Water from Dry Riverbeds - IRC...ASAL Consultants Ltd. P.O. Box 739, Sarit 00606, Nairobi, Kenya [email protected] [email protected] Fax/Tel: 254 (0) 202710296 Mobiles: 0733

Water from Dry Riverbeds

How dry and sandy riverbeds can be turned into water sourcesby hand-dug wells, subsurface dams, weirs and sand dams

Women drawing water from anunlined waterhole in a sandy riverbed.

Water being drawn from a hand-dugwell on a rivcrbank upstream ofa subsurface dam.

A weir constructed across Talek river in Maasai Mara was extended to be a sand dam toprovide domestic water for a tourist camp. The lush vegetation seen upstream is due tothe weir having raised the water table in the riverbanks.

Erik Nissen-Petersenfor

Danish International Development Assistance (Danida)2006

Water from Dry Riverbeds

How dry and sandy riverbeds can be turned into water sourcesby hand-dug wells, subsurface dams, weirs and sand dams

Women drawing water from anunlined waterhole in a sandy riverbed.

Water being drawn from a hand-dugwell on a riverbank upstream ofa subsurface dam.

A weir constructed across Taiek river in Maasai Mara was extended to be a sand dam toprovide domestic water for a tourist camp. The lush vegetation seen upstream is due tothe weir having raised the water table in the riverbanks.

Erik Nissen-Petersenfor

Danish Internationa] Development Assistance (Danida)2006

Technical handbooks in this series :

Titles1 Water for rural communities2 Water supply by rural builders3 Water surveys and designs4 Water from rock outcrops5 Water from dry riverbeds6 Water from roads7 Water from small dams8 Water from roofs

ContentsLessons learnt from Kitui pilot projectsProcedures for being rural contractorsSurvey, design and cost of water projectsRock catchment tanks, masonry and earth damsWells, subsurface dams, weirs and sand damsRainwater harvesting from roadsPonds and small earth dams built manuallyVarious types of roof catchments for domestic use

These handbooks can be obtained free of charge by either collecting them from the office ofASAL Consultants Ltd., or by paying through bank transfer the cost of sending the manuals bycourier. For further details, please contact: [email protected] with copy toasalconsultants @ yahoo, com

Published byASAL Consultants Ltd. forthe Danish International Development Agency (DANIDA) in Kenya

Text, photos and sketches byErik Nissen-Petersen

Computer drawings and tracing byCatherine W. Wanjihia

Editing byJoy Okech

Proofs byAmin Verjee and Steen S. Larsen

PrinterEnglish Press, P.O. Box 30127, Nairobi, Kenya

Website byEdwin Ondako

Distribution byASAL Consultants Ltd. P.O. Box 739, Sarit 00606, Nairobi, [email protected] [email protected]/Tel: 254 (0) 202710296 Mobiles: 0733 619 066 and 0722 599 165Website: www.waterforaridland.com

© CopyrightThe copyright of this handbook is the property of the Royal Danish Embassy In Kenya.Downloading from the internet and photocopying of the handbooks is permitted providedthe source is acknowledged.

Contents PageAcknowledgement ivAcronyms vForeword vi

Chapter 1. Riverbeds 11.1 Types of riverbeds 11.2 Water storage in sand 21.3 Water in riverbeds 31.4 Waterholes in riverbeds 41.5 Dowsing 51.6 Evaluation of riverbeds 5

Chapter 2. Survey of riverbeds 62.1 Probing 62.2 Probing of Mwiwe riverbed 72.3 Survey of Mwiwe riverbed 82.4 Plan and longitudinal profile of Mwiwe riverbed 92.5 Cross profiles of Mwiwe riverbed 102.5.1 The most suitable site for the Mwiwe intake 102.5.2 The most suitable site for Mwiwe Sand Dam 112.6 Trial pits and Survey Report 122.7 Plan and profiles of Nzeeu riverbed 132.7.1 The most suitable site for the Nzeeu intake 142.7.2 The most suitable site for Nzeeu Subsurface Dam 152.8 Volumes of sand and water in Mwiwe and Nzeeu riverbeds 16

Chapter 3. Design considerations 173.1 Sustainability and affordability 173.2 Waterdemand 183.3 Existing water yield 183.4 Three options for increasing water yield 183.5 Design decisions 19

Chapter 4. River intakes 214.1 Suitable seasons for surveys and construction works 214.2 Intakes in riverbeds 214.3 Intakes in riverbanks 234.4 Design and cost of a sinking well 244.5 Design and cost of a hydro-dynamic well head 254.6 Water lifts and pumps 274.7 Design and cost of large river intakes 294.8 Design and cost of infiltration pipes 304.9 Cost and capacity of pumps, generators and pump houses 324.10 Calculations and cost of rising mains and head tanks 334.11 Map showing pipes, tanks, valves and kiosks of Mbitini Water Project 34

Chapter 5. Subsurface dams built of soil 355.1 History of subsurface dams 355.2 Function of subsurface dams 355.3 Design and cost of Nzeeu Subsurface Dam 365.3 Guidelines on construction 37

Chapter 6. Weirs 396.1 Various types of weirs 396.2 Design and cost of Talek weir 416.3 Construction procedures 436.4 Maintenance of weirs 44

Chapter 7. Sand dams 457.1 History of sand dams 457.2 Functions of sand dams 467.3 Five types of sand dams 477.4 Criteria for construction of successful sand dams 487.4.1 Site criteria 487.4.2 Design criteria 517.4.3 Maintenance criteria 53

Chapter 8. Mwiwe Sand Dam 548.1 Yield of water from Mwiwe Sand Dam 548.2 Design calculations for Mwiwe Sand Dam 548.3 Design of Mwiwe Sand Dam 558.4 Construction procedures for Mwiwe Sand Dam 568.5 Bill of quantity and cost of Mwiwe Sand Dam 58

References 59

A small riverbed with a subsurface dam built of soilis flooded with rainwater run-off at Kibwezi, Kenya.

i l l

Acknowledgement

Sincere thanks are due to Birgit Madsen of the Royal Danish Embassy in Kenya who sawthe importance of financing the documentation of how to turn dry riverbeds intoperennial water sources in the semi-desert, arid and semi-arid zones of the world with agrant from the Danish International Development Agency (Danida).

Many thanks are also due to the persons whose agencies were interested in the subjectand financed training courses and utilisation of water sources in dry riverbeds. Theseagencies and ministries were;

> Danida with MoA and MoWI in Kenya froml978 to 2006.> Danish Television film Development in 1981.> UNDP/ILO/Africare/MoMW in Tanzania from 1990 to1992.> UNDP/Habitat in Myanmar (Burma) from 1995 to 1996.> Danida and BADC in Kenya in 1996.> CONCERN in Somaliland in 1997.> Sida and RELMA with MoA and MoWI in Kenya from 1997 to 2002.> CONCERN in Southern Sudan 2003.> South East Africa Rainwater Network (SearNet) with MoW in Eritrea 2004.> Danish Refugee Council (DRC) in Somaliland in 2005.> UNDP Somalia in Somaliland in 2005.> Royal Danish Embassy in Kenya from 2003 to 2006

The successful implementation of the above assignments was mainly due to the support,interest and co-operation of all the engineers, hydro-geologists, technicians, builders andcommunity groups that participated in these activities.

Special thanks are due to the team that assisted me in producing this handbook. The teamconsisted of Catherine W. Wanjihia, Civil Engineer Technician, who assisted with thedrawings, Joy Okcch, Assisting Information Programme Officer, who edited the draft,James Khamisi who kept the computers working, Amin Verjce and Steen S. Larsen whoproof-read the draft, Prof. Elijah K. Biamah who wrote the Foreword, the English PressLtd. who printed the handbook and Edwin Ondako who loaded the handbook onto theInternet.

Erik Nissen-PctcrsenManaging DirectorASAL Consultants Ltd.P.O. Box 739, Sarit 00606Nairobi, Kenya

IV

Acronyms

ASALASALCONBADCBQcmCONCERNcuDanidaDRC

gGLHardcoreILOITDGKmKVAKshKWKWSPmmmm2

m3

Max.MoAMoWINGOPVCRHLMARWSSSear.NetSidaUNDPUPVCY8Y10WL

= And and Semi-and Land= ASAL Consultants Ltd.= Belgium Assisted Development Council= Bill of Quantity= Centimetres= A British Non-governmental Organization= Cubic metres= Danish International Development Agency= Danish Refugee Council= gauge~ Ground level= Crushed stones= International Labour Organization= Intermediate Technology Development Group= • Kilometre= Kilo Volts Amphere= Kenya Shilling equal to US$ 72.00

Kilo Watt= Kenya Water and Sanitation Programme= Metre= Millimetre= Square metre= Cubic metre= Maximum ;

= Ministry of Agriculture= Ministry of Water and Irrigation= Non-governmental Organization= Poly Venyl Conduits= Regional Land Management Unit= Rural Water and Sanitation Programme= South East Africa Rainwater Network= Swedish International Development Agency= United Nations Development Programme= Ultra Poly Venyl Conduits= 8 mm twisted iron bar= 10 mm twisted iron bar= Water level

FOREWORD

In dry land areas of Kenya, episodic shortages of water are quite common. In many areas,women have to spend days traveling long distances with donkeys and camels looking forwater. Where the distances are too long, the people have been compelled to move with theirlivestock to the water sources. Due to the long distances covered searching for water,affected local people have had very limited productive time especially that devoted toagriculture and livestock production. Experiences in dry land areas have showed that whenthe water source is harnessed and conserved through watering points, local people andwomen in particular have more time to concentrate in other household chores and incomegenerating activities.

At present in most of Kenya, local people are compelled to live close to permanent watersources; agricultural activities are scattered and localized in transitional agro-climatic zones.Thus it is pertinent that the supply of water for domestic, livestock and supplementalirrigation uses be considered a top priority in managing hydrological and agriculturaldroughts in dry land areas. This supply should be followed by a well planned distributionnetwork of watering points which will meet the water requirements for various uses andminimize environmental degradation due to overgrazing or deforestation. Any developmentor rehabilitation of water supply schemes should aim to ensure reliable and adequate watersupply and sanitation.

The socio-economic consequences of inadequate water supply in dry land areas of Kenyaare costly and manifest themselves in low labor productivity, poor enrollment of children inschools, and livestock mortality due to drought and famine. Thus inadequate water supplyultimately decreases per capita water demand, and crop and livestock productivity.Consequently, water shortages reduce household water requirements, household food andwater security and incomes. The low incomes propagate the vicious cycle of poverty andfamine that must be overcome among affected rural communities.

The development of appropriate and affordable community water supply systems calls forinnovative rain and runoff water management technologies for domestic, livestock, andsupplemental irrigation uses. Some of the technologies that have proven to be effective andsustainable in water resource development and management in dry land areas include:runoff regulation and storage techniques (e.g. gully head dams, hafirs or waterpans, microdams, ground catchments (djabias), roof catchments, underground water storage tanks (shuijiaos), barkads, and macro and micro catchment systems; and ground water sources such ashand dug wells, shallow drilled wells, infiltration galleries of sand storage dams, and pipedwater from springs.

