IMPERIAL ETHIOPIAN GOVERNMENT
AW AMI VALLEY AITHORITY
ADDIS ABABA
MELKA SADI-AMIBARA PROPOSED
IRRIGATION PROJECT
FEASIBILITY STUDY
PART II
STUDIES AND SURVEYS
VOL 3 - GEOLOGY AND HYDROGEOLOGY
Il \ I I i > S M I I Rome. JulyIMPERIAL ETHIOPIAN GOVERNMENT
AWASH VALLEY Al THOR.1TV ADDIS ABABA
MELKA SADI-AMIBARA PROPOSED IRRIGATION PROJECT
FEASIBILITY STUDY
PART ! 1
STUDIES AND SURVEYS
VOL > - GEOIOGY AND HYDROGEOLOGY
Il \ I I 11 \ M I I
Rome. July 1969CONTENTS
Page
SCOPE 
1
WORK PERFOMED
2
2.1           ORGANIZATION OF THE INVESTIGATIONS
2
2.2  INVESTIGATIONS PERFORMED IN THE VARIOUS
AREAS
4
2.  3 DOCUMENTS AND MAPS SUBMITTED
7
RESULTS OF INVESTIGATIONS
3. 1 GENERAL GEOLOGICAL OUTLINE OF THE REGION
3.2          STUDY OF GEOLOGICAL AND ENGINEERING CONDITIONS IN THE MELKA SADI AREA
3.2.1        Geological Conditions
3.2.2        Choice of Site for Weir
3.2.3        Examination of Section A-A
3.2.4        Conclusions
3. 3 STUDY OF GEOLOGICAL AND ENGINEERING CONDITIONS IN THE MELKA WARAR AREA
3. 3.     I Geological Conditions
3. 3.     2 Choice of Site for Weir
3. 3.     3 Examination of Site for                 Weir
3.3.4        Conclusions
3.4           ENGINEERING PROPERTIES OF THE GROUND ALONG THE LINE OF THE DYKE 
3.5           NATURE AND HYDRAULIC BEHAVIOUR OF THE PLAIN ALLUVIALS
3. 5. 1       Nature of Ground Encountered 
3.5.2        Permeability 
3.5.3        Water Absorption Capacity 
3.5.4        Piczometric Map 
3. 5. 5       Chemical Quality of the Waters
3.6           POSSIBILITY OF FINDING GROUND WATERS CONCLUSIONS
8
8
9
9
10
11
13
13
1 3
14
14
16
17
1 8 1 8
18
19
19
20
20
22TABLES
I - CHEMICAL ANALYSIS OF THE WATERS
ANNEXES
1    - PETROLOGICAL EXAMINATION OF ROCK SAMPLES FROM BOREHOLES MW-1E, MW-1C, MS-3B AND MS-3D
2    - ANALYSES RUN BY THE ADDIS ABABA COLLEGE OF TECHNOLOGY
3  - ANALYSES RUN BY SAUTI CONSULTING ENGINEERS OF ADDIS ABABA
4     - ANALYSES RUN BY THE INSTITUTE OF FOUNDATION TECHNOLOGY, NAPLES UNIVERSITYFIGURES
LOCATION OF PIEZOMETRIC WELLS AND DAM SITES
1:2.000,000 GEOLOGICAL MAP
MELKA SADI GEOLOGICAL MAP SCALE 1:5,000
GEOLOGIC PROFILE THROUGH RODIO BOREHOLES
LOCATION OF BOREHOLES SCALE 1:1,000
LOGS     OF     7     BOREHOLES     DRILLED     BY     KELLER     AT     MELKA SADI
GEOLOGIC PROFILE AT INTAKE STRUCTURE MELKA WARAR GEOLOGICAL MAP SCALE 1:5,000 STUDY AREA MELKA WARAR SCALE 1:1,000 PROBABLE EXTENT OF SURFACE LAVA FLOW PROBABLE EXTENT OF DEEP LAVA FLOW
LOGS OF 10 BOREHOLES DRILLED BY KELLER
GEOLOGIC PROFILES THROUGH KELLER AND RODIO BOREHOLES
LOCATION OF HAND-DUG PITS
PIEZOMETRIC MAP
LOGS OF 14 PIEZOMETRIC WELLS
SECTIONS SHOWING PIEZOMETRIC GRADIENT LOGS OK DEEP PIEZOMETRIC WELLS1.    SCOPE
Geological and  hydrogeological investigations have been  carried  out within the framework  ol the Mrlka Sadi Amibara irrigation  project for the Middle Awash Valley.
The objectives          of these inves tigat ions were:
1)         To    c>ose    the    most    favourable    sections    for    siting    the    two    weirs    at    Mcl- ka Sadi and Melka Warar, and determination of the stability and water-
-^htness conditions at the sites.
2)          To    assess    the    engineering    properties    of    the    ground    along    the    line    of    the river dyke;
3)          To     ascertain     the     nature     and     hydraulic     behaviour     of     the     plain     alluvials, in order to assess the consequences of irrigation.
4)     To deteriai e the possibility of finding ground waters.2.     WORK PERFORMED
2. I            ORGANIZATION OF THE INVESTIGATIONS
2. 1. 1 Supervision of works and coordination of the investigations, to
gether with the geological surveys, were the responsibility of Italconsult. The Company assigned a geologist to undertake the field work oetween 5/1/69 and 15/5/69.
2. 1. 2         Drilling, the taking of undisturbed samples from the holes dril led, and water absorption tests were performed by the Johann Keller Drilling Company between 28/12/68 and 31/5/69. This company also dug 2-4 metre pits along the alignment of the dyke. A total of 986 m was drilled. 114 undisturbed samples were tak en, 26 Standard Penetration Tests and 96 water absorption tests were run and 5S pits dug.
2. 1. 3         The laboratory analyses of disturbed and undisturbed samples taken over the whole area were conducted partly in Addis Ababa at the Ethio-Swedish Institute of Building Technology and by Sauti Consulting Engineers, and partly in Italy at the Institute of Foundation Technology of Naples University, The results of those tests are given in the annexes.
Below Is a summary of the quantities and types of analyses run: At the Ethio-Swedish Institute of Building Technology
I he tollowing test were run on 51 undisturbed samples
. Natural moisture content (5 I tests) Permeability (38 tests) Grain size analysis (50 tests)- 3 -
26       undisturbed       samples       from       the       exploratory       boreholes       were subjected to the following checks:
. Specific gravity (22 tests)
Grain size analysis (24 tests)
Unit weight (13 tests)
Consolidation (20 tests)
Angle of internal friction (2 tests)
,        Cohesion (2 tests)
,        Liquid limit (9 tests)
, Plastic Limit (9 tests)
The     following     tests     were     run     on     108     disturbed     samples     from the hand-dug pits:
.        Liquid limit
,        Plastic limit
,        Natural moisture content
.        % finer than No. 200 sieve
•       Free swell (25 tests)
,        Grain siae analysis (3 teats)
Twenty  eight disturbed  samples taken  from the hand-dug  pits were dealt with  by  Sauti Consulting  Engineers*  Addis Ababa the following tests being run:
. Swelling
. Liquid limit
. Plastic limit
. Moisture content
, Wei sieve analysis.
1 hey al.uo carried out I I consolidation tests on undisturbed samples taken from hand-dug pits.- 4 -
Seventeen undisturbed samples were examined by the Institute of Foundation Technology. Naples University. These came
from 6 holes in M elka Warar and Melka Sadi, and the following tests were run:
. Specific weight (7 tests)
Grain size analysis (7 tests)
Plasticity index (7 tests)
,        Breaking strength (4 tests)
4            Consolidation strength (6 tests).
Unfortunately, only on 4 samples (4 boreholes) was it possible to run a Slow Triaxial Test (I), due to the lengthy period of time between the date of collection and date of delivery to the labora tory, and because of the poor state of conservation of the samples.
2. 1. 4 Chemical analyses of 1 0 samples of water collected in the piezo - metric wells were run by the Ethio-Swedish Institute of Building T echnology.
2. 1. 5 The petrological analysis of 4 samples of rock was performed by Italconsult's geologists in Rome.
2. 2 INVESTIGATIONS PERFORMED IN THE VARIOUS AREAS
The locations of the investigations performed are shown in Fig. 1.
2. 2. 1
To study the stability and watertightness conditions at the weirs.
(1) These tests should have been carried out by the Ethio-Swedish Institute, which however was unable to deal with them in the required time.
