Flood risk management in urban areas by using SUDS and a … · 2017. 12. 14. · 23 de Agosto 2014...
16
30/11/2017 1 Flood risk management in urban areas by using SUDS and a Reactive Early Warning System: Case study in Barranquilla, Colombia Humberto Ávila, Leandro Ávila & Augusto Sisa Instituto de Estudios Hidráulicos y Ambientales – IDEHA Universidad del Norte 2017 Introduction • The city of Barranquilla in Colombia has one of the major stormwater management issues in the world. • Due to the lack of a stormwater drainage system in more than 90% of its urban area, the runoff flows on the streets during rain events with dangerous flow rates • The permeable areas are little and the steep slopes of the city cause supercritical flow and flash floods on many streets with high velocities. • This has negative consequences such loss of human life, damage to infrastructure, and economic losses. May 9 th 2017
Transcript of Flood risk management in urban areas by using SUDS and a … · 2017. 12. 14. · 23 de Agosto 2014...
30/11/2017
1
Flood risk management in urban areas by using SUDS and a Reactive Early Warning System:
Case study in Barranquilla, Colombia
Humberto Ávila, Leandro Ávila & Augusto Sisa
Instituto de Estudios Hidráulicos y Ambientales – IDEHA
Universidad del Norte
2017
Introduction• The city of Barranquilla in Colombia has
one of the major stormwatermanagement issues in the world.
• Due to the lack of a stormwaterdrainage system in more than 90% of its urban area, the runoff flows on the streets during rain events with dangerous flow rates
• The permeable areas are little and the steep slopes of the city cause supercritical flow and flash floods on many streets with high velocities.
• This has negative consequences such loss of human life, damage to infrastructure, and economic losses.
May 9th 2017
30/11/2017
2
Watershed characteristics
Source: El Universal NewspaperSource: ADN Newspaper
(%)
SUPERCRITICAL FLOW
Typical flow rates and velocities
30/11/2017
3
Stormwater master plan: Channeling for ending segments of streets
Foto: Luis Rodriguez – El Herando
Foto: www.lachachara.org
Foto: www.semana.com
Street/Stream Lenght (Km) Cost (Millon US$)
Arroyo 84
2.4 33Arroyo 82
Arroyo 79
Arroyo 21 3.4 40
Felicidad 3.4 42
Calle 75 y 76 2.4 33
Calle 65 2.3 30
Hospital 2.4 30
Calle 91 y 92 1.5 21
Calle 58 0.9 15
Arroyo 21 3.4 40
TOTAL 19.7 251
60 Km left with flash flood issues!
Channeling for ending segments of streets
30/11/2017
4
Flood during construction: Calle 76 StreamMay 12th 2017
May 12th 2017
Purpose of the REWS
• The purpose of the REWS (Reative Early Warning System) is to account for an economic and effective EarlyWarning system that allows identifying flash flood riskbased on a deterministic and semi-hydrological model.
• The REWS is complementary to a FEWS (ForecastingEarly Warning System)
• The city of Barranquilla is currently on a channelingmaster plan attending up to 20% of the streams/streetswith flash flood risk.
30/11/2017
5
Fieldwork and monitoring
• Flow rate and rain gages• Topography• Flow direction• Land used
Flow directionElevation Land use
Fieldwork and monitoring
Tipo de Área Porcentaje %
Área permeable efectiva 27
Área impermeable efectiva 73
• Permeable/Imperviouspercentage area
• Satellite image analysis
30/11/2017
6
Type of soil
Type soil Area (Ha) Percentaje(%)
Sand 2984.02 86.71%
Clay 249.08 7.24%
Gravel 98.82 2.87%
Silt 75.43 2.19%
Other 34.15 0.99%
Land use
Residential
Comercial
Residential
Central
Institucional
30/11/2017
7
Modeling and Calibration
• Area• Width• Lenght• Permeability
Horton Model
ManningRoughnessCoefficent
PCSWMM
Model calibration – “Torcoroma” subcatchment
Error:ISE ratingISENSER²SEELSELSE dimRMSERMSE dim
SensorExcellent2.750.8940.9350.82699.530.86.710.376
0
1
2
3
4
5
6
7
8
9
10
11
14:00Aug 23 Sat 2014
14:30 15:00 15:30 16:00 16:30
Calibración cuenca Torcoroma23 de Agosto 2014
Tota
l inf
low (m
³/s)
Date/Time
Sensor Sensor (Obs)Error:ISE ratingISENSER²SEELSELSE dimRMSERMSE dim
SensorVery good4.410.8060.9370.933106767.590.49
0
1
2
3
4
5
6
7
8
9
14:00Aug 28 Thu 2014
14:30 15:00 15:30
Calibración cuenca Torcoroma28 de Agosot 2014
Tota
l inf
low
(m
³/s)
Date/Time
Sensor Sensor (obs)
Error:ISE ratingISENSER²SEELSELSE dimRMSERMSE dim
SensorVery good3.570.4180.6541.0714318.78.730.242
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
1p. m.Sep 3 Wed 2014
2p. m. 3p. m. 4p. m.
Calibración cuenca Torcorom3 de Septiembre 2014
Tota
l inf
low
(m
³/s)
Date/Time
Sensor J21 (obs)
Error:ISE ratingISENSER²SEELSELSE dimRMSERMSE dim
SensorVery good3.360.7350.930.64446.425.75.70.514
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
17:30Sep 7 Sun 2014
18:00 18:30 19:00 19:30
Calibración cuenca Torcoroma7 de Septiembre 2014
Tota
l inf
low
(m
³/s)
Date/Time
Sensor Enter name...