VI

Water from dry riverbeds as a hands on manual provides the requisite insights andinformation on how surface runoff that is released into seasonal rivers as flash floods canbe harnessed and abstracted through appropriate technologies such as hand dug wells,weirs and sand storage dams (sand and subsurface dams).

This hands-on handbook begins with a historical perspective/background information onsandy riverbeds as reservoirs of good quality water and then goes on to look into thesurvey and design considerations required before constructing hand dug wells, weirs andsand storage dams (sand and subsurface dams). This manual has also addressed theconcern over when to develop these water sources by giving the suitable seasons forsurveys and construction works. The latter information is important because experiencesfrom emergency relief operations in the arid lands of Kenya (i.e. Turkana) have showedthat the construction of sand storage dams in times of drought cannot provide the badlyneeded water to water deficit areas. Instead, these dams should be developed with a viewto being recharged during the rainy season and utilized in subsequent dry spells.

Another major contribution by this handbook to the development of water from dry andsandy riverbeds is that of providing the necessary details on design, bills of quantities andcosts, construction and maintenance procedures of the three water supply systems namelyhand dug wells, weirs and sand storage dams. Alongside this, this handbook gives thecosts and capacities of pumps, generator sets and pump houses as well as the calculationsand costs of water conveyance and distribution systems.

Indeed the information provided by this handbook will be of immense use to recentlyestablished (under the new Ministry of Water and Irrigation) water resource developmentand management institutions in Kenya such as the Water Resources ManagementAuthority (WRMA), the Water Services Trust Fund (WSTF), the Regional WaterServices Boards, and the grassroots based Water Resources Users Associations(WRUAs).

Prof. Elijah K. Biamah

Chairman/Head,

Department of Environmental and Bio-systems Engineering,

University of Nairobi,

P.O. Box 30197-00100,

Nairobi, Kenya

Email: [email protected]

vn

Chapter 1. Riverbeds

1.1 Types of riverbedsDry riverbeds are sandy and seasonalwater courses that transport run-offrainwater from catchment areas, alsocalled watersheds, into rivers or swampsonce or a few times in a year.

Dry riverbeds are also called ephemeralstreambeds, seasonal water courses orsand rivers. In Kenya they arc calledluggah and in Arabic wadi.

Most of the rainwater being transporteddownstream in riverbeds appears asflash-floods that can be several metreshigh, and that thunder downstream withthe sound of a steam locomotive.

Flash-floods can uproot trees and othervegetation growing on riverbanks andfields. Homes, villages and towns mayalso be flooded and people, livestock,buildings and bridges washed away bythe muddy maelstroms of flash-floods.

Riverbeds may be classified into 4classes of potential water sources asfollows:

1) The most potential riverbeds havehilly and stony catchments that producecoarse sand where up to 350 litres ofwater can be extracted from! cu.m. ofsand, i.e.an extraction of rate of 35%.

2) Gullies originating from stony milshave a potential for sand dams consistingof medium coarse sand where 250 litresof water can be extracted from 1 cu.m.of sand, i.e. an extraction rate of 25%.

3) Riverbeds having catchments of flatfarmland usually contain fine texturedsand, that can only yield a maximum of100 litres of water from 1 cu.m. of sand,i.e. an extraction rate of 10%.

Vj^g^-^25P

4) Stony riverbeds containing bouldersand fractured rocks have the lowestpotential for water extraction due toseepage caused by the boulders andfractured rocks. This seepage is,however, beneficial for replenishment ofboreholes situated on riverbanks.

1.2 Water storage in sandSand consists of small stone particlesthat originate from stones and rocksbeing broken down by the effects ofsunshine, rains and temperaturevariations.

Voids, which are empty spaces, arealways found between sand particles.When dry riverbeds are flooded by rainsand flash-floods, the air in the voids ispressed out by the water because it isheavier than air.

When a dry riverbed is being flooded, itlooks as if the riverbed is boiling as tensof thousands of air bubbles are beingpressed out of the sand. This process isknown as saturation.

Fine textured sand has tiny voids thatget saturated slowly with water. Onlyabout 10% of water can be extractedfrom the volume of fine sand.

Coarse textured sand has larger voidsand is therefore saturated much quickerthan fine sand. The volume of water thatcan be extracted from coarse sand isabout 35% of the volume of sand.

Silt and sand extractability was testedand classified as follows:

SizemmSatu-rationWater-extrac-tion

Silt

<0.5

38%

5%

Finesand

0.5 to1.0

40%

19%

Me-diumsand1.0 to1.5

41%

25%

Coar-sesand1.5 to5.0

45%

35%

Therefore, much more water can beextracted from riverbeds containingcoarse sand than from riverbeds withfine textured sand.

No water can be extracted fromriverbeds containing silt, such as in sanddams whose spillways were built instages higher than 30 cm.

Water storage in sand has severaladvantages, such as:

1) Evaporation losses are reducedgradually to zero when the water level is60 cm or more below the sand surface.

2) Livestock and other animals cannotcontaminate the water reservoir becauseit is hidden under a surface of dry sand.

3) Mosquitoes, and insects that carrywater-borne diseases, cannot breed inunderground water reservoirs. Frogs,snakes and other unpleasant reptiles andanimals are also unable to live in, and toand pollute, water reservoirs in sand.

Sand testingThe porosity and extractable capacity ofsand are found by saturating 20 litres ofsand with a measured volume of water.

The water is thendrained out of thecontainer andmeasured byremoving a plugfrom the bottom ofthe container.

1.3 Water in riverbedsSince time immemorial, riverbeds haveprovided water for people and animals.During extreme droughts, when all otherwater sources have dried up, water canstill be found in riverbeds.

Elephants, ant-eaters and some otherwild animals have a special sense bywhich they can locate such places wherewater is found in riverbeds.

Some rural people and most well-diggersknow that water can only be found atcertain places in riverbeds. They canalso give a rough estimate on how deepthey have to dig before reaching thewater-table.

Their knowledge is based on the fact thatcertain species of trees and vegetationmust have roots reaching down and intothe water-table in order to survivedroughts. This traditional information iscompiled in the table below.

Water-indicating vegetationBotanical name

CypcrusrotundusVangueriatomentosaDelonix elata

Grewia

MarkhamiahildebranditiHyphaenethebaciaBorassusflabellierferFicuswalkefieldii

Kiswahili &Kikambanames

KitndiuMuiruKikomoa

MwangiItiliku Itiliku

MuuChyooKikoko Ilala

MvumoKyatha

Mombu

Depthtowater-level3 m to7m5 m to10 m5 m to10 m7 m to10m8 m to15m9 m to15m9 m to15 m9 m to15 m

Ficus natalensis

FicusmalatocapraGelia aethiopica

PiptadeniahildebranditiAcacia seyal

MuumoMuumoMkuyuMukuyuMvungunyaMuatiniMgangaMukamiMgungaMunini

9 m to15 m9 m to15 m9 m to20 m9 m to20 m9 m to20 m

Since the above mentioned trees musthave their tap root in the water table, thedepth of a water-table can be found byknowing the depth of the tree?s tap root.

A rule of thumb states that the tap rootof a tree has a depth equal to about ofthe height of the tree.The height of a tree can be found bymeasuring the length of the shadow thetree is casting on the ground andcomparing it with the length of theshadow of a stick 100 centimeters long.

The two measurements should be takenin the sunshine of early morning or lateafternoon when the shadows are longest.

For example: If the stick? shadow is 80cm long, the ratio is: 80/100 = 0.8.If the tree?s shadow is 12 m long, thenthe tree is: 12 m x 5 / 4 = 15 m high andthe tap root and water level is at:15 mx = 11.25 m depth.

1.4 Waterholes in riverbedsThe reason why some waterholes areperennial and others are not, lies underthe sand, where it cannot be observed.

1) The floors under the sand inriverbeds can consist of soil, clay,murram, black cotton soil, boulders,fractured rocks or solid rock bars.

2) Where floors consist of permeable(water-leaking) material, such as sandysoil, fractured rocks or large boulders,water will seep into the undergroundthrough the floor. This may be beneficialfor deep boreholes but certainly not forextracting water from riverbeds.

3) If a floor consists of an impermeable(water-tight) texture, such as clay,clayey soil or murram, there is noleakage through the floor. Water cantherefore be extracted from the riverbed,unless it has drained downstream andleft the riverbed dry.

4) Riverbeds have downstreamgradients because they function asdrainage channels for rainwater run-off.Water in riverbeds is therefore alwaysmoving downstream, either on thesurface or between the sand particlesunder the surface of riverbeds. Surfaceflow of water can easily be observedduring floods but the subsurface flowcan only be observed when water isscooped out of waterholes.

The reason why some riverbeds havewater and others have not, lies in thefloor under the sand in riverbeds withimpermeable floors.

If riverbeds have impermeable floors butno water, it is because the water hasbeen drained downstream by gravity.

If riverbeds with impermeable floorshave water, then something must havestopped the water from being draineddownstream. What could that be?

The answer is found by hammering ironrods into the sand of riverbeds at certainintervals. Such probing shows that mostriverbeds have a floor under the sandthat bulges up and down. These naturalbarriers, called dykes, give the answer.

- /

the dyke prevents the water seepingdownwards in the voids between thesand in the riverbed.

I..;..;

Where the floor bulges upwards it actsas an underground dyke that stops theunderground flow of water, as seen onprobing points no. 10, 13, 15, 19 and 23.

Where there are a depression in the floorit accumulates water, as seen on probingpoints no. 9, 14, 16 and 21.

These are subsurface water reservoirs inthe sand, from where water can beextracted.

1.5 DowsingGifted persons can use dowsing to locateunderground water sources andunderground dykes.

-100

The tool consists of a 1 m long brazingrod cut in two halves and each halfhaving a 12 cm long handle.

The two dowsing rods are held looselyand pointing downwards so that they canswing freely. However, the hands mustbe held steady to allow gravity to pullthe rods down while they are parallel.

Here is water. Here is no water.

When walking slowly over an unknownunderground water source, the rods willswing inwards. The force of the pullindicates the depth and volume of waterin the ground.

Although these traditional techniques arerather successful when applied by reliablepersons, more precise data is required byprofessionals.

1.6 Evaluation of riverbedsThe most potential stretches of a riverbedare identified by walking in them togetherwith the community members who haverequested for assistance to improve theirexisting water source or to construct newones in the riverbed.

First draw a sketch of the riverbed.

Then walk and dowse, if capable of that,in the riverbed while plotting thefollowing information on the map:

1) Location and types of water-indicatingtrees and vegetation.

2) Location of water-holes and theirdepth to the water-table and quality of thewater.

3) Location and types of rocks andboulders.

4) Location of calcrete, which is a saltywhitish substance that turns water saline.

5) Coarseness of the sand.

6) Location of hand-dug wells, boreholesand weirs in the riverbed.

7) Names of houses, schools and roadcrossings near the riverbed.

After having compiled the informationlisted above, the community is informed ofthe various possibilities for improvingtheir water supply. When an agreement ismade with the community, a detailedsurvey with probing is implemented forthe purpose of providing data for drawingdesigns and estimating the yield of water,and the cost of construction, operation andmaintenance.

