OPTIAL CAPACITY INVESTMENTS AND FLEXIBILITY RESOURCES: AN INVESTMENT MODEL INTEGRATING THE SHORT-TERM REQUIREMENTS WITH THE LONG-RUN DYNAMICS PhD student, Manuel Villavicencio Chaire European Electricity Markets (CEEM) Université Paris-Dauphine Lulea 24/08/2016
Agenda 1 Introduction 2 The power system in context 3 Model presentation 4 Simulation results: optimal mix with increasing I-RES shares
1. Introduction DIFLEXO model: Dispatch, investments and flexibility optimization model A self-developed tool for market design benchmarking System cost and “multiple services” approach : investment and operational costs Hydrothermal optimization : when and how to use available hydro resources Operational constraints : Ramping limits, min/max capacities, part-load efficiencies, etc. Reliability issues : reserve requirements as a function of I-RES penetration (non-event control) LT ST RT Minutes (5 – 30 min) Source: Strbac. Imperial College London, 2012. 3
2. The power system in context The shock of increasing I-RES shares on electricity markets • Scissor effect Long-term • Capacity adequacy problems (years) • Energy security issues • Merit Order Effect Short-term • Less flexible dispatch but higher flexibility is needed (h) • Higher operational cost for committed units due to cycling • Missing money problem • Increased need for ancillary services Real time • Balancing issues (sec-min) • Need for an enhanced congestion management 4
2. The power system in context System dependent and interrelated issues • Scissor effect Long-term • Capacity adequacy problems (scissor effect) (years) • Energy security issues • Merit Order Effect Short-term • Less flexible dispatch but higher flexibility is needed (h) • Higher operational cost for committed units • Missing money problem • Increased need for ancillary services Real time • Balancing issues (sec-min) • Congestion management 5
2. The power system in context The cheapest power generation technologies might not deliver the greatest value to the system Flexibility for power and energy supply (short-term) • « Peak Shaving » (Intraday) • I-RES integration • Weekday/weekend arbitrage Multiple services, values and revenue sources Reliability and flexibility Capacity and for security supply flexibility adequacy (real time) (long-term) • Balancing and load following • Available capacity • Congestion management • Investment savings • System stability (generation and network) • Other ancillary system services 6
3. Model presentation: Research questions What electricity mix to minimize the total system cost and comply with operability and reliability requirements? 1. Do considering flexibility and reliability requirements matter while optimizing capacity investments? – What is their impact? – Do new flexibility resources have a role to play ? (e.i. electric energy storage ( EES ) and demand side management ( DSM) capabilities) – Are they complementary or in competition? 2. What is the real value of generation technology’s capacity (conventional, I - RES)? – Is that value dependent on the power system representation adopted? – How much flexibility is accounted from conventional units ? At what cost ? 7
3. Model presentation: Objective function System cost represented as Y : 𝒁 = 𝐽 𝑑𝑝𝑜 + 𝑃&𝑁 𝑑𝑝𝑜,𝑢 + 𝐺 𝑑𝑝𝑜,𝑢 + 𝐷𝑃2 𝑑𝑝𝑜,𝑢 +∆𝐻 𝑑𝑝𝑜,𝑢 S.T. operational min constraints : 𝑑𝑝𝑜 𝑑𝑝𝑜 𝑢 Ramping constraints + 𝐽 𝑆𝐹𝑇 + 𝑃&𝑁 𝑠𝑓𝑡,𝑢 + 𝑆𝐹𝐷 𝑠𝑓𝑡,𝑢 Min/max generation level ST 𝑠𝑓𝑡 𝑠𝑓𝑡 𝑢 Part load efficiencies Spinning and non- + 𝐽 𝑓𝑓𝑡 + 𝑃&𝑁 𝑓𝑓𝑡,𝑢 spinning reserve supply 𝑓𝑓𝑡 𝑓𝑓𝑡 𝑢 capabilities + 𝐽 𝑒𝑡𝑛 + 𝐸𝑇𝑁 𝑚𝑑,𝑢 + 𝐸𝑇𝑁 𝑚𝑡,𝑢 DSM and EES operation 𝑒𝑡𝑛 𝑚𝑑 𝑢 𝑚𝑡 𝑢 related constraints RT Clean energy policies… LT Set Element Description 𝑢 ∈ T T Time slice I ∈ I Generation technologies I I ⊇ Con 𝑑𝑝𝑜 ∈ I Conventional technologies I ⊇ RES 𝑠𝑓𝑡 ∈ I Renewable energy sources I ⊇ EES 𝑓𝑓𝑡 ∈ I Electric energy storage technologies DSM ⊇ LC 𝑚𝑑 ∈ DSM Demand side management able to supply load curtailment DSM ⊇ LS 𝑚𝑡 ∈ DSM Demand side management able to supply load shifting 8
