Dr. Ing. Emanuele Martelli - Politecnico di Milano, Dipartimento di Energia Attività di ricerca del gruppo GECOS sui sistemi trigenerativi ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
1 Attività di ricerca del gruppo GECOS sui sistemi trigenerativi avanzati, i multi-energy systems e l'efficienza energetica dei processi industriali Dr. Ing. Emanuele Martelli Politecnico di Milano, Dipartimento di Energia February 7th 2018 , Ferrara
THE DEPARTMENT OF ENERGY @ POLITECNICO, DIVISIONS & OUR GROUP 2 The Department of Energy at Politecnico joins researchers originally belonging to 5 divisions. It has 130 permanent researchers and professors 1. Chemical Technologies and Processes and NanoTechnologies Division 2. Electrical Division 3. Nuclear Engineering Division 4. Thermal Engineering & Environmental Technologies Division 5. Fluid Dynamic Machines, Propulsion & Energy Systems Division • Fluid-dynamics of turbomachines • Internal combustion engines • Propulsion and combustion • Group of Energy COnversion Systems (GECOS): 1 Emeritus professor (prof. Macchi) 3 Full professors (Lozza, Consonni, Chiesa) 9 Associate/Assistant professors 6 RTDA 4 Post docs 5 Temporary researchers 10 PhD students www.gecos.polimi.it
MAIN RESEARCH AREAS 3 1. CARBON CAPTURE TECHNOLOGIES 2. RENEWABLE ENERGY SOURCES AND WASTE-TO- ENERGY 3. ENERGY STORAGE, HYDROGEN AND SYNTHETIC FUELS 4. MICRO-GRIDS AND MULTI-ENERGY SYSTEMS COAL + MPGs TO F-T LIQUIDS + ELECTRICITY, WITH CCS Cequiv balances to atmosphere for F-T liquids OUT: photosynthesis (MPGs, soil&root C), electricity credit (2,852 tC/day) 5. ENERGY EFFICIENCY AND SYSTEMS OPTIMISATION IN: upstream emissions, vented at plant, fuels burned in vehicle,s (2,852 tC/day) transportation carbon vented 1,607 tC/day coal upstream emissions photosynthesis credit for e.e. prairie grasses upstream emissions 1,810 tC/day 223 tC/day 735 tC/day fuel for 225 tC/day 83 tC/day 1,032 MWLHV electricity production 6. ORC, S-CO2 AND ADVANCED POWER CYCLES 452 MWee prairie grasses 1,607 tC/day coal 668 MWLHV 5,328 tC/day 2,449 MWLHV polygeneration plant char 7. FUEL CELLS accumulation in 53 tC/day soil and root carbon storage 1,022 tC/day 4,337 tC/day arrows’ width proportional to C fluxes >10 ongoing EU projects (FP7-H2020) Tens of ongoing research contracts with industries
OPTIMIZATION OF MULTI ENERGY SYSTEMS (MES) AND CHP 4 Types of optimization problems associated to MES (for microgrids and DHC networks): 1. Design/retrofit of the system («investment planning») 2. Long-term operation planning accounting for yearly constraints (incentives/seasonal storage) 3. Short-term scheduling (unit commitment) 4. Optimal control (dynamic models of units and networks) 4
OPTIMIZATION OF MES AND CHP: ONGOING COLLABORATIONS 5 EPFL, MIT, Boston ETH, Zurich Lausanne Skoltech, Univ. of Malaga Moscow GECOS group DEIB, Polimi Univ. of Parma Univ. of LEAP Bologna 5
SHORT-TERM SCHEDULING PROBLEM 6 Given: ➜ Forecast of Electricity demand profile ➜ Forecast of heating and cooling demand profile ➜ Forecast of production from renewables ➜ Forecast of time-dependent price of electricity (sold and purchased) ➜ Performance maps of the installed units ➜ Operational limitations (start-up rate, ramp-up, etc) of units ➜ Efficiency and Maximum capacity of storage systems Objective: minimize the Daily/Weekly Operating Cost 24·7 24·7 24·7 24·7 CFuel,tot,t + & , , + C − , , ∓ , =1 =1 =1 =1 Indep. variables: on/off of units, load of units, storage level in each time period t
SHORT-TERM SCHEDULING PROBLEM 7 Constraints: ➜ Electric energy balance constraint Ɐ t (linear) ➜ Heating energy balance constraint Ɐ t (linear) ➜ Cooling energy balance constraint Ɐ t (linear) ➜ Start-up constraints Ɐ t, Ɐ unit (linear) ➜ Ramp-up constraints Ɐ t, Ɐ unit (linear) ➜ Performance maps of units Ɐ t, Ɐ unit i (nonconvex) Nonconvex MINLP Available MINLP optimizers cannot find the optimal solution Amaldi et al., 2017. Short-term planning of cogeneration energy systems via MINLP, in SIAM book: “Advances and Trends in Optimization with Engineering Applications”.
