The Yopougon-Koweït neighborhood is located in southern Côte d’Ivoire, between longitudes 4˚3'0'' and 4˚3'36'' West and latitudes 5˚18'15'' and 5˚19'20'' North (Figure 1). It covers an area of 1.7 km² and has an estimated population of 46,932 [11]. Yopougon-Koweït is bordered to the north by the Toit Rouge neighborhood, to the south by the Ebrié lagoon, to the east by Santé, and to the west by Camp Militaire. It is subdivided into five sub-neighborhoods. Johannesburg neighborhood to the north, Ravin neighborhood to the northeast, Konan Ferrand and Sante extension neighborhoods to the southwest, and Sub-Koweït and Gouro neighborhoods to the southeast. Figure 1 shows the different neighborhoods of Abidjan and the Yopougon-Koweït neighborhood.

Figure 1. Presentation of the study area.

Yopougon-Koweït has an equatorial climate, marked by four distinct seasons: two rainy seasons (March to July, with a peak in June, and October to November) and two dry seasons (August to September and December to February) (Figure 2). The climate is generally humid, with average monthly temperatures ranging from 24.2˚C in August, the coolest month, to 27.4˚C in March, the hottest month [12].

Figure 2. Ombrothermic diagram at the Adiopodoume station.

The data used in this study consists of demographic data [11] and AEP data for the years 2010, 2015, and 2022. This includes the number of subscribers, annual water consumption volumes, and a layout plan of the network showing its characteristics (pipe diameters and lengths). A differential GPS, a portable flow meter (Sewerin), and pressure sensors (Cello 4s) were used to collect the natural terrain elevations, distribution flows, and network pressures, respectively. The network was modeled using EPANET software. This simulation of the network’s operation involved the use of Mensura 12 software to link the elevations of the natural terrain to the network nodes and EpaCAD software to convert DXF files to INP format.

Populationestimate

The population of Yopougon-Koweït was estimated based on geometric population growth. The method used was trend extrapolation (Equation (1)). The years concerned are 2022, 2027, 2032, and 2037.

With:

P₀: Population in the initial year

P_n: Population in the year considered as the deadline

Ф: Growth rate

n: Deadline

Averagedailydemand

The average daily requirement (m³/day) or average daily consumption is the amount of water consumed by all active subscribers over a 24-hour period. In this study, requirements will be estimated using a specific consumption rate of 100 l/day/resident according to [13][14]. It is calculated using Equation (2) [15]:

With:

Q a . d : Average daily demand

Sc: Specific consumption

N: Population

Assessmentoftheincreasedaveragedailydemand

The average daily increase is calculated based on drinking water requirements, taking into account leaks in the network and waste. To do this, an increase is applied according to Equation (3) [15]:

With:

Q a . d . S u r p : Surplus average demand

α : Surplus coefficient expressing water losses.

Estimated water losses vary from 15% to 50%. In this study, a water loss coefficient of 30% is used in relation to age and network maintenance rate, as recommended [16].

Averagehourlydemandandpeakhourlyflow

The average hourly requirement is given by Nesrine’s Equation (4) [17]:

With:

Q a . h : average hourly flow en m³/h

Q m a x . d : maximum daily flow en m³/day

The peak hourly flow corresponds to the consumption flow rate during the hour of maximum network demand. It is given by Nesrine’s Equation (5) [17]:

With:

Q m a x . h : Maximum hourly flow

Q a . h : Average hourly flow

K m a x . h : Maximum hourly irregularity coefficient, which can be broken down into two other coefficients that depend on the characteristics of the urban area, namely α m a x and β m a x .

The maximum hourly irregularity coefficient is determined using Equation (6) [17]:

With:

α m a x : Coefficient taking into account the level of comfort and amenities. It varies between 1.2 and 1.4.

β m a x : Coefficient based on the size of the urban area (Table 1).

Table 1. Values of β_max as a function of the number of redents.

2.3.1. Simulation of the Operation of the Yopougon-Koweït Water Supply Network

Modeldesign

The physical characteristics of the pipes (diameters, lengths) extracted from the site plan in shapefile format in ArcGIS were integrated via conversion to DXF. A topographic survey of elevations using differential GPS enabled the generation of a Digital Terrain Model by interpolation. This made it possible to superimpose the network imported in DXF and assign elevation coordinates to the nodes using Mensura 12 software. EpaCAD software was used to convert the DXF file to INP format compatible with EPANET. The final configuration in EPANET consisted of materializing the structures (reservoirs, pumps) and specifying the simulation parameters.

