SEM, BETTA and Beyond Approach to Interconnector Modelling Energy Exemplar EU Conference Lisbon June 17 th 2013 Marie-Therese Campbell Dr. Eoin Clifford
Background Interconnector Modelling Difficult to master! Do modelled flows reflect reality? Trading Behaviour Price convergence, arbitrage, risk premiums Wheeling charges
Introduction Moyle 250 import/export Half of original capacity EWIC 500 import/export Fully operational since May Future IC to France? Renewable Export projects to UK
SEM & BETTA SEM BETTA Ireland & Northern Ireland Transparent bidding Short Run Marginal Cost, Start- up costs Capacity Payments 3 gate closures Ex-post Pricing England, Scotland, & Wales Non-transparent power trades Bilateral contracts Power Exchange and Balancing Market Hourly gate closures Ex-ante Pricing
Modelling challenges Representing BETTA Game Theory or Perfect Competition? APX as reference for prices? Modelling bilateral contracts? Lack of transparency SEM & BETTA BETTA = 10 x SEM 6 GW vs 60 GW Will this affect optimisation? We only care about IC flows!
24/10/2008 24/12/2008 24/02/2009 24/04/2009 24/06/2009 24/08/2009 24/10/2009 24/12/2009 24/02/2010 24/04/2010 24/06/2010 24/08/2010 24/10/2010 24/12/2010 24/02/2011 24/04/2011 24/06/2011 24/08/2011 24/10/2011 24/12/2011 24/02/2012 24/04/2012 24/06/2012 24/08/2012 24/10/2012 24/12/2012 Clean Spark Spreads 30 Clean Spark Spreads UK 25 20 15 10 Price GB MWh 5 0-5 Source: Thomson Reuters
Wheeling Charges Imports SP + uplift + CPM IC charges σ > GBP Exports SP + uplift + CPM + IC charges + σ < GBP Components SEM shadow price (SP) GB Price (GBP) SEM capacity payments (CPM) Uplift Risk (σ) IC charges
Wheeling Charges on Interconnectors 50 40 Wheeling Charges 30 20 10 0-10 -20-30 -40 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Wheeling charge forward Wheeling charge with 5 risk premium Wheeling charge with 10 risk premium Wheeling charge back Wheeling charge back with 5 risk premium Wheeling charge back with 10 risk premium Hour of Day
Imports (GWh) Carbon Price Floor Study IC Flows 11000 9000 SEM imports - No CPF SEM exports - No CPF SEM imports - with CPF SEM exports - with CPF 7000 5000 3000 1000-1000 -3000 2013 2014 2015 2016 2017 2018 2019 2020
Moyle flow (MW) Plexos Study Moyle Flows 250 Exports are positive flows 200 150 100 50 0-50 -100-150 -200-250 -300 Date
EWIC flow (MW) Plexos Study EWIC Flows 600 Exports are positive flows 400 200 0-200 -400-600 Date
Reality - Moyle
Reality - Moyle Moyle Always importing Flows often not price reflective SEM committee study It is clear that the correlation between the price differential and flows across Moyle was not close
Other Studies McInerney & Bunn Examined transmission rights and flows on Moyle SEM committee study Significant power flows against the efficient price spread direction
Reality - Moyle Prices Why are flows not price dependent? Congestion issue? Cheap gen in Scotland Flows between Scotland and England limited Will be fixed over next few years Wouldn t effect EWIC
Imports to SEM (MW) Reality - EWIC Same story Full imports all day Imports of 300 MW during night Dip at 6 a.m. Limited experience Fully operational for 1 month only Half capacity for 5 months 600 500 400 300 200 100 0 EWIC flows May 2013 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Hour of day
Price ( /MWh) Prices May averages 100 90 80 70 60 50 40 30 20 APX SEM Shadow Price 10 0 0 2 4 6 8 10 12 14 16 18 20 22 Hour of Day
Price ( /MWh) Prices May averages 100 90 80 70 60 50 40 30 20 10 0 0 2 4 6 8 10 12 14 16 18 20 22 Hour of Day APX SEM Shadow Price +Uplift +CPM
Reality EWIC Trader bids Traders bidding in different prices Not reflective of APX prices or gen costs Why? Can they not access APX prices? Ex-post pricing creates too much risk? Hedging strategies? Avoiding shutdowns?
System Operation High imports Reserve issues Curtailment of renewables Countertrading Countertrading agreements under discussion with NatGrid UK Third party now being used Accuracy of wind forecasting still an issue
Plexos modelling Market Flows Generate profile based on historic data Fix flows to profile using Line.Fixed Flow Countertrading Use constraint to fix flows Penalty Quantity represents countertrading volume Penalty Price gives associated cost
Discussion Modelling trader behaviour? Replace interconnector up into series of lines Each line represents trader Dummy generator with price based on trader behaviour As yet untested Future trading Will trader behaviour change in the future? How will carbon price in GB affect things? Trading with France? Target model? All suggestions welcome!