GOVERNMENT OF ETHIOPIA WATER RESOURCES DEVELOPMENT AUTHORITY MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS I irj VOLUME 6 ANNEX B : GROUNDWATER AND SALINITY ANNEX C :HYDROLOGY Wrer Resources Development Authority P O Box 5673 Addis Abeba Sir Willia’n Halcrow & Partners Consulting Engineer* Burderop Park Ethiopia S^-i.don Wilshire SN4 OQD ■ United Kingdom■ ■ ■ ■ < GOVERNMENT OF ETHIOPIA WATER RESOURCES DEVELOPMENT AUTHORITY MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS FINAL REPORT ■ 0 July 1985 ■ VOLUME 6 ANNEX B : GROUNDWATER AND SALINITY ■ ANNEX C : HYDROLOGY 0 ■ H 0 Water Resources Development Authority P O Box 5673 Sir William HsJcrow & Portner* Consulting Engineers Addis Ababa Ethiopia Burderop Park Swindon ■ Wiltshire SN4 OQD United KingdomI MASTER DRAINAGE PLAN FOR MELKA SADI 4 AMI3ARA AREAS ?■I N A L R £ ? 0 R l VOL 1 VOL 2 VOL 3 VOL 4 VOL 5 VOL 5 - SUMMARY - MAIN REBORT - ANNEX Al : SOILS AND DRAINABILITY CLASSES ANNEX A2 : SOILS AND DRAINABILITY CLASS MAPS ANNEX A3 ANNEX 3 DESCRIPTION OF AUGER BORES GHOUNDWATEH AND SALINITY VOL 7 VOL 8 VOL 9 VOL 10 ANNEX f* HYDROLOGY ANNEX D ENGINEERING ANNEX E AGRICULTURE ANNEX ■MARKETING AND PRICES ANNEX G PROJECT COST ESTIMATES ANNEXH FINANCIAL AJID ECONOMIC EVALUATION ANNEX I EVALUATION METHODOLOGY ANNEX J ENVIRONMENT AND HEALTHANNEX B GROUNDWATER AND SALLNITY CONTENTS INTRODUCTION 1.1 Location 1.2 Climate 1.3 Sella and Geomorphology 1.4 Geology and Hydrogeology 1.5 Irrigation Practices GROUNDWATER Page No 1 2.1 Background 2.2 Review of Previous Studies 2.3 Review of Available Data 2.4 Analysis of Data 2.4.1 Depth to Groundwater Contour Haps 10 2.4.2 Groundwater Salinity Contour Maps 11 2.4.3 Groundwater Contour Maps 11 2.4.4 Observation Well Hydrographs 12 2.4.5 Piezometric Sections 13 2.5 Perched Water Tables 14 2.6 Identification of Priority Areas for Subsurface Drainage 14 2*7 Predicted changes in Groundwater Depth with Time 15 2.8 Computerised Groundwater Mapping 16 2.9 Recommendations for Additional Observation Wells ig 2.10 Conclusions 17 19 19 19 3.3 Additional Field Investigations 20 3.4 Characterisation of Saline Soils SALINITY 3*1 Background 3.2 Review of Previous and Concurrent Studies 3.4.1 Inherently Saline Soils 3*4.2 Soils Affected by Secondary Salinity 3*4.3 Initially Alkali Conditions 3.4.4 Soils at Risk of Alkalisation Following Leaching 23 21 21 22 22 3.5 Distribution of Saline Soils 24 3.6 Irrigation Water Quality 24 3.7 The Development of Salinity in Irrigated but Undrained Land 75 3.7.1 Cotton 3.7.2 Bananas 3,7*3 Other Crops3.8 Control of Secondary Salinization by Subsurface Field Drainage - Principles 3.9 Reclamation Leaching After Drainage 3-10 Routine Leaching After Drainage 3 * 10 . 1 Cotton 3.10.2 Bananas 3*10.3 Other Crops 3.11 Requirements for Soil Amelioration REFERENCES APPENDICES A Observation Well Hydrographs B Rates of Rise of Groundwater and Predicted Groundwater Depth vs Time Relationships C Theoretical Model of Secondary Sallniaatlon in Cotton Fields D Outline Reclamation Procedure TABLES Page No 29 29 31 31 35 35 39 41 Table No. Title 1 Number of Observation Wells/Piezometers Installed by Individual Organisations and Identifying Codings 2 Status of Observation Wells and Piezometers, October 1904 3 Available Data Coverage 4 Priority Areas Delineated by Depth to Groundwater Contours 5 6 7 Reduction in Soil Salinity With Successive Irrigations in a Soil on the Soil Moisture Water Balance Threshold of Abandonment for Cotton Characteristics of Type Profiles (in mm) for two Soil Groups with Maximum Leaching 8 Effect of Leaching Regime on Different Soils with Different Salinity of Groundwater, and Corresponding Expected Yield Loss 9 Approximate Net Cotton Land Areas Within Drainability, Textural and Groundwater Salinity Classes Appendix B Rates of Rise of Groundwater and Observation Well Depth vs Time Relationships - Melka Sadi Area Rates of Rise of Groundwater and Observation Well Depth vs Time Relationships - Amibara Area 11Cl Di Appendix C Water Loss During Fallow from Vertisols and Alluvial Soils, with Drains at 1.0, 1.5 and 2.0 metres Depth Appendix D Reclamation Procedure for ten Cotton fields: Indicative Machinery Operations and Manpower Requirements inFLGURES Figure No. 1 2 3 4 5 6 7 8 9 Title Priority Areas For Subsurface Drainage Depth to Groundwater Contour Map (May 1984) Depth to Groundwater Contour Map (June 1984) Groundwater Salinity Contour Map (October 1984) Groundwater Contour Map Piezonetrie Sections - Melka Sadi Piezometric Sections - Amibara Projected Rate of Encroachment Without Drainage Proposed Locations of Additional Observation Wells 10 Leaching of Salts From Different Depths in Profile in Relation to the Cumulative Depth of Irrigation Applied 11 12 Leaching of Salts From Different Depths in an Undrained Soil Profile by Successive Irrigations as per Present Irrigation Schedule Expected Increase in Topsoil Salinity During Cotton Fallow Period as a Function of Salinity of Watertable and Soil Type. Drains at 1.8m or Waterrable at 0.75m at Start of 13 14 Fallow Expected Change in Level of Watertable During Cropping Season of Cotton with Drains at 1.Sm-Vertisol (Fine-Textured Soil) Expected Change in Level of Watertable During Cropping Season of Cotton with Drains at 1-8m-Alluvium (Medium to Coarse - Textured Soil) Appendix A A1-A15 Cl Observation Well Hydrographs Appendix C Variation in the Evaporation Coefficient (Et/E ) for Soil (Ks) with Depth of Watercable 0 C2 C3 C4 C5 C6 Fall in Available Storage Moisture with Time in Profiles of Stated Soil Texture and Drain (Watertable) Depth Variation in Daily Mean Eo (Pan Evaporation) from the Start of Fallow Variation in Crap (Kc) and Soil (Ks) Coefficients of the Et/Eo Ratio The Effect of Soil Salinity and Soil Moisture Content on Reduction in Vapour Pressure of Soil Solution Relationship Between Salinity of the Watercable and the Expected Rise Ln Topsoil Salinity During the Fallow Period for Stated Soil Type and Drain (Watertabie) Depth at Start DI D2 of Fallow Appendix D Location of Fields to Illustrate Typical Reclamation Procedure Schematic Illustration of Recommended Leaching Basin Layout IVANNEX B GROUNDWATER AND SALINITY1. INTRODUCTION 1.1 Location Anlbara Irrigation Project lias in the northern Rift Valley region of Ethiopia* in an area known as the Afar iriangle. The project area is situated adjacent to the right bank of the Awash River, and includes the settlements of Me 1 lea Sadi and Melka Warer. Access is provided by an all-weather link road, which connects with the Addia Ababa - Hille highway. The Rift Valley Escarpments rise some distance to the east and west of the site. 1.2 Climate The climate of the project area can broadly be described as semi-arid, with a bi-fflodal rainfall distribution amounting to some 500 - 600 am annually. (Ref.l). The rainfall pattern Is characterised by variability both on a month-to-month and year-to-year basis. Long rains extend from July to September, with short rains occurring between February and April. However, rainfall occurring in the intervening months as short duration, high intensity storm rainfall amounts to some 20% of the annual total. Temperatures range from mean minimas of 15”C and 23°C in December and June to mean maximas of 2l*C and 38*C in the respective months. Mean relative humidity is lowest in June (36%) and highest in August (58%), and mean daily sunshine hours amount to 8.5 hours on an annual basis. Evaporation rates, which have an important bearing upon soil sallnisatlon processes, have been estimated on the basis of the Penman method as ranging from a monthly mean of 170 mm in August to 252 mm in June. A more detailed discussion of the climate of the study area is presented In Annex C, 1.3 Soils and Geomorphology Within the project area three distinctive geomorphological units may be distinguished: the recent alluvial plain, a series of ancient fluviatile terraces trending roughly parallel to the present river course, and volcanic formations of varying age. The alluvial plain is a relatively recent geomorphological feature and shows abundant evidence of construction by fluvial processes, with old meanders and levees being readily identifiable on air photographs, Adjacent co the alluvial plain are a series of terraces which are most pronounced in the south of the project area, where two distinct terraces can be clearly seen. The higher of these is separated from the alluvial plain by a distinctive escarpment up to 10 metres high; the lower terrace is less pronounced, and both become discontinuous features cowards the north of the project area, where isolated remnants may be observed. Several irregular volcanic hills rise abruptly from the otherwise flat alluvial plain, and recent lava flows are a common feature In the vicinity of the project area. TThe generalised soil distribution within the project area reflects the recent geomorphological history. Coarser-textured levee soils occupy a discontinuous belt adjacent co the present river course, with finer textured ’basin* clays lying beyond this belt and extending towards the higher ground and river terraces in the east. Within this overall pattern, complex soil associations occur, and there is considerable soil variability, both spatially and vertically. Adjacent to the old terraces in the east of the area, occasional patches of saline soils remain, but have largely been excluded from the irrigation development. Elsewhere the soils are of variable but generally low salinity. 1.^ Geology and Hydrogeology The northern Rift Valley was faulted to its present formation during tectonic activity associated with the Tertiary and Eocene eras (Ref. 2). During the subsequent Plio-Pleistocene, outwash gravels containing clay lenses derived from the breakdown products of the elevated areas on the flanks of the Awash River were deposited on Che outwash plain adjacent to the river. Intermittent volcanic activity since the Eocene and associated outpourings of lavas has resulted in the construction of a gravel sequence interbedded with lavas* Towards the present flood plain of the Awash River, these deposits grade into an infill composed of a complex sequence of silts, sandy silts and clays which have In places been laid to depths exceeding 100m* The general geology of the area has been mapped by Halcrow (Ref* 3). The available drilling records show that all formations in the Awash River flood plain are water bearing. The sedimentary materials are consistently fine-grained and characterised by a low horizontal permeability, so chat they appear to form an aquifer chat Is essentially without flow. Aquifer-type response is mainly the result of inflow from rainfall and irrigation and outflow by evaporation. Such behaviour precludes the use of deep disposal methods for lowering the water table. Nevertheless, where boreholes have penetrated the gravels or fractured volcanics, water can be extracted in sufficient quantities for water supply purposes. Although confining layers may be encountered locally In the fine-grained sequences* the groundwater is believed to be in hydraulic contact regardless of geological formation. 1.5 Irrigation Practices Agricultural production under the Amibara Irrigation Project is based on the irrigation of cotton and bananas. The present gravity Irrigation system which supplies much of the currently irrigated area postdates earlier pump arrangements in which irrigation water was lifted directly from the Awash River. The irrigation of some 1300 hectares of bananas in the Melka Sadi State Farm Banana Unit was originally practiced using such a system, before being converted to a gravity supply. However, the old Algera, Ambash and Sublale State Farms, which produce cotton, still depend upon pumped supplies of irrigation water. 2The gravity irrigation system was designed on the basis of a 24 hour operation (Ref. 1), and comprises a network of secondary, tertiary and field canals which distribute water from offtakes on the main canal co the field. A limited number of flow control and measurement devices have been provided to assist in water management practices, by reducing conveyance, distribution and application losses. Furrow irrigation, which requires accurate land grading, is the method of field irrigation used in the project for cotton production in both gravity and pumped systems. For efficient operation, this application method necessitates the use of skilled irrigators to divide the water in the field canal into a number of furrow streams, and maintain the correct rates of flow until irrigation is complete. Bananas have a higher water requirement than cotton and require year-round irrigation under the climatic conditions which prevail at AnHbarg. A field irrigation arrangement suitable for tree crops, which consists of a series of small basins, is used to apply water to the crop in this case. The basin irrigation method is less sensitive to accurate land grading and requires less skilled irrigators than the furrow method. To ensure sustained agricultural production, a suitable drainage system must be regarded as an essential complement to the irrigation system, and provides a method of safe disposal of excess water. Surface drainage is required to remove excess water due to rainfall, irrigation and lateral seepage through canal banks, whilst subsurface drainage may be required under certain circumstances to control high groundwater tables and remove excess groundwater. At the time of conversion from pumped to gravity irrigation under Amibara Irrigation Project II groundwater tables in the Melka Sadi - Amibara area were generally at some depth and a subsurface drainage system was not an Immediate requirement. A surface drainage system was designed and installed in the project area, and the eventual need for a complementary subsurface drainage system was foreseen. 32. 2. L Amibara area on an intermittent basis since 1970 (Ref. !)*• These early measurements were derived from piezometers installetb-by—Tttflconsult (Ref* 4)» and also by the Awash Valley Authority. More recently, a network of observation wells was constructed in 1979/80 as part of the AIP II programme (Ref. 1). In recognition of the need to monitor movements of che groundwater table under irrigation. This network has subsequently been supplemented by additional wells and piezometers, the latter with the objective of studying vertical water movements and perched water tables. Unfortunately several of the early piezometers have subsequently been damaged and are now unserviceable, although some records do exist. Available information describing the number of observation wells and piezometers installed by individual organisations during successive phases of groundwater Investigations is presented in Table 1, To assist che reader when referring to the cited references, che identifying codings used by the various organisations on maps, to distinguish their monitoring devices, have also been listed. A number of the more recent observation wells have been damaged, and regular monthly monitoring has not always been possible, with the result that substantial gaps exist In the records. Moreover, the variety of observation well and piezometer designs which have been used necessitates particular care being taken in data interpretation. The operational status of observation wells and piezometers in the Melka Sadi - Amlbara area, as In October 1984, is summarised in Table 2. 2.2 Review of Previous Studies During their feasibility study of the Melka Sadi - Amlbara area in 1969, Italconsult prepared a piezometric map which indicated che presence of a trough in the water table on the eastern aide of the Awash alluvial plain (Ref. 4). On each side of this zone, higher water table elevations were observed Indicating some recharge from the river and from hills bordering the plain. The alluvial plain was believed to consist of a very deep and uniform interbedded sequence of silts, sandy silts, and silty sands, on the basis of piezometer logs. Sections showing che piezometric gradient were plotted by ICalconsult» with the following findings;- (1) A low hydraulic gradient towards che river in the Melka Sadi area; (2) A Low hydraulic gradient away from the river in Che Amlbara area; (3) A hydraulic gradient of approximately IX in a southwest — northeast direction across the Melka Sadi - Amlbara area. 5TABLE 1 NUMBER OF OBSERVATION WELLS/PIEZOMETERS INSTALLED BY INDIVIDUAL ORGANISATIONS AND IDENTIFYING CODINGS ORGANISATION TO WHICH OBSERVATION WELLS/PIEZOMETERS ATTRIBUTABLE: Italconsult identifying CODING NUMBER INSTALLED NOTES PPI, PP5 etc 4 P2, P3 etc 10 DW-1, DW-2 2 Deep plezometric well (50m) Shallow piezoroetric well (15o) Deep piezometric well (100m) Awash Valley Wp etc 4 1' No Information on Authority (A.V.A.) installation depths available Sir William Halcrow A Partners Foraky S.A. ABSl, 2 etc 15 Piezometers installed 1 Foraky S.A. Aoibara Irrigation Project II AIP 1, 2 etc 672/ Observation wells installed to variable depths depending on position of water table ATP 3/1, 3/2, 153' Piezometers installed 3/3 10/1. 10/2, adjacent to 10/3 etc. observation wells AIP 3,AIP 10 etc. 1/ Wells Irreparably damaged, March/April 1973 2/,3/ Several wells subsequently damaged and inoperative. Total number of planned observation wells not conscructed/handed over by October 1984. 6It was further concluded chat (.with groundwater flow velocities (Darcy velocity) In the order of 1(J m/sec (based on horizontal permeability measurements) rhe aquifer was essentially without flow varying in a vertical sense in response to inflow from precipitation and outflow by evapo t ranspirat ion. Subsequent studies in 1972/3 (Ref. 2) suggested rhe possible existence of two sets of aquifers below the Amibara plain. The deeper, unconfined aquifer occupying gravel beds interbedded with lavas, and shallow aquifers of variable depth occurring in alluvium above the confined aquifers, being supplied by water derived from rainfall, floods, the river and from irrigation. In 1975, the Angelele - Bolhamo feasibility study (Ref. 3) included hydrogeological investigations supported by the installation of fifteen additional piezometers. The earlier findings of Italconsult were largely confirmed. Sources of recharge were studied and the possibility of recharge occurring in the Melka Sadi area from the deep percolation of irrigation water and from canal seepage was recognised. More recently, drainage and salinity studies (Ref. 5) conducted during 1981/82 in the currently irrigated area indicated the existence of a perched water table within 1 to 2 metres of the surface in the Melka Sadi area, with locally artesian conditions. 2.3 Review of Available Data The available data coverage Is summarised In Table 3. The Awash Valley Authority piezometers were monitored from 1972, but all four piezometers in the Melka Sadi - Amibara area were destroyed In March/April 1973. Records for the Italconsult piezometers are available for the period 1971 to 1978. During thia period, several piezometers were destroyed and when monitoring was resumed In 1980 under AIP, several could not be located. Monitoring of the additional piezometers Installed by Foraky S.A. during the Angelele • Bolhatno feasibility study was apparently discontinued on completion of the study. The AIP piezometer network was monitored by AVA from August 1980 until September 1982, when responsibility was transferred to AIP. During this period, measurements of the electrical conductivity of the groundwater (EC) were also made. Regular monitoring was unfortunately not resumed until November 1983, and EC measurements were discontinued owing to the lack of a suitable Instrument.TABLE 2 STATUS OF OBSERVATION WELLS AND PIEZOMETERS, OCTOBER 1984 OBSERVATION WELL/ PIEZOMETER TYPE REFERENCE NUMBER STATUS Italconsult DW-1, DW-2 PPI, PP5 PP12. Not known Functioning P2, 5, a, 9, 11, P3, 7 Out of use Functioning Pl, 4, 5, 6. 10, 13 Out of use Not known PP 14 Not known Awash Valley Authority (A.V.A.) Wl. 2, 3, 4 Out of use Sir William Halcrow 4 Partners/Foraky S.A. ABS/I-15 Not known Amibara Irrigation Project II AIP 2-67 Excluding: AIP 1, 13, IB, 33, Functioning Out of use AIP 30, 58, 59, 60, 62, 63, 64, 65, 66, 67 Not yet constructed/ satisfactorily handed over.TABLE 3 AVAILABLE DATA COVERAGE ORGANISATION TO WHICH OBSERVATION WELLS/ PIEZOMETERS ATTRIBUTABLE 3/ ITALCONSUL1 AWASH VALLEY AUTHORITY (A.V.AJ2/ SIR WILLIAM HALCROW & PARTNERS + FORAKY S.A. L/ AMI BARA IRRIGATION PROJECT II ‘ Hi 68 69 1970 71 72 73 74 75 76 77 7B 79 1980 81 82 83 84 1/ Records of Individual wells/piezomoters intermittent 2/ Monitoring of observation wells prior to 1972 not known 3/ Intermittent monitoring prior to 1970 4/ Installation of observation wells commenced in 1980 as part of a phased construction programme; monitoring of flrec wells commenced in 1980Analysis of Data The available data was analysed with the primary objective of delineating priority areas for subsurface drainage, as specified In the proposal tor this study (Ref, 6), Further analyses were carried out to examine the nature of groundwater movement within the aquifer, and to predict rcal*tiM changes in groundwater levels as a basis for ’wlth-and without- project’ projections to be used in the economic evaluation of the recommended project (Annexes E and H)- Where contour maps based on point data have been required to assist in the analysis, individual contours have in all cases been interpolated assuming a linear gradient between adjacent observations. Areas have been measured at map scale using a zero setting compensating polar planimeter, with vernier scale, and stated areas are the average of at least three readings. 2,4.1 Depth to Groundwater Contour Maps The most recent and complete monthly observation well data (February to October 1984) were plotted on separate maps, and contours of depth to groundwater interpolated. Individual maps were then superimposed. and 'best fit' contour lines were drawn, so enabling the effect of missing data to be minimised. The approximate location of the 1 metre and 6 metre depth to groundwater contours, corresponding co the 'lnnnediate priority’ 1 and ’priority within next 5 years areas (Ref. 6) are shown on Figure 1 and the approximate areas enclosed by these contours are listed in Table 4. To illustrate the general arrangement of the contours, depth to groundwater maps for May and June 1984 have been included as examples (Figures 2 and 3), chosen as being generally representative of conditions during fieldwork. TABLE 4 PRIORITY AREAS DELINEATED BY DEPTH TO GROUNDWATER CONTOURS AREA (ha) LOCATION Melka Sadi Amibara TOTAL IMMEDIATE PRIORITY1 2247 1597 3644 PRIORITY WITHIN NEXT 5 YEARS2 4093 3897 7990 1/ Immediate priority defined by 3 metre depth to groundwater contour; these are referred to as Stage 1 areas. 2/ Priority within next 5 years defined by 6 metre depth to groundwater contour. 102.4.2 Groundwater Salinity Contour Maps Measurements of groundwater salinity were based on determinations of the electrical conductivity of water samples abstracted from observation wells, and were resumed in June 1964 following a break of some 16 months. Although the records are of limited extent, groundwater salinity contour maps were prepared for the months June to October 1964 inclusive, and the map for October 1984 has been included to illustrate the predominant arrangement of contours (Figure 4). Although successive maps are broadly similar in terns of contour arrangement, Individual maps are characterised by variability. Pronounced and sustained peaks occur In Melka Sadi area at observation wells AIP 25 and PH, and in Amlbara area at AIP 42. The overall configuration of contours appears to bear little resemblance to depth to groundwater and groundwater contour maps, and cannot be satisfactorily explained on the basis of the limited set of data available. The groundwater salinity data as obtained from observation well water samples was also compared with measurements of water salinity obtained during soil survey fieldwork, at sites where groundwater was encountered within augering depth (Annex A). This comparison revealed that the correlation between measurements taken at sites in close juxtaposition was very poor, and in several cases the observation well measurements were considerably lower than those at nearby augering sites. A similar phenomenon had been observed in earlier studies (Ref. 5), and the situation Is further complicated by the considerable spatial variability in groundwater salinity observed at successive augering sites. In developing the without-project' projection to be used in the economic evaluation of the recommended project (Annexes E 4 H), groundwater salinity data is required in predicting the rate of secondary sallnisation from a saline groundwater surface in an undrained situation. For the purpose of these predictions, groundwater salinity data has been obtained from observation veil records, because unlike the soil survey data, the former are available on a time-related basis. 2.4.3 Groundwater Contour Haps Observation well data for February and May 1984 were used in conjunction with topographic survey data to derive the reduced level of the groundwater surface at each observation well. Groundwater contours were then interpolated, and the predominant flow directions determined. In the tinned late vicinity of the Awash River, the observation well network is such that accurate positioning of groundwater contours was not possible, and hence contours were not extended to the river when preparing the maps. Although the river itself could therefore not be identified conclusively as either a source of groundwater recharge or a sink, especially in the Melka Sadi area, this was not considered to adversely affect rhe validity of the overall interpretation* TMaps were initially prepared for two months data on the assump on that substantial changes in the direction and rate of groun water flow would be most unlikely over a short tine interval. This assumption has subsequently been verified by preparing additions groundwater contour maps. These show that the general arrangement □f contours has changed little during the period February October 1984, the prevailing configuration being illustrated on Figure 5. Comparison with groundwater contour maps prepared during the study of groundwater and salinity conducted during 1961 (Ref, 5) confirmed that although a rise in groundwater has been sustained, the relative disposition of contours has changed little during the intervening years. 