Chapter 2. Survey of riverbeds

2.1 ProbingWhen the most promising lengths of ariverbed have been identified during theevaluation walk, they are probed usingprobing rods hammered into the sand.

The probing data is used for:

1) Drawing a plan and profiles of theriverbed to identify the deepest placefrom which water should be extractedand the most shallow place where thewall for cither a subsurface dam, a weiror a sand dam can be constructed.

2) Estimating the volume of sand inthe reservoir and the extractable volumeof water from the sand.

3) Providing the required data fordrawing the designs and estimating thecosts of construction.

The tools required for simple surveys asfollows:

1) Probing rods made of 16 mm (5/8?)iron rods for measuring depths of sand.

Notches should be cut in the probingrods for every 25 cm to bring sandsamples when the rods are pulled up.

2) A circularlevelling toolmade of atransparenthosepipe formeasuring thegradients ofriverbeds.

3) Two long tape measures, onehanging down vertically from thehorizontal one, to measure width anddepth of riverbeds.4) A tripod ladder for hammering

long probing rods into the sand.5) A mason hammer.6) A 20 litres jcrrycan with water.7) Half a dozen of transparent plastic

bottles with water.8) A knife and writing materials,9) The Probing Data Sheets shown on

the next page.

2.2 Probing of Mwiwe riverbedParts of the following contents are based on the actual implementation of the Mbitini andKisasi Water Projects situated about 30 km south of Kitui township in Eastern Kenya.

The Mwiwe riverbed, which is the water source for the Mbitini project, was surveyed inNovember 2004. The survey started where the Wingo riverbed joins the Mwiwe riverbed.The procedure was as follows:

1) A probing rod was hammered down in the middle of the riverbed until it hit the floorunder the sand with a dull sound. Then the level of the sand was marked on the rod and itwas pulled straight up without twisting. The following data was noted on the data sheet:

1) The depth of water is measured from the tip of the rod to the water indication mark.2) The depth of sand is measured from the marked sand level to the tip of the rod.3) The coarseness of sand is seen in the notches of the rod.4) The type of floor under the sand is seen at the tip of the rod.5) The width of the sand in the riverbed is measured.6) The height of the banks arc measured with two long tape measures.7) The presence of water-indicating vegetation, waterholes, roads, etc. is noted.8) The next probing is measured at regular intervals, for example 20 metres.

ProbingProbingnumber

123456789

1011121314151617181920212223

Data Sheet (m).Distancebetweenprobings

020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0020.0022.0022.00

Widthofriverbed

20.8024.2028.2025.5023.4030.5029.5033.0023.6223.6229.6032.9025.7020.0017.0026.0022.6918.4017.5020.0029.0030.0026.00

LocationDepth towater

No waterNo waterNo waterNo waterNo waterNo water

I No waterNo water

0.20No water

0.10No waterNo water

2.00No water

0.10No waterNo waterNo water

0.500.60

No waterNo water

: Mwiwe riverbedDepthof sand

0.500.250.280.320.450.850.761.001.250.501.000.590.553.000.751.250.930.500.501.751.750.880.45

Type ofsand

MediumFine

MediumMedium

CoarseCoarse

MurrainCoarse

MediumMediumMediumMediumMediumMediumMedium

CoarseCoarse

MediumCoarseCoarseCoarseCoarseCoarse

Date:Floorunder thesand

ClayClayClayClayClayRock

Soft rockClayClayClayClayClayClayClayClayClayClayClayClayClayClayClayClay

20/11/04Height ofriver banksLeft / Right1.50/ 1.801.30/ 1.601.30/ 1.701.42/1.841.30/1.651.32 /1.451.32/1.501.97/ 1.550.70/1.252.25/1.670.70/1.350.97/1.80

.33/

.50 /

.32/

.85/

.00/1.20/.20 /.46/.75/

.76

.68

.56

.60

.45

.53

.651.58.67

1.13/1.5811.20/ 1.45

Items seenon thebanks

Wingo

WaterholePawpaw

Acacia treeRock

WaterholeFig tree

Tele. PoleRoad

MukengekaKiindiu

WaterholesFence post

Orange treeOrange tree

MuninaFence post

MwangiFig tree

MusewaKiindiu

23 Survey of Mwiwe riverbed

The heights and shape of the riverbankswere measured using two long tapemeasures, one hanging vertically downfrom the other which was drawn tightlybetween the top of the riverbanks.

The gradient of the riverbed wasmeasured by a person standing atprobing point No. 1 while sightingdownstream using the two horizontalwater levels in a circular hosepipe thatwas filled hallway with water.

Other persons were standingdownstream at probing point No. 23while holding a long pole vertically.

The horizontal sighting line in thecircular leveling instrument hit the pole0.8 metres above the level of the sand.The difference of 0.8 m in heightbetween probing points No.l and No. 23being 440 metres apart, gives a gradientof: 0.8 / 440 m x 100 m = 0.18%.

Sand test

Samples of dry sand were taken andfilled in a 20 litres container for thepurpose of measuring the porosity andthe volume of water that can beextracted from the sand reservoir.

The saturation of the sand was reachedafter adding 8 litres of water, giving asaturation rage of 40% (8 / 20 x 100).

Then a small hole was made in thebottom of the container to drain thewater out of the sand. In one hour, 5litres of water was extracted from thesand. That gives an extractable capacityof 25% (5/20 x 100).

2.4 Plan and longitudinal profile of Mwiwe riverbedA plan and a longitudinal profile was drawn according to the probing data on a sheetof A3 mm graph paper which is 40 cm long and 28 cm wide.

The horizontal measurements were drawn to a scale 1:2000, which means that the440 metres length of the riverbed was drawn as 22 centimetres long and the 20 metreintervals between the probings were drawn as 1 centimetre long on the graph papers.

j i i l £ ' v v' v ft W^F^^yv y v t

160U0 120 10080 «0.40. 20S20 iOO 380 360 3«) 320 300 260 260 240 220DOWNSTREAMPlan of Mwiwe riverbed.

UPSTREAM

The plan shows that the probed river has a length of 440 m and width varying from17.5 m to 33.0 m. Water-indicating trees, e.g. Figs, Mwangi and Munina, grow along theriverbed. Waterholes having water 7 months after the last rains were located at probingpoint No. 10, 14 and 21 where the sand is deep. Water is trapped in the sand bydownstream dykes at points No. 11,18 and 23, as see on the longitudinal profile below.

DOWNSTREAMLongitudinal profile of Mwiwe riverbed.

UPSTREAM

The longitudinal profile on the former page shows that the sand is 3.0 m deep at point No. 14 and1.75 m deep at point No. 21. Since both places are holding water 7 months after the last rain theyare the best extraction points in that part of Mwiwe riverbed. The next phase in surveying theriverbed was to probe across at point No. 17 in order to learn of the volume of the depression andits water yielding capacity.

2.5 Cross profiles of Mwiwe riverbed

2.5.1 The most suitable site for Mwiwe extraction point

The riverbed was probed from bank to bank at 2 metre intervals at point No. 14. Toconfirm that point No. 14 was the deepest point, further probing was done berth upstreamand downstream from that point.

Probing data across the extraction point at No. 14 in Mwiwe riverbedProbingNo.

12345678

Distancebetweenprobings m

22222222

Depth of sand

m0.753.252.501.753.001.50

, 0.75NIL

Type of sand

Coarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sand

Type of floorunder thesandSandy claySandSandy claySandy claySandSandSandy claySandy clay

Ifr Sand level

LEFT BANK RIGHT BAN K.Probing profile of No. 14 in Mwiwe riverbed where the sand isdeepest and therefore the most suitable place for the intake.

10

2.5.2 The most suitable site for the dam wall of Mwiwe Sand Dam

The longitudinal profile and several cross probings showed that the underground dyke atpoint No. 18 was the most suitable for the Mwiwe Sand Dam as shown below.

ProbingNo.

01234567

7.5

Distance betweenprobings m

22222222

0.5

Depth ofsand m

0.851.351.301.701.301.601.401.100.40

Type of sand

Coarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sand

Type of floorunder the sandClayClayClayClayClayClayClayClayClay

'0-0 2 i, 6 8 10 12 K 16 16 m1 2 3 4 5 6 7 8 9 lOProbings

LEFT BANK RIGHT BANK

This cross profile shows the measurements of the most suitable underground dykeat No. 18 for Mwiwe Sand Dam. The design below was based on this profile.

r 3200r- - 750 -tW0| — 700 - 1150 — 1500-

Design of Mwiwe Sand Dam that was based on No. 18 cross profile

11

2.6 Trial pits

Trials pits were excavated down to Underground dykes can sometimes be locatedthe floor under the sand every 3 metres by vegetation growing in soil situated justto confirm the correctness of the below the sand surface,probing.

The surveyors prepared a Survey Report with all details and data compiled of Mwiweriverbed. The report .was handed over to a design engineer for his considerations whichare presented in the next chapter.

Thereafter the surveyors surveyed a much larger riverbed, Nzeeu, for 25,000 people inKisasi Water Project. The survey data is presented on the next pages.

On the left, a local expert isusing a pendulum to findwater under the ground.

On the right, a soda bottleis used for grading a sandsample

12

2.7 Plan and profiles of Nzeeu riverbed

SO 240 220 200 WO 160 140 120

L-oWINMKbAMPlan of Nzeeu riverbed

] f ? f U-4-4-4-420 100 80 60 40 20 0 20 £D 660 80 100 130 MO 160 180 m

U P S I K E A M

The surveyors started probing from the junction of Nzeeu and Kindu riverbed at point 0.Thereafter they returned to point 0 and surveyed upstream as shown on these sketches.

I?. . ¥ _ ¥ - ¥ - 7 - - - ? l - M . ' t • ? ' , ? • 1 P 1 f J M 1 1 1

^ ~ £ ^ " t z . '••• T °

•f-2'

- ! •

- - • 1

2t» iio l ioiui i io IOO JD sis ui A i ^ io BO ^B 4rfiA>440m

RIGHT BANKLongitudinal profile of the probing in Nzeeu riverbed.

LEFT BANK

The plan and profile show a large underground water reservoir at No. 3 and anunderground dyke with a narrow point at probing point No. 13 which is a perfectfoundation for either a subsurface dam or a sand dam.

13

2.7.1 The most suitable site for the Nzeeu extraction point

A cross profile of Nzeeu riverbed at probing point No. 3 on the longitudinal profileconfirms the large underground water reservoir from which water will be extracted.

Probing data across the deepest point at No. 3 of Nzeeu riverbedProbingNo.

123456789

101112131415


033333333333888

Depth ofsand

m0

2.503.50

4.50+4.50+4.50+4.50+4.50+4.50+4.50+

3.502.502.002.002.00

Type of sand

Fine sandMediim sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sand

Type of floorunder thesand

ClayClayNot reachedNot reachedNot reachedNot reachedNot reachedNot reachedNot reachedNot reachedNot reachedClayClayClayClay

r >f

«,$•$« 5? 5; :;t i t tS 3 3 * 3.

