Energy Analysis and Modelling: Power Sector Modelling with TIMES

Transcription

1 Energy Analysis and Modelling: Power Sector Modelling with TIMES Uwe Remme (IEA) Steve Pye (UCL) Energy Training Week April 2013

2 Module overview 11 April: Power Sector Modelling with TIMES (1) Introduction to energy systems modelling and TIMES Getting started in running a power sector model in TIMES 12 April: Power Sector Modelling with TIMES (2) Refining the analysis: Load curves and storage Using TIMES models to inform power sector strategy

3 Introduction to Energy Systems Modelling and TIMES Uwe Remme Energy Training Week April 2013

4 Interdependencies in the energy system Global energy flows in 2009

5 Electricity generation mix in 2010 in participants countries 2% 6% 2% 3% 7% 3% 1% 20% 7% 2% 2% 17% 1% 1% 7% 1% 20% 9% 1% 21% 98% 78% 72% 78% 71% 69% 4% 5% 7% 11% 26% 28% 34% 4% 27% 23% 3% 35% 16% 16% 93% 16% 1% 50% 9% 6% 3% 23% 46% 14% 1% 44% 54% 12% 12% 2% 3% 3% 5% 1% 94% 68% 10% 6% 40% 24% 20% 9% 6% 3% 23% 1% 99% 14% 1% 44% 11% 24% 100% 64% 7% 10% 1% 8% 1% 37% 47% 8% 80% 29% 21% 4% Coal Oil Natural gas Nuclear Hydro Wind Other renewables 46% Can you spot your country?

6 Electricity generation mix in 2010 in participants countries 2% 6% 2% 3% 7% 3% 1% 20% 7% 2% 2% 17% 1% 1% 7% 1% 20% 9% 1% 21% 72% 78% 71% 98% 78% 69% Algeria Brazil Cameroon China Colombia Egypt 4% 5% 7% 9% 12% 2% 11% 26% 6% 10% 6% 11% 24% 3% 3% 44% 40% 12% 3% 23% 24% 3% 28% 68% 23% 93% 14% 1% 20% 100% 64% 9% EU Georgia Germany India Indonesia 6% Mozambique Nigeria 1% 34% 4% 27% 35% 16% 16% 16% 1% 50% 46% 54% 5% 94% 1% 99% 47% 7% 10% 1% 8% 1% Pakistan Russia Saudi Arabia South Africa UAE Ukraine Uruguay 3% 23% 14% 1% 44% 8% 37% 80% 29% 21% 4% 46% Vietnam Coal Oil Natural gas Nuclear Hydro Wind Other renewables Current situation and future opportunities and challenges are quite diverse around the world.

7 Why energy planning? Integrated analysis is key for energy sustainable development (including linkages to land and water use) Economic way of covering energy needs Efficient use of domestic resources Enhancing energy security Access to energy Reducing environmental impacts Essential for sound decision-making and policy design Developing long-term strategies (avoiding stop-gap policies) Exploring and testing policy options Identifying the investment and financing requirements

8 Why energy modelling? Technology/Energy carrier level Long-term nature of planning decisions (e.g. lifetime of power plants) Future development of technologies Comparative assessment of technologies System level Infrastructure for energy (e.g. T&D for electricity) Interdependencies of technologies and sectors (e.g. electric heat pump or EVs with power sector) Integration of variable renewables Stakeholder level Tool to develop strategies involving wide range of actors in energy sector (from households industry to government) Communication Policy level How to reach policy goals Effectiveness and impacts of individual policy measures Energy system does not allow for real-world experiments Scenarios: Exploring the future Models: Developing consistent scenarios

9 Type of energy models Energy models Energy system models Economic models Focus on the energy system alone Description of large number of single energy technologies Capturing dynamics of the system such as substitution of energy carriers energy savings or technology deployment Typical results: energy consumption energy technology mix and capacities emissions Relationship of the energy sector to the general economy central element Energy system described in a highly aggregated way Technologies embedded in production functions Typical results: energy consumption emissions GDP employment trade balance

10 Energy prices Resource availability Energy system models: Technology-rich representation Cost and emissions balance Domestic sources Imports Coal Flows Coal processing Refineries Process Power plants and Transportation CHP plants 2.2 kwh and district heat networks 0.27 kg SKE Gas network Technology Supercrititcal coal plant () Flows 1 kwh Industry Commodities Commercial and tertiary sector Electricity Households kg CO 2 Transportation GDP Process energy Heating area Population Light Communication Power Passenger kilometers Freight kilometers Demands Primary energy Final energy Demand services

