Utility-Scale Solar 2016Utility-Scale Solar 2016 An Empirical Analysis of Project Cost, Performance,...

Project Site: http://utilityscalesolar.lbl.gov @BerkeleyLabEMP

Utility-Scale Solar 2016

An Empirical Analysis of Project Cost, Performance, and Pricing Trends in the

United States

Mark Bolinger, Joachim Seel, Kristina Hamachi LaCommare

Lawrence Berkeley National Laboratory

September 2017

This research was supported by funding from the

U.S. Department of Energy’s SunShot Initiative.


Presentation Outline

Strong growth of the utility-scale solar market provides increasing amounts of empirical project-level data that are ripe for analysis

1. Solar deployment trends (and utility-scale’s relative contribution)

7. Future outlook

2

Key findings from analysis of the data samples (first for PV, then for CSP):

2. Project design, technology, and location

3. Installed project prices

4. Operation and maintenance (O&M) costs

5. Performance (capacity factors)

6. Power purchase agreement (“PPA”) prices


Utility-scale projects have the greatest capacity share in the U. S. solar market

The utility-scale sector accounted for 72% of all new solar capacity added in 2016 and 61% of cumulative solar capacity at the end of 2016

3

Sources: GTM/SEIA Solar Market Insight Reports, Berkeley Lab

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 (trackingthesun.lbl.gov)

9 22 70 267784

1,8032,855

3,9224,149

10,636

8,6726,337

7,0848,735

10,44510,612

6975

250

877110

0

0

25

50

75

100

125

0

4,000

8,000

12,000

16,000

20,00020

07

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

E

2018

E

2019

E

2020

E

2021

E

2022

E

Cum

ulat

ive

Sola

r Cap

acity

(GW

)

Annu

al S

olar

Cap

acity

Add

ition

s (M

W) Utility-Scale CSP

Utility-Scale PV Commercial PV Residential PV

Columns show annual capacity additions,area shows cumulative capacity

PV is shown in WDC while CSP is in WAC


Solar power was the largest source of U.S. electric-generating capacity additions in 2016

Led by the utility-scale sector, solar power has comprised >25% of all generating capacity additions in the United States in each of the past four years

In 2016, solar made up 38% of all U.S. capacity additions (with utility-scale accounting for 26%), and was the largest source of new capacity, ahead of both natural gas and wind

4

0%

5%

10%

15%

20%

25%

30%

35%

40%

0

5

10

15

20

25

30

35

40

2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Sola

r Cap

acity

Add

ition

s (%

of T

otal

)

Tota

l Ann

ual C

apac

ity A

dditi

ons (

GW

AC)

Utility-Scale SolarDistributed SolarWindOther REGasCoalOther non-RE

Total Solar (right axis)

Distributed Solar (right axis)

Utility-Scale Solar(right axis)

Sources: ABB, AWEA, GTM/SEIA Solar Market Insight Reports, Berkeley Lab

Note: This graph follows GTM/SEIA’s split between distributed and utility-scale solar, rather than our 5 MWAC threshold


Solar penetration rates approaching or exceeding 10% in several states

5

• Solar penetration rate varies considerably depending on whether calculated as a percentage of generation or load (e.g., see Vermont)

• Contribution of utility-scale also varies (a minority in northeast states and Hawaii, a majority in other states and overall)


Utility-Scale PV

6

Photo Credit: Community Solar Amazon Solar Farm US East 1


Historically heavy concentration in the Southwest and mid-Atlantic, but now spreading to Southeast and Northwest

7

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 %

2016 2015 CA 54% 56% AZ 9% 12% NV 8% 7% GA 6% 3% NC 5% 6%

2015


Historically heavy concentration in the Southwest and mid-Atlantic, but now spreading to Southeast and Northwest

8

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 %

2016 2015 CA 54% 56% AZ 9% 12% NV 8% 7% GA 6% 3% NC 5% 6%

2016


Utility-scale PV continues to expand beyond California and the Southwest

Strong percentage growth outside the established markets: 7 new states added their first utility-scale solar project: OR, ID, MN, VA, AL, KY, SC Georgia added 726 MWAC – the second-largest amount of new solar capacity among all states in 2016 Texas doubled its annual new capacity with 263 MWAC Florida started growth spree with 229 MWAC – with substantially more planned for coming years

