Utility-Scale Solar 2015 An Empirical Analysis of Project Cost, - - PowerPoint PPT Presentation
Utility-Scale Solar 2015 An Empirical Analysis of Project Cost, Performance, and Pricing Trends in the United States Mark Bolinger & Joachim Seel Lawrence Berkeley National Laboratory August 2016 This research was supported by funding
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
This research was supported by funding from the U.S. Department of Energy’s SunShot Initiative.
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Annual solar “state of the market” reports
Derivative analyses (e.g., Academic Partnership Program)
2
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
3
Key findings from analysis of the data samples (We discuss PV projects first, then focus on CSP projects):
2.
Installed project prices
3.
Operation and maintenance (O&M) costs
4.
Performance (capacity factors)
5.
Power purchase agreement (“PPA”) prices
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Utility-scale solar had a 57% capacity share of 2015 installations and a 54% share of
cumulative installations at the end of 2015
4 Sources: GTM / SEIA Solar Market Insight Reports, LBNL Database
We define “utility-scale” as any ground-mounted project that is larger than 5 MWAC Smaller systems are analyzed in LBNL’s “Tracking the Sun” series.
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
RPS has historically been
a significant driver of utility-scale solar, particularly in the Southwest and Northeast
Recent RPS expansions
(e.g., in CA, OR) will ensure future RPS relevance for utility-scale solar markets
Increasingly, utility-scale
solar expansion into non- RPS states (Southeast) or continued deployment where RPS goals have been reached (e.g., Texas)
5
WI: 10% by 2015 NV: 25% by 2025 TX: 5,880 MW by 2015 PA: 8.5% by 2020 NJ: 22.5% by 2020 CT: 23% by 2020 MA: 11.1% by 2009 +1%/yr ME: 40% by 2017 NM: 20% by 2020 (IOUs) 10% by 2020 (co-ops) CA: 50% by 2030 MN: 26.5% by 2025 Xcel: 31.5% by 2020 IA: 105 MW by 1999 MD: 20% by 2022 RI: 16% by 2019 HI: 100% by 2045 AZ: 15% by 2025 NY: 30% by 2015 CO: 30% by 2020 (IOUs) 20% by 2020 (co-ops) 10% by 2020 (munis) MT: 15% by 2015 DE: 25% by 2025 DC: 20% by 2020 WA: 15% by 2020 NH: 24.8% by 2025 OR: 50% by 2040 (large IOUs) 5-25% by 2025 (other utilities) NC: 12.5% by 2021 (IOUs) 10% by 2018 (co-ops and munis) IL: 25% by 2025 VT: 75% by 2032 MO: 15% by 2021 OH: 12.5% by 2026 MI: 10% by 2015
Source: LBNL RPS Report 2016 Source: GTM 2016: The Next Wave of Utility-Solar
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Voluntary Procurement:
3rd party-ownership with competitive PPA deals (+ Hedge Value) Utility-Owned Generation (Florida Power & Light, Georgia Power, Dominion,
Duke, PNM) PURPA (“avoided cost” rates for “Qualifying Facilities”)
e.g., North Carolina (~2700 MW), Utah (~700 MW), Idaho (~500 MW), Oregon Potential for boom and bust cycles
Retail Procurement
Green tariffs / community solar Direct access (e.g., Apple 130 + 150 MW California Flats PPA) Community Choice Aggregation
Merchant Solar
e.g., Barilla Solar in Texas
Clean Power Plan???
6
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
7 Photo Credit: sPower SEPV Palmdale East
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
2015 Trends:
Strong growth in c-Si capacity (81%) relative to thin-film capacity (19%), driven in part by the completion of the
very large Solar Star project (594 MWAC). Largest c-Si manufacturers are SunPower (33% of c-Si market), Trina (20%), and Jinko (16%), while the thin-film market is dominated by First Solar (93% of the installed capacity).
