The Cannibalization Eﬀect of Wind and Solar in the California Wholesale Electricity Market

(1)

The Cannibalization Effect of Wind and Solar in the California Wholesale Electricity Market

Javier López Prol ^1,2 Karl W. Steininger ^1,2 David Zilberman ³

1

Department of Economics, University of Graz, Austria

2

Wegener Center for Climate and Global Change

3

Agricultural and Resource Economics, University of California Berkeley, USA

April 2018

(2)

Outline

I The rise of variable renewables

I The merit order effect

I The cannibalization effect

I

Absolute cannibalization (Unit revenues)

I

Relative cannibalization (Value factor)

I Conclusions

(3)

The rise of variable renewables in California

0 5 10 15

2013 2014 2015 2016 2017

P enetr ation (%)

source

solar

wind

(4)

The rise of variable renewables in California

(5)

The rise of variable renewables in California

0 5 10 15 20 25

2013 2014 2015 2016 2017

P enetr ation (%)

source

solar

wind

(6)

The merit-order effect

I Renewable energies pressure down electricity prices

Source: CLEW 2016

(7)

The cannibalization effect

The higher solar/wind electricity penetration, the lower its value

I Absolute cannibalization: (solar) daily unit revenues (p ^s _d ):

p _t ^s = P 24

h=1 p h q _h ^s P 24

h=1 q _h ^s

I Relative cannibalization: value factor (VF):

unit revenue (p ^s _t ) divided by daily avg. wholesale price (p _t )

VF _t ^s = p _t ^s p t

= P

24

h=1

p

_h

q

^s_h

P

24 h=1

q

_h^s

P

24 h=1

p

h

24 p = price; q = quantity ;

s : solar ; _h : hour ; _t : time (daily )

(8)

Unit revenue and value factor visualized

VF t > 1 ⇔ p _t ^s > p t

VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35 VF = 1.35

Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue Unit revenue

Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price Avg. electricity price

Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price Hourly electricity price

Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation Normalized generation 0

30 60 90

00 02 04 06 08 10 12 14 16 18 20 22 00 hour

$/MWh

2013−06−30

VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5 VF = 0.5

0 10 20 30 40 50

00 02 04 06 08 10 12 14 16 18 20 22 00 hour

$/MWh

2016−03−29

(9)

Absolute cannibalization: Unit Revenues

solar wind

2013 2014 2015 2016 2017 2013 2014 2015 2016 2017 0

20 40 60

$/MWh

(10)

Electricity prices

0 20 40 60

2013 2014 2015 2016 2017

Electr icity pr ice ($/MWh)

(11)

Relative cannibalization: Value Factors

solar wind

2013 2014 2015 2016 2017 2013 2014 2015 2016 2017

−50 0 50 100 150

%

(12)

Modeling the cannibalization effect

Dependent variables: solar & wind unit revenues & value factors

y ^{s,w} = α + β ₁ solar _sh _t + β ₂ wind _sh _t + β ₃ consumption _t + β ₄ gas _price _t + γ ⁰ D _t + _t

D = vector of daily, monthly and yearly time dummies

(13)

In summary

We estimate

I absolute (unit revenues) and relative (value factors)

I cannibalization (within technologies) and cross-cannibalization (between technologies) effects,

I of wind and solar technologies,

I for centralized-only and distributed-included data,

I with (diff.) OLS and Prais-Winsten FGLS

(14)

Results I: Absolute cannibalization

Solar Unit Revenue Wind Unit Revenue

−2.0 −1.5 −1.0 −0.5 0.0 −0.75 −0.50 −0.25 0.00 Solar

Wind

Solar Wind

Coefficient ($/MWh)

P enetr ation

Estimation

Prais−Winsten (Diff.) OLS

Data

Distributed included Centralized only

I Newey-West HAC standard errors 95% CI

(15)

Results II: Relative cannibalization

Solar Value Factor Wind Value Factor

−8 −6 −4 −2 0 0.0 0.5 1.0

Solar Wind

Coefficient (pp)

P enetr ation

Estimation

Prais−Winsten (Diff.) OLS

Data

Distributed included Centralized only

I Newey-West HAC standard errors 95% CI

(16)

Results III: exploring non-linarities I: Consumption

Solar Value Factor Wind Value Factor

600 650 700 750 600 650 700 750

0 1

−4

−2 0

Consuption (GWh)

Estimate (p .p .)

term

Solar penetration Wind penetration

I OLS estimation with NW HAC SE 95% CI

(17)

Results VI: exploring non-linarities II: penetration

Solar penetration Wind penetration

5 10 15 20 2.5 5.0 7.5 10.0 12.5

−5.0

−2.5 0.0

−5.0

−2.5 0.0 2.5

Penetration (%)

Estimate (p .p .)

dependent

Solar Value Factor Wind Value Factor

I OLS estimation with NW HAC SE 95% CI

(18)

