Validation of a lumped thermal parameter modelcoupled with an EnergyPlus model using BUBBLE data

Miguel Martin1 Afshin Afshari1 Peter Armstrong1

Leslie Norford2 Eberhard Parlow3 Roland Vogt3

1Department of Engineering Systems and ManagementMasdar Institute of Science and Technology, UAE

2Department of ArchitectureMassachusetts Institute of Technology, USA

3Department of Environmental SciencesUniversity of Basel, Switzerland

9th International Conference on Urban Climate, 2015

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 1 / 25

Agenda

Purpose of coupling an urban canopy model with EnergyPlus

Parameters considered by the urban microclimate model

Accuracy of urban temperature estimates in the Sperrstrasse

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 2 / 25

Agenda

Purpose of coupling an urban canopy model with EnergyPlus

Parameters considered by the urban microclimate model

Accuracy of urban temperature estimates in the Sperrstrasse

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 2 / 25

Agenda

Purpose of coupling an urban canopy model with EnergyPlus

Parameters considered by the urban microclimate model

Accuracy of urban temperature estimates in the Sperrstrasse

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 2 / 25

Detailed building energy model

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 3 / 25

Detailed building energy model

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 4 / 25

Detailed building energy model

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 5 / 25

Purpose

How boundary conditions of a building located in an urban area can bemore accurately evaluated to improve estimation of their impact on energyconsumption.

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 6 / 25

Urban temperature computation

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 7 / 25

Urban temperature computation

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 8 / 25

Urban temperature computation

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 9 / 25

Urban temperature computation

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 10 / 25

Urban temperature computation

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 11 / 25

Sensible heat gains

Waste releases: Qwaste = δwaste(1− η)Qc

Traffic: Qtraffic = ItrafficFtraffic (h)

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 12 / 25

Sensible heat gains

Waste releases: Qwaste = δwaste(1− η)Qc

Traffic: Qtraffic = ItrafficFtraffic (h)

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 12 / 25

Validation environment

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 13 / 25

Validation environment

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 14 / 25

Distances

Signal based distances:

I Root mean square errorI Mean bias error

Distribution based distances:

I Kolmogorov-Smirnov distanceI Root mean square of Hellinger distances for normal distributions

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 15 / 25

Distances

Signal based distances:I Root mean square error

I Mean bias error

Distribution based distances:

I Kolmogorov-Smirnov distanceI Root mean square of Hellinger distances for normal distributions

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 15 / 25

Distances

Signal based distances:I Root mean square errorI Mean bias error

Distribution based distances:

I Kolmogorov-Smirnov distanceI Root mean square of Hellinger distances for normal distributions

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 15 / 25

Distances

Signal based distances:I Root mean square errorI Mean bias error

Distribution based distances:

I Kolmogorov-Smirnov distanceI Root mean square of Hellinger distances for normal distributions

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 15 / 25

Distances

Signal based distances:I Root mean square errorI Mean bias error

Distribution based distances:I Kolmogorov-Smirnov distance

I Root mean square of Hellinger distances for normal distributions

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 15 / 25

Distances

Signal based distances:I Root mean square errorI Mean bias error

Distribution based distances:I Kolmogorov-Smirnov distanceI Root mean square of Hellinger distances for normal distributions

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 15 / 25

Distance of Hellinger

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 16 / 25

Root mean square of Hellinger

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 17 / 25

DNI estimation methods

DISC DISC+P DIRINT DIRINT+DP

RM

SE

[W

/m2 ]

0

50

100

150

Strasbourg

DISC DISC+P DIRINT DIRINT+DP

MB

E [

W/m

2 ]

-40-20

020406080

Strasbourg

DISC DISC+P DIRINT DIRINT+DP

RM

SE

[W

/m2 ]

100

200

300

400

Geneva

DISC DISC+P DIRINT DIRINT+DP

MB

E [

W/m

2 ]-100

0

100

200

300Geneva

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 18 / 25

DNI estimation methods

DISC DISC+P DIRINT DIRINT+DP

RM

SE

[K

]

0.5

1

1.5

2

2.5

3

3.5

4

4.5

DISC DISC+P DIRINT DIRINT+DPM

BE

[K

]

-4

-3

-2

-1

0

1

2

3

4

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 19 / 25

Winter temperature estimates

01/01 02/01 03/01

Tem

per

atu

re [

oC

]

-10

-5

0

5

10

15

20

25RMSE = 1.51 [K], MBE = -0.43 [K]

MeasurementsEstimates

Temperature [oC]-20 0 20 40

Fre

qu

ency

[%

]

0

1

2

3K-S = 8.47

LST5 10 15 20T

emp

erat

ure

[oC

]

-10

0

10

20RMSH = 0.04

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 20 / 25

Spring temperature estimates

04/01 05/01 06/01

Tem

per

atu

re [

oC

]

-5

0

5

10

15

20

25

30

35

40RMSE = 1.26 [K], MBE = 0.8 [K]

MeasurementsEstimates

Temperature [oC]-50 0 50

Fre

qu

ency

[%

]

0

1

2

3K-S = 11.03

LST5 10 15 20T

emp

erat

ure

[oC

]

0

10

20

30RMSH = 0.07

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 21 / 25

Conclusion

Urban canopy model coupled with a detailed building energy model

Validation protocol for mean urban temperature in the Sperrstrasse

Daily accuracy of temperature estimates using DISC or DIRINTmodels

Measurements are underestimated in winter and overestimated inspring

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 22 / 25

Conclusion

Urban canopy model coupled with a detailed building energy model

Validation protocol for mean urban temperature in the Sperrstrasse

Daily accuracy of temperature estimates using DISC or DIRINTmodels

Measurements are underestimated in winter and overestimated inspring

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 22 / 25

Conclusion

Urban canopy model coupled with a detailed building energy model

Validation protocol for mean urban temperature in the Sperrstrasse

Daily accuracy of temperature estimates using DISC or DIRINTmodels

Measurements are underestimated in winter and overestimated inspring

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 22 / 25

Conclusion

Urban canopy model coupled with a detailed building energy model

Validation protocol for mean urban temperature in the Sperrstrasse

Daily accuracy of temperature estimates using DISC or DIRINTmodels

Measurements are underestimated in winter and overestimated inspring

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 22 / 25

Thank you for your attention

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 23 / 25

Impact on energy consumption

5 10 15 203

4

5

6

7

8

9

Local time (hours)

Tem

per

atu

re [o

C]

5 10 15 203

4

5

6

7

8

9

10

11

12

Local time (hours)

Bo

iler

fuel

co

nsu

mp

tio

n [

kW]

CSIsolated

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 24 / 25

Winter night impact

+4% average temperature ⇒ -7.2% total fuel consumption

Miguel Martin (Masdar Institute) Coupled scheme validation in Basel July 20, 2015 25 / 25