9

69%

76%47%

40%

55%

20%

16%

24%

29%

22%

21%

8%

4% 7%

9%

0

2

4

6

8

10

12

14

16

0

1

2

3

4

5

6

7

8

<=2010 2011 2012 2013 2014 2015 2016

Annu

al P

V Ca

paci

ty A

dditi

ons (

GW

AC)

Installation Year

All Other States Southeast Southwest California

Cum

ulat

ive

PV C

apac

ity (G

WAC

)

Columns show annual capacity additions (left scale)

Areas show cumulative capacity (right scale)

PV project population: 427 projects totaling 16,439 MWAC


The eastward expansion is reflected in the buildout of lower-insolation sites

2016 was the 3rd year of declining median solar resource (measured in long-term global horizontal irradiance (GHI)) as the market expands to less-sunny states

Fixed-tilt PV is increasingly relegated to lower-insolation sites (note the decline in its 80th percentile), while tracking PV is increasingly pushing into those same areas (note the decline in its 20th percentile)

All else equal, the buildout of lower-GHI sites will dampen sample-wide capacity factors (reported later)

10

3.5

4.0

4.5

5.0

5.5

6.0

2010n=10

175 MW

2011n=34

478 MW

2012n=43

946 MW

2013n=38

1,344 MW

2014n=64

3,166 MW

2015n=87

2,870 MW

2016n=146

7,385 MW

Annu

al G

HI (k

Wh/

m2 /

day)

Installation Year

All PV Fixed-Tilt PV Tracking PV

Median values shown, with error bars indicating 20th and 80th percentiles


PV project population broken out by tracking vs. fixed-tilt, module type, and installation year

2016 Trends: Increasing dominance of tracking projects (79% of newly installed capacity) relative to fixed-tilt projects (21%)

Continued strong growth in c-Si capacity (77%) relative to thin-film capacity (23%). Largest c-Si manufacturers are Trina (22%), and Jinko (14%), Canadian Solar (14%) and SunPower (8%), while the thin-film market is dominated by First Solar (97% of the installed capacity).

11

PV project population: 427 projects totaling 16,439 MWAC

0

2

4

6

8

10

12

14

16

2007-2009 2010 2011 2012 2013 2014 2015 20160

1

2

3

4

5

6

7

8

Cum

ulat

ive

Capa

city

(GW

AC)

Installation Year

Annu

al C

apac

ity A

dditi

ons (

GW

AC)

Tracking Thin-Film Tracking c-Si Fixed-Tilt Thin-Film Fixed-Tilt c-Si

Columns show annual capacity additions (left scale)

Areas show cumulative capacity (right scale)

0.62

0.92

4.74

1.11

3.45

2.02

2.42

8.48


The median inverter loading ratio (ILR) has risen over time, though not much since 2013

12

As module prices have fallen (faster than inverter prices), developers have oversized the DC array capacity relative to the AC inverter capacity (i.e., the ILR) to enhance revenue

The ILR (DC:AC ratio) seems to have stabilized around 1.3 on average, though considerable variation remains

Fixed-tilt PV has more to gain from a higher ILR than does tracking PV; the highest ILR projects tend to be fixed-tilt

All else equal, a higher ILR should boost capacity factors (reported later)

1.00

1.05

1.10

1.15

1.20

1.25

1.30

1.35

1.40

1.45

2010n=10

175 MW

2011n=34

478 MW

2012n=43

946 MW

2013n=38

1,344 MW

2014n=64

3,166 MW

2015n=87

2,870 MW

2016n=144

7,325 MW

Inve

rter

Load

ing

Ratio

(ILR

)

Installation Year

All PV

Fixed-Tilt PV

Tracking PV



Median installed price of PV has fallen steadily, by over 65%, to around $2.2/WAC ($1.7/WDC) in 2016

