MEM Pipeline Project

A project to create a distribution pipeline network sourced from the Emwaniro Primary School well. [Explain and summarize further here]



The project began on a suggestion from Khwisero's Member of Parliament [Name] at the opening ceremony for the Emwaniro well in [month] 2008. He initially suggested that he could provide funding (through Khwisero's CDF program) for the system's water tower, tank and electric pump if EWB-MSU would cover the cost of installing piping and distribution taps at schools.

In spring 2008, three senior design groups in MSU's Civil Engineering program (one of which contained long-time member Chris Allen) developed preliminary designs for the project using rough GPS data gathered in 2008. [find a way to provide a link where those designs can be found] Those designs envisioned the project encompassing three phases:

Phase V - Summer 2009

During summer 2009, these designs were presented to the community and some assessment work continued. During this time, Emutsasa Dispensary was added to project due to its proximity to Emutsasa Primary School. The head teachers from the schools on Phase I and II of the project had formed a "special committee" for the purpose of raising funds to house Team 2.

In early meetings with this committee during the summer of 2009, let by JJ Larsen, it became clear that its members expected implementation of the project to begin immediately. EWB team members communicated that summer 2009 would be devoted to assessment work for the project with the intention of returning to implement its first phase in Dec. 2009 or Summer 2010.

Fall 2009 and Winter Assessment Trip

With Larsen remaining in Kenya through Novemeber, a group of EWB members organized by Eric Dietrich began to look at the possibility of beginning construction on the project in Dec. 2009. Fairly quickly, it became clear that more technical data, particularly more accurate elevation survey data of the proposed route, was necessary before the project's design could be finalized. In addition, concerns over the social feasibility of the project given the necessity of separate schools collaborating in its management lead to the conclusion that more social work needed to be done to help the community prepare itself for handling the project's management.

Over MSU's winter break, a team composed of Eric, Matt Smith, Hilary Fabich and Joe Thiel traveled to Khwisero. Using research-grade GPS equipment (composed of a base station unit and reciever) the team recorded elevation data along the three phases at sufficient detail (+/- 5 ft) to carry out hydraulic calculations. In addition, the team met with Muhammad Ali of Haikal Investments, determining that the firm had the capacity to construct the system.

Although commuting (by hour-plus walk) from Jackson's compound each day, they also began to work with the schools and local community members to facilitate the formation of a management structure for the project. Initially, the team met with a "special committee" composing of head teachers and institutional representatives (which had initially been founded before Summer 2009 to host Team 2), but chose to pursue the development of a new organization after it became clear that the special committee did not see itself as a management structure for the project.

In addition, it was discovered that Namasoili Primary School, though included in the designs for the system's third phase, had not yet been involved in the project (and had not contributed funds to hosting Team 2 in summer 2009). At this point, representatives from the school were brought on board.

The Mulwanda, Emutsasa, Mundekue (MEM) MEM Committee was founded during a community meeting at Mundeku Primary schools [on date?], with community members deriving the name from the three sub-locations the project was intended to serve portions of. Under the supervision of a local assistant-chief elections were held (but ultimately proved less-than successful due to the meeting's low turnout). EWB-MSU team members attempted to present the committee with a structure involving finance and communications sub-committees to encourage effective outreach and fiscal management efforts, but


[Departure—proposed timeline and set of requirements; poor communication due to time crunch]

More information can be found (hopefully) on the pipeline assessment trip page page.

Spring 2010

Phase V - Summer 2010

People Who've Been Involved


*Chris Allen - PM during 2008, Involved in CE senior design work
*JJ Larsen - PM during 2009
*Eric Dietrich - Work as travel team member in 2009, project lead throughout 2009-10 school year, PM on winter assessment trip, project lead in 2010.
*Kalen Ramey - Involved with project throughout 2009-10 school year, project manager of pipeline teams in 2010.
*Chris Maus - Travel team member in 2010, project lead from Fall 2010 onward.

