Decision-making for integrated energy systems 10th DS&OR Forum, - PowerPoint PPT Presentation

Decision-making for integrated energy systems 10th DS&OR Forum, Paderborn, 01/07/2017 Daniela Guericke ( dngk@dtu.dk ) Ignacio Blanco Henrik Madsen Technical University of Denmark Department of Applied Mathematics and Computer Science

The CITIES Project funded by Det Strategiske Forskningsr˚ ad through the CITIES research center (no.1035-00027B) CITIES = C enter for IT-I ntelligent E nergy S ystems in Cities and many partners from Danish Industry as well as international companies and research institutions www.smart-cities-centre.org 2 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

The CITIES Project The central hypothesis of CITIES is that by intelligently integrating currently distinct energy flows (heat, power, gas and biomass) in urban environments, we can enable large shares of renewables, and consequently obtain substantial reductions in CO2 emissions. Intelligent integration will enable lossless virtual storage. 3 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

The CITIES Project Continent Country City Geographical Scale District Power Building Appliance Data Communication Optimisation Models Complexity Integration based on data and IT solutions leading to models and methods for planning and operation of future flexible energy systems 4 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

The CITIES Project 5 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

  Project’s Methodology and Results The CITIES Project 6 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

CITIES WP7 - Decision-making and Support Methods Objectives: 1. Development of decision-making models for the optimal market participation of energy companies 2. Using the decision-making models to perform cost/benefit analyses for smart cities Methods: • Decision-making under uncertainty (stochastic programming, robust optimization) • Large-scale optimization (decomposition, heuristics) 7 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

CITIES WP7 - Decision-making and Support Methods Application areas: • Planning problems of energy companies • Evaluation of energy systems with special focus on: • Energy systems integration (mostly power and heat) • Flexibility and controllability of those sources • Portfolios of energy sources • Different sources of uncertainty 8 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Work in progress Decision support methods for - Biomass supply planning - Optimal heat production in district heating systems 9 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Work in progress Decision support methods for - Biomass supply planning - Optimal heat production in district heating systems Common background - Combined Heat and Power (CHP) plants ( Heizkraftwerk ) - District heating ( Fernw¨ arme ) 9 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Combined heat and power plants • produce electricity and warm water at the same time • are connected the electricity grid and the district heating network • in our case fueled by biomass Pmax Extraction Turbine ~ Power Generation Pmin Biomass Boiler Qmax 10 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply planning For a large scale producer the biomass is delivered based on supplier contracts which often have a runtime of one year. Our goal is to determine the optimal portfolio of biomass contracts minimizing the costs and taking uncertainty into account. 11 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply planning CHP plant Power Generation ~ Extraction Cold Water Turbine from DH Biomass Storage Biomass Heat Boiler Exchanger Hot Water to DH Heat Pump Tank Storage Natural Gas from Pipeline Gas Boiler 12 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply planning Biomass contracts • The biomass can be delivered by a number of suppliers • Each supplier offers one or more contracts defining the minimum and maximum amount of biomass per delivery, the total number of deliveries and frequency Biomass Storage Biomass storage • The biomass storage has a capacity and the maximum in- and outflow is restricted per period 13 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply planning Extraction-condensing CHP unit Pmax Extraction Turbine ~ Power Generation Pmin Biomass Boiler Qmax • Technical characteristics and limitations (e.g. min. up and down time, ramping constraints, capacity) • Commitment status 14 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply planning Gas Boiler • Handles heat demand peaks Natural Gas • Gas is bought instantaneously from the gas grid at from Pipeline Gas the natural gas spot price Boiler • Technical characteristics and limitations of the gas boiler (e.g. capacity) • Cheaper heat production than the CHP Heat storage • Stores heat that can be supplied to the DH Heat network Tank • Provides some flexibility to the system Storage • Is limited in capacity and flow 15 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply management Objective Cost minimization • biomass supply • operation ... and always satisfying the heat demand Uncertainty • Heat demand • Electricity prices • Natural gas prices 16 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply mass Heat demand 30 25 Heat Demand [MWht] 20 15 10 5 0 5000 10000 15000 20000 25000 Time [hours] 18 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply planning Stochastic programming with scenarios for heat demand, electricity prices and gas prices Hybrid model combining stochastic programming and robust optimization • Heat demand scenarios • Confidence intervals for gas and electricity prices, due to variability. Power prices for 1 week in June Forecasts from ARIMA(2,0,1)(1,0,1)[7] 45 100 Estimate Price 95% Confidence Interval 40 30 80 Power Price [Euro−MWhe] 35 25 Heat Demand [MWht] 60 30 Price [Euro−MWh] 20 40 25 15 20 20 10 15 0 5 10 0 500 1000 1500 2000 2500 0 50 100 150 0 5000 10000 15000 20000 25000 Time [hours] Time [hours] Time [hours] 19 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply planning Decision stages Operation of CHP plant and gas boiler in scenario 1 Selection of biomass Operation of CHP plant and contracts and gas boiler in scenario 2 delivery weeks . . . Operation of CHP plant and gas boiler in scenario Ω 20 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply planning Integrated model: • Time scale: Hourly for one year • Long computation times due to complexity • Apply decomposition techniques to reduce computation time Two-phase approach : • Divide in yearly planning and weekly planning • Forecasts can be updated every week • Loss of optimization potential, but closer to planning in practice 21 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Biomass supply planning Two-phase approach Long-term planning Operational planning Weekly production scheduling Selection of biomass contracts • Time scale: 1 week with hourly based on heat demand scenarios resolution • Time scale: 1 year with weekly resolution • Higher level of detail • Without most of the technical • Uncertainty of heat demand, requirements electricity price and gas price Resembles the planning process in practice 22 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018

Optimal heat production in district heating systems Demo project with EMD International A/S and ENFOR A/S Power Grid FB FB CHP CHP SC SC EB EB TS TS HP HP Heating Grid 23 Department of Applied Mathematics and Computer Science Decision-making for integrated energy systems 23.3.2018