Pap 19 Port Louis Is Flooding

Download as pdf or txt
Download as pdf or txt
You are on page 1of 20

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/342887395

Port Louis is Flooding (in Journal of the Institution of Engineers, Mauritius,


2014)

Article · September 2014

CITATIONS READS

0 2,572

2 authors, including:

Virendra Proag
University of Mauritius
69 PUBLICATIONS 371 CITATIONS

SEE PROFILE

All content following this page was uploaded by Virendra Proag on 13 July 2020.

The user has requested enhancement of the downloaded file.


Port Louis is Flooding
Nooriya JHINGUT and Virendra PROAG

Faculty of Engineering, University of Mauritius

Abstract

On 30th March 2013, some 152mm of rain fell in the city of Port-Louis and resulted in a severe
flood that caused eleven deaths and consequential property damage. According to witnesses,
photographs and videos, the flood level rose by up to 2 metres in some places and roads were
transformed into canals. Some explanatory causes put forward are: climate change, inadequacy of
drains, lack of maintenance of drains and massive unsustainable developments. However, historical
records revealed that Port-Louis has been flooded several times since 1840, with impacts similar to
those of March 2013. Based on this key finding, the effectiveness of the city's archaic drainage
system was suspected.
This paper assesses the adequacy of three specific canals which drain the lower regions of Port-
Louis: the Pouce Stream, Butte à Tonniers and Ruisseau des Créoles. Several techniques to quantify
the peak flow on the day of the flood were adopted: a direct method involving desk study whose
result was compared with an indirect method involving observed flood levels, survey of the three
canals and discharge calculations over structures acting as weirs during the flood. Even when
obstructions in canals were ignored, calculations revealed that the drains have an inadequate
carrying capacity at several places along their course. Two particular low-lying regions are prone to
ponding: the vicinity of Place d'Armes and Company Garden/La Chaussée Street. It is estimated
that the flood return period is probably more than 25 years, but is not the worst that could occur. To
prevent flooding in Port-Louis, a longer return period should therefore be adopted in the re-design
of the drains. Are we prepared to adopt more stringent design rules to feel safer in the future?

1. INTRODUCTION
Normally, drainage networks are meant to convey excess water due to a flood to nearby
watercourses and ultimately into the sea. If the drainage system is not properly designed, then the
concept behind stormwater planning itself is defeated. It is true that flood is a natural hazard, but,
man-made disturbances to the nature's right of way often amplifies the problem, resulting in the
economy and society being adversely affected (Suriya 2010).

Indeed, this was the case of Port-Louis, the capital of Mauritius, in the afternoon on Saturday 30th
March 2013. The city witnessed a severe flood which caused eleven deaths and several billions of
rupees of damage to infrastructure and property (Seebaruth 2013).It was reported that 152.6 mm of
rain fell over just five hours (Week-End 2013a), with the highest recorded rainfall intensity being
91 mm/hr from 14h00 to 15h00. Moreover, a high tide of 2.07mabove datum was recorded by the
Mauritius Meteorological Services at 15:13 hours (Kassamally 2013).

The city was significantly flooded especially downstream three main channels which discharge into
the Caudan Basin. These are the Pouce Stream, the Ruisseau des Créoles and the artificial Butte à
Tonniers canal (Hilbert 2013a).It was also reported that the water level had reached a height of
approximately 2 metres in the vicinity of the La Poudrière Street and 1 metre in the area of the La-
Chaussée Street (Le Dimanche/L’Hebdo 2013). These canals are shown on Figure 1.

1
Figure 1: Most significantly flooded places in Port-Louis on 30th March 2013.

Port-Louis had an estimated population of 148,780 inhabitants in the 2011 census (UN-Habitat
2011), and lies only 5 metres above sea level, which makes it vulnerable to storms, storm-surges
and flooding (Fairhurst et al. 2011). Moreover, the rapid development in the city has resulted in an
impermeability factor ranging from 90% to 100% in some places such as those near La-Chaussée
Street. This has resulted in greater runoff and reduced infiltration during heavy rainfall
events(MOW 1996).In fact, La Chaussée Street was previously a primitive crossing constructed
over a swamp and was rebuilt as a permanent causeway in 1779. During those times, the nearby
Company Garden which was a marshy area, was also filled up (Toussaint 1973, p.36).

