the use of artificial intelligence in energyefficiency and

16
SMART ENERGY The Use of Artificial Intelligence in Energy Efficiency and Transactions for Industry 4.0 and for Microgrid Management Systems Dr. Vincent SCIANDRA, CEO

Upload: others

Post on 17-Oct-2021

3 views

Category:

Documents


0 download

TRANSCRIPT

S M A R T E N E R G Y

TheUseofArtificial IntelligenceinEnergy Efficiency andTransactionsforIndustry 4.0andforMicrogrid

ManagementSystems

D r . V i n c e n t S C I A N D R A , C E O

About Us

METRON has imposed itself in the last five years as one of

the most innovative a leading tech companies in energy

for the industry. We were one of the first company

promoting that industrial data was paramount for energy

efficiency, thus industry 4.0 would revolutionize the way

we pilot energy in the factory.

Our vision became a reality, and we are proud today to

optimise industrial facilities world wild. Our teams are

composed of industrial experts, energy engineers, Artificial

intelligence researchers and project managers. They are

committed to the energy efficiency of our clients, and we

are proud to give them quantitative results of our work.

Our Company History

EnergyIntelligenceforIndustries©Allrightsreserved 3

AninternationalteamOurapproach

Upto2018 2019

OperationalCenters

§ Paris- HQ§ Milan§ Bogota 2020

70+membersofstaff11nationalities

100+clients

§ SãoPaulo§ Dubaï§ Singapore

1

2

3

We developed a technological platform able to create the

digital twin of any factory, by using existing systems installed

and no additional sensors. Our Artificial Intelligence

modelizes the factory, making it fully transparent in energy.

Once the factory digitalized, we are able to measure the

flexibility of the processes and utilities, making possible to

transact energy on the market or by peer to peer.

By knowing how to optimize the process and utilities and by

being able to forecast the energy consumption, Metron helps

its clients to reduce its carbon footprint by integrating

microgrids technologies as an integrated solution.

Digitalized

Decentralized

Decarbonized

Mission

Our technology enables any factory,

becoming fully transparent in energy.

That are factories able to optimize the full supply chain of

energy thanks to Artificial Intelligence.

Being able to monitor, modelize and forecast any energy

vectors in the industrial sites, Brings new methodologies

to optimize the factory. Real time optimizations adapted

to the real context of production are now possible with

digitalisation.

1. Energy Efficiency

Vision

Energy Markets are evolving to be able to manages

smaller assets of productions and virtual power plants.

New scenarios of peer to peer trading is developing

around the world.

2. Energy Markets

They have changed our way of consuming energy, but

their resiliency is still a complex equation. However AI

can help manage those resources.

3. Distributed Energy Resources

At METRON we believe that the energy of tomorrow

will drastically be managed differently from today.

This revolution is already on its way. We consider that

those changes will be possible by digitalizing each

electrons and molecules traded. We think the success

of that revolution resides at the crossroads of Energy

efficiency, energy markets and distributed energy

resources.

EnergyIntelligenceforIndustries©Allrightsreserved 6

Artificial Intelligence What is AI

AUTOMATEDatacollectionand

cleaning

Acquisition,Qualityandreliabilitytests

Processautomation:Digitalizetheassetsandtheircontext

REASON

Reasoning andknowledge beyond human:Understand theenergy andoperation context andcrossit with knowledge bases

ENGAGE

Proactiveactionandcommunication:NaturalLanguage Processing todirectlyinteract with humans andcontrol-commandtooperate themachine

Dataaggregationandcuration

Aggregation,Storage

Dataenrichment

BaselinesandKPIcalculation,Alarmcreation

Standardprocesses

ISO50001,IPMVP

Dataaggregationandcuration

Predictivemaintenance,Productionandstorageoptimisation,Risk

management

Optimisation

Identificationofoptimisationlevies,Prescriptionofimprovement actionswithROI,Ranking ofinfluential parameters,

Benchmarking,Driftdetection,Predictions

Flexibility

Energyassetsmanagement,Simulationofscenarios

Human collaboration

Proactivepersonalized recommendations,Adaptationtouserchoices andcontext,Ongoing

questionsanswering

Machineinteraction

Real-timecontrolandoptimisation,Knowledgecollectedfromthemachine

EnergyIntelligenceforIndustries©Allrightsreserved 7

Augmented Energy Manager

Follow-up,controlandvalidation

oftheAI’s results andproposals

Continuous improvement

ofAI’s knowledge andperformancesby

continuous feedbackandmanual optimisations

Leading change

byproject managementandcommunicationon

site

Artificial Intelligence

EnergyIntelligenceforIndustries©Allrightsreserved 8

ThebenefitsArtificial Intelligence

5-15%energy savings12months ROI

40%resources energyoptimisation

SAVINGS

Risks andmistakesreduction

RELIABLE

Volumeincreasingmultisitesdeployment

SCALABLE

Contextualadaptationtouser

preferences

ADAPTED

EnergyIntelligenceforIndustries©Allrightsreserved 9

METRON-EVA®VPPVirtualPowerPlantDemandResponseandflexibilityoperationTradingwithenergymarketsandbalancingmechanismsAssetsportfoliomanagement