11 Reference energy system Primary energy supply Lignite resources Oil resources Oil import Coal resources Coal import Gas resources Gas import Residual wood Area PV Solarthermal Energy crops Conversion sector Coal processing Refinery Gas processing Elec sector Coal cond. PP Lignite cond. PP Coal IGCC PP Gas CC PP Wind converter CHP sector Coal CHP Coal CHP Gas CC CHP Biomass CC CHP End-use sectors Residential Gas heating Oil heating Elc. heating Local heat grid Gas water boiler Elc. water boiler Elc. heat pump Commercial Room heat boilers Process heat boilers Warm water boilers Industry Industrial boilers Transport Gasoil car Area Wind Electrolysis Gasification Liquefication Gasoline car LH2 car Busses Trucks

12 Technology representation Commodities Process Coal PC Supercrit. Unit Size MW el Coal Electricity CO 2 Construction time Years Lifetime Years Flows Flows Efficiency (LHV) % kwh 0.27 kg SKE Supercrititcal coal plant () 1 kwh kg Max. availability h/a Spec. Investment costs (overnight) /kw el Fixed O&M /(kw a) Var. O&M /MWh el Efficiency eqn SPC COAL Emission eqn Activity definition Utilization eqn COAL CO 2 COAL CO 2 ACT ACT CAP Input parameter ε SPC COALCO2 Plant efficiency CO2 Emission factor Annual availability

13 Horizontal dimension of the RES: technology chains Appliances Coal imports Domestic mining Coal transport Supercritical coal plant Backward Loops Grid HV & Transf. Grid MV & Transf. Lighting Urban trains Industry SPC COL SPC COL COL COL CO 2 COL COL ACT ACT ACT CO 2 SPC COL COL COL CO 2 SPC COL COL COL CO 2 SPC COL COL COL CO 2 SPC COL COL COL CO 2 SPC COL SPC COL COL COL CO 2 SPC COL COL COL CO 2 SPC COL ACT COL COL CO 2 ACT ACT ACT ACT ACT ACT COL CAP COL CO 2 ACT CAP ACT CAP ACT ACT CAP ACT CAP ACT CAP ACT CAP CAP ACT CAP DEM DEM DEM DEM DEM ACT CAP ACT

14 Vertical dimension of the RES: competition Appliances Domestic mining Coal imports Coal transport Supercritical coal plant (exst) Ultrasupercritical coal plant (new) Natural gas GT (exist) Grid HV & Transf. Grid MV & Transf. Lighting Urban trains Industry Natural gas CC (new) Supercritical lignite plant (exst) Supercritical lignite plant (new) Competing options to produce electricity: between technologies between old and new plants Influenced by: Technology costs Efficiencies Emission factors Fuel prices

15 System aspects: Limited resources Primary energy supply Lignite resources Conversion sector Coal processing Refinery End-use sectors Residential Gas heating Biomass heating Elc. heating Local heat grid Gas processing Oil resources Oil import Elec sector Coal cond. PP Coal resources Lignite cond. PP Coal import Biomass IGCC PP Gas CC PP Gas resources Wind converter Gas import CHP sector Coal CHP Residual wood Coal CHP Gas water boiler Elc. water boiler Elc. heat pump Commercial Room heat boilers Process heat boilers Warm water boilers Industry Industrial boilers Area PV Solarthermal Energy crops Gas CC CHP Biomass CC CHP Transport Gasoil car Area Wind Electrolysis Gasification Liquefication Gasoline car LH2 car Busses Trucks

16 System aspects: substitution Primary energy supply Conversion sector Coal processing End-use sectors Residential Gas heating Oil heating Lignite resources Oil resources Oil import Coal resources Coal import Gas resources Gas import Residual wood Refinery Gas processing Elec sector Coal cond. PP Lignite cond. PP Coal IGCC PP Gas CC PP Wind converter CHP sector Coal CHP Coal CHP Elc. heating Local heat grid Gas water boiler Elc. water boiler Elc. heat pump Commercial Room heat boilers Process heat boilers Warm water boilers Industry Industrial boilers Substitution options Area PV Solarthermal Energy crops Gas CC CHP Biomass CC CHP Transport Gasoil car Area Wind Electrolysis Gasification Liquefication Gasoline car LH2 car Busses Trucks

