Photovoltaic and wind cost decreases implications for investment

Photovoltaic and Wind Cost Decrease:Implications for Investment Analysis

by

Ignacio Maulen.

Dept. of Economics and Business Management.Universidad Rey Juan Carlos, Madrid, Spain.

[email protected]

Index.Index.

1. Introduction.

2. Methodology.

2.1. Learning Rate & simulation.

2.2. Published PV & Wind LR.

2.3. Photovoltaic LR.

2.4. Wind LR.

2I.Maulen (Univ. Urjc, Madrid)

3. Results.

3.1. Framework.

3.2. Total investment.

3.3. Risk analysis.

4. Summary & Implications.

5. Pending research.

1. Introduction.Introduction.

Total Investment and paths implied by Renewable Energy targets.

COP 21IEA, estimates 2.5 tr. US $

Analysis:

3I.Maulen (Univ. Urjc, Madrid)

Price decreases with deployment.

Uncertain estimates: Simulations. Risk analysis

2.1. Learning Rate & simulation.2.1. Learning Rate & simulation.

Pt = k Ctb

Pt , Price (module, turbines)Ct , Capacity installed.b , Learning coefficient.

Learning Rate (LR):

Doubling capacity % price decrease

4I.Maulen (Univ. Urjc, Madrid)

Doubling capacity % price decrease

but:

b unknown estimated statistically uncertain

LR uncertain price decreases uncertain simulated

2.2. Published PV & Wind Learning Rates.2.2. Published PV & Wind Learning Rates.

REPORTED LEARNING RATES (%) (Rubin e.a, 2015)

N of studies Mean Range

Wind - onshore 18 12 -11 ; 23

Solar PV 16 22 10 ; 47

5I.Maulen (Univ. Urjc, Madrid)

Solar PV 16 22 10 ; 47

Too much variability.

Insufficient stat. detail.

2.3. The photovoltaic Learning Rate.2.3. The photovoltaic Learning Rate.

PV Cost Model Estimation.

Learning by doing, PV costs & Installed Capacity.

2.5

3

3.5

4

4.5

MODULE PRICES w.r.t. CAPACITY O.L.S. regression (logs.) & Learning Rate (LR)

Log(P)=3.98-0.33*Log(Cap) (45.) (26.)

R2=0.95, D.W.=0.44 LR=20% (23.1)

1978

1979

1980

1981

19821983

19841985

6I.Maulen (Univ. Urjc, Madrid)

-0.5

0

0.5

1

1.5

2

2.5

0 2 4 6 8 10 12

Log(

mod

ule

pric

es)

Log(Installed capacity)

19851986

1987

19881989

199019911992

199319941995

1996199719981999

20002001 20022003

20042005

2006 2007 2008

2009

2010

2011

20122013

2.3. The photovoltaic Learning Rate.2.3. The photovoltaic Learning Rate.

7I.Maulen (Univ. Urjc, Madrid)

2.4. The wind Learning Rate.2.4. The wind Learning Rate.

Insuficient reliable data world level IRENA working on data base.

Data

- Turbine prices Denmark- Turbine, LCOE, US

Overall results:

8I.Maulen (Univ. Urjc, Madrid)

Overall results:

Learning Rate ~ 13% Confidence bands estimated.

3.1. Framework.3.1. Framework.

CAPACITY FORECASTS

(world Gw.)

PV Wind Total

2020 390 640 1030

9I.Maulen (Univ. Urjc, Madrid)

2030 1720 1600 3320

2050 4670 2700 7370

IEA, International Energy Agency; Technology Roadmaps.

3.1. Framework.3.1. Framework.

Simulating Total Investment.

Total amount of funds vs. unitary price.

Price depends on investment.

( ) TIPI nn tt =1

10I.Maulen (Univ. Urjc, Madrid)

( ) TIPI ntt =1It , increase in capacity.

Pt , module / turbine price.

TIn , total accumulated investment; years 1 to n.

3.2. Total Investment.3.2. Total Investment.

TOTAL INVESTMENT (b. us $)

Random Not random

2030 mean 1580 1444

(50%) 1473

(90%) 2063

11I.Maulen (Univ. Urjc, Madrid)

2050 mean 2779 2513

Randomness => Median (50%)

Risk: 1.4 , 40% hi.

3.2. Total Investment.3.2. Total Investment.

TOTAL INVESTMENT (b. us $)

Slow Path Fast Path

2030 676 2492

2050 2765 2831

Discounting (3%)

12I.Maulen (Univ. Urjc, Madrid)

2050 1486 2066

2050, similar 2031, sharp drop (fast) Discounting

3.2. Total Investment.3.2. Total Investment.

13I.Maulen (Univ. Urjc, Madrid)

3.2. Total Investment.3.2. Total Investment.

14I.Maulen (Univ. Urjc, Madrid)

3.3. Risk analysis.3.3. Risk analysis.

Expected Investment at Risk (EIaR):

Expected Investment, if prices rise above a given high value.

[ ][ ]

TITIob =Pr

15I.Maulen (Univ. Urjc, Madrid)

[ ][ ] EIRTITITIE =|

e.g., valuexupperx ,,,%,90=TI, Total Investment

3.3. Risk analysis.3.3. Risk analysis.

16I.Maulen (Univ. Urjc, Madrid)

4. Summary & implications.4. Summary & implications.

Cost models efficiently estimated.

Learning Rates: PV > 23%

Wind ~ 13%

Parameter uncertainty SimulaDons Risk

Accelerated Investment paths: = Invest. 2050

2031 !.

17I.Maulen (Univ. Urjc, Madrid)

Caveat: Price decreases for ever ?.

4. Summary & implications.4. Summary & implications.

18I.Maulen (Univ. Urjc, Madrid)

5. Pending research.5. Pending research.

Costs.

- Balance of system costs (BoS).- Discounting. - Investment paths (speed, smoothness, )

Benefits.

19I.Maulen (Univ. Urjc, Madrid)

- GHG, - Value electricity.-

Thank you for your attention !

20I.Maulen (Univ. Urjc, Madrid)