Smart Grid Co-­‐Simulation with mosaik and the SESA-­‐Lab Rapid Prototyping of Future Energy Systems Sebastian Lehnhoff OFFIS – Institute for Information Technology Future Energy Systems (a.k.a. Smart Grids) MS
NS
u
NS
Integrating large amounts of active components into operation u
u
Environmentally dependent and hard to forecast Automated operation and (on-­‐line) optimization necessary OFFIS – Institute for Information Technology 2 ICT-­‐based Solutions Necessary Appropriate information, communication and automation systems are known from other domains u
u
u
But: long-­‐term use in safety-­‐critical energy systems mostly untested High risk for stakeholders in energy supply Rigorous testing necessary! u
u
u
u
Learning from other application domains... „Hardware in the Loop” Operation of the real electric controller hardware or a mechatronic component in a simulation of the real environment But: what belongs into this simulated environment? OFFIS – Institute for Information Technology Quelle: Daimler „Driving Simulator in Sindelfingen“, 2014 u
3 Influencing Factors of Future Energy Systems u
Relevant scope of „Smart Energy Systems” is hard to determine u
u
u
u
u
u
Renewable – fossil generation Distribution grid – transmission grid Users – consumers Markets ICT … u
u
Complex interactions Small effects gain relevance through scaling OFFIS – Institute for Information Technology 4 Simulation of Smart Energy Systems u
Adequate consideration of ALL facets required! u
u
Formal analysis (not feasible anymore) vs. simulation Integration of heterogeneous models and simulators Model Representations (non-­‐exhaustive) Partial Differential Equations (PDE) Ordinary Differential Equations (ODE) V
Algebraic Equations/Phasors (AE) Hz
Discrete Automata (DA) „trusted“ Time Series (TS) Probability Density Functions (PDF) „untrusted“ Static Rules/Descriptions (SR) 1 day OFFIS – Institute for Information Technology 1 hour 1 minute 1 second 1 cycle 1° at 50 Hz 5 Simulation of Smart Energy Systems u
Adequate consideration of ALL facets required! u
u
Formal analysis vs. simulation Major Challenges Integration of heterogeneous models and simulators u  Properly integrating hardware and software environments Model Representations u  Use case-­‐specific „functional” combination of coarse discrete (non-­‐exhaustive) models with high-­‐resolution (dynamic) models Partial Differential Equations (PDE) u  Quantifying aleatory (irreducible) and epistemic (reducible) uncertainties (black-­‐box models!) Ordinary Differential Equations (ODE) u
Rigorous testing schemes/strategies V
Algebraic Equations/Phasors (AE) Automatic composition and orchestration Hz of heterogeneous models, depending on budget (model availability, time etc.) Discrete Automata (DA) u  Design of Experiments (statistical scenario design, model Time Series (TS) „trusted“ exchangeability etc.) u
Probability Density Functions (PDF) Static Rules/Descriptions (SR) à Recurring processes… let’s get common and sound methods for that! „untrusted“ 1 day OFFIS – Institute for Information Technology 1 hour 1 minute 1 second 1 cycle 1° at 50 Hz 6 SESA-­‐Lab Smart Energy Simulation and Automation Laboratory (Hard-­‐ and Software Integration Platform)
sesa
Lab
sesa
Lab
sesa
Lab
Co-­‐Simulation Framework sesa
Lab
(OFFIS – Institute for Information Technology) Real-­‐time Automation Lab (University of Oldenburg) sesa
Lab
sesa
Lab
OFFIS – Institute for Information Technology 7 SESA-­‐Lab mosaik – modular Smart Grid Co-­‐Simulation u
u
u
u
Software suite developed at the OFFIS for automated composition and orchestration of heterogeneous energy system models Flexible interfaces for simulators (grids, markets, environment etc.) and controllers (users, „smart” ICT etc.) Powerful scenario description language (rule-­‐based instantiation and coupling of models) coordinated execution (simulation) u
After testing phase with international research partners now open source available (http://mosaik.offis.de) u