2.4.4 Observation Well Hydrographs Hydrographs were drawn from historic data for all observation wells for which data was available, including those for which the records are substantially incomplete. The resulting plots for Individual observation wells are presented in Appendix A. Each hydrograph was examined for evidence of seasonal movements about a mean trend, and several (eg AIP 39) predictably show a falling limb during the period October to March, with a rising limb during the irrigation season extending from May to September. This seasonal pattern was less clear in other hydrographs (eg AIP 36) , and several showed no evidence of seasonality (eg AIP 15). For each hydrograph, the approximate rate of rise of groundwater was also estimated by assuming a linear race of rise, and constructing visually the ’best fit’ straight line through the plots. The approximate average rate of rise was then determined 1 from the gradient of the ’best fit line (Appendix B). Although individual hydrographs may show considerable fluctuations, the following observations can be made;* (1) An overall rise in groundwater levels is being sustained. (2) The overall rate of rise is in the order of 1 metre per year. (3) On the basis of the limited records available, the aquifer appears to be responding in a vertical sense to deep percolation losses. Fluctuations in individual hydrographs for observation wells at which groundwater is currently at some depth are difficult to attribute to specific hydrological events, but in some cases (e.g* AIP 55) appear related both to seasonal irrigations and river discharges. To assess the possibility of rates of rise varying as a function of depth, several hydrographs were plotted for long-established observation wells which have been monitored Intermittently since 1970. Inspection of these showed that rates of rise wererelatively constant until the groundwater came within approximately 2 metres of the ground surface, when the rate of rise decreased markedly. In recognition of the importance of the projected rate of rise of groundwater as an index in determining the need for subsurface drainage, fluctuations in groundwater depth in areas where the groundwater was already within 2 metres of the soil surface were treasured for a limited period. The results showed considerable variations in response to individual Irrigations, such that a generalised terminal rate of rise, for predictive purposes, could not be reliably established- To estimate rhe rate of rise of groundwater above a depth of approximately 2 metres, reference was therefore made to the hydrographs of observation wells situated in areas where groundwater had risen close to the surface (observation wells Pll, AIP8, AIP3) for which the terminal race of rise between depths of 1.5 metres and 1 metre below the soil surface has been estimated at 0.3 metre per year. The prediction of changes in groundwater depth with time is discussed in Section 2-7 and the estimated magnitude of p salinlsatlon by capillary rise, which has an important implication in determining the need for subsurface drainage, is discussed in Section 3,7. 2.4.5 Piezometric Sections Cross sections approximately perpendicular to the groundwater contours and intersecting selected observation wells were plotted for both the Melka Sadi and Amibara areas to investigate the relationship between groundwater and Awash River levels (Figs 6 and 7), These sections are based on observation well data for May 1984, with river bed Levels based on available topographic survey data. The following observations can be made:- (1) In the Melka Sadi area, the piezomecric surface slopes towards rhe Awash River with a gradient which ilea approximately in che range 1 in 400 to 1 Ln 900 along the selected sections. With an assumed flow depth of 2 metres above bed level, representative of non-flood flow conditions, the river Levels are similar to those projected from observation wells such that its influence on recharge or drainage cannot be clearly determined. However, under prolonged high flow conditions, where river levels prevail at or near to bank level then an appreciable positive hydrostatic head is available for recharge from the river into groundwater. (2) Ln the Amibara area, the piezometric surface slopes away from the river, with a gradient lying approximately in the range 1 in 400 to I in 700 along the selected sections. The predominant flow depth in che river lies below che !3adjacent water levels in observation wells. At flood flows, however, the river may act as a source In thia area. The possible existence of perched water tables* particularly in the Melka Sadi area, is a further factor which must he taken into account in interpreting the piezometric sections, and is discussed in the following section* 2*5 Perched Water Tables Within the Melka Sadi area, the possible existence of a perched water table> separated from the deeper groundwater by an impermeable layer of varying thickness (Ref. 5) was noted in previous studies. The detailed soil survey fieldwork has demonstrated that the relative disposition of relatively impermeable horizons 17 in the Melka Sadi area is such that the occurrence of perched water tables is intermittent. but may nevertheless extend over a considerable area (Annex A). A continuous relatively impermeable horizon seems unlikely to exist. These more recent findings are consistent with the existence of localised conditions of partial confinement in the fine grained alluvial sequence underlying Melka Sadi, in which the groundwaters are in direct hydraulic contact (Ref. 1). Where relatively impermeable horizons exist close to the soil surface, the overall rate of downward movement of percolation water is likely to be considerably attenuated, with water moving laterally before moving downwards in areas where the Impermeable layer is absent. The apparently rapid lateral advancement of groundwater as has been observed in parts of the Melka Sadi area is probably explained by this phenomenon. In the Amibara area* the alluvial sequence is less complex, and perched water tables appear to be of very limited extent. 2*b Identification of Priority Areas for Subsurface Drainage The proposal (Ref* 6) specified that drainage works required for the reclamation of seriously affected areas, or areas likely to become so within the next 5 years (indicative hectarage of 4000 ha) would be designed in detail. In order to delineate these areas during the Identification Phase as a basis for the more detailed subsequent studies, criteria were established 1/ A relatively Impermeable horizon is defined as one whose hydraulic conductivity is less than ten tines that of the soil horizon immediately above, irrespective of soil textural group. In the highly stratified, generally fine-textured soils which predominate at Amibara, the relatively impermeable horizon can often be readily identified on a textural basis as a distinctive, compact layer of silty clay or clay. This horizon is characteristically a feature of considerable lateral extent, and should not be confused with minor horizons of similar texture which comtaonly occur closer to the soil surface, but which are spatially discontinuous features. Mwhich enabled those areas considered co be currently seriously affected (May 1984) to be differentiated from chose considered Likely to become so within the indicated 5 year period* The criterion used to delineate priority areas for subsurface drainage was expressed in terms of depth to groundwater, based on the following considerations:- (1) Critical watertable depth in relation to the capillary salinlsaclon hazard (a function of soil type) and groundwater salinity* (2) Watertable depth in relation to the depth of the root zone for specific crops* Cl) Rate of rise of groundwater. Taking into consideration the range of soil types encountered, the crops being produced, the rate of rise of groundwater and the groundwater quality, immediate priority areas (Stage 1 areas) were tentatively defined by watertable Levels currently at or within 3 metres of ground level. This criterion is comparable with similar specifications elsewhere (Ref* 7)* Areas likely to become seriously affected within the next 5 years were similarly tentatively identified* Assuming current rates of rise of groundwater are sustained, these areas are delineated by the 6 metre depth to groundwater contour. The extent of priority areas so defined are shown on Flg.l and the corresponding areas presented in Table 4* The tentative boundary of the immediate priority areas defined on this basis were largely confirmed during the detailed 2*7 Predicted Changes Ln Groundwater Depth With Time soil survey fieldwork* Predictions of changes in groundwater depth with time were required as a basis for implementation phasing and, in conjunction with corresponding yield predictions, for economic evaluation of the proposed drainage works. The approach to these predictions has been based on average rates of rise of groundwater determined from historic observation well records, and has been briefly outlined in Section 2.4.4. For each observation well, the ’best fit’ straight line from which the average race of rise was measured was used to predict the year (Y Yp ^2 etc^ which rising groundwater would reach a depth of approxSmartly 1 metre below soil surface (Y refers to September 1984). The extrapolated rate of rise was°coQtinued until the predicted depth to groundwater reached approximately 1,5 metres below soil surface, when a reduced rate of rise of 0,3 metres per year was applied (Section 2.4.4). The extrapolation was terminated either at Year 20, or when the predicted depth to groundwater reached a depth of approximately L metre below soil surface. This depth was chosen as being characteristic of land within which substantial areas were clearly observed to be affected by high groundwater cables. Above this depth; predictions in any event become very unreliable because of short-term fluctuations resulting from 15Individual irrigations (Section 2.4.4). The time-related groundwater depth predictions are sunuaarlsed in Table B.l (Melka Sadi area) and Table 3-2 (Amlbara area). Time-related boundaries between adjacent areas lying within the Y , Yj q .......... Y^ categories were determined in a spatial context by plotting on a base map at the appropriate observation well locations the years in which the depth to groundwater was predicted to reach a depth of approximately 1 metre below soli surface. Boundaries were then produced by linear interpolation between adjacent sites, and the resulting configuration is shown in Figure 8. The planimetered areas between adjacent boundaries have been used for both economic evaluation and implementation phasing purposes. Computerised Groundwater Happing During the Identification Phase, the suitability of groundwater mapping by computer, for use as a tool during the study, was investigated. An existing program was used, and was run on a Vax computer equipped with a plotter. Depth to groundwater contours were plotted using a program which enabled a set of randomly placed three dimensional data points to be interpolated onto a regular rectangular grid, the contours being plotted as a check on the resulting surface. The program consisted essentially of three sub-routines, the first performing the interpolation, the second plotting the contours and the third designed to plot the original data points onto the contour plot as an additional check. The first sub-routine performed the Interpolation using nested iterative schemes, fitting a surface to all the given data points lying within the required regular grid by using variable equations approximating to the Laplace and Spline equations. The grid was also used by the second sub-routine to produce a contour plot, the accuracy of the plot depending on the size of the grid. The third sub-routine plotted the observed random data points on the contour plot produced by previous sub-routines, thereby enabling an additional check to be made. Because of the effect of missing data, the resulting maps were found to be of limited value, and it was felt that more reliable contours could be drawn by hand. An additional disadvantage of computerised groundwater mapping was the need to send data away from site (to Head Office in England) for processing, the resulting maps being returned to site for interpretation. Following this exploratory exercise, further groundwater mapping was therefore carried out on site manually. When suitable computing facilities are available locally, the feasibility of computerised groundwater mapping should be re-examined. Recommendations For Additional Observation Wells The future monitoring of groundwater movements within the entire study area requires the installation of additional observation wells, and it is considered that a further 10 such wells would provide a minimum level of coverage in the Amlbara Extension area. The proposed locations of these additional wells are shown on Figure 9, and it is recommended that theirdesign should be similar to those of wells previously constructed under the Amibara Irrigation Project II Contracts. However, the provision of a lockable cap would help to prevent damage resulting from vandalism, which has often led to wells becoming inoperative in the past. In the Melka Sadi and Amlbara areas, all existing observation wells constructed under the AIP II Contracts which have subsequently been damaged should be repaired and returned to use. 2-10 Conclusions The magnitude of subsurface field drainage priority areas within the study area assessed on the basis of depth to groundwater contours has been shown to substantially exceed the initial estimates based on earlier work as stated in the proposal (Ref. 6). The criteria used to determine critical groundwater depths assume that current rates of rise of groundwater will be sustained. In the ins&ediate future this assumption is tenable* but during the projected 5 year period addressed by the study any measures which can be introduced to reduce the rate of rise will be beneficial. The predicted changes in groundwater depth with time are similarly based on the assumption of sustained rates of groundwater rise, as obtained from analysis of previous records. They should be regarded more as an Indication of the changes likely to take place rather than as a definitive statement* and should be the subject of revisions as more data becomes available. Perched water cables requiring remedial drainage works may not necessarily be reliably identified through review of observation well data* because of their frequently localised character. Nevecheless* information obtained during the detailed soil survey (depth to groundwater contours Interpolated from measurements recorded tn auger holes and profile pits) has shown good agreement with the delineation of Immediate priority areas. 17T0 n3 3.1 SALINITY Background Feasibility studies in connection with the proposed Amlbara Irrigation Project confirmed the suitability of the soils for irrigation development, but recognised the likely future hazard of salinlsation and rising water tables under sustained Irrigation. Since large-scale gravity surface irrigation commenced in 1981, the groundwater level has Increased and at the end of the 19S4 cropping season was close to the soil surface in substantial parts of the project area. Attendant with rising groundwater levels, there has been a marked Increase in the annual cultivated area abandoned through salinization, and in recent years several studies have been conducted specifically with the objective of addressing this problem. Detailed agronomic and soils fieldwork conducted during the current study have shown the observed salinity problems to be of a complex nature, influenced not only by physical conditions of the site but also by soil and water management techniques. Although this annex ia concerned specifically with salinity, reference should therefore also be made to appropriate sections of the Main Report, Soils and Drainability Classes and Agriculture technical annexes. 3.2 Review of Previous and Concurrent Studies The prefeasibility studies commenced by Sogreah in 1962 (Ref. 8) led to the identification of potentially irrigable alluvial and colluvial soils in the Melka Sadi* Amlbara and Boihamo areas, in which a semi-detailed soil survey was carried out. Feasibility soil studies in connection with the proposed Amlbara Irrigation Project were completed in July 1969 by Italconsult {Ref. 9) and covered an area of some 2B,000 ha. Soil and land classification maps at a scale of 1:20,000 were produced. In 1971, an additional soil survey was conducted (Ref.10) to increase the survey intensity and further characterise the soils in terms of permeability, infiltration, salinity and alkalinityi leaching behaviour and nature of the substrata. Irrigation suitability land classification maps at a scale of 1:20*000 were produced, but the 1969 soil maps were apparently not revised. These maps form an important basis for the currant studies, and cover the majority of the project area. The generalised soil distribution within the project area is discussed in Annex Al and shown on Map 1, based on the earlier Italconsult maps but revised on the basis of current soil survey fieldwork. The Angelele and Bolhamo feasibility study (Ref. 11) was carried out, from 1973-75 and included a soil survey of some 8,500 hectares of potentially irrigable land in Angelele adjacent to the northern boundary of the currant study area. With the exception of slight overlap in coverage at the boundary, this survey lies outside the present study area, but is nevertheless useful for comparative purposes. 19A hydropedological study was carried out in 197£ (Ref* 12) wit objective of evaluating available information on the soils and irr gat on water In reLation to the effects of intensive irrigation on water table levels salinity and waterlogging of surface soils. A further objective was co determine the infiltration characteristics of the soils and assess the leaching procedures chat might be required to maintain a favourable salt balance* One of the principal conclusions of this study was that deep percolation necessary to maintain the salt balance together with seepage losses would result in a rise of the phreatic layer, and the possible need for drainage about five years after implementation of the Irrigation scheme. In 1981,. high water tables and salinity problems were developing in the MAESCO banana plantation, Melka Sadi, and a more detailed drainage and salinity study was initiated leading to preliminary drainage design for this area (Ruf* 13). The study was subsequently extended, at a reduced level of detail, to Include the remainder of Melka Sadi and Amlbara, in anticipation of the eventual need for subsurface field drainage* A number of hydropedological maps were produced which are essential for comparative purposes. The need for drainage trials in the .Melka Sadi area was also identified, and a pilot drainage scheme was constructed in 1983 on a 35 hectare site within the MAESCO banana plantation (Ref. 14). The site was selected on the basis of proximity to saline areas within the banana plantation, the extremes of salinity within the site Itself, the existence both of abandoned salinised land and productive areas within the site and proximity to the primary drain. The soils did not include vertisols. These trials, which are continuing, included reclamation and leaching studies, and required a detailed soil survey of the site at an observation density of 1 angering (to 5 metres depth) per hectare. A large number of soil physical tests were carried out to study the relationship between drain performance and soil characteristics. 3.3 Additional Field Investigations During the current study,informatlon on soil salinity was obtained through the soil survey fieldwork, during which 350 profiles were described, and also from the agronomic fieldwork. Data from the former were used primarily for characterisation purposes, and to study the spatial distribution of salinity, The latter studies were designed to quantify the relationship between salinity and crop yield, and the effect of watertable depth on yield under specific site conditions prevailing at Amibara. This data was essential in formulating leaching requirements and the drainage design criteria. For detailed information describing observation densities and survey methods used during the fieldwork, reference should be made to the appropriate technical annexes. t 203.4 Characterisation of Saline Soils Saline and alkali soils (collectively referred to as rsalt-affected soils’) are defined in ten? of the electrical conductivity of the saturation extract (ECe) and exchangeable sodium percentage (ESP) as follows:- (Ref.15)* Soli Characteristic ECe ESP Saline > 4 mS/cD «•/ <15 f Saline-Alkali > 4 mS / cm >15 ' Nonsallne-Alkali < 4 mS/cm >15 „ Soils exhibiting each of these characteristics are represented In the study area. The inherent variability and patchiness of salt-affected soils makes accurate mapping even at derailed soil survey intensities problematical, so that during the Identification Phase, rhe extent of salt-affected soils was tentatively outlined by qualitative correlation with yield response. During the Study Phase, soil salinity maps at a scale of L;2Q 000 were prepared for the Immediate priority areas, based on detailed soil survey observations supported by Laboratory analysis of soil samples from profile pits. (Annex A2). Soil salinity mapping classes were selected to assist in Che interpretation of agronomic data» in addition to reflecting the spatial variations In salinity. Earlier findings (Ref. 9), which reported rhe nonsaline - alkali group to be of very limited extent in the study area, were largely confirmed. 3.4.1 Inherently Saline Soils Soluble salts occur in soils through the degradation and weathering of primary minerals, and also from ground and surface water. They consist principally of various proportions of the cations calcium, and magnesium, and the anions chloride and sulphate. (Ref. 15). Under climatic conditions in which evaporation exceeds precipitation. the predominantly upward movement of soli water results in a concentration of soluble salts in rhe surface horizons, indicated by white salt encrustations on the surface in extreme situations. For the purpose of the current study, salinity of this type has been termed ’inherent salinity’. Areas of inherent soil salinity were defined and mapped by Italconsult (Ref. 9) and were largely excluded from the Irrigation development. Localised soils of this type nevertheless occur in 1/ The term 'alkali is an older term which is now often replaced by the 1 word 'sodic’j both terms convey a similar meaning. 21Melka Sadi and are closely associated with poor co yields- Because of their very limited ext^^jC^__^4aRrt distribution, no attempt has been made soils as a separate mapping unit on the maps. 3.4.2 Soils Affected by Secondary Salinity to differentiate these 1120 000 soil salinity Secondary salinization results from the upward movement ofsalts from saline groundwater by capillarity» under conditions of Inadequate leaching. The rate at which salts are transported upwards and eventually deposited in the upper soil horizons Is dependent upon a number of factors, and is discussed in Section 3.7 and Appendix C. The critical capillary height, beyond which upward movement of soil water by capillarity effectively ceases, is dependent upon soil texture and particle size distribution, and is therefore related to soil type* For silt loams, which occur frequently in the study area, this critical height is In the order of 2 metres; stratification tn the profile reduces the critical capillary height. “In Melka Sadi and Amibara the excess irrigation water including Chat apparently required to maintain salt balances (Ref. 5) has resulted in a rapid rise of groundwater which Itself may be highly saline or alkaline. Where groundwater comes within the capillary zone, secondary salinisatlon takes place. Substantial areas of the Melka Sadi banana plantation (where groundwater is currently close to the surface) are suffering from sallnisation of this type, as shown by white salt encrustations in several abandoned fields. The alluvial soils in this area are of variable texture and the quality of groundwater varies considerably; the salinised areas therefore have an irregular distribution and are sometimes interspersed with productive fields* The pattern is further complicated by the apparent existence of perched water tables in this area. Further fields within the Amibara Settlement Farm are now suffering from salinity attributable to rising saline groundwater. 3.4.3 Initially Alkali Conditions Alkalization results from the accumulation of exchangeable sodium in soils, through the phenomenon of cation exchange. Sodium, calcium and magnesium cations are always readily exchangeable and calcium and magnesium are the principal cations found in the soil solution and on the exchange complex of normal soils in arid regions. 