Increasing dominance of tracking projects (70% of newly installed capacity) relative to fixed-tilt projects (30%)
8
PV project population: 278 projects totaling 9,016 MWAC
1,502 2,823 3,709 915 640 201 1,631 334 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 2007-2009 2010 2011 2012 2013 2014 2015 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500
Cumulative Capacity Additions (MWAC) Installation Year Annual Capacity Additions (MWAC)
Tracking Thin-Film Tracking c-Si Fixed-Tilt Thin-Film Fixed-Tilt c-Si Cumulative Tracking Capacity is 4,684 MW (52%)
(incl. hybrid projects with both thin-film and c-Si modules)
Cumulative Fixed-Tilt Capacity is 4,325 MW Columns represent annual capacity additions Areas represent cumulative capacity additions Cumulative c-Si Capacity is 5,211 MW (58%) Cumulative Thin-Film Capacity is 3,738 MW
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
9
Primarily fixed-tilt c-Si projects in the East Tracking (c-Si and, increasingly, thin- film) is more common in the Southwest
State Cumulative Capacity MW-AC % 2015 2014 CA 56% 59% AZ 13% 17% NV 7% 5% NC 6% 2% TX 3% 3%
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Strong percentage growth outside the established markets:
North Carolina (quadrupling previous capacity with 15 new projects) Georgia (nearly tripling previous capacity with 6 new projects totaling 177 MWAC) Nevada (more than doubling previous capacity with 4 new projects totaling 349 MWAC)
10 25% 25% 69% 76% 47% 51% 46% 55% 20% 16% 25% 41% 29% 19% 11% 8% 28% 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 <=2010 2011 2012 2013 2014 2015
Cumulative PV Capacity Additions (MWAC) Annual PV Capacity Additions (MWAC)
Installation Year
All Other States Southwest (NV, UT, AZ, NM, CO) California Columns represent annual capacity additions Areas represent cumulative capacity additions
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Historical trend of increasing solar resource quality for the average project site did not continue in 2015 – the first year where the Global Horizontal Irradiance declined
The wide 80/20 distribution of fixed-tilt PV reflects deployment throughout the US, whereas tracking PV is concentrated more in the high-GHI Southwest. However, 2015 shows an expansion of tracking into less-sunny areas (note the decline in the 20% percentile)
All else equal, higher GHI should boost sample-wide capacity factors (reported later). The effects of the lower GHI for new 2015 projects will be evaluated in next year’s report once they have been operational for a full year
11 5.01 5.39
5.28
3 4 5 6 2010 n=10 (175 MW-AC) 2011 n=35 (485 MW-AC) 2012 n=43 (946 MW-AC) 2013 n=38 (1,344 MW-AC) 2014 n=64 (3,166 MW-AC) 2015 n=83 (2,824 MW-AC)
Annual GHI (kWh/m2/day) Installation Year Fixed-Tilt PV Tracking PV All PV Markers represent capacity-weighted averages, with 20th and 80th percentiles.
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
12 1.33 1.31
1.31
1.0 1.1 1.2 1.3 1.4 2010 n=10 (175 MW-AC) 2011 n=35 (485 MW-AC) 2012 n=43 (946 MW-AC) 2013 n=38 (1,344 MW-AC) 2014 n=64 (3,166 MW-AC) 2015 n=83 (2,824 MW-AC)
Inverter Loading Ratio (ILR) Installation Year
Fixed-Tilt PV Tracking PV All PV
Markers represent capacity-weighted averages, with 20th and 80th percentiles.
