GREEN TECH CENTER ÅBNINGSEVENT DEN 24. JUNI 2014 WORKSHOP 2 SMARTERE ENERGILØSNINGER VIA BIG ENERGIDATA Anders Midtgaard Thomas Hune General Manager INSERO Software A/S [email protected] +45 4082 8811 Director of Energy INSERO Software A/S [email protected] +45 40158022 Mikkel Baun Kjærgaard Associate Professor ved Center for Smarte energiløsninger Mærsk Mc-Kinney Møller Institute [email protected] +45 2197 2447 Emil Holmegaard PhD Studerende ved Center for Smarte energiløsninger Mærsk Mc-Kinney Møller Institute [email protected] Baggrunden for Green Tech Centeret er omstillingen af samfundet til vedvarende energi. http://www.youtube.com/watch?v=lWuRxLNcCV4 Lean'Energy'Cluster'fokus' ' ' ' ' • A'cleantech'ini+a+ve'in'the'field'of' energy'and'climate'technology' ' • Work'to'promote'energy'efficiency,'based' on'sustainability'across'the'value'chain'of' the'en+re'energy'system.' ' • Work'to'create'na+onal'as'well'as' interna+onal'cluster'ac+vi+es.' ' • We'bring'together'companies,'public' authori+es'and'knowledge'ins+tu+ons'to' secure'growth'in'the'market'and'new' businesses.' ' Lean'Energy'Cluster'fokus' Green Tech Eksport Hubs Nantong Hvorfor Big Data Mængde af devices og hastigheden på dataproduktion stiger voldsomt I 2016 forventes verdens datacentre at håndtere 6.6 zettabytes pr. år Svarende til hver person på jorden streamer 2,5 timers HD video pr. dag Vi har teknologier til at lagre alle disse data Udfordringen er at strukturere data, søge og finde information Det udfordrer vores traditionelle SQL teknologier Krav til Scalebility: Mængden af data Antal brugere Typer/format af data Vore devices vil ‘tale sammen’ Brugeren vil ikke vide hvilke data information er baseret på Informationerne vil komme fra forskellige services Disse services vil IKKE blive leveret fra samme leverandør Stiller krav om åben kommunikation og adgang til fælles data Adgang til data vil skabe en ny platform for forretningsudvikling og nye services Hvordan sikrer man kommunikation mellem markedsdeltagere http://www.youtube.com/watch?featur e=player_embedded&v=OIRxKHZuIiw Hvordan sikrer man kommunikation mellem markedsdeltagere http://www.youtube.com/watch?featur e=player_embedded&v=OIRxKHZuIiw Energisektoren har samme udfordring som ATM og standarder vil blive udviklet til en åben kommunikation http://www.youtube.com/watch?featur e=player_embedded&v=OIRxKHZuIiw Green Tech Centeret Data fra Green Tech Centeret Detaljeret view af data og kalkulerede værdier Data fra Energilaug Vejle Nord Stor Skala legeplads for smart Grid løsninger VVM for 3-5 stk. 130 vindmøller p.t. 15 store virksomheder Stor Skala legeplads for smart Grid løsninger p.t. tilsagn fra AAB Landsby (Andkær) Stor Skala legeplads for smart Grid løsninger p.t. 50 familier Data i dataplatformen i øjeblikket… • • • • • • 15 industrivirksomheder Green Tech House Advice House 5 Testplatforme 20 En-familie huse 20 EL-biler • Antal fysiske målepunkter ca. 700 (10 sek. data) • Kalkulerede dataværdier ca. 400 • Historiske data for 3 år for nogle industrier Data i dataplatformen i nære fremtid… • • • • 10 industrivirksomheder Green Tech House (BMS system) 20-200 En-familie huse 40 lejligheder • Antal fysiske målepunkter ca. 5000 (10 sek. data) • Kalkulerede dataværdier ca. 2500 • Historiske data for 3 år for alle huse/industrier Big data Big data is high volume, high velocity, and/or high variety information assets that require new forms of processing to enable enhanced decision making, insight discovery and process optimization. - Laney, 2012 Volume and Velocity Fra manuel til automatisk data indsamling i høj opløsning Antal data punkter om året for en kontor bygning 1E+14 1E+12 1E+10 100000000 1000000 10000 100 1 Manuelt Smart Meters Sub Metering (0.02 Hz) Sub Metering (1 Hz) Fuldstændig Sub Metering (3600 Hz) Variety • Vejrforhold – Temperatur, Fugtighed, Vindstyrke, Vindretning, Solindstråling, … • Aktiviteter – Menneskelig tilstedeværelse og adfærd – Produktions aktiviteter • Udstyr • Andre typer af data Nye måder at behandle data på Big Data Machine Learning Information Visualisation Information Visualisation Forbrug per blok i Los Angeles Vinter Sommer http://sustainablecommunities.environment.ucla.edu/maproom/index.html Energi Effektivitet Mål: Forstå hvordan machine learning og informations visualisering kan hjælpe virksomheder med at optimere deres