Energy Management Systems Annabelle Pratt Power Research Engineer - - PowerPoint PPT Presentation

Energy Management Systems

Annabelle Pratt

Power Research Engineer Energy Systems Research, Intel Labs annabelle.pratt@intel.com

Contents

• Project goal
• Device-level management, e.g. plug-in electric vehicle
• Building-level management, e.g. home
• Collective level management, e.g. EV charging aggregator
• Collaboration opportunities

Proj roject ect go goal l

Our research aims to develop Energy Management Systems which shape the power demand of devices, buildings and collections of buildings in order to benefit individual consumers by minimizing their energy cost, and society at large by enabling efficient, reliable Smart Grids with significant renewable generation.

Multi-level management approach

• Device level (Device controller)

– Only in smart loads/DR, e.g. EV – Single device optimization (quadratic)

• Building level (HEMS/BEMS)

– Optimization of several devices – Multi-objective (search)

• Collective level (Aggregator)

– Optimization across collection of buildings & shared resources – Linear opt/multi-agent system?

Building automation level Device level

Collective level

Aggregator Aggregator Aggregator HEMS BEMS Device Ctrl Device Ctrl DR Agent Device Ctrl

*DR = Distributed Resource

Utility control center HEMS

Multi-level modeling tools

• Simulation platform in Matlab and Simulink being developed in

collaboration with the University of Colorado

– Short (sec/min) and long (hr/day) time scales – Microgrid can operate off-grid (islanded) and grid-tied

– ensure seamless disconnection & re-connection

Microgrid EMS Microgrid

Multi-level examples and results

• Device-level : plug-in electric vehicles (PEVs)

– significant and potentially intelligent loads

• Building-level : Home Energy Management System

– targeting next generation HEMS products

• Collective level : EV charging aggregator

– at early stages, with preliminary results

Demo at http://www.intel.com/embedded/energy/homeenergy/demo/index.html

Five residences charging EVs

Consumers billed based on time of use electricity pricing Simple timer to delay start time of charging

7 Collaboration with University

• f Colorado, Boulder

Utility Control Center Electric Vehicle Supply Equipment (EVSE)

Smart Meter

Ptxf

EV#1 EV#2 EV#3 EV#4 EV#5 Distribution transformer AMI

EVSE

Advanced Metering Infrastructure

Collective Building Device

Simple time delay

 Charges during minimum load period

8

Total power through transformer (Ptxf) Contribution of PEVs to Ptxf Note time scale : time plotted from noon

• n Day 1 through noon on Day 2

Collective Building Device

Smart charging of individual EVs

Electricity cost profile provided to vehicle Charging App in vehicle determines charging profile

9 Collaboration with University

• f Colorado, Boulder

Utility Control Center

Smart Meter

Ptxf

EV#1 EV#2 EV#3 EV#4 EV#5 Distribution transformer

EVSE

Home Energy Mgmt System (HEMS) AMI IP Internet Protocol

Collective Building Device

10

Intelligent vehicle optimizer

 Minimizes cost and charging rate of PEV

Collective Building Device

Simple time delay

 Charges during minimum load period

Home Energy Management System

Electricity cost profile provided to the HEMS HEMS determines optimal setpoints for all controllable appliances

11

Utility Control Center Smart Meter

Ptxf

EV#1

Distribution transformer

EVSE HEMS

AMI IP

Internet Protocol

Collective Building Device Fixed Load Min and max temps Max power, charge/discharge; P=0 when not plugged in; SOC=100% by target time Earliest and Latest start times; stove to start before dryer Power & ambient sensors

• 8.5% cost savings; balanced with user comfort

– EV charging not lower cost, but grid-friendly

Preliminary results : Single optimization result

5 10 15 20 25 10 20 30 Temp [degC] Initial guess Tstpt0 Troom Tout 5 10 15 20 25

• 5

5 10 PEV0 Pdryer Pstove Pavg-th Pfixed 5 10 15 20 25 10 20 Time [h] price [c/kWh] cum energy cost [$] 5 10 15 20 25 10 20 30 Temp [degC] Optimized results Tstpt Troom Tout 5 10 15 20 25

• 5

5 10 PEV PEV0-opt Pdryer Pstove Pavg-th Pfixed 5 10 15 20 25 10 20 Time [h] price [c/kWh] cum energy cost [$]

Collective Building Device

5 10 15 20 25 10 20 30 Temp [degC] Initial guess 5 10 15 20 25

• 5

5 PEV0 Pdryer Pstove Pavg-th Pfixed 5 10 15 20 25 10 20 Time [h] price [c/kWh] cum energy cost [$] 5 10 15 20 25 10 20 30 Temp [degC] Optimized results 5 10 15 20 25

• 5

5 5 10 15 20 25 10 20 Time [h] price [c/kWh] cum energy cost [$] PEV Pdryer Pstove Pavg-th Pfixed

Preliminary results : V2G enabled / home storage

$4.95/day 27%

• 27% energy cost savings with V2G enabled  higher load at night
• not direct comparison with previous

Aggregator Building Device

Proposed Aggregator

• Coordinates with all the HEMS/BEMS.
• May be implemented on a local device or as a cloud service
• Example functions :

– Determining the optimal solution for a collection of buildings. Most applicable to a campus with a single building owner – Interacting with the utility-issued demand response requests – Maximizing run-time when operating off-grid, e.g. for a microgrid. – Protecting local infrastructure (distribution transformer) through adjustment of local electricity price

Collective Building Device

EV charging aggregator

Aggregator sets local electricity price HEMS not yet included

15 Collaboration with University

• f Colorado, Boulder

Utility Control Center

Smart Meter

Ptxf

EV#1 EV#2 EV#3 EV#4 EV#5 PEV Aggregator Distribution transformer

AMI

EVSE Collective Building Device

With PEV Aggregator

 Limits total PEV power by adjusting local electricity price

16

Intelligent vehicle optimizer only

 Minimizes cost and charging rate of PEV

Collective Building Device

Collaboration opportunities

• Energy Management algorithms

– Optimization engines, load forecasting, thermal models for buildings, user behavior modeling and influencing, etc.

• Prototyping

– Device level: smart charging on PEV emulator and then on PEV – Building level : test HEMS optimization algorithms in a home

– as allowed by controllable appliances available, and EV capabilities

– Collective level : HEMS interaction with utility through aggregator

Electricity price

Device level : PEV

• Demonstrates operation of Charging App in conjunction with

Battery Management System to implement optimal charging

• Implementation on vehicle to follow

AC/DC DC/DC AC EMI filter

120/240 Vac

Gate drive signals

Utility Control Center model

EVSE Load

(emulates traction) Laptop

Battery Management System

Charger Vehicle emulator

IVI System

Charging App

Building level : HEMS

• Demonstrate HEMS algorithms in real (occupied) homes

– to the extent possible, determined by controllable appliance availability and PEV capabilities – First in ESR homes – Then Intel homes – Then external

Smart Meter

EV#1

EVSE HEMS Fixed Load Power & ambient sensors

Artificial price signal & service requests

Collective level : neighborhood

• Detailed simulations
• Field trial with external partner

20

Utility Control Center

Smart Meter

Ptxf

EV#1 EV#2 EV#3 EV#4 EV#5 Aggregator Distribution transformer

AMI

EVSE

Thank You

Please visit the Intelligent management of Electric Vehicles demo Contact me at : annabelle.pratt@intel.com