Fookes Et Al 1985 QJEG Dubai Etc Evaporites

Q. J. eng. Geol. London, 1985 Vol. 18, pp. 101-128.
Printed in Northern Ireland
The influence of ground and groundwater geochemistry on construction in the Middle East
P. G. Fookes,* W. J. Frencht & S. M. M. Rice~
*Consultant Engineering Geologist, 47 Crescent Road, Caterham, Surrey CR3 6LH; tDepartment of Applied Earth Science, Queen Mary College, London E1 4NS; ~:Sir William Halcrow & Partners, 25 Windsor Place, Cardiff CF1 3BZ (formerly Halcrow International Partnership, Dubai)
Summary
Many parts of the world's extensive hot drylands pose significant problems for the civil engineer because of chemically aggressive and changing ground conditions. The most important areas are generally the broad fiat coastal strips of more recent sediment, for these commonly contain abundant sulphate as gypsum and anhydrite and have widespread surface or near-surface crusts of halite. The coastal water and groundwater in these regions are usually sulphate and chloride rich, moving slowly through the sediments because of rapid evaporation of the groundwater via the capillary fringe into the drier atmosphere. This movement of groundwater accelerates solution and precipitation of the more soluble minerals, the greater the flow rate the more rapid the compositional change in the sediments--sufficiently rapidly to be of engineering significance. The engineering works themselves can effect significant changes in the ground chemistry and this is most often brought about by changing the position of the water table or the capillary fringe, by dewatering, or raising ground levels by depositing fill. In addition to the significant effect of salts on results of standard engineering tests, the other most likely problems created by the chemical changes are: settlement due to solution in flowing groundwater, heave due to crystallization from groundwater or by reaction between soils and groundwater, and the creation of a chemically aggressive environment in foundations. The properties of these groundwater-soil systems are reviewed from the point of view of field observations and some large-scale experiments which give indications of the magnitudes and rates of the chemical processes. Two case histories are cited by way of illustration-Dubai Dry Dock and the Harbour Works at Mina Jebel Ali, Dubai. The Dry Dock represents an example of settlement effected by the solution of gypsum from lithified sediments through which seawater was passing, and the Harbour Works show how groundwater and fill chemistry have been changed in the few years required for the construction works.
Introduction
The geology and geomorphology of the vast areas of the world's hot drylands are very diverse but particularly important from the civil engineering point of view are the long coastal zones which are backed by broad intertidal flats and are commonly fringed by
shallow seas of high salinity. The local groundwater here is generally extremely rich in sulphates and chlorides, and the shoreline sediments are commonly loose to moderately dense silty sands, shelly sands with thin carbonate silts and clays and other clay seams. The sediments are variable and may also include aeolian deposits and fluvial gravel. Inland much the same kind of sedimentary and groundwater regimes can occur in the broad, gently inclined areas around salinas or playas (Fig. 1). The open flat ground of these zones is frequently chosen as the site of major civil engineering works as well as substantial urban development, but the ground conditions can be particularly aggressive and rapidly changing. Difficulties arise for example from the low and variable load bearing capacity of the sediments which when dry can be satisfactory but when wet may not support heavy equipment. Commonly, problems relate to the chemistry of the sediments and their groundwater and it is the engineering geochemistry of these environments that is discussed in this paper. The geochemical hazards are mainly produced by the high levels of moisture evaporation that occur in these arid regions and the tendency for coastal and urban development to change the chemical equilibria that exist between soil and groundwater. Air temperatures are generally very high (around 40~ in summer) and the humidity can be low or high with marked daily and seasonal changes. There is also typically a dominant wind direction. These conditions promote evaporation from the surface and hence cause transport of groundwater through the capillary fringe for up to several metres above the water table. The result can be the precipitation of minerals at the water table, in the capillary fringe or at particular surfaces in rocks and soils. The flow also brings about the solution of soluble phases in older sediments beneath the water table. The rate of solution or precipitation of minerals varies with the individual properties of the rock-soil-water system and is commonly sufficiently rapid to be of engineering significance. Layers of new minerals can be formed in a matter of months or years and thin layers can be dissolved equally rapidly. Whether and where precipitation or solution takes
102
P. G. F O O K E S ,
W. J. F R E N C H
& S. M. M. R I C E
Areas of upward leaching