energiforbrug. Studie med virksomheder i energilauget. • Interview om energi effektivitet • Evaluering af tyve forskellige typer af visualisering af data Eksempel: Advicehouse 5500 m2 Ca. 100 medarbejdere Elektricitetsmålinger I 2014 (106.692 kWh) Data i spil Virksomheder Kommune Borgere Organisationer Hvem har adgang / må bruge data og til hvad? Hvordan får man adgang til data? Workshop Problemstillinger Husholdningen Kommunen Virksomheden Typer af Data Vejr Data Energi Data Adfærds Data Green Tech Center Sociale Medier Offentlige Data Opgave: Fortæl en historie 3 grupper 30 minutter Lav en tegneserie der håndtere en problemstilling hvor der benyttes data fra Green Tech Centeret og Big data teknikker. Vejle og Resilience http://100resilientcities.rockefellerfoundation.org/cities/entry/vejles-resilience-challenge Resilience Hvordan kan big data data hjælpe? Fra Rockefeller fonden: • Adaptive: changes based on new evidence • Reflective: Learns from past experiences • Robust: is organized & transparently managed • … Examples of (Big)Data analysis in FINESCE Thomas Hune, Director, INSERO Software Overview FINESCE (Future INtErnet Smart Utility ServiCEs) is the smart energy use case project of the 2nd phase of Future Internet Public Private Partnership Programme (FI-PPP) funded by the European Union within FP7. 7 trial sites combining Smart Energy Solutions with Future Internet technology. FINESCE Partners and Trial Sites trial site partner location Agenda • WP1 presentation from E.ON • WP2 presentation from Insero • WP2 presentation from SEnerCon FINESCE WP1 FI providing the sustainable smart city energy COMPANY LOGO David Lillienberg FI providing the sustainable smart city energy Scope • Elaborate how Future Internet technologies can contribute to an efficient and robust Demand Side Management system • Execute Demand Side Management and Demand Side Response tests with external buildings in Malmö, Sweden, based on an integrated approach of energy carriers Participants • E.ON, RWTH Aachen WP1 introduction Scope • The scope of the WP1 trial is to execute Demand Side Management and Demand Side Response tests with external buildings Malmö, Sweden, based on an integrated approach of energy carriers Desired outcomes • How Future Internet technologies can contribute to an efficient and robust Demand Side Management system • Evaluate and test different business model(s) according to defined use cases to obtain better view on Demand Side Management and Demand Side Response as well as ideas on customer’s potential to act as balancing power • Scale-up strategy for the trial, e.g. ability for other towns/regions/business sectors to use the results/functionality Participants • E.ON, RWTH Aachen • WP leader: David Lillienberg, E.ON Hyllie, Malmö WP1 Architecture Price data Weather data Generation data (CO2) GE prime candidates Platform Data context group • BigData Analysis • Complex Event Processing • Publish/Subscribe Broker Security and Access group • Access Control • Identity Management Optimization  Forecasting  Load steering  Baselining Internet Command signals Command signals BMS Command signals BMS / HEMS Heat loads Heat loads El loads El loads Commercial buildings Possible future scope: - Interface to Meter Data Management - Interface to Distribution management system Residential buildings Decentralized generation (possible future scope) - Large scale PV and Hyllie allocated wind turbines - Batteries Use cases, GEs, APIs Use cases • • • Cost optimization (electricity/heat) by price signals Optimization of demand (electricity/heat) by energy mix signals Instantaneous reduction of energy consumption GE prime candidates • • • • • Big Data Complex Event Processing Publish/Subscribe Broker Access Control Identity Management APIs • • • • getTemperature: This method provides temperature forecast for the Hyllie district over a time interval getPowerPrice: This method provides the Nord Pool power (electricity) price over a time interval getDistrictHeatingPrice: This method provides the district heating price over a time interval getDemand: This method provides the demand on load linked to the trial/demand response over a time interval Big Data in WP1 Purpose • Utilize the Generic Enabler “BigData Analysis” for processing large amount of data in order to validate optimization and find relevant insights concerning patterns