Professional Documents
Culture Documents
2 March 2022
Context
Basics on flow-based market coupling
TSOs discretionary actions and their impact
Recent evolutions
Pending Issues
Conclusions
Abbreviations
Context
Day-ahead Flow Based Market Coupling in Central West Europe
(DA FBMC in CWE)
• 2007: MoU on the implementation of a CWE DA FBMC
• 2015 (8 years later!)
• Conditional approval of the CWE DA FBMC methodology, 15 action points
listed in a Common CWE NRA position paper
• Go-live of the CWE DA FBMC
• First results successful, followed by drop in CWE cross-border exchanges
• 2015 - 2017: Monitoring shows CWE DA FBMC performance below
expectations and reveals impact of TSO discretionary actions
• 2017: CREG study on the impact of CWE TSO discretionary actions on the
design and functioning of CWE DA FBMC (published in English on CREG
webpage)
• 2018: several measures leading to an significantly improved DA FBMC
Context
Contrary to what was expected during the parallel runs, a significant drop
in CWE cross-zonal exchange was observed after the introduction of FBMC.
5000
4535 4553
4389 4409
4500 MW -615 MW
4018
4000 3828 3794
3500
3000
2500
2000
1500
1000
500
0
2011 2012 2013 2014 2015 2016 2017
CWE-DA CWE DA + LT
5
Zonal Net Positions (MW) / Total CWE exchange (MW)
2000
4000
6000
8000
-8000
-6000
-4000
-2000
0
201101
201103
Context
201105
201107
201109
201111
201201
201203
201205
NEP NL
201207
201209
201211
201301
201303
NEP FR
201305
201307
201309
201311
201401
201403
201405
201407
NEP DE/AT/LU
201409
201411
201501
ATC
201503
201505
What happened three months after the go-live?
NEP BE
201507
5643
201509
201511
FBMC
201601
201603
201605
201607
XB-vol CWE
201609
Evolution of monthly average CWE cross-zonal exchange
201611
201701
201703
201705
201707
2781
201709
5093
201711
201801
201803
5359
6
Inefficient market coupling leads to higher price spreads – day ahead
2017
7
Inefficient market coupling leads to higher price spreads – 1 year ahead
2007-2017
8
Context
Basics on flow-based market coupling
TSOs discretionary actions and their impact
Recent evolutions
Pending Issues
Conclusions
FBMC concept
10
NEP NL
11
Active constraint
NEP NL
12
13
How is the flow based domain built?
Set of Critical Branches – Critical Outages (CBCO)
NEP A
For each CBCO: For each CBCO:
Physical loading resulting
from cross-zonal exchange
Sum(NEPz*PTDFz)
(in N-1 condition)
≤ ≤
NEP B Available capacity RAM
Next slides:
1. PTDFz= zonal Power Transfer Distribution Factors (MW/MW)
2. RAM = Remaining Available Margin (MW)
3. CBCO - selection
14
15
2. Remaining Available Margin (RAM):ideal situation
RAM
Fmax
! RAM = Fmax
16
RAM
Fmax
Fref’
17
BE HU
+ Production forecast
18
Origin of reference flows: Commercial exchanges in the « base case »
« Domestic
flows »
« Loop flows » CZ
NL DE/AT/LU
SL
Wind generation in BE HU
northern Germany to
demand in southern
Germany creates high level
of loop flows through FR
Belgium (average 900 MW)
Priority access
19
2. Remaining Available Margin (RAM): actual #1
RAM
Fmax
Domestic trade
Fref’
gets priority
access
20
RAM
To account for
forecast errors
Fmax
FRM & locational
uncertainty
Fref’
RAM
To account for
Remedial Actions
FAV (negative value)
Fmax and for grid
FRM security (positive
value)
Fref’
22
3. Selection of CBCOs
RAM
FAV
Fmax
FRM
Fref’
Overview
Construction of Flow Based domain
Sum(NEP*PTDF) < RAM for all CBCO’s in the CBCO-set
With PTDFz the zonal power transfer distribution factor
PTDFz = GSK PTDFn
And RAM the available capacity for cross-zonal exchange:
RAM = Fmax – Fref’ – FRM – FAV
24
Context
Basics on flow-based market coupling
TSOs discretionary actions and their impact
Recent evolutions
Pending Issues
Conclusions
Impact of TSOs collective discretionary actions
PRICE SPIKES LINKED TO IMPORTANT LOOP FLOWS
CURRENT CWE FB DESIGN does not limit loop flows and gives them priority grid
access. This reduces capacity for import and export.