23 de Agosto 2014 28 de Agosto 2014
3 de Septiembre de 2014 7 de Septiembre de 2014
Event �� NSE ISE ISE rating Type
August 23rd 2014 0.93 0.9 2.91 Excelent Calibration
August 28th 2014 0.86 0.81 4.6 Very good Validation
September 3rd 2014 0.65 0.41 3.29 Very good Validation
September 7th 2014 0.92 0.70 5.93 Very good Validation
Adjustment of calibration and validation scenarios
• Coefficient of determination (R�)• Nash-Sutcliff efficiency model
(NSE) coefficients • Integral Square Error (ISE).
30/11/2017
8
Time of concentration equationA time of concentration equation was developed for Barranquilla, based on field data and modeling.
200
100
0
0
5
10
15
20
25
30
35
40
45
50
55
8:00Aug 23 Sat 2014
8:30 9:00 9:30
Hidrograma de salida cuenca Torcoroma
Rai
nfal
l (m
m/h
r)T
otal
inflo
w (
m³/
s)
Date/Time
8 T_CONCENTRACION Sensor T_CONCENTRACION
Hazard analysis for urban floodSeveral studies have been conducted regardingflood safety criteria for the implementation of earlywarning systems. Cox R.S. (2010) identified andevaluated the dangerous flow conditions associatedwith a person or a car based on hydrodynamicvariables.
The Hazard Level (m2/s) was defined as follow
���
�
�
� ∗ � ��
Source : Cox, 2010
Source : Cox, 2010Flow instability models for vehicles and people (Shand, 2011)
30/11/2017
9
REWS ModelA rainfall event is modeled cumulatively by rainfall increment pulses in order to know the runoff conditions on each street intersection each minute.
Felicidad Watershed: Results
Drainage area of 336 Ha and mainstream length of 4.8 Km.
The watershed was divided into 20sub-catchment with areas between5 and 30 Ha.
30/11/2017
10
Real time modeling: Rainfall-Runoff
The August 23th 2014 rainfall event was performed with 120 simulations and it was generated 120 flow, depth and velocity graphs.
Maximum flow conditions in Felicidad watershed
Variable Street 27-61 Street 44 -48 Outfall
Drainage area (ha) 63.4 252.2 336Peak Flow (��/�) 10 54 68
Max Depth (�) 0.24 0.96 0.8
Max Velocity (�/�) 4.10 5.6 8.2
Hazard level (��/�) 0.98 5.37 6.56
Carrera 44 – Calle 48
Reactive Early Warning System
30/11/2017
11
Results – Reactive Early Warning System
SUDS retrofitting: Pilot Tests
• A residential existing building wasretrofitted for improving itsstormwater system by usingSUDS.
• Number of floors: 5
• Underground parking
• Gardens located in front of thebuilding
• Existing non-used undergroundtank for drinking water
30/11/2017
12
Characteristics of the existing drainage system
Area of the property: 544m2
Effective area draining to the undergroundtank: 273m2
Tank´s storage capacity: 27m2
Runoff is discharged directly to the street
Schematic layout of the retrofitting
Recolección de aguas lluvias en
techo
Medidor de Caudal
Jardín
MotobombaTanque de aguas
lluvias parqueadero
Rejillas de recolección de aguas lluvias parqueadero
Tanque Subterráneo
Tanque de Infiltración
Calle
Antejardín
30/11/2017
13
Retrofittings
Water level gage
Garden retrofitting Flow measuring boxes
Pipe and valves systems for controling dischages to the garden or the street
Some resultsEvent:
Rainfall: 79 mm
Weir height: 1 cm
Runoff volume: 21,6��
Peak flow in the building: 21,5 L/s
Peak flow discharge: 12,2 L/s
Peak flow reduction: 43%
Storage runoff volume:2,7��
Runoff volumen reduction: 13%
30/11/2017
14
Results
(L/s) (L/s)
Using dispersed tank for reducing peak flowby land use
Avila, H., Avila, L., and Sisa, A. (2016). "Dispersed Storage as Stormwater Runoff Control in Consolidated Urban Watersheds with Flash Flood Risk." J. Water Resour. Plann.
Manage. , 10.1061/(ASCE)WR.1943-5452.0000702 , 04016056.
30/11/2017
15
Using dispersed tank for reducing peak flowEquivalent volumen estimation
Avila, H., Avila, L., and Sisa, A. (2016). "Dispersed Storage as Stormwater Runoff Control in Consolidated Urban Watersheds with Flash Flood Risk." J. Water Resour. Plann.
Manage. , 10.1061/(ASCE)WR.1943-5452.0000702 , 04016056.
Conclusions
• The REWS proposed in this study was based on a deterministic and semi-distributed hydrological model. Therefore, the runoff and hazard forecasting depend on the rainfall data collected each minute.
• The implementation REWS represents both an economic and effective measure to reduce flash flood risk. However, a Forecasting Early Warning System combined with the REWS would reduce the uncertainty and improve the prediction and reliability. Other applications will be implemented for public users, such as social networks, twitter, real time reports, citizen watch, among other actions to improve the application.
• SUDS are an excellent alternative for managing peak-flow in consolidadedurban watersheds if they are implemented in large scale. Feasablealternatives and social awarness are necesary to make it atractive forpublic and the local goverment.
30/11/2017
16
Thank you!
Humberto Avila, PhDInstituto de Estudios Hidráulicos y Ambientales – IDEHA
Universidad del Norte
![ReMAS-BaseDatos-16feb2012-v2 [Modo de compatibilidad] · Antecedentes 4 • 31 Agosto 2006 – Convocatoria Ideas de Mega‐Proyectos • 19 de Septiembre 2007 – Invitación a formar](https://static.fdocuments.in/doc/165x107/5bd61aaa09d3f2673e8d1a47/remas-basedatos-16feb2012-v2-modo-de-compatibilidad-antecedentes-4-31.jpg)