and dependencies Accessible data • Consumption • Production • Outside temperature • Sun radiation • Etc. Big Data CUSTOMER DISTRIBUTION PRODUCTION Thank you! Email: [email protected] Telephone: +46(0)702021113 FINESCE WP2 FI for end users of energy ecosystems Thomas Hune (INSERO Software) COMPANY LOGO Goals WP 2 goal • To test the mutual interaction of the technologies as well as the users’ experiences with the technologies. Furthermore, the coherence of the technologies with the entire energy system tested in an area outside the collective district heating. Stenderup WP 2 data Indoor climate Electricity consumption Electricity production EV charging and usage Heat production and consumption • +40 data points per house • Update frequency from 10 sec to 5 min WP 2 data analysis - plans House modelling – forecasting • Forecasting indoor climate and heat consumption • User behaviour – seasonal changes • Electricity consumption – user models EV usage • Driving patterns compared to “normal” car • Charing – battery usage – range Finding and understanding flexibility • Parameters influencing flexibility • How much and when - forecasting Impact of modernisation actions Multivariate big data analyses for consumption and modernisation events Dr. Johannes D. Hengstenberg Elmer Stöwer SEnerCon GmbH 06 14 Who is SEnerCon? Engineers, Programmers, Developers work on • interactive web tools, • web services and • mobile apps Focus on Efficiency of Heating Systems Online Energy Advise for 1 million households per year iESA - 78.000 interactive Energy Savings Accounts • monitor, display, analyse and benchmark meter readings of households Projects: • Web Energy Performance Certificates, • Climate Campaign (last campaign 44.000 Heating reports), • EU-Projects: ECCC, EPLACE, EECC, FINESCE. SEnerCon’s FINESCE Team 67 What is the iESA? iESA: Since 2006 an interactive monitoring and energy savings advice tool for residential energy consumption • 80.000 Users have registered and collect • manually 1,500 meter readings per day • Technical equipment • Base data of buildings Logging and Evaluation of modernisation events • Univariate analysis: saving effects of green modernisation 68 What is the iESA? EAC - Energy Analysis from Consumption / HEMON - Heating Energy Monitor: Heating energy consumption is analysed with regression analysis over outside temperature • Consumption signature of the building • Automatic allocation of • ‘base’ energy for warm water, cooking etc • heating energy • Predicts annual consumption within short time • Extended Benchmarking of heating energy consumption 69 Saving effects of green modernization Replacement of boiler: specific consumption before and after kWh/(m2[AN].a), weather adjusted 2006 - 2012, 24 04 14, N=98 Example Boiler replacement: 20 • 20 kWh per m2 average reduction • Half of modernizations disappoint investors Reduction in kWh/m2 • variation of +- 20 kWh (!) 0 -20 -40 -60 -80 -100 -120 In modernization praxis and theory are not the same -140 0 50 Challenges for our current Analysis • Samples are much too small for multivariate analysis • Algorithms and databases are too slow for upscaling 200 250 300 20 0 -20 Reduction in kWh/m2 • Low motivation of users to provide basic & event data 150 Replacement of boiler with solar heating: reduction specific consumption in kWh/(m2[AN].a), weather adjusted 2006 - 2012, 24 04 14 • Buildings data not fully structured and centralized • Data is fragmented and incomplete 100 Specific consumption before replacement -40 -60 -80 -100 -120 -140 0 50 100 150 200 250 300 Consumption before replacement 70 With big data analysis and FI-WARE we can have… Permanent comparisons of consumption before and after modernizations • Run multivariate analysis of all combinations of green modernization measurements Where should our money go? • Run cohort analyses to categorize building types, age, localizations etc • Create consumption patterns based on geographical distribution • Predict chances and returns of investments in single or combined modernization measures • Predict economic and ecological impact of subsidies and building regulation 71 Thank you! Elmer Stöwer SEnerCon GmbH Hochkirchstraße 11 · 10829 Berlin Telefon 030/ 76 76 85 - 0 Fax 030/ 76 76 85 - 11 [email protected] www.senercon.de, www.energiesparkonto.de, www.heizspiegel.de, www.enerplace.eu