13

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 $2.2/WAC ($1.6/WDC) in 2015 to $2.0/WAC ($1.5/WDC) in 2016 Minimum price among our 88 projects in 2016 was $1.5/WAC ($1.1/WDC) This sample is backward-looking and may not reflect the price of projects built in 2017/2018

0

1

2

3

4

5

6

7

8

9

10

2007-2009n=5

75 MW

2010n=10

175 MW

2011n=29

428 MW

2012n=40

915 MW

2013n=38

1,344 MW

2014n=64

3,166 MW

2015n=87

2,870 MW

2016n=88

5,497 MW

Inst

alle

d Pr

ice

(201

6 $/

W)

Installation Year

Median (DC) Individual Projects (DC) Median (AC) Individual Projects (AC)

1.0

1.5

2.0

2.5

3.0

3.5

4.0

2016

Inst

alle

d Pr

ice

(201

6 $/

W)


Pricing distributions have continuously moved towards lower prices over the last 5 years

14

Both medians and 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 Price spread is the smallest in 2016, pointing to a reduction in underlying heterogeneity of prices across

all installed projects

0%

10%

20%

30%

40%

50%

60%

≥ $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 (2016 $/WAC)

2016n=885,497 MW

2015n=872,870 MW

2014n=643,166 MW

2013n=381,344 MW

2012n=40915 MW

Proj

ect S

hare

of A

nnua

l Pric

e Sa

mpl

e


Tracking projects were $0.15/WAC more costly (at the median) than fixed-tilt projects in 2016

Tracking’s empirical cost premium has varied somewhat over time, but in general has declined considerably since 2010

Upfront cost premium usually compensated by higher annual generation

15

0

1

2

3

4

5

6

7

8

9

2007-2009n=5

75 MW

2010n=10

175 MW

2011n=29

428 MW

2012n=40

915 MW

2013n=38

1,344 MW

2014n=64

3,166 MW

2015n=87

2,870 MW

2016n=88

5,483 MW

Inst

alle

d Pr

ice

(201

6 $/

WAC

)

Installation Year

All PV

Fixed-Tilt PV

Tracking PV



2016 project sample hints at possible economies of scale (at least up to 100 MW)

Modest economies of scale evident in the sample, from $2.3/WAC for projects smaller than 20MWAC to $2.1/WAC for projects between 50 and 100MWAC

But higher costs for the 100+ MW projects, several of which have been under construction for several years, possibly reflecting a higher-cost past. In addition, larger projects may face greater development, regulatory, and interconnection costs that could outweigh any economies of scale.

16

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

5-20 MWn=35

505 MW

20-50 MWn=13

455 MW

50-100 MWn=25

1,952 MW

>100 MWn=15

2,585 MW

Inst

alle

d Pr

ice

(201

6 $/

WAC

)

Project Size Range (MWAC)

All PV

Fixed-Tilt PV

Tracking PV

Median values shown, with error bars indicating 20th and 80th percentiles.

Figure only includes 2016-vintage projects.


Project prices vary by region

Price differences driven in part by technology ubiquity (e.g., higher-priced tracking projects are more prevalent in the Southwest and California)

Other factors may include labor costs and share of union labor, land costs, soil conditions or snow load, and balance of supply and demand

17

Southwest: NV, UT, CO, AZ, NM

Southeast: AR, AL, FL, GA, KY, MD, NC, SC, VA

Northeast: NJ, NY

Midwest: IN, MN

Northwest: ID, OR

Not included: HI, TX

3.0

2.4 2.4

3.1

3.7

#N/A

2.42.1 2.1

#N/A

1.9

2.5

0

1

2

3

4

Californian=67

3,544 MW-AC

Southwestn=33

2,303 MW-AC

Southeastn=52

1,816 MW-AC

Northeastn=6

57 MW-AC

Midwestn=8

143 MW-AC

Northwestn=3

98 MW-AC

Inst

alle

d Pr

ice

(201

6 $/

WAC

)

Select Regions of the United States

2015 2016 U.S. national median 2016

Bars show median values, with 20th and 80th percentiles.