_
SALT
Limest~ - - p ~ ( ~ ' ~
~ , . . _
.._.
PLAYA
SALINA
I Prevailing wind Oceanic salt
Limestone~_.j~-'G r~undw.ater ~/~~',.".:',:.V:.":/.
:::-..:..-,:; : - . - . .
........ ... . . . . . . . :..--:, ,-.:--: . . . 1 . _ Sea .~.:. i~::i-~!::.'~i-:~-.; ::"!.!:i:;i:-.:~,~~ i:'... ;: i ' . ' ; . . , : ~
------~
Aeolian transport of salt Groundwater movement
~ ~
Salty surface run off Salt deposits

I , ' . J
Capillary rise
FIG. 1. Cross-section showing the usual relationship of the three main types of surface salt concentrations (salt playa, salina, sabkha).
place depends upon the individual circumstances, which can be changed by the engineering works themselves. Excavations, piling, the formation of embankments or the laying down of areas of fill, can effect changes in the groundwater-sediment regime which alter the positions at which mineral solution or precipitation can take place. From the engineering point of view, the importance of these processes is the possibility of heave generated by crystal growth in layers, the disruption of road or building materials through the crystallization of salts within them, and the creation of an aggressive groundwater and soil environment even in initially salt-free sediments or fill. Conversely, particular layers or other domains of sediments such as a general or local cementation can be dissolved and transported giving potential for settlement. The common minerals that can be precipitated and dissolved at significant rates include calcite, gypsum, anhydrite and celestine (strontium sulphate); problems can also arise from the oxidation and reaction of sulphides, such as pyrite, leading to the formation of sulphates such as gypsum.
origin can be spread over large inland areas, contaminating dust and surfaces of all kinds including those of engineering materials and works. This contamination is dissolved in rainwater and added to the groundwater resulting in dissolved salts concentrations of exceptionally high levels, ultimately leading to the precipitation of both sodium chloride and gypsum or sometimes more exotic mineral species such as borates and nitrates. Cooke (1981) summarizes the near-surface production of precipitated salts as follows: (i) Deposition of salt from groundwater. (ii) Deposition of saline aerosols. (iii) Reworking from exposures of fossil salt deposits including the diapiric rise of salt domes to the surface. In drylands diurnal humidities are often low and winds may be consistent and strong so that the rate of evaporation from the surface is extremely high. Once salt has been deposited at or near the surface the capacity of the sediments to transmit moisture upwards is enhanced so that salt accumulation is accentuated. Salt accumulates at a level in the soil at which the water vapour concentration is reduced to low values and, in areas with a very dry atmosphere, precipitation may may take place several centimetres below a dry loose ground surface and may not be easily detected in a reconnaissance survey. In regions of high humidity the salts will appear on the ground surface as a distinctive crust. Laboratory experiments made by the authors have shown that where the relative humidity is reduced to 30%, salt precipitation occurs about 20 mm below the ground surface. Rainfall leads to the rapid solution of this surface deposit, but once salt has been precipitated it can reappear within a day or two of dry conditions being re-established.
Relevant features of the dryland environment

Factors affecting salt a c c u m u l a t i o n
The salinity of coastal seawater adjacent to drylands is high and variable, sometimes exceeding 10% sodium chloride, and the contiguous groundwater commonly becomes saturated with sulphates and halides. The groundwater composition is mostly influenced by the proximity of the sea and concentration through evaporation and reaction, but there may also be accumulation of ions from the seaward movement of run-off or infiltrational groundwater. Spray of marine
Q. J. eng. Geol. London, 1985, Vol. 18
GROUNDWATER GEOCHEMISTRY AND CONSTRUCTION IN MIDDLE EAST
103
The coastal environment--Sabkha

