iMinds the conference 2012: Koen Vanthourhout

Koen Vanthournout, VITO (Flemish Institute for Technological Research)

Jan Van Ooteghem, Ghent University / iMinds

208/11/12

LINEAR

“Local Intelligent Networks and Energy Active Regions”

1. Flemish Government decision ‘dd. 8 mei 2009 VR 2009 0805 DOC.0671’

2. Official start: May 2009. Project runs until December 2014.

3. Total budget of 40M€: – Flemish government: 10M€– Industry: 30M€ (personnel, hardware, ...)

www.linear-smartgrid.be

308/11/12

Linear situated

• Increased share of intermittent renewables• Electrification of transport and heating• Stabilization/decrease classical production portfolio

⇒ Energy excesses and shortages⇒ Grid capacity issues

⇒ storage⇒ “consumption follows production” (i.e., demand response)

Classical control paradigm“production follows consumption”

no longer holds

Linear focuses onResidential Demand Response

408/11/12

Research Partners

Industrial Partners

Additional Members Steering Committee

Flemish Government

Partners

508/11/12

Core: large scale residential DR pilot

1. Focus on automated ‘smart’ devices.

2. Keep technical targets and BCs hidden for the user.

3. Keep interaction simple and stimulate good behavior financially.

Residential demand response with as little impact on the user comfort as possible

608/11/12

Linear Smart Devices

• Washing machines• Dishwashers• Tumble dryers• Electric hot water buffers• (Electrical vehicles)• (Heat pumps)

708/11/12

ServerFifthplay

TouchDisplay

Energy meas.

PlugsNetwork

…

Linear Server

VITO DHW buffer

GW GW GW GW

Mieledish

washer

BSH dish

washer

Mielewash.

machine

Mieletumble dryer

BSH wash mach

VITO controller

MieleGW

Smart Meter

Transformer measurements Historic

BC data

EV

Linear Control architecture

• One business case at the time • # aggregators with # control technology• Generic interfaces for smart devices• P1 smart meter interface definition• Interface definition between aggregator and user flexibility

Integration tests in lab ongoing, tests started at friendly users

Smart plug

GWINTERNETHOME

808/11/12

250 participating families

3 groups Friendly users Geographically dispersed

Geographically concentrated

Focus Technical testing BRP cases DSO cases, link with smart meter projects

Target 30 100 120Contract signed 33 81 13

Smart devices installed

25 203 25

Status Integration tests started

• Installed• Reference measurements ongoing• 15 HP owners in recruitment pipeline

Recruitment and installation ongoing in Hombeek& Leest (Eandis) and Genk (Infrax)

908/11/12

Use DEMAND RESPONSE for:

1. Portfolio Management Shift consumption in function of the day ahead market (corrected for 2020

production mix)

2. (Wind) Intraday balancing Shift consumption in function of the quarter hour unbalance of a BRP portfolio

3. Transformer aging Shift consumption to avoid accelerated aging due to high transformer temperatures

4. Voltage control for LV distribution feeders Shift consumption based on local voltage measurements to ameliorate voltage line

deviation

BRP/Retailer

BRP/Retailer

DSO

DSO

Linear Business Cases

1008/11/12

Linear - business cases

LV transformer loadFuture distribution grid:• More residential consumption

=> higher offtake peaks on LV transformers• More intermittent (renewable) generation

=> injection peaks on LV transformersÞ Aging of transformer will be accelerated during

these peak periods (increase of temperature)Þ If DSOs succeed in controlling temperature of LV transformer,

they are able to allow a higher share of local generation and extra demand without the need for new network investments

Objective: control the temperature on a transformer by the application of automated active demand to defer/decrease network (transformer) investments

Lifetime

Peakload

P1

Peakloadreduction due to DR

Deferrel ofInvestment

t1 t2

P2

1108/11/12

Linear - business cases

LV feeder voltage profileFuture distribution grid:• More residential consumption

=> voltage drops in feeders• More intermittent (renewable) generation

=> voltage increases in feedersÞ Predictability of the voltage profile of a feeder

decreasesÞ Situations can occur where the voltage levels

either become too high or too low

Objective: keep the voltage profile of the different LV lines within permitted limits by the application of automated AD (no coordinated local control at household level) to defer/decrease network (new lines / transformers) investments

Voltage profile with and without DG (Source: EELAB)

1208/11/12

Linear - business cases

Portfolio managementDynamic tariffs can be used to incentivize consumers to reschedule their consumptionTariffs used:• 6 time blocks a day• Based on Belpex DAM +

prediction sun/wind (scenario 2020)

• Communicated DA for the next day

Objective: coincide as much as possible periods with low generation costs (based on high availability of renewable generation by wind and/or sun) with consumption by applying dynamic tariffs to residential consumers

Dynamic Energy componentDistribution fee

Fixed Transmission feeCharges

1308/11/12

Linear - business cases

Wind balancing• BRP = responsible for the balance of

injection + offtake in their perimeter on a quarter-hourly basis

• TSO manages the instantaneous imbalances (that the BRPs are not able to control), by making use of power reserves

Þ Costs involved for making use of power reserves are transferred to the BRPs

Þ BRP with wind production in its portfolio: higher imbalance risk + higher imbalance cost

Objective: decrease the imbalance cost of the BRP by decreasing the unpredicted imbalance caused by wind production (intraday) in its portfolio by aggregating flexibility available at residential consumers’ premises (flexible cluster)

Balancing mechanism(Source: Elia)

1408/11/12

Linear - business cases

Control mechanism

DSO

LV transformer

load

LV feeder voltage profile

BRP/retailer

Portfolio management

Wind balancing

Automated active demand

Dynamic tariffs

Questions?Koen Vanthournout, VITO (Flemish Institute for Technological Research)

Jan Van Ooteghem, Ghent University / iMinds