, and T. Key, “Addressing solar photovoltaic operations and maintenance challenges: A survey of current knowledge and practices,” EPRI, Publication 1021496, 2010. [11] A. Vilagi, and P. Brown, “Effects of snowfall on solar power generation.” ACEP Report, University of Alaska Fairbanks, 2015.