22When excess soluble sales accumulate tn these soils, sodium frequently becomes the dominant cation in the soil solution. This may either be an Inherent situation, or it nay result from precipitation of calcium and siagr.eslute compounds. As the soil solution becomes increasingly concentrated through evaporation and water absorption by plants, the solubility 11mi rs of calcium sulphate, calcium carbonate and magnesium carbonate are often exceeded such chat they are precipitated. The result is a corresponding increase in the relative proportion of sodium in the soil solution, and under such conditions, part of the original exchangeable calcium and magnesium is replaced by sodium on the exchange complex. Calcium and magnesium cations in the soil solution are more strongly absorbed by the exchange complex than sodium, a factor which is of considerable importance in the reclamation of alkali soils. Adverse physical conditions (very Low infiltration and permeability rates, and poor structure) are normally associated with alkali soils, and result from deflocculation. They may often be recognised as black patches or * slick spots'* The extent of initially alkali soils within the study area was found during detailed soil survey fieldwork to be very limited, occurring within complex associations of saline-alkali and saline soils. As such, their distribution was too localised to permit reliable mapping, and similar to the ’Initially saline’ soil group, no attempt has been made to differentiate these soils as a separate mapping unit on the 1:20 000 soil salinity maps. 3.4*4 Soils at Risk of Alkalisation Following Leaching This category of sails includes saline-alkali soils, which although of limited extent, occur in complex associations with saline and alkali soils. On leaching saline-alkali soils, in which the soil particles may be flocculated prior to leaching, there is a risk that alkali conditions will develop as soluble salts are removed, and that adverse physical conditions normally associated with alkali conditions will develop. The successful leaching of saline-alkali (and alkali) soils relies upon a sequential process involving the following stages (1) Replacement of exchangeable sodium in the exchange complex (usually by calcium), followed byi (2) Removal of excess soluble salts by leaching, as in the case of saline soils. The cation used to replace exchangeable sodium during the reclamation process is partly dependent upon soil characteristics and is often calcium, applied to the soil as an amendment (e.g. calcium carbonate or gypsum). However, In certain circumstances the soil itself can, provide a source of exchangeable calcium, in which case the process outlined above takes place concurrently as the soil ie Leached. 23As part of the Melka Sadi Pilot Drainage Scheme programme, leaching trials were conducted on saline-alkali soils to investigate the possible need for soil amendments. The resu s these trials are summarised in Section 3.9. 3.5 Distribution of Saline Soils As part of the detailed soil survey programme, soil salinity maps for both Melka Sadi and Amibara areas were prepared at a scale of 1:20 000. The derivation of salinity mapping classes on which these maps are based is explained in Annex A. Soil salinity is widespread in the southern part of the Melka Sadi immediate priority area, but almost absent in the north. Where salinity does occur, the local distribution is often complex, with non-saline soils situated in close juxtaposition with highly saline soils. Terrace soils immediately south east of the currently developed area exhibit high salinity and high alkalinity. The majority of these soils were excluded from the irrigation development, with the exeption of some 170 ha close to the village of Melka Sadi. In this area, the weathered tuff material which forms the substratum is often saline, and in addition it has high to very high pH values. The general levels of salinity in the Amibara area are considerably lower than chose in the Melka Sadi area. An area of saline soils is situated in the Awash Settlement Farm between the primary canal and flood protection dyke within the Amibara immediate priority area. Other isolated areas of salinity have been observed in several locations in the vicinity of the primary canal. 3.6 Irrigation Water Quality In the context of soil salinity, Irrigation water quality is of Importance because salts dissolved in the irrigation water can accumulate in the soil profile, with damaging effects upon crops. It is normally evaluated with reference to the salinity hazard and sodicity hazard, and previous studies (Ref. 13) have shown Awash River water to be of good quality, ranging from C. to C. S? (low salinity and sodicity hazard to low salinity and moderate sodicity hazard, (Ref. 15) depending on time of year, with an average electrical conductivity of about 0.35 mS cm” . 1 Where groundwater is at depth, or where subsurface field drainage is Installed and irrigation is perennial, the damaging effects of irrigating with water containing dissolved salts can be minimised by providing a component of water in individual irrigations for leaching purposes. In concept this component is additional to the crop water requirement, and may be defined as that fraction of the irrigation water that must be leached through the root zone to control soil salinity at a specified level. Previous studies (Refs. 12 and 17) have shown that under these conditions, in which deep percolation losses are estimated to lie in the range 15-302, the leaching requirement is more than satisfied by deep percolation losses. 24Where groundwater is saline and close to the soil surface, leaching cannot be effective in the absence of drainage, and salts brought in with irrigation water accumulate in the soil profile with detrimental effects on crop yields. This situation is of importance for comparative purposes in economic evaluation of the recommended project, and is discussed in detail in Section 3.7. Once drainage has been installed and the saline groundwater is controlled at or near drain depth, salts will accumulate in the soil profile during the fallow through the process of secondary salinisation. During the cropping season, irrigation applications must be such as to remove these salts, in addition to removing salts introduced with irrigation water. An irrigation schedule has been developed which will achieve this once drainage has been Installed, and is discussed in detail in Section 3.10. 3.7 The Development of Salinity in Irrigated but Undrained Land The amount of secondary salinisation mainly is dependent upon the soil texture, depth of groundwater and its salinity, and the length of the dry period during which the mass flow of soil solution is upward. The model of secondary salinisation is described in Appendix C. The situation where the watertable is rising is one in which deep drainage is small compared with irrigation and evaporation. In such a nearly closed system, in which it is assumed that autochthonous release of salts by weathering is balanced by removal of salts in the crop product, the salt contained in the irrigation water is critical to the salt balance. The amount of irrigation water is dependent on the type. 3.7.1 Cotton The irrigation schedule currently in use applies a total of 1290 mm ^f water annually, at an average conductivity of about 0.35 mS cm . Supposing that 10Z of the salts contained in this water are . lost from the site in deep or lateral drainage, then the net delivery of salts is 3.15 me/1. This water passing through a nb^onal soil column of sectional area 50 cm* amounts to 6.45 1 in e and contains (net) 2O.3me of salts. If these salts are leached into the top 20 cm of the watertable, on average they are co^llined in about 500 ml of water. Therefore, the conductivity of the top 20 cm of the groundwater is alo^e is sufficient to explain the extremely saline parts of the estate. annual.rise in 4mScm . This groundwater in iIrf t thi hiss s saalltt m move ovess upw upwaarrdsds dur duriing t ng thehe f faallllowow pe perriiod a od nd is distributed in the pattern observed in the Pilot Drainage Scheme area (PO 4) then the salinity following amount: Depth from surface (cm) A ECe (annual) X profile would change annually by the 25 50 75 100 0.38 0.08 0.04 0 9 2 1 0 The weighted mean (plough depth) is incremental 3.15 -*-“r rise in ECe in the cop 20 cm of soil J • • 25This annual incremental increase in salinity is supplementary o the secondary salinity AECe which is predicted to occur rom a watertable of specified depth and salinity in a soil, of specified texture. The depth of the watertable depends mostly on the time of year as the general level approaches the soil surface, fluctuating in response to irrigation and receeding due to deep drainage and evaporation. The latter is markedly greater when the watertable is close to the surface and its rate falls as the watertable drops. There is a mean equilibrium level which is assumed to be at 75cm at the end of the cropping season. The secondary salinisation to be expected from this depth is illustrated in Fig. 12. The soils of the Pilot Drainage Scheme site are broadly representative of the estate as a whole. The mean salt release curves, determined in leaching basins, are shown in Fig.10, with the y-axis transformed to the percentage of initial level of salinity. These curves were determined during reclamation with drains at 2m and may not wholly reflect the leaching curve to be expected in an undrained soil with a high watertable. It is unlikely that leaching to 30 cm would be affected, but the lag-effect observed in the rate of leaching of lower horizons would be exaggerated in the undrained soil. This effect is shown by the dashed lines for the 60 and 120 cm leaching curves. (Error in the shape of the lower-profile curves has no influence on the argument used here). In Fig. 11 the data are redrawn to show the effect on the salinity profile of successive irrigations of the current schedule, beginning with the preirrigation of 330mm gross (230 mm net) and light applications of 120 mm gross (90 mm net). About 20% of salts remain in the ploughlayer after the preirrigation and at planting. Cotton is known to be most sensitive to salinity during germination. The relationship was investigated at the Institute of Agricultural Research Melka Warer and the results are shown in Annex E, Fig.14. There is serious suppression of germination if the topsoil ECe is greater than 10 mScm . The maximum acceptable topsop salinity before the preplanting irrigation therefore is 50 mScm The number of years before this critical threshold is reached is calculated by subtracting the increase in topsoil salinity during the fallow, AECe, from 50 and dlyiding the result by the annual incremental increase, 3.15 mScm . For an average soil with the watertable at 75 cm, the calculation for different watertable salinity classes is as follows: EC „ class 0-4 4-8 8-12 12-16 16-20 20-32 32-48 Mean EC 2 6 10 14 18 25 36 AECe Wt 6 18 28 42 55 75 115 Years 14 10 7 3 0 0 0 26The number of years until the critical salinity is reached and germination is seriously affected ranges from l^where the groundwater is of low salinity (a mean of^mScm ), to 3 years where salinity is moderate (mean 14 mScm conductivity is greater than about 16 mScm then the land is about to be or is already abandoned. To illustrate that the effect on germination is more serious than that on the growing crop, in Table 5 the salinity profile of a soil on the threshold of abandonment is calculated after each successive irrigation. The Salinity Index (Annex Ejection 2.3) begins very high but quickly falls to a level at which it has little effect on yield (Annex E, Fig. 15). If the groundwater These data are used in Annex E Section 6 to predict the decline in cotton yields with time, if nothing is done to overcome the problem of rising saline groundwater. 3.7.2 Bananas Analyses of irrigation practice in the past three years has shown that the water applied was less than consumptive use in dry months and more during the rainy season (Annex C). At present, the major part of the Melka Sadi Agricultural Estate is not being irrigated so that the natural regrowth of grass and shrubs is transpiring from the watertable with an Et/Eo coefficient (Kc) perhaps between 0.8 and 1.0, In effect, this more than doubles the evaporative loss from the irrigated banana and creates a marked hydraulic gradient and lateral flow away from them. The result of these two features is that the presently productive area appears to be in an equilibrium of both depth and salinity of groundwater. Banana plantations decline quickly on account of periodic water-stress, poor hygiene and other factors, and a watertable, the instantaneous rise of which after irrigation is too high into the root zone. The majority of the plantation on the point of abandonment is non-saline, indicating that other factors are responsible. It seems unlikely that the situation will change, or Indeed could change. Improvement in the irrigation schedule and the planting of a substantial portion of the abandoned land would lead to a marked rise in watertable which would hasten the decline in the plantations. In the majority of the area, the leaching factor in the irrigation schedule Is sufficient to minimise secondary salinity. 3.7.3 Other Crops The accumulation in salinity under maize and pasture as proposed for the Awash Settlement Farm would occur at much the same rate as under cotton. 27TABLE 5 REDUCTION IN SOIL SALINITY WITH SUCCESSIVE IRRIGATIONS IN A SOIL ON THE THRESHOLD OF ABANDONMENT FOR COTTON Depth below surface (cm) AECe . (mScm j Pre I 1/ ECe II 1/ ECe 12 ECe 13 ECe 14 ECe 15 ECe 0-20 50.5 0.20 10.1 0.18 1.82 0. 14 0.272/>- 25 18.4 0.21 3.86 0.19 0.73 0.15 0.27 ► 50 15.0 0.38 5.70 0.30 1.71 0.24 0.41 0.20 0.27 ► 75 14.5 0.72 10.44 0.63 6.58 0.53 3.49 0.43 1.50 0.34 0.51 0.24 0.27 100 14.03^ ► s?' 13.2 3.44 1.16 0.63 0.56 0.53 0.52 1/ Proportion of salinity remaining after preirrigation, post plant irrigation 1, 2, 3 etc. 2/ Minimum ECe produced by irrigating with water of EC 0.35mScm~ . 3/ Assumed to be EC of groundwater. 4/ Salinity Index - see Annex E3.8 Control of Secondary Salinisation by Subsurface Field Drainage-Principles In this section, the principles by which secondary salinisation can be controlled are outlined briefly, in order to provide a basis for subsequent sections concerned with leaching requirements. Under conditions which prevail at Amibara, a lengthy fallow period exists, extending from November until April. Once subsurface field drainage has been installed, the watertable can be expected to fall gradually following the final cropping season irrigation, and for much of the fallow period will reside close to, or perhaps a little below drain depth, in the absence of rainfall and lateral seepage. During the early part of the fallow, surface evaporation will result in the upward movement of residual soil moisture, until a position is reached where soil moisture tensions are such that further supply of water to the soil surface must be derived from groundwater. Where the groundwater is saline, salts will then be transported upwards, to be deposited either at the soil surface or within the profile itself, should the evaporation front retreat below the surface. A theoretical model has been developed describing this process of secondary salinisation (Appendix C). At the beginning of the following cropping season, leaching applications sufficient to reduce salt contents in the uppermost part of the profile, to a level suitable for seed germination, must be supplied. Subsequent irrigation applications must be designed to continue the leaching process, ensuring that as plant roots develop, an increasing depth of soil is leached to a salt content at which there will be minimal effect on the growing crop. Furthermore, on a long-term basis, there must be no net accumulation of salts in the soil profile. In designing the subsurface field drainage system, one of the objectives is to ensure that, within the range of site conditions likely to be encountered, drain depth is chosen such chat secondary salinisation during the fallow period is reduced to an acceptable level. The capacity of the field drainage system must be such as to remove excess water in the required time, such as to prevent a rise in watercable to levels at which there may be damaging effects on the growing crop. 3.9 Reclamation Leaching After Drainage Once subsurface field drainage has been installed, areas which by that time will be abandoned because of salinity may be reclaimed. In these areas, soils can be expected to exhibit salinity characteristics similar to those observed in abandoned areas of the Melka Sadi Pilot Drainage Scheme site prior to leaching. A feature of this site was the marked spatial variations in salinity, and this is typical of complex alluvial soil associations affected by secondary salinisation. Because of Chis, and because further salt accumulation can be expected prior to the installation of drainage, it is not possible to specify with accuracy the leaching applications required for the reclamation of specific fields in advance. 29An indication of the total depth of leaching water which will be required to achieve a given percentage removal of salts can however, be obtained trow the results of leaching trials conducted during the Pilot Drainage Scheme programme. Before Leaching, the salinity profiles of these soils were typically those of salinised lands, showing marked concentrations of salinity in the upper profile. Reference to Figure 10, which is based on the mean results obtained from two leaching basins, shows that BOZ of the initial salt content of a soil horizon at 60 cms depth can be leached by applying a cumulative depth of leaching water of 50 cms* For a soil horizon at 120 cms depth, a cumulative depth of leaching water of B5 cms would be required to achieve an equivalent reduction in salt content* When this cumulative depth of leaching water is applied, the salt content of the soil horizons above is leached to a considerably lower value, as is apparent from the shape of the curves,. Soil sampling in the moat saline part of the Pilot Drainage Scheme site prior to Leaching^showed that the salinity aC|a depth of 120 cms was typically IZmScm and in excess of 100 mScm at the surface. Reference to Figure 10 shows Chat to render this profile non-salin^ at a depth of 120 cms (EC after leaching equal to or less than 4mScm ), by removing approximately 70Z of the salts, would require a cumulative depth of leaching water of approximately 70 cms* This represents the depth of water which must pass through the soil profile; a greater depth roust be applied to the field to allow for evaporative losses during the reclamation period. Assuming an evaporation rate of I Ma per day, which is a conservative estimate, arid a terminal infiltration rate of 1 cm per day (comparable with findings during leaching trials), reclamation would require 70 days in this case, and would require the application of 126 cms of water (made up of 70 cms for leaching purposes and 56 erne evaporative loss during the period)- Reclamation leaching should be carried out in drained fields which have been subsoiled and land*planed (to remove high and low areas) prior to leaching and subdivided by bunds Co form a series of conveniently sized level basins. The undesirable effects of preferential flow into the drains from above, with consequent reduction in Leaching, can be reduced by positioning bunds coincident with drain lines. In fields 250 metres long with a drain spacing of 60 metres, a convenient size of leaching basin would be 3000 m , (60m x 50m) enabling 5 basLns to be accommodated within the field length- The choice of bund size must reflect a compromise between the type of machinery available for construction, the time available for leaching and management factors. A typical height of bund could be 30 cms, which would allow a depth of say 15 eras water to be ponded. 2 To leach continuously, water would need to be replenished at approximately 8 day intervals (assuming a terminal infiltration rate of 1 cm per day and evaporative losses of 8mm per day). However, continuous leaching is not a prerequisite provided secondary salinlsatlon is prevented by avoiding excessive intervals between Leaching applications. The leaching application for reclamation, together with evaporation losses, imposes a duty at field canal level of some 2 lit/sec/ha. 303 . 10 This in approximately double the normal irrigation duty and as such could overload the distributary canal systems were such reclamation co be extensive. However the total area requiring reclamation Is not expected to exceed lOOOha/year, which with continuous reclamation throughout the year, wouLd amount to some 200-25Qha maximum at any one time. This represents a small proportion of the overall irrigated area and would no pose any problems on main canal supplies, Similarly at field canal level Che normally adopted flow scream of 150 ilt/sec for a 25ba field would permit each leaching irrigation to be carried out in approximately 3 days. Ar secondary and tertiary canal level it may prove necessary to schedule the reclamation leaching to avoid excessive overload of any one canal system- Such scheduling would not in itself cause a reduction in the overall programme of reclamation. An ouline reclamation procedure for a selected sample area is given in Appendix D. Soil amendments, which are not expected to be required during reclamation, are discussed in Section 3.11. Routine Leaching After Drainage The reclamation of already saline soils at Amibara is has been discussed in the previous Section. There will be a steady increase in soil salinity during a fallow period, in all but the coarsest - textured soils, even after the installation of drains at 1.9®. This is due to secondary salinlsation, a process in which salts move from the watercable towards the surface in the mass flow of water for evaporation. Irrigation reverses this process by leaching the salt downwards. Furthermorei the irrigation water, although of low salinity, has a salt load which must be leached into the drain if a steady accumulation of salts in the root zone is to be avoided. The leaching requirement is variable, on account of crop tolerance to salinity. The length of the fallow period is dependant on crop type, and leaching efficiency and the amount of secondary salinity are a function of soil texture. 3.10.1 Cot con A revised cotton irrigation schedule is proposed in Annex E. In the early part of the season cotton requires a small amount of water applied fairly frequently because it lacks a well- developed root system to exploit the soil storage capacity. This theoretical requirement is somewhat leas than the design capacity of the canals. The balance represents the opportunity to apply an amount of water expressly for the purpose of leaching. During the six month fallow period, the increase in salinity in the surface 20cm of soil, AECe. is calculated to vary from some 27 rnScm in a fine-textured vertisol or compact silty clay over a very saline watertable of EC 50 mScm , to almost nothing in a coarse-textured alluvium over a non-saline watertable. This is illustrated in Fig.12, which is derived from the model of secondary salinisation discussed in Appendix C of this Annex.The reason for the difference due to soil texture Is that the maximum capillary rise in a coarse-textured soil is 160-lfl0cm, compared with 200 cm in a vertisol or compact silty clay. With drains at 1.8m and the watertable falling steadily, very little of the upward mass flow of saline groundwater reaches Che surface 20cm of coarse-textured soil to cause secondary salinisation. Constraints to Che amount of leaching which is possible are the design capacities of the irrigation and drainage systems. The former currently operates on a maximum single gross application of 286 turn before planting and about UOnnn during the longer intervals after planting. The average drain capacity is equivalent to 2.5 mm per day. Failure to leach successfully is most likely to occur in a fine-textured soil overlying very saline groundwater. To identify the limit of successful leaching the above constraints are used to calculate the maximum amount of leaching water which can be applied without the watertable rising so much above drain level that the crop is affected. Assuming (at present) that there is no constraint of river flow in April/May, two preplant applications of 286 mm (gross) would be possible on a limited area, the first to recharge the profile aa is currently done, and the second to ensure satisfactory leaching of the surface and germination. The first three post- plant applications could be amended from the present 130 mm gross to 150, 130 and 130 mm respectively, the the interval also being reduced. Using the soil characteristics given in Table 6 and the calculated evaporation from soil and crop (Annex E) the groundwater level changes as is shown in Fig. 13. The two preplanting applications leach salts from the upper profile but the wetting front fails to reach the watertable, which has fallen well below drain level during the fallow. The first poscplant irrigation completes the wetting and raises the watercable above drain level, to which it recedes rapidly, flushing most of the accumulated salts. The second and third post plant irrigations cause a substantial 'instantaneous' rise in groundwater but at that stage roots have not reached their maximum depth. Later irrigations using the revised schedule cause temporary saturation of the lower root zone but according to the Root Index model (Annex E) the effect on yield would be Insignificant. TABLE 6 SOIL TYPE Saturation Point rle]d Capacity Permanent wilting Point SOIL MOISTURE CHARACTERISTICS OF TYPE PROFILES VERTISOL X Moisture by volume 67.9 60.9 33.2 Z Porosity 7.0 27.7 ALLUVIUM I Moisture X Porosity by volume 48.0 39.0 9.0 20.7 18.2 Bulk Density (g/cm ) 3 Source: Water Management Modified by data 1.33 Manual, Sir William Halcrou and Partners from Pilot Drainage Area, Melka Sadi. 1.35 (1983), 32By contrast the amount of secondary salinisatlon on the coarser textured alluvium is quite slight and such heavy leaching applications are not required. The lower porosity between PWP and FC (Table 6) would also induce unacceptable saturation In the rootzone under heavy applications. A single pre-plant Irrigation only is proposed. The effect of such a schedule is shown in Fig. 14, The 'instantaneous rise in vatertable 1 only marginally intrudes into the lower root zone and flushing of salt only commences with the third post-planting irrigation. The water-balance for these soil types with their respective irrigation regime is detailed in Table 7. Net application is 70Z of gross to allow for Irregularity of water distribution and differential hydraulic conductivity between soils within the group. Small areas of very permeable soils have been observed which act as 'sinks co the watertable, transmitting more water than the least permeable areas by a large factor. Consumptive use is that estimated in the irrigation schedule developed for an average soil (only a single schedule is being recommended). The excess for leaching is the water which is surplus to that required to recharge the profile to field capacity and which percolates through the profile beyond rooting depth. It is calculated from the net application, but that alone would assume that the 'gross' amount (chat between Gross and Net applications) is completely Ineffective in Leaching. Since this is not true, the Excess is increased by 102 and 402 of the 'gross' fraction in fine and medium/coarse soils respectively. These allowances are considered to be conservative. 