Bottom-up models roughly consistent with LBNL’s top-down findings

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 overhead, fees, and profit margins Difficult to ensure consistency of scope in cost categories and time horizon (under construction vs. operation

date)

18

Prices are presented here in $/WDC for consistency with how they are presented by NREL, BNEF, and GTM 0.64 0.64 0.64

0.48 0.53 0.64 0.64 0.640.48 0.53

0.10 0.10 0.100.07 0.11

0.09 0.09 0.090.07 0.11

0.20 0.27 0.27

0.20 0.150.26 0.32 0.32

0.250.25

0.320.38

0.55

0.29 0.36

0.340.41

0.60

0.320.40

0.160.27

0.29

0.18

0.16

0.28

0.31

0.18

1.42

1.66

1.85

1.231.14

1.49

1.75

1.96

1.30 1.28

LBNL Fixed-Tilt: 1.55LBNL Tracking: 1.73

$0.0

$0.5

$1.0

$1.5

$2.0

NREL 2016100 MW-DC

NationalAverage

Non-UnionLabor

NREL 201625 MW-DC

NationalAverage

Non-UnionLabor

NREL 201625 MW-DC

NationalAverage

Union Labor

BNEF 2016NationalAverage

c-Si

GTM 201610 MW-DC

NationalAverageEPC Only

NREL 2016100 MW-DC

NationalAverage

Non-UnionLabor

NREL 201625 MW-DC

NationalAverage

Non-UnionLabor

NREL 201625 MW-DC

NationalAverage

Union Labor

BNEF 2016NationalAverage

c-Si

GTM 201610 MW-DC

NationalAverageEPC Only

Fixed-Tilt Tracking

Proj

ect C

ost o

r Pric

e (2

016

$/W

DC)

Other (Developer Overhead + Margin, Contingencies, Sales Tax) Design, EPC, Labor, Permitting, Interconnection, Transmission, Land Tracker / Racking, BOS Inverter Module


O&M cost data still very thin

Only a few utilities report solar O&M costs, slow emergence of project-specific O&M costs

O&M costs appear to be declining over time, to $17.8/kW-year and $8.2/MWh in 2016 (slight increase from 2015)

Cost declines may reflect economies of scale

Cost range among utilities continues to be large

19

Year PG&E PNM Nevada Power Georgia Power APS PSEG FP&L

MWAC project # MWAC project # MWAC project # MWAC project # MWAC project # MWAC project # MWAC project #

2011 #N/A #N/A #N/A #N/A #N/A #N/A #N/A #N/A 51 3 #N/A #N/A 110 3 2012 50 3 8 2 #N/A #N/A #N/A #N/A 96 4 #N/A #N/A 110 3 2013 100 6 30 4 #N/A #N/A #N/A #N/A 136 6 #N/A #N/A 110 3 2014 #N/A #N/A 55 7 #N/A #N/A #N/A #N/A 168 7 #N/A #N/A 110 3 2015 150 9 95 11 #N/A #N/A #N/A #N/A 191 9 #N/A #N/A 110 3

2016 150 9 95 11 16 1 36 2 237 10 44 3 110 3

predominant technology Fixed-Tilt c-Si 4 Fixed-Tilt,

7 Tracking Tracking c-Si Fixed-Tilt c-Si Tracking c-Si Fixed-Tilt c-Si mix of c-Si and CSP


25.8% average sample-wide PV net capacity factor, but with large project-level range (from 15.4%-35.5%)