Having identified the most vulnerable places in Port-Louis, it was judicious to limit the study as
shown in Figure 1 to three main channels: Pouce Stream, the Ruisseau des Créoles and Butte à
Tonniers canals to determine whether they are adequate to carry flood flows effectively into the
Caudan Basin, using the following steps:
• To estimate the peak discharge (Qp) from the catchments
• To obtain the cross-sections and slopes of the canals
• To estimate the carrying capacities of the canals (Q)
• To compare the carrying capacity of each canal at its outfall, QOUT,with its corresponding Qp
• To identify the possible causes of flooding based on the design of the canals and propose
flood mitigation solutions
2
2. LITERATURE REVIEW
2.1 Possible causes of the March 2013 flood
Opinions on the causes of the catastrophic flood in Port-Louis on the 30th March 2013 are sharply
divided. The reasons given by authorities, private consultants, journalists and the public at large are:
• Climate change
• Inadequacy of existing drains
• Modification of existing drains
• Lack of maintenance
• Massive unsustainable development
• The waterfronts
• New infrastructure projects
• Inadequate road side drains, with the overflow being directed downstream towards Port-
Louis, through the motorway which acts like a well designed channel (Proag 2013; Rainer
2013). This is illustrated in Figure 2. This overflow was then naturally carried towards the
Place d’Armes and is suspected to be the principal cause of inundation of the subways
(L’Express 2013; Seebaruth 2013; Week-End 2013b).
• Illegal constructions
• Debris from road construction (see Figure 3).

Figure 2 (left): Roads near the Caudan Flyover behaving as drains during flood
Figure 3(right): Blockage of Pouce stream at its outlet by debris (L’Express 2013)

2.2 Estimation of flood flows


2.2.1 Design flood
Peak floods need to be determined for the safe design of drains, bridges, culverts, channels and
other flood control structures(Proag 1995).The design flood (Raghunath2006)is selected after
considering the project features, a proper balance between the cost of the flood control structures
and the cost of the possible flood damages such as loss of life and property and interruption of
communication and business. Protection against the most severe floods might be an uneconomical
investment due to the rare occurrence of such floods (SANRAL 2006).

2.2.2 Return period


The degree of flood protection is reflected in the design return period or recurrence interval of the
flood. Return periods that are generally adopted for different types of structures are given in the
Table 1.
3
Table 1 : Guidelines for the selection of return periods (Ponce[no date] )

What is high risk or medium risk is subjective. The person who has gone through a flood experience
has very marked feelings about the matter!!

Based on the current practice in Mauritius, a 5 or 10-year return period is adopted for the design of
gutters, ditches and storm drain laterals. On the other hand, the return period considered for the
design of culverts is 25 years whilst 50 years is used for structures like bridges (Kassamally 2013).

2.2.3 Event probability


Even if a flood of a long return period is selected for the design of a flood control structure, there is
always a flood risk to the structure during its lifetime. The probability, R, of an event occurring or
being exceeded during the stated design life can be determined by the equations below:

R=1-(1-P) N
R=1-[1-(1/Tr)]N
where:
P= Probability of occurrence= 1/Tr
Tr= Return Period (years)
N= Useful lifespan of the structure (years)
This concept is further illustrated in Table 2.

Table 2 indicates that a 1 in 1000 year flood has a probability of 7% occurring during a man’s
lifetime of 75 years! Would a sensible person insure himself against a 7 % risk?

It can be observed that the greater the return period, the smaller is the probability that a rainfall
event will be equalled or exceeded. This guideline is often considered in the design of huge
structures and where public safety is threatened (Proag 2013). According to the South African
National Roads Agency Limited (SANRAL 2006), an optimal design standard is usually based on a
minimum recurrence interval that is equal to the design life of the hydraulic structure However,
based on Table 2, the use of a 50-year return period for bridge design means that the structure still
run a high risk of experiencing floods with higher magnitudes during its lifetime (50 years or
greater).

Therefore, a conservative design approach should be adopted where protection of human life is
crucial, even though the structure will be more costly. The designer is also liable to choose a design
return period based on his engineering judgment, that can cater for future developments (SANRAL
2006; Proag 2013). The same logic applies for the design of urban drainage systems. According to a
basic policy of the Urban Drainage and Flood Control District (UDFCD) of Denver, any major
drainage system should be able to convey water without flooding the surroundings and remain fairly
stable during a major storm, such as a 100-year flood event (UDFCD 2008).
4
2.2.4 Peak flood calculation methods
There are many calculation techniques that can be used to determine the peak discharge for the
design of drainage systems. Some of the most commonly used methods (Raghunath 2006;
SANRAL 2006) are categorised as:

• Statistical Methods (e.g. Flood frequency analysis)


• Empirical methods
• Deterministic methods (e.g. Unit hydrograph method and Rational method)
• Indirect hydraulic methods (e.g. Post-flood surveys and peak flow estimates from nearby
hydraulic structures)

2.2.5 Rational method


The Rational method gives a simple rainfall-runoff relationship:

Q= 0.278 CiA
where Q = Design flood (m3/s)
C = runoff coefficient (dimensionless)
I = rainfall intensity (mm/hr)
A = catchment area (km2)

Table 2: Risk R, that a flood of a given return period will be equalled or exceeded during
periods of various lengths(Proag 2013)

Although C may have low values of C=0.15, a conservative value of C=1 is recommended in
situations where drainage facilities have to be designed for maximum flood in urban areas
(SANRAL 2006), along paved streets(UDFCD 2007),for buildings with inclined roofs (NCDENR
2007) and easily saturated soils under cyclonic conditions, for clayey sites located in tropical
countries (Proag 2013).

