Potential of Low-Impact Development for Stormwater Management in - - PowerPoint PPT Presentation

Potential of Low-Impact Development for Stormwater Management in Hong Kong

Kit Ming LAM, Ting Fong May CHUI, Ze-ying LI

Department of Civil Engineering, University of Hong Kong

Flooding  Urban drainage demand

Culprits for Urban flooding in Hong Kong  Severe (short and intense) rainstorms  Storm drain extension in harbour reclamation  Ageing of storm drains  Steep topography

Urbanization  reduced permeability of land

Need for stormwater drainage  Hard engineering solutions

Additional Runoff

Source: http://en.wikipedia.org/wiki/Stormwater

• Erode watercourses
• Water body pollution
• Flooding

Large infrastructures, e.g., :

• Upstream interception and diversion for flood prevention,

e.g., Hong Kong Western Drainage Tunnel

• Flood storage scheme, e.g., Tai Hang Tung storage tank
• Average Annual Rainfall
• f Hong Kong ~ 2200 mm
• Concentrates in May -

September

Stormwater LIDs:

• Devices, practices, or methods used to manage stormwater runoff

locally

• Mitigate changes to water quality of urban runoff caused by

increased impervious surfaces from land development

• Can also reduce stormwater volume and peak flows
• Work through evapotranspiration, infiltration, detention, and

filtration or biological and chemical actions.

Low impact development (LID)  Environmental friendly solutions

Also known as:

• Stormwater best management practices (BMP)
• Treatment train
• Sustainable urban drainage systems (SUDS)
• Water sensitive urban design (WSUD)

Common LID devices

Permeable/porous pavement Bioretention (rain garden) Vegetated swale Green roof

Challenges in Hong Kong:

×

Steep topography  low retention rate of runoff

×

Very heavy rainfall

• Existing drainage design and layouts
• Lack of spaces
• Inadequate awareness of public and engineers

Stormwater LIDs in Urban Drainage

Retrofit

Retrofit stormwater LIDs

Page et al. (2014), Schlea et al. (2014)

Experience overseas:

Retrofit stormwater LIDs

Possibilities for Hong Kong Rain Gardens  Effective design for road-side applications

landscape garden  bio-retention

Potential LID applications in Hong Kong

Bioretention and Porous pavement:

Working principles  Infiltration structures: Retention & Infiltration General benefits

• Reduce adverse stormwater impacts due to urban development
• Easy to incorporate into urban landscaping
• Can be applied on a wide scale

Few previous applications in Hong Kong

• Rainfall patterns in Hong Kong?
• Ease of retrofit?
• Installation and maintenance problems?

Bioretention:

• Depressed planter with engineered soil (soil

media)

• Receive and store stormwater from the

vicinity of large area.

• Allow stormwater penetrate through it.
• Encourage stormwater seeping into the

ground. Examples of ponded bioretention cell

Depression storage Inflow Overflow Infiltration into soil media

• Absorption by soil media
• Percolation into in-situ soil
• Underdrain (optional)

(assumed constant)

Bioretention - Introduction

Rooftop Driveway Park Car park

Source: Curtis Hinman (2007), Rain Garden Handbook for Western Washington Homeowners, Designing your Landscape to protect

• ur streams, lakes, bays, and wetlands, Washington State University Extension Pierce Country

Bioretention usually serves adjoining impervious catchment area

Bioretention - Introduction

Bioretention - Introduction

• Uncontrollable factors
• Local rainfall pattern
• Controllable: factors
• Area of the rain garden vs. area of

whole catchment, it determines the amount of inflow

• Storage depth of the rain garden,
• r the volume between the soil

surface and the controlled

• verflow level
• Infiltration rate, or the rate of

water dissipation through infiltration

• 1. Inflow
• 2. Overflow
• 3. Infiltration

Performance of rain garden can be determined by

Pilot-scale physical model of bioretention cell at HKU

• Monitor inflow and overflow under local rainfall events.
• Quantify peak reduction and volume retention against rainfall events.
• Provide data for verification of numerical model.

Physical Experiment

 A planter-box type rain garden constructed for collecting and treating runoff from a rooftop  Measure inflow,

• verflow, and

underdrain for real rainfall events

Bioretention - Experiment

• Rooftop area = 26 m2
• Rain garden area = 0.9 m2

Bioretention - Experiment

• Rates of inflow, overflow, and underdrain

were monitored in 5-second intervals

• Data monitored during (1) August 2 to

September 5, 2013; (2) March 20 to April 8, 2014

(Infiltration rate ∼ 50 mm/hour) Freeboard = 50 mm Storage depth = 95 mm Soil media = 450 mm Gravel and pipe = 100 mm

Bioretention cell  Planter box

Flow measurement of

• verflow and underdrain

Bioretention - Results

58 rainfall events identified during the 55 days of monitoring periods  Typical examples….

Rainfall depths ranged from less than 1 mm to over 90 mm.

Observation: Performance of the bioretention cell is affected largely by the rainfall depths.

Full interception

(1) 04:57-05:16, August 3, 2013. Effective rainfall depth = 4.2 mm

Bioretention - Results

Partial interception

(2) 00:40-01:22 on Aug 14, 2013. Effective rainfall depth = 6.8 mm.

