Energy Distribution Systems with Multiple Energy Carriers

Transcription

1 Symposium Gas and Electricity Networks May 2002, Brasilia Energy Distribution Systems with Multiple Energy Carriers Bjørn H. Bakken, PhD Michael M. Belsnes, MSc Jarand Roynstrand, MSc N-7465 Trondheim Norway

2 Energy Distribution Systems with Multiple Energy Carriers Background New technologies for energy conversion, storage and transport are emerging better possibilities to design sustainable energy systems for the future more complex energy systems to design, operate and maintain An overall system perspective is necessary More flexible and comprehensive planning tools are needed New tool under development with detailed technology models combined in a generic linear network

3 Customer needs are related to use and comfort, not specific forms of energy Hot water Communication Data Lighting Heating Cooking

4 Possible energy sources to end-users Electricity Solar Biomass District heating Hydrogen Gas Fuel oil Ambient heat

5 Possible local energy resources Hydropower Solar Waste Gas / Oil Windpower Hydrogen Biomass Ambient heat

6 Energy Distribution Systems with Multiple Energy Carriers Methodology OBJECTIVE: develop a flexible tool for analysis of complex energy systems with multiple energy carriers generate system model with standard modules for transport channels, energy conversion processes and storage capacity detailed models of different energy technologies and energy carriers connection to superior system model through simple and uniform set of variables superior system analysis in a general network model multi-criteria optimization

7 Sample: Energy supply to industrial site ELECTRICITY DISTRICT HEATING GAS TRANSPORT GAS PIPELINE ENERGY DEMAND: ELECTRICITY SPACE HEATING AIR CONDITIONING / COOLING STEAM ETC.

8 System model for industrial site RESOURCES SITE AREA BUILDING NO i END USE EL AC SYSTEM EL HP 2 HP 1 DISTRICT HEATING DH SYSTEM / PIPELINE HEAT BOILER 1 COGEN 1 HEAT EXCH 1 GAS TRSP STORAGE COGEN 2 GAS GAS PIPE BOILER 2 STEAM SYSTEM / -PIPE STEAM

9 Network model for industrial site SOURCES SINKS EL T1 T2 T3 T4 EL DISTRICT HEATING P1 T5 T6 T7 T8 P2 HEAT GAS TRSP P3 P4 P5 P6 T9 T10 P8 P10 GAS PIPE P7 T11 T12 T13 P9 STEAM

10 Case: Hylkje suburb, Bergen Currently: RollsRoyce engine engine factory: factory: GWh GWh e / e / MW MW Foundry: GWh GWh e / e / 6 MW MW Weak Weak kv kv grid grid Planned: New New suburb suburb including public public services Gas Gas fired fired co-generation and and district district heating heating network using using surplus surplus heat heat

11 Hylkje system model Hylkje DC load flow model Foundry Bogen RollsRoyce Nesjavegen End-use Nesjavegen Gas CHP Langmyren End-use Langamyren District heating model

12 Hylkje sample results Energy price System Load Øre/kwh Electricity Gas MWh/h Sum Heat Electricity Time Time Energy purchase from sources Energy for space heating MWh/h Gas Electricity Sum MWh/h Distr. heat Electricity Sum Time Time

13 Energy Distribution Systems with Multiple Energy Carriers Summary Introduction of new technologies for energy conversion and transport creates more complex energy systems to design and operate More flexible and comprehensive tools are needed to handle multiple technologies New tool under development with standard modules for transport channels, energy conversion processes and storage capacity connection to superior system model through simple and uniform set of linear variables superior system analysis in a general network model multi-criteria optimization

14 Symposium Gas and Electricity Networks May 2002, Brasilia Energy Distribution Systems with Multiple Energy Carriers Bjørn H. Bakken, PhD Michael M. Belsnes, MSc Jarand Roynstrand, MSc N-7465 Trondheim Norway

15 Energy Distribution Systems with Multiple Energy Carriers Special Report Questions - 1 Which are the dominating objectives and criteria in the planning process of the total energy distribution system? Who should make the weighing factors? Tool for scenario studies to evaluate different investment alternatives, price sensitivity, public taxes, environmental constraints etc Use of objectives and criteria depending on who is using the tool and which aspects they are considering Private investors: Minimize costs subject to existing and future public demands like environmental taxes and/or constraints How sensitive are the alternatives to different demands? Official staff: Obtain public goals regarding environment and energy efficiency by using taxes and constraints What means are best suited to reach the goals without destroying the economy of the projects?

16 Multi-criteria optimization Economy (min C) Environment (min U) In In addition: Risk Risk Security Public Public opinion Regulatory framework etc etc Energy efficiency (max E)

17 Energy Distribution Systems with Multiple Energy Carriers Special Report Questions - 2 Is a multi-criteria approach still valid under liberal market conditions? In general, an actor in an ideal liberal market might have a single criterion optimization (max Profit / min Cost), but he would still have to consider aspects like financial risk, security, environment, public opinion etc Note that even though market operations are liberalized, the system expansion planning is still highly regulated with numerous public demands, constraints and taxes Multi-criteria approach is a systematic approach to handle non-quantifiable and non-comparable objectives related to energy system planning like security, aesthetics, public opinion etc In addition, multi-criteria methods are suited for fuzzy techniques with soft constraints

18 Optimization Operation planning level System database Topology alternative i t = 1 to N District heating model Electrical transmission model Waste transport model Detailed LP model YES t < N? Results

19 => Expansion planning level GUI Generate topology alternatives i = 1 to J System database Results OPERATION PLANNING Results YES More alternatives? Choose best expansion plan

20 Mathematical tools Linear Programming with possible extension to MIP AMPL as matrix generator, CPLEX for solver(s) + Mathematical equations rather than program code + Easy to modify and expand during testing and development + Easy to use in iterations - Difficult to implement GUI - Possible limitations in tool difficult to bypass - Difficult to combine several solvers in AMPL - Expensive! Final implementation environment not decided