26
Impact of TSOs collective discretionary actions
MORE DIFFICULT INTEGRATION OF RENEWABLES
EXAMPLE: SITUATION OF 1TH OF MAY 2017
>10000 MW 5000
4000
40
CURRENT CWE FB DESIGN, with the current configuration of bidding zones, does not
allow the maximisation of energy produced by renewables
27
28
Data 2015 - 2016
29
Data 2015 - 2016
30
Impact of TSO individual discretionary actions
EVENTS NOVEMBER 2016
Simulations performed by ELIA on CREG’s request
• Scenario 1: Application of winter limits (Fmax +20%) on Amprion CBCOs,
• Scenario 2: Removal of the internal lines added after go-live and removal of
positive FAV values on the four interconnection lines between Amprion and
Tennet NL area,
• Scenario 3: Use of PSTs on Belgian-Dutch border to keep the D-2 loop flows
expectations through Belgium below 500 MW, if possible.
CWE Day Ahead Net Position [MW] CWE Day Ahead Prices [€/MWh] CWE vol Max Price Spread
BE DE FR NL BE DE FR NL MW €/MWh
Ref -1231 3192 -2077 115 62,4 38,2 65,2 42,9 3694 29,3
Scenario 1 -1258 3862 -2794 190 61,9 39,2 62,6 43,6 4372 28,4
Scenario 2 -1269 4202 -3155 222 62,6 39,7 61,9 43,8 4697 29,1
Scenario 3 -1194 2867 -1686 14 62,6 37,8 66,2 42,4 3360 30,3
Scenario 2: Monthly averaged CWE cross-border trade would have been 4697
MW instead of 3694 MW (+1003 MW or +27%).
French import increases by 1078 MW and the French-German price spread
decreases by 4,8 €/MWh.
31
Amprion added internal critical branches just after the FBMC go-live. This single TSO
discretionary action has increased the average price spread between Belgium and
Germany over 2015 – 2016 by almost 2€/MWh.
Source: CWE TSOs, CWE Consultative Group meeting Brussels, February 28th, 2017
32
Key messages
33
Improvements
Several measures taken in 2018 leading to an
improved FBMC
• Increasing pressure on TSOs in CWE: common understanding of problematic situation
➢ High level meetings in 2017 with NRAs and EC (DG ENER) discussing solutions
1. Obligation to give minimum 20% of RAM on transmission lines
➢ End of April 2018 implementation; political pressure on TSOs in October 2018 due to non-
compliance of min20%RAM by German TSOs possibly leading to shortages in Belgium
2. Increase of Belgian import limit to 5500 MW (now 6500 MW, going to 7500 MW)
➢ In June 2018, Belgian TSO increases import limit from 4500 to 5500 MW
3. Zone split of DE/AT on 1 October 2018 following Acer decision
➢ All flows resulting from exchange from DE to AT are now put in competition with other
exchanges between countries in CWE => lower loop flows
35
2,000
In 2017, 46% of all active CBCOs
had less than 20% RAM.
1,500
Of these, 64% were Amprion
lines.
1,000
500
0
Distribution of active CBCOs
Elia Tennet DE TransnetBW Amprion RTE Tennet NL
by TSO & by RAM in 2017
<10% RAM 10 à 20% RAM 20 à 30% RAM 30 à 40% RAM +40 %RAM
Source: CWE TSOs, CREG
36
Fmax
RAM
Application
FAV of negative
FRM FAVs
Fref
RAM of active CBCOs in % of Fmax and the applied FAVs since the
implementation of the 20%minRAM.
Source: JAO, CREG postprocessing
37
38
39
40
Impact of measures
41
42
More efficient market coupling leads to lower price spreads – 1 year ahead
2007-2019
43
FBMC improved and extended
44
Re-dispatching « Domestic
flows »
= increase production in South
+ decrease production in North PL
=> “pushing back” domestic flow
CZ
NL DE/AT/LU
Re-dispatching decreases
domestic flow, but hence also SL
the loop flow.
If re-dispatching is structural, BE HU
this action can be foreseen =>
impact on flows can also be
« Loop
foreseen => should be taken
FR flows »
into account in DA capacity
calculation (“using the best Priority access
available forecast”)
45
Volume of Re-dispatching in Germany
• These data have been used to calculate the impact of re-dispatching on the
zonal price
• New CREG-study on “Best forecast of remedial actions to mitigate market
distortion”
46
• By removing the units re-dispatch down, we shift the German supply curve to the left
and increase the zonal clearing price
• Applied to the months of October 2018 till end February 2019, this gives an average of
6,15 €, and 7,8 € only for the hours with re-dispatching!
• It is assumed here, for simplicity reasons, that volumes of cross-zonal trade are not
affected by the price increase in Germany
47
CONCLUSION
Take home messages
• Several flaws in CWE day-ahead flow-based market coupling (DA FBMC)
design identified in the CWE NRAs decision in 2015, have not been resolved.
In 2015-2017, low RAM and high loop flows getting priority access to the
grid, leading to price spikes on the day-ahead market.
• Increased pressure on CWE TSOs in 2017 + several measures taken in 2018
leading to a significantly improved DA FBMC in CWE.
• Regulation 2019/943: 70% minRAM
• Improvements go slow, but are possible: try to understand the complexity
of power markets, analyze the data and keep the pressure on
• Don’t forget the TSO: this monopolist has a very important role in market
functioning and price formation => next battle on re-dispatching
49