17 Partial equilibrium Primary energy supply Lignite resources Oil resources Oil import Coal resources Coal import Gas resources Gas import Residual wood Area PV Solarthermal Energy crops Conversion sector Coal processing Refinery Gas processing Elec sector Coal cond. PP Lignite cond. PP Coal IGCC PP Gas CC PP Wind converter CHP sector Coal CHP Coal CHP Gas CC CHP Biomass CC CHP End-use sectors Residential Gas heating Oil heating Elc. heating Local heat grid Gas water boiler Elc. water boiler Elc. heat pump Commercial Room heat boilers Process heat boilers Warm water boilers Industry Industrial boilers Transport Gasoil car Area Wind (Source: GianCarlo Tosato) Electrolysis Gasification Liquefication Gasoline car LH2 car Busses Trucks

18 Interdependencies At optimal solution: Equilibrium between electricity supply and demand Conversion Changes in Primary the system (e.g. introduction CO 2 price) sector yield new equilibrium energy e.g.: 1) Shift supply to low-carbon technologies Coal processing 2) Increase in electricity price Lignite resources Refinery 3) Substitution and saving of electricity in the end-use sectors End-use sectors Residential Gas heating Oil heating Elc. heating Local heat grid Oil resources Oil import Coal resources Coal import Gas resources Gas import Residual wood Area PV Solarthermal Energy crops Gas processing Elec sector Coal cond. PP Lignite cond. PP Coal IGCC PP Gas CC PP Wind converter CHP sector Coal CHP Coal CHP Gas CC CHP Biomass CC CHP Gas water boiler Elc. water boiler Elc. heat pump Commercial Room heat boilers Process heat boilers Warm water boilers Industry Industrial boilers Transport Gasoil car Area Wind Electrolysis Gasification Liquefication Gasoline car LH2 car Busses Trucks

19 Mathematical formulation Input data Cost data Efficiencies Full load hours Emission factors Demands Objective function: Model equations: Minimizing discounted system costs = Sum of Import/Extraction costs variable and fixed O&M costs Investment costs Transformation relationship ( e.g. efficiency relationship for power plant) Energy and emission balances Inter-temporal constraints on new capacity additions ( e.g. early investments needed for using technology on larger scale in later periods) Capacity-activity constraint ( e. g. available capacity limits elec gen of power plant) Decision/Output variables Process activities Energy/Emission flows New capacities Fundamental prices Cumulated constraints over time ( e. g. available fossil resources) Peaking constraint ( Ensuring reserve capacity at peak load) Load curve equations Storage equations (e.g. pumped storage) Scenario specific constraints ( e. g. bound on CO2 emissions quota for renewables)

20 Not only about modeling but also data analysis Energy security Outputs from macro and demand models Behaviour Existing policy Statistics Experts Assumptions Other energy system models Stakeholder workshop Literature review Demand projections Technology database Policy Scenarios Constraints Energy system model Resources Government Statistics Calibration of energy & emissions Expert knowledge International statistics Reports Reference Source: William Usher 2010

21 Energy Technology Data Source of ETSAP implementing agreement Consistent data set for more than 50 energy supply and demand technologies Free access on

22 Example of an energy system model: TIMES Development By ETSAP (Energy Technology Systems Analysis Program; Implementation in GAMS Model generator Methodology Bottom-up Model Perfect competition Perfect foresight (or myopic) Optimization (LP/MIP/NLP) Min/Max Objective function s.t. Equations Constraints Decision Variables <=> Solution Input parameters TIMES (The Integrated MARKAL EFOM System) Advanced Features/Variants Multi-regional Inter-temporal Elastic demands Endogeneous learning Discrete capacity expansion Early retirement of capacity Macroeconomic linkage Climate extension Stochastic programming Alternative objective functions Multi-criteria optimization

23 Applications of MARKAL/TIMES around the world Contracting parties Model users Used by more than 150 institutions in 63 countries

24 Applications: Island of Renunion (TIMES) (Source: Maïzi N. et al. Flexibility and reliability in long-term planning exercises dedicated to the electricity sector XXI World Energy Congress Montreal September )

25 Discussion Experience with energy models in your organisation? Which energy models are used in your country or region?