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

20

0%

5%

10%

15%

20%

25%

30%

35%

40%

1ILR

2ILR

3ILR

4ILR

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

Cum

ulat

ive

Net

AC

Capa

city

Fac

tor Simple Mean

Individual Project

14 p

roje

cts,

165

MW

8 pr

ojec

ts, 1

20 M

W

15 p

roje

cts,

202

MW

3 pr

ojec

ts, 3

7 M

W

4 pr

ojec

ts, 5

4 M

W

7 pr

ojec

ts, 1

14 M

W

10 p

roje

cts,

170

MW

4 pr

ojec

ts, 8

9 M

W

11 p

roje

cts,

237

MW

8 pr

ojec

ts, 3

11 M

W

13 p

roje

cts,

332

MW

1 pr

ojec

t, 23

MW

4 pr

ojec

ts, 6

05 M

W

8 pr

ojec

ts, 1

06 M

W

16 p

roje

cts,

945

MW

12 p

roje

cts,

580

MW

6 pr

ojec

ts, 6

1 M

W

13 p

roje

cts,

146

MW

3 pr

ojec

ts, 5

3 M

W

9 pr

ojec

ts, 3

25 M

W

2 pr

ojec

ts, 2

74 M

W

10 p

roje

cts,

323

MW

15 p

roje

cts,

464

MW

14 p

roje

cts,

183

MW

6 pr

ojec

ts, 7

36 M

W

10 p

roje

cts,

1,1

73 M

W

4 pr

ojec

ts, 7

9 M

W

14 p

roje

cts,

350

MW

4 pr

ojec

ts, 1

58 M

W

4 pr

ojec

ts, 1

32 M

W

Sample includes 260 projects totaling 8,733 MWAC that came online from 2007-2015

ILR Quartile ILR QuartileILR QuartileILR QuartileILR QuartileILR QuartileILR QuartileILR Quartile

7 pr

ojec

ts, 1

76 M

W

1 pr

ojec

t, 10

MW


For those who prefer to think geographically rather than in terms of insolation quartiles…

Not surprisingly, capacity factors are highest in California and the Southwest, and lowest in the Northeast and Midwest

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

21

Regions are defined in the map on slide 8

18.0% 19.0%20.7% 21.0% 21.6%

25.9% 25.2%

18.8%20.5%

23.7% 24.0%

29.3% 30.2%

0%

5%

10%

15%

20%

25%

30%

35%

Northeast Midwest Southeast Hawaii Texas Southwest California

Aver

age

Cum

ulat

ive

Net

AC

Capa

city

Fac

tor Fixed-Tilt Tracking

21 p

roje

cts,

219

MW

1 pr

ojec

t, 6

MW

10 p

roje

cts,

96

MW

3 pr

ojec

ts, 2

6 M

W

27 p

roje

cts,

474

MW

15 p

roje

cts,

375

MW

3 pr

ojec

ts, 3

0 M

W

15 p

roje

cts,

1,0

69 M

W

2 pr

ojec

ts, 4

4 M

W

10 p

roje

cts,

262

MW

33 p

roje

cts,

2,2

36 M

W

70 p

roje

cts,

2,7

60 M

W

50 p

roje

cts,

1,1

36 M

W

No

trac

king

proj

ects

in H

I yet


More recent PV project vintages have higher capacity factors on average

Average capacity factors driven higher from 2010- to 2013-vintage projects by an increase in ILR (from 1.17 to 1.28), tracking (from 14% to 54%) and average site-level GHI (from 4.97 to 5.29).

But since 2013, average long-term site-level GHI has decreased while tracking has increased (with ILR roughly unchanged), leading to stagnation in capacity factors among 2014 and 2015 projects.

22

21.5%23.6% 24.6%

26.7% 26.4% 26.5%

22.0%24.1% 24.8%

26.9% 26.2% 26.5%

0%

5%

10%

15%

20%

25%

30%

2010 Vintage 2011 Vintage 2012 Vintage 2013 Vintage 2014 Vintage 2015 Vintage

7 Projects 30 Projects 37 Projects 48 Projects 53 Projects 78 Projects

144 MW-AC 440 MW-AC 892 MW-AC 1,720 MW-AC 2,785 MW-AC 2,660 MW-AC

2016 Cumulative

Mea

n N

et A

C Ca

paci

ty F

acto

r

ILR = 1.17

14%Tracking

GHI = 4.97

ILR = 1.23

49%Tracking

GHI = 5.13

ILR = 1.18

50%Tracking

GHI = 5.17

ILR = 1.28

54%Tracking

GHI = 5.29

ILR = 1.29

60%Tracking

GHI = 5.19

ILR = 1.30

67%Tracking

GHI = 5.11


Performance degradation is evident, but difficult to assess and attribute at the project-level

