Pr

1
EWEC 2009, Marseille
Strategies for Wind Power Trading in Sequential
ShortTerm Electricity Markets
Franck Bourry and George Kariniotakis Center for
Energy and Processes
2
Wind Power in Europe

• Installed capacity

EWEAs 2007 scenario for wind energy
installations up to 2030
Source EWEA

• Challenges

• Economic and secure operation of a power system
• Competitiveness of wind energy in a liberalized
  electricity market

3
Objective

• To develop a methodology for the optimal
  participation of independent wind power producers
  in dayahead and intraday electricity markets
• Aim reduction of imbalance penalties resulting
  from low predictability of wind generation

4
Hypothesis Balance Responsible Wind Power
Producer

• Selling bids A wind power producer is assumed
  to be an energy producer participating in the
  electricity market with only selling bids.

5
Dayahead and intraday market description

• Dayahead market
• Submission of bids the day before the energy
  delivery
• Power exchange sessions

• Intraday market
• For enabling encouraging selfbalancing of
  market parties (adjustment market )
• Different mechanisms such as power exchange
  sessions, power exchange continuous trade or Over
  The Counter (OTC)

Examples
OMEL, Spain Elbas NordPool, Nordic Countries
6 power exchange sessions with the gate closure time 2h15 before the start of the energy delivery period. Continuous trading till one hour before delivery.
6
Sequential bids in dayahead and intraday markets
D1
D2
D



Day-ahead market participation




Intraday market participation





Example of a combined participation in the Elspot
and Elbas markets (NordPool). Bids in the Elbas
market are proposed 6 hours before the delivery
time.
7
Model of the participation in the dayahead market

• Dayahead bid Eb0,?b0
• Price taker hypothesis price independent bid,
  at zero price
• Bid quantity based on the available wind power
  forecast.

• Market settlement
• Unique market price resulting from meeting
  aggregated purchase/sale curves (marginal
  pricing)
• Accepted energy quantity dependent on the bid
  price all price taker bids are accepted.

Bid (price-taker) Market settlement ? Contract
Quantity Eb0 Ppred ?t Market settlement ? Quantity Ec0 Eb0
Price ?b0 0 Market settlement ? Price ?c0 ?c0 market
8
Model of the participation in the intraday market

• Intraday bid Eb1,?b1
• Eb1,?b1 are the results of a decision making
  method which aims to reduce imbalance penalties.

• Market settlement (for continuous trade market)
• Trade occurs when the selling and buying bids
  match
• The contract price is the bid price (pay as bid
  pricing)
• The contracted energy quantity depends on the
  buying bids of the other participants a
  proportion of traded energy over the bid quantity

Bid Market settlement ? Contract
Quantity Eb1 Market settlement ? Quantity Ec1 a Eb1
Price ?b1 Market settlement ? Price ?c1 ?b1
9
Model of the participation in the intraday market

• Intraday market price distribution model
• Triangular distribution model
• Estimated from min(?c1), mean(?c1), max(?c1)

• Modeling of a
• Ec1 a Eb1
• a probability(?b1 ?c1 ) 1
  probability(?c1 lt ?b1) 1 F(?b1)

Example Elbas (NordPool ) 11/04/2004 at 11.00 AM
(prices in DKK/MWh)
F
f
a 0.91
?b1
?b1
?b1 181.2
mean 184.35
10
Imbalance penalty model (1/2)
?c0 Dayahead contract price Ec0 Dayahead
contract energy ?c1 Intraday contract price Ec1
Intraday contract energy ?/-
Positive/Negative Imbalance price E Delivered
energy

• Dayahead market participation

Contract Income Ec0 ?c0 E ?c0

Regulation costs (Ec0 E) ?/- d (E )
Revenue
- - -


Production Income
Imbalance Penalty

• Dayahead and intraday market participation

Ec1 (?c0 ?c1) d(E Ec1)
-
Revenue
E ?c0

d(E) Ec1 (?c0 ?c1 ) d (E Ec1))
11
Imbalance penalty model (2/2)

• Imbalance price
• Proposed model (valid in the NordPool system for
  example)

System upregulation
System downregulation
?lt?c0
?- ?c0
?- gt?c0
??c0
12
General Formulation of the Intraday Bidding
Problem
Model for participating in the intraday market
for the hour hi
Eb1, ?b1 argmin( d(?b1 , Eb1) )
hi
hi
Eb1, ?b1
Eb1, ?b1 argmin( (?b1 ?c0 ) ahi Eb1
d hi (E hi ahi Eb1) )
hi
hi
hi
Eb1, ?b1

• ?c0 is the dayahead market price for the hour
  hi
• E hi is the estimated energy delivery for hour
  hi
• d hi is the estimated imbalance penalty function
  for hour hi (? hi,?- hi)
• Ec1 a hi Eb1, where a hi is the estimated
  proportion of traded energy over the bid quantity
  for hour hi

hi
13
Proposed intraday bidding approach
d(Eb1, ?b1) (?b1 ?c0 ) a Eb1 d (E a
Eb1)
System downregulation
System upregulation
Eb1
Eb1
d
With a
3
E Ec0
1
0
?
?c0
?
?b1
?
?b1
?
? ?c0
?c0 ?
14
Simulation methodology
Dayaheadbid
Up/down regulation price
Market price
B.
WP forecasts
Up/down regulation price forecast
M. S.
WP measures
Dayaheadcontract
Balance settlement
Intradaybid
B.
M. S.
Intradaycontract
Delivered energy
Imbalance penalties
B.
M. S.
Bid
Market Settlement
15
Case Study

• Test case based on real world data
• 18 MW wind farm in Denmark
• Wind power forecasts from a statistical model,
  based on power curve modeling
• Participation in the Nord Pool dayahead
  (Elspot) and intraday (Elbas) markets from
  January to March 2004 (3 months)
• Penalties associated to the balancing mechanism
  are applied.

16
Results
?b1
?b1 ? ß (?c0 ?), with 0.2 ß
1.2
17
Results
?b1
Ec1
?b1 ? ß (?c0 ?), with 0.2 ß
1.2 Ec1 a Eb1,
with a 1 F(?b1).

(E Ec0) if E Ec0 0 else.
Eb1
18
Results
-18
19
Conclusions

• A model for the settlement of continuous trading
  market is proposed. This model is based on the
  available data of market prices
• The participation in an intraday market is
  formulated as a hedging method which aims to
  reduce the imbalance penalties.
• The present case study shows that the
  participation in the intraday market can reduce
  the imbalance penalties by up to 18 .

20

• Thank you for your attention

This work was performed in the frame of the
Anemos.Plus project (FP 6).
ARMINES participates in
20