Nipun Batra, Amarjeet Singh, Kamin Whitehouse14 May 2016

Exploring The Value of Energy Disaggregation

through actionable feedback

1

General eco feedback vs Actionable Feedback

Eco feedback

Misc.22%

Light10%

Fridge11%

HVAC56%

General eco feedback vs Actionable Feedback

Eco feedback

Misc.22%

Light10%

Fridge11%

HVAC56%

General eco feedback vs Actionable Feedback

Eco feedback

Misc.22%

Light10%

Fridge11%

HVAC56%

Pow

er (

W)

0175350525700

Pow

er (

W)

0175350525700

Home 1

Home 2

Actionable feedback

Fridge consumption over 24 hours

Misc.22%

Light10%

Fridge11%

HVAC56%

General eco feedback vs Actionable Feedback

Eco feedback

Misc.22%

Light10%

Fridge11%

HVAC56%

Pow

er (

W)

0175350525700

Pow

er (

W)

0175350525700

Home 1

Home 2

Actionable feedback

Fridge consumption over 24 hours

Misc.22%

Light10%

Fridge11%

HVAC56%

High power state

General eco feedback vs Actionable Feedback

Eco feedback

Misc.22%

Light10%

Fridge11%

HVAC56%

Pow

er (

W)

0175350525700

Pow

er (

W)

0175350525700

Home 1

Home 2

Actionable feedback

Fridge consumption over 24 hours

Misc.22%

Light10%

Fridge11%

HVAC56%

High power state

High power state

General eco feedback vs Actionable Feedback

Ecofeedback

Misc.22%

Light10%

Fridge11%

HVAC56%

Pow

er (

W)

0175350525700 Home 2

Actionable feedback

Fridge consumption over 24 hours

Your fridge defrosts too much, wasting 30% energy

Misc.22%

Light10%

Fridge11%

HVAC56%

Approach overview- How to give feedbackPo

wer

(W

)

0

175

350

525

700

Specific features of trace to infer why energy usage is high

Length of duty cycles

Approach overview- How to give feedbackPo

wer

(W

)

0

175

350

525

700

Specific features of trace to infer why energy usage is high

Actual power value

Feedback methods on Fridge and HVAC

Both appliances commonly found across homes

Others38%

Fridge8%

HVAC54%

Evaluation overviewSubmetered

traces

Pow

er

(W)

0350700

Submeter sensor

Can we give such feedback?Disaggregated

traces

Pow

er

(W)

0350700

NILM

Householdaggregate

Submetered traces

Pow

er

(W)

0350700

Submeter sensor

020004000

020004000

Smart meter

Do disaggregated traces provide features needed for providing feedback?

Disaggregated traces

Pow

er

(W)

0350700

NILM

Householdaggregate

Submetered traces

Pow

er

(W)

0350700

Submeter sensor

020004000

020004000

Smart meter

0

125

250

375

500

Fridge is a duty cycle based appliance; compressor turns ON and OFF

periodically

Defrost cycles occurs periodically and consume high amount of power

0

125

250

375

500

0

125

250

375

500

Defrost introduces heat increasing ON duration of next cycles

Fridge usage increases compressor ON durations (and reduce compressor OFF

durations)

0

125

250

375

500

Night hours typically have “baseline” usage

0

175

350

525

700Baseline duty % = Median

duty % in the night

Defrost energy

0

175

350

525

700

Defrost energy = Energy consumed in defrost state + Extra energy consumed in next few compressor cycles

Defrost energy

0

175

350

525

700

Defrost energy = Energy consumed in defrost state + Extra energy consumed in next few compressor cycles

Usage energy

0

175

350

525

700

Usage energy = Extra energy consumed over

baseline

Experimental setupWiki Energy data set

1. 97 fridges2. 58 HVAC

13 out of 95 homes can be given feedback based on usage energy saving

upto 23% fridge energy

13 out of 95 homes can be given feedback based on usage energy saving

upto 23% fridge energy

NILM algorithms show poor accuracy in identifying homes which can be given feedback based on usage energy

17 out of 95 homes can be given feedback on excess defrost saving upto

25% fridge energy

Such feedback can’t be given with disaggregated traces, since these techniques

fare poorly on defrost detection.

Benchmark NILM algorithms on our data set give accuracy comparable or better than

state-of-the-artKolter 2012

Parson 2012Parson 2014

Batra 2014CO

FHMMHart

Error Energy %

0 17.5 35 52.5 70

0

125

250

375

500

“Average” error in energy would be low even if NILM predicted this

But, we wanted to predict..

0

125

250

375

500

It’s the details that we care about

0

125

250

375

500

Like fridge, HVAC duty cycles to maintain the set temperature

0

1000

2000

3000

4000

As temperature increases during the day, more energy required to cool the home

0

1000

2000

3000

4000

0

1000

2000

3000

4000

People typically turn up the temperatures when they leave home

Reco

mm

ende

d M

in. Te

mp

(F)

7779818385

2 4 6 8 10 12 14 16 18 20 22 24

EnergyStar.gov recommended HVAC setpoint schedule

Sleep Morning Work Evening

Reco

mm

ende

d M

in. Te

mp

(F)

7779818385

2 4 6 8 10 12 14 16 18 20 22 24

Setpoint schedule score

Sleep Morning Work Evening

Reco

mm

ende

d M

in. Te

mp

(F)

7779818385

2 4 6 8 10 12 14 16 18 20 22 24

Setpoint schedule score

Sleep

Sleep score = 1 if sleep temp. > 82, (82-temp.)/4 if 78<sleep temp. <82 0 otherwise

Learning HVAC setpoint

01000200030004000

7785

5 10 15 20

HVAC trace

Weather

Learnt setpoint

Giving feedback

7785

5 10 15 20

Features from HVAC trace

6975

5 10 15 200

1000200030004000

7785

5 10 15 20

Learnt setpointDon’t need feedback

Need feedback

84% accuracy on giving feedback using submetered traces

39

NILM methods give 15-30% worse accuracy for feedback

40

Benchmark NILM algorithms on our data set give accuracy comparable or better than

state-of-the-art

Batra 2014

CO

FHMM

Hart

Error Energy %

0 7.5 15 22.5 30

Erro

r in

pred

ictio

n of

m

inute

s of H

VAC

usag

e (%

)

0

6

12

18

24

Hart FHMM CO

NightMorning

Morning hours which have lesser NILM accuracy are important for HVAC feedback

Conclusions

Appliance level data does enable actionable energy saving feedback

Conclusions

Appliance level data does enable actionable energy saving feedback

BUT Results show that we need to revisit the metrics by which we measures progress