'I
* •*kXSclka Sadi:                 Geological survey of the zone,
-    7 holes drilled totalling 1 64.85 m  by the rotary
t
method with continuous coring (diameters 146-
76 mm). These holes were set out on two align ments running N-S and E-W,
Twenty three undisturbed samples taken.
The following laboratory tests were run on these undisturbed samples: consolidation, grain size distribution, liquid limit, plastic limit, unit weight, wet sieve analysis, slow triaxial.
7 standard penetration tests.
5 permeability tests (Lugeon method).
7 permeability tests (Open End method).
Mclka Wararr                 Geological survey of the zone.
10 holes drilled for a total of 262.50111, of which one borehole inclined at 25°, by the rotary meth od with continuous coring (diameters from 148 to 76 mm), These holes were concentrated in the weir site area.
Thirty six undisturbed samples taken.
The same laboratory tests as listed for Mclka Sadi were performed,
19 standard penetration tests.
18 permeability tests (Lugeon method). 3 permeability tests (Open End method).
To     assess     the of the dyke.
engineering properties of the ground along the line
Twenty pits were hand-dug to a depth of between 3.50 and 4.00 m.
Thirty eight pits were hand-dug to a depth of about 2.00m. One hundred and thirty five disturbed samples were taken.Seventeen undisturbed samples were taken.
The following laboratory tests were run on the above samples:
liquid limit, plastic limit, natural moisture content, wet
sievp analysis, swelling, grain siae distribution, consolida
tion.
To ascertain the nature and hydraulic behaviour of the plain al-
luvials:
Fourteen piezometric wells were sunk for a total depth of
357. 55 m; )0 reached a depth of about 15 m and 4 a depth 
of about 50 m.
Th■    --e    holes    were    drilled    by    the    rotary    method    with    continuous coring (dia. 146-86 mm).
They     were     cased     with     50     mm     galvanized     iron     piping     slotted against the water-bearing levels.
Fifty five undisturbed samples were taken.
The following laboratory tests were run on the above samples grain size distribution, natural moisture content, permeabi lity.
4 p rmeability tests were run by the Le Franc method
I i permeability trusts were run by the Lugeon method. 27 permeability lusts were run by the Open End method. Ten samples of water were taken and the following tests were carried out: pH. total dissolved solids, temporary hardness as CaCO-.
i
To determine the possibility of finding ground waters. Two piexornetric wells were sunk to a total depth of 201. 60 m, by the rotary method (diameter 146-76 mm).
Two permeability tests were run using the Lug con method.- 7 -
2, 3
DOCUMENTS and maps submitted
The results of the laboratory tests are shown in the annexes.
The results of the investigations set out in the following para
graphs are illustrated by a series of figures in the text, namely:
Fig,       1     -        Location      of      piezomclric      wells      and     dam      sites, 
scale 1:100, 000
Fig.       z    -     1:2, 000, 000 geological map
Fig.       3    -    Melka Sadi geological map, scale 1:5, 000
Fig.       4    -    Geologic profile through Rodio boreholes, Melka
Sadi
Fig.        5    -    Location of boreholes, scale 1:500
Fig.       6    -    Logs of 7 boreholes drilled by Keller at Melka Sadi
Fig.       7    -    Geologic profile at intake structure
Fig.        8    -    Melka Warar geological map, scale 1:5, 000
Fig.       9    -     Study area, scale LI. 000
Fig.     30    -    Probable extent of surface lava flow, scale 1:1, 000
Fig.     1 I    -     Probable extent of deep lava flow, scale ]:1,000
Fig.    12    -    Logs of ]0 boreholes drilled by Keller
Fig.    13    -     5     geologic     profiles     through     Keller     and     Rodio     Bore- ho] e s
Fig.     14    -     Location of hand-dug pitsr scale 1:1 00. 000 Fig.     15  —■  Piezometric map, scale 1:1.00,000
Fig.    16    -     Logs of 14 piezomclric wells
Fig.     17    -     3 sections showing piezometric gradient Fig.    16    -    Logs of 2 deep piezometric wells..- 8 -
3, results of investigations
3. 1 GENERAL GEOLOGICAL OUTLINE OF THE REGION
The vast plain where the project is sited consists essentially of very 
thick alluvial deposits (over 100 m) (1). Deposition probably starled 
at the end of the Pliocene and is still in progress, These deposits oc
cur in terraces which reflect their age (these terraces are crossed 
along the road from Awash Station to Melka Sadi). They consist es
sentially of very fine silty material though interbedded gravels and 
coarse sands are also present; these are sometimes slightly cement
ed by argillaceous silt, They are to be seen in the area south of
Melka Sadi.
This alluvial cover probably overlies the basic volcanic complex 
outcropping along the margins of the plain (Z). It should be noted 
that volcanic activity continued into the Quaternary at the time the
alluvial silts were being deposited so some flows are encountered 
in the alluvial rover (as observed at Melka Sadi and Melka Warar). 
The various ridges rising from the plain (Asada, Bahadamu, Ghed-
diaro) are also manifestations of this recent volcanic activity, as
are the hot mineral springs at Unda Bilen and in the sector west of
Melka Sadi.
rhe area is still subject to  volcanic activity  though  conditions are quiescent at the moment.  The explosions of Mount Fentale and  of the adjacent cones occurred only 100 years ago.
■ I) Boreholes D^S-1 and DW-2 sunk during the project reached a depth of 100 m and were still in alluvium,
2) However, the presence of Mesozoic limestones and sandstones analogous to those present in the Abbai River and Webi River areas is not altogether to be excluded,.GEOLOGICAL MAP
-L.&QENQ.
L
A den vole ante serie*
Antal* hmeiten*
Ad<jral ssndilsee facies
Tea? series
Unmet a merphoied intrusives
Upper landatonr faciei
lltnoitonea    Somalia
in
< Stcelral
Met a mor pSici
fr
Anin© Irmeit one
SCALE F 2 000 000
No. 28 138
ITALCOMSULT- 9 -
This volcanic activity is linked with the large-scale fracturing as sociated with the Ethiopian Rift Valley, which runs approximately NNE-SSW along the 40th meridian and accompanies a zone where
the basalt crust bends down to the east. Intense seismic activity 
occurs in this zone. The most active sector is encompassed by a circle having a radius of 50 km around Kara Kore (see Fig. 2) (this activity has been going on for 10 years). Statistics denote an aver
age frequency of one catastrophic earthquake every 4 years along the Rift Valley (i. e. magnitude greater than 5) (1).
It is observed that several stretches of the Awash River run in the
same NS direction as the principal faults, while other shorter sec
tions run in a perpendicular direction, denoting an orthogonal system
of faults. The risk of instability is probably greatest at the points
where the two systems meet.
