GOING OFFSHORE: INVESTMENTS IN GERMAN WIND ENERGY UNDER UNCERTAINTY - - PowerPoint PPT Presentation

GOING OFFSHORE: INVESTMENTS IN GERMAN WIND ENERGY UNDER UNCERTAINTY

Yu-Fu Chen University of Dundee Michael Funke Hamburg University 2nd October 2015, Brescia, Italy

1

Introduction (1)

• Wind energy: the fastest-growing segment of green energy

worldwide due to abundant and reliable wind resources.

• Offshore wind is a relatively new form of renewable energy that has
• nly recently spread beyond Europe to China and a handful of other

countries.

• By December 31, 2014, 258 offshore wind turbines in the German

North Sea and Baltic Sea with a total capacity of 1,049.2 megawatts (MW). Another 220 foundations are currently under construction and further projects are in the pipeline.

• See Green and Vasilakos (2011) for comprehensive overview

2

Introduction (2)

• Germany’s roadmap for renewable deployment aims to redesign

Germany’s energy system within the next few decades.

• By 2022, all nuclear power plants are to be switched off. In addition,

by 2050 about 80 percent of electricity is to come from renewable sources, compared with 22 percent now.

• Furthermore, CO2 emissions are supposed to fall from those in 1990

by 50 percent in 2030 and 80 percent by 2050.

• The Renewable Energy Sources Act promised 20 years of

guaranteed prices to wind, solar or other renewable sources.

3

Introduction (3)

• One of the most ambitious elements is to build 6.500 MW of wind

turbines off the North Sea and Baltic Sea coasts by 2020. (Due to anti-onshore wind farm protests on their size and the noise of their blades.

• Offshore wind farms are more expensive but offer much higher

energy yields. However, these potential gains of capturing economies of scale are counterbalanced by higher investment and maintenance expenditures.

• With this background in mind, we analyze the investment in a

German offshore wind farm bearing in mind the specific German support policy framework through the prism of a real options modelling framework.

4

The Analytical Framework (1)

• Offshore investment decisions are option-rights and renewable

energy investments can be seen as an application of real options.

The Investment Incentives of the EEG Feed-In Tariff System

• The EEG 2012 aimed to increase the share of renewable energy

significantly within the next decades.

• The operating company has the opportunity to receive initial

remuneration of 19 ct/kWh for a shortened period of 8 years, provided that the wind farm is put into operation before 2018.

• In short, 19 ct/kWh for a period of 8 years; 15 ct/kWh is for the

extension period. Afterwards, a lower bound flat rate of 3.5 ct/kWh is guaranteed until the end of 20 years; and market prices for the last 5 years.

5

The Analytical Framework (2)

Figure 1: Timeline of the Multi-Stage Offshore Wind Energy Feed-in Tariff Scheme

t t =

At t= 0

t , the

windpark starts to

• perate, generating

cash flow.

1 t Time

1

τ : period of

guaranteed price

h

P , higher

guaranteed price.

2

τ : period of

guaranteed price

m

P ,

medium guaranteed price, where

m h

P P <

.

2 t

3

τ : period of guaranteed

floor price

f

P , if market

prices fall below

f

P ,

where

f m h

P P P < <

.

3 t

At 4

t , the

windpark ceases

• perations.

At t=0, the investment project decision is made and sunk cost finished at t= 0

t .

t=0 4 t

4

τ : period of

market prices received.

6

The Analytical Framework (3) Figure 2: Electricity Prices Received in the Three Operating Phases

Received nominal

electricity prices

t

Z Payoff over

three periods

1

t

2

t

3

t

time

h

P

m

P

( )

max ,

f t t

Z P P = t

4

t

Market electricity prices

t t

Z P =

are received

f

P

7

The Analytical Framework (4) Modelling of electricity prices

• Assume that the wholesale electricity prices follow the mean-

reverting stochastic process: = ( − ) +

• where

is the nominal electricity prices,

• denotes the long-term constant prices,

 is the parameter related to the mean-reverting speed of electricity prices returning to long-run equilibrium and is the volatility parameter of the standard Wiener process .