•- — ,. . ..U-L. . . • ..J^Ls

f—+—I—t- 4—\—H-in v i * t s , . . . « ? • • • . : : • * • • • • * • i 3 : * " » .

LEFT BANKRIGHT BANK

Cross profile of Nzeeu riverbed at point No. 3 (A-A) which will be the extraction point.

14

2.7.2 The most suitable site for the dam wall of Nzeeu Subsurface Dam

A cross profile of the underground dyke and narrow riverbed at point 13 confirms thelocation of the most shallow place for either a subsurface dam, a weir or a sand dam.

Probing data across the shallow dyke at No. 13 (B-B) of Nzeeu riverbedProbingNo.

1234567


0333333

Depth of sandm

1.501.502.001.751.751.501.00

Type of sand

Coarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sandCoarse sand

Type of floorunder thesandSandy claySandSandy claySandy claySandSandSandy clay

m1« 15 12 9 § 3 8 mRIGHT BANK LEFT BANKCross profile of Nzeeu riverbed at probing point No. 13 (B-B). The profilewas used for the design of the subsurface dam as shown below.

m181—i'.ilM i—i—i8 1512 9 t 3 00m

15

2.8 Volumes of sand and water in Mwiwe and Nzeeu riverbeds

The porosity and extractable percentage of water from the sand in Nzeeu riverbed wasfound to be 35% and 30% respectively.

The approx, volume of a dam reservoirs can be found by this rule of thumb:Maximum depth x maximum width x maximum throw-back / 6 = Volume.(A through-back is the horizontal length of water in a dam).

The volume of sand and water in the reservoirs of the two riverbeds were found to be:

MwiweNzeeu

Max. depthm

3.255.00

Max. widthm

25.7066.00

Throw-backm

40.00380.00

1/6

1/61/6

SandVolume m"f

55720,900

% Waterextraction

2530

Watervolume m3

1396,270

An unknown volume of subsurface flow of water in the sand replenishes the tworeservoirs which should be added to the volume of water shown above.

In order to increase the yield of water from Mwiwe riverbed it was decided to construct aa sand dam upon the dyke at No. 18, because that would increase the storage capacityfrom 139 m1 to 2,997mJ> as shown under WL 4 in the table below.

Three o

WL1WL2WL3WL4

ptions for increasing the volume of sand andMax. depth

m3.25

0.8 + 3.251.4 + 3.252.3 + 3.25

Max. widthm

25.7026.00^28.0030.00

Throw-backm

40.00260.00300.00360.00

water in Mwiwe riverbed1/6

1/61/61/61/6

SandVolume m3

5574,5636,5109,990

% Watcrextraction

25252730

Watervolume m3

1391,1411,7582,997

Three o

WL1WL2WL3WL4

ptions for increasing the volume of sand andMax. depth

m5.00

0.8 + 5.001.4 + 5.002.3 + 5.00

Max, widthm

66.0068.0070.0072.00

Throw-backm

380.00420.00440.00480.00

water in Nzeeu riverbed1/6

1/61/61/61/6

SandVolume m3

20,90027,60832,85342,048

% Waterextraction

30303030

Watervolume m3

6,2708,2829,856

12,614

Key:WL 1 - Existing natural reservoir in the riverbedsWL 2 = Increased volume by a subsurface dam wall built to 30 cm below sand levelWL 3 = Increased volume caused by a weir built to 30 cm above sand levelWL 4 = Increased volume caused by a sand dam wall built to 150 cm above sand level

A Survey Report was then given to the engineer to determine whether the dam should bea cheap subsurface dam or a low cost weir or the more expensive sand dam. Theconsiderations for this assessment are described in the next chapter.

16

Chapter 3. Design considerations

3.1 Sustainability and affordabilitySustainabilHy and affordability are two key words that should always be rememberedwhen designing water projects.

Sustainability of water projects can be achieved by designing tested and simplestructures that the community can construct, operate, maintain and repair themselves.

For example, initially it was planned to drill two boreholes and equip them withsubmersible pumps at Mbitini and Kisasi water projects at a total cost of about Ksh 3million. Instead, some local builders were contracted to sink two hand-dug wells withinfiltration pipes at a total cost of Ksh 425,000. all inclusive. Not only was theconstruction cost reduced by Ksh 2.5 million but it was much easier and cheaper for thecommunity to service the pumps in the hand-dug wells than it would have been in theboreholes.

Affordability is also achieved by designing tested and simple structures using amaximum of locally available skills and materials that the community can afford toprocure, construct, operate, maintain and repair. The example described above withboreholes versus hand-dug wells describes exactly what affordability is all about.

The supply capacity of a water project should be able to satisfy a community s waterdemand for the next 20 years. If the supply capacity is less than the demand, people willcomplain that their water supply is insufficient and they have wasted their money andtime. Where a water project is over-designed, the construction works and operation costsare higher than necessary and the extra cost has to be levied on the sale of water. Thesupply capacity and the water demand should therefore balance.

Another factor to consider is the population increase. When water is available, manypeople living outside the project area will be attracted to settle in the project area. Thesupply capacity must also include this population in addition to the normal increase.

Water is sold at kiosks iroin K.sh 2 to Ksh 4per 20 litre jerrycan. The price of waterdepends on the cost of operation, main-tenance and repairs.

A Kiosk Attendant s sales record ischecked. The income from sale of watermust be sufficient to pay for salaries,fuel, operation, maintenance and repairs.

17

3.2 Water demandWith reference to an appraisal report produced by Development Impact Consulting (DIC)in August 2004, the combined water demand for Mbitini is projected as shown in thetable below.

Daily Water Demand for Mbitini. The water demand by Kisasi is similar to Mbitini.Category

General populationLivestock unitsSchoolsHealth and Adm. CentresShopping centresTotal daily water demand (nrVday)Total annual water demand (nr'/year)Demand in 6 months without rains

mVday2004

16920115

12217

80,00040,000

m/day2005

17220115

14222

81.00041,000

mVday2015

21023145

18270

98,00049,000

m3/day2025

25626175

21325

118,00059,000

3.3 Water yield in year 2004

Volume of sand and water in the underground reservoirs

MwiweNzeuu

Max. depthm

3.255.00

Max. widthm

25.7066.00

Throw-backm

40.00380.00

1/6

1/61/6

Sandvolume m3

55720,900

% Waterextraction

2530

Watervolume m3

1396,270

3.4 Three options for increasing the water yieldsThe three available options for increasing the water yields are to construct dam walls foreither:

1) A subsurface dam built of soil to 30 cm below the surface of sand in a riverbed, or

2) a weir built of concrete or rubble stone masonry to 30 cm above the sand level, or

3) a sand dam built of concrete or rubble stone masonry to a maximum of 5 metresabove the level of sand in a riverbed.

These dam walls should be constructed on the identified underground dykes situateddownstream of existing underground water reservoirs.

In other riverbeds, the three types of riverbed dams should, preferably, be constructed onunderground dykes. The dams can, however, also be built in riverbeds not havingunderground reservoirs or underground dykes.

18

3.5 Design decisionsThe following decisions on the designs were taken on the basis of the evaluation reportand the considerations described above as well as during several meetings with the Kisasiand Mbitini project committees:

3.5.1 Decision on extraction point in Mwiwe riverbedAs mentioned above, it was found more sustainable and affordable to lay infiltrationpipes in the riverbed to drain water into a hand-dug well to be sunk in the riverbank,instead of drilling a borehole and equipping it with a submersible pump.

The extraction point will consist of 72 metres of infiltration pipes made of perforated 160mm UPVC laid in crushed stones as deep in the riverbed as the water level would allowat point No. 14 in Mwiwe riverbed. The infiltration pipes will be laid with a gradienttowards the riverbank to facilitate water flowing into a hand-dug well that will be sunk inthe riverbed.

The hand-dug well will have a diameter of 3 metres and be sunk as deep as the watertable will allow. The well will be built of curved concrete blocks reinforced together on acircular foundation/cutting ring of reinforced concrete.

Extraction of water from the well will consist of a surface pump powered by a dieselgenerator that will be installed in a pump house to be constructed.

The pump will deliver the water to a 100 m3 head tank to be constructed of concreteblocks and situated 0.96 km from the pump at an elevation of 80 metres above the pump.From this head tank water will gravitate through 2 water tanks and 22.9 km of pipelinesto 11 water kiosks.

3.5.2 Decision on a sand dam in Mwiwe riverbedFor Mwiwe riverbed it was decided to construct a sand dam of rubble stone masonry to aheight of 1.5 metres above the existing sand surface at probing point No. 18. The sanddam will raise the water table from 1.7 metres below the sand level to 1.5 metres above,which is a total of 3.2 metres. This will increase the storage capacity from the existing139 m3 to 2,997 m1 as shown under WL 4 (Water Level 4) in the former chapter.

After the infiltration pipes have been laid in the riverbed and the hand-dug well has beensunk as deep as possible, the water table in the riverbed will be raised by construction thesand dam in the required stages.

19

3.5.3 Decision on extraction point in Nzeeu riverbedThe extraction point in Nzceu riverbed will be a replica of the extraction point in Mwiweriverbed, except for the pump, which will deliver the water to a 50 m" elevated head tankto be constructed of steel plates. The tank will be situated 5.3 km from the pump at anelevation of 154 metres above the pump. From this head tank water will gravitate through2 water tanks and 16 km of pipelines to 10 water kiosks.

3.5.4 Decision on a subsurface dam in Nzeeu riverbedThe selected option for Nzeeu was a subsurface dam whose wall will be built of soil to 30cm below the existing sand surface, which will raise the water table 170 cm. This willincrease the water storage capacity from the existing 6,270 m3 to 8,282 m3'

The subsurface dam will have two purposes, namely, a) to increasing the yield of water,and, b) to compare its construction cost and performance with the sand darn at Mwiwe.

someSurvey and design data are being compiled into a Survey Report by Swiparticipants of a training course dealing with water from dry riverbeds.

20

Chapter 4. Riverbed intakes

4.1 Suitable seasons for surveys and construction works

Surveys of riverbeds should always be carried out in the driest seasons when water levelsare at their lowest. Information and mobilization of communities should also beimplemented in the dry seasons when water supplies are scarce and people are concernedabout their water supplies. Designs, bills of quantities, costs and work plans can be drawnup during rainy seasons.

Construction of water projects should always start with sinking of the wells and laying ofthe infiltration pipes and that should always take place during a dry season. By doing so,the minimum yield of water can be known and used for calculating the required capacityof the pump and generator, pipelines, water kiosks, etc.

The minimum yield is also used to determine whether the riverbed can supply sufficientwater or whether a subsurface dam, weir or sand dam has to be built to increase the yieldof water.If so, the construction work of such structures should take place in the beginning of shortdry season or in the middle of long dry season when there is no risk of an unexpectedthunder shower that can flood the construction site.

Excavation of trenches for pipes and the laying of pipes should also take place in the dryseasons when the fields have been harvested to avoid spoiling crops in the field. Otheractivities that demand much manual labour should also be implemented when the farmersare less busy in the dry seasons.

Another big advantage of constructing water projects during dry seasons is that theprovision of water is usually the biggest issue during those periods. Tt is therefore fairlyeasy to mobilize and organize people to carry out manual work.