In the hight of the foregoing, allowance must be made for seismic activity when designing the works (in particular the writs' It is considered that a series of expansion joints should suffice to guar
antee stability sincethnthick alluvial cover should considerably attenuate 1 nr energy transmitted.
2 STUDY OF GEOLOGICAL AND ENGINEERING CONDITIONS IN THE MELKA SADI AREA
2. 1 Geological Conditions
Fig. 3 shows the surface geology of the area.
The following formations art* present (2):
Recent alluvium: silts and sandy silts
1) Data supplied by Prof. Guoin, Director, Geophysical Laboratory, Addis Ababa.
(2) The loose soils have been classified on the basis of the grain size analyses of the samples collected. The percentages have ’'Ccn expressed by the terms used in the U. S. Bureau of Reclamations triangular diagram, namely: silty sands, silts, sandy silts, fine silts, and fine clayey silts.MELKA SADI
GEOLOGICAL MAP
SCALE 1:5,000
LEGEND
RECENT    ALLUVIUM    SANDY    SILTS AND SILTS
OLD ALLUVIUM. ALTERNATIONS OF CONGLOMERATES WITH SANDY-SILTY CEMENT, SILTS, SANDY SILTS AND SANDS BASALTS
MS2
0
2
KELLER BOREHOLES
RODIO BOREHOLES
-
’ J SV
ITALCONSULTOld alluvium: alternating conglomerates with sandy-silty cement, silts, sandy silts and sands
Basalts
Old clayey silts (which outc rop only on the east of the map under
a basalt flow).
The boreholes have shown the recent alluvials to vary from 5 and 
10 m in thickness.
The old alluvium overlying the basalt flow has a thickness of between 
5 m and a few dozen metres
The basalt flows may be up to 20 m thick.
Under these flow’s it is probable that the same formations repeat for
some hundreds of metres in varying alternations.
It is probable * hat there is a series of N-S and E-W faults or fractures along the course of the river and in particular a N-S frac? ire 'etween the
profiles of boreholes A-A and B-B, and E-W fractures to the north of AA transversally to B-B.
Choice of Site for Weir
Prior to this study, an initial position explored by Rodio boreholes
had been chcs-n as the site for the weir. This lay to the south of the area explored by Italconsult. Fig. 4 shows the geologic profile as evidenced by these boreholes. The an: a chosen by Rodio however had the following disadvantages.
Position lay at an elbow in the Awash; this is probably of tectonic origin
1 he formations on the right and left banks are not homogeneous (alluvials on right, bas tits on left).
The basalt outcrops on the left bank are very jointed.Therefore, investigations w- re concentrated an the straight section north of this position, the whole length of which was explored (circa 350 m). A series of borehol- s was sunk parallel to th< river with 
■i view to locating soils havi; z: suitable mechanical characteristics for a dam sit. . Witurally, morphological characteristics were also hot. i- in mind when deciding ->n the location of the weir and the di* version canal
rhe section d* scribrd below ulfills both requirements.
3. 2. 3 Kxamin a ti□ n o  Section A-_A
1
The     results     obtained    from    th-.-    5    boreholes    drilled    are    set    forth    in the log* in Fig. h and tn the profile in Fig- 7-
Sij i <j 5_‘'■!nn_ot _bed s
r 1 ■ in ’."Lye r of rccen' it- bed of - glomerales ii
east to owe i 15 m in the wet.
i in cmigiotLu* rate r ests or and wedges o t to the west„
-ills outcropping on both banks there
reusing in thickness from 5 m in the
this bed is absent in borehole MS-3).
a sail flow about 7 m thick to the ear. The whole overlies old silts meta mor
ph u •. d along '.he line of com                 zt with the flow. Mr hamcai c .aracter iatics
Standard Pen- : ration Tests of between 3 • nd 5. They tl: are to be cun: ide red absolu* we i r.
,n on the surface silts gave N valu. ±
* have very low bearing strength <■ url y unsuitable for the foundations of the
The    conglomt    -ate#    and    the    h-    salt    flow,    on    the    other    hand,    are    cam pact and thus form an excel!nt foundation.MELKA SADI BOREHOLES
NOTES
-        In    the    Lugeon    tests,    the    absorption    and    permeability values   obtained   at   1   .5   kg/sq,   cm   are   shown   on   the   left of     the     column     while     those     obtained     at     3     kg/aq^cm aT6   shown   on   the   right.   When   the   test   was   run   at   at   - mosphcric pressure, only one value is shown.
-        [n    the    absorption    diagrams    the    broken    line    indicates values     recorded     at     1.5     kg/sq.cm     and     the     unbroken line those recorded nt 3 kg/sq.cm.FEATURE MELKA SADI
BOREHOLE No MS
FIG 6’T
H31V#
IE»nd iad u -b'q |,
13'drtVS
□aaai-Siawri
-| 6 q lJAil 33.7V
H IHhr'HlHffT.H
0% I*ItiO aL JrtViri
HDI.VA J13
UlkJ
a3-9 OH
FEATURE MELKA SADI
BOREHOLE No MS-FEATURE MELKA SADI
BOREHOLE No M5-]
X
c*
>
sE
UJ
▼tel U
DEPTH
m. UElOW
GROUND 1 EVE.
-LA!
7 1- uaTER J Al
PLHMEABUiH *E$T5 r
■J
3
Zl
u
> 14
■_
_4
A
I
l/l
LUGEO*
lit fmirLfrnt
opt N
EhD TEST
Ut?min /ml
ABSO«PTiL>w DIAGRAM
M /n-i i rt /ml
?*• ■
t rt
••I M
MI
II Tte* M
1 10
J Au
1M
aw
1
t* >n
> pc
Aw
L ]0
*<:
a ic
'«
au
iw
i«a
Cl* yay Sdt
Silty Sami
Fine 51.
Sindy 5J1:
Cang;3urnvr*1u Si It v - i^n'hr C*n-.»ai
Cl*yay 5LH
TH AL
•a 14
4K
CwTr-p*cC Bililt
T» tei
1UB
t ta
Ven Cid* r EL***Lt, nmc Ilmen Juialvd
fir 11_____
5.04
Fine Cl*yvf 5ilt
U'M
ilpo
i*w
2 Ml
ikiuiy Sill
HIM
___
E-9 OUI
I
I
I
I
I I
■
I
■
ft
**
*a i.umu nn
i«iir -
F MURE UELKA SAt>
BOREHOLE HoMSJb3
' i£ *
elevation
SIMPLE
DiAME’E*
□ RkllMG METHOD
WARR LEVEL IT t.C I
m
UNDISTURBED
SAMPLES
PEWE1HMI0N
RESISU'.tf
I blnwi pr' fact
WATf P
LOSU '
I 1 II I 1 1 ]
I
I
,1 I
I
I
9-9 OUEl F VAI 13 h
sample
I
I
DI*mf >
I
DRU.>j
Ml IK]
mill    Hl
KIUIIII
I
LJ.LLLllL
,HJ4l  HI
II :
V
I
I
I
UNDI5'UPStD SAMPLED
HMLThilfOh At JIMXnCE
. Ib^wi r<‘ iki
WA*rfi LDEEES
LI I I I I I I IHe. 2 B 14912 -
Pe.rmeabi_l£tiej (1)
The Open End Tests carried out in the conglomerates indicated per meability values of between 3 and 5x10 cm/aec. Thus the forma tion is pra tically impermeable. During drilling operations, in falct, the total absence of water from this formation was noted*
The Lugeun tests run in the basalts (with a pressure of 1.5 atmo* spheres) indicated permeability values of between 2xl0'^and 
8x1 0” cm sec which. although higher than in the preceding case, 
r
is still very low in the absolute sense.
(l J Hate on permea bdily rnr < ■ u.ttTWf
Varioui typrwid lf»>te were run. depending on the nature of the eallv and th a type
of prrmeabilill j Iw bn
(horiioniat or vertical).
in she baish* compact, dr? allluviala (villi). the Luge on method *4i applied.
In the ha ealt a two auccreaive !«•**<■ re run at pressurel of 1.5 and J. 0 kg/x^cm 2a lhe a|luvia]i the Lug eon teste were run at umo vph* rlc preevur*. The Le Franc method wii adopted in rhe loose alluviale and where there wae a water table.
Th- Open End method waa applied to determine vertical permeability.
The values of pi> rmeabiUt* coefficiento K were calculated with the following for-
molar