As module prices have fallen (faster than inverter prices), developers have oversized the DC array capacity relative to the AC inverter capacity to enhance revenue
Fixed-tilt PV generally has a higher average ILR than tracking PV (fixed-tilt has more to gain from boosting ILR), dip in 2014 is skewed by several very large projects
The apparent decline in the capacity-weighted average ILR from 2013 to 2014 is related to several large projects – the median ILR held nearly constant in 2014
All else equal, a higher ILR should boost sample-wide capacity factors (reported later)
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
13 1 2 3 4 5 6 7 8 9 10 2007-2009 n=5 (75 MW-AC) 2010 n=10 (175 MW-AC) 2011 n=29 (428 MW-AC) 2012 n=39 (895 MW-AC) 2013 n=34 (1,287 MW-AC) 2014 n=59 (3,050 MW-AC) 2015 n=64 (2,135 MW-AC)
Installed Price (2015 $/W) Installation Year
Capacity-Weighted Average (DC) Median (DC) Individual Projects (DC) Capacity-Weighted Average (AC) Median (AC) Individual Projects (AC)
Installed prices are shown here in both DC and AC terms, but because AC is more relevant to the utility sector, all metrics used in the rest of this slide deck are expressed solely in AC terms
The lowest 20th percentile fell from $2.3/WAC ($1.8/WDC) in 2014 to $2.2/WAC ($1.6/WDC) in 2015
Capacity-weighted average prices were pushed higher in 2014 and 2015 by several very large projects that had been under construction for several years (but only entered our sample once complete)
This sample is backward-looking and may not reflect the price of projects built in 2016/2017
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
14
Not only pricing medians but also pricing modes have continued to fall (moving towards the left) each year
Share of relatively high-cost systems decreases steadily each year while share of low-cost systems increases
Interquartile price spread is the smallest in 2015, pointing to a reduction in underlying heterogeneity of prices across all installed projects
0% 5% 10% 15% 20% 25% 30% 35% 40% 45% >= $1.25 < $1.75 >= $1.75 < $2.25 >= $2.25 < $2.75 >= $2.75 < $3.25 >= $3.25 < $3.75 >= $3.75 < $4.25 >= $4.25 < $4.75 >= $4.75 < $5.25 >= $5.25 < $5.75 >= $5.75 < $6.25
Installed Price Interval ($2015/WAC)
2015 n=64 (2,135 MW-AC) 2014 n=59 (3,050 MW-AC) 2013 n=34 (1,287 MW-AC) 2012 n=39 (895 MW-AC)
Project Share of Annual Price Sample
Sample Statistics for each Installation Year:
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Pricing has converged among the various mounting/module configurations over time
The two CPV projects built in 2011 and 2012 were priced similar to PV at the time, while the 2014 CPV project was at the very low end of price distributions (unfortunately, no price data was available for the new 2015 C7 project)
15 1 2 3 4 5 6 7 8 9 2007-2009 n=5 (75 MW-AC) 2010 n=10 (175 MW-AC) 2011 n=29 (428 MW-AC) 2012 n=39 (895 MW-AC) 2013 n=34 (1,287 MW-AC) 2014 n=59 (3,050 MW-AC) 2015 n=64 (2,135 MW-AC)
Installed Price (2015 $/WAC) Installation Year Fixed-Tilt c-Si Tracking c-Si Fixed-Tilt Thin-Film Tracking Thin-Film CPV All PV Markers represent medians, with 20th and 80th percentiles.
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Not surprisingly, tracking appears to be slightly more expensive than fixed-tilt
New EIA summary statistics from project-level data for 2013 consistent with LBNL medians
16
2.5 2.8
1 2 3 4 5 6 7 8 9 2007-2009 n=5 (75 MW-AC) 2010 n=10 (175 MW-AC) 2011 n=29 (428 MW-AC) 2012 n=39 (895 MW-AC) 2013 n=34 (1,287 MW-AC) 2014 n=59 (3,050 MW-AC) 2015 n=64 (2,135 MW-AC)
Installed Price (2015 $/WAC) Installation Year Fixed-Tilt, all modules Tracking, all modules EIA Fixed-Tilt Estimate 2013 EIA Tracking Estimate 2013 All PV Markers represent medians, with 20th and 80th percentiles.
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Modular/scalable “power block” solutions from manufacturers like SunPower and First Solar may have already wrung out most of the cost savings otherwise available to larger projects.
Potential savings may not fully be passed through from EPCs to developers, or procurement savings may occur at the portfolio rather than project level
Several of the 100+ MW projects have been under construction for several years, possibly reflecting a higher-cost past. We find a correlation between increased COD-PPA lag and higher project prices (additional year of lag time results in premium of $0.5/WAC). Larger projects may face greater development, regulatory, interconnection costs that outweigh any economies of scale.
17 1 2 3 4 5 5-10 MW n=12 (108 MW-AC) 10-20 MW n=30 (475 MW-AC) 20-100 MW n=19 (853 MW-AC) 100-1,000 MW n=3 (699 MW-AC)
Installed Price (2015 $/WAC) Project Size (MWAC)
All PV Fixed-Tilt c-Si Tracking c-Si Fixed-Tilt Thin-Film Tracking Thin-Film
Markers represent medians, with 20th and 80th percentiles. Figure only includes 2015-vintage projects.