• Overflow
• bserved but

behind the first peak inflow

Negligible interception

(3) 12:09-14:11 on Aug 13, 2013. Effective rainfall depth = 27.6 mm.

• Most inflow bypassed the cell through overflow
• Overflow observed and lasted till the end of inflow
• Magnitude of underdrain was much smaller than overflow

Bioretention - Results

Bioretention - Results

Peak flow reduction ratios for all 58 rainfall events:

• 100% reduction for events
• f depth < 3 mm.
• Little reduction for events
• f depth > 3 mm.

Same observation on the Volume retention ratios 

Bioretention - Results

Volume retention:

• 2.7 mm was retained statistically

from the monitored results. Physical set up:

• 95 mm storage depth in

bioretention cell, serving catchment at 1 / 3.3%

• that is 95×3.3% = 3.1 mm effective

storage depth Overflow depth vs. inflow depth for events less than 30 mm.

• The bioretention cell can retain rainfall

close to its storage depth only.

• (Infiltration rate is not very large)

Bioretention - Results

Given a fixed setting (storage depth) of a bioretention cell, its retention performance is governed by:

• Rainfall depth (and pattern)
• Catchment area ratio (spreading thin the storage depth)
• Infiltration rate of soil media

With the monitoring results from the physical model, numerical model (e.g., SWMM) is used to estimate the hydrologic performance of a bioretention cell (storage depth = 150 mm) for local rainfall climate and patterns

Example: 10-year averaged retention ratio of stormwater in Hong Kong rainfall conditions (2004-2013) 

Porous Pavement - Introduction

 Similar infiltration BMP structure as bioretention

Storage from pores Inflow Overflow Infiltration into in-situ soil and underdrain

• commonly used to collect rainfall
• n its surface and some apron

area.

• can provide storage layer for

stormwater below the surface.

• allow stormwater to percolate

into in-situ soil at the bottom.

General configuration: 1) Permeable surface course and bedding layer (permeable enough to allow inflow to penetrate through) 2) Open-graded gravel base course (storage layer to provide additional storage below the surface) 3) Underdrain and in-situ soil (to drain the stored water away) 4) Overflow structure (to avoid flooding the pavement)

Porous Pavement - Introduction

(http://www.icpi.org)

Major differences from bioretention:

• large watershed area ratio, or the catchment is restricted to the area of

permeable pavement, thus allowing a considerable storage.

• can be designed to capture all the stormwater, leading to a good

reduction in peak discharge and high volume retention. Major hydrologic design parameters:

• Design storms
• Gravel storage
• Infiltration capacity of in-situ soil

Porous Pavement - Introduction

Site trials of porous pavements in Hong Kong (2014-15)

• A study initiated by DSD (Drainage Services Department, Hong Kong

Government)

• To explore the hydrologic performance of porous pavements under the

local rainfall conditions of Hong Kong

• Test site in Western Kowloon

Porous Pavement – Field Tests

3 test panels, each 4 m × 3 m & 1 control panel (impervious concrete cover)

Porous Pavement – Field Tests

Three surface course materials: 1) PICP: permeable interlocking concrete pavers (with fine loose aggregates in between gaps) 2) OCP: open-cell pavers (with fine loose aggregates in open area) 3) PB: permeable paver blocks (with fine sand filling the gaps) (a) PICP (b) OCP (c) PB

Porous Pavement – Field Tests

Vertical structures: 1) Surface course: paver blocks (60 to 90 mm thick) 2) Bedding layer: fine aggregates (10 mm), or sand, thickness ∼ 40 mm 3) Base course: washed uniformly graded open aggregates (37.5 mm), Storage layer of thickness 500 mm 4) Sub-soil, with geotextile on top

• On pavement surface, berms to contain rainfall on pavement surface
• nly, routed to overflow gully at corner

Weir box Piezometer well Gully inlet box

Porous Pavement – Field Tests

• Surface runoff falls to weir box for

flowrate measurement

• Piezometer well to measure water

level in storage layer

Positive finding from the monitored results: 100% peak reduction and 100% volume retention achieved. For all rainfall events

• Very small surface runoff

measured

• Very shallow water level

recorded in the storage layer.

• Suggesting very high

infiltration capacity of subsoil at test site.

• Design effective storage

depth at 500×0.4 = 200 mm may be too large

Field Tests – Preliminary Results

Daily rainfall depths during June to December 2014

* Red rainstorm signal was issued on the day with 160 mm precipitation.

