smarthor wim cardinaels - energyville€¦ · gas dc grids inverter battery charger wind pv...

SmarThorTowards Energy as a Service

Wim Cardinaels

SmarThor

Towards Energy-as-a-Service

Tourism

MobilityRetail

Business as Usual

Communication

SmarThor

Towards Energy-as-a-Service

Energy

SmarThor

Multi-energy optimisation

AC

HCG

gas

DC

grids

inverter battery

charger

wind pv batteries

geothermal

HP

ORCCHPstorage PCM

II000III0

0II0IIIII0

III0

0II000III0

IIIII0

00IIIIIIIIIII0

00I0

I0I0

00III0

0II000III0

IIIII0

00IIIIIIIIIII0

00II000

III0

0II0IIIII0

III0

0II000III0

IIIII0

00IIIIIIIIIII0

00I0

I0I0

00III0

0II000III0

IIIII0

00IIIIIIIIIII0

0011011100

boiler

ICTservices

LaaS(light)

EVCaaS(Electric Vehicle

Charging)

HaaS(Heating)

CaaS(Cooling)

HWaaS(Hot Water)

WaaS(Washing)

SmarThor

Preparing a Living Lab

P4

MoThorMoThor

MoThor

C4

Central

IncubaT

T2

EV1EANEAN

EAN

EANEAN

EAN

vEAN

EV2EAN

Building

Mgt Syst

sub

metering

PV

monitoring

markets

balance

Meteo

forecast

SmarThor

Aggregator

meteo

station

lab

server

DSO

IoT services

TSO

RTDS

Electric Mobility as a Service

• 4G district heating & cooling grids, storage, …

• Monitoring & Controls

• Regulation

• Envisioning the Future Pieter Valkering & Erik Laes

• Multi-carrier Energy MarketsKris Kessels

• SmarThor Data PlatformKlaas Thoelen

SmarThor

Agenda

EnvisioningOpportunities and barriers for multi-energy services in

the future Flemish energy landscape

Pieter ValkeringErik Laes

• Explore the role of multi-energy services in the future energy system, given various uncertain developments

• Identify main opportunities and barriers

• Facilitate dialogue and open innovation

SmarThor

Objective

• Horizon scanning

• factors influencing the evolution of the Flemish energy system towards 2030

• major uncertainties concerning these factors

• Development of four future visions on the Flemish energy system

• scenario-axis technique

• Interviews with key players in the innovation system

• barriers and opportunities for multi-energy innovation

• Innovation potential for the Thor site and beyond

SmarThor

Methodology

SmarThor

Energy Visions

Trends & uncertainties:

• Energy policy

• Energy demand side

• Energy supply

• System integration

• Energy users

SmarThor

Multi-energy services: Today

Selling kWhrs

Feed-in PV electricity Local use of waste heat

Regional wind cooperatives

SmarThor

Multi-energy services: Tomorrow

Selling kWhrs

Heating as a service

Feed-in PV electricity

Trading individual flexibility

Self-sufficiency

Local use of waste heat

Local multi-energy communities

Regional wind cooperatives

Community Virtual Power Plants

Geothermal heat

Interviews Helicopter viewers

Opportunities:

• Individual technologies largelydeveloped → challenge is integration

• Positive attitudes among policy andusers

• Regulatory environment adapts (e.g. heat networks)

• In some case, viable business cases exist (e.g. agregation residential flex)

Barriers:

• Perception and awareness among residential / industrial users

• Difficult business case (e.g. low gas price)

• Regulatory barriers (e.g. obligatory individual connection points, direct DC lines)

• Financing, uncertainties in technology and policy

SmarThor

Opportunities and barriers

Advocates:

• Locally optimal solutions

• Lower connection capacity to the distribution grid

• Especially under high electrification

• Attractive for the socially engaged

• Development potential in new district development

Sceptics:

• Slow decision-making in a community context

• Privileged households profit most (‘Mattheus effect’)

• Distributed flexibility services and P2P trading more efficient?

• Difficult to reconcile with free choice of energy provider.

Where are we heading?

Example of Smart Energy Districts

• Enabling:

• Removal of regulatory barriers, alternative pricing schemes, …

• Benefit from upscaling effects

• Innovation:

• Exploring business cases for smart energy district solutions

• Exploring what motivates people/businesses to engage

• Developing standardized planning tools, control algorithms, and platforms

• Dialogue:

• Explore and quantify potential for multi-energy solutions in Flanders

SmarThor

Conclusions: What is needed?




SmarThor

Agenda

Multi-Carrier Energy MarketsA Multi-Energy System Calls for Multi-Energy Market

Models

Kris Kessels

Multi-Carrier Energy Markets

Context• Drivers multi-carrier energy systems

• Climate goals: low-carbon supply

• Increased need for flexibility

• (Local) energy independence

• Technological developments

• Move toward energy services

• Objective• “Develop a multi-energy market platform so that the interaction between the energy

carriers on a given level (building, industrial site, district, city, region, country,…) is optimal in terms of a certain objective.”


Design challenges

Physical Market Regulation

• Coupling of carriers (conversion technologies)

• Different scales (local vs. global)

• Different networks(network constraints!)

• …

• Different trading times / practices

(coupling!)• Compatibility with fut.