If ■ 4h£ _LL. _ I J Al" 2.1_ Mr Lh<>4
e
K•
QI
i 5 Tjfc
K = pflrmcj L.lily xn m *#t
£j > Injected J-aiharge in. tu. s/ltc
P            r»*l pretnarr rx;iiiag In the o«Oton in rr of *ll«r A            half thi- JrngEh of the section in m
R             ridim ir m
U AI£•< itlA.        F T! Hi _±<
DKh
C             pOC^tl coefficient
□             divcbarge In cu.rn/ane
L.1  ■ diarresrr cf the per ket Ln m
K ■ petmea!- ility Ln m/«*c
h = dltferri-Le in m br-.«4«n italic lw«l and dynamic level
i
Pp = ri
MeLhpy,
5, 5H
K * peirnei'-.kiy in m/aec
Q = absorbed ovvehirg- in CM m/|ft
r rkdjiie oi caving in. m
U = height nr column of water In fii.- 13 -
2. 4 Conclusion 5
The    weir    should    be    sited    exclusively    on    the    conglomerates    which    have excellent mechanical characteristics and zero permeability.
The structure should incorporate a series of joints to allow for the
rismic nature of the region and the probable presence of a fault or
fracture which here run parallel to the course of the river.
3 STUDY OF BIOLOGICAL AND ENGINEERING CONDITIONS IN THE 
MELKA WAR. AR AREA
3. - Geological 1 'unditio^a
Fig*    9    shows    the    surface    geology    of    the    area.    The    following    forma tions are pr*-ient:
Recent alluvium: silts and sandy silts
OLd allu ium:               fine clayey sandy silts
Various basalt flows interbedded wi*h the older silts.
The Recent silts cover the whole oi the valley bottom for a width ol
250-300 mj while the old silts outcrop on both lulls bordering the
valley ,
There appear to be three or four basalt flows. The most recent
covers the crest oj the hills, the middle one occurs at the foot of
these and th- oldest outcrop in the river.
.hesr’    latter    have    been    picked    up    in    various    boreholes    and    the    pro bable   extent   of   the   surface   flow   is   shown   in   Fig,   10.   The   top   of   the iluw   lies   at   an   elevation   of   between   726   and   7   34   m   and   it   dips   from north    to    south.    It    varies    from    70    cm    to    10m    in    thickness.    Beneath this    flow    the    □□rehales    have    picked    up    yet    another    which    is    evert    old er.    The    extent    of    this    is    shown    in    Fig.    11.    It    appears    to    be    border ed    to    the    west    by    a    fault    which    follows    the    course    of    the    river.    The top elevations are .28 m to the north and 714 m to the west.MELKA WARAR
GEOLOGICAL MAP
SCALE 1:5,000
LEGEND
□
□
MW1
□
RECENT ALLUVIUM SANDY SILTS
ANO SILTS
OLO ALLUVIUM ALTERNATIONS OF CONGLOMERATES WITH SANDY-SILTY CEMENT, SILTS, SANDY SILTS AND SANDS BASALTS
KELLER       BOREHOLES
1 RODIO BOREHOLES
•
N. 28ISOA
WARAR
FIG. 9
STUDY
AREA
LEGEND
RECENT ALLUVIUM SANOY SILTS
AND SILTS
OLD ALLUVIUM ALTERNATIONS OF
CONGLOMERATES WITH SANOY - SILTY
CEMENT. SILTS. SANDY SILTS ANO SANOS
basalts
MW1 O
KELLER BOREHOLES
RODIO BOREHOLES
CROSS SECTIONS
SCALE
I ’,000
N. 28151mEL
ka warar
FIG. 11
probable
extent of deep lava flow
MW1b-»
El
LEGEND
MW1O
KELLER BOREHOLES
2»
RODIO BOREHOLES
A--------------
CROSS SECTIONS
TU TO
ELEVATION TOP OF BASALT
DEEP FLOW
-------------
PROBABLE FAULT
SCALE 1        ’.000
he 28151
ftTAlCQMWLTj14 -
The profiles in Fig. 13 «how the trend of the two flows and of the
silts lying above, below and between them in the Study Area.
3. 2 Choice of Site for Wuir
An initial line of holes run across the valley by Rodio indicated a pos sible position lorthe wrir sonv (00 m south of the island inthe river. However it has been necessary to discard this because of the low load bearing capacity of the silts (3. F. T. gave N value of 19 as shown by the Ra
dio boreholes). Furthermore, from the design aspect the weir must be sited in a wider part of the river so as to keep the height of the dykes to a minimum and to provide a better site for the intake works (set- Vol. IV).
Thus the survey was concentrated a little farther north where the river broadens out considerably (to about 6 0 m as against the 30 m to the south.; Boreholes have been put down here to ascertain the nature of th subsoil.
The results obtained from the 9 boreholes drilled at the site for the weir (plus an additional hole outside this zone) are set forth in the Logs in Fig. 12 and in the profiles in Figs. 15-1 and 13-2.
Sue£ _r s_si qn_i u f^Bed s
The cross sections indicate the following sequence of beds:
Recent silts on the banks to a depth of 6-8 m
b irst basalt flow some 4 m thick. This wedges out on the left bank.
Fine sandy clayey and sandy silts, very well developed on the Left bank. These pinch out between two basalt flows on the right bank*
Second flow, or scries of flows with interbedded silts, These basalts are over 10 rn thick.,MELKA W1RARBOREHOLES
lib
.
14-ii«n
th*i ibmrptikjn
-i.iiiih'4 Hi 1 k 4 ' tq ijm m #hw
IP* le:'t
j 0».« +hki'4H, whilifl Umtt# ik'tnjuned >i < v< *4 fri
r»n thm fight Wh^R ihu W»t •** * rW it it-
'tllitoi’h-
Hilv "ii*
1*
H th» '1 *i*t 1*1 hi
di j Tj 1 mv* i>* >. *
, 1 tr ■
■ ■ ■ ^. • -i t
*n-
1-*
1 n4 ■ h ** l; ril j r ■ t c* '
<«•& H ... ,
<- V 1;I
I
I
I
I
I
I
I
I
I
I
i
a
a
a
a
a
a
a
a
a
fig T3-iELE VATIDW
EiAME'EP
A‘AT i H ! Fkfl tn h g I
1/MDISrU^BEt' SJt M J'L FS
jw rNiiiDK
W|r SiVftNLf
fa.n-S fl I?!- fcjl
WAIF k -LT.5' j
=
I
I
■
■FIG.12-3r.     'P< r.Gl i
i £j
Di
Pi
4«
.
sr
SlSSDl WlfM
|Wj Uld Ufc*wl| j
3JH¥ISh>1H
NCll J 1H3d
SJIdUVJ ajeiiHLSiiiHA
qohjih
gwiriihia
alLSitVia
'inwnj e
11 ■ x
-401 LtAJIl
-     „• <
... j i mI
I
I
I
I
I
I.s-n sij
SC1RLHQLE No MW I «NOir/A l ■ j
IX ki"
d. 
k t- 'JT
J . _
Z-Zl OldHI
HIREHOLE ><  MW I g
&-r-          -
SJJcx-irouFIG. 13-1
(jTALCONlUtT ]
N 2B164FIG. 13-2
mELka *arar
GEOLOGIC PROFILES THROUGH KELLER AND RODIO BOREHOLES
□
□
•‘SALTS15 -
As indicated in Fig- 11* it is possible that a fault runs along the line of the river, truncating the old lava flows Id the west.
ALrktn\nica 1_ Gharj.c,ter isti_cs_
Standard Penetration Tests run on the surface silts gave N values
of between 12 and 20. With thesefigures the safe loading is as sessed to be less than I kg sqxm.
A■ the expocted level of the foundation - 728 m a. 9. 1* ™ N was 40 to 60 and the safe load could thus be over 2 kg/sqcnru Despite the great number of undisturbed samples taken at around this elevation from holes MW1A, MW1B and MWlD. there are not sufficient reliable triaxial compression test results available because the Contractor
did not carry out his obligations as regard the performance of labo ratory tC!5tS.
Tin* results of the samples taken from MW1B at elevations of 724. 50 and 721 :n have been assumed to be indicative.
Tin■s u are:
26L
C = CL 3 7 kg/ sqjcm
- i . i>4 kg/ l.u. d-n
l
Iti&nat ^onfficfered prudent to consider the results obtained from the sample taken from MWlD at an elevation of 728 m since the value of ^fcp(39°20 ) is we 11 above the average. This high figure may well h- due to the state uf preservation of the sample which was not tested until four months after it had been taken.
