CHALLENGES OF REPRESENTING ELECTRICITY SYSTEM FLEXIBILITY IN - PowerPoint PPT Presentation

CHALLENGES OF REPRESENTING ELECTRICITY SYSTEM FLEXIBILITY IN ENERGY SYSTEMS MODELS Vera Silva, EDF R&D Co-authors: Gregoire Prime, Timothee Hinchliffe, Dominique Lafond, François Rehulka, Miguel Lopez-Botet Petten, 04 December 2014

T HE PROBLEM OF FLEXIBILITY ADEQUACY  A system that has sufficient capacity to meet peak load is adequate but if this capacity is composed mostly of low flexible plants it can experience problems for handling demand and generation variability.  Flexibility has always been an essential ingredient to handle the variability and uncertainty in the demand-generation balance. It is required at the operational time scales but needs to be considered at planning stage.  Assessing the flexibility adequacy will probably emerge as a new task in power system planning and metrics and models are being developed to help with this task.  The representation of flexibility at the energy and power systems planning will help to deliver a system that can handle variability in a cost effective way. | 2

I NTEGRATION OF ELECTRICITY SYSTEM FLEXIBILITY TO ENERGY SYSTEMS MODELS Goal: to obtain an a long term electricity system expansion solution that ensures a flexible system by solving a problem that includes : 1) the interaction between the energy and the electricity systems 2) the long term uncertainties and 3) the relevant short term operation constraints . Long term forecasts of demand, Reliability and flexibility requirements commodity prices, etc MADONE Multi-energy system model Transmission expansion options and candidate generation technologies Electricity system model Generation and Transmission planning Production cost simulations and operation flexibility assessment Continental Model with FlexAssessment Adequate and flexible transmission and Investment loop | 3 generation expansion solution

D IFFERENT APPROACHES TO ADDRESS THE INTEGRATION OF ENERGY AND ELECTRICITY SYSTEMS PLANNING Option 1) Representation of electricity system flexibility in the Times model by increasing the simulation granularity and including additional constraints=> MADONE Option 2) Coupling energy system models with electricity system models using a chain of simulation tools with the possibility of back feeding relevant information • Energy system optimization : Madone • Electricity system planning : Continental Model with Investment loop • Detailed near term flexibility assessment : Continental with FlexAssessment | 4

M ADONE PERFORMS A MULTI - ANNUAL AND MULTI - ENERGY SIMULATION OF EACH OF THE 29 INTERCONNECTED E UROPEAN COUNTRIES  Perimeter EU27+NO+CH (Europe 29)  with different levels of detail depending on the country 275 Mtep / 549 TWh  Trans-national networks represented 1339 Mtep / 2114 TWh  electricity and gaz, CO2 258 Mtep / 580 TWh  Storage capacities  hydro (one lake per country + hydro-pumping), gaz, CO2  Pipelines/electricity injections at the frontier of EU29  NordStream, Southstream, Nabucco , DESERTEC… detailed interm. aggregated  National resource potentials & limits Capacités d’interconnexion en 2005 (enGW)  wind off-shore: km2(depth, wind speed, distance to coast) X Légende: TO / FROM Capacity density; Wind on-shore km2 ( area potentially available) X 2.05/ 1.65 Capacity density; solar PV: area available, roofs surfaces; CO2 2.3/ 2.2 storage; biomass resources 0.95/ 0.95 0.35/ 0.35 1.98/ 2.44  Period and simulation time step 0.75/ 0.75 0.6/ 0.1 1.3/ 1.5 0.6/ 0.6 2.05/ 1.5  yearly from 2005 to 2010, every 10 years from 2010 to 2050 0.3/ 0 3/ 3.85 1.75/ 0.8 1.1/ 1.2 2.5/ 0.8 2.4/ 2.4  representation of each year with load curves eg: 24, 288 points 0.5/ 0.5 2/ 2 1.9/ 0. 3.2/ 1.5 3.2/ 2.2 2/1. 9 1.5/ 0.8 2/ 1 1.9/ 2.75 8 1.2/0.47  Outputs 0.8/ 0.7 0.9/ 0.6 2.3/ 3.2 4.24/ 1.81 0.65/ 0.65 0.75/ 0.75 0.08/ 0.02 0.43/ 0.16  technology mix & detailed energy balances , energy dependency, 0.995/ 2.65 0.5/ 1.4 0.5/ 0.6 and environmental indicators, balance for electricity (including 0.5/ 0.5 1.2/ 1.3 exchanges), association of energy uses and activities…. 4 12/11/2005 EFESE/OSIRIS lot MADONE / Enerbat / EPI / MFEE/ EIFER | 5