3. Model presentation: The energy only market (EOM) Clearing the power market including new flexibility resources (i.e. DSM and EES ): 𝑢𝑢=𝑢+𝑀 𝑚𝑡 𝑒𝑑ℎ − 𝑇 𝑓𝑓𝑡,𝑢 𝑡𝑧𝑜𝑑 𝑣𝑞 base 1 + 𝜀 = 𝐻 𝑚𝑠𝑓𝑡,𝑢 − 𝐻 𝑑𝑣𝑠𝑓𝑡,𝑢 𝑑ℎ + 𝐸𝑇𝑁 𝑚𝑚𝑑,𝑢 𝑒𝑝 L t + 𝐻 𝑑𝑝𝑜,𝑢 + 𝑇 𝑓𝑓𝑡,𝑢 + 𝐸𝑇𝑁 𝑚𝑡,𝑢𝑢,𝑢 − 𝐸𝑇𝑁 𝑚𝑡,𝑢 ∀ 𝑢 𝑠𝑓𝑡 𝑑𝑝𝑜 𝑓𝑓𝑡 𝑚𝑑 𝑚𝑡 𝑢𝑢=𝑢−𝑀 𝑚𝑡 𝑚𝑡 Supply side flexibility Demand side flexibility EOM Variable Unit Description Synchronized power level of technology con 𝑡𝑧𝑜𝑑 𝐻 𝑑𝑝𝑜,𝑢 [GW] on time t 𝐻 𝑚 [GW] Power level of RES unit res 𝑠𝑓𝑡,𝑢 𝑑𝑣 𝐻 𝑠𝑓𝑡,𝑢 [GW] Power curtailed by res on hour t 𝑑ℎ 𝑇 𝑓𝑓𝑡,𝑢 [GW] Power demanded by storage unit ees on time t 𝑒𝑑ℎ 𝑇 𝑓𝑓𝑡,𝑢 [GW] Power supplied by storage unit ees on time t 𝐸𝑇𝑁 𝑚 [GW] DSM curtailment of load lc on time t 𝑚𝑑,𝑢 𝑣𝑞 𝐸𝑇𝑁 𝑚𝑡,𝑢 [GW] DSM shifting up ls on time t 𝑒𝑝 𝐸𝑇𝑁 𝑚𝑡,𝑢,𝑢𝑢 [GW] DSM shifting up ls on time tt from t 9
3. Model presentation: The balancing market Balancing markets: tackling variability and uncertainty of net load with the FRR Residual imbalance due to variability 𝑏𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 + 𝑇 𝑓𝑓𝑡,𝑢 𝑑ℎ,𝑏𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 + 𝑇 𝑓𝑓𝑡,𝑢 𝑏𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 𝑀 𝑢 𝑏𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 𝑄 𝑒𝑑ℎ,𝑏𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 𝑐𝑏𝑡𝑓 1 + 𝜀 + 𝜁 𝑠𝑓𝑡 𝜁 𝑚 𝑠𝑓𝑡 = 𝐻 𝑑𝑝𝑜,𝑢 𝑠𝑓𝑡 𝑑𝑝𝑜 𝑓𝑓𝑡 Residual forecast error due to uncertainty 𝑡𝑞 𝑜𝑡𝑞 𝑛𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑞 𝑛𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 𝑑ℎ,𝑛𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 + 𝑇 𝑓𝑓𝑡,𝑢 𝑛𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 𝑀 𝑢 𝑛𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 𝑄 𝑒𝑑ℎ,𝑛𝐺𝑆𝑆 𝑣𝑞/𝑒𝑝𝑥𝑜 𝑐𝑏𝑡𝑓 1 + 𝜀 + 𝜁 𝑠𝑓𝑡 𝜁 𝑚 𝑠𝑓𝑡 = 𝐻 𝑑𝑝𝑜,𝑢 + 𝐻 𝑑𝑝𝑜,𝑢 + 𝑇 𝑓𝑓𝑡,𝑢 𝑠𝑓𝑡 𝑑𝑝𝑜 𝑓𝑓𝑡 Total forecast error Forecast error covered by BRPs Residual forecast error Residual imbalance Source: NREL 2011, “Operating Reserves and Variable Generation” 10 “Operating Reserves and Variable Generation”
3. Model presentation: The balancing market Not-event secondary control driven by variability and uncertainty of RE generation Bidders: Synchronized units only + EES units aFRR down aFRR up Bidders: All spinning and non-spinning units + EES units mFRR down mFRR up 11
3. Model presentation: Solving algorithm LT ST Optimal investments Capacity adequacy level EOM Investments : set installed capacity (I-RES, Defining investments (capacity and conventionals, EES) and fixed cost cost) given operational constraints Defines annual net load Defining annual net load Defines reserve requirements Defining reserve requirements Optimal dispatch Optimal scheduling of available capacity Min Total Cost Capturing flexibility needs (generation, EES and DSM) regarding Assets competing to operational constraints and variable cost Optimal investments Optimal dispatch for unit supply multiple scheduling to minimize Optimal dispatch services at least cost operational cost Optimal reserve supply aFRR up aFRR down RT Optimal FRR allocation Verifying Verifying reliability compliance reliability given dispatched unit regulating compliance capabilities mFRR up mFRR down 12
4. Simulation: Valuating flexibility resources by the difference Experimental setup: optimal mix increasing I-RES shares progressively allowing for investments on flexibility resources (e.i. EES and dsm ) • Greenfield system without interconnections • Perimeter and dataset: Load (Lt) and I-RES capacity factors of France on 2013 • Hourly time slice and 8760 hours simulated • Considered portfolio of technologies: endogenous investment on – Power generation technologies: Nuclear, reservoir hydro, hard coal, lignite, CCGT, CT (high peak), wind and solar (including curtailment decisions). – Bulk storage technologies ( EES ) : PHS, CAES, VRFB, NaS, Li-ion – DSM : Load curtailment and load shifting less than 1% and 2% of Lt respectively – Other RES: Fatal hydro, Biomass and other RE accounted on the residual load. • Cost and parameters compiled from reports of DIW, Black and Veatch, IEA, EPRI, NREL and other scientific publications. Solved on GAMS under CPLEX 12.5 using the barrier algorithm 13
Recommend
More recommend