Fleetwide degradation appears to exceed the 0.5%/year benchmark commonly assumed in PPAs and pro forma models

Contributing factors (other than actual degradation) could include inter-year resource variability (e.g., several bad solar years in a row), curtailment (which has become an issue in California – the largest market), and an inconsistent sample (which drops off quickly) in each successive year

23

Graph shows indexed capacity factors in each full calendar year following COD. No attempt has been made to correct for inter-year resource variation or other factors.

80%

85%

90%

95%

100%

105%

1 2 3 4 5 6 7 8 9

260 182 129 81 43 13 6 3 1

8,733 6,073 3,288 1,567 667 227 83 29 7

Median (with 20th/80th percentile error bars) Capacity-Weighted Average Simple Average Representative 0.5%/year degradation rate

Years post-COD:

Sample projects:

Sample MWAC:

Inde

xed

Capa

city

Fac

tor (

Year

1=1

00%

)

Sample includes projects with COD from 2007-2015


Combination of falling installed prices and better project performance enables lower PPA prices

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 2016 dollars

Top graph shows the full sample; bottom graph shows a sub-sample of PPAs signed post-2014

CA and the Southwest dominate the sample, but in recent years the market has expanded to other regions

Hawaii projects (included here for the first time) show a consistent and significant premium over the mainland

Three PPAs featuring PV plus long-duration battery storage do not seem to be priced at a prohibitive premium to their PV-only counterparts

Smaller projects (e.g., 20-50 MW) are seemingly no less competitive

>90% of the sample is currently operational

24

$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

Jan-

17

PPA Execution Date

California

Southwest

Texas

Southeast

Midwest

Hawaii

Leve

lized

PPA

Pric

e (R

eal 2

016

$/M

Wh)

550 MW

210MW

50 MW

Sample includes 189 contracts totaling 11.7 GWAC

12 MW

$0

$20

$40

$60

$80

$100

$120

$140

Jan-

15Fe

b-15

Mar

-15

Apr-

15M

ay-1

5Ju

n-15

Jul-1

5Au

g-15

Sep-

15O

ct-1

5N

ov-1

5D

ec-1

5Ja

n-16

Feb-

16M

ar-1

6Ap

r-16

May

-16

Jun-

16Ju

l-16

Aug-

16Se

p-16

Oct

-16

Nov

-16

Dec

-16

Jan-

17Fe

b-17

Mar

-17

Apr-

17M

ay-1

7Ju

n-17

Jul-1

7Au

g-17

Sep-

17

PPA Execution Date

California Southwest Texas Southeast Midwest Hawaii

Leve

lized

PPA

Pric

e (R

eal 2

016

$/M

Wh)

100MW

Sample includes 52 PPAs totaling 3,276 MWAC that were signed since the start of 2015

12 MW

includes battery storage

50MW


On average, levelized PPA prices fell by >75% from 2009 through 2016

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

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 COD (orange bars in bottom graph) rather than by PPA execution date (blue bars)

Though the average levelized price of PPAs signed in 2016 is ~$35/MWh, the average levelized PPA price among projects that came online in 2016 is significantly higher, at ~$60/MWh

2017 is provisional and currently reflects a very small sample and a high proportion of high-priced Hawaiian PPAs, plus several PPAs with long-duration battery storage

25

0

50

100

150

200

250

200617

200715

20083

770

200916

1,030

201030

1,746

201120

1,790

201217

1,073

201319

478

201430

1,503

201539

2,383

20166

556

20177

337

Generation-Weighted Average

Individual PPA (Hawaii PPAs shaded orange)

PPA Year:Contracts:

MW:

Leve

lized

PPA

Pric

e (R

eal 2

016

$/M

Wh)

$0

$50

$100

$150

$200

$250

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Generation-weighted average based on the year in whichcommercial operation was fully achieved