Table 3 gives rainfall intensities in Mauritius, for different durations and corresponding to different
return periods.

5
Table 3: Rainfall Intensity(mm/hr) in Mauritius for different durations and return periods
(Mauritius Meteorological Services)

The Kirpich formula, most commonly used to determine the time of concentration, tc, for small
watersheds that are dominated by channel flow (Thompson 2006; Proag 2013), is given by:

𝒕𝒄 = 𝟎. 𝟎𝟏𝟗𝟒𝟕 𝑳𝟎.𝟕𝟕 𝑺−𝟎.𝟑𝟖𝟓


Where
tc= time of concentration (minutes)
L = maximum length of travel of water (m)
S= slope of watershed = ΔH/L in which
ΔH= difference in elevation between the most remote point on the watershed and the outlet.

It must be noted that estimates of i based on tc is inappropriate for durations longer than 300 minutes
since the assumption of steady rainfall rate becomes invalid. Conversely, the estimates become
unacceptably large for durations lesser than 10 minutes.(Thompson 2006). However, the UDFCD
(2007) recommends a minimum tc of 5 minutes for urban areas.

2.2.6 Indirect hydraulic method: Post-flood surveys and Flow estimate weirs
Post-flood surveys are usually carried out to identify flood marks and to collect historical flood
levels through local enquiry (Raghunath 2006; Proag 2013). The surveys also include the
determination of the flood marks’ elevations, cross sections of the channels under study and a visual
assessment of their roughness (Rantz 1982).

This approach is an indirect method for flood discharge estimations. Topographic levels are
measured to connect the flood marks and to plot a water profile with the highest flood level marked
on it. The peak discharge can then be determined by the slope-area method. This technique involves
the calculation of the cross-sectional area (A) of the profile and its velocity (V) though empirical
formulae such as the Manning’s formula. The discharge (Q) is then calculated as Q=AV (Chow
1959; Rantz 1982; Raghunath 2006; Lumbroso and Gaume 2011).

However, witnesses often tend to overestimate the flood levels because of the turbulent flow.
Furthermore, the flood levels witnessed may not always be the maximum which correspond to the
peak stage but can nevertheless be confirmed by high water marks.

6
A weir is a device for flow measurement and control in open channels, such as canals and streams.
The measured discharges can range from a trickle in a ditch to a flood in a river (Salmasiet al.
2012).Thus, a weir in the vicinity of the flooded area can be used to estimate the flood discharge.
This approach is also an indirect method, whereby the flood flow can be determined through a field
survey of flood marks and the geometry of the weir (Rantz 1982; Raghunath 2006).

Sharp-crested weirs are usually used for the measurement of irrigation water and thus rarely present
in urban areas (Smajstrla and Harrison 2002).

The broad-crested weiris a solid flat-crested weir of rectangular shape which spans the full width of
a channel. The basic equation to calculate flow over the weir is given below:

Q= CdbH3/2

Where,
Q =discharge (m3/s)
Cd= coefficient of discharge, which takes into consideration the weir shape and caters for minor
energy losses. For the broad-crested weir, Cd= 1.6 (Kay 2008; Claydon 2011).
b = width of weir normal to the flow (m)
H = head on weir measured from crest (m).

A free overfall(free-overfall weir)occurs when flow in a long channel that is not submerged by
tailwater undergoes a sudden drop, resulting in the separation of the flow from the solid boundary.
(Subhasish and Kumar 2002).

In gently sloping open channels, the approaching flow is sub-critical and the ratio of the end-depth
to the critical depth can be used to estimate the discharge over the weir. However, the approaching
flow developed in channels with steep slopes is super-critical. In that case, the discharge is
dependent on the end-depth, channel slope as well as the channel roughness (Beirami et al. 2006).

In a channel of rectangular cross-section, the equation below is used to predict the discharge over a
free-overfall weir:
Q = Cd B g 0.5 he1.5(ISO 3847,1977 cited in Zbyneket al. 2013)
Where,
Cd = 1.66 for confined nappe and 1.69 for unconfined nappe.
B = width of weir > 0.3 m
he = end depth > 0.04 m
g = gravitational acceleration= 9.81 m/s2

Two simple validation tools (Lumbroso and Gaume 2011) are :


(a) Cross-checking the estimated discharge value with that obtained from another method,
such as the rational method, using C=1 and the maximum rainfall recorded for the flood
event over its duration.
(b) Use of photographs and videos taken during the flood, can give more clues on flood
levels reached in different places.