Kenya Team Members and Others
*Ronald Omyonga
*Jackson Nashitsakha
*Patrick Otwoma
*Wellington Lanya
*Stephen Olieka
*James Anzabwa

MEM Officials & Members
William Arthur Ashioya - Initial Chairman
Wycliff Lukale - Mundeku SMC Chair, Host father for Teams 1 & 5 during Phase V
Billah Owiso - Initial Treasurer
Auggrey Oluchiri - Initial Secretary

*MP Evans

Lessons Learned

[A significant amount of stuff]

Pipeline Documents

MEM Pipeline Water Project Completion Report, January, 2012
This report is the final document that the contractor, Haikal Investments, provided after the project was completed.
Download the PDF here


Design Data Compendium

This has a detailed report on the design process and phases of construction for the MEMU Pipeline.

MEM system cost breakdown

View here

material cost estimates

Technical Information [currently copied from 525 report—need to summarize, reformat tabular information, update for design changes from summer 2010]

The design’s first iteration was created as a senior design project by members of MSU’s civil engineering program in the spring of 2008 using rough elevation data collected with recreational handheld GPS units. However, questions about the validity of the profile data and the community’s desire for the system to include access points along the route necessitated a re-design. The current design was created using water demand estimates and more exact location and elevation data collected during the assessment trip this past winter.
The distribution system will pump water from the previously installed well at Emwaniro primary school using a submersible pump Grundfos SP8A-7. The water will be pumped into a 32,000 L, 3-meter water tower located at Emwaniro, constructed using a local design. The water will gravity feed from the tower through to the rest of the system.
As noted above, the system will be constructed in three phases. Although only the first phase is planned for construction this coming summer, hydraulic calculations were carried out for all three phases to ensure the design for phase I has the capacity for expansion.
Demand Estimates
EWB-MSU utilized several methods to estimate the demand for water along the proposed pipeline. Estimates were made on both the consumption of water at the institutions (i.e. schools or health clinics) and at the kiosks. The following section details the methods and data used to make these estimates.

I. Review of Similar Projects
To estimate community demand, similar projects were analyzed. The Khwisero North Water Project, recently installed near Khwisero Market, provided the most useful information. The project includes two 30,000 liter water towers that supply water to four kiosks. The management committee for this system provided the following estimates for consumption at each kiosk during the dry season.
The Khwisero Market: 100,000 liters a month (about 3300 liters a day). This market is the largest in Khwisero (probably double or triple the size of the one on the proposed distribution system route) and provides an estimate for high population density water use.
By FSA: 23000 liters a month (about 800 liters a day) — This is near the market by the local government offices and provides an estimate for consumption near larger residential areas.
Past SEK Khwisero: 600/ day — another residential area.
Final kiosk: 200/day — another residential area.

II. School Water Use Accounting Records
EWB‐MSU also looked into the water sold at the schools where wells have already been installed. The most successful of these, Ikomero P.S. and Emwaniro P.S., had very complete records that were very useful in determining water demand. Records indicate that the community peak water usage during the dry season is 2000‐2500 L of water per day.