METRON-EVA®MicrogridMicrogrid Management

Energygenerationandconsumptionforecasts

ManagementandcontrolofDistributedEnergyResourcesand

Ancillaryservices

METRON-EVA®OEMSmartMachinePerformanceandup-timeoptimisationPredictivemaintenanceDecentraliseddecision-makingEmbeddedartificialintelligence

Applications Aplatformservingallactorsusingenergy

METRON-EVA®FactoryEnergyTransparentFactoryReal-timemonitoringandforecastofenergyusagesIdentificationofbestsetofparametersandenergyefficiencyprojectsSmartcontrolofindustrialsystems

EnergyIntelligenceforIndustries©Allrightsreserved 4

Technicaloverview Thedigitalplatform

CloudservicesWebinterface

DataandknowledgeintegrationVisualisation - Exploration- Piloting

EnergyprojectmanagementDataScienceandAItool

OntologiesKnowlege integration

DataScienceMachinelearning

ETL

BigDataUnlimiteddatastorage

ComplexEventProcessing

IIoT

EnergyIntelligenceforIndustries©Allrightsreserved 11

UseCase Processoptimisation:Meltingfurnace/Glasswaremanufacturer

§ Processing of energy and process datacoming from the furnace supervisionsystem (more than 120 variables)

§ Visualisation of a 12% gasconsumption drift in 2 years

§ Modelling of the variables impact onconsumption:

§ Detection of the most influentialvariables, using a ranking algorithm

§ Determination of the variables forwhich the operator has some flexibilityto pilot (crossing of data and businessknowledge): cullet, electric/gas mix,air/gas mix

§ Definition of the optimal configurationfor these parameters according to D+1production planning so as to maintainthe product quality while minimisingenergy cost

§ Creation of a predictive model toforecast gas consumption based onthe 15 most influential variables (errorrate < 3%)

Actionplan

§ 4.5% reduction in the annualenergy consumption of thefurnace (eq. €250,000/year)

§ Daily optimal settings advice providedto operators: continuous optimisationof the furnace operation andreduction of its energy efficiency drift

§ Prediction of future consumption(accuracy > 97%): operational teamsare notified if an abnormal drift isforecasted

Results&benefits

§ Implementation of an automaticcontrol-command system so that theoptimal regulation of influentialparameters is directly integrated intothe supervision via the METRONplatform

Nextsteps

§ Production of glass flasks for thepharmaceutical industry

§ Perimeter: smelting furnace, themost energy-intensive equipment ofthe plant

>100GWhofgasconsumptionperyear

>>Stake:Limitgasconsumptiondriftsandpilotoperational

parametersofthefurnacewithefficiency

Context

EnergyIntelligenceforIndustries©Allrightsreserved 12

UseCase Optimisationofelectricself-generation/PaperIndustry

§ Determination of the energyanalysis methodology and validationby the client’s process knowledgeapproach

§ Collection and analysis of more than500,000 points over 1 year

§ Sensitivity study made to theminute granularity (previously madeto the hour): research of theoptimum offset and calculation ofthe associated economic gain

§ Definition of the optimum offset at125 kW (instead of 500 kWpreviously)

§ Implementation of therecommendation and real-timemonitoring of the turbine activity

Actionplan

§ USD 31,000 of annual savingswith no investment

Results&benefits

§ Continuous re-evaluation of theoptimum with context integration(production cycles, energy prices onthe markets…)

§ Dynamic offset control

Nextsteps

§ Plant producing paper from sugarcane residues

§ Perimeter: backpressure steamturbine producing electricity

§ Variation in the electrical siteconsumption happening to quicklyto allow an ideal balance betweenconsumption and production

§ Production of the turbine dependson the plant power consumption(set point = consumption + offset)

§ One-site self-generated electricitythree times cheaper than the gridone but the surplus of producedelectricity is not valued

180GWh ofannualelectricconsumption

>>Stake:Determinetheoptimaloffsetforoperatingtheturbinesoastoreachtheeconomicoptimum

(betterbalancebetweenelectricityimportsandexports)

Context

EnergyIntelligenceforIndustries©Allrightsreserved 13

Ourapproach microgrid

Optimisemicrogridmanagement

○ Moreaccurateforecastsofenergyproductionandconsumptionoftheassets(basedontheknowledgebases)

○ Real-timebestdecisionmakingbetweenstorage,produceandconsume(off-grid),andsell/buyfromthegrid(on-grid)

○ Dynamicoptimisationandpredictivemaintenance:theevolutionoftheassetsperformanceisintegratedintothealgorithms+predictionofthemomentwhentheprofitabilitythresholdwillnotbeachievedanymore,allowingtotriggeramaintenanceaction.

EnergyIntelligenceforIndustries©Allrightsreserved 14

REIDSProject- SingaporeMETRON-Microgrid

Our Contact

+ 3 3 6 2 8 8 2 0 9 1 7

A D D R E S S102 rue Réaumur, 75002, Paris

P H O N E

V i n c e n t . s c i a n d r a @ m e t r o n l a b . c o m

E M A I L