2.3 Open channel design


One commonly used empirical formula which describes the relationship between channel size,
shape, slope and roughness is the Manning’s formula, given by:

7
𝑨 𝟐/𝟑 𝟏/𝟐
𝑸= 𝑹 𝑺
𝒏

Where,
Q= flow in channel (m3/s)
A= wetted area (m2)
N = Manning's coefficient of roughness
R= hydraulic radius (m) = wetted area/wetted perimeter (m)
S= slope of energy line= channel bed slope

A properly designed channel should be based on two rules. Firstly, the carrying capacity of the
channel at its outlet should always be greater than the peak discharge from the catchment. Secondly,
the downstream carrying capacity at a given cross-section of the canal should be greater or equal to
the upstream flow (Proag 2013). It is important that there is no stagnant water – as with a tide -
under a bridge because the clearance under the bridge is reduced, and the Manning’s formula no
longer applies.

3. DATA COLLECTION
The aim of the data collection was to estimate the peak flow during the flood and the carrying
capacity of the drains under study.

3.1 Desk Study


This comprised investigating the defined area through:
• Aerial maps from Google Earth software to locate relevant streets, drains, and buildings, and
to obtain rough estimates of lengths and elevations.
• Topographical map on AutoCAD software to observe the terrain relief and identify low-
lying contours, and to delineate catchments under study, to determine their length, slope and
area.
• Soil map to determine the main types of soil which affect the catchments' response to runoff.
• Pictures and videos that were taken during the flood to get an indication of flood marks and
the extent of inundation in different areas.
• Archives and books written by local authors to gather information on historical floods in
Port-Louis. Historical records (Chelin 1989) reveal that Port-Louis has witnessed many
floods occurring during cyclones and torrential rains. The most severe past floods and their
impacts (Toussaint 1973; MOW 1996) in the city are summarised in Table4.
3.2 Site reconnaissance
A site survey was also carried out at an initial stage of the project to be well versed with the study
area and to locate, on site, the water levels observed in photographs and videos.
3.3 Interviews
Useful information such as the highest water levels reached during the flood and possible causes of
inundation were obtained by interviewing the local residents, shop owners, contractors and others.

3.4 Land surveying


The land surveying comprised the following:
• Recording ground and flood marks' levels from Pouce Stream next to Majestic cinema to the
esplanade Bissoondoyal of Port-Louis waterfront.

8
Table 4: Historical floods in Port-Louis since 1840
•High rainfall intensity that caused most streams of the capital to get flooded.
1840 •Company’s Garden flooded by up to 1 m of water.

•Torrential rains that caused flooding in the vicinity of the Pouce stream and Butte à Tonniers canal
1865 •La Chaussée Street and the Company’s Garden flooded by up to 1.5 m of water.
•30 deaths were recorded.

•Storm that caused serious flooding in the region of the Company’s Garden and La Chaussée Street
1888

•2 major floods occurred in that year.


•La Poudrière Street flooded by up to 1.2 m of water.
1896 •La Chaussée Street was also seriously flooded.
•Main bridges along Pouce stream and Butte à Tonnier canals collapsed.

1904 •Area surrounding La Chaussée Street was inundated up to a depth of 1.2 m.

1929 •2 occurrences of floods were recorded, with consequential damage.

•Pouce stream and Butte à Tonnier canals were completely submerged.


1995 •Depth of inundation recorded at La Poudrière and La Chaussée Streets were 1.5 m and 1.0m
respectively.

•2 major floods occurred on 13th February and 30th March, the latter being more severe.
•On both occasions, Place d'Armes & Caudan area were submerged.

•Flood of 30th March:


•11 deaths caused, most victims trapped in the water retained in the 2 subways at Place d'Armes
and the underground parking of the Harbour Front building.
•Rainfall intensity of 91 mm/hr recorded by the MMS.
2013 •High tide also recorded at time of peak rainfall.
•Pouce stream almost entirely blocked by debris & domestic wastes near the Company Garden.
•Depth of inundation recorded at La Poudrière and La Chaussée Streets were 2.0 m and 1.0m
respectively.
•At Caudan outlet, Pouce Stream was partly blocked by debris from road constructions.
•Caudan flyover (near Ruisseau des Créoles) & Ring road motorway were transformed into a
large water drainage system, with vehicles trapped in the water.
•Floodwaters were finally evacuated into the sea over the esplanades of Port-Louis waterfront.

