©C

opyright, G

rontm

ij A

/S 2

011

Grontmij

Modeling runoff from a green roof

Vis

ualis

ering:

Cura

Vita

DBarbara Kleinlercher

©C

opyright, G

rontm

ij A

/S 2

011

Outline

Motivation

Objectives

Theoretical background

Experimental set-up

Results

Conclusions

©C

opyright, G

rontm

ij A

/S 2

011

Motivation

There is a need for modeling tools, able to reproduce green roof runoff

Urban centers: need for sustainable technologies managing

runoff

Odense: more than 35 overflows per year

©C

opyright, G

rontm

ij A

/S 2

011

Objectives

1. Develop and build a mathematical model, able to reproduce rainfall-runoff relationships.

2. Test the model on field data.

3. Examine the impact of green roofs on runoff hydrographs with respect to rainfall depth and initial condition.

©C

opyright, G

rontm

ij A

/S 2

011

Theoretical Background

Identification of dominant processes

Processes:• Rainfall (P)

• Infiltration

• Subsurface flow - subsurface runoff (SSF)

• Saturated surface flow – surface runoff (SF)

• Evapotranspiration (AET)

A. 30mm Sedum-herb-grass had

B. 30 mm Roof soil

C. 40 mm Mineral Wool

D. 0.8mm Root barrier foil

E. Roof

Layering:

• Interception (I)

©C

opyright, G

rontm

ij A

/S 2

011

Theoretical Background

Conceptual model – 2 reservoir model

©C

opyright, G

rontm

ij A

/S 2

011

Experimental Set-up

Measuring devices:• BL1: Level meter • BL2: Level meter • BL3: Level electrode • QX2: Filling valve• QX1: Drain valve• 2 soil moisture sensors

Study Period:April 2009- June 2010

Recorded data:• Rainfall• Runoff• Soil moisture

©C

opyright, G

rontm

ij A

/S 2

011

Results

Experimental Results

Modeling Results

©C

opyright, G

rontm

ij A

/S 2

011

Experimental Results

Retention capacityTotal analyzed rainfall- runoff samples= 56

©C

opyright, G

rontm

ij A

/S 2

011

Experimental results

Retention capacity of the green roof

©C

opyright, G

rontm

ij A

/S 2

011

Detention, runoff and retention

Initially: Dry Total Rainfall 43.2mm7th of June 2010

Initially: Wet Total Rainfall 35.6mm11-12th June 2009

Experimental results

©C

opyright, G

rontm

ij A

/S 2

011

Surface flow does not occur on the roof, since moisture peaks in the soil layer are well defined by one data point and do not stay constant.

Dominant flow dynamics

Experimental results

©C

opyright, G

rontm

ij A

/S 2

011

07/06 08/06 09/06 10/06 11/06 12/06 13/06 14/06 15/060

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

Date [day/month]

Q [

l/s]

07/06 08/06 09/06 10/06 11/06 12/06 13/06 14/06 15/06

0

2

4

6

8

10

12

14

16

Rain

[m

m]

Qobs.

Qsim.

Rainfall: 10 min. interval

• General impression: model overestimates runoff

• Underestimates peak flows

• Nash-Sutcliffe coefficient: 0.44

07/06 08/06 09/06 10/06 11/06 12/06 13/06 14/06 15/060

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

Date [day/month]

Q [

l/s]

07/06 08/06 09/06 10/06 11/06 12/06 13/06 14/06 15/06

0

2

4

6

8

10

12

14

16

Rain

[m

m]

Qobs.

Qsim.

Rainfall: 10 min. interval

Calibration

• Total volume error: 1.78%

Total amount of rainfall: 91mm

Experimental results

©C

opyright, G

rontm

ij A

/S 2

011

Modeling Results

• Peak flow errors: 1.21% to -81.03%

• Volume errors: -32.68% to 21%

• Nash- Sutcliffe: 0.23 to 0.91 -> 3/5 higher than 0.8.

-> Timing and drainage behaviour: very well reproduced

©C

opyright, G

rontm

ij A

/S 2

011

Modeling Results

06:00 09:00 12:00 15:00 18:00 21:000

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

Time [hh:mm]

06:00 09:00 12:00 15:00 18:00 21:00

0

2

4

6

8

10

12

14

16

Qobs.

Qsim.

Rainfall: 10min interval

24/05 31/05 07/06 14/06 21/06 28/06 05/07 12/07 19/07 26/07 02/08 09/080

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

Date [day/month]

Runoff Q

[l/s]

24/05 31/05 07/06 14/06 21/06 28/06 05/07 12/07 19/07 26/07 02/08 09/08

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Rain

[m

m]

Rain: 10 min. interval

Qobs.

Qsim.

• Peak flow errors: -76% to 16.52%• Volume errors: -69.4% to 19.54%

Model validation: 69 days – 34 rain events

• Total volume error: -49.81%

21:00 00:00 03:00 06:00 09:00 12:000

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

Time [hh:mm]21:00 00:00 03:00 06:00 09:00 12:00

0

2

4

6

8

10

12

14

16

Rain

[m

m]

Rainfall: 10min interval

Qobs.

Qsim.

00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:000

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

Time [hh:mm]

Runoff Q

[l/s]

00:00 06:00 12:00 18:00 00:00 06:00 12:00 18:00 00:00 06:00 12:00

0

2

4

6

8

10

12

14

16

Rainfall: 10min. interval

Qobs.

Qsim.

©C

opyright, G

rontm

ij A

/S 2

011

Conclusions

Substantial amount of rainfall is stored and evaporated

Layer saturation is the key factor determining retention capacity

A 2 storage linear reservoir model is a promising approach to simulate green roof runoff!

07/06 08/06 09/06 10/06 11/06 12/06 13/06 14/06 15/060

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

Date [day/month]

Q [

l/s]

07/06 08/06 09/06 10/06 11/06 12/06 13/06 14/06 15/06

0

2

4

6

8

10

12

14

16

Rain

[m

m]

Qobs.

Qsim.

Rainfall: 10 min. interval

Total amount of rainfall: 91mm

©C

opyright, G

rontm

ij A

/S 2

011

THANK YOU FOR YOUR ATTENTION!