More detailed information on survey and design is given in another handbook of thisseries called Water Surveys and Designs.

4.2 Intakes in riverbedsThere are several technologies for lifting water from riverbeds that are affordable andsustainable for individuals and community water projects.

The oldest and simplest method is to dig awaterhole in a riverbed and use a calabashcut in half to scoop water into a calabash or ajerrycan.

21

A safer and cleaner option is to sink a hand-dug well next to the waterhole in a riverbed.

Hand-dug wells situated in riverbeds must beequipped with a well-head to prevent the wellshaft being filled with sand during floods suchas this hydro-dynamic well-head used inSudan.

Another type of hydro-dynamic well-headshaped as a wedge to break the force offloods.

A well-shaft can also be lined with worn-outlorry tyres and closed by a lid made from thebottom of an oil-drum.

22

4.3 Intakes in riverbanksAlthough sinking of wells directly into riverbeds is the cheapest way to extract water, thewells might be damaged by unusually large flash-floods. It is therefore safer to buildhand-dug wells on the riverbanks and draw water from there provided the undergroundsoil is sandy soil that allows water to seep into the hand-dug wells.

If underground water cannot seep into a hand-dug well, an infiltration pipe can be laid in atrench to drain water into the hand-dug well.The design can be made cheaper by filling thetrench with stones covered with polythenesheeting instead of laying the infiltration pipe.

Well shafts can be lined with burnt brickswhen built upwards from the bottom of hand-dug wells. However, this procedure is riskybecause collapsing soil can bury the buildersalive.

A safer method is to build sinking wellswhereby curved concrete blocks are reinforcedtogether onto a foundation ring made ofconcrete. The foundation ring and infiltrationcourse are built at ground level in a riverbed.

Then sand is removed from the inside of theshaft causing it to sink into the sand. When theshaft has sunk to ground level, more blocksare added onto the shaft. Sand is then removedand the shaft sinks. The procedure is repeateduntil the shaft has reached its final depth.

Concrete culverts can be sunk in a similar waybut they arc more difficult to handle.

The cost of lining well shafts can be reduced byplacing a large perforated PVC pipe in the centre ofthe excavation and filling stones and gravel allaround it upto the ground level.

23

4.4 Design and cost of a sinking well

A tool made of iron for cutting grooves

Plan of well-shaft and curved concrete block

Section of well-shaft

A groove being cut lor a foundationring with the tool seen above. Thegroove will be filled with reinforcedconcrete.

Curved concrete blocks are made ofmortar in the ratio 1:5 and compactedinto a mould made of iron sheets.

It takes 16 blocks to make a circularcourse with a diameter of 140 cm and112 blocks to build 1 metre of shaft.

7LI11 mm rod M >

VA-A e 40 mmIran pip*

4.5 Design and cost of a hydro-dynamic well head

A hydro-dynamic well head can be built around a well shaft directly into riverbeds toprevent the well shaft being damaged or filled with sand from flash-floods.

y - - : • • - •

The lay-out of a hydro-dynamic well headin the middle of a riverbed.

Foundation being built of rubblestones reinforced with barbed wire.

A well Head is construcicu ol rubble stones An open steel cover the well shaftset in mortar, 1:4, and reinforced with barbed should be in the position shown towire and chicken mesh. allow flash-floods to close the lid if

people have forgotten to close it.

A hydro-dynamic well head in the middle of a riverbed.

25

Bill of quantity and cost of a wellDescription6 metre deep hand-dug well with ahydro-dynamicwell head

Labour costSupervisorContractorArtisanTraineesLabourersCost and value oflabourMaterialsCementWeldmeshG.I. wireChicken meshBarbed wireWindlassIron barRiver sandCrushed stonesRubble stonesWaterCost and value ofmaterialsTransport costTrailer loadsCost of transportTotal cost andvalueTotal cost

Unit

SupervisorContractor1 mason2 trainees4 labourers

50 kg bags1.2 m. x 2.4 m4mm0.9 m x30mGauge 12.5Two handlesY12mmTonnes2.5 mmTonnes210 litre drums

3 tonnes

with hydro-dynamic well head, KshQuantity

3 days6 days

24 days24 days24 days

26 bags1 sheet20 kg4 meters10 kg1 unit1 length4 tonnes2 tonnes4 tonnes10 drums

4 loads

Unit cost

1,200/day800/day200/day100/day100/day

600400

5010080

3,000500200600200100

900

Totalcost

3,6004,8004,8004,800

28,000

15,600400

1,000400800

3,000500

1,200600

23,500

3,600

45,10069,700

Value ofcommunitycontribution

4,8004,8009,600

19,200

1,200600

1,000

3,600

1,800

24,600

The first part of a well shaft has been builtonto the foundation ring in an excavation dugas deep as safety allows in a riverbed.

Insulated electric wires bent as U s are insertedin the mortar between the blocks in the lowestpart of the well shaft. When the sinking of theshaft has been completed, the wires are pulled out,thereby creating many small holes through whichwater can seep into the well shaft. Also, note thebent iron bars that are mortared into the shaft tofunction as a ladder.

26

4.6 Water lifts and pumps

Water can be drawn by several typesof lifts and pumps. The simplest andcheapest is a bucket tied to a rope.

A windlass made of a few Sisal poles.

A windlass with an iron handle.

A rope and washer pump.

A hand-operated pump with a cylinderhead made of concrete.

life'

A foot-operated money-maker pump.

A broken down India Mark II hand pumpreplaced with a windlass made of Sisal.

27

Shallow water pumps (Reference Technical paper No. 13, TTDG).

New No. 8 1km

Deep water pumps

4.7 Design and cost of large river intakes

According to the evaluation report, each of the intakes at Mwiwe and Nzeuu riverbeds isrequired to supply 217,000 litres (21.7 m3) of water daily, which amounts to 80,000 m3

annually equal to 40,000 m3 during a 6 month period without flooding of the riverbeds.

For each of the two intakes, this yield of water was obtained by sinking a well shaft madeof curved concrete blocks on a foundation ring with an internal diameter of 300 cm. Shortlengths of insulated electric wires were bent into a U/ shape and inserted in each courseof the concrete blocks for every 30 cm with the U facing inwards. The well shaft wassunk as deep as a petrol-powered suction pump could remove the inflowing water.

When the capacity of the pump was reached at about 4 metres depth, the internaldiameter of the shaft was reduced to 110 cm in order to reduce the inflow of water. Thewell-diggers then sunk the well shaft to an additional 3 metres before the inflow of waterbecame too much for the diggers and the pump. When the shaft has reached its finaldepth, the insulated bent wires were pulled out of the shaft, each wire leaving two smallinfiltration holes in the shafts for replenishing of the well.

A small well shaft being sunk inside a widewell shaft of a river intake.

The completed well head with thepipe from the well to the pump.

Plan and section of the intake wells at Mwiwe and Nzeeu (mm).

29

Bill of Quantity and cost of the Nzeeu intake wellDescription8 m deep intake well

Labour cost1 Surveyor1 Supervisor1 Contractor with2 artisans and2 trainees10 labourersCost of labourMaterialsCementRiver sandCrushed stonesCurved well blocksGalvanised wire, 4mmIron bar, Y8Dewatering pumpCost of materialsTransport ofmaterialsTractor trailer loadsCost of transportCost and valueTotal cost

Unit

SurveyorSupervisorContractorArtisanTraineesLabourers

50 kg bagsTonnesTonnesBlocksKg20 m lengthDays

3 tonnes

Quantity

2 days10 days

1 x 20 days2 x 20 days2 x 20 days10 x 20 days

1033

700501010 days

I load

Unit costKsh

1,200/day1,200/day

800/day200/day100/day100/day

600200600

50150500800

900

Total costKsh

2,40012,00016,0008,0004.000

42,400

6,000

1,80035,0007,5005,0008.000

63,300

900900

96,600145,900


8,0004,000

20.00032,000

6001,800

14.000

16,400

900900

49,000

4.8 Design and cost of infiltration pipesIn order to increase the recharge even further, 72 metres of 160mm PVC pipes, intowhich thousands of small holes were burnt with red-hot nails, were connected to theintake. The infiltration pipes were laid in crashed stones as deep as possible and with agradient towards the intake well to enable the water to drain from the sand into the intake.

Infiltration pipe connecting to the intake. A flooded trench for infiltration pipes.

30

L 1t

;;; 4. . • • • • * • > ) . " ' • • • • . ' * • . . ' . • • . ' • • ' ' ' ;

UPVC infiltration pipes

.. Intake well*HJ • • . . -

Plan of the infiltration pipes laid deep in the sand of the riverbed.

Bill of Quantity and cost of 72 metres of infiltration |DescriptionPerforating and laying 72meters of 160 mm PVC pipedeep in a riverbed and slopingtowards a well in the riverbankLabour cost1 Surveyor1 Supervisor1 Contractor with2 artisans and4 trainees10 labourersTotal cost of labourMaterialsDewalcring suction pumpPerforated PVC pipes, 160 mmCost of materialsTransport of materialsTractor trailer loadsCost of transportCost and valueTotal cost and value

Unit

SurveyorSupervisorContractorArtisansTraineesLabourers

8 days6 m length

3 tonnes

Quantity

4 days6 days

1x15 days2x 15 days4x15 days

10 x 15 days

14 pipes

lload

jipe

Unit cost

Ksh

1,200/day1,200/day


8003,048

900

Totalcost

Ksh

4,8007,200

12,0006,0006,000

38,000

6,40042.67249,072

900900

85,972109,422


Ksh

6,0006,000

15,00027,000

450450

27,450

31

4.9 Cost and capacity of pumps, generators and pump houses

The pump in Nzeeu pump house is aGrundfoss CR15-17, 15.0 KW, 3 Phasebooster pump with a capacity of 19nvVhr at154 m head. Together with control panel and60A isolator, float switch, chlorine doserand installation the unit costKsh 560,305 in December 2004.

The pump seen in the photo is the pump inMwiwe and is a Grundfoss CR32-6, 11.0KW, 3 Phases electric booster pump with acapacity of 32m3m/hr at 71 m head.Together with accessories and installationthe cost was Ksh 558,200 in December 2004.

The diesel generator seen in the photo is anAtlas Copco 41, KVA, QUB41 that powersthe Mwiwe pump. The cost was Ksh876,550 in December 2004.

The Nzeeu pump is also powered by anAtlas Copco diesel generator with a biggercapacity of 41 KVA, QUB41. The cost wasKsh 876,550 in December. 2004.

The Nzeeu pump house seen in the photo issimilar to the Mwiwe pump house. Bothhouses have a private room for the PumpOperator.

The cost of the pump house with fence wasKsh 793,866 including value of local labourand materials.