26 Summary Energy models provide a consistent analysis framework Energy models are simplified representation of the real-world system Level of detail depends on questions to be addressed and available data Choice of model type depending on analysis question Energy system models: focus on role of technologies and interactions within the energy sector Economic models: focus on interdependencies of the energy sector with the remaining economy Energy modelling is a long-term and continuous process involving experts from different disciplines Limitations of individual model approaches important when interpreting model results (no analysis is perfect) Energy planning not about predicting the future but supporting decision making under inherent uncertainties

27 Getting Started with TIMES: Running a Power Sector Investment Model Uwe Remme Energy Training Week April 2013

28 GW Key challenges Decommissioning scenario for existing global capacity Demand? How to replace? Geothermal Wind Solar Biomass and waste Nuclear Oil Natural gas Coal Hydro Aging power capacity has to be replaced Investment characterized by long lifetime and high upfront capital demand Decisions have to take into account long time horizon with uncertain or unknown conditions: Long-term energy prices? Economic development? Market conditions? Technology development? Climate policies? Operational aspects (variable renewables electrification)?

29 Typical structure of a power sector model Fuel supply Generation T&D Demand Coal supply Gas supply Wind potential onshore Coal plant subcritical Coal plant supercritical NGCC Gas turbine Wind turbine onshore Transmission and distribution Industry demand Transport demand Residential demand Wind potential offshore Wind turbine offshore Commercial demand Hydro potential run-of-river Hydro run-of-river Approach: Optimisation vs. simulation Foresight: Perfect foresight vs. myopic Horizon: Short-term (dispatch planning) vs. long-term (investment planning)

30 Least-cost optimisation Objective function: Minimising total discounted costs of power system including investment operating and fuel costs Constraints Coverage of given electricity demand Technical characteristics of power technologies Capacity constraints Resource constraints (domestic imports) Renewable potentials Decision variables Dispatch of power plants Operation of storage processes Construction of new power plants

31 Basic model constraints based on fundamental principles Principle TIMES equivalent Decision variables Energy conservation Energy conversion Generation of a technology limited by available capacity Commodity balance Transformation equation Capacity-utilisation constraint Objective function: Minimise system costs Flow variables Generation variables New capacity variables Cost coefficients of decision variables

32 Factors influencing power technology choice Existing capacity Inv. costs New capacity Tech. lifetime Econ. lifetime Reserve margin Capacity factor Total capacity Bounds FOM costs Bounds Firm capacity during peak load Full load hours Model variables Load curves Electricity generation Var. OM costs Inputs parameters Efficiency Fuel input Emission factors Emissions Fuel prices Fuel availability Taxes Sectoral bounds (annual or cumulative)

33 Exercise: Reference energy system IMPREG COL IMPCOL0 GAS IMPGAS0 NUC IMPNUC0 Coal USC plant ECOLSC0 Gas turbine EGASTUR Gas combined-cycle EGASCC0 Electricity ELECTR DEM_PRC Electricity demand DEM_TR WIN RNWREG RENWIN0 Nuclear LWR ENUCLWR Wind onshore EWINON0 REGION1 Reference energy system (RES) i.e. network of commodities and technologies used to describe real-world system RES can be adapted to considered case study RES contains existing and future technologies

34 Model interface: ANSWER-TIMES

35 ANSWER-TIMES: Technology representation

36 VEDA-BE: Data cube for result analysis

37 Excel interface for hands-on exercises

38 Hands-on exercises (ETW2013_TIMES_V1.xlsm) Question 1: Which gas price increase by 2050 is needed to see a switch from coal to gas in the electricity mix in 2050? Question 2: Introduce the CO 2 constraint and explore the mix of wind and nuclear by varying the nuclear investment costs. Question 3: Explore how a lower or higher discount rate changes the solution?