III. Water Use Surveys
Each of the institutions along the pipeline was given a water use survey to fill out. The survey contained questions about the population around the institution, their estimated water use for the institution, their expectations as to how readily available water would change consumption and their estimates of community consumption.
IV. Application to Proposed Distribution System
The above analyses lead to the conclusion that 1,500 L/day at each primary school and 1,000 L/day at each kiosk is needed.
The system will be designed with distribution points at four primary schools, one secondary school, two health clinics and one market. Kiosks for community use will be used as additional distribution points and will be located at each primary school as well as along the route. To ensure sufficient water supply, a factor of safety of two was applied to the estimated demands, and the factored demands are shown in Table 1 below. In addition to demand data, Table 1 also summarizes the elevations of each distribution point and the distances between points.
Table 1. Site elevations, length of pipe needed between distribution points and anticipated water demand throughout the proposed water distribution system. Note that, as in other tables in this section of the report, phases II and III are included in calculations for purpose of gauging the feasibility of future expansion only.
Phase Location Elevation Distance From Previous Point (m) Anticipated Demand (L/Day) Anticipated Average Demand (L/s)** Peak Anticipated Demand (L/s) Phase Total Demand
1 Emwaniro Primary 1453 0 3000 0.035 0.104
Emwaniro Primary Kiosk 1453 NA 2000 0.023 0.069
Kiosk 1 1450 665 2000 0.023 0.069
Mundeku Primary 1440 911 3000 0.035 0.104
Mundeku Primary Kiosk 1440 NA 2000 0.023 0.069
12000 L
2 Kiosk 2 1421 634 2000 0.023 0.069
Ebuyonga Primary 1427 694 3000 0.035 0.104
Ebuyonga Kiosk 1427 NA 2000 0.023 0.069
Kiosk 3 1414 486 2000 0.023 0.069
Emutsasa Primary 1415 464 3000 0.035 0.104
Emutsasa Kiosk 1415 NA 2000 0.023 0.069
Health Clinic 1408 331 2000 0.023 0.069
16000 L
3 Khumutibo Market 1438 792 5000 0.058 0.174
Kisosk 4 1437 638 2000 0.023 0.069
Namasoli Health Center 1429 571 3000 0.035 0.104
Namasoli Health Kiosk 1429 NA 2000 0.023 0.069
Kiosk 5 1416 439 2000 0.023 0.069
Namasoli Secondary 1409 355 2500 0.029 0.087
Namasoli Primary 1402 240 3000 0.035 0.104
Namasoli Primary Kiosk 1402 NA 2000 0.023 0.069
21500 L
Ultimate Daily System Total= 49500 L
Selection of Pump
A Grundfos SP8A-15 pump was suggested by Haikal Enterprises, the firm that initially drilled the well. However, after looking at the manufacturer's pump curve and efficiency curve, it was determined that the Grundfos SP8A-7 was the ideal pump for our situation (calculations are attached in Appendix F). This pump will supply the approximately 35 m of necessary head at a flow rate of 11.3 m3/hr and be close to the maximum efficiency of the pump. Since this flow rate is in excess of the 10 m3/hr flow rate recommended by the test pumping report conducted by Joseph Siboe of the Ministry of Water and Irrigation, the pump will be used intermittently.
Selection of Storage Tanks & Water Tower
The system consists of two types of distribution points: kiosks and school distribution points. At each school, there will be a water tank for water supply to the school and a kiosk without a water storage tank to supply water to the community. The level of water in the tank at the school will be controlled by a float valve. The kiosks will be constructed the same way as the kiosks that are located at midway points along the distribution line.
In order to provide backup water supply in case problems are encountered and the pump has to be shut off, it was decided that the system would be designed such that each school would have one day's worth of water in a storage tank on site. In addition, the water tower would be designed to hold one day's worth of the un-factored demand on the entire system. The kiosks were not designed with water storage tanks, since they are not as critical as the water supply at the schools.
HDPE Kentanks will be used at the schools for storage. A steel water tower, elevated 3 m off the ground, will provide the necessary head for the system. The pump specified in the previous section (the Grundfos SP8A-7) will be used to fill the water tower at Emwaniro.
The water tower will be constructed to a locally approved design provided by the contractor. Similar designs are widespread in the region and EWB-MSU team members felt utilizing local knowledge would be a more effective design option to bringing in an outside design.
Control System
Storage tanks will contain float valves to prevent overfilling. The water tower will contain a float valve to local specifications in order to control the pump.
The program EPANET was used to model the system. Due to the size of the file used,it could not be attached to this document. A copy can be obtained by e-mail from moc.liamg|usmbwe#moc.liamg|usmbwe.
Within EPANET, a model was created with distribution points at each of the schools and kiosks. The elevations for the distribution points were based on survey data from the winter assessment trip, and the water tower height was modeled at 3 meters based on the available prefabricated water tower system options. The water tower was modeled as a constant head reservoir in EPANET and each of the distribution points was simply a node with a demand. The system was connected using sch40 PVC with a roughness of 0.007 mm. To account for head loss at fittings a loss coefficient of 2.0 was used for pipes between distribution points. The program uses the Darcy-Weisbach equation for head loss.
The figures below show the demand patterns used at the schools and the kiosks. The average flow rate for each tap was determined over a 24-hour time period, and the demand pattern accounts for the fact that water will only be used over a 15-hour time period. Since the actual water use pattern for the region is unknown, two different demand patterns were used at schools and kiosks. At the schools, it was assumed that the demand would be twice as high during early morning, noon, and late day. At the kiosks, it was assumed that water flow would be constant over the 12 hours during which the kiosk will be open.

