Electricity markets with a predominant share of renewable generation - - PowerPoint PPT Presentation

Dipl.-Ing. Niels Ehlers

Berlin University of Technology Department of Energy Systems www.ensys.tu-berlin.de

Electricity markets with a predominant share of renewable generation Can competition survive?

Dipl.-Ing. Niels Ehlers Department of Energy Systems Technical University of Berlin niels.ehlers@tu-berlin.de www.ensys.tu-berlin.de

Niels Ehlers- 2 -

Two roads diverged…(Robert Frost 1875) 1996

Directive 96/92/EC of the European Parliament: 1. enable competitive electricity markets (2) 2. priority may be given to the production of electricity from renewable sources (28)

www.epexspot.com www.n-pv.de

2010

Niels Ehlers- 3 -

4 52 36 10 9 20 40 60 80 100 120 20

41

73 32 49 100 200 300 400 500 600

German National Renewable Energy Action Plan

• Electricity

Source: Nationaler Aktionsplan für erneuerbare Energie gemäß der Richtlinie 2009/28/EG zur Förderung der Nutzung von Energie aus erneuerbaren Quellen

Demand

Power [ GW] Energy [ TWh]

Hydropower Geothermal Photovoltaics Wind Onshore Wind Offshore Biomass

Explicit promotion of specific technologies by means of feed-in tariffs

?

Niels Ehlers- 4 -

1000 2000 3000 4000 5000 6000 7000 8000 10 20 30 40 50 60 70 80 Hours Power [GW ]

Demand W ind Onshore W ind Offshore Biomass PV Residual load Export/Storage

Scenario 2030 - Residual load with growing shares of renewable generation

Lost sales of fossile power plants

Little change in peak demand! Powerplants need to be more flexible How can this capacity be financed?

Niels Ehlers- 5 -

• 200
• 100

100 200 300 400 500 600 700 800

• 5

5 10 15 20 09/12/23 09/12/24 09/12/25 09/12/26 09/12/27 09/12/28 09/12/29 09/12/30 Load / Infeed [GW] Windpower Real Windpower Prognosis Vertical Gridload (East Germany) EPEX Day Ahead Spotprice Day-Ahead Price [Euro/MWh]

The result: Grid situation Christmas 2009

Wind >20.000 MW !

Spot Price: Minus 200 €/MWh

Even baseload plants (Nuclear) must reduce

Dispatch of fossile powerplants (different scale)

Niels Ehlers- 6 -

Back to the basics: market fundamentals

• All markets are based on the trade of limited (constrained) goods or

resources.

• Each constrained resource can be assigned a shadow price of relaxing this

constraint (i.e. expand production) by one unit.

• The price should be made equal to marginal cost.
• When average costs are decreasing, marginal costs are less than average

costs, the total amount paid for the product will fall short of total costs

See for example: R. H. COASE 1946: The Marginal Cost Controversy

Niels Ehlers- 7 -

A market perspective on electricity I

Specifics of the electricity market:

• Different technologies with highly varying cost factors
• Short –run marginal costs (SRMC):

– consumption of primary energy (coal, gas, U3O8 ) or use of emission certificates – opportunity costs of (pumped-)storage power plants – near zero for most renewable generation units – negative for fossil units with binding technical constraints (must-run)

• Long –run marginal costs:

– capacity expansion (capital costs) / building new power plants – capacity maintenance and repair / maintaining existing power plants

Niels Ehlers- 8 -

A market perspective on electricity II

Theory:

• A cost-minimal and CO2-constrained system has a definite set of

shadow-prices for electricity and CO2 that leads to full cost-recovery of all market participants. (Holds only for the case of a linear cost function)

• Additional min/max capacity constraints for certain technologies raise

the total costs of the system and lead to shortfalls or windfall profits. Price components:

• Case 1 (Germany): No capacity markets:

– One price for short-run marginal costs – Long-term costs can only be recovered while the system is either

• not cost-minimal (expansion restrictions for nuclear power

plants, lack of new market entrants) or

• in the event of scarcity (Value of Lost Load Pricing)
• Case 2 (PJM): Capacity payments:

– Long-term costs are recovered on the capacity market

Niels Ehlers- 9 -

Our market model

• Linear Optimization Model

– Target Function: Minimize Costs!

• Construction Costs
• Fuel Costs
• Costs for load gradients (to represent startup costs)

– Subject to constraints:

• Load-Serving (Renewables allowed to be curtailed)
• CO2-Emission Cap
• Minimum uptime requirements (linear

representation)

• Simulation Period

10/2008 – 10/2009

• Input data for Germany

– Weather data – Wind/Solar/Temperature (very low wind activity during this period – worst case) – Real electricity demand data – Averaged commodity prices 2008-2010 – Annualized capital costs for different technologies

Niels Ehlers- 10 -

Scenario Results: The way to carbon-free generation

• Scenario including nuclear generation

50 100 150 200 250 300 300 200 100 50 25 Installed Capacity [GW]

CO2-Emissions per year [Mio.t/a]

Storage Peaking Unit Gas Turbine CCGT Coal Nuclear (max 20GW)

First step: Fuel switching from coal to gas Second step: Extension of offshore wind backed up by CCGT

Photovoltaic (min 30GW) Wind offshore Wind onshore (min 30GW)

Third step: Expansion of storage capacities

Niels Ehlers- 11 -

Price spike that covers the fix costs

1 10 100 1000 10000 1 8760 Price [€/MWh] Hours per year SRMC of fossile power plants Shadow Price of load constraint