Discussion: dissentangling the positive effect of solar penetration on wind value factor

Hourly solar & wind penetration per month in 2013 and 2016

September October November December

May June July August

January February March April

0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0

500 1000 1500 2000

0 500 1000 1500 2000

hour

MWh

2013

0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0

2000 4000 6000

0 2000 4000 6000

hour

MWh

2016

(19)

Discussion: dissentangling the positive effect of solar penetration on wind value factor

Hourly wholesale electricity prices per month, 2012-2017

0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20

25 50 75

electricity price (USD/MWh)

factor(year) 2012 2013 2014 2015 2016 2017

(20)

Conclusions

I Confirmed absolute cannibalization and cross-cannibalization (weaker) effects for both solar and wind (weaker)

I Confirmed relative cannibalization:

I Gas prices increase wind but decrease solar VF

I Omitting distributed generation overestimates cannibalization

(21)

Relevance & further discussion

I Cannibalization jeopardizes competitiveness despite LCOE decline

I CE mitigation: storage, demand flexibility, intercontinental

interconnections

(22)

Thank you

(23)

US electricity system

US interconnections

(24)

US electricity system

US electricity markets. Source: FERC

(25)

Appendix Solar power in California

PV penetration. Source: IEA

(26)

Appendix Solar power in California

Installed capacity. Source: SEIA and EIA

(27)

Appendix: California electricity mix

California electriciy mix, 2011 - 2016. Source: IEA

(28)

Appendix: Actual vs. forecasted demand

20000 30000 40000

demand_act

demand_fct

Demand forecast error

(29)

The Cannibalization Eﬀect of Wind and Solar in the California Wholesale Electricity Market

The Cannibalization Effect of Wind and Solar in the California Wholesale Electricity Market

Javier López Prol 1,2 Karl W. Steininger 1,2 David Zilberman 3

Department of Economics, University of Graz, Austria

Wegener Center for Climate and Global Change

Agricultural and Resource Economics, University of California Berkeley, USA

April 2018

Outline

I The rise of variable renewables

I The merit order effect

I The cannibalization effect

Absolute cannibalization (Unit revenues)

Relative cannibalization (Value factor)

I Conclusions

The rise of variable renewables in California

0 5 10 15

2013 2014 2015 2016 2017

P enetr ation (%)

source

solar

wind

The rise of variable renewables in California

The rise of variable renewables in California

0 5 10 15 20 25

2013 2014 2015 2016 2017

P enetr ation (%)

source

solar

wind

The merit-order effect

I Renewable energies pressure down electricity prices

Source: CLEW 2016

The cannibalization effect

The higher solar/wind electricity penetration, the lower its value

I Absolute cannibalization: (solar) daily unit revenues (p s d ):

p t s = P 24

h=1 p h q h s P 24

h=1 q h s

I Relative cannibalization: value factor (VF):

unit revenue (p s t ) divided by daily avg. wholesale price (p t )

VF t s = p t s p t

= P

p

q

P

q

P

p

24

p = price; q = quantity ;

s : solar ; h : hour ; t : time (daily )

Unit revenue and value factor visualized

VF t > 1 ⇔ p t s > p t

Absolute cannibalization: Unit Revenues

solar wind

2013 2014 2015 2016 2017 2013 2014 2015 2016 2017 0

20 40 60

$/MWh

Electricity prices

0 20 40 60

2013 2014 2015 2016 2017

Electr icity pr ice ($/MWh)

Relative cannibalization: Value Factors

solar wind

2013 2014 2015 2016 2017 2013 2014 2015 2016 2017

−50 0 50 100 150

%

Modeling the cannibalization effect

Dependent variables: solar & wind unit revenues & value factors

y {s,w} = α + β 1 solar _sh t + β 2 wind _sh t + β 3 consumption t + β 4 gas _price t + γ 0 D t + t

D = vector of daily, monthly and yearly time dummies

In summary

We estimate

I absolute (unit revenues) and relative (value factors)

I cannibalization (within technologies) and cross-cannibalization (between technologies) effects,

I of wind and solar technologies,

I for centralized-only and distributed-included data,

I with (diff.) OLS and Prais-Winsten FGLS

Results I: Absolute cannibalization

Solar Unit Revenue Wind Unit Revenue

Javier López Prol ^1,2 Karl W. Steininger ^1,2 David Zilberman ³

I Absolute cannibalization: (solar) daily unit revenues (p ^s _d ):

p _t ^s = P 24

h=1 p h q _h ^s P 24

h=1 q _h ^s

unit revenue (p ^s _t ) divided by daily avg. wholesale price (p _t )

VF _t ^s = p _t ^s p t

s : solar ; _h : hour ; _t : time (daily )

VF t > 1 ⇔ p _t ^s > p t

y ^{s,w} = α + β ₁ solar _sh _t + β ₂ wind _sh _t + β ₃ consumption _t + β ₄ gas _price _t + γ ⁰ D _t + _t