Figure 1: Typical School Demand Pattern Figure 5: Typical Kiosk Demand Pattern
Once the model was created, different flow conditions were analyzed to check for potential failures:
• Two critical conditions were examined to check maximum and minimum pressures. First, the static pressure for the system was determined by analyzing the pressures under no flow conditions.
• The maximum water pressure that sch40 pvc can handle is 93.5 m of head, and the maximum encountered in the system was 55 m.
• The minimum expected pressures were determined by analyzing the system under four times maximum flow conditions.
• The minimum pressure in the system was 4 m of head.
The schematic does not include the storage tanks at each school, which do not affect the hydraulics of the distribution system and will help to buffer the demand at the schools. The system was analyzed at the base demand using two different time-use patterns (one pattern for kiosks and one pattern for schools). A minimum flow of 0.167 L/s was desired since at this flow rate a 20 L bucket could be filled in 2 minutes. The pipes used will be 102.26 mm (4") for phase I and 77.93 mm (3") sch40 pvc for phases II and III.
While not included in the EPANET model, valves will be used at junctions to ensure that if a problem were encountered with a part of the system, that portion could be shut off and repaired without disturbing the use of the rest of the system.
The tables on the following pages summarize the input demands and the pressures output by the program EPANET.

Figure 6: EPANET Model Schematic

Table 2: Demands and Pressures in Pipes
Network Table - Links at 7:00 Hrs (peak demand)
Phase Location Link ID Piping Distance (m) Outside Pipe Diameter (mm) Inside Pipe Diameter (mm) Flow (LPS) Velocity (m/s) Unit Headloss (m/km) Friction Factor
1 Water Tower - Emwaniro Junc Pipe 1 10 114.3 102.26 1.15 0.14 0.27 0.028
Emwaniro Junc-Kiosk1.1 Pipe 3 675 114.3 103.26 1.03 0.13 0.23 0.030
Kiosk1.1 -Mundeku Junc Pipe 4 911 114.3 104.26 0.98 0.21 0.77 0.028
Emwaniro Junc - Emwaniro PS* Pipe 24 10 88.9 77.93 0.07 0.01 0.01 0.106
Mundeku Junc -Mundeku PS* Pipe 25 10 88.9 77.93 0.07 0.01 0.01 0.106
2 Mundeku Junc -Kiosk2.1 Pipe 5 634 88.9 77.93 0.37 0.08 0.14 0.036
Kiosk 2.1-Ebuyanga Junc Pipe 6 694 88.9 77.93 0.32 0.07 0.11 0.038
Ebuyanga Junc-Kiosk2.2 Pipe 7 486 88.9 77.93 0.21 0.04 0.05 0.037
Kiosk2.2-Emustasa Junc Pipe 9 464 88.9 77.93 0.16 0.03 0.02 0.030
Emustasa Junc-HC Junc Pipe 11 331 88.9 77.93 0.05 0.01 0.01 0.089
Ebuyanga Junc- Ebuyanga PS* Pipe 26 10 88.9 77.93 0.07 0.01 0.00 0.000
Emustasa Junc -Emustasa PS* Pipe 27 10 88.9 77.93 0.07 0.01 0.01 0.106
Health Clinic Junc-Heath Clinic Pipe 28 10 88.9 77.93 0.05 0.01 0.00 0.000
3 Mundeku Junc-Khumutibo Market Junc Pipe 14 792 88.9 77.93 0.50 0.10 0.24 0.033
Khumutibu Market Junc-kiosk3.1 Pipe 16 638 88.9 77.93 0.38 0.08 0.15 0.036
Kiosk3.1-NamasoliHC Junc Pipe 17 571 88.9 77.93 0.34 0.07 0.12 0.037
NamasoliHC Junc-Kiosk3.2 Pipe 19 439 88.9 77.93 0.22 0.05 0.05 0.039
Kiosk 3.2-Namasoli S Junc Pipe 20 355 88.9 77.93 0.17 0.04 0.03 0.031
Namasoli S Junc-NamasoliPS Junc Pipe 22 240 88.9 77.93 0.12 0.02 0.01 0.035
Namasoli HC Junction-Namasoli HC* Pipe 30 10 88.9 77.93 0.07 0.01 0.01 0.106
Namsoli S Junc- Namasoli S* Pipe 31 10 88.9 77.93 0.06 0.01 0.00 0.000
NamasoliPS Junc-Namasoli PS* Pipe 32 10 88.9 77.93 0.07 0.01 0.01 0.106
*These links are minor pipes which connect the schools distribution points to the main distribution line