Applying Terzag hi's. Caquot-Ke risel's and Frohlich's formulae to 
the results fr he MW13 samples it ensues that the safe hcirin^
load lies Jome where between 1*1 and 1, 85 kg/sq, cm ( Frohlich 1,1, Caquot-Kerisel 1,47. Te raaghl 1. 85). This rings is iiigoodajree-
with the figures derived from the Standard Penetration Tests.
Fr ni the series of T values avaiUhLc for the three holes it is noted 
that the value biueath the 7ZH m elevation is always over 40. Thus
it would Seem that a load of I kg. sq.cm or a little more would be
p - rmissible (the N value would even allow for a loading of 2 kg/ sq.cm tut bi view of the incomplete knowledge of conditions in this area
it would not bij prudent to adopt this figure). The basalts seem to 
tt reasonably ompact and capable of taking a load of at least 2 kg.Aq.cm.
116 -
Pe
Lc Franc tests run on the Recent silts gave values of between 1 and 5 x ID"  cm/sec. These tests and Open End Tests run on the old 
4
silts gave values of between 3 x I O’® and 1 x 10 cm sec* Thus while the permeability of the Recent silts is very low, the old silts arc virtually impermeable.
Lug. on Tests run on the basalts at a pressure of 1-5 kg/ sq- cm in dicate permeabilities of between I x 10-3 and 1 x 10   cm. sec* With permeabilities of this kind, there could be serious losses from the reservoir if extensive areas of basalts were involved- However, is only limited areas of outcrop will be under water# this eventuality 
4
appears unlikely.
jL 4 Conr lusionh
The Recent surface siits must be eliminated from the foundations. Th« weir will lie partly on the basalts, on the right bank, and part
ly on the old silts on the left. It is held that both formations have suitable mechanical properties. However, during the Final Design stage it would be as well to supplement the data through sinking further boreholes and running geophysical surveys to determine
the thickness and extent of the basalt flows and to learn more about the mechanical properties of the silts.
The possible presence of an erosion surface between the: two forma tions and of a fault in the western part of the section should also be borne in mind- A sufficient number of construclicn joints should be if corpora ted in  h.r weir to allow tor the seismicity of the region and 
b
the probable presence of a fault or a fracture*
Thu water-retention properties of the reservoir do not give rise to any great cause lor concern Since the old silts are virtually im- permcable while the more permeable basalts do not outcrop over a particularly large area. Even if losses occur, they will not be great. However, the basalts could be rendered watertight by grout ing'17 -
3. 4 ENGINEERING PROPERTIES OF THE GROUND ALONG THE LINE OF THE DYKE
Twenty hand-dug pits varying in depth from 3 to 3,5 metres
have been sunk along the Line of the dyke to ascertain the mechanical properties of the soil. The position of these holes
is shown on Fig. 14 where they are marked with the symbol
"A". The pits are spaced at intervals of 2, 000 to 2, 500 metres,
A further 38 pits averaging 2 metres deep were sunk in the area between the river and the line of the dyke at a distance not more than 50 metres from the latter. These pits which are marked with the letter "R" on Fig. 14 are spaced at 1, 000-1,200 metre intervals; they were designed to ascertain the engineering pro perties of the soils which will be used to build the dyke.
Two or three disturbed samples were taken from each pit and a series of undisturbed samples was also taken from the repre sentative horizons. The results of the Laboratory analysis are given in the Annex.
On the whole, the soils are fine, the percentage passing a 200 mesh A5TM sieve almost always being in excess of 90%. They may be classified as silts, sandy silts and fine silts. The plasticity index is normally in excess of 15% and the liquid limit is almost always above 40%. This means that these are A6 and 
A7 soils according to the AASHO classification and OH soils fol lowing the Cassagrande classification.
Their cohesion is between 3 and 6 tons/sq.m and the angle of internal friction lies between 6° and 12°.
Five metre high dykes built from materials lying nearby and having side slopes of one to two would have a safety factor of over two.
On the basis of the consolidation tests run on undisturbed samples total settlement would be in the region of 3 0 cm, 50% of this
5 cm' occurring in about 6 years.ja -
It maybe bxmcludad that the Jtibillty of the dykes will be •*«•- factory provided that sufficient freeboard is left between the maximum waler Level and the top of the dyke t© Uke account of the expected 30 cm of settlement.
5 NATURE AND HYDRAULIC BEHAVIOUR or THE: PLAIN ALLUVIALS
The 14 pier.ometric wells which are sunk provide information on the nature, permeability and water absorption capacity of the plain alluvia Is to a depth of 50 m  and on the configuration 
r
□ £ the water table.
5. 1 Nature >f Ground Encountered
Am will be seen from the 14 logs given in Fig. 16. the plain alluvia Is consist of an alternation of silts., sandy silts, clay 
gilts and silty sands.
The cover is extremely uniform, the only exceptions being noted in Piezometer P 9 which encountered basalt (12 m from ground 
t
level) and in P. P. 12 where a bed of conglomerate was picked up
.6.2 Permeability
Lugcon  and  Le Franc tests indicate an  average horizontal per meability  of 2.4  x  10-**  cm/sec while the Open  End  Test points to vertical permeability being 1 x I O’  cm/sec,
5
Laboratory tests gave an average figure of 3.2   L0“  cm/sec
x
4
which is in. good agreement with the results obtained by the Lug eon method.- 19 -
It is interesting to note that vertical permeability as indicated by the Open End Test is considerably lower than horizontal per meability fas is usually the case), With a vertical permeability of L x KO”’  cm/sec the seepage rate is 3 m per year.
1
3.5. 3         Water Absorption Capacity
Comparison of the moisture contents of undisturbed samples taken from above and below the static level indicates an aver age difference of l«%. This 18% represents the volume avail able for ibsorbing waters which infiltrate.
If; as indicated in Vol. IV part II. the annual infiltration rate is not Less than 3 0 cm  the level of the water table would rise
r
each year by °a
’ i ■ m- Thus in 5-6 years the silts
would be saturated. However, it should be noted that a similar
phenomenon has not been observed in the case of the old silts
lying beneath the river level. Indeed there is a difference of
around 2 in between the level of the river and that of the water
table at Molka Sadi and around 4 m at Melka Warar. It is thus
likely that evaporation is sufficiently great to decidedly Limit the
inflow of water to the aquifer, in view of the low seepage rate in 
the deeper strata.
None the Less, it is possible that the more permeable recent
silt-* (K = 10’- to 10"3 4 cm/sec as indicated in the Soils Report)
will become more rapidly saturated by irrigation water, and 