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Price differences driven in part by technology ubiquity (higher-priced tracking
Other factors may include labor costs and share of union labor, land costs, soil
18
Southwest
NV, UT, CO, AZ, NM
Southeast
AR, TN, MD, NC, GA
Northeast
NJ, MA
Not included
HI, IN, TX 3.2 3.0 2.3 2.5 2.9 2.7 2.3 2.3 1 2 3 4 California n=53 (3,419 MW-AC) Northeast n=12 (105 MW-AC) Southeast n=28 (700 MW-AC) Southwest n=21 (736 MW-AC)
Installed Price (2015 $/WAC) Select Regions of the United States 2014 median prices 2015 median prices U.S. national median 2015
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Prices presented here in DC terms, to be consistent with how presented by NREL, BNEF, GTM
LBNL’s top-down empirical prices are fairly close to modelled bottom-up prices
GTM project represents only turn-key EPC costs and excludes permitting, interconnection, transmission, developer
Difficult to ensure consistency of scope in cost categories and time horizon (under construction vs. operation date)
19
0.65 0.65 0.65 0.66 0.67 0.65 0.65 0.65 0.67 0.11 0.11 0.11 0.11 0.14 0.11 0.11 0.11 0.14 0.32 0.33 0.33 0.27 0.21 0.38 0.39 0.39 0.33 0.39 0.41 0.60 0.75 0.18 0.44 0.47 0.67 0.21 0.31 0.36 0.40 0.20 0.24 0.33 0.39 0.43 0.31 1.78 1.86 2.09 1.99 1.44 1.91 2.00 2.25 1.66 LBNL Fixed-Tilt 2.01 LBNL Tracking 2.10 $0.0 $0.5 $1.0 $1.5 $2.0 $2.5 NREL 2015 100 MW-DC National Average Non-Union Labor NREL 2015 25 MW-DC National Average Non-Union Labor NREL 2015 25 MW-DC National Average Union Labor BNEF 2015 Unknown Size California GTM 2015 10 MW-DC California EPC Only NREL 2015 100 MW-DC National Average Non-Union Labor NREL 2015 25 MW-DC National Average Non-Union Labor NREL 2015 25 MW-DC National Average Union Labor GTM 2015 10 MW-DC California EPC Only Fixed-Tilt Tracking
Project Cost or Price (2015 $/W-DC)
Other (Developer Overhead + Margin, Contingencies, Sales Tax) Design, EPC, Labor, Permitting, Interconnection, Transmission, Land Tracker / Racking, BOS Inverter Module
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Only a few utilities
O&M costs appear
Cost range among
20
30.6 29.3 28.1 18.0 15.6 19.0 14.1 13.7 8.3 7.3 4 8 12 16 20 24 28 10 20 30 40 50 60 70 2011 2012 2013 2014 2015
2015 $/MWh 2015 $/kWAC-yr
FPL CSP ($/kW-yr) Mean of all PV projects ($/kW-yr) Mean of all PV projects ($/MWh)
Whiskers represent least and most expensive utility Year PG&E PNM APS * FP&L MW-AC project # MW-AC project # MW-AC project # MW-AC project # 2011 #N/A #N/A #N/A #N/A 51 3 110 3 2012 50 3 8 2 96 4 110 3 2013 100 6 30 4 136 6 110 3 2014 150 7 55 7 168 7 110 3 2015 150 7 95 11 191 9 110 3 predominant technology Fixed-Tilt c-Si 4 fixed-tilt / 3 tracking thin-film, 4 tracking c-Si primarily tracking c-Si mix of c-Si and CSP
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Project-level variation in PV capacity factor driven by:
Solar Resource (GHI): Highest resource quartile has ~8 percentage point higher capacity factor than lowest
Tracking: Adds ~4 percentage points to capacity factor on average across all four resource quartiles
Inverter Loading Ratio (ILR): Highest ILR quartiles have ~4 percentage point higher capacity factor than lowest
21
0% 5% 10% 15% 20% 25% 30% 35% 40% 1 ILR 2 ILR 3 ILR 4 ILR 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 Fixed-Tilt Tracking Fixed-Tilt Tracking Fixed-Tilt Tracking Fixed-Tilt Tracking 1st Quartile Solar Resource 2nd Quartile Solar Resource 3rd Quartile Solar Resource 4th Quartile Solar Resource
Cumulative Net AC Capacity Factor Simple Mean Individual Project