• Noting topographic levels of the esplanade Bissoondoyal and that along Classic cinema to
obtain their ground profiles.
• Recording ground and flood marks' levels in Caudan Waterfront.
• Noting topographic levels in the region of Place d'Armes.

9
Figure 4: Pictorial survey of drains under study

• Survey of canals to obtain their cross-sections and slopes at different locations for
subsequent carrying capacity calculations. An assessment of the drains’ roughness was
carried out visually.

4. DATA ANALYSIS
Different techniques were used to process and analyse the gathered data. The main findings from
the investigation and brief discussions on the results obtained are given below.

4.1 Flood recurrence


Table 4 shows that at least 8 serious floods occurred in Port Louis over a period of 173 years from
1840 to 2013, i.e. an average occurrence every 20-25 years.

4.2 Determination of peak discharge – direct method


Table 5 gives the flood estimate, Qp ,in the study area by the Rational method.
Table 5:Peak discharge from catchments under study

For a duration equal to a tc of 30 minutes, the return period for a rainfall intensity of 91 mm/hr was
found to lie between 10-25 years (see Table 3).
10
4.3 Determination of canal carrying capacity
The different canals’ cross-sections surveyed are illustrated in Figure 4. Corresponding survey data
such as the canal dimensions and slopes were substituted into the Manning's formula to estimate the
carrying capacity, Q, of the canals at those locations. These are given in Table 6.

Table 6: Estimated carrying capacity of canals under study at different locations

4.4 Determination of peak discharge- indirect method


The Esplanade Bissoondoyal and that along Classic Cinema were assumed to behave as broad-
crested weirs during the flood, whereby excess flow from the Pouce and Butte à Tonniers canals
was being discharged into the sea. Thus, the indirect approach to estimate Qp consisted of plotting a
flood profile on that stretch, to determine the head and hence the flow, as illustrated in Figure 5.
Qp was finally calculated as the summation of the discharges obtained over the weirs and the
carrying capacities of the drains at the outlet of the watershed.

Table 7: Flow estimates at Caudan

11
Hence the total flow over the weirs during the flood was estimated as:
Ʃweirs= 229 m3/s
From Table 6, the carrying capacity of Pouce Stream (merging with Butte à Tonniers canal) at its
outlet was estimated at 182 m3/s whilst that of Ruisseau des Créoles was 89 m3/s. Thus, the total
flow that these canals could carry during the flood was:

Ʃdrains= (170+89) m3/s


Ʃdrains = 259 m3/s
Hence, from indirect method,
Qp = (Ʃdrains + Ʃweirs)
Qp = (259+229) m3/s
Qp = 488 m3/s

5. RESULTS AND DISCUSSION


5.1 Validation of estimated peak discharge
This section compares, Qp obtained from both the direct and indirect methods. The previous section
shows that:
• Qp from Rational method = 224 m3/s
• Qp from indirect approach = 488 m3/s > 224 m3/s

The estimated discharge from the indirect method should normally be less than that obtained from
the rational method (Lumbroso and Gaume 2011), which is not the case here. This discrepancy
might be due to the following limitations:

Inaccurate carrying capacities of drains


Firstly, the carrying capacities of the drains were overestimated since they were assumed to be
smooth under normal conditions. In reality, during floods, they get rougher due to the accumulation
of debris and other obstructions. Secondly, channel alignment, irregularities and meandering which
affect the roughness coefficient (Chow 1959) were also ignored. It was also difficult to measure the
slopes of the canals accurately since they were not uniform. Thirdly, the drains were assumed to be
empty when calculating the carrying capacities. In fact, large volumes of the channels are occupied
by stagnant water at rising tides, and debris (see Figure 7).

Taking these issues into account, the worst case is when the canals are blocked completely by the
stagnant water, debris and other obstructions. The overflow is then equal to:
(Qp from indirect method -Ʃdrains) = (488-259) = 229 m3/s

This overflow being very close to the runoff estimated from the Rational method indicates that the
drains functioned poorly during the flood of March 2013.
Inaccurate flood levels
Some water levels were measured based on statements from witnesses. However, people tend to
overestimate water levels due to the turbulent nature of the flood. Moreover, flood levels observed
on photographs and videos might correspond to different flows at different periods of the flood.
Furthermore, during the topographic survey, a straight alignment could not be established owing to
the high urbanisation in Port-Louis. Hence, the flood levels had to be transferred perpendicularly
onto a virtual alignment drawn on a map on AutoCAD software. All of the above might have
12
induced errors in the flood profile drawn, notwithstanding that the value of Cd= 1.6 adopted for weir
discharge calculations was based on literature. Ideally, Cd must be derived from laboratory
experiments. Thus, this may have added some inaccuracies in the flow estimates.