Many coastlines in arid areas are backed by broad coastal flats and have offshore islands which tend to make for lagoons with very small tidal ranges (e.g. Evans et al. 1969; Bush 1973). Clastic materials (such as the quartz sands and silts found on the northern shore of the Arabian Gulf, or the carbonate silts and sands of the southern Gulf shoreline) accumulate in the sub-tidal areas. In lagoons and other protected tidal areas terrigenous material and fine carbonate muds of algal origin are often deposited and the water salinity is greatly enhanced by evaporation. This can lead to the formation of rapid alternations of algal, clastic and evaporitic sediments. The supra-tidal flats, which have a gently inclined surface of around 1 : 3000, contain pinkish-brown, silty, shelly sands with surface crusts of halite. The flats are periodically inundated by
the sea and may have persistent shallow ponds close to normal high water mark but their landward end may be reached by the sea only once or twice a year. The surface is usually damp and salty and algal matts form. These marts have also been seen on made ground and can develop within months of the fill being placed. The kind of matt produced is related to its position on the sabkha and the frequency of marine incursions (Fig.
2).
The sabkha fiat is often interpreted as a deflation surface, from which the wind removes dry small particles, which develops parallel to the water table at a level controlled by the dampness of the sediments. Fragments of porous rock or concrete placed on this surface are rapidly comminuted through the crystallization of halite and gypsum within their pores and discontinuities so that the surface is returned to its equilibrium level. Small blocks can be virtually
FIo. 2. (a) Algal matts on the surface of the landward end of a sabkha in Dubai, UAE. (b) Layers of gypsum and carbonate sand in a sabkha near Benghazi. (c) Salt polygons and algal and gypsiferous crusts on a sabkha surface in Dubai, UAE. (d) Dune on a sabkha surface, Dubai, UAE, showing the rise of capillary moisture into the dune sand.
104
t,. G. FOOKES, W. J. FRENCH & S. M. M. RICE
FIG. 3. (a) Deterioration of Tertiary limestone blocks caused by salt crystallization (physical salt weathering). (b) Damage to foundation concrete by salt precipitation. (c) Salt damage to railway lines, Egypt. (d) Destruction of blacktop surfacing due to gypsum precipitation from underlying sabkha, Bahrain. (e) Deterioration of concrete caused by salt crystallization (physical salt weathering). (f) Deformation of three-year-old pipeline (rhs) due to salt damage to concrete supports with replacement pipeline alongside.
GROUNDWATER GEOCHEMISTRY AND CONSTRUCTION IN MIDDLE EAST destroyed in this way in a few years and concrete structures of small dimension also deteriorate rapidly (Fig. 3). The landward termination of the sabkha flat may be rocky or consist of blown sand. The rocky eminences show evidence of salt weathering with steps being eroded into the solid rock, while dunes often develop gypsum crystals (desert roses) high in the capillary fringe where moisture rises almost imperceptibly into the sands. Gypsum crystals have been observed in stockpiles of sand dumped onto sabkha surfaces where the capillary fringe reaches the surface. In Bahrain, desert rose growth formed a complete crust over such a pile of sand in what must have been at most a year or two. The engineering significance of this is obvious, for the sulphate can be precipitated as easily and rapidly in fill or concrete as it can in the natural occurrences. In areas which are only intermittently covered by the sea, halite makes a surface crust often in the form of polygonal ridges. The chloride may be deposited for some decimetres below the ground surface, in amounts which decrease downwards, and gypsum occurs close to the surface but often makes rhythmic alternations of layers. Anhydrite also occurs in these environments and tends to be precipitated in the higher, landward parts of the sabkha while gypsum occurs near the sea. From the engineer's point of view dehydration of gypsum could cause settlement while hydration of anhydrite might cause heave. The surface evaporation which produces the salt crusts and promotes a steady transference of coastal water into the area can also lead to the development of secondary minerals in the carbonate sediments. These include anhydrite, celestine, dolomite and magnesite with celestine accompanying the formation of dolomite from aragonite. The importance of recognizing the presence of these processes from the engineering point of view is that they point to an active chemical environment which can be affected by engineering works such that the equilibrium is disturbed and where solution may occur in one place and precipitation may begin in another. Experiments and field observations show that the establishment of new chemical equilibria can be achieved in two or three years though it might take a decade for substantial chemical change to result. The large-scale laboratory studies carried out by the authors have included the direct solution and reprecipitation of carbonate, sulphate and halite in soil bodies up to a metre thick; in these cases it required almost a year for the moisture profile to reach equilibrium. As a consequence of alternations in sea level many of the deeper, older sediments have features characteristic of the various modern sedimentary regimes. It is sometimes seen, for example, that sabkha and windblown sand facies alternate vertically
105
and that clastic marine sands and fluvial gravels interdigitate with the coastal sediments and their evaporite and algal layers. It is also common to find thin beds of gypsum, dispersed gypsum crystals, or a gypsum cement in foundation rocks excavated for major structures. These salts are usually in equilibrium with the local groundwater and are being neither dissolved nor enlarged in their natural state but the engineering works can change the groundwater composition or conditions and eliminate the equilibrium condition.
Inland saline environments~Playa and Salina