hence the calculated saturation Emit of 5-6 years may well occur
Thus,  provision  must be made for an  effective drainage network which  must be constructed  in  the year*immediately  after project implementation.
3. 5,4 Piezometric Map
The Piezometric Map (Fig. 17) indicates the presence of a trough in the water table on the extern side of the plain. High
er water table elevations arc to be observed on both sides of
this 2one indicating that a certain amount of recharge in received from the river and from the hills bordering the plain.MELKA SADl-AMIBARA
PIEZOMETRIC MAP
SCALE I 104000
71664.
720 JO,
.PIO
A72656
LEGEND
CROSS SECTIONS
72444
PIEZOMETRIC WELLS
J AO v >
Jii
FIG.PIEZOMETERS.. . ..-'2;: .*
U>3 M
nM/vtui/in 1531 0M3
N3dO
iw/-uiwn’i Noaom
saidHvs
oiGHrusiOMn
"l fl u>
i3A3i aaivw
■nvmnrnnn^n< niH.iu ihiii
T.'IWI ::
v .• 4 I -
2e i*eSM/un k
-luifiMuinn 1$31 W 3
M340
juyv'wn’l Noaom
S3AdrtfS oaemusiQNn
OHPSVD
I I q ui 13AJ1 a3i¥*
001
OOH13K ONiniua
h 1,1 IIIIHlillllllH ELRIiJIllIflJlUCFlIIII
TIHlLIlJI Lr i!l IIIIL1.1l1IlHIIIMHIIJLPilrH4Hitl‘H.!t.llFl JAtr■. I* I.L, . ——UJ li l l i.
,
sOIir" 11?
LEt£ rfAf IOK
p bA7URE AM I BARA
DEPTH m BELOW
GROUND LEV£.
PIEZOMETER NoP3
FIG. >6-3
1
F-
Uf
I
<
□
DESCRIPTION ano
ClASSlFrCATiON
OF MATERIAL
PEPHE ABILITY
TtSTs
1o
_j X'
—J fr— a"
a1
Fl
•J
Wl
*
sH
i
JC
Sa Mr Silt
5b Mv 51 n
Fin* Sb It
Silly S-nd
Fih*
Sils:*-i
-Fl
LT
o
in
Cl
*-■
u*
r
FLt VAI ION
msI
r
«or*Rr
CUMUl simple
DlAMETf ft
□KILLING METHOO
LOG
c*>
O r~
►
O
X
r~ O
Z
>
z
o
1
B7i !&’*
i i.tf ’
■
WATER LEVEL m b g, I,
S CASING
UNDI5TURBED SAMPLES
LUGEON liUrnin /ml
OPEN END TEST UUrnm/mt. K cm/>«c.
-o
rnlM
A•—  —A
r
[
V-
** *
rr
w
B"b
ELftfi! Ik,
m*kI
L-
**
w
5R
A
-
te
•
-
OI
*
CVMUL
•
*■
o>
o
rp
fT>
Ma
■>
In m —4 cz
XI m ID
> X > o
SIMPLE*
DIAMtTEM
DRILLING MEIHQO
c
c
WOTARV
,
i 7 TIT         J .1 1 i         1 T 111   nirii i        ..:.i 1 lllliII l.r 1,11.I.I.IIHI IIlill.l. IllIIJlill>II IbilllllHitJH. IIII11 II
'irmiiiN illliiiiKiHl JiriiiiinitllilJliiinriiilillni'ii.r
..UIIUIIUU' lIJLjUII II l.l,.1 1.................. mi
.      irr.ir.i      .      ,.i;      r      it.—mi      -"nmT1 liiiHrrjiiiihiutiHrnnimitidFA
i! ’ F IF1 HiJf
■1! d HI I
PT'rrr
r. 11 i 1111 fH 111 II !! '
Fl H
, mrTnrTT'TTT—r-tTr illItHlriltlHIIIIUIH hrkliiHiiiHlilMniil
*I111HII1I HUH IUL!
E
&
CL
LZi
—
;_l
D
*
3
VJ-
1“
Vi
'S
a.
a
Ul
H-
A
wk
*•
-i
rR ft
c
►
r- O >
XX
o
>
►X
-•
p
u
-+
D
1;
— — ---- - ■—     • — -------- .——       —
- — 4---------------------- ---------------------- --------------------------- ------------------ ----------------------- ----------------- - ---------- —------------------------ --- ■ ----------------
— --------  —> --------------------------------------------------------- --------- ------- :------- =---------=------ -- ---------------------------------------------------------------------------
J
I®
• 1
'I
• 1
,= l«
1* **
p=
WAIER LEVEL m A i I.
Jj CASINO
UhDISTUHDED SAMPLES
IGGEOH
tkl/mkH ,*M
U.l***
1
•li id’4
■J
Dall
££
13 >
O
LllIO1
HUB 1
OPtN
t< IC1 MJnwnJmL
R cm/Mt.
-
--------------------- -----------------------
Fin* CUy«y 54leELEVATION
o
o
C7    O
5
,
r mm
Q
X
m
iJIAMFTEn
SHILLING METHOD
T
*AFER LEVEL
m b fl I CASING
UNDISTURBED
SAMPLES
J l|>
3 I * 10 1« ■ 10
LUGEDN
ht/minJrnl.
Of*€N
ENG TEST
UUfflkMml
K em/MC.■<*
>■
t
■w^ mi aw
NldO
B JM" c1
Olitl
t .DI *« I
f _orxi’*
v m »rt
a. jw/tHiijfPl
LU
NO3$fn
saiduvs 099Hni5iaM.n
ONISVD £
I 1 q mj 13A31 U31VW
Rl
*11
_____ -________________              ________ —
*
'1
s
R
M
s
--
JL
n
r
Ut=
Ui— ■—’
r
—
i
hr
■-+
IT.
■.
>
4
□
U
£
k
3h
F-«
Si
5
in
>L.
c
•
*
«
>■
4
D
4
C
£
■p
E
4
W>
Sk
in
5h
T
E
MY
k
■
>
4
b
«j
iZ
in
e
c
001
QGHL3« OHiniMa
fejgjlWTia
SldrtlS
Lm i'i1 Vi
Lij
iiiiiiiii
.
__
I■ =' 1 • 1U LIHdll1 MttiOffll
iiiiii rmiMliu
i ij i j i i llii;i
UIISI1IUIIIIIHH i i Him mi ill
HUI IIIHII Illi II
LEI                  : !  1 J
;<
.    !1;1i
nwj twinr i ' V ii.i
h innmni
IH 1JIU1 II
1 |l|i ‘l 1,1.1
ta
WWPWHHiW^
■i r nr uh di ii hi ip ii uh 'Him iiinyi h ii iiii i 1    i\ t 1 1 . . 1 j 1 .Id Lz.  J  - L LLJ j □. L J uL ■ 1, ,11:—.-JI-—-J !
_                                  _  __        _
AHVlDM
IT)
r
EC. r
-
Fb
!FL
BB
*|
IV
■
:
irmn;
dJ
R
■
*4
4-                        V                            *
tr                        m
♦
I f r lu
NOllVAlil
#
’
:t
*-
■i
r-
R
iO*
F—
2
1-----------------------
—--------------- V--------------------- 4------------------- i
.
F
....
HQ
inn
□ t 11L.UI
«E 1
F . t 7.MI0 S
-jf 3m
□ mu
r- oi
CD
SIM4-H
HDTiRt TmiriT
DIAMETER
DRILLING Mt- rnon
ryi-p-rrrri'l! < ■ i *1IIET1
Mil 1.1 I.OBllir III
*
m
1:
■
r
73Alfl'1
: .imf1
15.
WATER LEVEL m bgl CASING
UNDISTURBED SAMPLES
LUGEOM lil/min /ml
OPEN EMG TEST LiUmiru'mL
IK cm /sic,1 * AlJ
|1
E LE RATION
J. t I
CUMUi SAMPLE
DIAHt 11 H
C3
€J d -1
C C»
--
o?
=
OI
DRILL ING
ME fHDD
T TT'-'rn—nr---'7T Ji. CH H I Illi. I
WATER LE^EL
m b. q |, uASING
UNOI5TURBED
SAMPLES
3»i IO"1 64 ■ IO"1
um’’
Il W
LUGEON IkUmwv /mt
O^EN END TEST bt/mirUmt, K cm/ifc.B
6
I I I 1
I I I n
i i i i i i t I I
DEBU depth
TI below
GROUND LE^El
PIEZOMETER No F it
FIG.16-11
PERMEABlLlTt TESTS
OESCHI P *' 0*» AhO c l * ss-iF icaTien
MiTERKL
1
i T m • n % i iif ir
■ tt'MllDN
ft J $ i
I ; ML
y rpf-i
■j
w r- -•
r’ <j I
o
>
UlJLMETFB
DRIUIMG
METHOD
2!
’£
X
j"
n    r. <1 c.
u3 D ” 3
4 R to
1
5
5
F"
I
■
□ Oi
■<
2                  IS
n
17
:1
<■
-4
r*        1
s
VM ~4 —
»■»
1
WATER LEVEL m, 6. f I.
CASING
UNDISTURBED
SAMPLES
LUGEON lll/min I’fnA
OPEN ENG TE51 (lUmnJfnl.
K cm ll«-
3
o <n
?!*>/. Ki
frIt'S 4'MlS
T
■*                         -
i
-1 '*•1
=     iS
u.
.
JTi
■ ----
t-
__ ——■'               --—[---------------- I£
-
1“
-!
I’ *
1-
!«■
wi
1.
M
r
L
.7
------------------------------ r
i-
1r
r
1.
*■
ft
»■
Ml.
■.
■■           1-
«=■
■r
-
■1
b!MF»LE
]a
H l££
HU11Hf
LJRILUNG
ME rnoD
-------- 1----
III '
rmr
1
rr'—'i-n
II I-. HMIHI I
HHIII l.’illlll II ii l.. r
I LlLHL1l ill 111 IHI ■ I
i
WATER LEVEL
mbI
I
2 CASING
UNDISTURBED SAMPLES
LUGEON
hl/min /ml
l OPEN
C< TEST liUrninSmt.
■m
±j
>
J q - ro
h cm/ifcFEATURE UNO A BllEN                       PIEZOMETER No PF
1
w
e
Z
> *—
ar
* **»
w
u,
X •“
«
Ui
Cl
■iwru'u»ni
1*31 QN1
|U4f VW/VI
N030H1
WdHYS oinunivONn
OMIlVX UMl M31VM
o
faJ u
001
OOMlirr OWIllIMU
H 11IHYTCI
.<w, 31d«1S XO_lj
a uj ° |-
ujE |
° 3'
r
IHROO
t 4T .
i % r uj NOHVA313
mi: nFIO.17
melka warar
SECTIONS SHOWING PIEZOMETRIC GRADIENT
i
AWASH RIVER
c
*"« ll.
no- no nO- TIO-
010
PP 12
P 11
P3 X
P6
I
P9
^1
■ no
■HO
•no
■no
on
I
[itauconoult
N. 2818220 -
It    will    also    be    seen    that    there    is    a    low    hydraulic    gradient    running
southwest-northeast
Hence with a horizontal permeability
of ]0"  l cm/sec the flow rate is around 10’"*^ m/sec.  The aquiier is thus virtually  without flow and in all probability  merely  varies in a  vertical sense with inflow from the rain and outflow by  evapo transpiration.
3.5.5 Chemical Quality of the Waters
The analyses show (see Table 1) that the waters generally have a high silt content (from 0. 4 to 10 gpl \ This is no doubt due to the static nature of the aquifer and associated evaporation.