Figure 5: Ground and water profiles from Majestic cinema to outfall at Esplanade Bissoondoyal

Observations from graph:

1. Based on the equation of the flood line, the head of water over the weir (Chainage = 535.5 m)
was calculated as H= 0.73 m.
2. Place d'Armes region lies below the esplanade level.

5.2 Adequacy of drains


Calculations were performed to estimate the carrying capacities of the three channels under study,
as given in Table 6. Two important checks were performed on the drains’ carrying capacities. These
are whether:
• QOUT ≥ Qp
• Downstream flow ≥ Upstream flow
The results from Table 6 show that this is not always the case. Thus, the different streams studied
have inadequate capacities in specific location, e.g. around Chaussée bridge and the Company’s
garden for the Pouce stream. As a result, unless remedial action is taken, the Pouce stream will
always overflow at this location, when the discharge exceeds a certain value – even if it is free of
debris as in Figure 6 and is not blocked as in Figure 7.
5.3 Debris in Pouce stream
Figures 6 and 7 show contrasting sections of the Pouce stream. While the first conclusion would be
that the trash refers to the garbage thrown in by inhabitants, a more reasonable interpretation is that
the high water discharge flowing on the roads and within premises, during heavy rains, have carried
litter bins and their contents in the water courses (e.g. Pouce Stream). These deposit – earth, rock
and bags – on the shallower parts of the stream, reinforced by the columns acting as good props,
which then seriously impede the flow.
13
Figure 6 :Pouce Stream below hawkers' stalls under normal conditions

Columns in Pouce canal causing accumulation of debris,


thereby blocking the channel significantly

Figure 7: Pouce Stream below hawkers' stalls after flood of March 2013
SEA
Esplanade
Bissoondoyal

Place d'Armes

CAUDAN
Old
BASIN
Mokabridge

President John
Kennedy Street

KEY
Chausséebridge
Direction of overflow G Butte à Tonniers
canal
Bridge H

M Majestic Cinema
M
G Company Garden
Rue Des
H Hawkers' stalls Rochesbridge

Low- Low-lying area (ponding) Pouce


Stream

Figure 8: Routes taken by flood as a result of overflow from Pouce Stream and Butte à Tonniers
6. RECOMMENDATIONS
On 30th March 2013, 152.6 mm of rain fell in the city of Port-Louis, resulting in a fatal flood. The
most significantly flooded places were found to be those downstream three canals, namely: Pouce
Stream, Butte à Tonniers and Ruisseau des Créoles. These are long-existing channels that drain the
lower parts of Port-Louis and discharge into the sea at the Caudan Basin, as illustrated in Figures 1
and 8. Extensive background information on the study area revealed that there has been no major
improvements to the archaic drainage system in Port-Louis during the last 150 years, whilst massive
urbanisation has taken place. Historical data since 1840 also indicated that Port-Louis has been
flooded many times in the past, on average every 20 years. It was therefore suspected that there has
always been a deficiency in the drainage network of Port-Louis. Thus, this study aimed at assessing
the effectiveness of the three main canals mentioned above during high intensity rainfalls.
The study has shown that the Pouce Stream, Butte à Tonniers and Ruisseau des Créoles canals have
an inadequate carrying capacity at several places. These canals overflowed during rainfall events
with return periods of 20-25 years and are therefore expected to get flooded with higher flows. In
other cases, even when the canals are capable of evacuating certain flows, obstructions in the
channels cause blockages that significantly reduce the carrying capacities which results in flooding.
Taking into consideration the fact that Port-Louis is already a built-up area, the possibility of
channelling storm water from the canals to new drains is also very limited. In view of the above, the
following flood mitigation measures are proposed:

6.1 Enlargement of existing canals


This can be achieved by either increasing the depth, width or slope of the canals. Re-profiling the
canals, especially near the outlets, may not be sufficient since the downstream invert levels will be
governed by the sea and tides. The widening of existing canals is a more feasible solution since the
depth of the channels is also limited by the level of existing roads and bridges. This may entail the
temporary demolition of certain roads and buildings. Another possibility is a general raising of
levels – roads, pavements, etc.

6.2 Further studies


The problem of stagnant water in the canals due to tides remains unsolved. The equation commonly
used to determine the carrying capacity of a channel is the Manning's equation, which is applicable
when the channel is empty. The influence of the channel's linings on the carrying capacity is known
but that of stagnant water should be investigated. Therefore, further research on this special topic is
recommended.
Furthermore the study could be expanded to the whole catchment of Port-Louis, instead of limiting
to the three canals investigated. Presently, flooding occurs with a 20-25 year return period.
Unless more stringent design rules are adopted, to feel safer in the future, with an increase in the
carrying capacity of the canals, to cater for flows with a safeguard level of a 1 in 100 years return
period at least, (given that there is a 7 % chance of a 1000 year flood occurring during a man’s
lifetime), some may take an appointment to watch the next flooding in some 20-25 years’ time:
“Port Louis is flooding ! Port Louis is flooding” !!