In topographic depressions of hot deserts (Fig. 1) the capillary fringe of the groundwater may reach the ground surface. Intermittent lakes may occur and evaporation may produce salt deposits analogous with those of the coastal sabkhas. The areas are sometimes termed inland sabkhas though the name playa is more often used, or salina where groundwater is of greater importance than rainfall. The water in these environments comes from two sources---periodic rainfall and groundwater (capillary rise and springs) and both sources bring soluble salts into the area. Halite and gypsum are again important minerals in these evaporites and there is sometimes an aerial zonation of the precipitated phases. Vertically variable sequences involving evaporites are also commonly seen, as for example in Death Valley, California, where massive gypsum is capped by anhydrite and overlies halite and a calcitic sediment containing trona (sodium carbonate). Borates and nitrates may also occur in this kind of deposit. The playa surface is much like that of the sabkha with salt polygons again being characteristic, but the salts are often strongly influenced by the local bedrock and therefore differ in composition from those related to marine coastal conditions.
Groundwater chemistry
The solubility of minerals in coastal water and groundwater
Typical compositions of coastal and sabkha waters are illustrated in Table 1. These waters are characteristically low in Ca 2+ and have high Mg 2+, SO42- and of course Na 1+, K 1+ and C11- but are not normally saturated in these ions. The solubilities of minerals such as calcite and gypsum are strongly affected by the ionic strength of the water and by conditions under which solution takes place. Figure 4 shows for example how the solubility of gypsum varies with the sodium chloride content of the
106
P. G. FOOKES, W. J. FRENCH & S. M. M. RICE The sodium chloride content of coastal sea water in the Arabian Gulf and similar arid regions is commonly around 4% by weight but the sulphate and calcium levels are normally much lower than is required to saturate the solution with CaSO4. Hence solution of sulphate will be brought about by the ingress of seawater into sulphate-bearing soils. Dilution of the contaminated seawater with fresh water will cause further solution of calcium sulphate but might lead to precipitation of calcium carbonate from the fresh water fraction. Evaporation of the contaminated seawater in drier sediments will bring about precipitation of the sulphate. In the presence of calcium carbonate the form of the solubility curve for CaSO4 is changed to sharpen the peak of maximum solubility in NaC1 solution but to reduce the solubility more rapidly at high NaC1 contents. Hence a solution nearly saturated with CaSO4 in siliceous soils for example will begin precipitating gypsum (or anhydrite) if it passes into carbonatebearing soils. Thus, if the behaviour of a groundwatersoil system following engineering works is to