1 For reclamation the leaching requirement is estimated from the equation: OW/DS - b. log (ECf/ECi) where DS is the depth of soil to be leached and DW is the depth of water required to leach it (in nm) _|Cf and ECI are the ECe values, final and initial respectively (in mScm ), and b is a coefficient dependent on the soil leachability. In fine, compact soils the value of b is likely to be as low as -2.2. In the leaching basins of the Pilot Drainage Area at Melka Sadi, the average value of b was -1.1 for a medium textured soil and in a coarse soil the value of b could be greater than -0.5. Using this equation and the appropriate values, ECf is calculated as shown in Table 8 for different soils and salinity of groundwater and for the two Leaching regimes above. The table shows the approximate yield loss which may be expected if the leaching efficiency is as predicted. Within the Stage 1 drainage project described in the Main Report, the total area may be subdivided on soil texture into drainability classes: undrainable, fine and coarse. The 'undrainable' class conforms closely to case 1 in Table 8. This land within the project area (there is more which will feature in subsequent drainage plans) is restricted to the NW corner of Algeta farm, and is the subject of proposals for further investigation. It is likely that the present irrigation and future drainage systems may need to be modified to be satisfactory. t 33The ’fine* drainability class is mixed in its Teachability. and is likely to range from case 3 to case 5 on account of the range in texture, compaction and salinity of the groundwater. Land which is fine/medium in texture with a leaching coefficient (b) of about -1.5 is unleachable by the heavy leaching regime where the groundwater is more saline than 20 mScm . and as such behayes like the very fine textured soils more saline than 12 mSc^i (case 1). This texture and groundwater between 16 and 20 mScm would require the heavy leaching regime, but the moderateregime would suffice where the groundwater is less chan 16 mScm The medium/coarse textured soils classified in the ’fine’ drainability class (case 5) and the land classified as 'coarse* are all adequately leached by the moderate leaching regime. The approximate areas within these classes are shown In Table 9 for the net cotton producing area of Melka Sadi and the whole of the Stage 1 area at Melka Warer. By leaching requirement, they are summarised as follows: Leaching Requirement No of Approximate Net Area (ha) preirrlgns Melka Sadi Melka Warez Redesigned systems 150 90 Heavy leaching 2 50 360 Moderate leaching 1 1340 1560 The model used to predict rhe accumulation of salinity in undrained soil is based on the annual contribution of salts in the irrigation water (Section 3,7). In a Largely undrained system, this annual accumulation alon^ produces an annual Increase in topsoil salinity of 3.2 m£cm at the end of the fallow. Irrigation subsequently leaches these salts into the upper layer of the groundwater, which is likely to become markedly stratified in salt concentration. After several years of irrigation the surface layer of the groundwater will become very much more saline than the watey at depth. Surface groundwater salinity of more than 50 tdSctd has been observed in auger holes which confirms the model. Where the system is not closed and is draining more rapidly through aquifers then this accumulation does not occur, which explains the considerable variation encountered in groundwater salinity. After the Installation o£ drains and several years of satisfactory leaching the salinity of the groundwater surface layer will fall to an equilibrium level. This equilibrium is most likely to be below 10 mScm , which Implies that all the soils of the project area will be satisfactorily leached by the moderate regime with a single preplant irrigation. 343.10.2 Bananas A revised Irrigation schedule for bananas is proposed in Annex E With a gross annual application of nearly 3000 mtn* the ’gross1 fraction of 30Z Is 900mm- Using the same effectiveness factors as for cotton, then 90nun in fine textured and 3 6 Oran in medium textured soils would leach the profile. Since the intervals range fron 8 to 15 days, secondary salinisatlon would be minIm1. In medium and coarse-textured soils no additional leaching requirement would be necessary. In very fine and compact soils a periodic leaching application may be necessary according to the following schedule: Month November January March May Gross (mm) 126 139 127 134 Net (hud) as 97 8? 94 Consumptive use (mm) 64 73 65 70 Excess (mm) 28 28 28 28 The drainage system is without the watertable designed for this amount of rising into the root zone. discharge Only small areas of this soil type exist in the banana estate and a very detailed survey would be necessary in order to locate them. 3.10.3 Ocher Crops It is proposed to increase maize production on the Relief and Rehabilitation Commission (RRC) Settlement Scheme. In order to justify investment in irrigation and drainage for this crop Lt will be necessary to produce two crops on the same land each year. Vegetables likewise should be produced continuously except during the two to three months of canal closure. The relatively high and frequent irrigation requirement of these crops should ensure chat secondary salination is □lnlnal. On finer-textured soils it will be necessary to Leach effectively by a preirrlgation before replanting after the short fallow period. An amount of 286mm, similar to that for cotton, would be adequate. Pasture on fine-textured soils could suffer from secondary salinisatlon from a watercable at L.8m. However, settlers adhere to an adequate irrigation regime. A convenient Irrigation schedule for pasture is presented in Annex- E. An annual leaching supplement is recommended, in which 286 mm is applied at the first after the canal is reopened on completion of maintenance. 35TABLE 1 WATER BALANCE (in mm} FOR TWO SOU. GROUPS WITH MAXIMUM LEACHING Preplanting Post Planting Irrigations Irrigation Number 1 2 1 2 3 4 5 6 Total Interval (days) 20 10 2D 20 40 30 30 Fine - Textured Soil Cross 330 330 150 130 130 260 185 185 1700 Net 23) 231 105 91 91 182 no 130 1191 Consumptive use - 60 30 60 80 185 130 125 670 Excess 241 181 80 35 15 5 6 11 574 Medium and Coarse - Textured Soil Gross — 330 150 130 130 260 185 185 1370 Net ■— 231 105 91 91 182 130 130 960 Consumptive use - 60 30 60 80 185 130 125 670 PExc!/ ess 211 93 47 27 28 22 27 455 1/ See text for calculationTABLE 8 EFFECT OF LEACHING REGIME ON DIFFERENT SOILS WITH DIFFERENT SALINITY OF GROUNDWATER, AND CORRESPONDING EXPECTED YIELD LOSS Soil DW Case Type 1 / Salinity of Groundwater (ECu In aS cm S 5 10 15 20 25 30 40 50 I VERT b—2.2 546 Heavy AECe=ECi ECf SI 2 yieLd loss 2/ 2.5 3/1.5 4/ 0.7 5/ 2 5.0 3.0 1.4 7 7.5 4.5 2.1 15 10,0 6.1 2.9 26 12.6 7.6 3.6 34 15.2 9.2 4.4 43 21*0 12.7 6.0 62 27.0 16.4 7.8 84 2 VERT 430 AECe-ECi 2.5 5.0 7.5' 10,0 12.6 18.2 21-0 27.0 b»-2.2 Nod ECf 1.7 3.4 5.0 6.7 8.5 10.2 14.1 18.2 SI 0.8 1.6 2,4 3.2 4.0 4.9 6.7 8.7 2 yield loss 2 9 18 29 38 49 70 95 3 ALVM 546 AECe-ECi 6/ 1.9 3.8 5.6 7.5 9.5 11.4 15.8 20.3 b’-1.5 Heavy ECf 0.9 1.8 2.7 3.6 4*6 5.5 7.6 9.7 SI 0,4 0.9 1.3 1.7 2.2 2.6 3.6 4.6 X yield loss 1 3 6 11 16 22 33 45 4 ALVM 430 AECe-ECi 6/ 1.9 3.8 5.6 7.5 9.5 11.4 15*8 20.3 b—1.5 Hod ECf 1.1 2.1 3.1 4.2 5.3 6.4 8.8 11.3 SI 0*5 1.0 1.5 2.0 2*5 3.0 4.2 5.4 X yield loss 1 4 8 14 20 26 40 55 5 ALVM 430 AECe-ECl 7/ 0.6 1*3 1.9 2.5 3.2 3.8 5.3 6.8 b=-l.l Mod ECf 0.3 0.6 0.9 l.l 1.4 1.7 2*4 3*1 SI 0. 1 0.2 0.3 0.4 0.5 0.7 0.9 1.2 X yield loss 0 0 0 1 1 2 3 5 1/ DW is the depth of leaching water in nan 2/ See text* AECe is expected secondary salinity during fallow and assumed equal to initial salinity in leaching equation, 3; ECf calculated assmaing DS to be rhe_coverage rooting depth, 1200™, 4/ SI is the Salinity Index (Annex E), EC e - 2.1xSI In vertisols and 2.6xSI in alluvium (approx*), where ECe is the weighted mean ECe in the profile. 5/ Relationship of yield with SI in Annex E* 6/ Less fine and compact soli than vertisol so 25Z less AECe. 7/ Less fine and conpact soil chan cases 3 and 4 so 751 less AECe than vertisol. 37TABLE 9 approximate net cotton land areas within drainability, textural and groundwater salinity classes Drainability Class Soil Texture Leaching Coefficient (b) Conductivity of Groundwater (mScm ) Approximate Land Area (ha) Melka Sadi Melka Water Leaching Regime Type mm water Required No preirrgns Undralnable Fine -2.2 >12 0 50 Redesigned system <12 0 340 Heavy 546 2 Fine Fine/nedlum -1.5 >20 ISO 50 AO 20 Redesigned system 20-16 <16 490 670 Heavy Moderate 546 430 L 1 Fine Medjum/coarse -I. J all 680 720 Moderate 430 1 Coarse Coarse -0.5 all 170 170 Moderate 430 1 Tota 1 1540 20103.11 Requirements for Soil Amelioration Soil investigations in the Melka Sadi - Amibara area have shown that the soils of the project area contain varlble amounts of calcium carbonate and gypsum. (Ref. 9,13). As part of the Melka Sadi Pilot Drainage Scheme programme (Ref. 16), leaching trials were conducted on saline - alkali alluvial soils to Investigate the feasibility of reclaiming these soils without the use of amendments. The results have demonstrated that saline - alkali soils similar to those in the trial area can be successfully leached, without the use of amendments. The field trials were supported by comprehensive laboratory analyses of soil samples, representative of leaching basin sites, prior to and on completion of leaching. These showed that both gypsum and calcium carbonate were present In variable amounts in all samples analysed. Calcium carbonate percentages varied generally in the range 1.5 to 3.5Z, and gypsum percentages were more variable, ranging from 0.07 to 2.69%. The latter constituent is primarily responsible for exchange reactions with adsorbed sodium under leaching conditions, calcium carbonate being only slowly soluble in water. By making assumptions concerning the soils initial exchangeable sodium percentage (ESP) and final value required, it is possible to tentatively estimate the percentage of gypsum in the soil below which additional amendments are likely to be required. For calculation purposes, the following data have been assumed based on experience gained from the leaching trials outlined above: Initial soil ESF - 40Z Final soil ESP required - 10Z (i.e. non-alkali conditions) Cation exchange capacity ■ 37 meq/LOO gm Amount of Na X (adsorbed Na + Ion) to be replaced by gypsum a 11,1 meq/100 gin. Assuming non-quantitalive replacement (the reaction between an amendment such as gypsum and exchangeable sodium is an equilibrium reaction and therefore does not go entirely to completion) the Indicative gypsum percentage in the soil below which additional amendments may be required is some 1.1Z. Over a long period of time, however, the slowly soluble calcium carbonate may also cake part in exchange reactions. This calculation reflects the serious soil conditions at the leaching trial sites resulting from prolonged abandonment, and as such they can be regarded as extreme conditions. In previous sections the spatial variability of soil salinity has been emphasised, and localised areas may exist where in situ reserves of calcium carbonate and gypsum could prove to be insufficient for exchange purposes during reclamation leaching. The addition of these amendments would be required for full reclamation of such affected areas. pj p f - >^JL XL- 3? il <40REFERENCES 1. Amibara Irrigation Project. Vater Management Manual. Sir William Halcrow and Partners, July 1903. 2. Additional Geomorphlc, Geological and Groundwater Studies in the Awash Valley. Informal Technical Report No 14, FAO, May 1983* 3. Angelele * Bolnamo Feasibility Study Report. Annex 1 - Hydrology. Sir William Halcrow and Partners, October 1975. 4. Melka Sadi - Amibara Proposed Irrigation Project. Feasibility Study* Part II Studies and Surveys, Vol 3 - Geology and Hydrogeology. Italconsult, Rome, July 1969. 5. Amibara Irrigation Project II. Drainage and Salinity Study and Recommendations for Field Drainage. Vol 1 — Study and Recommendations. Sir William Halcrow and Partners, April 1982. 6. Revised Proposal for Master Drainage Plan for Melka Sadi and Amibara Areas and Agreement for Consultancy Services. Sir William Halcrow and Partners. April 1984* 7. Salinity Seminar Baghdad. Irrigation and Drainage Paper No 7t FAO, Rome 1971, 8. Report on Survey of the Awash River Basin, Vol II. United Nations Special Fund, FAO 1965. A 9. Melka Sadi - Amibara Proposed Irrigation Project, Feasibility Study, Part II Studies and Surveys. Vol 2 - Soil Science - Italconsult, Rome, July 1969. 10. Melka Sadi - Amibara Irrigation Project. Additional Soil Study. Italconsult> Rome, September 1971, 11. Angelele and Bolhamo Feasibility Study Report, Annex II - Soils and Land Classification, Sir William Halcrow and Partners, October 1975. 12. Amibara Irrigation Project. Hydropedological Studies, Sir William Halcrow and Partners, October 1975. 13* Amibara Irrigation Project II. Drainage and Salinity Study and Recommendations for Field Drainage. Vol 1 * Study and Recommendations. Sir William Halcrow and Partners, April 1982. 14. Amibara Irrigation Project II. Melka Sadi Pilot Drainage Scheme Revised Working Paper. Sir William Halcrow and Partners, September 1983* 15. Diagnosis and Improvement of Saline and Alkali Soils* Agriculture Handbook No 60. U.S,D.A. 4116* Amlbara Irrigation Project IX. Melka Sadi Pilot Drainage Scheme. Draft Final Report. Vol I. Analysis and Conclusions. Vol 2: Environmental Background and Experimental Design* Sir William Halcrow and Partners. October 1984. 17. Angelele - Bolhamo and Amlbara Irrigation Expansion Project. Reappraisal and updating of previous feasibility studies. Final Report, Vol II. NEDECO, June 1982. 18. Marshal T.J. and Holmes J,W. Soil Physics. Cambridge University Press,1979. 19. Smedema L.K.t Rycroft D.W. Land Drainage. Batsford Academic Educational Limited, London, 1983. 42APPENDIX A OBSERVATION WELL HYDROGRAPES Figures Al - A15FIGURE Al MASTER DRAINAGE PLAN FOR MELKA SADI ANO AMI0ARA AREAS OBSERVATION WELL HYDROGRAPHS SIR WILLIAM HALCROW a PARTNERS JULY I9S5I K I I FIGURE HZ MASTER DRAINAGE PLAN FOR MELKA SADI ANO AMI6ARA AREAS OBSERVATION WELL HYDROGRAPHS AIP2S r AJP2& i r n Al P 27 K Aipze A 11 11 I t I I I I AtPS AiP? Al P 23 SIR WILlI4M UALCROW ft PlRTMCRs JULY 1965FIGURE A3 MASTER DRAINAGE PLAN FOR MELKA SAOI ANO AMI0ARA AREAS AIP12 0 2 4 6 a io 12 AlP 13 AIP18 SIR WILLIAM HALCROW ft PARTNERSII DDEPTH TO GROUND WATER - metres FIGURE A4 MASTER DRAINAGE PLAN FOR MELKA SAOI ANO AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS. AIP19 AIP20 SIR WILLIAM HALCROW 3 PARTNERS JULY 1985FIGURE A5 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS AIP15 AlP 16 SIR WILLIAM HALCROW 3 PARTNERS JULY 1985ACtPIH TQ GROUNL WATER - metres FIGURE A6 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS AIP 22 T AIP 38 SIR WILLIAM HALCROW 6 PARTTCRS JULY 1985FIGURE A7 MASTER ORAINAGE PLAN FOR MELKA SAOI AND AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS 0 2 4 6 AIP6) AIP32 AlP 36 AIP 41 AlP 49 SIR WILLIAM HALCROW B PARTNERS JULY 1995FIGURE A8 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS -U31VM ~UNnO«9 01 Hid 3d SIR WILLIAM HALCROW a PARTNERS JULY 1985■ ■J J rFIGURE A9 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS - saj U31VW llNnOHO 01“Rld3'0 SIR WILLIAM HALCROW 8 PARTNERS JULY '9853 ■FIGURE AIO MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS <D 2? cO 2? — r-> <D 3 in <d 2 >r CD C“» n cn (D 1 < • M <D CD 0 2' J 41 CD s 1\ I i \ “1 g c ■X 2l o <D cd 1 CD CT CD- 2 ♦ t r* 2? — u> r- a> \ I 1 I 4 i i • •i i ”1 <JD F 2? co Q- g 2 i o- 2k- I <4 > 1 < * % [ 1 O- I 2 LT> 2? ! ______ 1 1 a ) s 1 1 • • 1 pj cd r*. 2 r m r- CD L......_ 1 I 1 i 1 j 1 f m 2 -J r* CM 2 •— rv 2 I 1 1 2 -4 c <o 2 £ ci kujiauj-usivM awnoHO oi maau SIR WILLIAM HALCROW ft PARTNERS JULY 1985FIGURE All MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS sir william halcrow a partners JULY 1985FIGURE AI2 MASTER DRAINAGE PLAN FOR MELKA SADI ANO AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS AIP 47 16 18- AIP48 AIP50 AIP51 AIP52 SIR WILLIAM HALCROW 9 PARTNERS JULY 1965FIGURE AI3 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS 0 2 4 6- 8 10 12- AIP 35 3 SIR WILLIAM HALCROW a PARTNERS JULY 1985DEPTH TQ GROUND WATER metre FIGURE AI4 MASTER DRAINAGE PLAN FOR MELKA SAOI ANO AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS g 8- 10 12- AIP53 ’4 I6i IfrL ’ T T AIP54 AIPS5 AIP56 AIP57 SIR william HALCROW a PARTNERS JULY 1985r I1FIGURE AIS MASTER DRAINAGE PLAN FOR MELKA SAOI ANO AMIBARA AREAS OBSERVATION WELL HYDROGRAPHS SIR WILLIAM HALCROW & PARTNERS JULY 1985J II II ■I ■I ■I ■ ■APPENDIX B RATES OF RISE OF GROUNDWATER AND PREDICTED GROUNDWATER DEPTH vs TIME RELATIONSHIPSTABLE Bl RATES OF RISE OF GROUNDWATER AND OBSERVATION WELL DEPTH VS TIME RELATIONSHIPS - MELKA SADI AREA Obser Rate of Predicted vation Ri se Depth fm) Well No. (m/yr) Yr.O 1Z Yr.l Yr.2 Yr.3 Extrapolated Depths Below Ground Surface (m) Yr .4 Yr. 5 Yr. 6 Yr.7 Yr.8 Yr . R Yr. 10 Yr.11 Yr.12 Yr. 13 Yr. 14 AIP 21 0.65 9.55 8.9 3.25 7.6 6.95 6.3 5.65 5.0 4.35 3.7 3.05 2.4 1.75 1.1 0.8 16 0.55 7.7 7.15 6.6 6.05 5.5 4.95 4.4 3.85 3.3 2.75 2.2 1.65 1.1 0.8 - 1! 0.7 7.0 6.3 5.6 4.9 4.2 3.5 2.8 2.1 1.4 0.9 - - -- - 14 0.6 1.9 1.3 0.85 - ---- -- - ---- 17 1.1 3.1 2.0 1.3 1.0 - - - ------ - - 19 1.1 5.0 3.9 2.8 1.7 1.4 1.1 0.8 - ■ - - - --- 9 1.0 6.3 5.3 4.3 3.3 2.3 1.3 - - - - - -- -- 7 0.8 3.4 2.6 1 .8 1.25 0.95 -- - - -- -- - 20 1.6 4.4 2.8 1.85 1.55 1.25 - - - - -- --- - B 0.5 6.4 5.9 5.4 4.9 4.4 3.9 3.4 2.9 2.4 1.9 1 .4 1.0 - - - 5 0.6 4.8 4.2 3.6 3.0 2.4 1.8 1.2 0.9 - --- - -- P 11 0.2 - -- --------- -- - 8 0.3 - -- -------- -- - 3 0.3 -- --------- -- - 22 1.2 5.7 4.5 3.3 2.1 1.35 1.05 - - - -- - - - \J Yr.O refers to September 1984TABLE B2 RATES OF RISE OF GROUNDWATER and OBSERVATIQn_wElL DEPTH VS TIME RELATIONSHIPS - AMIBARA AREA Obser Rate of Predicted Extrapolated Depths Belov Ground Surface (m) vation Ri se Depth (n» Well No (m/yr) Yr.O U Yr.l Yr.2 Yr.3 Yr.4 Yr.5 Yr.6 Yr.7 Yr.8 Yr.9 Yr.10 Yr.11 Yr.12 Yr.13 Yr. 14 Yr. 15 Yr. 16 Yr.17 Yr.18 Yr.19 Yr. 20 37 0.6 7.4 6.8 6.2 5.6 5.0 4 .4 3.8 3.2 2.6 2.0 1.4 0.96 - -------- 42 0.85 9.9 9.05 8.2 7. 35 6.5 5 .65 4.8 3.95 3.1 2.25 1.4 0.96 - - - - ---- 46 0.8 11.8 11.0 10.2 9.4 8.6 7 .8 7.0 6.2 5.4 4.6 3.8 3.0 2.2 1.4 0.85 ---- - 0.45 12.5 12.05 11.6 11.15 J0.7 10 .25 9.8 9.35 8.9 8.45 8.0 7.55 7.1 6.65 6.2 5.7 5.3 4.85 4.4 3.95 3.5 - 50 0.25 11.4 11.15 10.9 10.65 10.4 10 15 9.9 9.65 9.4 9.15 8.9 8.65 8.40 8.15 7.9 7.65 7.4 7.15 6.9 6.65 6.4 51 0.35 13.1 12.75 12.4 12.05 11.7 11 . 35 11.0 10.65 10.3 9.95 9.6 9.25 8.9 8.55 8.2 7.85 7.5 7.15 6.8 6.45 6. 1 52 0.25 12.1 11.85 11.6 11.35 11.1 10. 85 10.6 10.35 10.1 9.85 9.6 9.35 9.1 8.85 8.6 8. 35 8.1 7.85 7.6 7.35 7.1 53 0.95 10.1 9.15 8.2 7.25 6.3 5. 35 4.4 3.45 2.5 1.55 0.93 - - -- -- -- 54 0.25 14.7 14.45 14.2 13.95 13.7 13. 45 13.2 12.95 12.7 12.45 12.2 11.95 11.7 11.45 11.2 10.95 10.7 10.45 10.2 9.95 9.7 55 1.0 3.5 2.5 1.5 0.85 - - - --- -- - ----- - 48 0.35 5.1 4.75 4.4 4.05 3.7 3. 35 3.0 2.65 2.3 1.95 1.6 1.25 0.93 - -- -- - - 57 0.7 12.4 11.7 11.0 10.3 9.6 8.5 8.2 7.5 6.8 6. 1 5.4 4.7 4.0 3.3 2 .6 1.9 1.2 0.8 - - - P 9 1.5 6.0 4.5 3.0 1.5 0.9 - - - - - - - - -- --- -- PP 5 0.22 7.5 7.28 7.06 6.84 6.62 6.4 6.18 5.96 5.74 5.52 5.3 5.08 4.86 4.64 4 42 4.2 3.98 3.76 3.54 3.32 3.1 AIP.38 0.65 2.75 2. 1 1.45 I.15 0.85 - - - - - - - - -- - --- - 39 0.3 2.0 1.7 1.4 1.1 0.8 - - - - - - - - -- - - -- - 45 0.4 2.4 2.0 1.6 1.3 1.0 - - - - - - - - --- --- - 32 0.5 2.5 2.0 1.5 1.2 0.9 - - - - - - - 1---- - 36 1.4 4.0 2.6 1.45 .15 0.85 - - -- - •- --- -- 41 1.1 3.1 2.0 1.3 .0 0.7 - - --- --- --- - 49 0.9 4.8 3.9 1.0 2 .1 1.3 1 .0 - - -- -- --- -- P 2 0.8 5.9 5.1 4 .3 3 .5 2.7 1 .9 1.1 - - - -- -- -- - P 3 1.5 6.1 4.6 3 .1 1 .6 1.3 1 .0 - - - -- -- - AIP 44 0.65 6.4 5.75 5 .10 4 45 3.80 3 .15 2.5 1.85 1 .2 0.81 - -- -- - 35 1.0 5.9 4.9 3 .9 2. 9 1.9 I 4 0.75 - - -- - -- 33 1.3 5.2 3.9 2 6 1. 3 1.0 0. 7 - - - -- - £/ Yr.O refers to September 1984.V 0 0 d u LJAPPENDIX C THEORETICAL MODEL OF SECONDARY SALINISATION IN COTTON FIELDSJappendix c THEORETICAL MODEL OF SECONDARY SALINISATION IN COTTON FILEDS The increase in salinity (AECe) in a soil horizon due to the upward movement of salts dissolved in an upward mass flow of soil water is simply calculated as follows: AEC - lOOme e where m and e are the amount of the mass flow of water from a watertable and its salinity (EC ) respectively, and s is the average saturation percentage by volume of the^horizon in question. Units are 1q cm and mS/cm. The mass flow term (m) is difficult to estimate, however, since according to Darcy’s Law it is a function of the pressure gradient and the hydraulic conductivity at the evaporative surface and they vary as a function of the depth of the watertable. Furthermore, the evaporative surface is not neatly at soil surface but occurs in voids and fixtures deep into the profile and, when a transpiring crop is present, through the root system. Lastly, the pressure gradient (g) is Influenced by the vapour pressure of the soil solution, which is reduced by increasing salt concentration. The availability of terms in the Darcy equation relevant to the situation at Amibara is limited. The most useful data which are available are ETo and Eo values (Ref. 1) and a generalised relationship between rate of evaporation in relation to depth of watertable (Ref. 18). The most satisfactory approach then is to estimate mass flow as a varying function of Eo. Since the presence of a cotton crop has a significant influence on mass flow of water in the profile it is assumed that the last irrigation nominally is on 10 October and that the crop is destroyed at the end of December. The ’fallow is considered to begin on 1 November and end 180 days later, at the end of April, with the crop being present for the first 60 days. 1 The variation in Eo through the fallow period is shown in Fig.Cl. Evaporation from the fallow system will be related to it through a coefficient, K. Since some of the evaporation will be through the plant and the remainder from the soil then there are two coefficients Kp and Ks respectively. The coefficient is the normal crop coefficient Kc until crop eradication but its plant component gradually falls relative to its soil component as the crop matures. The microclimate beneath the leaf canopy, which normally suppresses evaporation from the soil surface, changes and Ks gradually Increases as Kp declines. As the watercable falls the hydraulic conductivity also falls. As the soil surface dries out the pressure gradient increases but proportionately less, so chat through Darcy’s Law mass flow and Et are reduced by a falling watertable. When the soil is wet, evaporation is known to equal that from an open water surface (Ks *1) but when the watertable is below about 50 cm the soil surface dries and Ks falls. The effect is shown in Fig. 3 for extremes of soil texture encountered at Amibara. In moist soil Et and mass flow are equal but as the watertable falls diffusion of water vapour becomes an increasingly important component of evaporation, relative to the mass flow. Mass flow to the soil surface through capillarity is known to cease when the watertable is below 2m in a clay and silty clay and below 1.6m in a coarse textured soil (Ref. 19) hence Ks = 0 at these depths.TABLE Cl WATER LOSS DURING FALLOW FROM VERTISOLS AND ALLUVIAL SOILS, WITH DRAINS AT 1.0, 1.5 AND 2.0 METRES DEPTH. Inter Mean Io in Vertisol (P - 38X1 Mean (Eoltc) Ffvb Mean ( EaO) Froa val to tat'I Kt AV WT VTd r> •4 AM VT WTd^ (d) («/d) ( — ) (») (■■) (*) (as) (t.) Allxrviua (P - J1.5X) Froa Mean (Beta) Frxzn AM UT WTd U •4 AW WT) (w) WT<) □rein □eoth 1.0 n 1 20 1.75 175 0.58 102 30 7.80 234 0.35 144 15 7.13 107 0.53 64 IS 8.95 104 0.51 61 IS 5.95 104 IS 7.05 106 JO 7.40 222 JO 8.00 240 30 8.90 257 1 ‘Jeter lone (an) 102 0 1000 72 72 1189 24 40 1294 15 44 1413 • 102 0 1000 50 94 1298 12 52 1463 7 54 1634 0.37 46 11 35 1507 0.31 40 9 31 1389 0.26 73 11 62 1752 0.23 70 8 62 1915 0.20 68 4 64 2083 0.26 35 J 32 1736 0.21 30 2 28 1825 0.14 46 1 45 1968 0.11 41 0 41 2098 1 0.08 36 0 36 2212 668 559 Drain □eptr. 1.5 ■ 20 8.75 175 0.58 102 30 7.80 234 0.S5 164 IS 7.15 107 0.53 64 15 6.95 104 0.51 61 15 6.95 104 IS 7.03 106 30 7.40 222 30 8.00 240 30 1.90 267 Water loaa (a) 102 0 1300 94 so 1632 33 JI 1714 27 34 1803 102 0 1500 76 68 1716 25 39 I860 15 46 1986 0.26 37 24 13 1837 0.23 34 20 14 1874 0.21 62 33 29 1950 0.20 63 26 37 2047 0.19 66 20 46 2168 0.22 30 13 17 2040 0.18 27 9 18 2097 1 0.09 35 6 29 2189 0.07 32 9 23 2262 0.05 28 5 23 2335 633 (523) □Tain □•pth 2.0 a 20 9.75 175 0.58 102 30 7.80 234 0.55 144 IS 7.15 107 0.53 64 15 6.95 104 o.Sl 61 15 6.95 104 15 7.05 106 30 7.40 222 30 8.00 240 30 8.90 267 water lota (m) -< 102 0 2000 94 50 2132 33 31 2214 27 34 2303 102 0 2000 76 68 2216 25 39 2340 15 46 2486 0.25 34 26 to 2329 0.18 27 20 7 234 7 0.14 46 S3 13 2381 0.13 46 26 20 2434 0.12. 