16
REFERENCES
 Beirami, M., Nabavi, S., and Chamani, M., 2006.Freeoverfall in channels with different cross
sections and sub-critical flow.Iranian Journal of Science and Technology, Transaction B,
Engineering [online], 30(1). Available from:
www.shirazu.ac.ir/en/files/extract_file.php?file_id=493 [Accessed 21 September 2013].
 Chelin, A., 1989. Maurice: Une île et son passé. Editions du CRI, Ile de la Reunion.
 Chow, V.T., 1959. Open Channel Hydraulics. Tokyo: McGraw-Hill.
 Claydon, J., 2011. The Broad Crested Weir [online]. London, Claydon.
http://www.jfccivilengineer.com/broad_crested_weir.htm [Accessed 28 August 2013].
 Fairhurst, L., Rowswell, P., AndJochlinger, I., 2011. Sub-Saharan African Cities: A five-City
Network Pioneer Climate Adaptation through Participatory Research and Local Action
[online]. Cape Town, ICLEI. Available from:
http://www.resilientafrica.org/Resources/Final%20Resources/ICLEI%20Africa_5%20City%20
Adaptation%20Network_Port%20Louis%20Baseline%20Study.pdf [Accessed 26 January
2014].
 Hilbert, P., 2013. “Quand la modification des drains fait des morts”. Le Défi Quotidien, 4 April,
p.4.
 Hilbert, P., 2013. “Risques d’inondations : L’urbanisation à outrance pointée du doigt ”. Le
Défi Quotidien, 4 April, p.4.
 Ilmiaty, R.S., SUSANTO, R.H., SETIAWAN, B., SURYADI, F.X., AND BONAR, G., 2012.
International Journal of Biological, Ecological and Environmental Sciences. Drainage effects of
Sangkuriang Indah residential area on condition existing drainage system in water shed
Borang river Palambang city. [online], 1(4). Available
from:http://psrcentre.org/images/extraimages/IJBEES014032.pdf [Accessed 26 January 2014].

 Kassamally, R., 2013. Port-Louis Surface water flooding. Mauritius: Kassamally Engineering
Consultant, (SWF-PL-30-03-2013).
 Le Dimanche/L’hebdo, 2013. “Six noyés aux Caudan : Le tunnel de la mort’’ Le
Dimanche/L’Hebdo, 31March, p.10.
 Lumbroso, D., And Gaume, E., 2012. Reducing the uncertainty in indirect estimates of extreme
flash flood discharges. Journal of Hydrology [online], 414-415(16-30). Available from:
http://eprints.hrwallingford.co.uk/604/1/HRPP558-Uncertainty_in_flash_flood_prediction.pdf
[Accessed 30 August 2013].
 Ministry Of Works (Mow), 1996. Investigation of flood prone areas in Port-Louis by the fact
finding committee. Mauritius: MOW.
 NCDENR., 2007. Stormwater BMP Manual [online]. North Carolina, North Carolina
Department of Environment and Natural Resources. Available from:
http://portal.ncdenr.org/c/document_library/get_file?uuid=a85cdd45-7f1f-44d8-9921-
babea9478516&groupId=38364 [Accessed on 16 September 2013].
 Ponce, V., no date. Q&A on the return period to be used for design[online]. Mexico, Ponce.
Available from: http://returnperiod.sdsu.edu/ [Accessed on 5 February 2014].
 Proag, V.. 1995. The Geology and Water Resources of Mauritius. MGI, Mauritius.
 Proag, V., 2013. Drain design for dry feet. Journal of The Institution Of Engineers Mauritius,7-
18.

17
 Raghunath, H.M., 2006. Hydrology:Principles, Analysis, Design. 2nded. New Delhi: New age
international (P) Limited Publishers.
 Rantz, S.E., 1982. Indirect Determination of PeakDischarge [online]. United States, U.S.
Geological Survey (USGS). Available
from:http://wwwrcamnl.wr.usgs.gov/sws/fieldmethods/Indirects/rantz.htm [Accessed 30
August 2013].