> DA > ID > RT• Energy as a service• …

• Different regulations for different carriers

• Access to markets• Free trade within MES• Smart metering, settlement,

(flexible) tariffs, etc.• …

Compatible with (expected evolutions of) regulationFocus on coupling of carriers and day-ahead market


Conversion technologies and storage

• Single-carrier (i.e. existing order types)

• Block orders (consecutive time steps)• Linked block orders, exclusive block orders, flexible hourly orders

• Multi- carrier (i.e. new order types)

• Linked multi-carrier orders

• Exclusive multi-carrier orders

• Dependent multi-carrier orders

• Conversion multi-carrier orders

• Storage orders


Multi-carrier order types time

(P,Q)carrier

P

Q

A

ELECTRICITY

P

Q

1

GAS

e.g. gas turbine

demand

supply

• Two designs

• Two-step approach: a local electricity-gas-heat market before a national electricity and gas market

• Integrated approach: a single national electricity-gas-heat market

• Case study

• Belgian national gas and electricity markets, and a single heat market of the size of the Thor campus

• Reference scenario: sequential day-ahead market clearing (heat > electricity > gas)


Implementation of multi-carrier energy market


Case study: two-step vs. integratedOption 2 chosen

• A simultaneous day-ahead multi-carrier market is able to increase the social welfare compared to a series of sequential day-ahead markets as:

• It eliminates the need for price forecasting, and therefore also the associated errors that occur in sequential markets

• It is able to use the flexibility in one energy carrier to clear another one through multi-carrier technologies

• It allows for a type of order acceptance configurations that cannot be realized in a sequential market set-up


Results




SmarThor

Agenda

ICT: Platform, Forecasting, Smart ChargingSmarThor Data Platform

Klaas Thoelen

Internal dataExternal data

• A single Web interface for often used data at EnergyVille

SmarThor Data Platform

A Central One-stop Shop for Energy Data

Energy Markets

Power Generation

Forecast

Weather Forecasts

& Observations

Building Management

Systems

Local Energy Production

& Consumption

EV Charging Stations

PV Production Forecast

• Accelerate data-driven algorithmic research at EnergyVille

• Facilitate access to data, increase TRL of research deliverables

• Repository for multi-year, multi-feature data sets

• Gain insight into the energetic operation of the EnergyVille 1 building and other buildings at Thor Park, e.g.:

• Integrate with Building Management Systems

• Monitor renewable energy production

• Reusable ICT infrastructure for (future) research and cooperation with project partners


Goals


RelationalDB

TableStorage

D

a

t

a

A

P

I

Application Environment

Project Z.1

Project Z.2

Project X

Project Y

Thor Park

Monitoring & Data ViewersOperations & Billing

Captor 1

Captor 3

Captor 4

Real-time

Control

WebProject

DB

P

r

o

j

e

c

t

A

P

IOther

Captor 2

• SmarThor project: from 7 to 25 charging points

• EV charging capacity > 500kW

• > capacity of electrical panel of the parking

• > half of the grid connection capacity of the building

• Thus: need for smart charging to reduce peak grid consumption

• Main objective for optimization: increase self-consumption of PV production


Use-case: Smart Charging @ EnergyVille


EnergyVille 1: Building Consumption vs. PV Production

• 369kW PV installation at EnergyVille 1

• Total production in 2017: 228MWh

• Total grid injection in 2017: 63MWh

PV Production 2017

Hours

in

a d

ay

Grid Injection 2017

Days in 2017

Days in 2017

Hours

in

a d

ay

• Precisely predict the PV production at EnergyVille 1

• Next day, hourly resolution

• Optimize charging of electrical vehicles and heating at EnergyVille 1


PV Production Forecaster

• Data-driven expert system:

1. Historical PV production data

2. Irradiance predictions for nearby sites

3. Numerical weather predictions

Empowered by: KU Leuven, VITO, imec & UHasselt

SmarThor PV ForecasterGeneric PV Forecasting, Based on Historical/Weather

Data

Description

Scope:

• Data-driven, generic PV Forecasting module, based on ‘black-box’ machine learning approach.

• No custom installation or configuration required. Only requirement: access to historical and/or weather data. (This is

facilitated by the SmarThor data platform.)

Purpose:

• Universal solution, that is not tailored to specific hardware.

• Implicitly take into account impact of specific local circumstances (shadow, azimuth, gradient) and panel-specific properties

(kWph, maximal inverter output, …)

• Useful for smart homes and/or larger installations. Test-Case EnergyVille: PV power predictions for a 24-hour window,

which serve as a guideline for smart charging of electrical vehicles.)

Technical Features:

• Built in Python, on top of in-house forecasting library, extending existing machine learning libraries (such as scikit- learn

and keras).

• Expert System, weighing and combining predictions from three different sources:

• Immediate future forecast, based on endogenous data: look at recent history, and make future prediction based on

window of previous time values. [Blue]

• One-on-one mapping from closest-by weather station predicting irradiation data. [Red]

• Forecast based on weather data: train on historical weather measurements + PV power generation, then use predictions

for those weather variables to generate PV power forecast. [Green]

• For each data source, one or more regressors of choice provide us with an estimate of the next PV value. These values are

then used by the expert system to make a final prediction.

Weather Feature #1 ### ### ### ### ### ### ### ###

Weather Feature #2 ### ### ### ### ### ### ### ###

Weather Feature #3 ### ### ### ### ### ### ### ###

Weather Feature #4 ### ### ### ### ### ### ### ###

Weather Feature #5 ### ### ### ### ### ### ### ###

PV Power (Invertor #1) 4000 4500 5100 5650 6300 7100 7250 ?Nearby Irrad. Forecasts ### ### ### ### ### ### ### ###

Timestep #1 Timestep #2 Timestep #3 Timestep #4 Timestep #5 Timestep #6 Timestep #7 Timestep #8


Posters & Demos

Questions?

[email protected]

Envisioning the Future [email protected]

[email protected]

Multi-carrier Energy [email protected]

SmarThor Data [email protected]

Welcome to our poster sessions in the labs !

Take your belongings with you

We do not come back to Thor Central

Eager to find out more?The scientific publications developed during the project can

be found using the QR-code on the posters

smarthor wim cardinaels - energyville€¦ · gas dc grids inverter battery charger wind pv...

Documents