9 projects, 106 MW 6 projects, 87 MW 8 projects, 83 MW 4 projects, 67 MW 3 projects, 36 MW 3 projects, 45 MW 7 projects, 101 MW 2 projects, 40 MW 8 projects, 116 MW 4 projects, 113 MW 10 projects, 269 MW 1 project, 23 MW 6 projects, 725 MW 6 projects, 87 MW 7 projects, 89 MW 13 projects, 606 MW 2 projects, 19 MW 12 projects, 135 MW 5 projects, 98 MW 2 projects, 38 MW 1 project, 242 MW 7 projects, 513 MW 5 projects, 84 MW 10 projects, 115 MW 4 projects, 596 MW 7 projects, 984 MW 5 projects, 98 MW 7 projects, 283 MW 1 project, 18 MW 4 projects, 72 MW
Sample includes 170 projects totaling 5,907 MWAC that came online from 2007-2014
ILR Quartile ILR Quartile ILR Quartile ILR Quartile ILR Quartile ILR Quartile ILR Quartile ILR Quartile
1 project, 20 MW
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Not surprisingly, capacity factors are highest in California and the Southwest, and lowest in the
Northeast and Midwest (with the Southeast and Texas in between)
Although sample size is small in some regions, the greater benefit of tracking in the high-insolation
regions is evident, as are the greater number of tracking projects in those regions
22 17.5% 18.3% 20.2% 20.1% 25.6% 25.6% 18.7% 19.9% 21.9% 22.7% 29.5% 30.6% 0% 5% 10% 15% 20% 25% 30% 35% Northeast Midwest Southeast Texas Southwest California Average Cumulative Net AC Capacity Factor
Fixed-Tilt Tracking
16 projects, 173 MW 1 project, 6 MW 8 projects, 86 MW 2 projects, 16 MW 13 projects, 179 MW 6 projects, 101 MW 1 project, 14 MW 7 projects, 151 MW 14 projects, 819 MW 38 projects, 773 MW 27 projects, 2,128 MW 37 projects, 1,462 MW
Regions are defined in the map on slide 9
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Higher capacity factors by vintage driven by an increase in tracking (most notably in 2011 and 2014),
average inverter loading ratio (in every year), and long-term global horizontal irradiance at project sites (in 2011 and 2013)
The fact that single-year 2015 capacity factors (blue columns) show same trend as cumulative
capacity factors (orange columns) suggests that inter-year resource variation is not much of a driver
23
21.0% 23.4% 24.1% 26.8% 26.7% 22.1% 24.2% 24.6% 27.1% 26.7% 0% 5% 10% 15% 20% 25% 30% 2010 Vintage 2011 Vintage 2012 Vintage 2013 Vintage 2014 Vintage 7 Projects 32 Projects 36 Projects 48 Projects 41 Projects 144 MW-AC 464 MW-AC 891 MW-AC 1,728 MW-AC 2,597 MW-AC 2015 Cumulative
Mean Net AC Capacity Factor
Mean ILR = 1.17 14% Tracking Mean GHI = 4.97 Mean ILR = 1.23 50% Tracking Mean GHI = 5.07 Mean ILR = 1.18 53% Tracking Mean GHI = 5.15 Mean ILR = 1.28 54% Tracking Mean GHI = 5.30 Mean ILR = 1.30 61% Tracking Mean GHI = 5.28
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Over the 8-year period shown in the graph, a degradation rate of 0.5%/year would reduce a 30%
capacity factor to 29%—a modest decline that could easily be swamped by inter-year variation in the strength of the solar resource
Though degradation is no doubt present in the graph above, the 2013-2015 decline (evident among
Western projects in particular) is more likely attributable to below-normal summer insolation
24 15% 16% 17% 18% 19% 20% 21% 22% 23% 24% 25% 26% 27% 28% 29% 30% 31% 2008 2009 2010 2011 2012 2013 2014 2015 Net AC Capacity Factor Calendar Year NV CO CO CA NV NM NV FL FL IL FL TX OH
Graph shows a time series of capacity factors by calendar year for the 13 projects in our sample that have been
least 5 (and for as many as 8) years
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Graph progresses from a 2007 fixed-tilt project with average GHI and ILR (on the far left) to a 2013 fixed- tilt project with a higher GHI and ILR (in the middle) to a 2013 tracking project with the same higher GHI and ILR (on the far right).