Generation-weighted average based on the year in whichthe PPA was executed

Leve

lized

PPA

Pric

e (R

eal 2

016

$/M

Wh)

2017 isprovisional


The value of solar has declined in America’s largest solar market

With increasing solar penetration in California, solar curtailment has increased and solar’s wholesale energy value has declined

In 2012, when solar penetration was ~2%, solar earned 126% of the average wholesale power price

In 2016, with solar penetration at ~12%, solar earned just 83% of the average wholesale power price

Based on data for the first half of the year, this value decline is likely to continue in 2017 (1Q17 was particularly bad – bottom graph)

Most other markets are not yet facing this value decline, due to lower levels of solar penetration

26

0%

15%

30%

45%

60%

75%

90%

105%

120%

135%

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

2012 2013 2014 2015 2016 1H2017

Solar Penetration Rate (left axis) Solar Curtailment Rate (left axis) Solar Value Factor (right axis)

Ener

gy V

alue

of S

olar

Rel

ativ

e to

a 2

4x7

Flat

Blo

ck

CAIS

OSo

lar P

enet

ratio

n an

d Cu

rtai

lmen

t Rat

es

CAISO curtailment data not available prior to 2015

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

20%

2015 2016 2017 2015 2016 2017 2015 2016 2017 2015 2016 2017

1Q 2Q 3Q 4Q

Solar Penetration Rate (left scale) Solar Curtailment Rate (left scale) Solar Value Factor (right scale)

Sola

r Cur

tailm

enta

nd P

enet

ratio

n Ra

tes

Ener

gy V

alue

of S

olar

Rel

ativ

e to

24x

7 Fl

at B

lock


Levelized PPA prices track the LCOE of utility-scale PV reasonably well

Using empirical data from elsewhere in the report, along with a number of assumptions (e.g., about financing), we calculated project-level LCOEs for the entire sample of projects for which we have CapEx data (14.3 GWAC)

Central estimates of LCOE track median PPA prices (levelized over 30 years in this case, and shown by COD rather than by execution date) reasonably well, suggesting a fairly competitive PPA market

PPAs are lower than LCOEs because they reflect receipt of the 30% ITC and perhaps also state-level incentives

27

NOTE: LCOE calculations do NOT include the 30% ITC (whereas PPA prices do reflect the ITC, and perhaps also state-level incentives)

0

50

100

150

200

250

300

350

201010

175

201128

423

201238

863

201338

1,344

201463

3,160

201586

2,883

201683

5,407

2017TBDTBD

2016

$/M

Wh

Capacity-Weighted Average LCOE Median LCOE Simple Average LCOE Individual Project LCOE Median Levelized PPA Price (by COD)

COD:Projects:MW-AC:


PV PPA prices generally decline over time in real dollar terms, in contrast to fuel cost projections

Two-thirds of PV sample has flat annual PPA pricing (in nominal dollars), while the rest escalate at low rates

Thus, average PPA prices tend to decline over time in real dollar terms (top graph)

Bottom graph compares recent PPA prices to range of gas price projections from AEO 2017, showing that…

…although PV is currently priced higher than the cost of burning fuel in a combined-cycle unit, over longer terms PV is perhaps 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

2050G

en-W

eigh

ted

Aver

age

PPA

Pric

e (2

016

$/M

Wh)

2006 (7 MW, 1 PPA)

2008 (770 MW, 3 PPAs)

2010 (1,746 MW, 30 PPAs)

2014(1,503 MW, 30 PPAs)

2007(5 MW, 1 PPA)

2015 (2,383 MW, 39 PPAs)2016 (566 MW, 6 PPAs)

0

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

2041

2042

2043

2044

2045

2046

2047

2048

2049

2050

2016

$/M

Wh

Overall range of AEO 2017 gas price projections (converted to $/MWh terms) AEO 2017 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 bars) over time

PV PPA sample includes 29 PPAs signed 7/2015-8/2017 and totaling 2,184 MWAC(PPA sample excludes Hawaiian projects)


Utility-Scale Concentrating Solar Thermal Power (CSP)