8

The Analytical Framework (5)

• The investor’s problem is to maximize the discounted value of profit

for the firm, , which is given by = (1 − ) ( − )()

• ,

where is the constant tax rate on profits, − , is the electricity price received in various phases,

• is the maximum wind farm electricity output capacity per year,

is the capacity utilization rate, is the constant annual maintenance per maximal output and is the discount rate.

9

The Analytical Framework (6)

• Pulling the deterministic parts out of the integral,

= + + (1 − ) max, − ()

• +(1 − )

− ()()

• ,

where

= (1 − ) ( − )()

• =

(1 − )( − )1 − ()() +

= (1 − ) ( − )()

• =

(1 − )( − )()() − ()() + .

10

The Analytical Framework (7)

• Approximate analytical solutions for period 3 and period 4 can be
• btained

by numerical integration:

max, −

• () &

− ()()

• The real options are only related to the uncertain part of the
• integrals. Note that the duration of the third operating phase starts

from = and lasts until . The value-matching condition for the

• ffshore wind farm at = is denoted by

=sunk cost + real optionsmax(; ), , (; ), sunk cost uncertainty, where is denoted by equation (3).

11

The Analytical Framework (8)

• Problem of mean-reverting of electricity prices uncertainty from

markets: electricity prices uncertainty affect payoffs, but not real

• ptions: uncertainty is at 12 years away with fast mean-reverting.

Value-matching implies that offshore investment is undertaken

• This implies that only the sunk costs’ uncertainty exerts an influence

upon the real options. (sunk cost uncertainty) ≡ real options(sunk cost uncertainty).

12

The Analytical Framework (9)

• The cost of electricity from offshore wind farms can be reduced by

about one-third until 2023.

• The costs for support structures and other components as well as for

the installation will also decline.

• A straightforward functional form for such declining sunk costs due

to learning by doing and experimentation is = + , where denotes the sunk cost of the sophisticated state-of-the-art technology and represents the declining variable sunk costs over time with uncertain and h > 0 determines the cost reduction over time.

13

The Analytical Framework (10)

• We assume

sunk cost = + , where follows a geometrical Brownian motion with a negative

• trend. Thus, is governed by

= −ℎ + , where is the uncertainty parameter for stochastic process and is the standard Wiener process for .

• Real Options
• = −ℎ

+ 1

2

• ,

() =

• ,

14

The Analytical Framework (11)

• Value-matching condition and its corresponding smooth-pasting

condition:

| , , , , , = + +

• .

0 =1 − − 1 2 − ℎ

• + 1

2 + ℎ

• + 2
• .

| , , , , , = +

• 1 +
• .

15

The Analytical Framework (12)

• > 0. The higher the expected payoffs from electricity prices, the

lower sunk costs are needed to undertake the investment; hence any changes in parameters that increase the value of (such as rises in

• , , , , or a fall in ).
• < 0. The higher the uncertainty about the sunk costs, the lower

the thresholds of the sunk costs.

• > 0. The higher the cost reduction over time, the lower the sunk

costs are needed to undertake the investment.

16

The Analytical Framework (13)

• We can compute the aggregate sunk cost (

) needed to undertake the offshore wind farm investment decision as

• = +

. We can then compare the required implied by the model set-up with the actual aggregate sunk cost estimates.

• We discuss the sunk costs instead of the revenue flows (“sunk

benefits”). This implies that firms should undertake the offshore investment if is larger than the estimated market sunk cost. On the contrary, if is smaller than the estimated market sunk cost, then the offshore wind farm investment will be put on hold until the market sunk cost drops below .

17

The Analytical Framework (14) Offshore Tender Bidding

• The difficulty in setting feed-in tariffs is that if the level is too high,

firms will make excessive profits. On the contrary, if it is too low, no deployment of offshore wind energy will take place.

• A solution to the problem would be to run a tender for new projects,

requiring firms to bid for the right to develop a new project.