be understood it is essential to establish the following: (i) the composition of the water likely to be in or to enter the soils (or fill) and whether mixing of water or contact between different groundwaters are likely to occur; (ii) the mechanism and rate of flow of moisture through the soils and the nature of the flow paths; (iii) the mineral composition of the soils; (iv) the pattern of distribution of the minerals in the soils and whether contrasts in lithology occur; (v) whether evaporation of the groundwater is likely; (vi) whether the capillary fringe reaches or is close to the ground surface; (vii) whether the system can be described as chromatographic. Thus if groundwater is motionless, an equilibrium between solution and precipitation will be reached. If the groundwater is moved from one sedimentary environment to another it will eventually change its composition to come into a new equilibrium state. Any lithological boundary is likely to be significant from this point of view, but specially significant is the boundary between beds of strongly contrasted composition such as the junction between sulphaterich and sulphate-free layers or between rocks containing water which is anaerobic and those with water which contains oxygen. Examples of changes that occur at these boundaries are the formation of iron oxide layers at the oxygen surface and of crusts such as calcrete, gypcrete, dolocrete, ferrocretes and hardpans. The movement of groundwater into a new equilibrium state can be brought about by changes in
TABLE 1. Typical experimental chemical compositions of seawater and sabkha water Coastal, seawater, Open sea Arabian Gulf 420 1320 10 700 380 2700 19 300 75 420 1550 20 650 660 3300 35 000 170
Sample location Parts per million of: Ca Mg Na K SOn CI HCO3
Sabkha 1250 4000 30 000 1300 9950 56 600 150
solution at a given temperature and pressure, reaching a maximum at around 14% NaC1 by weight. The solubility is further affected by other phases present in rocks such as, say, calcite which reduces solubility because of the common ion effect (see French et al. 1982), by the presence of dissolved gases, such as CO2, and by the pH and Eh of the water.
1.0. Solubility curves : Gypsum only In presence of calcium carbonate
o~
>
0 u~ "ID
u~
r
0.5-
"---L__.
~ Results of field U Sabkha, N Africa 9 Sabkha, Gulf El Sea, N Africa 0 Sea,Gulf :~ Groundwater, N Africa :
O" o
lb
2'0
a'o
%NaCI
FIG. 4. Solubility of gypsum in saline water. Gypsum is precipitated from water compositions above the solubility curves and dissolved by water compositions below them. Q. J. eng. Geol. London, 1985, Vol. 18
GROUNDWATER
GEOCHEMISTRY
AND
CONSTRUCTION
IN M I D D L E
EAST
107
drainage pattern or by loading the surface with consequent upward transport of moisture by compaction.
and sandstones can also act as chromatographic materials. The chromatographic process can be described quantitatively (e.g. Hoffman 1972).
The influence of the carrier soils--chromatography