29

Photo Credit: Solar Reserve: Crescent Dunes


Sample description of CSP projects

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


Not much movement in the installed price of CSP

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 thermal storage and/or new technology (power tower) in some cases, making comparisons difficult

31

0

1

2

3

4

5

6

7

8

9

10

11

2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

Inst

alle

d Pr

ice

(201

6 $/

WAC

)

Installation Year

CSP Trough CSP Tower Median PV (for reference)

68 MWAC

250 MWAC with 6 hours of storage

75 MWAC 250 MWAC each

377 MWAC

110 MWAC with 10 hours of storage


Several newer CSP projects continued to underperform relative to long-term expectations

32

The two “power tower” projects (Ivanpah and Crescent Dunes) were hit with closures in 2016 that negatively impacted capacity factors. The Crescent Dunes closure lasted into 2017.

Solana was at reduced capacity for part of 2016 due to micro-burst storm damage, and for part of 2017 due to a transformer fire.

Genesis and Mojave were both largely on target in 2016 Most newer CSP projects generally performing better than older CSP projects, but not necessarily

any better than (and in some cases worse than) local PV projects

0%

5%

10%

15%

20%

25%

30%

35%

2008 2009 2010 2011 2012 2013 2014 2015 2016

SEGS I & II

SEGS III-IX

GenesisSolana

Ivanpah

MojaveNevada Solar One (dashed)

For reference: average PV in CA, NV, AZ (red diamonds)

CrescentDunes

Net

Cap

acity

Fac

tor(

sola

r por

tion

only

)


Though once competitive, CSP PPA prices have failed to keep pace with PV’s price decline

When PPAs for the most recent batch of CSP projects (with CODs of 2013-15) were signed back in 2009-2011, they were still mostly competitive with PV

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 in the U.S. since 2011

– though the technology continues to advance overseas

33

$0

$50

$100

$150

$200

$250

Jan-

07

Jan-

08

Jan-

09

Jan-

10

Jan-

11

Jan-

12

Jan-

13

Jan-

14

Jan-

15

Jan-

16

Jan-

17

PPA Execution Date

PV in CA, NV, AZ (for comparison)

CSP trough w/o storage

CSP trough w/ storage

CSP tower w/o storage

CSP tower w/ storage

Leve

lized

PPA

Pric

e (R

eal 2

016

$/M

Wh)

250 MW


0

5

10

15

20

25

30

35

40

California Southeast Southwest Northeast Texas Central Northwest

2013

2014

2015

2016So

lar i

n Q

ueue

s at Y

ear-

End

(GW

)

020406080

100120140

Wind Solar Gas Other Storage Nuclear Coal

Entered queues in that year

2016

GW

in Q

ueue

s at Y

ear-

End

Entered queues in prior years

Looking ahead: long-term ITC extension should support continued growth in the utility-scale solar pipeline

121.4 GW of solar was in the queues at the end of 2016—up from 56.8 GW at end of 2015, and more than six times the amount of installed capacity at the end of 2016

83.3 GW of the 121.4 GW total first entered the queues in 2016 Very strong solar growth in all regions, with the possible exception of the Northwest The Southeast moved ahead of the Southwest for the number two position behind California

34

Graphs show solar and other capacity in 35 interconnection queues across the US:

• Inset compares solar to other resources (2016 only)

• Main graph shows location of solar (2013-2016)

• Not all of these projects will ultimately be built!


Questions?

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

Follow the Electricity markets & Policy Group on Twitter:

@BerkeleyLabEMP

35

Contact:

Mark Bolinger: [email protected]

Joachim Seel: [email protected]

This research was supported by funding from the U.S.

Department of Energy’s SunShot Initiative.

http://utilityscalesolar.lbl.gov/

http://emp.lbl.gov/reports/re

Date post:	14-Aug-2020
Category:	Documents
Upload:	others
View:	3 times
Download:	0 times

Utility-Scale Solar 2016Utility-Scale Solar 2016 An Empirical Analysis of Project Cost, Performance,...

Documents