MILP WITH PWL APPROXIMATION OF MAPS 8 Basic idea: conversion into MILP via linearization of the performance maps Advantages of MILP formulations: - Guarantee on the global optimality of the solution - Super-efficient commercial MILP solvers (e.g., CPLEX, Gurobi) 2-D PWL approximation with the «triangular method» 1-D PWL approximation D’Ambrosio et al. 2010. Op. Res. Letters Usefull effect Bischi et al. 2014. Energy, Vol. 74
MILP PWL FOR DAILY AND WEEKLY PROBLEMS 9 E High temperature heat H elcust l 7000 e elHP,comp e Heat Pump, qlow,HP,comp 6000 a llow,stor c compression t 5000 High Temperature Heat [kWh] t U 4000 r elpur fAB, LT qlow,AB S S 3000 i Aux. Boiler, LT t e 2000 c elsold o r elICE r. qlow,cust 1000 qlow,ICE fICE 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 G ICE qhigh,ICE qlow,los -1000 Hour r qdeg Cogenerated ICE Cogenerated GT GT Post Firing i elGT qlow,GT Auxiliary Boiler Downgraded Demand d fGT qhigh,cust fGT,PF,Not,cog GT+PF qhigh,GT Bischi et al. 2014. Electricity qhigh,GT,PF,Not,cog fAB,HT Aux. Boiler, HT Energy, Vol. 74 5000 qhigh,AB 4000 3000 Computational time: 2000 Electric Energy [kWh] 1000 1 day operation: < 1 sec 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 -1000 1 week operation: < 2 min -2000 -3000 Up to 18% primary energy saving compared -4000 Generated ICE Hour Generated GT Consumed HP to usual operation strategies! Purchased Sold Demand
OPTIMAL OPERATION WITH COGENERATION INCENTIVES 10 1. COGENERATION INCENTIVES: if CHP units allow to save fuel compared to “business as usual systems”, subsides are granted (additional revenue) Conditions to be met for ICE CHP: First Principle Efficiency Primary Energy Saving (PES) ℎ + _ − _ ≥ 75% ≥ 10% _ Easily rearranged as linear constraints but yearly-basis! 2. SEASONAL STORAGE SYSTEMS: Underground systems (water tanks, aquifers, etc), sorption systems (e.g., sodium sulfide), thermochemical storage systems and H2 storage. Challenge: operation must be optimized considering the whole year! IDEA: Rolling-horizon algorithm (Bischi et al. 2018. Energy, in press)
ROLLING HORIZON ALGORITHM FOR OPTIMAL OPERATION WITH INCENTIVES 11 Test case: CHP system of a large hospital Computational time: 21 hours (worst case) applicable for day-ahead scheduling elcust E qdiss,low, AB 1 fAB, LT 1 H l qlow,AB 1 Aux. Boiler LT 1 e e qdiss,low, AB 2 a llow,stor fAB, LT 2 qlow,AB 2 c Aux. Boiler LT 2 t t elpur U qdiss,low,ICE 1 S r S elsold fICE 1 qlow,ICE 1 t i qlow,cust e ICE 1 o c elICE 2 qhigh,ICE 1 r. r qhigh,diss,ICE 1 qlow,los qdiss,low,ICE 2 qdeg G fICE 2 qhigh,cust r qlow,ICE 2, i elICE 1 ICE 2 qhigh,ICE 2 d qhigh,diss,ICE 2 fAB, HT 1 Bischi et al. 2018. Energy (in press) Aux. Boiler HT 1 qhigh,AB 1 qhigh,diss,AB 1 fAB, HT 2 Aux. Boiler HT 2 qhigh,diss,AB 2 qhigh,AB 2 About 7% higher revenue (yearly basis) compared to the weekly optimal operation.