47 20 27 2505 0.16 24 13 11 2521 0.38 16 9 7 254) 3.01 17 6 11 2578 0 15 9 6 2397 0 15 5 10 2629 S71 (458)The conceptual model changes significantly with crop eradication. Until that point mass flow and deep drainage will be the only significant loss of water from the profile below 50 cm and little mass flow will reach the soil surface. Furthermore the origin of evaporated water mainly will be from that stored In the profile from the last Irrigation and therefore non-saline. After crop eradication the site of greatest water evaporation moves to the soil surface and the origin of the main flow shifts Increasingly to the watertable, the level of which is falling. The application of this model to a vertisol and an alluvium is shown in Table Cl. Although three drain depths are featured, their significance is that the watertable is assumed to be at drain depth at the start of the fallow. The model therefore applies equally to undrained soil. The fallow period is shown as divided into Intervals of 15 or 30 days, with 20 days for the watertable to recede to drain depth by the nominal start of the fallow (day 0) after the final irrigation. The mean daily Class A pan evaporation (Eo) is shown for each interval, and its product with number of days is the mean evaporation from an open water surface. The crop coefficient Kc is given for each interval until eradication, and it is assumed to be unaffected by soil type and drain depth within 80 days of the final irrigation. The evaporation by the crop during the interval is the product (EoKc) and an amount (d) for deep drainage is added. The value of d is assumed to be 0.5 nnn/day from observation well hydrographs, but there is considerable variation between wells. The root activity is mainly in surface horizons (see Annex E) so that the origin of water for evaporation prior to the start of the fallow is assumed to be wholly that stored in the profile after irrigation. The available water (AW) between field capacity (FC) and permanent wilting point (PWP) is taken to be 283 and 205 mm/m in vertisols and alluvium respectively (Ref.l). From this, curves were constructed (Fig. C2) to show the depletion of stored water from the profile with time. A maximum of 90Z depletion of stored moisture is assumed since the lower profile will remain above PWP and the vapour pressure of evaporating water will be reduced by salinity and maintain mass flow at a higher level. After eradication of the crop the coefficient becomes Ks with a value dependent upon the height of the watertable as shown in Fig. Cl. However, it is assumed that residual storage moisture will initially produce a higher Ks than if the origin of the water for evaporation were solely the watertable. This is illustrated in Fig.C. .4 where the variation in coefficient with time 19 shown for the extremes of a vertisol with the drain at 1.0m and an alluvium with the drain at 2.0m. The values of Kc are the same in both situations. The value of Ks is dependent only on the watertable depth and Is shown by the dashed lines Ks’. The effect on these of the suppression of evaporation from the soil surface by the crop canopy is shown by the dot-dashed lines Ks11. However, the influence of residual stored moisture on the coefficient is shown by the solid lines Ks. After crop eradication the water loss from the watertable is the product of (EoKs) plus the deep drainage loss (d). Since Ks is dependent on watertable depth the procedure is stepwise, the calculation of depth of wacertable for each period determining the value of Ks for the next. The change in depth of the watertable is the incremental waterloss divided by the porosity of the particular soil between saturation and PWP. Porosity (P) i 38Z for a vertisol and 31Z for an alluvium (Ref. 1). ?The total waterloss from each profile Is shown in Table Cl. The values for the deeper-drained alluvium are theoretical because mass flow to the top 20 cm of soil, which is important in secondary salinisation, ceases when the watertable falls below about 2m. The horizontal lines in the table indicate the limit of mass flow to the soil surface. Subtracting from these amended totals of water-loss the deep drainage loss and loss from stored water produced the following totals (mm): Drain depth These apparently are the 1.0m 1.5m 2.0m Vertisol 323 161 99 Alluvium 284 128 0 values of m in the equation for calculating AECe given above. There are two adjustment to be made however: firstly for the increase in mass flow due to vapour pressure reduction, and secondly for the difference in the EC of the storage water between the start and end of the fallow period. The influence of salinity on vapour pressure (VP) is described by Raoult s Law. r The reduction in VP is the mole fraction of the salt solution, S/(S+W), where S and W are the molar concentrations of salt and water. The mole fraction is a function of the moisture content since salt concentration in moles is E/W, where E is the ECe of the soil and W is the moisture content. Assumed values of W are shown in Fig. C5 together with the calculated values of percentage reduction in VP. The soil solution at the start of the fallow will be non-saline at the surface with an EC similar to that of the irrigation water. Its salinity increases with depth towards the watertable. Through the fallow period, this solution will move upwards to be replaced by generally more saline solution from the watertable. For the purpose of estimating depth of watertable after each interval, a distinction is made in the above calculation, between the components of mass flow orginating in stored water and watertable. However, a variable proportion of the initial low salinity water, the residual 10Z of available water (AW) plus that below PWP, will become Involved in the mass flow by replacement by higher salinity groundwater. The overall effect will be a function of the relative salinity, before and after, and a characteristic related to the leaching efficiency factor. The effect therefore will increase with drain or watertable depth and be smaller in fine than coarse-textured soils. The reduction of secondary salinisation by the differential in salinity of soil solution is somewhat greater than the increase by the effect on vapour pressure. The net result, expressed in terms of a percentage reduction in mass flow is as follows: Drain depth 1.0m 1.5m 2.0m Vertisol 0 19 37 Alluvium 15 36 N/A The adjusted values follows: Drain depth (m) Vertisol Alluvium of mass flow (m) and the corresponding estimated AECe are as Mass flow to surface (mm) Salinity rise AECe (mScm’S 1.0 1.5 2.0 1.0 1.5 2.0 323 130 62 110 44 21 241 82 0 126 43 0The relationship between the amount of secondary salinisation and groundwater salinity for different soils and drain (watertable) depths is shown graphically in Fig. C6.aEf/Eo Coefficient for Soil (Ks) FIGURE Cl MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS VARIATION IN THE EVAPORATION COEFFICIENT (Ef/E ) 0 FOR SOIL(Ks) WITH DEPTH OF WATERTABLE SIR WILLIAM HALCROW a PARTNERS JULY 1985Et/Eo Coefficient (Kc a Ks) Daily mean Eo(mm) Moisture avoitable from storage (mm) FIGURES C2,C3 a C4 MASTER DRAINAGE PLAN FOR MELKASADI AND AMI0ARA AREAS FIGURE C3: Variation in Daily Meon Eo ( Pon evaporation) from the start of Follow FIGURE C4: Variation in Crop (Kc) and Soil (Ks) Coefficients of the Et / Eo ratio .'1 J<c ^s' — Vertisol 1 Om 1 ---- Ks" Alluvium 2 1 ■■ 1 L I.' -30 0 30 60 90 120 150 180 Days from start of fallow (nominally 0 = 1 November) SIR WILLIAM HAUCROW a PARTNERS JULY 1985FIGURE C5 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMI8ARA AREAS THE EFFECT OF SOIL SALINITY AND SOIL MOISTURE CONTENT ON REDUCTION IN VAPOUR PRESSURE OF SOIL SOLUTION SIR WILLIAM HALCROW a PARTNERS JULY 1985FIGURE C6 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS RELATIONSHIP BETWEEN SALINITY OF THE WATER - TABLE ANO THE EXPECTED RISE IN TOPSOIL SALINITY DURING THE FALLOW PERIOD FOR STATED SOIL TYPE AND DRAIN (WATERTABLE) DEPTH AT START OF FALLOW SIR WILLIAM HALCROW8 PARTNERS JULY 1985APPENDIX D OUTLINE RECLAMATION PROCEDUREAPPENDIX D OUTLINE RECLAMATION PROCEDURE To Illustrate the procedure and resources required In reclaiming saline areas, a reclamation methodology has been developed for a typical group of ten fields in the Melka Sadi area which, although not yet abandoned, are currently becoming salinised. The fields chosen are F3/2/19 - 25, F3/3/33 - 35, and their location is shown on Figure DI, The approach is hypothetical in that it is assumed these fields have been abandoned, the EC at 120 cms depth being approximately 12nSaB (comparable to conditions at Melka Sadi Pilot Drainage Scheme site prior to drainage trials). It is also assumed that drainage has been installed prior to leaching. The objective of conducting a leaching programme is to reduce the salinity of the soil profile to levels which are more favourable for crop growth. Reducing salinity in the upper part of the soil profile is particularly Important. For calculation purposes, this objective has been expressed in quantitative terms as the achievement of a reduction in EC at 120 cms depth in the soil profile to AinScm ; for cotton, which is a salt-tolerant crop, this is an exacting requirement. As explained in Section 3.9, a total depth of water of some 125 cms must be applied to achieve this, Including losses by evaporation. The essential stages of the procedure which is equally applicable to future areas which may become salinised prior to the installation of drainage, are as follows: 1. Identify fields requiring reclamation leaching In order to conduct a leaching programme, it may not be necessary for the entire field to display characteristic symptoms of severe salinisation, which normally include white surface salt encrustations and a lack of plant growth. Monitoring of yields from individual fields, which is currently undertaken by State Farms, would provide an alternative to visual assessment of the need for reclamation leaching. 2 Subsoiling This operation Is required prior to leaching to remove compact layers and to ensure that water can move freely to the field drains. In fields which have recently been subsoiled (within I year of reclamation) this operation may not be required. Subsoiling should be carried out to a depth of approximately 80 cms. The spacing between adjacent clues should be no more than 1 metre to achieve the required degree of soil disturbance. 3 Land Planing The achievement of a well-graded, smooch surface is a prerequisite for successful leaching using the basin method. Laser levelling equipment for use with tractor-mounted land scrapers is recommended as one of several opportunities for agricultural improvement (see Annex E) and this equipment would be suitable for land smoothing prior to leaching.The operation can be efficiently undertaken in large fields such as those at Amibara using a single laser-transmitter mounted on a tripod, and four 80-120 bhp tractors, each with a simple drag-scaper, laser-receiver and control box. The use of laser levelling equipment for land smoothing operations is discussed further in Annex E. Bunding Bunds are required to subdivide fields into a series of conveniently sized basins. The size chosen will depend upon a number of factors, including the geometry of the individual field, the field slope, the bund height, the number of irrigators (labourers) available, and size and number of siphons to be used. In the ten fields under consideration the recommended drain spacing is 45m. The bunds are chosen to directly overlie the subsurface field lateral drains in order to minimise direct seepage into these laterals. The basin size has accordingly been fixed at 45m x 50m in each field. A suitable height of bund is considered to be 30 cm, and can be constructed with machinery currently operating on site. A disc plough should be used to form the bunds, occasional wheel damage where the tractor crosses bunds being repaired by manual labour. For a field of size 1000m x 250m, the total linear distance travelled by the tractor (including turning) during bund preparation is estimated at some 20 km, which at an assumed work rate of 4 km/hr would require 5 hrs. Once the basins have been filled, some settlement of soil and minor breaching is inevitable, but this can readily be repaired using manual labour. The bunding operation completes land preparation requirements prior to leaching. 5 Initial Filling of Basins It has been assumed that siphons of 5 cms diameter, with a discharge capacity of 2.5 1/s under a 25-30 cms head, will be used to fill the basins initially. It is recommended chat basins should be filled to a depth of approximately 15 cms (allowing 15 cms freeboard), and to achieve this 30 siphons would be required to fill one basin in 1.25 hrs. However, this flow rate (i.e. 30 x 2.5 1/s, or 75 1/s) is only half the field canal supply capacity, and therefore to utilise the whole design capacity of 150 1/s, which is desirable for practical reasons, two basins should be filled simultaneously. To simplify Che field operation as far as possible, two adjacent basins at Che lower end of the field should be filled simultaneously (basin No. 5, Rows 1 and 2, in Figure D2). This would require siphoning water into the basins adjacent to the field canal, and breaching the bunds downslope to allow water to flow to basin No. 5 in each row. Once the basin has been filled to approximately 15 cms depth, the adjoining bund can be closed and basin No 4 filled, until all basins in one row have been filled.Whilst the downslope basins are filling, some ’losses' of water will take place upslope through deep percolation because of the greater intake opportunity time here. However, these 'losses are beneficial in that they achieve additional leaching. During the initial filling process, when the soil profile is likely to be at a very high moisture tension, it is advantageous to use as high an intake rate as practicable, consistent with minimising erosion. For this reason the stream size of 75 1/s has been chosen, and this expedient will help to ensure that initial filling of the basins does not become unduly protracted. Manpower requirements for the basin filling operation have been assessed with reference to current practices, in which a team of six irrigators handle the field supply canal capacity of 150 1/s, and work on a 12 hour shift basis. The first shift would require approximately 6.25 hours to fill basin Nos. 5, 4, 3, 2 and 1 in rows 1 and 2, and would proceed to rows 3 and 4. The schedule should be smooched such that 6 hours are given to each row, with 4 rows being completed over a shift. The second shift will require a similar time to fill basin Nos. 5, 4, 3, 2 and 1 in rows 5 to 8. Continuing this operation, a 25 hectare field (1000 m x 250 m) with approximately 22 rows will require 3 days to complete with two shifts (one day, one night) per 24 hours. Assuming a terminal infiltration rate of 1 cm per day and evaporative losses of 8 urn per day, water would need to be replenished at approximately 8 day intervals were leaching to be continuous. A short period of drying out between leaching applications would not adversely affect the efficiency of leaching. At the above projected rate of progress a second field of approximately the same size can be treated by the same team of irrigators before basins in the first field are replenished. Two teams of Irrigators (one day shift, one night shift) would be needed to maintain the leaching programme on two fields, and ten teams would be required for the ten fields under consideration, each team attending to two fields. The control of a flow of 75 lit/sec through breached bunds may present some practical difficulties. To some degree these may be alleviated by having multiple breach points through each bund such that individual flow streams are small and more easily managed. Alternatively the field canal flow of 150 lit/sec could be sub-divided to serve 4 or 6 basins with individual flow streams of 37.5 or 25 lit/sec respectively. Minor modifications to the filling technique would form part of the smoothing required to adjust the programme to particular field conditions. 6 Subsequent Replenishment of Basins The procedure for replenishment of water in the leaching basins should be the same as that for initial filling, each set of basins being filled as soon as possible, subject to operational constraints. Once natural consolidation of the bunds through wetting has taken place, little time should be required for repair and maintenance work, and filling should also require less time as the moisture content of the soil profile increases. rApproximately nine irrigation cycles are expected to be required in order to reclaim a soil profile of salinity as defined for the purposes of this programme. As leaching progresses, colonisation by weeds and other plants serves as a useful indicator that salts are being successfully removed from the soil profile. Reclamation is unlikely to be achieved in less than 2-3 months, and may take longer if local high spots are revealed in the fields. 7 Irrigation Duties and Canal Capacities The leaching application for reclamation, together with evaporation losses is estimated at 18mm depth per day (Section 5) or 180m /ha/day. s This represents a field canal duty of 2.1 lit/sec/ha; approximately twice the peak irrigation duty for cotton. At field canal level the capacity of 150 lit/sec serving an area of generally less than 25 ha is equivalent to a potential duty of 6 lit/sec or more. The reclamation duty of 2.1 lit/sec is thus easily accommodated without overloading the field canals. At tertiary canal level a continuous duty of 2.1 lit/sec could overload the canal capacity if:- (i) The entire area commanded by a tertiary canal required reclamation, and (ii) All reclamation was effected simultaneously. The pertinent data relating to the tertiary canals serving the sample block are tabulated below; Tertiary Canal T3/2 T3/3 Net Irrigable Area - ha Canal Capacity lit/sec 257 240 440 440 Considering tertiary T3/2, the simultaneous reclamation of all land commanded by this tertiary would require 257 x 2.1 - 54Q lit/sec, that la, some 100 lit/sec in excess of its designed capacity . The maximum area which could be reclaimed simultaneously, without interruption of the 1 Source: Amibara Irrigation Project, Water Management Manual. Halcrow. July 1983. 2 Ignoring tertiary losses due to seepage and operation.normal irrigation applications in the remaining fields commanded by the tertiary can be calculated by the formula noted below: Va Aa r + *i < ~ P " Q r r where q ■ reclamation duty - (2.1 lit/sec/ha) a ■ reclamation area - ha q • irrigation duty - (1.0 lit/sec/ha) A1 - area served by tertiary - ha Q - tertiary capacity - lit/sec Applying this formula to tertiary T3/2, 2.1a + 1.0 (257 - a ) - 440 giving a - 166 ha f r For tertiary T3/3 the equivalent value would be 182 ha. In the example quoted the areas requiring reclamation are 122 ha and 51 ha for tertlarles T3/2 and T3/3 respectively, totalling 173 ha. Such reclamation could thus be carried out simultaneously without surcharging the tertiary canals. At secondary canal level surcharge could occur during periods when reclamation is concurrent with peak irrigation demand. Secondary canal S3, serving the sample area, has a design capacity of 1080 lit/sec and serves some 1005 ha net. With all 173 ha in the sample area being reclaimed simultaneously the maximum surchage at the primary canal offtake would be in the order of 1.1 x 173 - 190 lit/sec (ignoring distribution losses). As all reaches of the secondary canal are designed to accept the full capacity at the intake this surchage would only affect the initial short section upstream of the offtake for tertiary T3/1, during the peak irrigation period. This level of surcharge can be accommodated at the structure such that simultaneous reclamation of the entire sample area could be undertaken during the peak irrigation period. For planning purposes a surcharge limitation of 20X should be taken. The actual level of surcharge can be established on site, within limitations of available head, scour considerations, freeboard etc. At Primary canal level the proportion of area to be reclaimed, relative to the total irrigated area, is likely to be small and no limitation on canal capacity is anticipated. In Melka Sadi the area projected for abandonment by the end of 1988 is some 1275 ha (Annex E, Table 16). By this time some 1850 ha of sub-surface drainage is programmed to be completed over the initial 18 month installation period (Annex D, Fig. 20). Assuming reclamation of abandoned land were to keep pace with drainage installation the annual maximum rate of installation could be 1275/1.5 • 850 ha/year. Allowing a 3 month reclamation period then the maximum amount being reclaimed during the peak irrigation season would be in the order of 200 — 250 ha which would have a negligable influence on the Primary Canal capacity.The above computations relate to the worst case of superimposing a reclamation duty on the peak irrigation duty. For both cotton and bananas the peak Irrigation duty only occurs for a limited period during the year and reclamation activities could be intensified outside these periods such that no conflict need occur. 8 Land preparation prior to next cropping season Once reclamation is complete, the normal sequence of tillage operations should be carried out prior to planting. It is recommended that the opportunities for additional agricultural improvements described in Annex E are incorporated at an early date: however, as subsoiling is a prerequisite for leaching, this operation could possibly be omitted as part of land preparation for the following cropping season on recently reclaimed fields. Where reclaimed fields are left fallow during the canal closure period, some secondary salinisation will Inevitably cake place, but can be controlled by the following cropping season irrigation applications. The requirements for the reclamation programme in terms of machinery operations and manpower, are summarised in Table DI.<n sj m vo TABLE - DI RECLAMATION PROCEDURE FOR 10 COTTON FIELDS: INDICATIVE MACHINEY OPERATIONS AND MANPOWER REQUIREMENTS FIELD Net Approx Approx Filling Irrigator Subsoiling A Land^ Bunding6 Total No F3/- Area (ha) No j Time (HRS) Team Basins Allocation Planing ----- (HRS)------------ 2/19 19.18 85 53 la 29 19 4 52 2/20 13.63 61 38 lb 20 14 3 37 2/21 12.70 56 35 2a 19 13 3 35 2/22 15.66 70 44 2b 23 16 3 42 2/23 20.08 89 56 3a 30 20 4 54 2/24 20.34 90 56 3b 31 20 4 55 2/25 20.04 89 56 4a 30 20 4 54 3/33 11.82 53 33 4b 18 12 2 32 3/34 18.12. 81 51 5a 27 18 4 49 3/35 21.42 173 95 59 5b 32 21 4 57 467 (2.7hrs/ha) 1 Calculation assumed all assumes a basin size of 45m x fields are 250 in Length but 50m; in all these of variable width. fields Che recommended drain spacing is 45m» and it is Assumes basin volume 337.5m3 (45m x 50m x 0.15m) is filled in 1.25 hrs at a rate of 75 1/s. a and b denote day and night shift teams. Assumed rate of work 1.5 hr/ha (Table H9, Annex H). Assumed rate of work 1.0 hr/ha (Table H9 Annex H). Assumes linear bunding of 0.83 km/ha, and rate of work 4km/hr. tFIGURE DI MASTER ORAINAGE PLAN FOR MELKA SAOI AND AMIBARA AREAS LOCATION OF FIELDS TO ILLUSTRATE TYPICAL RECLAMATION PROCEDURE x------- f3/3/34 F3/3/33 F 3/2/25 F 3/2/24 * X-X F 3/2/23 IT' EXISTING PILOT DRAINAGE PROJECT AREA LEGEND IIIIIIIU X------X- MELKA SADI BANANA UNIT boundary Dyke Service Rood Primory Conal Secondary Canal Tertiary Canal Field Canal Secondary Drain Tertiary Dram Field Drain Melka Sodi Banana Unit Boundary Contour in metres Bunds Locotion of chosen fields 05 I Okm SIR WILLIAM HALCROW a PARTNERS JULY 1965RFIGURE 02 MASTER ORAINAGE PLAN FOR MELKA SADI ANO AMI0ARA AREAS SCHEMATIC ILLUSTRATION OF RECOMMENDED LEACHING BASIN LAYOUT DETAIL 'A' ( Refer to Figure DI) Scale I 10 000 LEGEND Subsurface Field Lateral —x— Field Drain ------ — Field Canal ii || Bunds 30cm high =jpr= constructed by II 11 Troctor and Disc Plough Siphons (5cm dia. 2 5 l/s) discharging into Leoching Basins Schematic indication I of need to breach 8 repair Bunds for downslope basrs I P£Taii,'9‘ Scale. 1 = 2000 SIR WILLIAM HALCROW 8 PARTNERS JULY 1985r 11 n ■ ■ ■ ■B bFIGURE I MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS PRIOWTY AREAS FOR SUBSURFACE DRAINAGE Sir Wlllkom Malcrow A Poftrwf* July <985riwunc. c. MASTER DRAINAGE PLAN FOR MELKA SADI AND AM I BAR A AREAS DEPTH TO GROUNDWATER CONTOUR MAP (MAY 1984 ) Sir Will Ha I crow & Portners i9«5 JFIGURt 3 o ■ ■ I I I I I MASTER DRAINAGE PLAN FOR MELKA SAO( ANDAMIBARA AREAS DEPTH K) GROUNDWATER CONTOUR MAP (JUNE 198A) I SlrWIlhom Hakrow I Partn«rc July 1985I I FIGURE 4 master drainage plan for melka SAOl M AND AMlBARA AREAS GROUNDER SALINITY CONTOUR MAP (0CT06ER 198A) 9rWlU«n Halero- t Partners JuiT r9«>MbURt O Sir William Hal crow ft Partners July 1988FIGURE 6 MASTER DRAINAGE PLAN FOR MELKA SADI ANO AMI9ARA AREAS PIEZOMETRIC SECTIONS^ MELKA SADI SIR WILLIAM HAUCROW 8 PARTNERS JULY 1985 FLOOD PROTECTION DT*€ MARKED FPDFIGURE 7 MASTER DRAINAGE PLAN FDR MELKA SAOI AND AM18ARA AREAS PIEZOMETRIC SECTIONS AMI0ARA a 3 REDUCED LEVELS (ml SIR WILLIAM HAUCROW a PARTNERS JULY 1965 FLOOD PROTECTION DYKE MARKED FP D.UliUHt b F ■i k k 0 I master drainage plan for MELKA SADI ANOAMIBARA AREAS projected rate of encroachment WITHOUT DRAINAGE 5ltWllliom Halcrow 1 Portrwrs ’CTliFIGURE 9 MASTER DRAINAGE PLAN FOR MELKA SADI AND AM I BARA AREAS PROPOSED LOCATIONS OF ADDITIONAL OBSERVATION WELLS SIR WILLIAM HALCROW a PARTNERS JULY 1985FIGURES 103 II MASTER CRANAGE PLAN FOR MELKA SADI AND AMlSARA AREAS FIGURE 10 LEACHING OF SALTS FROM DIFFERENT DEPTHS IN PROFILE IN RELATION TO THE CUMULATIVE DEPTH OF IRRIGATION APPLIED I ■ tv \ X I V \ 11 \ \ 1V X \X (Mean of two basins in Pilot Dromoge Scheme, i.e, land with drains at 2m shown as bold lines und rained Io nd assumed to be os shown by doshed lines.) r'\v 1 \ x \ X ’120 cm , 60cm Hhr- 30 o tfJ O 20 O c Z u 10 o Qj uCCrTi □% Cum Depth of Irrigation water applied (mm) Percentage of salts leocned FIGURE II LEACHING OF SALTS FROM DIFFERENT DEPTHS IN AN UN DRAINED SOIL PROFILE 8Y SUCCES SIVE IRRIGATIONS AS PER PRESENT IRRIGATION SCHEDULE (From Figure IQ) Percentage of salts remaining SIR WILLIAM HALCROW ft PARTNERS JULY 1905FIGURE 12 MASTER DRAINAGE PLAN FOR MELXA SADI AND AMIBARA AREAS EXPECTED INCREASE IN TOPSOIL SAL IN IT Y DURING COTTON FALLOW PERIOD AS A FUNCTION OF SALINITY OF WATERTABLE AND SOIL TYPE DRAINS AT !Bm OR WATERTABLE AT 0 75 m AT START OF FALLOW SIR WILLIAM HAUCROW a PARTNERS JULY 5985 i i “ *^-**^- — Increose in Salinity (AECe rnScm J)Depth below Soil Surfaco tmm) FIGURE 13 MASTER DRAINAGE PLAN FOR MELKA SADI AND AMI0ARA AREAS EXPECTED CHANGE IN LEVEL OF WATERTA8LE DURING CROPPING SEASON OF COTTON WITH DRAINS AT I 8m, VERTISOL ( Fine Textured Soil) Sail Surface ------- *— \ 1 4-------------- I 1 1 NOTE- Irrigation includes a heavy leaching factor 2 pre-plant and sJightly increased posr- planting applications ------------------------ I 1 1 1 1 \ \ \ \ \ oca Ilf ^Rooting Depth Soil at Field Capacity (FC) IWF>- Droin ufl^el Saturated Soii(S) 2000 2400 IWF) 1989 Past-Planting Irrigation and Gross Application (mm) Doys before and after planting (P) SIR WILLIAM HALCROW ft PARTNERS JULY 1985FIGURE MASTER DRAINAGE PLAN FOR MELKA SADI AND AMIBARA AREAS EXPECTED CHANGE IN LEVEL OF WATERTABLE DURING CROPPING SEASON OF COTTON WITH DRAINS AT I 9m. SIR WILLIAM HAU2ROW ft PARTNERS JULY I9Q5MASTER DRAINAGE PLAN FOR MELKA 5ADI & AMI3ARA *REAS FINAL REPORT VOL 1 VOL 2 VOL 3 VOL 4 VOL 5 VOL 5 - - - - - SUMMARY MAIN REPORT ANNEX Al : SOILS AND ORAINABILITY CLASSES ANNEX A2 : SOILS AND DRAINABILITY CLASS MAPS ANNEX A3 ; DESCRIPTION OF AUGER BORES ANNEX 9 : GROUNDWATER AND SALINITY - ANNEX C : HYDROLOGY VOL 7 VOL s vol a ■ * - - - • VOL 10 - ANNEX D : ENGINZESING ANNEX E AGRICULTURE ANNEX F : MARKETING AND PRICES ANNEX G : PROJECT COST ESTIMATES ANNEX H : FINANCIAL AND ECONOMIC EVALUATION ANNEX I EVALUATION METHODOLOGY - GUIDELINES ANNEX J : ENVIRONMENT AND HEALTH4ANNEX C - HYDROLOGY Contents Fa Se 1 INTRODUCTION 2 AVAILABLE DATA 3 CLIMATE 3.1 Rain rail 3.2 EvnporatLon 4 WATER QUALITY 4.1 Existing Water Salinity, 4.2 Projected Salinity of Return 5 IRHIGATLON WATER UTILISATION 1 2 4 4 Flows to Awash River 6 0 8 10 11 5.1 Evaluation of Present Recomended Application A&uunts 5.2 Contribution to Groundwater Recharge of Excess Irrigation Applications 5.2.1 Cotton Irrigation 5.2.2 Banana Irrigation 6 FLOOD ESTIMATION FOR THE EASTERN CATCHMENT AREAS 6.1 The Catchments 6.2 Flood Estimation 6.3 Design Return Period 7 FLOOD ESTIMATION WITHIN THE PROJECT AREA 8 FLOOD ESTIMATION FOR THE AWASH RIVER 8.1 Estimation of Peak Flow Rates 8.2 Construction of Design Hydrographs 11 1^ 12 13 I4 „ . 