32

4.10 Calculations and cost of rising mains and head tanks

The head tank for the Mbitini Water Projectis a 100 nr1 tank built of concrete blocks fora cost of Ksh 797,590. The tank is situatedat a distance of 0.916 km from the Mwiweintake and at an altitude of 35.1 m above theintake. The pumping head with 100 mm G.Ipipe delivering 30 mVhr is:

Delivery headFrictional lossesTank height10% residual (extra) headTotal pumping headPumping head of the pump

= 35.1m= 25.7 m= 3.0 m= 6.4 m

70.2 m71.0 m

The head tank for the Kisasi Water Project isa 50 m3 elevated tank made of steel platescosting Ksh 1,277,621. The tank is situatedat a distance of 5.3 km away form the Nzeeuintake and at an altitude of 74.2 m abovethe intake. The pumping head with a 100mm and 80 mm G.I. pipe delivering 19m'/hr is the following:

Delivery headFrictional lossesTank height10% residual (extra) head

= 74.200 m= 57.023 m= 8.000 m= 13.922 m

Total pumping head 153.922 mPumping head of the pump 154.000 m

The cost of the rising main pipes were:

Intake to head tankN/euu to steel lankMwiwe to block tank

Km5.3100.916

Ksh7,461,4351,878,867

Water is delivered from the head tanksthrough Distribution pipelines to the waterkiosk by means of gravity as shown on amap of the Mbitini project area on the nextpage.

33

4.11 Map showing pipes, tanks, kiosks and valve chambers of MbitiniWater Project.

KATWALA -MBITINI WATER PROJECTIN

A

0 STORAGE TANK+ AIR VALVE• WASHOUT CHAMBER| M PUMP HOUSE

0 1

ROADS

Loase Surfara RoadsMain tracks (MrtorabW

RIVERBEDS

A/1* SLUICE VALVE * SETTLEMENTS

^ SAND DAM • WATER KIOSK

• GATE VALVES

8 9 10 km

More technical details and costs on the Mbitini Water Project are given in anotherhandbook o f this series: Water Surveys & Designs.

34

Chapters. Subsurface dams built of soil

5.1 History of subsurface dams

The oldest subsurface dams known were constructed in the almost waterless area aroundDodoma, in then Tanganyika and now Tanzania, when the railway was built around1905. The purpose of the subsurface dams was to provide water for the steamlocomotives which they successfully did so for many decades.

A subsurface dam was built of soil and documented by Bihawana Mission near Dodomain the early 1950s and is still functioning well. A similarly successful subsurface damwas built in the same area in the 1920s. Several subsurface dams were built of variousmaterials, such as soil, burnt bricks, concrete blocks and reinforced concrete during atraining course for TLO in the Dodoma region in 1991. Many more subsurface damscould probably be located in many other places, but if so, they would be invisible belowthe sand surface of riverbeds.

5.2 Function of subsurface damsShould the water yield from a river intake be insufficient for the demand, then asubsurface dam can be constructed cheaply of soil. The function of subsurface dams is to:

a) Block the underground flow of water between the voids in the sand, andb) to raise the water level in the sand to about 30 cm below the surface of riverbeds.

With regard to Nzeeu riverbed, it was decided to construct a subsurface dam of soil thatwould increase the yield from the existing 6,270 m3 of water to 8,282 m3 which, togetherwith the substantial underground replenishment would be sufficient to meet 6 monthsdemand of 40,000 m3.

Longitudinal profile combined with a three dimensional view of a subsurface dambuilt of soil.

35

5.3 Design and cost of Nzeeu Subsurface Dam

Depth of 600

I50JL- ̂ j._i5M,.u '5i<L

Plan and section of Nzeeu subsurface dam

Bill of Quantity and cost of Nzeuu Subsurface DamDescription

Labour costSurveyor/designerSupervisorContractorArtisansTraineesLabourersCost of labourMaterialsClayey soilCost of materialsTransport of materialsHiring suction pumpTractor trailer loadsHiring de watering pumpCost of trans, and pumpCost and valueCost of subsurface dam

Unit

SurveyorSupervisorContractorArtisansTraineesLabourers

Tonnes

3 tonnes Days

Quantity

1 x 2 days1 x 6 days1 x 22 days2x 20 days4 x 20 days

10x20 days

69

23 loads4 days

Unit costKsh

1,200/day1,200/day


100

900800

Total costKsh

2,4007,200

17,6008,0008.000

43,200

6,9006,900

20,7003,200

23,900

74,000115,250


8,0008,000K.000

24,000

6,9006,900

10.350

10,35041,250

36

The sand in a riverbed has beenremoved and the 60 cm deep keyhas been excavated in the floor.

The excavation is being filled withcompacted moist clayey soil bya community in Myanmar (Burma)

Subsurface dams being constructed by a self-help community in Makueni.

5,3 Guidelines on construction ,

In order to get a maximum volume of water for a minimum of work and investment,subsurface dams, weirs and sand dams should, whenever possible, always be constructedon underground dykes that are situated downstream of underground water reservoirs,such as waterholes and water-indicating vegetation.

When a suitable site has been identified in a riverbed, the most clayey soil forconstruction of the dam wall has to be found. This is done by collecting some soilsamples from nearby riverbanks and fields. The equipment for analyzing the soil samplesconsistsof some plastic bottles of equal size of which the caps have been removed and thebottoms cut off.

The bottles are placed upside down in the sand, or sloping against a wall, and filledhalfway with soil samples. Water is poured on top of the soil samples several times. Aftersome minutes it can be observed which soil sample has the slowest infiltration rate due tohaving the highest clay content. This soil is the most suitable for building the dam wall.

37

Soil samples being tested for their clay content by pouring water onto the soil samples.The sample having the slowest infiltration rate is the most suitable for dam walls.When a suitable site has been identified, all the sand in the riverbed is removed in a 3metre wide stretch between the two riverbanks so that the floor under the sand is fullyexposed.

Thereafter a 100 cm wide trench, called a key, is excavated into the floor right across theriverbed and into the two riverbanks. The depth of the key must be at least 60 cm intosolid soil to prevent seepage under the dam wall.

The clayey soil identified by the testing mentioned above, can now be transported to thedam site by means of sacks, donkeys, ox-carts, wheelbarrows or a tractor with a trailer.

First the whole length of the key is filled with a20 cm thick layer of soil that is moisten withwater and compacted using either short lengthsof tree trunks or cows or a tractor driven backand forth until all air is forced out of the soil.Thereafter other 20 cm thick layers of soi I arelaid out along the whole length of the key anddam wall, moistened and compacted until theheight of the dam wall has reached to 30 cmbelow the surface of the sand in the riverbed.

The upstream and downstream sides of the dam wall, having a slope of about 45 degrees,are then smoothened using shovels and wooden floats. Especially, the upstream side ofthe dam wall should be plastered with clay or cow dung to prevent water from seepingthrough the dam wall.

Finally, the excavated sand is back-filled against both sides and the top of the dam wall.

It is wise to hammer two short iron bars into the riverbanks at each end of the dam wallbecause the dam will be invisible after the first flooding.

38

Chapter 6. Weirs

6.1 Various types of weirsWeirs are water-tight structures built across flowing water streams, dry sandy riverbedsand gullies. Only weirs built across dry sandy riverbeds and gullies are described here.

Although such underground water reservoirs might only be recharged with floodwater afew times annually, they may be perennial water sources provided the storage capacity issufficient large and without underground leakages. Weirs can be constructed of variousdesigns and materials, such as;

A wall of soil supported on both sideswith plastic sacks filled with soil and aspillway at one end of the dam wall.

Rubble stones mortared together with alillway at the middle of the dam wall.

A similar but more sophisticated weir.

The above weirs were built acrossgullies to provide for sand harvesting.

This long weir was built of rubble stonemasonry across a sandy riverbed. Itsupplies water throughout the year.

Once or twice a year, the undergroundwater reservoir is replenished by floods.

A hand-dug well is sunk upstream of thisweir for extraction of water

The above weirs were built acrossriverbeds to supply domestic water.

39

This simple and low-cost weirwas constructed across a smallstream by a farmer himself inCentral Kitui. He uses thewater for bucket irrigation ofvegetables for sale.

Many similar weirs have beenbuilt by fanners themselves.Some of the farmers use smallpetrol-powered pumps forirrigation of vegetables andfruit trees.

This is a weir constructedacross a small stream on theSagalla Hill at Voi long timeago.

The water is gravitated down toa piped water scheme in thelowland near the foot ofSagalla Hill.

This weir was built across Voiriverbed by Mau Mau prisonersin the 1950s.

The purpose was to divert floodwater for large scale seasonalirrigation of farmland situatedat a lower elevation.

The required volumes of waterwere regulated by opening thegate covering the hole seen inthe wing wall. Released waterwould pass from the holethrough a canal to the field.

40

6.2 Design and cost of weir

CROSS SECTION B- B

Plan and profiles of Talek weir.

r / / N>w*Bnd& water livtl: •w

longitudinal profile combined with a three-dimensional view of a weir built on anunderground dyke and stretching across a riverbed.

A hand-dug well equipped with a hydro-dynamic well head is sunk at the deepestpoint of the underground water reservoir.

41

The Talek weir under construction. Some PVC pipes were inserted in thedam wall to discharge the water accumulating upstream. After completion of thewall, the pipes were removed and the holes in the dam wall were sealed.

Bill of Quantity and cost of Talek weir built of rubble stone masonry. Ksh.

DescriptionA weir with wingwalls having a totallength of 18 metresLabour costA surveyor/designerA supervisorA contractor withhis/her team ofbuilders andcommunity workersCost of labourMaterialsCementBarbed wire, g 12.5River sandCrushed stonesHardcoreWaterCost of materialsMaterial transportTractor trailer loadsCost of transportCost and valueTotal cost of weir

Unit

1 Surveyor1 Supervisor1 Contractor3 Artisans4 Trainees10 Labourers

50 kg bags25 kg rollsTonnesTonnesTonnesOil-drums

3 tonne loads

Quantity

6 days12 days26 days24 days24 days24 days

105 bags2 rolls

15 tonnes15 tonnes40 tonnes50 drums

17 loads

Unit cost

1,200/day1,200/day


6002,500

200600200100

900

Total cost

7,20014,40020,80014,4009,600

80,800

63,0005.000

68,000

15.30015,300

134,800185,500


14,4009,600

24.00048,000

3,0009,0008,0005,000

25,000

7,65050,700

42

6.3 Construction procedures

A 60 cm wide trench was dug across theriverbed and into the banks of Talekriverbed bordering the Maasai Mara.

A small stream of water in the riverbedwas diverted over the trench in somePVC pipes. A petrol pump sucked waterout of trench while it was beingexcavated to its final depth of 60 cm intosolid and impermeable soil.

Thereafter the trench was filled withconcrete packed with rubble stones andreinforced with double lines of barbedwire, g.12.5.

f

When the trench had been filled withconcrete, rubble stones and barbed wirefor every 30 cm height, large flat stoneswere mortared onto the concrete to formthe outside shuttering of the dam wall.

Concrete, rubble stones and lines ofbarbed wire were filled in between theflat stones. The PVC pipes wereremoved and the dam wall plastered.

43

6.4 Maintenance

Flood water passing over spillways always try to remove the sand from the downstreamside of weirs. It is therefore always important to place as large boulders as possibleagainst the downstream side of weirs. It might be necessary to use reinforced concrete tobond the boulders together and place them properly in the event of violent flooding.