Currently ~1.000 downloads per month OFFIS – Institute for Information Technology 8 Practical mosaik Workshops u
Regular mosaik courses and user workshops abroad Hands-­‐on model integration (simulators and hardware) u
Most recent workshop on 24.09.2014 u
u
u
u
DTU Denmark 20 international participants from el. engineering, physics, mathematics and computer science Next workshops Tomorrow (CMU) u  April, 28-­‐29 (AIT) …registration still open! u
OFFIS – Institute for Information Technology 9 SESA-­‐Lab Smart Energy Simulation and Automation Laboratory (Hard-­‐ and Software Integration Platform)
sesa
Lab
sesa
Lab
sesa
Lab
Co-­‐Simulation Framework sesa
Lab
(OFFIS – Institute for Information Technology) Real-­‐time Automation Lab (University of Oldenburg) sesa
Lab
sesa
Lab
OFFIS – Institute for Information Technology 10 SESA-­‐Lab sesa
Lab
Topology-­‐free Interconnection and Assignment of I/O (analog and digital) sesa
Lab
Supervisory Control SCADA IEC 61850 UML CIM (61970/61968) Protocol Switch BECKHOFF CX2020 sesa
Lab
BECKHOFF C6920 analog I/O-­‐Switch D/A D/A Substation Control SCADA Protection DER/Component Models Agents Controllers ... OPC UA (62541) IEC 60870 ... Standard-­‐compliant Information and Process Chains RT-­‐Unit-­‐Targets DER/Component Models Agents Controllers digital EtherCAT Ethernet Communication Simulation OPAL-­‐RT eMEGAsim RT-­‐Network-­‐Target Grid Model DER/Component Models OFFIS – Institute for Information Technology High-­‐End Computing Cluster sesa
Lab
D/A VPN Gateway ... Router/Firewall ... External Sessions @ 1Gbit/s Virtualization Server sesa
Lab
Station Control sesa
Lab
Internet Switching States Utilization/Configuration LabView Console Virtualization Server mosaik-­‐Simulation Licensing Servers IDE 11 SESA-­‐Lab Smart Energy Simulation and Automation Laboratory (Hard-­‐ and Software Integration Platform)
sesa
Lab
Steady State Dynamic sesa
Lab
(large scale) (small scale) sesa
Lab
u
u
sesa
Hardware (
emulation) i
n t
he s
imulation l
oop… Lab
Co-­‐Simulation Framework Real-­‐time Automation Lab (OFFIS –
I
nstitute f
or I
nformation T
echnology) (University of Oldenburg) Studying dynamic effects of a large amount of active components
u
u
u
sesa
Dynamic stability of distributed Labcontrol Oscillatory effects of market interactions sesa
Dynamic effects of synchronized user behavior due to DR/DSM schemes Lab
OFFIS – Institute for Information Technology 12 SESA-­‐Lab Mission u
u
u
Systematic design of operational concepts/controllers in more and more complex/extensive energy systems Development of tools/models for systematic integration and handling of heterogeneous/external models and processes SESA-­‐Lab is NO replacement for existing tools and models u
u
Goal of energy informatics at OFFIS/University of Oldenburg u
u
u
Integration platform for established tools and approaches Interdisciplinary collaboration with domain experts from electrical engineering, economy, social sciences etc. Creating system competence, developing system intelligence International network u
u
CO-­‐simulation-­‐based energy SYstem Modeling plAtform (COSYMA) UC Berkeley/Berkeley National Lab (us), NREL (us), CMU (us), TU Delft (nl), AIT (at), DTU (dk), OFFIS (de) OFFIS – Institute for Information Technology 13 References Integration of Aristo (Real-­‐Time el. Grid Simulation) Integration of Power Hardware (-­‐in-­‐the-­‐loop), Integration of Modelica Models through FMI Integration of SYSLAB (Flexible Intelligent Energy Laboratory) Substation Modeling and Simulation, Integration of MAS-­‐based Transmission System Control Integration of Real-­‐Time Low-­‐Voltage Grid Simulation and State Estimation Hybrid Simulation of large-­‐scale Distribution Networks Statistical Scenario Design and Validation Smart Grids Joint Programme What about you? OFFIS – Institute for Information Technology 14 Thank you! Prof. Dr. Sebastian Lehnhoff Energy Information Systems lehnhoff@offis.de OFFIS – Institute for Information Technology 15