S MART C AP : Flattening Peak Electricity Demand in Smart Homes - - PowerPoint PPT Presentation

▶
s mart c ap
SMART_READER_LITE
LIVE PREVIEW

S MART C AP : Flattening Peak Electricity Demand in Smart Homes - - PowerPoint PPT Presentation

Jul 09, 2023 351 likes •582 views

S MART C AP : Flattening Peak Electricity Demand in Smart Homes Sean Barker , Aditya Mishra, David Irwin, Prashant Shenoy, and Jeannie Albrecht University of Massachusetts Amherst Williams College Department of Computer Science

slide-1
SLIDE 1

Department of Computer Science

SMARTCAP:

Flattening Peak Electricity Demand in Smart Homes

Sean Barker, Aditya Mishra, David Irwin, Prashant Shenoy, and Jeannie Albrecht†

University of Massachusetts Amherst Williams College†

slide-2
SLIDE 2

Sean Barker (sbarker@cs.umass.edu)

Pervasive Computing in Smart Homes

! Smart homes: efficiency, automation, convenience ! Enabled by pervasive computing

  • Smart meters, energy sensors, load controllers
  • Appliance integration

! Greening smart homes

  • Why? 73% of U.S. electricity

! Economic benefits:

  • infrastructure, energy costs

! Environmental benefits:

  • carbon footprint, renewables

2

slide-3
SLIDE 3

Sean Barker (sbarker@cs.umass.edu)

Demand-Side Energy Management

! Control consumer-side energy demand

  • Respond to energy availability
  • Reduce peak usage, fluctuations

! Components of DSEM

  • Monitoring (data collection)
  • Control (e.g., load shifting)

3

Peak Usage Off-Peak

Power

Time

Heater Dryer Light

Shiftable Load

! We focus on performing peak load reduction

slide-4
SLIDE 4

! For utilities:

  • Lowered peak grid demand
  • Infrastructure savings
  • Transmission & distribution

Sean Barker (sbarker@cs.umass.edu)

Benefits of Peak Load Reduction

4

(loss ∝ current2)

47%!

! For consumers:

  • Variable pricing cost savings
  • Battery efficiency
  • Assist with capping
slide-5
SLIDE 5

! Change user behavior

  • Users don’t want to!
  • Maintain household routines

! Inflexible loads

  • Unacceptable: lights, TV
  • Inconvenient: dishwasher

! Goal: transparent peak reduction

  • No user cooperation
  • No negative impact

Sean Barker (sbarker@cs.umass.edu)

Challenges of Peak Load Reduction

5

Lights off until 9 pm!

slide-6
SLIDE 6

Sean Barker (sbarker@cs.umass.edu)

Outline

! Motivation ! Home measurement study ! SmartCap scheduler ! Evaluation on home data ! Conclusions

6

  • What can we schedule transparently?
  • How can we perform transparent peak reduction?
  • How effectively does SmartCap flatten demand?
slide-7
SLIDE 7

! Interactive loads

  • Controlled by users
  • TV, lights, microwave
  • Little scheduling freedom!

! Background loads

  • Not controlled by users
  • A/C, refrigerator, heater
  • Don’t care how objective is met
  • Significant scheduling freedom!

Sean Barker (sbarker@cs.umass.edu)

Types of Loads

7

Not OK to interfere! OK to change

  • n/off times!
slide-8
SLIDE 8

Sean Barker (sbarker@cs.umass.edu)

SmartCap Measurement Study

! Home monitoring deployment

  • 3 occupants, one year (so far)

! Instrumented all outlets and switches

  • Lights, TV, dishwasher, freezer, A/C, etc.

8

SmartCap Gateway grid power renewables

Outlet/Switch Meters

Panel Meter power readings Programmable Switches

Appliances

commands power readings

slide-9
SLIDE 9

! Few major background loads

  • heat recovery ventilator
  • refrigerator
  • freezer
  • dehumidifier
  • air conditioner (x3)

! 8% of loads but 59% of energy use

Sean Barker (sbarker@cs.umass.edu)

  • 1. Background Loads are Significant

9

Load Peak Average Quantity Refrigerator 456W 74W 1 Freezer 437W 82W 1 HRV 1129W 24W 1 Dehumidifier 505W 371W 1 Main A/C 1046W 305W 1 Bedroom A/C 1 571W 280W 1 Bedroom A/C 2 571W 141W 1 Background 4715W 1277W 7 Interactive 9963W 887W 85

! Background loads are few but major energy users

slide-10
SLIDE 10

50 100 150 200 250 300 350 400 450 500 5 am 9 am 1 pm 5 pm 9 pm 1 am 5 am 20 40 60 80 100 120 140 5 am 9 am 1 pm 5 pm 9 pm 1 am 5 am 50 100 150 200 250 300 350 400 450 5 am 9 am 1 pm 5 pm 9 pm 1 am 5 am 20 40 60 80 100 120 140 160 5 am 9 am 1 pm 5 pm 9 pm 1 am 5 am

refrigerator freezer dehumidifier HRV

Sean Barker (sbarker@cs.umass.edu)

  • 2. Background Loads are Periodic

10

! Periodicity: regular on/off intervals ! Mostly (but not fully) independent of user behavior ! Background loads vary but have useful periodicity

unusual event

irregular intervals

regular intervals

slide-11
SLIDE 11

50 100 150 200 5 am 9 am 1 pm 5 pm 9 pm 1 am 5 am

coffeepot

10 20 30 40 50 60 70 80 90 100 5 am 9 am 1 pm 5 pm 9 pm 1 am 5 am

entertainment center

2 4 6 8 10 12 14 16 18 5 am 9 am 1 pm 5 pm 9 pm 1 am 5 am

lamp television

200 400 600 800 1000 1200 5 am 9 am 1 pm 5 pm 9 pm 1 am 5 am

500 1000 1500 2000 2500 3000 5 am 9 am 1 pm 5 pm 9 pm 1 am 5 am

Power (watts) Time

interactive mealtime peaks

! “Peaky” total load

  • Brief, high-power devices

! Many unpredictable individual loads

  • Human usage patterns
  • May change over time

Sean Barker (sbarker@cs.umass.edu)