OPTIMAL OPERATION OF MES AND CHP SYSTEMS UNDER 12 FORECAST UNCERTAINTY Given: ➜ Forecast of power production from renewables and its uncertainty ➜ Forecast of energy demand profiles and its uncertainty ➜ Forecast of electricity prices and its uncertainty ➜ Performance maps of the installed units ➜ Operational limitations (start-up rate, ramp-up, etc) of units ➜ Efficiency and Maximum capacity of storage systems Determine: Nominal set-points: on/off of units, nominal load of units, storage man. Correction rules: how to adjust units loads during operation Minimizing the daily operating cost in the most probable scenarios Constraints: meet energy demands, technical limitations of units, performance maps, etc EFFICITY PROJECT (LEAP-Polimi, CIDEA, CIRI-EA, CERR) co-funded by Emilia Romagna Region (POR-FESR 2014-2020)
OPTIMAL OPERATION OF MES UNDER FORECAST UNCERTAINTY: ROBUST MILP 13 Pel [kW] Time [h]
OPTIMAL DESIGN OF MES AND CHP SYSTEMS 14 Given: A catalogue of possible units (e.g., CHP ICEs, HP GTs, boilers, heat pumps, etc) and the list of available sizes (discrete or continous) A catalogue of heat storage systems Forecast of future energy demand profiles (whole year) of each building/site Forecast of future energy prices and their time profiles of each building/site Determine: Which units and heat storage system to install in each district/building The sizes of the units and storage systems The required energy connections between sites/buildings Considering: Nonlinear size effects on investment costs and efficiency On/off and part-load operation of the units Objectives: Maximize the NPV/Minimize the energy consumption/CO2 emissions
DESIGN OPTIMIZATION APPROACHES 15 Linearization without decomposition Design-scheduling decompositiom (Gabrielli et al. 2018, Applied Energy, in press) (Elsido et al., 2017. Energy Vol. 121) (Zatti et al, 2017, Comp. Aided Chem. Eng., 40) Upper level (evolutionary alg.): optimizes Units’ selection, sizes and operation optimized in design variables (selection/sizes) a single large scale MILP Lower level: MILP scheduling problem Advantages: Avantages: • Linear problem (computationally efficient) • Size effects accounted for on both • Global optimality guarantee performance and costs • Uncertainty can be rigorously handled • Possibility of considering many operating periods solved in sequence Disadvantages: Disadvantages: Scale effects on investment costs must be linearized Slow convergence rate of evolutionary algorithms Size effects on efficiency must be approximated No optimality guarantee
OPTIMAL DESIGN OF MES: DESIGN OF DH NETWORKS FOR A CITY 16 Deterministic MILP solution Stochastic MILP solution CAPEX+OPEX= 21,92 M€ CAPEX + OPEX= 18,22 M€ boiler site 2 site 2 heat pump 1,4 MW 1,4 MW gas turbine 1,4 MW 1,4 MW HS heat storage 32,6 MWh HS 41,5 MWh thermal connection HS Q23,max = 0,9 kW Q12,max = 1,8 MW Q23,max = 1,3 kW Q12,max = 0,9 MW site 1 site 3 site 1 site 3 1,4 MW 4,5 MW 6,7 MW 11 MW 4,5 MW 1,4 MW 5 MW Q13,max = 4,1 MW Q13,max = 2,5 MW HS ES 78,5 MWh 13,7 MWh ES HS 6,2 MWh HS HS 78,5 MWh Accounting for uncertainty of forecasts leads to a cost saving of about 20%!
OPTIMAL DESIGN OF MES: RETROFIT OF DH NETWORKS 17 DH network of Brescia University of Parma Campus Storage Tipo 1 PdC Fumi TU Linea1 Storage Tipo 1 PdC Fumi TU Linea2 GR3 Caldaie Storage Tipo 2 Lamarmora Lamarmora PdC Fumi TU Linea3 Recupero Calore: Ori Martin Recupero Calore: Altre P.d.C industrie (Verz.+Caff +Concesio) Caldaie Extra Solare Storage Termico Tipo 3 Caldaie Storage Tipo 3 Centrale Nord EFFICITY project: Tritorno Tmandata Collab. with LEAP, CIDEA & SIRAM Utenze RTR Collab. with Univ. Of Brescia
LABORATORIES 18 LAB for MICRO-COGENERATION (LMC) Test bench for performance and emissions of small scale co- and tri-generation systems (ICE, Stirling engines, PEM, SOFC, etc with < 100 kWel, 300 kWth), electrolizers and H2 production systems (at steady-state and transient conditions). SOLAR TECH Facility for developing and testing photovoltaic, photovoltaic-thermal and concentration systems. MICRO-GRIDS LAB Facility for testing control strategies of hybrid micro-grids (solar panels, CHP engines, batteries, thermal storages, etc). LABORATORIO ENERGIA AMBIENTE PIACENZA (LEAP) Research and consultancy activities in the areas of energy efficiency, waste-to- energy plants, biomass-fired plants, and pollutant emissions.