Field Tests – Preliminary Results

Example: runoff monitored on 22 June 2014, 144 mm rainfall in 5 hours

Porous Pavement – Laboratory Tests

A replica of one field test panel:

• To test the full retention capacity of porous pavement
• Under design rainstorm events
• Under controllable exfiltration rate into subsoil

Wooden rectangular open box, 1.7 m × 1.2 m, 0.65 m tall,  test panel

• Same vertical structures,

but base course at 280 mm thickness

• Except that there is no

sub-soil

• Instead, exfiltration into

sub-soil is represented by a bottom drain fitted with a valve to control the exfiltration rate

Porous Pavement – Laboratory Tests

• Precipitation onto pavement

surface is simulated by draining water from an overhead perforated pipe

• Valve and rotameter to control

the hydrograph

• Setting and measurement every

5 minutes 280 mm 400 mm

Porous Pavement – Results

(1) Continuous steady rainfall events

Rainfall intensity, dr (mm/h) Exfiltration rate into soil, i (mm/h) Time to surface runoff, tf (min) Experiment 50 3.5 165.0 50 8.6 185.0 50 35.3 590.0 100 3.5 75.0 100 8.6 80.0 100 35.3 100.0 150 3.5 50.0 150 8.6 50.0 150 35.3 60.0

• Exfiltration rate

into sub-soil largely governs the retention performance

Porous Pavement – Results

(2) Design synthetic rainstorms

• Rainstorms constructed from the IDF specified in the

Stormwater Drainage Manual by DSD

• Return period (year) = {2, 10, 50}
• Rainstorm duration (h) = {2, 4}
• Exfiltration rate into sub-soil, i (mm/h) = 3.5

Example: 10-year return, 2-hour rainstorm  Storage full after 100 min.  Overflow occurs

Porous Pavement – Results

(2) Design synthetic rainstorms

Return period: Storm duration: 2 year 10 year 50 year 2 hour (210/400 mm) 105 min. 75 min. 4 hour (390/400 mm) 135 min. 115 min.

*Depth of pavement:

• 80 mm paver block
• 30 mm bedding layer
• 290 mm base course (280 mm above bottom drain)

 The storage layer of 280 mm can retain very severe rainstorms

• Bioretention and porous pavement work under similar principles of

stormwater retention and infiltration

• Pilot-scale model tests, field trials and laboratory experiments are carried
• ut to explore their effectiveness in managing the quantities of stormwater

runoff under Hong Kong local conditions

• The tests confirm that the effective storage depth (volume) and the

exfiltration rate of stored water into the subsoil are critical factors in the hydrologic performance

• Numerical model (SWMM) can be used to estimate the statistical

hydrologic performance in Hong Kong over several years

Concluding Remarks

• Bioretention and Porous Pavement

Bioretention

• Usually serves a large impervious watershed area  storage depth can

be quickly filled up during design rainstorms

• Recommended storage depth of 150 mm, when serving catchment 10×

larger catchment, and with exfiltration at 40 mm/h, can retain over 60%

• f annual runoff in Hong Kong

Porous pavement

• Usually receiving runoff from its surface and some apron area
• A dry porous pavement with base course thickness 280 mm can retain

fully rainfall depth ≈ 100 mm , which includes 2-year return rainstorms in Hong Kong for a duration up to 4 hours. Even for highly impervious subsoil condition.

• The 500 mm storage in the field test panels should be able to retain 50-

year return rainstorms

• Porous pavement has great potential for LID in Hong Kong
• Bioretention will be quickly ponded during Hong Kong rainstorms,

but can combine with landscape gardens

• Green roof is expected to have similar hydrologic performance as

bioretention, with small watershed area and small storage depth

Concluding Remarks

• Bioretention and Porous Pavement

Nullah Decking

Decking an open nullah with a green channel cover (GCC)

Other potential LID devices for Hong Kong

Pond Catchmen t Sub1 Sub2 Sub3 SwaleSub 3 SwaleSu b2 SwaleSub1 Pond Catchment

• utlet

SWMM model of a generic Singapore urban catchment (10 ha in area):

• Drain (open channel) dimensions 2.0 m x 2.5 m
• Green cover: drains 1 -3 topped with green

cover

• Green cover atop Drains 1-3 are modeled (1-D)

as Bioretention cells in SWMM

Other potential LID devices for Hong Kong

• The green channel cover delays the release of runoff to drain, and

produces a hydrograph with a wider receding portion

• The GCC decreases and delays the peak flow of a large storm by 14%
• GCC’s performance is particularly encouraging given its small space

requirement

Other potential LID devices for Hong Kong

SWMM model of Fo Tan Nullan

Other potential LID devices for Hong Kong

SWMM model of Fo Tan Nullan

Other potential LID devices for Hong Kong

3000 mm soil depth

44

Other potential LID devices for Hong Kong Green roof

Potential LID benefit

If all impervious surfaces are fitted LIDs  ??

Tin Hau catchment

Potential LID benefit

Shatin catchment Kowloon City catchment

Potential LID benefit

 Numerical models, field trials and laboratory experiments all demonstrate the potential of LID in reducing stormwater runoff.  For frequent storms, bioretention, porous pavements and green roofs are all capable of nearly full retention of stormwater  Deep porous pavement and bioretention can also reduce stormwater runoff during severe storms.

The Way Forward

LID implementation -->

• Socio-economic benefits
• Public policy (by government departments)

Success of LID depends on

• Objectives (e.g., reduce peak flow, maintain low flows)
• Catchment choice (e.g., land uses, existing drainage networks)

 Water quality improvement??  Rainwater harvesting??

Thank you

Acknowledgement: DSD, Hong Kong & Islands Division Mr Lawrence CC Li, Mr KK Choi