The photo below shows a weir that was built 20 years ago in the dry area of Mutha. Theweir supplies water all year round for the hand-dug well into the riverbank as seen on thefront cover of this handbook.

The boulders that were concreted to the downside of the weir were washed away a longtime ago. The weir and riverbed are now exposed to erosion that may cause the weir to beundermined by flood water. That in turn will result in the weir being washed away,thereby leaving the hand-dug well dry until a new weir can be built.

44

Chapter 7. Sand dams

7.1 History of sand dams

Sand dams have been constructed and used in India for centuries. The first sand damsknown in Kenya were built by the District Agricultural Officer, Eng. Classen, as partof a development project named African Land Development Board (ALDEV).That project constructed many earth dams, rock catchment dams, sand dams,boreholes and rangcland schemes in Ukambani from 1954 to independence in 1964.

Some of the sand dams built 50 years ago arestill functioning despite lack of maintenance.An example of the successful sand dams builtin the 1950 s is Manzui at Kyuso.

Some 200 sand dams have been constructed in Machakos, Makueni, Kitui, Mwingi,Embu and Meru by various agencies and ministries since the early 1970s. Anevaluation showed that only about 5% of the dams built during the last 40 years werefunctioning. This high failure rate is not surprising upon learning that the engineersof today had not learnt about subsurface dams, weirs and sand dams during theirstudies.

This handbook is based on the techniques used for survey, design and construction ofsand dams built according to the ALDEV design in Tanzania, Burma, Somalia,Eritrea and Kenya since 1978.

Should a reader find that he or she still needs more information on the survey, design,construction, etc. before starting on construction of sand dams, please turn to theReferences on the last page.

Hopefully, this will assist relevant agencies and ministries to realise the big potentialfor turning some of thousands of dry riverbeds in the arid and semi-desert regions ofKenya into perennial water sources for people, livestock and irrigation.

45

7.2 Functions of sand dams

Water supplyThe main function of sand damsis to increase the volume of sandand water in riverbeds.

The Mwiwe sand dam increasedthe volume of extractable waterfrom 139 m' to 2,997 m3. Despitethe fact that the reservoir has onlybeen recharged once by a smallrain shower during the last 18months, 75 m3 of fresh water isextracted daily from the riverbed.

Sand harvesting and rehabilitation of gulliesSand dams can also be built ingullies to supply both sand andwater, while healing the gullies.10 sand dams built in gulliesharvest sand instead of the sandsilting up Lake Victoria. Theconstruction cost of the sanddams was recovered in less than18 months through the sale ofsand.

If the purpose of building sanddams is to harvest sand andrehabilitate gullies, then weirs builtof plastic bags filled with soilare more profitable.

Riverbed crossingsRural roads often cross riverbeds.In remote areas, people simplydrive across such riverbeds.Passable crossings can be madeby concreting a curved andreinforced concrete slab over theriverbed called ? Trish Bridges? byhumorous Englishmen.

Riverbed crossings can be madeinto sand dams capable of holdingwater upstream of the crossing asseen in this photo from Mulangoat Kitui where SASOL hasconstructed about 480 weirs.

46

...

7.3 Five types of sand dams

This photo is of a sand dam builtaccording to the ALDEV designwhich has functioned for 50 yearswithout failure. 35% of water can beextracted from its sand reservoir ifthe spillway is raised in stages of 30cm height above the level of sanddeposited by floods. The designcriteria and construction procedureare explained on the following pages.

The spillway of this sand dam inTharaka, Meru, was not built instages. Therefore its reservoircontains fine-textured sand and siltfrom which no water can beextracted.

>^mtm

The sand reservoir of this sand damin Taveta was also silted up with siltand fine-textured sand due to notbuilding the spillway in stages. Nowater can be extracted from this dam.

USAID built this interesting sanddam in Kitui in the 1980s. The damreservoir contains coarse sand fromwhich people draw water.

Although the dam has two spillways;one made of steps and being higherthan the other without steps, the leftside of the spillway is heavily erodedby floods.

47

7.4 Criteria for construction of successful sand dams

The major reasons for many failed sand dams are due to insufficient knowledge of:a) Identifying suitable sites for dam walls and dam reservoirs.b) Design criteria and flood dynamics.c) Construction procedures.d) Maintenance requirements.

Should a specific site not meet all the criteria shown on the following pages it is notadvisable to construct a sand dam.

7.4.1 Site criteria

1) Suitable riverbeds must have twohigh riverbanks to enable the wing wallsto keep over-flowing flood water withinthe spillway and not flowing over theriverbanks.

If flood water is allowed to flow over thewing walls and riverbanks, it will erodethe riverbanks and cause the river tochange its course, thereby leaving the sanddam as a ruin, as seen on one of theriverbanks here.

2) Dam walls should never be built onfractured rocks or large boulders becausesuch walls cannot be made water-tight.Water will always seep out between theboulders as seen in this photo.

3) Dam walls should therefore always bebuilt either on a solid bedrock base orkeyed 1 metre into solid and impermeablesoil.

If dam walls are keyed less than1 metre into solid and impermeable soil,water will find its way out under the damwall and cause the wall to hang in its wingwalls over the riverbed.

48

4) Suitable sites for dam reservoirsshould be without boulders andfractured rocks because they will drainwater from the reservoir into theground below.

While such seepage is unwanted forsand dams it can be beneficial forrecharge of ground water and nearbyboreholes.

5) Riverbeds with fine-textured sandoriginating from flat land are alsounsuitable for sand dams, because lessthan 5% of the water stored in thevoids between the sand particles canbe extracted.

6) Wide riverbeds, say of more than25 metres width, are also unsuitablefor sand dams because they becometoo expensive due to the reinforcementrequired for the long dam walls.

Riverbeds exceeding 25 metres inwidth are suitable for subsurface damsbuilt of soil because that material isplastic and do not crack like concrete.

7) The whitish stones seen on theright side of the photo are calledcalcrete. Calcrete is licked bylivestock and wild animals because itcontains salt and other minerals neededby them.

If calcrete is situated upstream on theriverbanks where the dam will belocated then the water will be salineand therefore only useful for livestock.

49

8) Water-indicating vegetation, thatrequires water all year round, should begrowing on the banks where thereservoir will be located as proof of theriverbed capacity to store water.

A list of water-indicating vegetation,such as the Mukuyu (wild fig) seen nextto a hand-dug well in a dry riverbed inthe photo.

9) Waterholes, even temporary ones,should preferably also be located wheredam reservoirs will be located to provethat the riverbed has no leakagesdraining water into the ground below.

10) Catchment areas should containstones or stony hills, from whererainwater and floods can transporteroded stone particles, which coarsesand is made of, to the riverbedsand deposit them into the dam reservoirs.

11) As mentioned earlier, sand dams,weirs and subsurface dams should alwaysbe constructed on natural undergrounddykes to gain free storage of water whilealso reducing construction costs.

50

7.4.2 Design criteria

1) The most important design criteria isto ensure that flood water will depositcoarse sand into the dam reservoir fromwhere up to 35% of water can be extracted.Most reservoirs contain silt and finetextured sand from where little or no watercan be extracted as seen in the photo.

Flood water contains all sizes of silt andsand particles. Since sand is heavier thansilt, sand will always be transported in thelowest part of the current of the flood water,while silt being lighter will float in theupper part of the flood water.

Therefore, if flood water is made to passover a barrier being about 30 cm high, thebarrier will trap the heavy coarse sand whilethe lighter silt will pass over the barrier onits way downstream to the riverbed.

Practically, this method of harvestingcoarse sand is obtained by raising thespillway of sand dams in stages of 30 cmheight over the sand level in riverbeds.

When flood water has deposited coarsesand to the first stage of 30 cm height, thespillway is again raised by another phase of30 cm. When the sand from next flood hasreached the height of the second stage, thespillway is again raised by yet another stageof 30 cm and so on until the final height ofthe spillway is reached as seen in thesketch.

The photo shows the spillway of a sanddam at Makueni that has reached the secondstage of raising the spillway in stages of 30cm above the sand level.

Judging from the height of the personstanding on the spillway, another 6 stages,each oi" 30 cm, are to be raised when thenext 6 floodings have deposited their coarsesand.

51

2) The force of water and sand in damreservoirs against the wall of sand damsis counter balanced by making the widthof the base for dam walls 0.75 (3/4) ofthe height of its dam wall.

3) As mentioned earlier, dam walls mustbe keyed at least 1 meter into solid andimpermeable soil. The thickness of the keyshould be 0.55 (1/22) of the height of its damwall.

4) The width of the crest and its height onthe downstream side should be 0.2 (1/5) ofthe height of its dam wall.

5) In the ALDEV design shown here thefront of the dam wall is leaning downstreamwith a gradient of 0.125 (1/8) of the height ofits dam wall.

6) The spill-oVer apron ontowhich the over-flowing waterwill fall with force, must bereinforced onto the dam wall.

The apron should be of thesame width as the dam walland extend up along thewing walls.

Large stones should be setinto the concrete to break theforce of water.

hxO.75

7) The key under the dam wall mustextend all the way up to the end of thewing walls, otherwise water might seepunderneath them, thereby eroding theriverbanks,

8) Water can be extracted from sanddams by one of the types of river intakesmentioned earlier or gravitated througha galvanised pipe inserted into the damwall to a tapping point somewheredownstream. Note the stage of closing thespillway.

52

7.4.3 Maintenance criteria

Sand dams require careful maintenance and immediate repair during flooding wherehundreds of tonnes of water tall over the spillway and onto the spill-over apron. Floodwater may also spill over and erode the wing walls and, perhaps, even over theriverbanks during heavy rains.

Owners of private sand dams can usually mobilize repair works quickly and keeptheir sand dams functioning well for many years. Maintenance and repair ofcommunity-owned sand dams take much longer time because the committees have tomeet and decide what to do and how to pay for it.

This sand dam was surveyed, designedand constructed successfully accordingto all criteria during a training course atWote in Makucni in 1996.

During a study tour to the sand dam acouple of years later, the community wasadvised to repair the eroded apron at thewing wall. It didn t. The dam wal I waswashed away during the next flood.

The lower part of the spill-over apronwas washed away by heavy floods due toinsufficient reinforcement.

If the repair had not been carried outpromptly, water would have seepedunder the dam wall and destroyed it.

Note the sand harvesting activity and thephases of the closed spillway.

This sand dam at Kibwezi has a numberof problems due to having been built onlarge boulders and complete lack ofmaintenance since it was built by MoA in1978.

Despite all the criteria listed above, sanddams have a great potential for supplyingwater from small riverbeds in the semi-arid, arid and semi-desert zones ofAfrica, provided the criteria are adhered to.

53

Chapter 8. Mwiwe Sand Dam

8.1 Yield of water from Mwiwe Sand DamAs mentioned earlier, Mwiwe sand dam was constructed to increase the volume of thewater reservoir from 139 m3 to 2,997 m3 from a small riverbed, Mwiwe, in Kitui.

Althougn the spillway was closed only to the second stage, which is 60 cm above Lheoriginal sand level, 75 m3 of fresh water was pumped from the reservoir every day for 18months amounting to a total 40,500 m3. Part of the recharge came from a small floodingand a small continuous underground flow of seepage from some weirs situated a fewkilometres upstream in the riverbed.

