Co st Re duc tio n Po te ntia ls in Co st Re duc tio n Po te ntia ls in the Ge rma n Ma rke t fo r Ba la nc ing Po we r K K ai F ai F linke rbusc h (MSc E linke rbusc h (MSc E c ) Dr Mic hae l He ute rke s c ), Dr. Mic hae l He ute rke s Unive rsity o f Münste r
Supply = Transmission system operator Demand Demand (TSO) (TSO) • Provides ancillary services • Monopsony Three types of balancing power Consumers Un ‐ • Primary balancing power Balancing Balancing + anticipated power power • Secondary balancing power Secondary balancing power Events Events Producers • Tertiary balancing power Control areas • Four areas: EnBW, e.on, RWE, Vattenfall Supply • Each area is controlled ≠ separately separately Demand Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 2
Pro c ure me nt o f balanc ing po we r 1. T SO spe c ifie s de ma nd 2. T SO buys quantitie s by auc tio n 3. Supplie rs submit bids (de mand rate , e ne rg y rate ) 4. T SO o rde rs bids by de mand rate ( sc o ring rule ) Demand Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 3
Ac tiva tio n o f ba la nc ing po we r 1. Co ntro l are a imbalanc e o c c urs SO o rde rs supplie rs by e ne rg y rate ( se ttle me nt rule ) 2. T Imbalance Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 4
Co st re duc tio n po te ntia ls 1. E ffe c t: Ne tting antipodal use of balanc ing powe r 2. E ffe c t: Mor e e ffic ie nt pr oc ur e me nt auc tions 3. E ffe c t: L e ss pr ovision of balanc ing powe r Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 5
Da ta Obse rve d Pe rio d • 12 mo nths – De c 2007 – No v 2008 – So urc e • We bsite s o f T SOs – T wo distinc t datase ts • 1. Auc tio n data (bids fro m pro c ure me nt auc tio ns) ( p ) 2. Ac tivatio n data (po we r flo ws) Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 6
1. Auc tio n Da ta Pr imar y Balanc ing Powe r Se c ondary Balanc ing Powe r Bids are valid fo r1 mo nth • Po sitive / ne g ative g • Pe ak- / o ffpe ak • T e r tiar y Balanc ing Powe r y g Bids are valid fo r4 ho urs • Po sitive / ne g ative • Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 7
Se c o nda ry ba la nc ing po we r: Bid size s Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 8
De ma nd ra te s Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 9
E ne rg y ra te s Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 10
2. Ac tiva tio n Da ta Ac tual po we r flo ws Data fre que nc y Quarte ro f an ho ur • 35,136 o bse rva tio ns • Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 11
Se c o ndary balanc ing po we r [MW] E E nBW nBW Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 12
e on e .on Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 13
RWE RWE Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 14
Vatte nfall Vatte nfall Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 15
T e rtia ry ba la nc ing po we r [MW] RWE RWE Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 16
Market design is known • Scoring Rule • Settlement Rule S ttl t R l • Control Areas Data is available • Auction Data • Activation Data A ti ti D t Simulation of the market for balancing power with GAMS • Scenario 1: Status quo • Scenario 2: One united control area • Scenario 2: One united control area Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 17
T e sting the mo de l T e c hnic al re stric tio ns • Data fre que nc y • Mo de l spe c ific atio n • Ave ra g e we ig hte d e ne rg y ra te s: AWER ENBW AWER EON AWER RWE AWER Vattenfall DATA SIM DATA SIM DATA SIM DATA SIM Mean 60.10 58.24 36.74 36.43 52.40 49.09 53.02 47.19 Median 61.00 61.00 0.00 0.00 2.00 2.00 3.00 2.75 Correlation 0.98 0.99 0.97 0.97 Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 18
Scenario 1: "status quo" 120 80 80 Activation tertiary Procurement tertiary M € Activation secondary Activation secondary Procurement secondary 40 Primary 0 0 December ‐ 07 November ‐ 08 Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 19
Scenario 2: United control area 120 80 80 Activation tertiary Procurement tertiary M € Activation secondary Activation secondary Procurement secondary 40 Primary 0 0 December ‐ 07 November ‐ 08 Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 20
Co mpa riso n o f a bso lute le ve ls Primary Primary Secondary Secondary Tertiary Tertiary Procurement Activation Procurement Activation Sum Scenario1 (M€) ( ) 114.8 230.4 351.3 227.5 30.7 954.8 Scenario2 (M€) 114.8 229.4 190.6 227.3 29.8 792.1 0.00% ‐ 0.43% ‐ 45.72% ‐ 0.07% ‐ 3.06% ‐ 17.04% difference Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 21
Co mpariso n o f share s Scenario 1 Scenario 2 3% 12% 4% 14% Primary 24% P Procurement secondary t d 29% Activation secondary 24% Procurement tertiary 29% 29% Activation tertiary 24% % 37% 37% Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 22
Conclusion Conclusion •When sufficient grid capacity is assumed, antipodal use of balancing power is inefficient balancing power is inefficient •Netting area imbalances major cost reductions (160 M€) •This is mostly on account of secondary balancing power Further research Further research • Co ‐ operation of e.on, EnBW and Vattenfall • Efficiency of scoring rule and settlement rule c e cy o sco g u e a d sett e e t u e Data The Model h d l Results l Conclusion l Motivation Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 23
Co ntac t: K ai F linke rbusc h L e hrstuhl für Vo lkswirtsc haftsthe o rie We stfälisc he Wilhe lms-Unive rsität Münste r We stfälisc he Wilhe lms Unive rsität Münste r K a i.F linke rbusc h@ wiwi.uni-mue nste r.de Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 24
Ba c kup Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 25
Mo de l se que nc e Step 1: Procurement min! min! • Scoring rule: PC dr m b , t , c b , t , c b t c • s.t. m m b , t , c t , c b Step 2: Activation min! • Settlement rule: AC er x b , t , c b , t , c b t c • s.t. x m b , t , c b , t , c c c • and x CAI b b , , t t , , c c t t , , c c b b Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 26
AWER ENBW AWER EON AWER RWE AWER VET DATA SIM DATA SIM DATA SIM DATA SIM Mean 60.101 58.240 36.749 36.434 52.404 49.094 53.027 47.192 Median 61.000 61.000 0.000 0.000 2.000 2.000 3.000 2.750 Maximum 475.000 289.370 322.000 282.060 599.000 331.760 501.000 262.190 Minimum 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Correlation 0.984 0.993 0.977 0.975 Obs Obs. 35136 35136 35136 35136 35136 35136 35136 35136 35136 35136 35136 35136 35136 35136 35136 35136 Table 3: Descriptive statistics of average weighted energy rates Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 27
Co -o pe ratio n o f E nBW, e .o n and Vatte nfall De c e mbe r2008 Re duc tio n o f antipo dal use – May 2009 Re duc tio n o f pro c ure me nt – June 2009 Pre qualific atio n fo runite d c o ntro l are a – Oc to be r2009 Jo int me rit o rde r – Oc to b e r10, 2009 K a i F linke rb usc h, Dr. Mic ha e l He ute rke s 28
Recommend
More recommend