14 lb 18 20 20 23 (i)Contents eon cd 9 HYDRAULIC MODEL STUDY OF THE AWASH RIVER 9*1 General 9*2 Extent of present Model 9.3 Flood Hydrographic Inputs 9.4 Hydraulic Properties Adopted tor the Flood Routing 9*5 Testing of the Model 9.6 Estimates of Flood Levels REFERENCES Page No 24 na 25 25 26 2A (Li)Contents contd tables 1 Details of Gauging Stations in the Awash Valley* 2 Monthly and Annual Rainfall Statistics far Melka Water. 1965-1983. J Monthly Rainfall Distribution ar Melkfl Warer as a Percentage of the Mean Annual Rainfall, 4 Monthly and Annual Reference Crop Evaporation Computed frora the Climatological Records for the Melka Warer Research Station. 1981-1904. 5 PhysicaL and Chemical Analysis of water in ATP area. 6 Chemical Analysis of Water Samples taken on 29 October 1984. 7 Sunwnary of Results from Electrical Conductivity Monitoring. 8 Contribution of Excess Irrigation Applications co Groundwater Recharge on Melka Sadi Units 1 and 2 for Three Cropping Seasons. 9 Summary of Flood Estimates for the Eastern Catchments. 10 Correlation Matrix of the Annual Floods for the Awash River at Awash Station, rhe Reas™ River at Awara Melka and the Kebena River at Kebena* Numbers in brackets indicate the number of pairs of points used in computing the correlation coefficient* 11 Suooary of flood estimates (m^/s> and Growth Fjctors for SeLected Return Periods for Awash Station. Awava Melka and Kebena* 12 Maximum Predicted Flows for 1 in 10, I in 20 and 1 in 50-Year Floods for SeLected Points along the Awash River. A.l Estimates of Wetted Perimeter for Secondary and Tertiary Canals in Melka Sadi Units 1 and 2. A.2 Water delivery to fields on Melka Sadi Units 1 and 2 from Primary Offtake Ho 4 during 1981. 1982 and 1983 Irrigation Seasons. A* 3 Planting Schedules for Irrigation Seasons. Melka Sadi Units 1 and 2 for the 1981 , 1982 and 1933 A-4 Monthly Net during 198L A.5 Monthly Net during 1982 A* 6 Monthly Net during 1983. Irrigation Water Requirement a for Melka Sadi Units 1 and 2 irrigation Water Requirements for Melka Sadi Units 1 and 2 * Irrigation Water Requirements for Melka Sadi Units 1 and 2 (Hi)Contents coned A. 7 Suninary of Contributions to Groundwater Recharge from Melka Sadi Units I and 2 during the 1981. 1982 and 1983 irrigation Seasons. A.8 Annual Water Ba Lance for Banana Plantation at Melka Sadi for the Years 1981-1983. A.9 Monthly Wet Sadi during Lrrigat ion 1981. Wat er Requirements for Banana Plantation a t Helka A. 10 Monthly Net Irrigation Water Requirements for Banand Plantar ion at Melka Sadi during 1982. A. 11 Monthly Net 1 rr igation Water Requirements for Banana Plantation at Melka Sadi during 1983- B. 1 Variation in Drainage Water and Awash River Salinity Development. for 14,200 ha Cotton ( iv)Contents contd FIGURES 1 Location of SaiapLing Points for Water Quality Studies 2 Typical Hydrograph Shapes for the Awash River 3 Layout of the Awash River Hydraulic Model 4 Design Hydrographs for Awash River Model 5 Surface Water Profiles for the Awash River under Various Design Conditions APPENDICES Appendix a: Trial Water Balances for Cotton and Bananas A-l Introduction Seepage Losses A-2-2 A-2’3 A’2-4 A-3 A-4 Evaporation Losses froa Canals Rejection Losses Evaporation Losses from Uncultivated Area Water Balances for Cotton Water Balances for Bananas Appendix B; Projected Salinity of Drainage Water (v)1 INTRODUCTION The hydrological studies that have been undertaken during this Project may be broadly divided into those concerned with the estimation of crop water r equ ir eiaents rind irrigation water utilisations and those concerned with the assessment of flood flows and drainage discharges. The former have been directed principally to the implications of the guidelines for managing and operating the Amibara Irrigation project in terms of the overall water balance. Crop water requirements estimated from the prevailing climate have been compared with rhe amounts of water actually applied to a representative section of the irrigated area in order to quantify the contribution to deep percolation of excess in the volume supplied. Water quality implications have also been considered. The flood studies may be sub-divided into three parts: consideration of floods in the Awash River and the design of protection works along the western aide of the project area; the behaviour of the catchments draining to the eastern side; and the design of the internal drainage system. Both of these major lines of investigation have been heavily dependent upon both hydrometric and climatalogical records obtained from Water Resources Development Authority (WRDA) archives and other sources. This Annex therefore begins in Section 2 with a sunnaary of the available data and continues in Section 3 with a description of the prevailing climate. Particular attention is paid to rainfall and evaporation* and the development of a rainfall depth-duration—frequency 1 DDK) relationship and the estimation of reference crop evaporation for the Amibara area. Attention is then turned in Section 4 to water quality considerations prior to the discussion of irrigation water utilisation in Section 5. The final four sections of this Annex are devoted to the flood studies. The evaluation of the eastern catchments is described in Section 6 and the determination of internal drainage duties for the project area in Section 7. Section B discusses the estimation of floods far the Awash River and Its tributaries* and Section 9 outline* the development of a computational hydraulic model fojr the reach of the Awash River from Melka Sadi downstream along the western aide of the Amibara area and into Angalele. 12 available data The principal lource of climatologLcaL data tor che project area was the net eoro lap ica I station at rhe Nel'ka Water Research Station. Those data were employed primarily co iMSiM reference crop evaporation and effective rainfall totals for the 1981. 1982 and 198J irrigation reasons; ind to derive a rainfall lepth-durat ion-frequency (DDE) relationship to bn used in estimating floods and drainage flows. For the farmer, record* of temperature, relative humidity, sunshine hours and wind speed were required in order to Apply Che mod Hied Penman method, and effective rainfall was eitiraaced on the basis □£ 0Dsei-v?d rainfall and computed evaporation depths using US Department of Agriculture procedures (see Ref I),* Values of the climatic variables were abstracted from Jaily iuramari«s provided by the Research Station. Far the analysis of rainfall totals shorter than che observer day, access was provided to the original weekly recorder charts, which have a time scale of 54 ram to one day. Records of river flows for gauging stations tn the vicinitv of the project area war* supplied by Cha Water Resource* Development Authority (W£DA> in Addis Ababa, Details of these stations are presented in Table I. for station numbers 10,11 and Jb, luanaries of son dally stages and flows were provided from the beginning of record to December, L982. dimilir inforradc i on for station 12 was only available up to mid-Auguac, H80. Surrwnarias of monthly flow volumes, maSimuiH mean daily flows, ins taut aiteoua peak -discharges and luumm mean daily flows ware also obtained along with rating curves for station* 10,11 and 36. Values of monthly run'-off volumes* and raasimum >q<ih Jailv flows were abstracted frutu WRDA records tor station 40. Owing to the short length of record, no data lusMsarias were available for station 14, out of th- uriguial water level recorder tbstad in table I apart from Koka Dam. but a detailed *i am nation was carried charts for this site and all thoseFurther background information vn obtained fro® the Awash River and Lake Banins Hydrometeorological Summaries for the years 1976-1979 (Ref 2-4). TABLE 1: Details of gauging ata;ions in the Awash Valley Station He 1 1 River I 1 Station 1 1 Area (4sa*> Records fr» 10 1 Awash 1 Awash station ( itno Attgxrit, 1962 11 K es s era Avar a **el<a 11)5 Rty. 1962 12 | Kebtena Cefi«a 12M Jniy. 1962 11 34 36 40 1 Awash i! I Awash i I Awash M«1U UJi ■Mika Mirer IL<a j*a 213OC vcven&er. 1982 2 MIO Aa|-v«:. IV2 11*2^ Aa<utt. 19M i1 CLIMATE the Middle Awash Valley Lies in in area of semi-arid cliaate with 1 bi-annual rainfall pattern typical of the uomoon conditions associated with moveci ent of the inc *»r-tr op ica I convergence cone at this Latitude. Only one meter*logical station, Melka Warer Research Station, ha* long Cera records. Very high temperatures, sometimes exceeding 40*C are reached during May. June and July. Coolest months are December and January when temperHures at low as 6*C have been recorded. Mean monthly humidity varies tr?ni 58* in August to J6t in June. Minds are generally strongest in June and July when the run of wind cm exceed 200 kin/day. cloudy turner months with annual 1.1 Hamfall regional rainfall pattern is affected The considerably from year co year. Sunshine Coralt are high even sunsh ine hours being in excess of during the SH>r 3000 hours. Rainfall Ethiopia is usually attributed to seisonaL mov-snent of Che inter occurrence in the eastern highlands of opical convergence cune. eastern Ethiopia, th* movement of chi* sone, modified by the orography of reaches its northern limit in July th* only station Ln the area with long term rainfall (L965). Th a iii»4ii nun ch Ly annual rain tai I during chi* raintall recorded u 226 iran in August per LJd • nd August records i« is S60 it about 18 *N. Melka War er The highest 1W, rhe Characteristics of are given in Table 2. A study of che Malka Merer record shows a di sei net nonrhly rainfall partem, The typical monthly rainfall distribution percent■** of the annual muan is shown in at this station can be expressed as Table I. is fairly - during Ot the cw, 4tA&LL 2 Muiithly Jud Annual tuitifalJ Sratuuci. for Mel kA tlarcr, 1965*1983 -J < -□ z ■fcr < Lf*l ■* '3 Zl a J3 iT* S’ 1— — — u a — — —• e > 5 H § 3= fR ■a s — S“ r* <■ o Ri z*1 •* •C c n— ooa x r* a ** ■x £OX —' “1 j < lift —> g u M a X- — - = <* z < J* s XX rwi L *" r J a *—o r* _. w* =* :i • il 5 3. s L_ — — ** j0 E <— * a *»» —X 8 X 3 ft, -■ Ri > 5 0* Xa «— Ji — — « ft «- Z— - ;3 — *'- o z-1 z 5 X ? JI*. '■ - u*. — c ** *« *3 -G sG XX * ! ;: 5 *< ■ «— • — _J ' o .» a! fk- -0 •0 0* A a £ 0t 3a 'i c □ hT. & £> Xk •0. o — s1 i 11 !4 * -* I c * o: • w —J 1 K s _. — >» i3 -* 5« g:: ii 3 X• - 33Another aspect of chn climate of Che project: and of particular importance in flood estimation and drainage design is the frequency of heavy rainfalls within short durations. An autographic rainfall recorded has been in operation at the Malka War er ttesearch Station since July, 1968. Individual frequency analyses of )-h, 6-h, 12-h and one and two observnr-day rainfall totals were carried out by Halcrow (Ref 5) using rhe data up co April, 1974. These results were updated using the records from both daily and autographic raingauges at Melka Warer up co December, 198J. For each selected rainfall duration, the "Peaks-ovtr-a- Chmhold" or partial duration aeries approach was* applied to the recorded depths. The one 4nd two observer-day rainfall depths were adjusted to 24-n and 48-h totals using the standard factors of l.LJ and 1.34. to differ little from the 6-h totals, an effect that was attributed to sampling variability. The estimated depths for selected durations and return periods were therefore pooled and subjected to a multiple linear regression analysis with depth 0 (r»J, as the dependent and both Juration, t (h), and return period, t (y), a* independent variables. The resultant equation was: D • 16.29 t°--3 r°-^J The form of this egpreeeion, with depths roughly proportional to the quarter-root of the duration of rainfall, i* m broad conformity with recorded espar Leilce ali^whore m similar climates. 1.2 gvapurar inn the daily records of temperature, relative humidity. sunshine hours and wind ruts et the He Ike War er e I inato lag id e I station were employed to compute mrinaces of reference crop evaporation using the modified Penman method. These calculations were cunt n. ml to th- period January 1981 to April 198-, over which records are alau available for the planting and irrigation schedule* for different units within th- lrrU.,t-l area. a <h.„, of munch!, evaporation rates and annual totals is presented m fable 4. fh« mean -nfluxl rH r 12-n depths were found . „,„ efnp wvapor.ttan over the three rear. 19MH19EJ given bv ihlerw, 19 Ji i K-r V end he ’all j , which compare, well with the 2181 » t m given by ME DECO, 198? (Ref 6), *i.<■ #1 5*n lild far pr av itbikia ffiri.TABLE 5 - PHYSICAL AM) CM Ml CAL ANALYSIS OF HATER IN ATP AREA All LKtita are in iaq/1 unleoa otherwise stated 1 LOCATION 1 11 1 DATE | 11 1 PH | [ 1 EC* 1 1 1 Cl 1 1 Ce2+ I I 1 2 Ng * 1 1 MCOj | I 1 $o42- | 1 1 Ml2* 1 1 Fe2* 1 | PO5 1 1 1 *2 1 1 1 NOj 1 1 *3 1 17 | Color2 1 Tirblditjl 1 DIU 11 1 lesp 1 1—1------------------------- 1 1 1 1 R1 1 R7 1 R4 1 1 co 1 06 I 1 R1 1| I 13/10/84 ) 1 8.1 1 1 0.32 I1 11.25 I 1 1 20.04 | 4.97 1 1 100 | 1 1 4 1 Nil | 0.15 1 0. » 1 I Nil 1 I 2.64 1 I 0.732 1 | 540 11 1 15C | 6.72 11 1 ?2 *C | | 13/10/84 | 8.0 1 0.56 11 37.5 | 20.04 | 4.26 1 165 | 16 | 0.3 | 0.17 1 0.4 1 Nil I 1.76 1 0.64 | 580 1 150 | 6.4 1 25 *C I 1 15/10/04 | 7.9 | a.64 1 43.75 | 22.1 1 4.97 1 185 | 19 1 0.4 | 0.19 I 0.48 1 0.0066 1 3.96 | 0.549 | 510 1 1 160 | || 6.08 1 23 ’C | 11 I 13/10/84 | 8.1 1 0.28 1 12.5 | 21.042 1 1 4.96 | 1 110 | 1 31 1 Nil 1 0.14 1 1 0.39 1 I Nil 1 3.52 I 0.61 1| 760 1 225 | 6.4 | 1 22 *C | 1 13/10/84 | || 7.7 1 0.68 1 85 | 32.1 1 7.81 | 125 | 36 | 0.1 1 0.08 1 0.4 I 0.01 1 2.2 I 0.22 1I 290 1 80 | 6.4 | 1 21 «C I 1 I 11 || I 16/10/84 | 8.0 | 1 0.2B 1 9.5 1 1 21.04 | 1 6.W | 1 100 | 1 4I 1 Nil | 0.08 1 1 0.32 1 1 Nil 1 1 0.88 1 0.92 11 510 1 140 1 6.4 1 1 21 ’C 1 1 R2 1 16/10/84 | 7.9 11 0.385 1 21.25 1 20.04 | 4.26 1 125 I 1 R3 I 16/10/84 | 8.1 | 1 0.44 1 25 | 20.04 | 4.26 | 14 Q | 5 1 Nil 1 0.1 1 0.28 1 0.017 | 1 D.Bfi 1 1.04 | 515 1 144 | 6.4 1 1 25 «C 1 8 1 Nil 1 0.10 1 0.19 1 Nil 11 8.8 1 0.85 1 625 1 200 1 6.72 | 1 25 •£ 1 I RO 1 16/10/84 | 8.1 1 1 0.50 1 31.25 | 20.04 | 4.97 | 140 | 15 1 Nil 1 0.5 1 0.35 1 Nil 1 Nil 1 0.7J | 600 1 175 1 6.4 | 24 *C 1 1 RS I 16/10/84 | 8.1 1 0.52 1 35 | 20.84 | 4.4 | 150 1 15 I Nil 1 0.12 1 0.3 1 Nil I Nil 1 D.85 | 612.1• 1 lea | 6.4 | 25 °C | 1 C4 ( I 16/10/84 | || 9.1 1 0.28 | 1 11.25 I 20.04 | 4.97 | 110 1 3 1 Nil 1 0.2 1 0.29 1 Nil I DO 1 16/10/84 | 8.1 1 1.80 1 310 I 53.11 | 12.1 1 160 | no i Nll 1 .1.04 1 0.3 1 11 : o.ob i I 0.88 I J.96 | 11 1 iso | u 9.68 1 0.15 | 140 1h J9 i 7.04 1 25 °C 1 1 25 «C 1 1 D5 I 16/10/84 | 7.9 1 0.725 1 93.75 | 35.07 | 7.81 | 120 | 43 1 Nil 1 0.08 I 0.35 1 Nil | 2.2 1 0.366 I 150 I 45 1 6.54 | 24 «C 1 I D6 1 16/10/84 . 7.9 1 0.48 1 42.5 | 29.06 1 12.36 1 115 | 22 1 0.1 1 0.1 1 0.3 I 0.01 I 4.4 I 0.67 | 360 1 ™ 1 7.0) 1 23 °C 1 I D16 | 19/10/84 | 7.7 1 5.00 1 1025 | 64.13 | 9.91 | 635 | 560 • | 0.2 | 0.04 1 1.05 I 0.02 | 6.16 1 Nil | 65 1 1 I - 1 J2.5 °C 1 I 018 I 19/10/84 | || 7.4 I 14.22 | 1 3250 | | | 284.57 | 145.3 | 520 | 1250 • | 0.1 1 O.D 1 1.1 1 0.1 1 >1.68 1 0.122 | 11 10 1 J 1 - 1 M t 1 I All* 5 I 19/10/84| 7.9 1 0.94 1 15 I 6.41 | 11. u I 335 | 24 | 0.7 | 0.02 1 24 1 Nil | 6.6 1 0.146 I 150 1 1 43 1 1 -1 1 1 I AIP 8 I 19/10/84 I 7.4 I 0.705 1 27.5 | 42.1 I 12.8 | 235 1 B | 0.6 I 0.02 1 0.5 I 0.026 | 4.4 1 0.061 | I AlP 10 I 19/10/84 | 7.2 1 5.52 1 1250 | 84.168 1 8.451 | 330 | 280 | Nil | a.02 1 0.17 i 0.0627 1 4.84 1 0.2074 | I AIP 25 I 19/10/84 | 7.4 I 1.40 1 75 I BO. 16 1 21.31 1 380 | no i 0.8 | 0.05 1 0.35 1 0.0033 | 4.4 1 0.183 | 45 | 18 1 I AIP 27 I 19/10/84 | 6.8 1 7.20 1 1275 I 76.55 | 39.89 | 245 1 1200 • | 0.36 1 0.05 1 0.28 1 1.63 | 29.48 1 0.183 | 35 1 U i 40 | 10 1 160 I v> I 1 AIP 39 | 23/10/84 | 7.9 1 2.70 1 105 1 1 60.12 1 22.77 | 175 I 750 • | 0.5 I 0.02 1 1 0.2 1 O.Q165 | 3.96 1 0.09/6 | -1 1 - 1 32 -C 1 1 - 1 27 «C 1 I AIP 51 | 23/10/84 1 7.1 1 25.92 I 5800 | 1 745.49 | 558.91 | 220 | 4000 • | 14.8 | 0.25 11 0.39 1 a.oi | 1 19 1 - 1 28 »C 1 5.72 1 7.32 | 110 | 40 1 - 1 30 »C 1 I l | Pll I 24/10/84 | 7.6 I 12.42 I 2700 I 1 36.1 | 51.37 | 370 | 19.8 I From pilot 1 1 1 1 1 1 1 1450 • | 1 1150 •! 11 1 Nil | 80 I1 | 25 1 92.48 | 1 | | I scheme al I 24/10/84 | 7.8 I 9.00 I 2050 | 1 260.52 | 460 | I puaping site 1 1 1 1 1 1 1 1 1 J L 1 11 LL LL Nil | 1 0.2 I 1 L 0.05 1 0.2 1 0.3 I 0.8 I 1 Nil | 11 J 0.246 | 11 11 28.16 1 Nil | '1 1L | 210 1 5’ ’ 1‘ IL 1 -1 1 -1 1 1 1 31.1 ”C 1 1 25 'C | | •la aS/c* •2s Platinua cobalt color unit • 3 = Fcraarin Turbidity Units (FIU)TABLE 6 - CHEMICAL ANALYSIS OF WATER SAMPLES TAKEN ON 29 OCTOBER 1904 1 1“ --------------- r | Sample No | 1 1 2 3 i---------------------------------- r 141 1------------------------ 1 5 16 1 1 1 1 1 11 1 r 1 1 1 1 Tea C ’ Sampling I Awash River at 1 Awash River u/s ot 1 Primary Drain Pilot Drainage Observation Wei 1 1 Observation Well I i I I Location 1 Diversion Headworks 1 Primary Drain Outlall 1 Out fall I Scheme — drain | AT? 39 1 ? 11 1 11 1 11 1 t i 1 l 1 pH i 1 Electrical Conductivity mS/cm l 11 1 1 1 1 1 1 1 1 8.0 .24 76.0 11 8.2 1 8.1 1 1 1 8.2 I 1 |1 .16 1 5 76.0 I 1 1 8.43 1 | 780.0 I 8.1 2.13 2$2.0 I I f I 1 8.2 1 | 9.03 I || 1 Total Hardness as Ca CO3 .60 120.0 1 1 | 1 356.0 1 I I Total Silica as SiO2 1 1 | 1 I 28.8 3U.B 1 I 1 30.8 1 51.3 1 10.0 16.3 1 11 1 Cations 1 1 1 1 | 1 11 1| 1 I 1 11 | Acnmon l urn NHa 1 9.68 0.9 | 1.43 1 0.13 1-1 Sodlum Na 11 47.7 121.0 173.0 I 1901.5 1 507.4 2723.4 | Potasslum K 11 6.1 1 6.6 i 7.8 | 24.8 Io.8 108.9 | Calclum Ca 11 24.1 19.2 36.9 1 195.6 1 67.3 I 52.9 Magnesium Mg 11 3.9 6.8 6.8 | 71.0 1 25.3 54.5 1 Iron (total) Fe 1 1 0.6 | 0.22 | 0.56 | 0.1 1 0.14 1 I 11 | 1 1 I | I 1 Anions 11 1 1 1 i 11 11 1 1 1 1 | Chlor ine Cl 11 14.2 I 42.5 | 127.6 | 2226.3 I 127.6 2892.7 | Nitrite no2 NOj 11 0.05 0.07 | 0.03 I 0.11 1 0.05 I 0.06 | Nitrate 1 0.2 1 0.7 | 0.4 9.9 1 0.2 0.2 | Flourine F 11 1.11 I 1.42 1.42 1.95 1.11 1.6 Bicarbonate HCO3 11 170.8 | 219.6 | 219.6 9 76.0 I 341.6 707.6 Carbonate Sulphate CO3 SO4 11 1 1 | - 141.3 1 - | 1444.0 Boron B 11 11 30.5 0.08 | 1 1 101.ft 0.14 0.25 954.7 3.75 I | I 897.1 1.08 Sodium Adsobtion Ration (SAR) 11 11 55.3 76.0 | 74.3 | 83.6 79.0 1 2.11 92.3 I I 11 _1I4 WATER quality 4.1 Existlag Water Salinity The quality of irrigation water supplies from the Awash River has been monitored and reported on in previous studies (Ref 5, 5, 7 and 8). Ln general these studies have concluded that good quality water is available and that only limited Leaching applications are required Co maintain a favourable salt balance within the soils. Ln Phase II of the Project, detailed chemical analyses were carried out on site during the visit of the WRDA chemist on 25 water samples. Eight samples were taken from the Awash River, two from the Main Canal, seven from project drains and eight from observation wells. The results of these analyses are given in Table 5. A further six water samples were analysed at the Central Laboratory and Research Institute (CLRI) in Addis Ababa and the results are reproduced in Table 6. These latter samples comprised two from the Awash River, one at the Primary Drain Outfall, one from the Pilot Drainage Scheme and two from observation wells. These results, supplemented with those obtained from previous studies, were used to determine Leaching requirements under conditions of sub-surface drainage (Annex B). The samples of sub-surface water, analysed on site, taken from boreholes and the Pilot Drainage Scheme indicated high suLphate ion concentrations. Only four of the eleven samples analysed returned values of less than 160 m/1 which ls considered to be the upper threshold for the tolerance of Ordinary Portland Cement in concrete pipes (Ref 9). These high results were also confirmed by the three subsurface water samples analysed at the CLRI. A programme of monitoring electrical conductivity (EC) and pH values in the project canals and drains was carried out between Aug 1984 and Nov 84 with the objective of determining variations in sale concentration within che system and 8identifying sources of saline water intrusion. Seven sampling points were seLected on the Main Canal and six on the Primary Drain. A further five points on the Awash River were also monitored, covering the reach from the Melka Sadi Diversion Weir to the Primary Drain Outfall. The locations of the sampling points are shown on Figure I and the results are reproduced in Table 7. The electrical conductivity of water samples at the diversion weir and in the upper reaches of the Mam Canal exhibited little variation, ranging from 0.2 to 0.3 mS EC values at the lower end of the canal system generally indicated an increase in salinity, with values ranging from 0.2 to 0.5 mS cm-1, the average increase being in the order of 0.1 mS cm“l• No significant trends in values were observed, within the limited sensitivity of the conductivity meter empLoyed. Samples of river water similarly did not exhibit any consistent pattern although an increase in conductivity between sampling points at Melka Sadi Diversion Weir and Melka Warer, downstream of the confluence of the Kessem and Kebena tributaries was apparent. The detailed chemicaL analysis of river water samples carried out on site included two sets of observations in which the progressive increase in salinity downstream is clearly exhibited on samples taken on the same day. Both sets of observations were carried out during mid-October with low river discharges (13 and 16 Oct), such Chat Che influence of contamination from other sources would be more discernible than earlier in the season when flows were higher. Similar increases in conductivity are noted in the observations of Awash River waters carried out by Italconsult in early 1971 (Ref 8). The electrical conductivity measurements of the project drain flows showed large variations, reflecting the different origins and volumes of drainage discharges. The highest values occurred in the Primary Drain downstream of the outlet from the Pilot Drainage Scheme. Samples taken upstream of this point indicated a range of 0.4 to 0.6 mS cm^l to be typical for surface drainage flows, largely comprising rejection flows trom fields and irrigation canals. With the combined flows from the Pilot Drainage Scheme the conductivity values m the Primary Drain increased appreciably, a maximum reading of 5.5 mS cm being recorded. Downstream of this point the readings indicate a gradual reduction in conductivity value, as a result of dilution from tributary inflows. Values at -i 9TABLE 7 - SUMMARY OF RESULTS FROM ELECTRICAL CONDUCTIVITY MONITORING 1 1 DATE: 1 | 1 1 16/8/84 1 EC - raS/ca 20/9/84 24/9/84 25/9/84 29/9/84 1/10/84 1 1 1 AWASH 1 Rl 1 0.3 0.2 0.2 0.2 0.3 1 RIVER 1 1 - 1 R2 1 0.4 - 0.2 0.3 0.4 — 1 1 1 1 1 1 1 R3 0.2 0.2 0.2 0.4 - - 1 1 1 1 R4 0.3 0.2 0.3 - - 1 1 11 1 R5 0.2 0.4 0.2 0.3 - - 11 1 1 MAIN I CANAL 1 1 Cl 1 0.3 0.2 0.2 0.2 0.3 1 1 1 1 1 C2 | 0.3 0.2 11 0.2 0.2 0.3 - 1 1 1 1 1 11 1 1 1 1 1 1 1 I MAIN 1 1 C3 I | 1 1 0.3 0.3 0.2 0.2 - - C4 0.3 - 0.2 0.2 - - 1 C5 1 1 1 C6 1 I C7 1 1 1 0.4 0.2 0.3 0.2 - - 0.4 0.3 0.2 0.2 0.4 0.5 0.3 0.2 - - - - DI I 0.6 0.5 0.4 | DRAIN 1 1 1 1 1 )1 1 1 1 1 11 02 , 0.5 - 1.2 - 0.8 - D3 I 1 i D4 I | 0.4 0.6 1.3 5.5 2.92) • - 0.3 1.6 2.3 2.6 2.3 -------------- n 0.6 2.6 l.l l.i 1 D5 I | - 0.4 0.5 11 0.5 0.8 2.4 1.2 1 1 1 06 | 1 0.6 0.4 0.3 0.4 - 1.0 0.6 1 I pH RANGE - 7.4-8.1 7.6-8.1 7.3-7.6 6.8-7.6 7.9-8.2 7.7-8.1 1 1) Combined reading taken downscream of secondary drain 2, with resultant contamination with flows from Pilot Drainage Project 2) Additional observation from secondary drain 2, with flow from Pilot Scheme project, gave EC value of 5.1, pH = 7.0the Primary Drain outtall ranged from a minimum of 0.3 mS cm I to a maximum of 1.0 mS cm’l. ?H values for water samples taken during the above programme ranged from 6.8 to with no discernible pattern being exhibited between the origins of the various samples. 4 * 2 Projected Salinity of Return Flows to Awash River An analysis of the seasonal salinity balance of drainage return flow and Awash River water is described in Appendix B. The analysis is based on assumed final development of 14,200 ha irrigated cotton with sub-surface. drainage. The results indicate that, with a probable long term groundwater salinity of 10 mS ctd~1 • salinity of Awash River water just downstream of the outfall from Main Drain 2 would reach a seasonal maximum of 0.68 mS cm^l , From the point of view of use by downstream farms chis salinity level falls in the medium salinity hazard range, as defined by USDA. With an extreme groundwater salinity of 40 mS cm~l over Che whole area ot 14200 ha the salinity of downstream Awash River water would fall within the high salinity hazard range for part of the season, reaching a peak of 1.24 mS cm ^• It should be stressed, however, Chat a - groundwater salinity of 40 mS cra-1 f most unlikely Co persist over a s significant proportion of the project area and the analysis using this value lm intended to provide a guide to the upper limit of the effects of drainage water on Awash River salinity levels. 