  • 3. Interactive Loads are Unpredictable

11

! Peak reduction must compensate for interactive loads

slide-12
SLIDE 12

! Background loads cycle on and off ! Exact on/off times (mostly) don’t matter ! Schedule cycling of background loads

Sean Barker (sbarker@cs.umass.edu)

Flattening in SmartCap

12

! Interleave background loads to flatten peaks

power peak = 1000W A/C 1

(b) with scheduling

A/C 2 A/C 3

  • ne hour period
  • ne hour period

power peak = 3000W A/C 3 A/C 2 A/C 1

(a) no scheduling

slide-13
SLIDE 13

! Periodic background loads exhibit ‘slack’

  • Measure of how long background load can remain off
  • Based on guardband (e.g., fridge temperature range)

Sean Barker (sbarker@cs.umass.edu)

Scheduling Loads: Slack

13

20 40 60 80 100 120 140 160 37 37.5 38 38.5 39 39.5 40

Power (watts) Temperature (F) Time (6 hours)

Power Temperature

Passive warming: consuming slack

Active cooling: accumulating slack

! Control slack by modifying device duty cycle

slide-14
SLIDE 14

power peak = 2000W A/C 1 A/C 2 A/C 3

  • ne hour period

interactive loads

(c) offline scheduling

! Schedule based on remaining slack ! Least Slack First (LSF)

  • Operate loads in order of ascending slack

! Online scheduler

  • Respond to foreground (interactive) loads

Sean Barker (sbarker@cs.umass.edu)

SmartCap Scheduler

14

! Preempt background loads by interactive loads

power peak = 1000W

(d) online scheduling

  • ne hour period

interactive loads

slide-15
SLIDE 15

! Evaluate LSF on home data

  • Computed per-period slack
  • Linear slack model

! Flattening metric

  • Average deviation from mean

! Flattening period

(a) One day (b) Four hours

! High deviation period (mealtimes) or low (nights)

Sean Barker (sbarker@cs.umass.edu)

SmartCap Evaluation

15

20 40 60 80 100 120 140 160 37 37.5 38 38.5 39 39.5 40

Power (watts) Temperature (F) Time (6 hours)

Power Temperature

Nighttime Daytime

slide-16
SLIDE 16

! Day-long periods

  • Flattening on 91% of days
  • 16% average flattening

! Four-hour periods

  • High variance (31%)
  • >20% flattening
  • Low variance (69%)
  • <3% flattening

Sean Barker (sbarker@cs.umass.edu)

Evaluation: Smart Home

16

  • 10

10 20 30 40 50 60 70 10 20 30 40 50 60 70 80

% Deviation Decrease Days

% LSF Improvement No Improvement

(a) Each Day Up to 50% improvement

  • 5

5 10 15 20 25 5 10 15 20 25 30 35 40

% Deviation Decrease 4-hour Periods

% LSF Improvement No Improvement

>1kW Variance

! LSF especially good at flattening high peak periods

Untimely interactive loads

slide-17
SLIDE 17

! Electric Vehicle (EV)

  • Peaky typical usage
  • Grid unreadiness at scale
  • EV load added to home data

Sean Barker (sbarker@cs.umass.edu)

Evaluation: Electric Vehicle

17

Peak power periods reduced

1000 2000 3000 4000 5000 6000 7000 10 20 30 40 50 60 70 80 90 100

Power (watts) Percentage of Time

No Scheduling LSF (3.0kW)

! EV is good candidate for LSF: 22% flattening

slide-18
SLIDE 18

! Live testbed for active control, repeatability

  • ‘Smart appliances’ via programmable Insteon switches

! Active background scheduling in LSF

Sean Barker (sbarker@cs.umass.edu)

Evaluation: Lab Testbed

18

! 23% flattening (background + interactive) on testbed

500 1000 1500 2000 2500 3000 3500

Aggregate Power (watts) Time (4 hours) No Scheduling LSF

Background

Background loads pushed forward

slide-19
SLIDE 19

Sean Barker (sbarker@cs.umass.edu)

Related Work

! Demand-side energy management

  • Load shifting [Keshav, GreenNet 10]
  • Prediction [Schülke, SmartGridComm 10]
  • Batteries [Zhu, BuildSys 11], [Bar-Noy, WEA 08]

! Background schedulers

  • Optimizing for renewables [Taneja, SmartGridComm 10]
  • Offline scheduling [Bakker, SmartGridComm 10]

19

slide-20
SLIDE 20

Sean Barker (sbarker@cs.umass.edu)

Conclusions

! Demand-side energy management for peak reduction ! SmartCap flattens transparently

  • Modifies only background loads
  • Interactive loads unaffected

! 20-30% flattening using Least Slack First ! Additional savings possible with modest user changes

  • Subject of ongoing work

20

slide-21
SLIDE 21

Department of Computer Science

Questions?

Sean Barker sbarker@cs.umass.edu

slide-22
SLIDE 22

! Threshold power to preempt loads

  • Start scheduling when threshold is reached

! Adaptive threshold (moving average of past use)

Sean Barker (sbarker@cs.umass.edu)

Least Slack First Threshold

22

1000 2000 3000 4000 5000 9 am 8 am 7 am 6 am

Power (watts) Time (Hours)

No Scheduling LSF (2.2kW) A/C's 1, 2, 3 activate Stacking triggers scheduling