LAB FOR MICRO-COGENERATION (LMC) 19 1 kWel CHP unit based on Stirling cycle 10-20 kWel CHP units based on internal combustion engines 20-30 kWel CHP units based on PEM fuel cell with steam reformer 1 kWth membrane reactor test stand (metallic membranes for hydrogen separation and application to fuel processing ) 0.5 kWth Thermo-photo-voltaic (TPV) generator 4 x 2.5 kWel SOFC CHP unit 230 170 30 kWel - 18 kWLHV H2 electrolyzer unit At steady-state and transient / cyclic conditions Controlled and measured fuel / electricity exchange, temperature, pressure and mass flow of all I/O hot/cold circuits. 19
LABORATORIO DI SISTEMI SOLARI - SOLAR TECH LAB 20 PV system optimization: - improvements of commercial devices/technologies (e.g., PV, PVT, inverters, etc) - development of innovative concepts/prototypes (e.g., solar-driven heat pump) - development of accurate forecasting tools Facilities: - PV and PVT panels with adjustable tilt angle and distance - Thermal oil loop to test PVT panels - Meteorological station 20
LEAP 21 Attività LEAP nel settore cogenerazione ed efficienza industriale • Ricerca nel settore MES e sistemi CHP in collaborazione con il gruppo GECOS • Consulenza scientifica per il recupero di calore/energia da processi industriali (BP, GE, etc) • Consulenza tecnica e procedurale per l’accesso a sistemi incentivanti sulla produzione efficiente di energia termica: CAR (Cogenerazione ad Alto Rendimento) e TEE (Titoli di Efficienza Energetica) Audit energetico degli impianti; Valutazioni tecnico-economiche sui meccanismi di remunerazione dell’energia; Verifica della rispondenza alle normative di riferimento; Validazione e supporto tecnico dei modelli per il calcolo del PES (Primary Energy Saving); Assistenza per la presentazione dei documenti verso il GSE. • Consulenza tecnica e procedurale per il riconoscimento degli incentivi per la produzione di energia elettrica da impianti a fonte rinnovabile: In particolare: per impianti a biomassa e impianti ibridi alimentati da rifiuti parzialmente biodegradabili. 21
MAIN PUBLICATIONS ON MES/CHP/MICRO-GRIDS OPTIMIZATION 22 1. Gabrielli et al., 2018. Optimal design of multi-energy systems with seasonal storage. “Applied Energy” (in press). 2. Zatti et al., 2017. A three-stage stochastic optimization model for the design of smart energy districts under uncertainty. “Computer Aided Chemical Engineering” Vol. 40, pp. 2389-2394. 3. Elsido et al., 2017. Two-stage MINLP algorithm for the optimal synthesis and design of networks of CHP units. Energy, Vol. 121, pp. 403-426.. 4. Taccari et al. 2015. Short-term planning of cogeneration power plants: a comparison between MINLP and piecewise-linear MILP formulations. Computer Aided Chem. Eng., 37, pp. 2429-2434. 5. Bischi et al., 2014. Tri-Generation Systems Optimization: Comparison of Heuristic and Mixed Integer Linear Programming Approaches. Proceedings of ASME Turbo-Expo 2014. 6. Bischi et al., 2014. A detailed optimization model for combined cooling, heat and power system operation planning. Energy, Vol. 74, pp. 12-26. 7. Bischi et al. 2016. Distributed cogeneration systems optimization: multi-step and mixed integer linear programming approaches. International Journal of Green Energy, Volume 13. 8. Mazzola et al., 2015. A detailed model for the optimal management of a multigood microgrid. Applied Energy, Vol. 154. 9. Mazzola et al., 2017. Assessing the value of forecast-based dispatch in the operation of off-grid rural microgrids. Renewable Energy, Vol. 108. 10. Mazzola et al., 2016. The potential role of solid biomass for rural electrification: A techno economic analysis for a hybrid microgrids in India. Applied Energy, Vol. 169.
23 Thank you for your attention! emanuele.martelli@polimi.it www.gecos.polimi.it 23
OPTIMAL DESIGN OF A SWISS ENERGY DISTRICT (WITH ETH) 24 Possible units Selected units Gabrielli et al. 2018, Applied Energy, in press
You can also read