105 IRRIGATION WATER UTILISATION 5.1 Evaluation of Present Recommended Application Amounts During the Phase I Studies, a review was carried out of the crop water requirement a for cotton in which a comparison was made between the recommended irrigation water amounts and those computed on the basis of climatological records for the 1991. 1982 and 1983 irrigation seasons. The application of 714 um was obtained using the method proposed Management Manual (Ref 10) based upon an average climate and exceeded in 3 years out of 4 (the 75Z exceedance rainfall). and 1983 irrigation seasons, estimates of reference crop evaporation were recommended in the Water the rainfall depths For the 1981. 1982 computed by the modified Penman method using the data from Melka Warer climatological station. Irrigation water requirements were produced using crop coefficients for cotton and a representative cropping schedule, and effective rainfalls derived from total rainfalL using the procedure outlined by FAO (Ref I). The results of this comparison showed that recommended net applications significantly exceeded net water requirements in all three seasons, the accumulated excess amounting co 526 mm. The comparison also demonstrated the desirability of deriving crop water requirements by soil moisture accounting, although its implementation within the project area is unlikely within the immediate future. An alternative proposal, that of using the median rather Chao the 75Z exceedance rainfall in the computation of application amounts, was advanced as a means of reducing the estimated excess. Implementing this proposal in the 1981. 1982 and 1983 irrigation seasons was shown to reduce the net application to 618 com and the accumulated 3-year excess co 238 ran; in 1981. there was a small under-application of 25 mm. During the Phase £1 studies, the investigation was extended with trial water balances for both cotton and bananas in order to assess deep percolation Losses resulting from the excess uf actual water applications over estimated net crop water requirements for the same three irrigation seasons. 115.2 Contribution. to Groundwater Recharge o£ Excess Irrigation Applications 5.2.1 Cotton Irrigation For the cotton crop, the trial water balance was carried out for MeLka Sadi Units 1 and 2, which are served by Primary Offtake No 4. The total area of these two units is 634 ha. Appropriate allowances were made for canal seepages, surface runoff and losses from uncultivated areas. Crop water requirements were derived from analysis of the planting schedules for each cropping season which gave appropriate monthly crop coefficients for cotton. Reference crop evaporation was again computed from the records for Melka Warer climatological station, and rainfall data were taken for the same site. Details of these calculations are presented m Appendix A. The estimated contributions to groundwater recharge for all three cropping seasons are summarised in Table 8. These amounts of recharge would only occur where the level of the table is appreciably below ground, such that evaporation from bare soil in the fallow season is negligible. Where water tabLes are Less than 1*0—1-5 m below ground, the evaporation losses during the fallow season could exceed the estimated recharge (see Appendix A). Under these conditions, the net annual recharge to groundwater wouLd effectively be zero and the groundwater table would tend to stabilise about the average annual value, rising during the irrigation season and falling during the fallow. TABLE 8: Contribution of E xcess Irrigation Applications to Groundwa ter Recharge on Melka Sadi Units 1 and 2 for Three Croppi ng Seasons | cropping sea son 1981 1 1982 1 1983 1 Average 1 1 1 I estimated re charge, mm ' 244 1 1 226 1 1 I 208 1 1 1 226 1 1 1 1 . 1 125.2.2 Banana Irrigation A similar trial water balance was carried out on data relating to the Banana Unit for the same three years. Annually cultivated and irrigated areas were based upon the following data provided by .MSaE: Year 1981 1982 1983 Irrigated area (ha) 863 799 769 An extra 70 ha was added co the above totals co allow for irrigated areas under pasture and other crops. Details of the calculations are again presented in Appendix A. The results obtained indicated that the monthly water releases to the Banana Unit have been, with few exceptions, appreciably less than the computed water requirements, with apparent annual net deficits of 373 mm, 200 mm and 82 mm for 1981, 1982 and 1983 respectively. A portion of the 1981 and 1982 deficits was caused by closure or the main canal during March and April, together with zero flow requisitions in August. During 1993, both supply and demand were continuous,, resulting in a markedly Lower deficit. Assuming that information on the extent of irrigated areas to be correct, such deficits can only be considered apparent rather than real since, in the presence of the high water tables which generally prevail throughout the Banana Unit, any deficit would have been reduced appreciably by groundwater abstraction. Conversely, even under conditions where monthly irrigation water requirements exceed rhe supply, all water supplied is unlikely to be available to the crop because of inherent application inefficiencies, and some contribution to the groundwater table would occur. The net annual contribution to groundwater within the Banana Unit would therefore appear to have been negligible during the period 1981-1983 inclusive. Under conditions of no water shortages, in which the net irrigation supplies equal the computed demand, the effective contribution to groundwater could be in the order of 25-30 per cent of gross field application. For mature bananas under project conditions, the gross field application would be some 3000 nm. making the contribution to groundwater between 750 and 900 m per annum. 136 FLOOD ESTIMATION FOR THE EASTERN CATCHMENT AREAS 6.1 The Catchments The catchment areas whose flows collect in Che drain running alongside the eastern flood dyke of the project area are typically ephemeral, with few signs of the well-developed dendritic drainage networks found in spore temperate climates. Some traces of drainage channels may ba found where rock bars and riffles in the local geology have provided stable hydraulic controls, upstream of which the occasional floods have eroded a path into the overlying silts. Elsewhere, flow paths are marked by a broad ribbon of vegetation sustained by infiltration. In some parts of the Alledeghi Plain, deep-rooted grasses have clung to the soil cover while flows have eroded between the clumps to create in uneven surface in direct contrast to Che flatter, poorly-vegetated areas into which no flows have concentrated. The flatness of the Alledeghi Plain makes the watersheds ot individual catchments difficult to delineate. In broad terms, the headwaters of the areas draining south-east towards the project area lie at about 1500 id on the slopes of Mount Aseboc. Within 20 km. their altitude is lowered by some 600 m as they reach the edge of the Alledeghi Plain where slopes are markedly reduced. After a further 20 km, a fault scarp running in a north-east - south-west direction is reached, marking the edge of the Awash River flood plain. The Awash-Tendaho highway running along and parallel to this scarp has interrupted natural drainage lines locally, but is nevertheless well provided with crossing points. To the north-west of the scarp, slopes are slightly increased and flow paths turn northwards before joining the drainage chaoneL alongside the eastern flood dyke. 6. 2 Flood Estimation In the absence of flow measurements from the eastern catchment areas, 10-year and 25-year design floods were estimated using the Richards method. This technique is a modified version of the Rational Method in which ttje time of cone entr at ion of the drainage area is allowed to reduce as the intensity of rainfall increases (see Ref 11). as £n the Rational Method, one of theprincipal parameters upon which the flood estimate depends is the runoff coefficient, values of which were chosen on the basis of inspections of the catchment area, observations of floods subsequent to the building of Che eastern flood dyke and experience of percentage runoffs recorded elsewhere in similar climates. Rainfall intensities for selected durations and return periods were obtained from the rainfall DDF relationship derived from the data for the Melka Warer climatological station. Owing to the nature of the catchments as described in Section 6.1, and the characteristics of heavy rainfall in the project area, which is predominantly of short duration and convective in origin, the whole of the 800 plus km^ of drainage basin is most unlikely to be subjected to a single storm of similar areal extent. Two design conditions were therefore considered: i a contributing catchment consisting of the area to Che south-east of the eastern flood dyke, extending some 6 km into the Alledeghi Plain beyond the highway (317 km^); and ii the catchment of the Sullalaidla Creek, augmented by the area between the Amibara link road, the highway and the eastern flood dyke (590 km^). The results are summarised in Table 9, which shows that, with the particular choice of runoff coefficients, both the above cases yield similar results. TABLE 9: Summary of Flood Estimates for the Eastern Catchment I case 1 return period, y 1 1 runoff coefficient 1 peak flow m^/s | I1 1 10 1 0.6 ) 94 1 1 1 25 i 0.6 1 128 | 1 2 1 10 1 0.45 1 93 | 1 1 25 I 0.45 1 126 |6.3 Des ign Re cu rn Pgri ud The selection of an appropriate design return period for a flood dyke can be based on an optimisation exercise in which the annual equivalent cost of a flood protection dyke designed to resist a particular flood return period is compared to the estimated loss in revenue or in damage caused by the dykes being breached by such floods. Such exercises have been carried out in the earlier assessments of the appropriate design return period for the Awash flood dykes (Halcrovr 1975 Ref 5 and HEDECO 1982 Ref &)* The conclusions of these assessments are of interest in that the results of both indicated chat low return period values could be used in rhe design. In the Halcrow 1975 analysis Che most economic dyke design related to a 7 year return period flood. Similarly in the Nedeco 1962 analysis no clear optimum value was obtained although the results presented would tend cowards a return period of less than 10 years. In ooth instances, tn view of the very limited accuracy of inch an approach where crop loss/damage estimates are hypothetical and at best approximate, a conservative interpretation was taken, Leading to the adoption of a 1 in 20 year return period design flood. Based on the results obtained from the analysis related to the Awash dykes it is Likely that a similar analysis for the eastern catchment dyke, were adequate data available, would indicated even lesser return periods in that the flood hazard is far less as the drainage system has large channel storage and flood absorption capability. This latter point is evidenced by the limited damage which occurred from actual overtopping of the existing eastern dyke in 1931 . However, such an analysis is not warranted as insufficient actual data exists to reliably predict the Magnitude or behaviour of flood flows or Co aasess what damage might be induced. At the other end of the spectrum for infield agricultural drainage a storm return period of 2 to 5 years is generally appropriate for design purposes; less if Low value crops are grown (1 year) and mure with specialised high value horticultural crops (10 years). (Ref 12 and 13). 16Within the above range a 10 year design return period is recommended by the consultants to be appropriate for the eastern catchment dyke, with the additional proviso that the freeboard allowance should be sufficient to accommodate a 25 year frequency flood without overtopping. The corresponding discharges are 94 m^/s and 130 m^/s. The recommended freeboard is 0.5 metre above the design flood level. The 25 year return period flow increases the flow depth by 0.24 metres which is well within the freeboard allowance. 177 FLOOD ESTIMATION WITHIN THE PROJECT AREA In estimating the allowance for drainage of tha runoff from alarm rainfall within Che irrigated area, NEDECO (Ref 6) employed a aeries of 'maximum' rainfall depths for different durations quoted from Halcrow (Ref 5). The 2~'h figure of 40 mm was taken to drain in 24 h with an assumed runoff coefficient of 0.4 to give a drainage duty of 1.95 l/s/ha. Applying the rainfall DDE relationship derived for Melka Warer (see Section 3*2) shows Chat this am xi aura rainfall depth of 40 mm in 2h has a return period of approximately 20 years. Moreover, the runoff coefficient of 0.4 may be regarded as too high bearing in mind both the very mild ground slopes and the in-field bunding practices within the project area. Repeating the calculation using the 2-h, 5-year rainfall depth of 28.8 sen and a runoff coefficient of 0.2 gives a drainage duty of 0.67 1/s/ha. The NEDECO estimates may also be compared with chose of Halcrow (Ref 5), who assumed 10 mtn of runoff from the 24-h, 5-year rainfall depth occurring over 24 h. Using these figures, a drainage duty of 1.16 I/s/ha was obtained. However, a 30-h drainage time can be regarded as a more reasonable allowance for a 24-h rainfall total, which reduces rhe required duty to 0.93 l/s/ha. During the Phase I Studies, preliminary estimates were made of the peak flow rates that would be generated at both the outfall of the present system on the Awash River and the link road bridge from a 5—y storm. A duration of rainfall equal Co the time of concentration of each catchment area was assumed along with a runoff coefficient of 0.2., and storm hydrographs ware construct cd using time-area diagrams. Allowance for the attenuation by storage in the drainage system was made by Assuming the channel network acted as a reservoir with the water surface parallel to the bed sLopes of each individual channel length. Althougn these estimates were undertaken primarily to check the available treeboard of the existing channels, the results also provide some guidance on the required magnitude of the drainage duties. The estimated pea* discharge of 10.65 n.3/s on the irea O f 3353 ha gives 1. 7 I/ /ha. However, this figure may 23 b< regarded as too high by at least 20 per cent because of the inherent underestimation of attenuation provided by the approximate routing procedure adapt ed. 18Taking into account the above results and the variety of assumptions upon which they are based, an allowance of 1.0 1/s/ha as the drainage duty for the project area is recommended. 192 2 8 FLOOD ESTIMATION FOR THE AWASH RIVER 8.] Estimation of Peak Flow Rates At the upstream end of the project area. the Awash River has a catchment area of some 21,500 km , over half of which is tributary co Koka Dam and therefore subject to regulation. Two tributaries. the Kessem and the Kebena, enter the Awash between Melka Sadi and Melka War er, adding a further 43fl0 km of area. The design gf flood protection works for the project area must therefore take into account the behaviour and interaction between all three rivers. Table 1 shows that the Awash has been gauged upstream at Awash Station since 1962. and that records for the Kessem and Kebena began in the same year. Observations of annual floods were available for Awash Station for the years 1963-1982 inclusive. However, for the Kessem River at Awarn Melka, no data were available for 1965. 1969, 1972 and 1975, and for the Kebena River at Kebena no annual instantaneous maxima ware provided for 8 of the 21 years from 1962-1982. The possibility that extreme floods might occur simultaneously on all three rivers is of considerable importance when considering the design of embankments to protect the Melka Sadi portion of the project area, a correlation analysis was therefore carried out on the annual floods from all three gauging stations, and Che correlation matrix obtained is presented in Table 10. This table shown that the correlation between sites is very low and that inverse reTat ionships were indicated in all cases, ie low ranking floods on one river tend to coincide with higher-ranking floods on another, and vice vers*. Examination of the original recorder charts for all three sites has revealed the 'flashy* nature of the response: of each river. High flows on the Kebena in particular are characterised by short times of rise and rapid recessions. Coincidence of flood peaks could only arise from storm conditions that affected the Awash Valley sufficiently to induce high flows prior to moving over the smaller tributary catchments. Examination of the dates of the annual maxima (where available) shows only two occasions where the Kessem and Kebena floods coincided, and no instances of the Awash flood occurring simultaneously with the other two during the 20 years 1963-1982. Coincidence ot flood peaks can therefore be discounted as 4 basis for establishing design standards appropriate to the Aimibara. area.. 20TABLE 10: Correlation Matrix of th*> Annual Floods for the Awash River at Awash Station, the Kessem River at Avar a Melka and rhe Kebetia River at Kebena. 1 1 1 1 i Awash I Kessem Kebena 1 1 1 I 1 1 1 1 1.00 1 1 11 i Awash 11 11 tI 11 1 1 1 1 I Kessem -0.09 (161* 1 1.00 1 1 1 1 11 1 Kebena 1 1 -0.L8 (12)* 1 -1.10 (0) 1 1 1.00 | 1 I * Numbers in brackets indicate that number of pairs of points used in computing the correlation coefficient. Prior to analysing the available records of annual floods, the data were examined for evidence of non-stationary effects, such as trends. For the Kessem River at Awara Melka, both the mean and standard deviation of the instantaneous maxima were Lower during the period 1973-1982 than during the previous decade, although the changes were not significant at the 5 per cent confidence level. However, for the Awash River at Awash Station, the opposite tendency was observed over the same two periods, with a 29 I increase in the mean and a 175 2 increase in the standard deviation. The latter change is highly significant in the statistical sense, but its causes are not readily identifiable. a similar analysis was carried out on the same two decades of observations for the Awash River at Koka Dam. At the latter site, the mean and standard deviation of Che daily maximum flows increased by 56 and 114 Z respectively. Part of the increased variability at Awash Station may therefore be ascribed to the changing pattern of releases from Kaka Dam, although in terms of annual runoff volumes, the amount of variability has remained remarkably consistent. The changes in the annual floods at Awash Station are most probably associated principally with Land use changes within the catchment area, but insufficient data are available 21tn confirm this possibility. Further examination of the Awash Station records shoved that as the period of observations has increased both mean and standard deviation have grown in magnitude, thereby accounting ror the increase in the size of the design floods provided in successive studies (see Refs 5 and 6). Ln the absence of further information with which co correct the available annual flood series, or at least to isolate that portion of the records that might be regarded as rapresentative, the total period of observations for Awash Station was analysed along with those for Awara Melk.i and Kebena. A variety of statistical distributions was fitted to the available data using both the method of moments and the method of maximum likelihood. An extreme value Type 1 distribution was adopted for both the Kes sew and Che Kebena and an extreme value Type 2 distribution was chosen for Awash Station on the basis or graphical plots and values of the Ko 1 mogorov-Smirnov statistic* The results are sunaaarised in Table 11 in the form of summaries of floods of a given recurrence interval and growth factors, ie ratios of the higher return period floods to the mean annual flood. As might be anticipated from the use of different distributions, the growth curve for Awash Station is much steeper than those for the ocher two sites. Moreover there is no obvious relationship between the mean annuaL flood and size of catchment area, even when allowances are made for the portion of the Awash River basin regulated by Kaka Dam. TABLE 11: Summary of Flood Estimates and Growth Factors for Selected Return Periods for Awash Station, Avara Melka and Kebana | Return I Period 1 AWASHSTATION | KESSEM 1 KL3ENA 1 Flood Factor | Flood Factor Flood Factor 1 I 2.33 1 1 365 1 1 850 1 ’ 320 1 5 1 561 1.545 1 1170 1.376 1 478 1.494 1 io 1 757 2.005 1 1393 1.639 1 587 1.834 1 20 ’ 1012 2.788 [ 1606 1.889 1 691 2.159 1 50 ! 1465 4.036 | 1882 2.214 1 827 2.584 1 100 1 1939 5.342 | 2088 2.456 1 928 2.900 229.2 Construction of Design Hydrographs Estimates of the heights of the embankments required to protect Che project are from Awash River flooding have been provided in two previous studies. Halcrow (Ret 5) adapted a 20-year design standard and proposed a design flood consisting of the sum of the 20-year floods at Awash Station, Avars Melka and Kebena less 202. A triangular hydrograph shape was assumed with a time of rise of 30h and a time base of 72h, NEDECO (Ref 6) also adopted a 20-year design standard and estimated the flood peak from the sum of the Awara Melka and Kebena flows plus a 200 m^/s baseflow. Their design hydrograph had a 2-h time of rise and a 20-h time base. Using the results presented in Table 11, these design floods would 3 3 amount to 2647 m /s (Halcrow) and 2497 m /s (MEDECO). However, both estimates are based upon the assumption of coincidence of flood peaks, which Table 10 shows to be highly unlikely. In addition* the two estimates differ markedly in hydrograph shape and therefore in runoff volume (343 MCM - Halcrow; 90 MCM-NEDECO). Two issues must therefore be addressed in constructing the design hydrograph: 1) determination of typical hydrograph shapes, and therefore times of rise and time bases of representative floods; and 2) estimation of the peak flow race immediately upstream of the project area. Subsequent to the studies referred to above, a gauging station has been commissioned at Melka Sadi, immediately upstream of the project area headworks (see Table 1), The available records from this site were examined and three major events, which occurred on 19 February, 24 April and 19-20 May, 1983. were abstracted, The ordinates of each hydrograph were expressed as a proportion of the recorded hydrograph peak, and the results superimposed with their peaks coinciding. An envelope curve was then drawn through the plotced points with greatesr weight being given to the event with the largest peak, which occurred in May, 1983. The results presented in Figure 2 were taken co provide the required representative flood flow hydrograph, and has a time of rise of 12h and a time base of 40h . 239 HYDRAULIC MODEL STUDY OF THE AWASH RIVER 9.1 General Flood flows in she River Awash have been analysed using Halcrow's open channel flow model ONDA which is based upon die Preissman finite difference box operator (Refs 14 and 15). This operator may be used for solving Che St Venant open channel flow equations. 9.2 Extent of Present Model A L :5QOOQ plan of the reach of the Awash River that has been model led, is shown on Figure 3. the river and the location of existing and proposed flood dykes on the flood plains are shown. Also marked are the river-flood plain cross sections used in the numerical model. These cross sections have been obtained from the Awash Valley Survey. The upstream end of the model has been fixed at Awash Valley Survey chainage 745.8 km. This point is just downstream of the confluence with the Kebena River. The downstream end of the model has been fixed at chainage 642.0 ksi which is to rhe north-west of the Dijilu Irrigation Area. Survey information for Che flood plains, particularly on the western side of the Awash River, is limited but is nevertheless sufficient to provide general flood plain ground levels for the hydraulic model. Reference has also been made co the Awash Valley Authority's 1:20000 mapping. 9.3 Flood Hydrograph Inputs The Awash River has been analysed hydraulically in two earlier studies. The first of these was carried out by Halcrow (Ref 5) in 1975 and the second by NEDECO (Ref 6) in 1982. am discussed in Section 8, estimation of peak flood flows in earlier work was based upon the assumption of coincidence of flood peaks from.the Awash, Kess«o and Kebena rivers. The data available indicates this situation to be highly unlikely and therefore in this study, the Awash Station flows alone have been used to provide estimates of design flood peaks. 24The I Ln 10, 1 in 20 and I in 50-year flow hydrographs that have been adopted for flood routing are shown in Figure 4. Hal crow also carried out some computational flood routing in June, 1979. However, Che downstream end of the numerical model has fixed at chainage 664 km, Ln addition, lateral spillage was modelled in a natural floodway that was originally proposed on the western side of the river between chainages 694 km and 691.4 km. In practice, the continuous dykes were constructed on both sides of the river in this location and the floodway blocked off. 9.4 Hydraulic Properties Adopted for the Flood Routing l A Hanning’s *n' coefficient of 0.040 ( /n ’ 25) was considered suitable for the river channel. This value is typical of a natural channel, with irregular side slopes and some variation in cross section. For the flood plain a Manning's 1n’ of 0.150 has been adopted. This figure assumes an irregular ground surfaces with many roots, bushes and trees impeding but not arresting flood plain flow* Composite crass sections across both river and flood pLains have been used in the model. Allowance has been made for Che greater hydraulic conveyance of the river channel, as compared to the flood plain, by dividing river channel and flood plain into separate hydraulic 'panels’. The sinuousity of the river within i’ts flood plain has also been taken into account by adjusting the flood plain Manning coefficients using a function of the relative path lengths of fLood plain flow, beLween adjacent, cross sections divided by river channel path Iength. 