25
More information at: https://emp.lbl.gov/publications/maximizing-mwh-statistical-analysis
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
PPA prices are levelized over the full
term of each contract, after accounting for any escalation rates and/or time-of-delivery factors, and are shown in real 2015 dollars
Top graph shows the full sample;
bottom graph shows a sub-sample
CA and the Southwest dominate the
sample, but 2014 and 2015 saw a broadening of the market to TX, AR, AL, FL—and even MN and MI
Strong/steady downward price
trend since 2006 to <50$/MWh in 2015
Smaller projects (e.g., 20-50 MW)
seemingly no less competitive
>75% of the sample is currently
26
$0 $50 $100 $150 $200 $250 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Jan-14 Jan-15 Jan-16
PPA Execution Date
California Southwest Texas Southeast Midwest
Levelized PPA Price (Real 2015 $/MWh)
550 MW 210 MW 50 MW
4 of 5 regions now have PPA prices <$50/MWh (Midwest <$60/MWh) Sample includes 136 contracts totaling 9.1 GWAC
$0 $10 $20 $30 $40 $50 $60 $70 $80 Jan-14 Feb-14 Mar-14 Apr-14 May-14 Jun-14 Jul-14 Aug-14 Sep-14 Oct-14 Nov-14 Dec-14 Jan-15 Feb-15 Mar-15 Apr-15 May-15 Jun-15 Jul-15 Aug-15 Sep-15 Oct-15 Nov-15 Dec-15
PPA Execution Date California Southwest Texas Southeast Midwest Levelized PPA Price (Real 2015 $/MWh)
150 MW 45 MW
Sample includes 40 contracts totaling 2,562 MWAC that were priced in 2014 or 2015
7 MW 100 MW
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Top figure presents the same data as
previous slide, but in a different way: each circle is an individual contract, and the blue columns show the average levelized PPA price each year
Remarkably steady downward trend in
the average PPA price over time has slowed in recent years as average prices approached and then fell below $50/MWh
Price decline over time is more erratic
when viewed by commercial operation date (orange columns in bottom graph) rather than PPA execution date (blue columns)
Though the average levelized price of
PPAs signed in 2015 is ~$40/MWh, the average levelized PPA price among projects that came online in 2015 is significantly higher, at ~$85/MWh
27
50 100 150 200 250 2006 1 7 2007 1 5 2008 3 770 2009 16 1,030 2010 26 1,640 2011 16 1,584 2012 14 931 2013 19 568 2014 16 918 2015 24 1,644
Generation-Weighted Average Individual Contract
PPA Year: Contracts: MW: Levelized PPA Price (Real 2015 $/MWh) $0 $50 $100 $150 $200 $250 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Generation-weighted average based on the year in which commercial operation was fully achieved Generation-weighted average based on the year in which the PPA was executed Levelized PPA Price (Real 2015 $/MWh) 2016 is provisional
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
~70% of PV sample has flat annual
PPA pricing (in nominal dollars), while the rest escalate at low rates
Thus, average PPA prices decline
graph)
Bottom graph compares 2015-
vintage PPA prices to range of gas price projections from AEO 2016, showing that…
…although PV is currently priced
higher than the cost of burning fuel in a combined-cycle unit, over longer terms PV is likely to be more competitive, and can help protect against fuel price risk
28
$0 $50 $100 $150 $200 $250 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 2032 2034 2036 2038 2040 2042 2044 2046 2048 2050 Gen-Weighted Average PPA Price (2015 $/MWh)
2006 (7 MW, 1 PPA) 2009 (1,030 MW, 16 PPAs) 2008 (770 MW, 3 PPAs) 2010 (1,640 MW, 26 PPAs) 2011 (1,584 MW, 16 PPAs) 2012 (931 MW, 14 PPAs) 2013 (568 MW, 19 PPAs) 2014 (918 MW, 16 PPAs) 2007 (5 MW, 1 PPA) 2015 (1,644 MW, 24 PPAs)
10 20 30 40 50 60 70 80 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 Overall range of AEO 2016 gas price projections (converted to $/MWh terms) AEO 2016 reference case gas price projection (converted to $/MWh terms) Generation-weighted average PV PPA price over time Median PV PPA price (and 20th/80th percentile error bars) over time
2015 $/MWh PV PPA sample includes 24 contracts priced in 2015, totaling 1,644 MWAC
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
29 Photo Credit: Solar Reserve: Crescent Dunes
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
After nearly 400 MWAC built in