Modelcalibration

Calibration is a step without which a model is meaningless. Its purpose is to ensure that the model is representative and can be used to understand and predict the behavior of the network outside of normal operating conditions. Calibration is the final step in modeling a drinking water network [18]. This step aims to adjust the model’s pressure parameters so that the model’s results are as close as possible to reality. In this study, calibration consisted of installing six (06) sensors over a period of 24 hours at a rate of one sensor per sub-neighborhood. At the end of this stage, pressure calibration was performed. This involved integrating the data recorded (pressure and flow rates) at the sensors into the model to achieve a perfect match between the pressure data simulated by the model and that recorded in the field. The results are considered satisfactory for a correlation coefficient close to 1 and an RMSE of less than 10 [7].

2.3.2. Sizing of Water Supply Structures

Pumpsizing

The sizing of hydraulic pumps involves the total manometric height (TMH) (Equation (7)), which takes into account the geometric height (Equation (8)) and pressure losses (Equation (9)):

Avec:

H g e o : Geometric height;

Δ H : Pressure drop in the supply pipe;

Δ P : Pressure variation between suction and discharge point (bar);

With:

λ : Pressure drop coefficient;

V : Flow velocity (m/s);

D : Pipe diameter (m);

L : Length (m);

p : pressure (bar);

g : gravity (m∙s⁻²).

Determinationofthehydraulicpowerofpumps

Hydraulic power represents the energy transferred to the pumped fluid to increase its velocity and pressure in the RePEP. It is calculated using Equation (10) [19]:

With:

ρ : Fluid density (kg/m³);

g : gravity (environ 9.81 m/s²);

Q : Flow (m³/s);

η : Pump efficiency (on prend 80%);

HMT: Total Manometric Height (m).

Reservoirsizing

In this study, the storage tank capacity is sized to hold 30% of the total daily demand, which is within the recommended range of 25% to 50% for large communities. Indeed, due to the lack of reliable statistics on supply systems in West Africa, [20] recommends using 30% of the daily peak flow rate for tank sizing.

Determinationoftheelevationofthefoundationslab

The base elevation refers to the minimum piezometric elevation required at the head of the network to guarantee flow rates and operating pressures at the various nodes [21]. It is determined by the Equations (11) and (12):

With:

H i m p i : Initial height imposed (m);

Z a v a l i : Initial downstream slope (m);

p d c : Pressure drop (m);

P m i n i m a l   s e r v i c e i : Minimum operating pressure.

The minimum operating pressure is calculated using the Equation (13):

With Equation (14):

Reservoirsizing

The dimensioned tank is cylindrical in shape. Its diameter is calculated by setting an arbitrary height for the tank. This is given by Equation (15) as follows [22]:

D_ins: Inside diameter (m);

V: volume (m³);

h: arbitrary reservoir height (m).

Determinationoftheouterdiameter

e v : veil thickness (30 cm).

Optimizationofthedistributionnetwork

Optimizing the drinking water network will make it possible to adjust pressure and flow velocity to ensure a sustainable supply for the residents of Yopougon-Koweït. The parameters have been set to maintain: 1) pressure between 1 and 6 bars and 2) flow velocity between 0.5 and 1.5 m/s, in accordance with the standards authorized for an optimal network.

The daily water demands of the population of Yopougon-Koweït are estimated at 4829 m³/day for a population of 48,293 residents in 2022. These demands will be 9640 m³/day by 2037, representing an anticipated satisfaction rate of 80%. An increase in this demand gives 6278 m³/day in 2022 and 9640 m³/day in 2037. The water consumption rate of subscribers is estimated at 2407 m³/day in 2022, representing a water demand satisfaction rate of 38%. This satisfaction rate will fall to 25% in 2037, a decline of 13%. Average daily production is set at 2407 m³ (Table 2).

Table 2. Drinking water demand.

The calibration of the hydraulic model indicates that the correlation coefficient between the observed and simulated values is 0.97, while the root mean square error (RMSE) is 8.17 (Table 3).

Table 3. Correlation report between simulated and observed pressures.

Figure 3(a) confirms the strong linear correlation (r = 0.97) between simulated and observed values. The vast majority of points are aligned along the first bisector. Furthermore, a comparison between simulated and observed data indicates overall agreement between them (Figure 3(b)).

Figure 3. Comparison of values. (a) Comparison of simulated and observed values; (b) Linear correlation between simulated and observed values.