 Salmasi, F., Poorescandar, S., Dalir, A., And Zadeh D., 2012. Discharge Relations for
Rectangular Broad-Crested Weirs. Journal of Agricultural sciences [online], 17(8). Available
from: http://tarimbilimleri.agri.ankara.edu.tr/2011/17_4/7_makale[Accessed on 18 September
2013].
 Sanral, 2006. Drainage Manual [online]. Pretoria, SANRAL. Available
from:http://www.citg.tudelft.nl/fileadmin/Faculteit/CiTG/Over_de_faculteit/Afdelingen/Afdeli
ng_Bouw/_Secties/Sectie_Weg_en_Railbouwkunde/_Leerstoelen/Leerstoel_Wegbouwkunde/-
_Onderwijs/-_College_Dictaten/doc/Drainage_Manual.pdf [Accessed on 28 August 2013].
 Seebaruth, J., 2013. Public demand for the resignation of Anil Bachoo as Minister.L’Express, 5
April, p.8.
 Smajstrla, A., And Harrison, D., 2002. Weirs for open-channel flow measurement. [online].
Florida, Institute of food and Agricultural Sciences. Available from:
http://edis.ifas.ufl.edu/pdffiles/AE/AE10800.pdf [Accessed on 23 September 2013].
 Smithers, J., 2012. Methods for design flood estimation in South Africa [online]. South Africa,
University of KwaZulu-Natal. Available
from:www.ajol.info/index.php/wsa/article/download/81042/71267[Accessed on 13 September
2013].

 Subhasish, D., And Kumar B.R., 2002. Hydraulics of free overfall in triangular channels.
Sadhana[online], 27(15). Available from:
http://www.ias.ac.in/sadhana/Pdf2002Jun/Pe1005.pdf [Accessed on 28 September 2013].
 Suriya, S., 2010. Studies on Chennai Drainage System-Rivers, Canal, Creeks, Estuaries, lakes
[online]. Tamil Naidu, Anna University. Available
from:http://www.cmdachennai.gov.in/pdfs/SeminarOnWaterways/7.pdf [Accessed 1 February
2014].
 Theodore, G., Thompson, D., And Fang, X., 2011. Use of the Rational and Modified Rational
Method for hydraulic design. [online]. Texas, Texas Tech University. Available from:
http://library.ctr.utexas.edu/hostedPDFs/TechMRT_0-6070-1.pdf [Accessed on 9 September
2013].
 Thompson, D., 2006. The Rational Method. [online]. Texas, Texas Tech University. Available
from: http://www.drdbthompson.net/writings/rational.pdf [Accessed on 9 September 2013].
 Toussaint, A., 1973. Port-Louis: A Tropical City. London: George Allen &UnwinLimited.
 UDFCD, 2007. Drainage Criteria Manual(V.1): Runoff [online]. Denver, Urban Drainage and
Flood Control District. Available from:
http://www.udfcd.org/downloads/pdf/critmanual/Volume%201%20PDFs/050%20Chapter%20
05%20Runoff%202008-04%20Rev.pdf[Accessed on 21January 2014].
 UDFCD, 2008. Drainage Criteria Manual(V.1): Major Drainage [online]. Denver, Urban
Drainage and Flood Control District. Available from:
http://www.udfcd.org/downloads/pdf/critmanual/Volume%201%20PDFs/070%20Chapter%20
07%20Major%20Drainage%202008-04%20Rev.pdf[Accessed on 21January 2014].
18
 UN-HABITAT, 2011. Mauritius: Port-Louis Urban Profile [online]. Kenya, UN-Habitat.
Available from:
http://www.unhabitat.org/pmss/listItemDetails.aspx?publicationID=3384[Accessed 21 January
2013].
 Weaver, J., 2003. Methods for Estimating Peak Discharges and Unit Hydrographs for Streams
in the City of Charlotte and Mecklenburg County, North Carolina [online]. North Carolina,
U.S. GEOLOGICAL SURVEY. Available from:
http://pubs.usgs.gov/wri/wri034108/pdf/report.pdf [Accessed on 25 September 2013].
 Weekend, 2013. “Caudan Waterfront: L’hécatombe”. Weekend, 31 March, p.9.
 Weekend, 2013. “ 152.6mm de pluies enregistrés en moins de cinq heures à Port-Louis”.
Weekend, 31 March, p.18.
 Weekend, 2013. “ Alan Ganoo : Les autorités sont responsables, elles n’ont pas étè à la
hauteur !”. Weekend, 31 March, p.19.
 White, T., 1952. A Master Plan for Port-Louis Mauritius. Cape Town, University of Cape
Town.
 Zbynek, Z., Bohm, P., Parilkova,J., Safar,R., And Sulc, J., 2013. Shape of the nappe during free
overfall from a rectangular channel with zero bed slope. Journal of Hydrology and
Hydromechanics [online], 3(61). Available from:
http://147.213.145.2/vc_articles/2013_61_3_Zachoval_222.pdf [Accessed 25 September 2013].

19

View publication stats

You might also like