39 Power Sector Modelling with TIMES: Load curves and storage Uwe Remme (IEA) Steve Pye (UCL) Energy Training Week April 2013

40 Overview Power sector analysis: Off-model approaches for assessing cost-competitiveness of power technologies: LCOE & Screening curves + load duration curves Hands-on exercise: Load curves and storage in TIMES

41 LCOE (USD/MWh) Levelised costs of electricity (LCOE) Fuel costs Capital costs 50 0 USC Gas turbine NGCC Onshore wind PV utility PV rooftop CSP LWR Specific investments costs (USD/kW) Capital recovery factor crf LCOE n i1 i n 1 i 1 inv crf af Interest during construction factor idc 8760 fom Fixed operating and maintenance costs (USD/kW) 1000 fuelprice eff 3.6 Fuel price (USD/GJ) USD/MWh Availability factor as fraction Efficiency as fraction

42 Technology assumptions for LCOE Technology Overnight investment costs (USD/kW) Fixed O&M costs (USD/kW)) Technical lifetime (a) Construction time (a) Load factor (%) Efficiency (%) Fuel costs (USD/GJ) USC coal Gas turbine NGCC Onshore wind PV utility PV rooftop CSP LWR

43 Discussion What are the advantages and disadvantages of the LCOE approach?

44 Total annual power plant costs (USD/MW) Screening curves - Concept F i Hours of operation during the year Total annual plant cost fuelprice 6 eff inv crf idc fom 3. t Annual fixed costs F i Variabe costs V i

45 USD/(MWh/yr) Screening curves - Example USC Gas turbine NGCC % 20% 40% 60% 80% 100% Load factor Operating ranges Gas turbine NGCC USC coal

46 Variability of the electricity system Gas Oil Firm capacity during peak time Wind Nuclear Coal Hydro Load Reserve margin Demand load curves influence technology choice Need for flexible generation Storage to balance between peak and base load demand Demand side response Fluctuations also on supply side: some renewables (wind PV) and CHP plants (if heat-driven)

47 GW GW Load duration curve Chronological load curve over a year hours 8760 Reordering by load Load duration curve hours 8760

48 GW USD/(MW/yr) Combining screening and load duration curves USC 400 Gas turbine 300 NGCC % 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 30 Load factor Load duration curve 25 8 GW 20 Gas turbine 4 GW 15 NGCC GW USC 5 0 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Load factor Load factor

49 Discussion What are the advantages and disadvantages of the screening curve and load curve approach?

50 Hands-on exercise: Load curves and storage IMPREG COL IMPCOL0 GAS IMPGAS0 NUC IMPNUC0 WIN RENWIN0 RNWREG Coal USC plant ECOLSC0 Gas turbine EGASTUR Gas combined-cycle EGASCC0 Nuclear LWR ENUCLWR Wind onshore EWINON0 Electricity ELECTR DEM_PRC Pump ESTGPMPIN ESTGPMPOUT Turbine Electricity demand DEM_TR Electricity stored ELECTR_STG REGION1 Storage reservoir ESTGPMP000 ETW2013_TIMES_V2.xlsm

51 Hands-on exercise model: Load curves for typical days SP_01 SP_02 SP_03 SP_04 SP_05 SP_06 SP_07 SP_08 SP_10 SP_11 SP_12 SP_13 SP_14 SP_15 SP_16 SP_17 SP_19 SP_20 SP_21 SP_22 SP_23 SP_24 Percentage of annual demand Period 1 Period 2 Model horizon Period 3 Period 4 Example: Hourly load profile for four typical days 1.8% Representative year (milestoneyear) Spring SP SU FA WI Seasons 1.6% 1.4% 1.2% Weekday (level left out in exercise) 1.0% 0.8% 0.6% Winter day 0.4% 0.2% Fall day Spring day Summer day Hourly 0.0% Hours

52 Hands-on exercise model: Background on wind assumptions Capacity factor (-) Frequency Wind speed (m/s) Generation (MWh) Wind speed distribution Wind speed (m/s) Hours Assumed normalised operational curve for wind turbine Distribution of wind generation for 1 MW Wind speed (m/s) Wind speed (m/s)

53 Hands-on exercises (ETW2010_TIMES_V2.xlsm) Question 1: Run the model with and without storage. How does the addition of storage change the solution?

54 Key features of power sector analysis in energy system models Strengths Detailed technology representation of power sector Integrated optimisation of investment planning and system operation Deriving least-cost strategies to cover given electricity demand taking into account possible additional policy goals (how-to) Analysing impact of individual power technologies on costs energy use and emissions (what-if) Analysing interdepedencies between different policy goals/instruments e.g. CO 2 pricing and renewable support Weaknesses Operational aspects represented in an approximated fashion: No detailed representation of electricity grid infrastructure Ramp-up ramp-down constraints start-up costs minimum operation time part-load efficiency generally not considered Uncertainty can only be included to some extent (computational limitations) Perfect competition on electricity markets i.e. no market power of individual players