3. 6 POSSIBILITY OF FINDING GROUND WATERS
The two deep piczometric wells(logs given in Fig, 18) drilled near the volcanic hills at Melka Warar encountered various horizons of sand and gravelly sand which are water bearing. Permeability tests run on DW -1 indicate an average value of 2.5x10 cm/sec. In theory a maximum discharge of around
5 litres per second could be obtained (1). However, before making my plan* to exploit these ground waters, investigations would have to be performed to ascertain recharge possibilities
md to determine the chemical nature of the waters themselves.
(I) Discharge calculated by means o£ Sichard'a formula Q V c x S
where;
Q ib the discharge
Vc i > the critical velocity
5 well diameter, assumed in this ca(je to
be 300 mm.CHEMICAL ANALYSIS OF THE WATERS
Sample Marked as
Total Dissolved Solids
at 11O°C
Tempora ry Hirdnes s
as CaCO^
pH
Well N° p. p. 1
Hole N° P- ?
Borehole bL P. 4
Borehole N  P. P. 5
r
Hole P. 8
P. 9
Hole P. 10
P, 11
P. P. 12
P. P, 14
1,130 mg/1
I, 644 mg/l
10, 624 mg/1
932 mg/1
I, 390 mg/1
486 mg/l
5. 854 mg/1
396 mg/1
412 mg/1
740 mg/1
38 mg/l
I 08 mg/l
395 mg/1
14 mg/1
68 mg/1
I 06 mg/1
769 mg/1
96 mg/1
104 mg/l
8, 8 mg/1
8. 75
7  68
r
7. 03
8, 65
7. 60
8. 10
7. 48
8. 05
5. 25
8. 40
Table 1FEATURE MAHADO
f IE ZOM E TER No DW-1FEATURE   MELKA WARA'*                PIEZOMETER No OW-2
FIG.18-2
z
e» _
«:
DEPTH
•n BELOW
ground LEVEL E
-j 1 UJ             t!
a1                  *
—
description and
classification
PERMEABILITY
TESTS
t£
uj □ Bio
U1*
?°
-*x o
±- o
3
a                                             OF MATERIAL
u
si
85
-J -e
.Bi UJ W. c
u3
V
•
M
i
M
Sa r 1 f Silt
<2*?* w»Mi Basalt Fragments
tl/Sihy sand
Clay Silt with’ Basalt
tmesis
c        hUy Silty Sand
N 21 II4. CONCLUSIONS
The Melka Sadi intake will li<- on conglomerates which have good load bearing properties and are lrpermeablc.
At Melka Warar part of the intake will lie on basalts (right side) and part on old alluvium (left side) below an elevation of 72S. The safe bearing capacity here is a little over 1 kg/sq.cm. Permeability is virtually nil in the? alluvium and low in the basalts.
At both Melka Sadi and Melka Warar provision must be made for con struction joints to allow for the seismicity of the region and the pre sence of fractures running parallel to the river. At Melka Warar ac count should also be taken of the non-homogeneous nature of the forma tions on the two banks.
Examination of the soils which occur along the line of the dykes indicates these to be suitable for construction (safety factor greater than two, and settlement in the region of 30 cm over a period in excess of 6 years).
The plain soils have a very low permeability both in a vertical and a horizontal direction. The aquifer may be considered as practically
— tatic+
I he dry ground above the wat< r table will, it is thought, become saturat ed with irrigation water in 5 i years, hence a drainage system should be built ir. the years immediate! after the land is brought into cultivation.
Ground waters may be found : ear the Melka Warar volcanic hills how ever, discharge tests must b run and more information obtained as to recharge and water quality. y and large it would seem that the wells would have to go to a depth of 1 00 m or more; they would have a delivery
"n
- i-ANNEX 1
PETROLOGICAL EXAMINATION OF ROCK SAMPLES FROM BOREHOLES MW-IE, MW-1C, MS-3B AND MS-3DAl -1
Sample from MW-1 E at 4,50 m
Macros copica I Ly the rock is grey to dark grey in cokur and has flmall
roundish or ovalish vesicles which tend to be elongated in one direction 
ant! are lined with secondary minerals. The cavities vary from 0. 5 to 
5 mm in diameter.
Under the microscope the rock is seen to consist basically of calc-alkali
felspars and pyroxene. It has a porphyritic structure. There arc oc
casional phenocrysts of olivine, labradorite and augite having an average
diameter of 400-500 microns. These are set in a ground mass of micro-
lithic felspars and pyroxene grains.
A fair number of opaque granules of iron oxides and hydroxide^ occur
evenly throughout the rock. Calcite is also present in places as an ac
cessory mineral. The olivine which is always rather rounded has a thin 
reddish brown fringe of iddingsite.
The rock may be classed as olivine basalt.
Sample from MW -1C at 13.50 ni
The rock  is grey  to  dark  grey  in  colour with  some small roundish  vesicles
0. 5 to 5 trim in diameter.
Under     the     microscope     it     is     seen     to     consist     of     a     few     phenocrysts     of     calr- alkali felspar and olivine having an average diameter of around 7Q0-
800 microns. These are set essentially in a ground mass of labradorite with subsidiary augite, olivine and opaque iron hydroxide grains. The average dimensions of the ground mass crystals ia around 2 00 microns. Thu olivine crystals are generally rounded and partially decomposed to id 
• i .it- around the margins and along the fractures. Thr rock may be classed as basalt.ANNEX Z
ANALYSES RUN BY THE ADDIS ABABA COLLEGE OF TECHNOLOGYAl -2
&i:n_?Le from MS-3B at g, 70 m
The rock  is dark  grey  and  has numerous small vesicles with  a round  d or elongated  section,  varying  in  diameter from a few hundred  microns to several millimetres.
In thin section the rock is seen to consist of a few phenocrysts ck calc
alkali felspar and olivine, whose dimensions are of the order of a few
hundred microns, set in a minutely vesicular groundmass consisting 
mainly of closely interknit calc-alkali felspars with a markedly tabular
elongated habit. The interstices between the felspar crystals are filled 
mainly with more or less isometric crystals of pyroxene and subordinate
ly with opaque granules and glassy materials. The olivine always has a
reddish border of iddingsite.
The rock may be classed as basalt.
S.iinple from MS-3D at 17.60 m
Dirk grey or black compact rock with a few vugs partially filled with 
secondary” minerals. In the sample examined, one af these vugs was
about one centimet re in diameter, the others were all much smaller
(around a millimetre).
In thin section the rock is seen to have a porphyritic structure and a sub-
ophitic texture. Crystals of pyroxene are present; these partially en
close felspar crystals and also occur interstitially (1).
In order of magnitude, the rock consists of labradorite augite, olivine
and iron oxides and hydroxides. Chlorite is also present in a fair amount
as a secondary mineral and is frequently associated with caicitn.
The rock may be classed as basalt.
(1)    The    pyroxene    crystals    may    be    from    200    microns    to    1    mm    in    size while the felspars are normally around 500-600 microns.ADDITIONAL COPIES ARE AVAILABLE ON REQUESTANNEX 3
ANALYSES RUN BY SAUTI CONSULTING ENGINEERS OF ADDIS ABABAANNEX 4
ANALYSES RUN BY THE INSTITUTE OF FOUNDATION
TECHNOLOGY, NAPLES UNIVERSITYADDITIONAL COPIES ARE AVAILABLE ON REQUEST