9.5 Testing at the Model 3 An in-bank flow of approximately 300 m /s can be inferred from the flows presented in Table 11. This flow has an associated return period of between 1*5 and 2 years. The base flows associated with the 1 in 10. I in 20 and 1 in 3 3 50-year hydrographs are 22? m /s 304 m /s and 440 m /s respectively. It was 3 t found that predicted water levels for a steady flow of 304 m 3/S coincided very well with the bank top levels. Predicted water levels for this flow have been 25shown on the longitudinal profile of the Awash River in Figure 5. This result confirms the selection of a Manning’s coefficient of 0*04 as appropriate for the river channel. 9 * 6 Est imates of Flood Levels The maximum stages associated with the I Ln 10, 1 in 20 and 1 in 50—year floods along the Awash River are shown on Figure 5. These maximum: stages are bounded by a predicted upper and Lower stage relationship for two selected steady flows* The lower stage, which is predominantly within bank, is for a flow ot 227 m^.'s, which is the base flow of the I in 10-year hydrograph. The upper bound stage is for a steady flow of 1465 m^/s, which is the peak flow of the 1 in 50-year hydrograph* The upper bound stage is generally up to 3 m above the bank top levels of the river in its upper reaches. The three flood hydrographs themsaLves are subject to substantial attenuation. Maximun predicted flows at selected points along the river for each of the three floods are also given in Table 12. As a result of this flood wave attenuation, the predicted maxi mum flood stages do not rise so far above banktop at the downstream end of the model as they do at the upstream end. Typically, 3 m high dykes are required to contain the 1 in 20 year design flood at the upstream end of the existing Amibara development near Melka Sadi. At the northern end of the Amibara development the dyke height requirement is reduced to 2.0 m above river bank level. At the downstream end, dykes to protect Dijilu and Angelela need only be 1.0 m high (excluding freeboard allowances). The dyke heights given above are only indicative of general requirements for flood protection and actual heights will be subject to local topography. The required dyke heights are very dependent on the assumed flood pathway width. In general, proposed dykes have been located approximately not less than 300 m apart* Should the distance between dykes be substantially changed, it would be necessary to re-estimate flood levels. z 26Ideally, the design flood peaks required to make the typical hydrograph of Figure 2 dimensional should have been based upon the Helka Sadi data. However, as Table 1 shows, records at this site are too short-for frequency analysis. Since the available records at Awash Station, Awara Melka and Kebena did not yield a basis for consistent regiona1isation in terms of growth factors and a relationship between mean annual flood and catchment area, alternative design assumptions were required. Bearing in mind the independence of these three sites, as indicated in Table 10, flood estimates should preferably be based on a single site, and Awash Station* being upstream of Melka Sadi on the Awash River, was chosen for this purpose. Since the catchment area increases by 2590 km^ between Awash Station and Melka Sadi, some enhancement of flows recorded at the former site might be expected. However, examination of major floods at both sites provided evidence of attenuation rather than increase in peak discharges. The Awash Station flows were therefore employed without change to provide estimates of design flood peaks. The 20-year flood of 1012 ra-^/s shown in Table 11 was used to scale up typical hydrograph shown in Figure 2 and provided the design input to the the computational hydraulic model of the Awash River described below in Section 9. 2"TABLE 12 Maxi in hue Predicted Flowti tor I in 10, 1 in 2U and 1 in 50 - year Floods for Selected Points along Lhe Awash River I I MAXIMUM 1 PHHDlLTEi) 1 FLOWS (m /s) 1 CHA IMAGE (Kilometres) 1 3 1 I 748.8 1 725.7 701.0 681.0 1 1 666.0 | 11 650.0 1 | 1/lU Year 1 Flood 1 1 1 1 757.0 1 “i ------ 376.1 338.9 334.3 1 1 324.0 1 315*0 f 1 1/20 Year I Flood 1 1------ 1 1 1012.0 T1 505,6 1 438.6 419.3 1 1 1 409.7 1 399. J 1 1 r 1 1/50 Year I Flood 1 1 I 146b.0 1 1 776.3 603.0 646.1 1 1 1 624.2 1 r 1 608.2 1 r-j IDREFERENCES 1 Doorenbos, J and Pruitt, W Or Guidelines for Predicting Crop Water Requirements. FAO Irrigation and Drainage Paper 24, 144 pp, 1977. 2 Valleys Agricultural Development Authority, Awash River and Lake Basins Hydrometeorologic Summary, 1976, 110 pp, 1977. 3 Valleys Agricultural Development Authority, Awash River and Lake Basins Hydrometeorologic Summary, 1977, 104 pp, 197S. ft 5 6 7 a Water Resources Development Authority, Awash River and Lake Basins Hydrological Year Book. 1978, 102 pp, 1979. Sir William HaLcrow & Partners, Angelele and Bolhamo Feasibility Study Report, Awash Valley Authority, October, 1975. NEDECO, Angelele-Bolharao and Amibara Irrigation Expansion Project - Reappraisal, Water Resources Development Authority, June, 1982. Ltaleonsult, Melka Sadi-Ami.bara Proposed Irrigation Project, Feasibility Study, Awash Valley Authority, July, 1969. I taiconsult, Additional Soil Study, Awash Valley Authority, September, 1971, 9 Building Research Establishment, Concrete in Sulphate Bearing Soils and Groundwaters, BRE Digest 250 , June, 1981. 10 Sir William Halctow 4 Partners, Water Management Manual, Amibara Irrigation Project, 1983* 11 Richards, B D, Flood Estimation and Control, Chapman A Hall, 3rd Edn, 187 pp. 1955.12 Smedoma L K and Rycroft D W, Land Drainage, Bats ford Academe, London, 1983. 13 Castle D A, McCunnall J, Tring I M, Field Drainage, Batsford Academic , London, 1984. 14 Preissmann, A, Propogation des intumescences dams les canaux et rivieres, Lire Congres de 1’Association Francaisc de Calcul, Grenoble, I960. 15 Cunge, J A, Applied Mathematical Modelling of Open Channel Flow, Ch 10. Unsteady Flor in Open Channels, Water Resources Publications, Fort Collins Colorado, 1975. anappendix a TRIAL WATER BALANCES FOR COTTON AND BANANAS A* 1 Introducr ion During the Phase LI Studies, trial water balances were computed for cotton using iniorwation tor Meika Sadi Units I and 2 and hr bananas based upon data for the plantation at Melka Sadi- The resuLts obtained from these calculations, which were carried out for the L981, 1982 and 1983 seasons, are discussed in in Section 5*2* In this AppendLX, details of the assumptions made in computing these water balances are iuzirra.ir ised cLofluj with tabulations of their conspcnen ts . Losses in the conveyance system are discussed iu Section A* 2, and the water balances tor cotton and bananas are described separately in Sections A-3 and A* A respectively . A. 2 Losses in the Conveyance System A. 2-1 Seepage Losses Field trials on seepage Losses from ponded reaches ot tertiary canals indicated 2 figures of the order of 0.25 1/a/1000 m of wetted perimeter* Values of the latter for secondary and tertiary canals were estimated to be as shown in Table A-l- TABLE A. 1 Estimates at Wetted Permecer for Secondary and Tertiary Canals in Malka Sadi Units 1 and 2 1 can* 1 1 i length, 1 welted perimeter 1 m , a2/m I m- 1 S4 I LOGO i 4*3 | 4300 1 1 14/1 | 3240 .3.0 I 9720 1 T4/2 t 2200 I 3.8 8360 1 | 14/3 1 3900 1 3.8 1 1 14820 1 1The above total wetted perimeter leads to an average seepage loss of 9,3 1/s. Given the intermittent usage of the field canals, the average seepage loss was assumed to be double that measured in the tertLary canals, ie 0.50 l/s/1000 Using an average field canal length of 50 m/'na and a totaL canal length of 2 31700 rn gives an approximate wetted perimeter of 63000 m . Assuming a 25 per cent utilisation of field canals leads to a seepage loss of 0,25 x 63 x 0.5 - 7.9 L/s. A.2.2 Evaporation Losses from Canals The secondary and tertiary canals have an average top width of 3.4 m, giving a 2 totat area contributing to evaporation of some 35000 qi , On the field canals, the total area for a 2,0 m top width and 25 per cent utilisation is 2 approximately 16000 tn . Assuming an average daily evaporation loss of 6 mm, the equivalent continuous loss is found to be 3.5 l/$, A.2.3 Rejection Losses Based upon limited spot observations and measurements of drainage discharge in the Primary Drain, 10 per cent rejection was assumed for the secondary and tertiary canals and 2.5 per cent from fields (including irrigation runoff). A. 2.4 Evaporation Losses from Uncultivated Areas Assuming 15 per cent of the gross area lies outside the field boundaries, the total gross area ot Units I and 2 is 746 ha. With an average net evaporation of 2 rnrn/d from bare soil, headLands, canal banks etc outside the project area the equivalent loss on the 112 ha area is some 26 l/s. A.3 Water Balance for Cotton The amount of water delivered to the fields in Melka Sadi Units 1 and 2 during the 1981, 1982 and 1983 irrigation seasons is sunmarised in Table A.2. Using the planting schedules shown in Table A.3 and average monthly crop coefficients from Che Water Management Manual (Section 4.7.2), weighted monthly crop 32coefficients were derived. The c1imatological data Erora the Helka Water Research Station were used to estimate reference crop evaporation by the Pentium method, and these figures together with the recorded monthly rainfall totals provided the basis for the computation of irrigation water requirements as shown in Tables A.4 - A. 6. Detailed analysis of the planting schedules as given in Al? record* showed that in 1931 all field* with the exception of 5A (17.5 ha) received a pre-planting irrigacian. In the absence of recorded data on this irrigation^ rho recuuHaandcd gross amount of 286 nun (220 ton net) was assumed* Ln 1982 » a pre-planting irrigation was applied to only 19 per cent of the area, and in 198 3, only 49 per cent was treated. Average gross applications of 54 w and 140 *20 were therefore assumed in these two years. Any pre-planting irrigation was taken effectively to recharge soil moisture up to field capacity with negligible contribution to deep percolation. Following the final irrigation, the crop will continue to transpire at a reduced rate until removed. The residual soil moisture on removal of che crop was □ ssuHied to be half the available soil moisture, ie 110 non in che top metre of soil. The results from Tables A.2 - A. 6 are drawn together in Table a.7 which summarises contributions to deep percolation for all three irrigation seasons. A.4 Water Balances for Bananas The total volumes of water delivered to the banana plantation from Primary Offtake No 1 and estimates of the equivalent depths of water retained on the fields for rhe three years 1981-1933 are shown in Table A.B. Monthly water balances for the same three years are presented in Tables A.9 - A.11. Reference crop evaporation was again computed by the Penman Method using data from Melka Water Research Station. Effective rainfall was estimated from monthly rainfall and evaporation totals using US Department of Agriculture methods. The apparent excess applications, which overall amounts to annual deficits are Listed in the List row of each Cable. 33TABLE A.2 Water Delivery to Fields on Helka Sadi Units 1 and 2 from Primary Offtake No 4 during 1981, 1982 and 1983 Irrigation Seasons 1 1 19a i 1982 1983 | 1 1 1 1 r 1 1 1 I Total water delivery at primary offtake, MCM 1 1 I Equivalent Continuous Flow Over 8 Months^ 1/s Seepage and Evaporation Loss in Canals, 1/s 10 Per Cent Rejection Flow from Canals, 1/s 1 T" 1 10.432 1 i" 7.736 i 1 1 9.081 I 496 368 1 432 1 ! -21 1 -21 1 -21 1 -50 1 -37 1 -43 I | 1 1 I Gross Water Applied to Fields, 1/s 1 1 1 425 310 “1 " 1 1 368 1 I 2.5 Per Cent Rejection & Irrigation Run-off, 1/s 1 -11 - 8 1 - 9 1 1 1 1 1 I Gross Water Retainted on Fields, l/s Equivalent Depth Retained on Fields, imn 1 1 1 414 1 1372 i “I 302 1 359 1001 1 1190 1 TAELE A-3 Planting Schedules for Melka Sadi Units L and 2 tor the 1981 t 1982 and 1983 Irrigation Seasons 1 1 AREA PLANTED, ha 1 1 PERIOD 1 1 1 r 1 1981 1 1 1982 1 1983 | 11 I 1 1 5-27 May 1 1 - 1 1 - r i 1 242 I 15 Hay-15 June 1 1 | 14 June-9 July I 15 June-15 July 1 1 1 L6 July-15 Aug I 255 1 1 390 1 1 - 1 F" 1 - 1 386 1 288 237 1 [1 1 1 1 - 1 1 1 1 1 84 ------ 1-------------- 1 - 1 j 1---------------1 1 ' 1 1 1 1 34Table; A.4 Monthly SeL Irrigation Water Requirement* for Melka Sadi Unita 1 and 2 during 1981 T 1 JUN r r------ MONTH Penman ET , -nn O Weighted Kc Cotton ET , flram 1 JUL 1 AUG SEP 1 OCT i----------- 1— 1 NOV ; DEC i 1 279 1 ' 229 1 211 1 136 1 210 1 i 189 1 167 J ' TOTAL [ [ r 1471 C RainfalL P , mm Effective rf PA, mni Net [rrlg Req, mm I 0.24 1 0.68 I 0.99 | 1.01 I 0.85 1 0.52 I 0. 18 1 - 1 67 1 156 1 209 1 188 1 179 1 98 1 30 ! 927 I 1 1’ 120 1 131 1 28 1 6 1 o 1 o 1 236 1 0 1 09 I 107 1 23 1 5 ,! o 1 0 224 1 67 1 1 ! 67 [ 102 I 1 165 174 98 1 30 1 1 1 1 1 703 i TAB LE A. 5 Monthly Het Irrigation Water Requirements for Melka Sadi Units 1 and 2 during L932 1 1 1 1 1 1----------- 1■ MONTH 1 JUN 1 1 JUL 1 1 AUG 1 1 SEP 1 , OCT I NOV 1 1 1 1 DEC 1 1 TOTAL 1 1 1 261 1 1 238 rt r | 188 1 1 165 1 Penman £T , ram q 1 172 1 149 1 1 149 1 1 1322 Weighted Kc J 0.38 1 0.84 1 1.06 1 0.99 1 0.78 I 0.65 0.25 Cotton ETC, nun i~ 99 1 200 1 182 ! 186 1 129 1 97 1 37 1 930 Rainfall P, inm ]1 1 46 1 226 1 30 1 106 i 65 2 i 476 Effective rf P ram S1 1 0 1 41 1 165 1 26 ! 73 1 42 1 1 1 348 Net Irrig Req. 1 99 1 159 1 17 1 160 1 56 I 1 55 1 36 582 TABLE A.6 Monthly Het Irrigation Water Requirements for Melka Sadi Units 1 and 2 during ISfltj 1 MONTH Penman £1^, urn Weighted Kc Cotton ETCr no Rainfall P mm 1 MAY 1 JUN 1 JUL | AUG “n------------- r 1 sep : 1 l ------ r 1 T~ l 242 1 1 1 [ 1 248 | Ld4 OCT I 1 1 NOV [ 1 DEC 1 134 0. 20f 0.41 / 0.851 1 . 071 0.981 0. 771 □ .401 0.40 1 27 1 99 1 211 197 1 191 1 149 I 66 I 30 1 22 I 16 1 108 | 149 1 24 | 12 ! 0 1 n 13 1 15 1 88 1 116 20 ! 10 1 0 o ~T----------- r 1 195 193 | “T 165 1 75 t Effective rf Pe>on Net Lrrig Req, mm 1 9 1 84 1 123 81 1 171 J 1J9 66 | 30 I total; I- 1 [ 9 70 I I 333 ! I 267 I I 703 I 35TABLE A. 7 Swannacy of Contributiong to Groundwater Kachar^e from Melkn Sadi Units land 2 during the 19811982 and 1983 Irrigation Seasons 1 i 1981 1 1 1 1 1372 J -702 1982 1 ' 1 Ln-Field Contribution [ Gross Water Retained on Field, mm [ Crop Water Use, mm 1 Pre-Lrrigation Amount, mm I Residual Soil Moisture, mm 1001 1 1983 I 1 ------------------ 1 1 1 1190 -582 1 -703 1 -314 1 - 60 1 -154 | 1 -110 -110 1 -110 1 1 1 i 1 1 I Excess Available for Deep Percolation, mm 1 245“ I ii 1 1 249 1 1 22T“ 1 1 1 Other Contributions 1 1 I 1----------------- 1 1 1 1 Canal Seepages, mm f 57 57 1 57 I Evaporation Losses Outside Cultivated Area, mm 1 1 -86 1 1 -86 1 1 -86 1 1 I Excess Available for Deep Percolation, mm -29 -29 1 1 -29 1 1 I Net Contribution to Groundwater, ™ [ 216 220 1 1 1 194 1 1 36TP,BLl a. 8 Annual Water Balance for Banana plantation at M«llca Sadi for che fearJ 1981-1933 Illi 1981 I 1982 I 1983 1 1 1 1 1 1 1 Total water delivery at primary offtake, MCX 19.830 1 16.199 ; 19.189 ; 20 per cent seepage, evaporation and rejection 1 3.966 1 3.280 1 3.838 1 1 losses I I I 1 Gross Water Applied to Fields, MCM 15.864 | 1 13.119 1 1 1 1r r 15.351 1 1 3 I 2.5 per cent rejection & irrigation runoft, MCM 1 0.39 7 I 0.328 I 1 I Equivalent deprh retained on field*, rain 1i 1 0.384 I I Gross water retained on fields, NCM 1 15.467 I 12.791 I 14.967 I ' Total irrigated area, ha 933 I 869 339 | 1 1 I 1658 . 1472 I 17341AHL> A.9 Hunthh X Irrigation Matar Reouir—ntn fur Banana Plantation lit MaUa 'during lfrfi) 1 □Cl 1 J q i JAM _____ r-------- i *<F 1 11 -J < B FX *fl sfl i“ U*1 2J P, ■fl1 1 <r </* N Sm 03 J-’i 3i —1 c 3 j- a a* a « \3 i 5 "• CM T’- CM ** SJ -O <1 EM CM -F 1“ r» Q -?' -■'^ 3 aoaxo ■hQ -■? 2X O B sf S3 » Ch cz u-i■ — <T Fd PM —■ — — — I 03 a* o <r a o -<» co -0 « -M —■ -—■ 3* idrt ««■ •> 2 A"-. s c* O E5 S3 r-4 £ 4= 3 m S8 — —1 1? o d fl ££ 5 n* fl a Mrt as a —■ S3 r-*> e> r-J w*> - c —» wi r- s Fm 9 X e a a a ?« <! ■s, ff* £ X d r- C « — “ r-. m XFwf-r-A — 1 -- s- w g « F* a* -• a -p —i j-i — « s^ = i •o S E L* «x C> ?* O *■» <j n -* “■- tS fS ?■ — Of •■* V 1—i. a 9* —• ‘fl <3 rH — FJ —■ *m o ■* «M F* « — r-, flr FM 1 p s —r — a-* « « st fl fl fl S m pm ih r» — ’ * —* •"> ST K i* s aS s -FWA 0k P -Rtf * M r a’i=2i a r- F- Fs i i *1 *» 3? s CE, 4 1- — fl “* p* « •», ** — — r* —r £ °£ ■ =a 1 s -* a* a> — r* rs ™ J * = F i 8 ?■?.==? i ’? s t 3 ’3-*S = •* --S w & 2 d5t»« •* F. c —• t 3 Jf Sf p* a » a a > w• r ■v S fl' L- - ; ___________ a *? 1 '■ « > • ,1 J 11 *u- S f =i S *- lL F* X U.I 13 UJ * "* U —1 id c T- ffl -J B- H|'c £" C ”* C -* M-r •--* £ £J 1 T> %. » 1. > d K uj z a h I1. il V D -r — H — < d lll& *- ft *• * i- * i— Xjl — 5 « X 5 Pp’E 4U J~i l’ ii Q V- ?i 21 is l I.—J 1 | D30 1 1_ —J n AWf 1 ._____ 1i-j I I ZB- tsh CVT 1 HI -1 1 Ct 1 62 cn £9- nr wt 91 6Z" 70 /zr 1 T“ 1 611- i 99' ■ •- Vi 6“ r -T 1 99 0T 969 m 1 Ml 1 m 0! 0 01 n srz 9fl 1 99T OT n 9/1 <61 19t i X 1 {ft! ■n J . . .J w ^twajr] fi*r«dE?y I •» "'***V I —__________ _- ■ -_______ J IB * f U&T JJ | i ? tz (Hl til! M- 907 i KZ Z(1 191 1 W htt ftfH cm nri 96 0 /6*0 OH 561 tpi F»Z L 1 9| 1 * tf zn 1 OT 1 M 1 f<r K K I o I Z9 7------ F ImI I WI r I wrt~4 | Ini I I cm j OT1 1 IMVEJ If ro 6Zf Z 54T ( U- I 97 1 i one 1 otz 6« ■w S9T Z9Z 1 1 r« •T TTJ zot 66'0 WZ ZOZ f TT’#*TMI Binsajn r w Jl ’’ll ’*-»J I | f «• IWICIJ 30 | AIR J ZD LI dJS mv HIT _______ MY 1 MV HW wt 03J J L J________ ■J r mt I fwi l i rm i I1 Iwi I J. _________ _________________ ...___________________ J I HjM> | _________________________________ J _ ____ ___________________ ____ g _____ i* <J0 T1 <1 1*14 JDJ «)rjav9J1Fi6wH JW» UO1>*ftT||T ]■! T * ‘ t 11311■ ■ ■JAPPENDIX 3 PROJECTED SALINITY OF DRAINAGE 'MATERAPPENDIX 2 PROJECTED SALINITY 0? DRAINAGE WATER An estimation of the quality of return fid* to the Awash River, once subsurface drainage throughout the entire project is comnlete, is Important in the context of salinity hazards for downstream users, A seasonal salinity balance has been prepared for drainage return flows into the Awash River just downstream of the proposed Main Drain 2 outfall. It assumes full development of Irrigation and sub-surface drainage over a net area of 14,200 ha ±nri full cotton production. The components end results of the calculations are shown in fable Bl. The balance is based on the irrigation applications and water table movements illustrated in Figures 13 and 14 of Annex 3 (Ground^ater and Salinity). Each irrigation application over the entire project area is assumed to take 60 days, following an assumed 60 day planting period commencing Ln mid-May. On a total project area basis there is a differential overlap Ln irrigations dependent on the irrigation interval, The drainage resulting from each irrigation Is then, over the entire project arsa, Assumed to take place over a period of 60 days plus the number of days required for drainage from the final application at ar. assumed discharge af 2.5mm per day. For irrigation number 2 in vertisol the depth oT drainage water is calculated from the rise of water table above drain depth at the estimated effective porosity of 7%. For irrigation number 3 In alluvium the depth of drainage water is assumed to be the full depth resulting fror. the irrigation, as this assumption gives more conservative results than if residual salts are aas umod to be removed by the following (large) irrigation 40 days later. The depth of drainage from all other irrigations is assumed to be 25% of gross field irrigation application (with 5% assumed to run off the field and 70% used by the crop). This simplifying assumption iS slightly conservative, as it under-estimates the volume of water in wtilch leached salts are removed - hence slightly over-estimating the salinity af the sub-surface drainage water»The drainage from irrigations numbers 3 (Alluvium) and 2 (Vertisol,' is assumed to remove all salts which accumulate during the fallow period and are flushed down to the groundwater during the first irrigations of each season. The factors described above enable calculation of the unit discharges shown in Table 01, over periods determined by the overlaps of successive irrigation and drainage periods. The period flow volumes for the project area are determined from the unit discharges with the soils of the total net area of 14,200ha estimated to comprise 80% alluvium and 20% vertisols. Surface drainage volumes at the outfall to the Awash River are assumed to comprise rejection flow amounting to 15% of gross irrigation diversion plus the 5% of gross field application mentioned earlier. A conservative estimate for Awash River flows upstream of the diversion weir has been derived by using the flow records from the Melka Werer gauging site, which give lower monthly discharges than the records from Awash Station. Contributions from the Kessem and Kebana rivers have been ignored, again a conservative assumption from the point of view of salinity downstream of the drainage outfall. A salinity of 0.35 mS cm-1 has been adopted for irrigation and surface drainage water and a value of 0.50 mS cm-1 for the Awash River upstream of the outfall. The latter value reflects the apparent slight increase in river water salinity downstream of the Kessem and Kebana confluences. The salinity balance has been prepared for two assumed levels of groundwater salinity. A value of 10 mS cm-1 is intended to reflect a probable long term level and that of 40 mS cm-1 reflects a possible extreme. The salinity of the drainage water removing accumulated salts derived from the groundwater has been determined from the ratio of drainage water depth to depth of fall of groundwater during the fallow season. This gives salinity levels as follows: Groundwater Drainage water salinity - vertisol Groundwater Drainage water salinity - vertisol 10 mS cm-1 23 mS cm-1 - alluvium 14.5 mS cm-1 40 mS cm-1 91 m3 cm—1 alluvium 57 mS cm-1The salinity of the drainage water deriving iron subsequent Irrigations has been Obtained by aEsuming a long-term steady state salt balance in the anil profile above the sub-surfacc drains - ie salt deposited aa a result of secondary salinization during the fallow is removed by drainage water during the rollowing cropping season. In the context of Awash Water used by farms downstream of Main Drain 2 outfall and the passible re-use of drainage water Ln the Angeiele area rather than return to the river, the results of the calculations may be compared with guidelines quoted in the USDA Agricultural Handbook Notes □n the suitability of water for Irrigation: Low Hazard Medium Hazard High Hazard Very High EC m§Zem D/10 ~ 0.25 0.25 - 0.75 0.75 - 2.25 2.25 With groundwater at 10 mS cm—1 the calculations indicate that Awash River salinity downstream of the Main. Drain 2 outfall would, on a period basis, remain acceptable for use. With an extreme groundwater salinity of 40 nS cm-1 the Awash water would fall in. the high hazard range until about mid-Septemhor, when drainage water carrying residual salts would cease to contribute "o the total drainage discharge. During reclamation leaching, in which areas abandoned through saliniaation art reclaimed, relatively smaLl areas will be reclaimed in any one season (ref Annex 3 - Groundwater and Salinity)- Because of the small size of these reclamation areas In proportion to the remainder of the project area being irrigated, it Is considered that the quality of return flows to the Awash River will not be adversely affected.TABLE Bl VARIATION IN DRAINAGE WATER A»<) AWASH RIVER SALINITY FOR 14,200 HA COTTON DEYELOPKNT month OAYS TRQM PLANTING (Planting start aid-May) M 1 JULY 1 AUGU5T 22 4 2 62 75 82 | SEPTEMBER I OCTOBER I MIVEMflER I DECEMBER 119 135 142 165 170 PER1DD LENGTH (Oeya) 20 20 13 7 12 7 PCRIOO UNIT DISCHARGES 5 (M /SCI€»€ HA) Gross irrigation diversion 1149 1585 568 304 701 523 434 520 1187 520 1416 308 616 370 0 Surface drainage 213 293 105 56 144 27 80 96 220 96 262 57 114 68 0 Drainage with residual salts (Alluviua) 0 112 73 39 67 45 56 39 - - - - - - • Drainage with residual aalta (Vertisol) ai 81 52 28 41 - - - - - - - - - lotal sub-surface drainage (AL) 0 112 73 19 158 105 132 133 215 94 446 96 236 71 106 Total eub-aurface drainage (VS) 81 193 125 67 206 105 132 133 215 94 446 96 236 71 106 6 PERIDO FLOW VOLUTES (M\lQ ) AwBah upstreas of diversion Gross irrigation diversion A-aah upslreaai of outfall Combined drainage of outfall Sub-Surface drainage Surface drainage SALINITIES (siSca“i) (Irrigation and surface drainage 0.35) (Awash upatreun of outfall 0.5) WITH GROUNDWATER 10 ■Sea-1 SUB-SURI ACE drainage water COMB INC0 drainage at outfall AWASH downstreaa of outfall WITH GROM) WATER 40 "6cm'1 SUB-SURFACE drainage water C0MBKD drainage at outfall AWASH downstream of outfall 46.9 141.9 103.6 102.9 170.; 16.3 22.5 8.1 4.3 11.1 10.6 119.4 95.5 98.6 159.4 3.2 6.0 2.7 1.4 4.4 0.2 1.8 1.2 0.6 2.A 3.0 4.2 1.5 0.8 2.0 113.H 112.4 79.7 149.3 34.1 7.4 6.2 7.4 16.7 2.4 106.4 106.2 72.3 132.6 26.7 2.9 3.0 3.3 6.2 2.7 1.5 1.9 1.9 3.1 1.3 1.4 1.1 1.4 3.1 1.4 116.9 14.0 56.0 32.7 48.2 20.1 4.4 3.7 5.3 0 96.8 9.6 47.3 27.4 48.2 10.1 2.2 5.0 2.0 1.5 6. 3 1.4 3.4 1.0 1.5 3.8 0.8 1.6 1.0 □ 23.0 13.4 13.3 11.1 6.6 2.0 4.3 6.0 6.0 3.7 0.64 0.68 0.65 0.5B 0.9 91.0 52.0 52.0 51.0 24.0 7.8 16.0 23.0 23.0 13.1 1.08 1.24 1.11 0.82 O.fr 5.7 5.4 4.1 1.3 1.3 3.1 3.5 2.5 0.8 D.B 0.57 0.58 0.59 0.51 0.53 20.0 20.0 14.0 1.3 1.3 10.8 12.5 9.2 0.8 0.6 0.77 0.83 0.88 0.5! 0.53 0.7 0.8 0.6 1.2 1.3 0.6 0.6 0.5 0.8 1.3 0.51 0.52 0.50 0.52 0.53 0.7 0.B 0.6 1.2 1.2 0.6 □ .6 0.5 0.8 1.3 0.51 0.52 0.50 0.52 0.53FIGURE > MASTER DRAINAGE PLAN FOR MEL^A SADI AND AMI0ARA AREAS LOCATION OF SAMPLING POINTS FOR WATER QUALITY STUDIES — --------»"« ... w 1 1 I J [1 4 legend R River Sampling Point C Canal Sampling Point D Mam Drain Sampling Point I [i william halcrow a PARTNERS JULY 1905FIGURE 3 •KUH <««!G*TON FMJtCT ANGEllLE ■—co o» tortaA trr»l (*B* •if* SW*M> MASTER DRAINAGE PLAN FOR MELKA SADI AND AMrBARA AREAS LAYOUT OF THE AWASH RIVER HYDRAULIC MODEL S*r Wilborn Hoicrow 8 Ryfrw»» July 1985sir william halXPCw a partners JULY 1385 DIMENSIONLESS HYDROGRAPHS 19/2/83 • 24/^/83 « 20/5/83 a 2o r < o c -n i in D D t- 2 K O u s* u X s o pi 3 o “I X ¥ < »n u t* z 3 u £ |B Lfl o E« j a E* X 01 E* If k k H fll r------- T »- r * I * 20 30 I h UJ D& n C n piFIGURE 4 I I I I I I I I I I I G G C f G G G I I ANO AMI0ArA ar£AS MASTER DRAINAGE PLAN FOR MELKA SADI design hydrographs FOR awash RIVER MODEL OC c uu a in o m SIR WILLIAM HALCROW & PARTNERS JULY 1985 Time,
🔊 Read Aloud
⏹ Stop
<
Summery
<
×
Summery