the late-1980s (and early- 1990s), no new CSP was built in the U.S. until 2007 (68 MWAC), 2010 (75 MWAC), and 2013-2015 (1,237 MWAC)
Prior to the large 2013-15
build-out, all utility-scale CSP projects in the U.S. used parabolic trough collectors
The five 2013-2015 projects
include 3 parabolic troughs (one with 6 hours of storage) totaling 750 MWAC (net) and two “power tower” projects (one with 10 hours of storage) totaling 487 MWAC (net)
30
CSP project population: 16 projects totaling 1,781 MWAC
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Small sample of 7 projects (5 built in 2013-15) using different technologies makes it hard to identify trends That said, there does not appear to be much of a trend (in contrast to PV’s steady downward trend) To be fair, newest projects are much larger, and include storage and/or new technology (power tower) in
some cases, making comparisons difficult
31
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
32
Capacity factors at Solana and Ivanpah improved in 2015, but were still below long-term, steady- state expectations (the ramp-up is still in progress)
Genesis maintained its 2014 capacity factor (at expectations), but similar Mojave trough project fell a little short
Newer CSP projects generally performing better than older CSP projects, but not necessarily any better than PV projects
SEGS I & II have been decommissioned (and may be replaced with PV)
0% 5% 10% 15% 20% 25% 30% 35% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Net Capacity Factor (solar portion only)
SEGS I & II SEGS III-IX Genesis Solana Ivanpah Mojave Nevada Solar One (dashed) For reference: average PV in CA, NV, AZ (red diamonds)
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
When PPAs for the most recent batch of CSP projects (with CODs of 2013-15)
But CSP has not been able to keep pace with PV’s price decline Partly as a result, no new PPAs for CSP projects have been signed since 2011
33 $0 $50 $100 $150 $200 $250 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12 Jan-13 Jan-14 Jan-15 Jan-16
PPA Execution Date
PV (for comparison) CSP trough w/o storage CSP trough w/ storage CSP tower w/o storage CSP tower w/ storage
Levelized PPA Price (Real 2015 $/MWh)
250 MW
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
5 10 15 20 25 California Texas Southwest (NV, AZ, UT, CO, NM) Southeast Central Northeast Northwest Nameplate Solar Capacity (GW) Entered queue in 2015 Total in queue at end of 2015 40 80 120 160 Gas Wind Solar Nuclear Coal Other Nameplate Capacity (GW)
December 2015’s extension of the 30% ITC through 2019 (along with the switch to a “start construction” rather than “placed in service” deadline), with a gradual phase down to 10% thereafter, should ensure continued momentum for the foreseeable future
56.8 GW of solar was in the queues at the end of 2015 (up from 44.6 GW at end of 2014): more than 5 times the installed solar capacity in our project population at the end of 2015
Solar was in third place in the queues, behind natural gas and wind
34
Graphs show solar and other capacity in 35 interconnection queues across the US:
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
The utility-scale solar pipeline has been replenished and has even grown in recent years, despite the record buildout in 2014 and 2015
Although California and (to a lesser extent) the Southwest still dominate the interconnection queues, recent growth in the queues has come largely from outside of those two traditional markets—e.g., Texas and the Southeast, Central, and Northeastern regions
Not all of these projects will ultimately be built (some will undoubtedly fall by the wayside)
35
10 20 30 40 50 60 Total California Southwest Texas Southeast Central Northeast Northwest purple shows new capacity that entered the queues in year shown 2013 2014 2015 2013 2014 2015
Solar in Queues at Year-End (GW)
green shows legacy capacity that entered the queues in prior years
Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP
Download the full report, a data file, and this slide deck at:
http://utilityscalesolar.lbl.gov
Download all of our other solar and wind work at:
http://emp.lbl.gov/reports/re
36
This research was supported by funding from the U.S. Department of Energy’s SunShot Initiative.