C ONTINENTAL M ODEL WITH I NVESTMENT LOOP FOR MODELING THE E UROPEAN INTERCONNECTED ELECTRICITY SYSTEM The model simulates a “realistic” European electricity system, including:  description of different countries generation mix and key transmission corridors  interconnection capacities between countries  management of water reservoirs and pump storage  a large number of scenarios of climate years represented by demand and variable generation across the European system => time-synchronise data with hourly (or lower) resolution  several scenarios of generation availability Some key challenges of this problem:  hydro and storage flexibility optimization => stochastic problem  generation scheduling needs to be performed across the whole Europe including interconnection and key transmission constraints => high performance computing  analysis of system static and dynamic security => hierarchical approach | 6

C ONTINENTAL M ODEL WITH I NVESTMENT LOOP : ELECTRICITY GENERATION INVESTMENT MODEL FOR INTERCONNECTED SYSTEMS The objective is to obtain the thermal generation INPUT DATA mix that ensures that for every new unit the Demand Storage revenues equals its annuitized fixed costs : Variable generation Investment costs profiles Commodity prices  Fixed costs include investment and O&M Interconnection CO2 price constraints  Variable costs include start-up and fuel costs Investment loop The conventional generation mix is optimized in two iterative steps:  Load duration curve based heuristic to propose a candidate solution CONTINENTAL  Validation of the heuristic solution solving the hourly load-generation dispatch => creates a price OUTPUT signal that feeds the investment loop Optimal thermal Market clearing prices generation mix CO2 emissions The generation mix needs to respect an adequacy Production dispatch Hydro stock level paths criterion Production costs Interconnection uses  Maximum of 3h/year with marginal price = VOLL | 7

C ONTINENTAL MODEL HYDRO - THERMAL GENERATION SCHEDULING Scenario based representation of stochastic parameters : Large number of annual scenarios of demand, wind and PV generation, water inflows, fuel costs, thermal unit availabilities Scenarios Water data values Minimize global production cost for each Stochastic hydro-generation zone scheduling Unit commitment and economic dispatch minimizes thermal and Maximize the reduction in terms of generation costs hydro generation cost over all the scenarios using dynamic optimization to obtain the « water value » for each time step Constraints include primary, secondary and tertiary reserve and generation dynamic ratings Define a set of strategies of the optimal use of hydro reservoirs in order to minimize the global Multi area optimization with interconnection constraints generation cost represented by NTC For each dispatch period and for each zone the dispatch solution and the market clearing prices are obtained to access the revenues of generation units Reference: Langrene, N., van Ackooij, W., Breant, F., « Dynamic Constraints for Aggregated Units: Formulation and Application », Power Systems, IEEE Transactions on , vol.26, no.3, Aug. 2011 | 8

O PTION 1 - REPRESENTATION OF FLEXIBILITY IN THE ENERGY MODEL : C OMPARISON MADONE - CONTINENTAL MADONE Continental Model with Investment loop  Bottom-up TIMES model : all 29 interconnected  EDF R&D’s Elec Production Cost model European countries  Electricity generation portfolio optimization  Renewables national resources potentials & limits detailed: Wind off-shore, wind on-shore, roof  Stochastic simulation of hourly system for PV etc… operation  Horizon = 2005 to 2050 with a perfect foresight  Demand-generation balancing solved for one of each year year with hourly resolution  Time-slices: 24 or 288  Interconnection constraints included  24 =Peak and Off-Peak for each month  Stochastic parameters: (T°, hydro, wind, PV and  288 = 2 representative day (Week/W-E, bi- generation outages) hourly)/month  Peak equation: additional demand constraint  Dem ( t ) Dem ( t ) max  With or without renewable contribution  Deterministic  Or multi-scenarios for one chosen year : testing with 4 annual scenarios | 9

C OMPARISON OF CONTINENTAL AND MADONE OPTIMAL THERMAL GENERATION MIX CONTINENTAL Capacité installée globale : MAD Global (GW) Base Mid-merit Peak - Consistency of REF_Def 20k € _Seuil 5k € _200MW base capacity needs 91 63 96 between the 2 Def 20k € _Seuil 10k € _200MW 91 70 88 models Def 20k € _Seuil 10k € _600MW 90 68 92 Def 20k € _Seuil 20k € _200MW 91 71 86 - Mid-base capacity Def 5k € _Seuil 5k € _MC2all 90 68 90 underestimated Def 10k € _Seuil 5k € _MC1pt with TIMES model 90 40 125 Def 50k € _Seuil 5k € _200MW 89 65 106 - Peak capacity Def 10k € _Seuil 5k € _200MW 90 63 90 dependant on peak Def 5k € _Seuil 5k € _200MW 89 67 77 equation MADONE (TIMES) 0 50 100 150 200 250 calibration Base Mid-merit Peak 24p 81 40 10 24p + peak 81 43 96 24p + peak ss EnR 81 42 224 288p 89 41 34 288p + peak 89 41 101 288p + peak ss EnR 89 43 234 - 50 100 150 200 250 300 350 | 10