Scenario Results: Price Duration Curves - BAU

300 Mio.t CO2 (28 billion €/a)

Price spike of shadow prices that covers the fixed costs

• nly generation costs

Niels Ehlers- 12 -

Scenario Results: Low Emission Scenario

25 Mio.t CO2 (34 billion €/a)

1 10 100 1000 10000 1 8760 Price [€/MWh] Hours per year SRMC of fossile power plants Shadow Price of load constraint

Pumped-Storage plants set the price  Opportunity costs

Niels Ehlers- 13 -

Scenario Results: Zero Emission Szenario

1 10 100 1000 10000 1 8760 Price [€/MWh] Hours per year SRMC of fossile power plants Shadow Price of load constraint

0 Mio.t CO2 (42 billion €/a)

Ramping Costs of nuclear generation units Including 20 GW nuclear generation Windfall Profits of nuclear generation if deprecated

Niels Ehlers- 14 -

Scenario Results: Market results 50 Mio.t CO2/year (including 20 GW nuclear)

Fixed costs covered by price spike: 775 Mio. Euro

In sum, revenues cover the total costs, but while nuclear power plants generate huge profits, fossil power plants and especially renewables cannot recover their total costs with market prices based on SRMC.

[Mio. Euro] Total Marginal Costs (Shadow prices) Short-Term Marginal Costs (Spot-Market) Explanation Onshore Wind 164,8 160,1 Losses Offshore Wind 0,0 85,4 Losses Photvoltaics 1.185,8 1.164,0 Losses Nuclear Power

• 5.089,7
• 4.484,6 Profits

CCGT 0,0 1.205,9 Losses Gas Turbine 0,0 293,0 Losses Peaking Unit 0,0 49,7 Losses Pumped-Hydro 0,0 359,8 Losses CO2 Market

• 3.239,9
• 3.239,9 Taxes

Customers 39.130,6 36.616,8 Sale revenues Market Results(Dual Solution) 32.151,6 32.210,0 Costs (Primal Solution) 32.151,6 32.151,6 System costs Difference 0,0

• 58,4

No opportunity costs of storage plants considered in spot prices!

Niels Ehlers- 15 -

Scenario Results: Market results 50 Mio.t CO2/year (nuclear phase out)

Fixed costs covered by price spike: 735 Mio. Euro

[Mio. Euro] Total Marginal Costs (Shadow prices) Short-Term Marginal Costs (Spot-Market) Explanation Onshore Wind 212,7 755,0 Losses Offshore Wind 0,0 2.557,0 Losses Photvoltaics 1.045,1 1.180,5 Losses Nuclear Power CCGT 0,0 1.121,0 Losses Gas Turbine 0,0 275,1 Losses Peaking Unit 0,0 66,8 Losses Pumped-Hydro 0,0 168,2 Losses CO2

• 4.289,5
• 4.289,5 Taxes

Revenues Load Serving 40.553,1 35.839,7 Sale revenues Sum (Dual) 37.521,4 37.673,8 Costs (Primal)

• 37.521,4
• 37.521,4 System costs

Difference 0,0 152,3

In this scenario, the revenues do not cover the total costs including CO2.

Niels Ehlers- 16 -

Conclusions I

• Many market participants in Germany can incur losses in the future due to two

effects: – Change of the merit-order by subsidized market entries

• Not a problem of market design!
• Just similar to changing demand.
• Danger of sunk costs and stationary higher system costs.

– Losses due to pure SRMC-pricing

• Inherent to the market design
• In the range of ~2-5% of the system costs
• Can be reduced by the use of peak load DSM with low fixed costs
• An enhanced operating reserves market might cover remaining costs
• Capacity payments might become necessary but less of the fact of wrong

market design but more because the pace of installation of renewable resources highly exceeds regular investment cycles of fossil generation.

Dipl.-Ing. Niels Ehlers

Berlin University of Technology Department of Energy Systems www.ensys.tu-berlin.de

Thank you for your attention!

DI Niels Ehlers Department of Energy Systems Technical University of Berlin niels.ehlers@tu-berlin.de www.ensys.tu-berlin.de

Niels Ehlers- 18 -

Additional conclusions from the model

• Non-market based deployment of renewable generation raises the

system costs:

– Onshore Wind: 2.000-5.000 €/MW/year – Photovoltaics: 50.000-100.000 €/MW/year without sunk costs of fossil power plants and costs for grid extension

• Curtailment of renewable resources is often preferable to storage.
• Pumped-storage power plants will become increasingly price-setting

and market power mitigation will be more difficult

• To avoid negative external effects, a more coordinated dispatch of

storage plants (one that is not only based on price forecasts) might become necessary

Niels Ehlers- 19 -

Rational Bids of Market Participants

• Rational bids at the day-ahead market:

– free market: = marginal costs of production – feed-in tariff (active) < minus feed-in tariff – feed-in tariff (passive) =

• infinity, only bounded by market

restrictions: EPEX:

• 3000 €/MWh

Regulator: -350 to -150 €/MWh (willing to reduce output): – feed-in tariff (passive):

Niels Ehlers- 20 -

Results: Shadow Prices for Load Serving

48,9 48,9 64,6 76,9 80,9 94,1 63,2 64,8 76,2 79,784,1 94,7 40 60 80 100 50 100 150 200 250 300 CO2-Emissions per year [Mio. t] With Nuclear Without Nuclear

Load- Shadow-Price [Euro/MWh]