• In Denmark, the offshore wind tender process kicked off in 2013:

price of 10-14 ct/kWh over the next 12 years, which is the period during which the offshore wind farm will receive subsidies.

18

The Analytical Framework (15) Offshore Tender Bidding Figure 3: Timeline of the Tender-Bidding Scheme

Received nominal

electricity prices

t

Z Payoff over

three periods

1

ˆ t

time

T

P : fixed tender price t

2

ˆ t

Market electricity prices

t t

Z P =

are received

19

The Analytical Framework (16) Offshore Tender Bidding

• The expected intertemporal payoffs of the tender-bidding

scheme : (| , , , ) =

+ (1 − )

− ()

• ,

where

• = (1 − )

− ()

• =

(1 − )( − )1 − (

)

+ .

• The integral (1 − )

−

• represents the

expected intertemporal payoffs from electricity prices in phase ̂, and all the other parameters are the same as in the previous section.

20

The Analytical Framework (16) Offshore Tender Bidding

• The rest of the set-up of the tender-bidding scheme is the same as

the one of the EEG tariff scheme.

(| , , , ) = +

• 1 +

1 − 1 2 − ℎ

• + 1

2 + ℎ

• + 2
• .

The threshold is found by solving the above equation

21

Model Calibration and Evaluation (1)

• We use the historical spot prices from the European Energy

Exchange (EEX) in Leipzig from 2009 to 2014: a mean-reversion rate = 3.16, a risk parameter of = 0.4366 and a long-run mean electricity price of = 3.99 ct/kWh.

• T = 25 years. The offshore wind farm has a capacity of 400 MW and

the investment expenditures cost estimate is 3750 EUR/kW in 2014. Thus, the total construction sunk cost is 1.5 billion EUR.

• The median market sunk cost: 1.43 billion EUR. Capacity utilization

rate : 0.45. = 0.9 billion EUR with a cost reduction potential parameter ℎ = 0.06 and = 0.05 . = 19 ct/kWh, = 15 ct/kWh and the floor price = 3.5 ct/kWh, with = 0, = 8, = 10, = 20 and = 25 in the EEG scenario. In the tender- bidding scheme, we assume = 12 ct/kWh within ̂ = 12 years and ̂ = 13. = 0.05 applies. : 0.03.

is assumed to be 0.10.

22

Model Calibration and Evaluation (2) Figure 4: The ASC Thresholds as a Function of

23

Model Calibration and Evaluation (3) Figure 5: The ASC Thresholds as a Function of the Running Time of the Wind Turbines

24

Model Calibration and Evaluation (4) Figure 6: The ASC Thresholds as a Function of the Learning-by-Doing Parameter ℎ

25

Model Calibration and Evaluation (5) Figure 7: The ASC Thresholds as a Function of the Standard Deviations and

26

Model Calibration and Evaluation (6)

27

Model Calibration and Evaluation (7) Figure 8 The ASC Thresholds of Declining Prices , and (in %)

28

Conclusion and Future Directions (1)

• This paper strived to contribute theoretically sound perspectives to

the offshore wind discussion. We scrutinized the incentives of the current EEG feed-in regulation for investing in offshore wind farms as well as an alternative tender-bidding scheme.

• The modeling approach provides several new insights into the

impact of the EEG 2012 feed-in tariffs on offshore wind farm investments in Germany and guides our thinking about appropriate policy design.

• Policy implications: first, a shift from interventionist policy

measures towards more market-based instruments is a particular

• priority. Second, the renewable offshore wind sector should be

29

subject to competition and the guaranteed prices should be phased

• ut.

30

Conclusion and Future Directions (2)

• We have neglected policy uncertainty.
• The two subsidy scenarios were treated as independent from one
• another. On closer inspection, however, it is clear that they are

intertwined and smart models should not treat them in isolation.

• Our partial-equilibrium modeling set-up also does not address issues

related to competitive interactions in the offshore wind industry.

• These findings can serve as guidance for academics, policymakers

and practitioners alike.