3.3.1. Size of Pumps

The pumps (P1 and P2) have been sized with total manometric heights (HMT) of 68 m each. They will be positioned at the various entrances to the RePEP in the Yopougon-Koweït neighborhood (P1 “military camp” and P2 “Koweït pharmacy”). They will deliver 100 l/s and 70 l/s respectively. Pump (P1) will use 83,017 W of power compared to 58,112 W for pump (P2) to propel water into the RePEP (Table 4).

Table 4. Characteristics of sized pumps.

3.3.2. Size of Water Reservoir

The reservoir in the Yopougon-Koweït neighborhood has been designed with a capacity of 3854 m³. The inner and outer diameters are 26 m and 27 m, respectively. The depth of the foundation slab is estimated at 43 m (Table 5).

Table 5. Water reservoir size.

Pressurevariationsinthenetwork

Variationat3a.m.,8a.m.,and5p.m.(pumpoption)

All pipes (100%) in the RePEP will have good pressure at 3 a.m. This pressure will be between 22 and 52 mWC (meter of water column) for all six sub-neighborhoods (Figure 4(a) and Figure 4(b)).

Figure 4. Pressure variation on the Yopougon-Koweït RePEP (pump option). (a) Pressure variation on the RePEP at 3 a.m.; (b) Pressure distribution at 3 a.m.; (c) Pressure variation on the RePEP at 8 a.m.; (d) Pressure distribution at 8 am.; (e) Pressure variation on the RePEP at 5 p.m.; (f) Pressure distribution at 5 p.m.

At 8 a.m., pressure variations in the pipes will peak at 63 mWC. The majority of pipes (88%) will have good pressure (15 and 60 mWC). Only 12% of pipes will have pressure above 60 mWC. The majority of neighborhoods (Johannesburg, Ravin, Gouro, Sante Extension, Sub-Koweït, and Konan Ferrand) will have good pressure. However, a small portion (2% of pipes) of the Santé Extension (Sikasso) and Konan Ferrand neighborhoods will have estimated pressure of 62 mWC (Figure 4(c) and Figure 4(d)).

At 5 p.m., most pipes (82%) will have good pressure between 10 and 60 mWC. However, 18% of RePEP pipes will have pressure above 60 mWC. The Johannesburg, Ravin, and Gouro neighborhoods will have good pressure. However, the Konan Ferrand neighborhoods and certain areas of the Sante Extension and Sub-Koweït neighborhoods will have estimated pressure of 68 mWC (Figure 4(e) and Figure 4(f)).

Variationat3a.m.,8a.m.,and5p.m.(reservoiroption)

Almost all of the pipes (95%) in the RePEP will have a pressure between 39 and 60 mWC at 3 a.m. Only 5% of the pipes will have a pressure between 60 and 64 mWC. All sub-neighborhoods of Yopougon-Kowët will have good pressure. However, part of the Sub-Koweït sub-neighborhood will have high pressure above 60 mWC (Figure 5(a) and Figure 5(b)).

Figure 5. Pressure variation on the Yopougon-Koweït RePEP (reservoir option). (a) Pressure variation on the RePEP at 3 a.m.; (b) Pressure distribution at 3 a.m.; (c) Pressure variation on the RePEP at 8 a.m.; (d) Pressure distribution at 8 am.; (e) Pressure variation on the RePEP at 5 p.m.; (f) Pressure distribution at 5 p.m.

At 8 a.m., the pressure variation in the pipes will peak at 41 mWC. The majority of pipes (99%) will have good pressure (33 and 56.5 mWC). Only 1% of pipes will have pressure below 10 mWC. The neighborhoods of Yopougon-Koweït will have good pressure. However, a small part of the Sub-Koweït sub-neighborhood will have low pressure below 10 mWC (Figure 5(c) and Figure 5(d)). Almost all pipes (98%) will have pressure between 10 and 45 mWC. However, 2% of pipes will have pressure below 10 mWC.

At 5 p.m., almost all pipes (98%) will have a pressure between 10 and 45 mWC. However, 2% of pipes will have a pressure below 10 mWC. All Yopougon-Koweït will have good pressure. However, a small part of sub-Koweït will have low pressure (Figure 5(e) and Figure 5(f)). The variation in pipe velocities will reveal that, on average, 72% of pipes will have a velocity between 0 and 0.5 m/s, 12% will have a velocity of zero, and 16% will have a velocity greater than 1.5 m/s.