If the groundwater reaches a new static position then precipitation and growth of nodules or hardpans depends upon diffusion of ions to a growth centre and a counter diffusion of unrequired ions. This is a slow process commonly being of geological rather than engineering significance. However, if the groundwater is in steady motion, the solute is transferred by both diffusion and by mass transport. The influence of flow rates on growth of calcite concretions is illustrated in Fig. 5. Gypsum growth rates are more rapid; with a given water velocity, gypsum will accumulate about one hundred times faster than calcite. The curves also give an indication of the rate of solution of equidimensional crystals in undersaturated water. However, the position of the curves is strongly influenced by other ions present such as sodium, chloride or carbonate and by acidity. Growth of new phases can also occur through reactions which remove ions from migrating groundwater. Where the flow is through a microporous inert (insoluble) medium it can become chromatographic and highly localized concentrations of solute can be created which eventually create strongly layered and often monomineralic precipitation. An example of such a system would be a dense, poorly sorted, siliceous or carbonate clayey, silty sand of low porosity but with many micropores and intergranular surfaces. Microporous limestones
The engineering aspects of this environment Mechanical behaviour of sabkha soils

The mechanical behaviour of the sabkha sediments varies with sediment type and position. The typically shelly silty sands are poorly sorted (i.e. well graded, e.g. Fig. 6 compared with windblown sand) and can therefore be compact. Generally, cementation is absent or weak and Standard Penetration Test values tend to be moderate to very low. At topographic levels close to high water mark the upper beds might not sustain heavy construction equipment and N values can be zero and certainly less than 5. Where the sediments are somewhat drier they can become very weakly cemented by evaporites such as halite or gypsum and this, together with the usual good engineering grading of the sediment, increases the SPT results to 30 or 40. Stronger beds can be produced by the development of coarsely crystalline gypsum; layers around 500 mm thick, which consist of alternations of gypsum and sand, can be formed in this way. These beds can have a moderate to high ability to sustain loads when dry, but may be much weaker when wet. They can also be thoroughly impermeable and are able to withstand a modest head of water, at least in the short term. Movement of water under such a head can lead to the formation of
---- 40 E E
l-
V = flow velocity of groundwater
I00-Aeolian f
50
=
.g
9 30U
C 0 U
/V--lO-3m/s
20.
/e----- Sabkha sand
13 IIC
10-
~.o 0.06
,50 100 1~i0 200 Years of growth
i. /
/
0:2 0'.6 2'0
Porticle size
60 (ram) i Fine J M ISAND e d u im J C o a r s e FineJ IGRAVEL M e d u ir n J C o a r s e
FIG. 5. Influence of groundwater velocity on growth rate of spherical calcite concretions.
FIG. 6. Typical grading of sabkha sand and aeolian dune sand.

]08
P. G. FOOKES, W. J. FRENCH & S. M. M. RICE
small caverns, channels and surface holes, structures which can often be seen in fossil sabkha deposits (Fig. 2b). Excavations in normal sabkha sediments can be free standing, especially if they contain gypsum layers, and can be made readily by mechanical excavation. They tend however to collapse at about water table level and sand layers tend to be washed out from the sides. Older sabkha sediments can be variably cemented by fine-grained carbonate and thus develop a uniaxial compressive strength of a few megapascals. Thin gypsum beds preserved in such sequences are sometimes missed or are underestimated in drilling operations, but their presence or the presence of a gypsum cement may be highly significant in greatly
(a) Measured moisture content Single result for isolated band ~ ~" 2"6.3~3 45
reducing the rock strength below that anticipated from measurements of the dominant carbonate cemented materials. Much the same can be said of deeper and more thoroughly lithified sediments: the recognition of important layers of sulphate is often impaired by losses in sample recovery.
Capillary rise and engineering works

The restoration of the capillary rise after excavation and refilling or other engineering works brings with it the addition of soluble salts to the new materials and can restore the original potentially aggressive ground conditions. As an illustration of this, Fig. 7 shows the profile of sulphate and chloride developing in
(b) Schematic variation of moisture content v. depth (after Meyboom 1967) 0Moisture transmission mainly by vapour -L Field capacity, i 4--- moisture potential ~ ~ [ t Moisture transmission
" ~ m a i n l y by liquid
\ ~._ rRange of water ~'=" I leve,'ls after "~"~.".~.~= ~ ~ r e c l a m at|on \\ Water-lOm deep before reclamation
' 10 ' 210 ' :~0
CARBONATE 1SAND AND SANDSTONE A E 2DREDqED FILL ~r~ ~ , ~,-~ i~"3FINE DUNE SAND 45
v G.W.L
"1
MC ( % )
MC (~
(c) Relation of chloride and moisture contents
(d) Relation of sulphate and moisture contents

) .I" t k I
1o
E 2l-
/..J /
i
~3E3 45 / i
c)
~34-1
t w
'
1'2
MC (%)
5-
o
Key :
~o
Chloride (as NaCl) x 100 ....
160 240 Sulphate as (SO3) x 100 MC(%)
Measurements immediately after
dewatering stopped (about 24

months after reclamation) after reclamation
Measurements about 33 months __.__ Measurements about 40 months after reclamation