< <

Summery

Summary of Melka Sadi-Amibara Irrigation Project Feasibility Study

Melka Sadi-Amibara Irrigation Project Feasibility Study - Summary

Document Overview

Title: Melka Sadi-Amibara Proposed Irrigation Project Feasibility Study

Volume: 3 - Geology and Hydrogeology

Prepared for: Imperial Ethiopian Government, Awash Valley Authority

Prepared by: Italconsult, Rome

Date: July 1969

Study Objectives

Select optimal sites for weirs at Melka Sadi and Melka Warar, assessing stability and water-tightness
Evaluate engineering properties of ground along the river dyke alignment
Analyze nature and hydraulic behavior of plain alluvials to assess irrigation impacts
Investigate potential for groundwater resources

Key Findings

General Geology

Region consists of thick alluvial deposits (over 100m) overlying basic volcanic complex
Area is seismically active with frequent earthquakes (magnitude >5 every 4 years)
Volcanic activity continues with recent eruptions (100 years ago)

Melka Sadi Site

Recommended weir site on conglomerates with excellent mechanical properties (N values 40-60)
Conglomerates are practically impermeable (permeability 3-5×10^-6 cm/sec)
Construction joints needed due to seismic activity and potential faults

Melka Warar Site

Weir would be partially on basalts (right bank) and old alluvium (left bank)
Safe bearing capacity estimated at >1 kg/cm²
Basalts show permeability of 1×10^-3 to 1×10^-4 cm/sec
Additional investigations recommended before final design

Dyke Construction

Soils classified as silts, sandy silts and fine silts (A6, A7, OH classifications)
Estimated total settlement of 30cm (50% within 6 years)
Safety factor >2 for 5m high dykes with 1:2 side slopes

Hydrogeology

Plain alluvials show average horizontal permeability of 2.4×10^-4 cm/sec
Vertical permeability significantly lower at 1×10^-5 cm/sec
Aquifer is virtually static with low hydraulic gradient (1%)
Water table may saturate in 5-6 years of irrigation - drainage system recommended
Groundwater potential near Melka Warar volcanic hills (potential yield ~5 l/sec)

Water Quality Analysis

Sample Location	Total Dissolved Solids (mg/l)	Temporary Hardness as CaCO₃ (mg/l)	pH
Well P.1	1,130	38	8.75
Borehole P.4	10,624	395	7.03
Hole P.10	5,854	769	7.48
P.11	396	96	8.05

Waters generally have high salt content (0.4 to 10 g/l) due to static aquifer conditions and evaporation.

Conclusions

Melka Sadi site suitable for weir construction on conglomerates
Melka Warar site requires additional investigation but appears feasible
Dyke construction materials and design parameters identified
Drainage system needed within 5-6 years of irrigation commencement
Limited groundwater potential identified near volcanic hills
Seismic considerations important for all structures