Table 3: Demands and Pressures at Nodes
Network Table - Nodes at 7:00 Hrs (peak demand)
Phase Location Elevation (m) Change in Elevation (m) Demand (LPS) Length along phase (m) Static Pressure (m) Residual Pressure (m)
1 Emwaniro WaterTower 1458 0 -1.15 0 0 0
Emwaniro Kiosk 1454 4 0.05 0 4 4
Emwaniro PS 1454 4 0.07 0 4 4
Kiosk1.1 1451 7 0.05 675 7 6.84
Mundeku Kiosk 1441 17 0.05 1586 17 16.14
Mundeku PS 1441 17 0.07 1586 17 16.14
2 Kiosk2.1 1422 36 0.05 634 36 35.05
Ebuyonga Kiosk 1428 30 0.05 1328 30 28.97
Ebuyonga PS 1428 30 0.07 1328 30 28.97
Kiosk2.2 1415 43 0.05 1814 43 41.94
Emutstasa Kiosk 1416 42 0.05 2278 42 40.93
Emutstasa PS 1416 42 0.07 2278 42 40.94
Health Clinic Junction 1409 49 0.00 2609 49 47.93
Health Clinic 1409 49 0.05 2609 49 47.93
3 Khumutibo-Market 1439 19 0.12 792 19 17.95
Kiosk3.1 1437 21 0.05 1430 21 19.85
NamasoliHC Kiosk 1430 28 0.05 2001 28 26.78
Namasoli HC 1430 28 0.07 2001 28 26.78
Kiosk3.2 1417 41 0.05 2440 41 39.76
NamasoliS Junction 1410 48 0.00 2795 48 46.75
Namasoli SS 1410 48 0.06 2795 48 46.75
NamasoliPS Kiosk 1403 55 0.05 3035 55 53.75
Namasoli PS 1403 55 0.07 3035 55 53.75

Table 4: Distribution System Summary
Water Tower
Location: Emwaniro PS
Size: 32,000L
Height: 3m
Type: Sch40 PVC
Size: 4" Phase I
3" Phase II & III
Model Information
Pipe Roughness: 0.007 mm
Minor Losses: loss coef. of 2 for each stretch of pipe
Equation for head loss: Darcy-Weisbach
Highest Static Pressure: 55 m (Namasoli PS)
Lowest Pressure: 4m (Emwaniro PS)
Checking Peak Velocity
Once peak flow was determined (Table 5), peak velocity was determined by dividing the peak flow by the cross sectional area of the pipe. Dimensions were determined from published values. Peak flow velocity was compared to published values in order to check that peak flow velocity was within allowable limits and that the maximum pressure and head loss at the determined velocity were acceptable.
Table 5: Flow rate calculations
Anticipated Volumetric Flow Rate and Flow Velocity In Phase 1 Mainline
Total Daily System Demand 49500.000 liters
Total Daily Demand After Emwaniro 44.500 m3/Day
Average Flow Rate Over 10 Hrs 0.074 m3/min
Average Flow Rate Over 10 Hrs 0.001 m3/sec
Peak Instantaneous Flow Rate 0.004 m3/sec
Peak Instantaneous Flow Velocity 0.452 m/s
Peak Instantaneous Flow Velocity 1.481 ft/s
Peak Instantaneous Factor 3.000
Interior Diameter Schedule 40 PVC 4.026 inches
Interior Diameter Schedule 40 PVC 0.102 meters
Cross Section Area 0.008 m2