When more floods have deposited sand in the reservoir the spillway will be raised instages, each of 30 cm height.

8.2 Design calculations for Mwiwe Sand Dam

The Mwiwe Sand Dam was designed according to the specifications of ALDEV (AfricanLand Development Board) design that has proven to be successful since the mid 1950s.The main specifications of this particular design are:

The height of the spillway is determined by the maximum height of flood water that cansafely spill over the spillway without eroding the upper ends of the wing walls and theriverbanks. The maximum height of floodwater can usually be known by interviewingelderly residents who can point out how high the water level had reached during thehighest flooding in their time.

In Mwiwe riverbed the highest flood level was found to be 100 cm above the sand level.Since the two riverbanks were 250 cm above the same sand level the maximum height ofthe spillway was found be:

Height of riverbanks: 250 cmHighest flood level: 100 cmMaximum height of spillway: 150 cm

54

The Mwiwe dam wall has the required criteria of:

a) A base that is 0.75 (3/4) of the height of the spillway equal to:height 150 cm x 0.75 = 112.5 = 113 cm.The key under the dam wall has a depth and width of:100 cm into solid and impermeable soil and have a width of:height of dam wall x 0.55 equal to 150 cm x 0.55 = 82.5 cm - 83 cm

b) The upstream side of the dam wall has a gradient of:height x 0.125 equal to: 150 cm x 0.125 = 18.75 cm = 19 cm.

c) The crest has a width of: height x 0.20 equal to 150 cm x 0.20 = 30 cmand is vertical on the downstream side for a similar distance of 30 cm.

d) The spillway was only raised in stages 30 cm above the level of sand in thereservoir.

The keys under the two wing walls have:A width of 45 cm and a depth of 60 cm into solid and impermeable soil

The spill-over apron is:Reinforced together with the base of the dam wall.Has a width being equal to the base = 113 cm.Stretch all along the downstream side of the dam wall and wing walls withlarge rubble stones set in the concrete.

8.3 Design of Mwiwe Sand Dam

55

8.4 Construction procedures for Mwiwe Sand Dam

1) The outline of the wing walls, dam wall andspill-over apron was marked with wooden pegsand nylon strings according to the measurementsgiven on the design criteria.

2) All top soil, sandy soil and roots within themarked areas were removed until solid andimpermeable soil was reached.

3) The key and base were excavated into solidand impermeable soil below any layer of sand andsandy soil encountered as follows:

Structure of sand damBase of dam wallBase of spill-over apronKey under dam wallKey under wing walls

Depth30 cm30 cm

100 cm60 cm

Width113 cm113 cm83 cm45 cm

4) Two templates were made of timber. Theinner sides of the templates gave the outline of thespillway and dam wall given under the designcriteria. The height of the wall was 150 cm. Theupstream wall had a gradient of 19 cm to thevertical. The crest was 30 cm wide. Thedownstream wall had a vertical drop of 30 cmfrom where it sloped towards the base that was113 cm.

5) The completed excavation was then filledwith 30 cm high layers of cement mortar ofmixture 1:4 into which many boulders werecompacted without touching each other.

6) The nibble stone masonry was reinforcedwith Y8 twisted iron bars laid in the concrete forevery 30 cm height.

7) After the excavation was filled with rubblestone masonry, the two templates were erected atthe ends of the spillway for the purpose of givingthe outline of the dam wall, spillway and wingwall.

56

8) Nylon strings were drawn tightly from theinner corners of the templates to pegshammered into the soil next to the upper end ofthe wing walls.

9) Flat stones were then set in cement mortar1:4 along the inner lines of the strings to makethe outer sides of the wing walls.

10) Next day the space between the flat stoneswas filled with mortar 1:4 into which roundrubble stones were compacted. Thereafter flatstones were mortared onto the wing walls sothat they could be filled with mortar andstones next day.

11) The base of the dam wall, the spill-overapron and the spillway, the latter being situatedbetween the two templates, were only raised to30 cm above the original sand level in theriverbed.

12) Luckily, a small flooding deposited a 30cm thick layer of coarse sand that reached thefirst stage of the spillway. The spillway wastherefore raised another 30 cm above the sandlevel for the next stage of the spillway.

13) Large boulders were concreted into thespill-over apron to reduce the velocity ofsurplus water falling over the spillway andwing walls.

14) The wing walls were thereafter plasteredwhile the spillway was left open until the nextflooding

15) The next flooding deposited coarse sandto the level of the spillway. The spillway wasraised another 30 cm above, the new sandlevel.

16) The process of rising a spillway in stagesof 30 cm height may be completed in onerainy season provided the required numberof flooding occur and builders are ready fortheir work without delay.

57

8.5 Bill of quantity and cost of Mwiwe SandDescriptionA sand dam being 2 m high and 12m longLabour costA surveyor/designerA supervisorA contractor, inclusive his/herteam of builders and communityworkers

Cost of labourMaterialsBags of cementRiver sandHardcore 2" to 6"WaterBarbed wire, g 12.5y 8 twisted iron barsCost of materials

Transport of materialsHardware lorriesTractor trailer loadsCost ol transport

Cost and valueTotal cost of Sand Dam

Unit

1 Surveyor1 Supervisor1 Contractor3 Artisans4 Trainees10 Labourers

50 kg bagsTonnesTonnesOil-drum25 kg rollsLength

7 tonne3 tonnes

Quantity forMwiwe sanddam

10 days18 days

1 x 42 days3 x 40 days4 x 40 days10 x 40 days

2451.50

120304

30

2 loads44 loads

DamUnit cost

Ksh

1,200/day1,200/day800/day200/day100/day100/day

600200200100

3,000350

5,000900

Total cost

Ksh

12,00021,60033,60024,00012.000

103,200

147,000

12,00010,500

_169,500

10,00039.60049,600

322,300476,600


24,00012,00040,00076,000

30,00024,0003,0001.500

58,500

19.80019,800

154,300

Bill of Quantity and cost of closing spillwayDescriptionClosing last 3 stages of 30 cmheight of spillway _ ^ _

Unit Quantity forclosingspillway

Unit cost

Ksh

Total cost

Ksh


Labour costSupervisorContractor, inclusive his/herteam of builders and communityworkers

Cost of labour

1 Supervisor1 Contractor1 Artisan2 Trainees10 Labourers

2 days1 x 3 days1 x 6 days2 x 6 days

10 x 6 days

1,200/day800/day200/day100/day100/day

2,4002,4001,2001.200

7,200

1,2001,2006,0008,400

MaterialsBags of cementRiver sandHardcore 2" to 6"WaterCost of materials

50 kg bagsTonnesTonnesOil-drum

204

1015

600200200100

12,000

8002,0001.5004,300

Transport of materialsHardware lorriesTractor trailer loadsCost of transport

3 tonne3 tonnes

lload3 loads

3,000900 1,350

1.350

Cost and valueTotal cost of spillway

24,90038,950

14,050

Total cost of Mwiwe Sand DamSand dam with open spillwayCompleting spillway in 3 stagesCost and value of sand dam with spillwayGrand total of sand dam with completed spillway

322,30024,900

347,200515,550

154,30014,050

168,350

58

References

Agarwal, A. and Narain, S. 1991.Dying Wisdom.Centre for Science and Environment, India.Ahnfors, O. 1980. Groundwater arresting sub-surface structures. Sida/ Govt. of India.Backman, A. and Isakson. 1994. Storm water management in Kanye, Botswana. Swedish University of Agricultural ScienceBurger, S.W. 1970. Sand storage dams for water conservation. Water Year 1070. S. Africa.Faillacc, C. and K.R. 1987. Water Quality Data Hook. Water Development Agency, Somalia.Fewster, E. 1999. The feasibility of Sand Dams in Turkana. Loughborough University, UK.Finkel & Finkel Ltd. 1978. Underground dams in arid zone riverbeds. Haifa, Israel.Finkel & Finkel Ltd. 1978. Underground water storage in Iran. Haifa. Israel.Gould, J. and Nissen-Petersen, E. 1999. Rainwater Catchment Systems, IT Publications, UK.Hudson, N.W. 1975. Field Engineering for agricultural development. Oxford, UK.Longland, F. 1938. Field Engineering. Tanganyika.Newcomb, R.C. 1961. Storage of groundwater behind sub-surface dam. US Gfcoi.Survey, USNilsson, A. 1988. Groundwater Dams for small-scale water supply. IT Publications. UK.Nissen-Petersen, E. 1982. Rain Catchment and Water Supply in Rural Africa. Hodder/Stoughton.Nissen-Petersen, E. and Lee, M. 1990. Subsurface dams and sand dams. No. 5. Danida Kenya.Nissen-Petersen, E. 1995. Subsurface and Sand-storage Dams. UNDP/Africare, Tanzania.Nissen-Petersen, E. 1990. Harvesting rainwater in semi-arid Africa. Danida, Kenya.Nissen-Petersen, E. 1996. Ground Water Dams in Sand-rivers. UNCHS, Myanmar.Nissen-Petersen, E. 2000. Water from Sand Rivers. RELMA/Sida, Kenya.Muliso, G. and Thomas, D. 2000. Where there is no water. SASOL, Kenya.Raju, K.C.B. 1983. Subsurface dams and its advantages.Ground Water Board, India.Sandstrom, K. 1997. Ephemeral rivers in the tropics. Linkoping University, Sweden.Slichter. 1902. Sub-surface dams.USGS Water Supply and Irrigation. USA.Wipplinger, O. 1958. The storage of water in sand. Water Affairs, South West Africa.Wipplinger,O. 1974. Sand storage dams in South-West Africa. South Africa.

59

Training, implementation and documentation ofrainwater harvesting in ASAL regions

IntroductionASAL Consultants Ltd. promotesaffordable water in dryland usinglocally available skills andmaterials to construct affordableand sustainable water projectstogether with communities inASAL (Arid and Semi-aridLand).

Rock catchment tank and dams.

An earth dam built manually.

We are specialised inimplementing a combination ofpractical and theoretical trainingcourses for engineers, techniciansand artisans.

The participants learn hands-onhow to survey, design andconstruct water projects in co-operation with self-help groups.

Services offeredASAL Consultants Ltd. offertechnical services on:* Ponds and earth dams.* Subsurface and sand dams.* Hand-dug wells, river intakes.* Roof, rock, road and ground

catchment systems.

A hydro-dynamic well-head on adeep hand-dug well in a lugga.

A subsurface dam being constructedof rubble stone masonry.

ASAL Consultants Ltd. wasregistered in Kenya in April 1990as a consulting firm.

Erik Nissen-PetersenASAL Consultants Ltd.P.O.Box 739, Sarit 00606,Nairobi,Kenyaasalconsultants @ yahoo.comasal @ wananclii.comTel. 0733 619066 & 02 2710296

60

Water from Dry Riverbeds - IRC...ASAL Consultants Ltd. P.O. Box 739, Sarit 00606, Nairobi, Kenya [email protected] [email protected] Fax/Tel: 254 (0) 202710296 Mobiles: 0733

Documents