FIG. 7. Moisture, chloride and sulphate contents changes in hydraulic fill reclamation at Mina Jebel Ali; see trial pit 1, Table 6 and Fig. 16. Dredged fill consists of dredged weak calcarenite and marine carbonate sands.
GROUNDWATER
GEOCHEMISTRY
AND
CONSTRUCTION
IN M I D D L E
EAST
109
hydraulically placed marine fill underlain by a relatively consistent water table. The fill consisted of dredged carbonate sand and calcarenite and was placed while the local water table was lowered by dewatering. After dewatering ceased, the water level rose and stabilized rapidly and a series of trial pit excavations was started to investigate the variations in moisture, chloride and sulphate content with time. Plots (c) and (d) compare measured chloride and sulphate content with moisture content. Where the plotted result exceeds the solubility of the relevant mineral solvent containing these and other ions it is assumed that the salts must be present as solids. The concentrations at which precipitation of sulphate could occur can be estimated from Fig. 4. The data for the trial pits shows that there is some evidence of a general increase in chloride level with time and apart from higher values near the ground surface, the concentrations of chloride in the pore water are similar to those found in sea water. The sulphate results clearly indicate that substantial amounts of calcium sulphate have been deposited in the short time since the water table restabilized. Figure 8 summarizes some data obtained from 31 sampling points taken on a rectangular grid in an urban area of silty sandy sabkha. The information was required to assess the amount of water table drawdown which would be needed to prevent salt attack on existing concrete structures. The work consisted of examining the site to classify the ground condition at each point and measuring the depth to existing groundwater by hand-auger holes. In this way the capillary rise was ascertained and the necessary water level drawdown determined. The results appear to show that the capillary rise over the site varied from (a) 0.5 m to (b) 1.3 m.
Apparent height of capillary rise above ground water level (m) 4 |
I I
I I
result |
i i i i 1 , | t i| I , , , = i i i ,I
0.01
O.1 __1 C.Silt _1_

I
I~,
M.Silt
F.Sand
1.0 D10mm _l_ M.Sand _[_C.Sand

I I
Capillary rise Symbol: estimated by: |
Material Coarse sand ) to silt /
Notes
After Lane & Washburn (1946) (lab conditions) Coarse gravel| Little or no to silty gravel,I evaporation. 20"C Surface dampness Coastal sebkha, 19 Field Arabian gulf measurements temp. range 15-20"C Coastal dunes, 5 Field Arabian gulf measurements mean temp 25* Dredged carbonate sand and sandstone Coastal dune sand, Araoian gulf Results from Fig.7 temp.range 20-45"I2 High sulphate levels at ground surface temp.range 20-45*(2
r,,\\-~
x 6
Dampness on trial pit sides Levels of sulphate concentration Levels of sulphate concentrat ion
FIG. 9. Estimates of capillary rise by various methods.
No.of results 101 a
~ Sabkha Not Sabkha
5-
i / / / /
/7
v/~
1.o
2.0
Depthto GWLlml
3'.o
[-q
i
FIG. 8. Observed relationship between depth to Ground Water and occurrence of Sabkha. In these particular materials, Sabkha develops only when depth to Ground Water is less than about 1.3 m.
Figure 9 compares estimates of capillary rise by various methods with the grading of the soil. Laboratory results agree reasonably with field observations where these are based on the occurrence of visible dampness, although capillary rise does not fall off as rapidly with increasing grain size in the field as in the laboratory. However, sulphates can be deposited well above the visible limit of dampness and the existence of visible dampness is not itself a reliable guide to the extent of potential contamination in borrow pits or to potential attack on structures. Intermittent rain or watering during road construction, for instance, would assist the transport of salts in normally dry soil. It is also not uncommon to find desert rose gypsum growing in loose apparently dry sands well above sulphate-rich groundwater zones. This effect is brought about by the production of thin hydrous coatings to soil particles during moist atmospheric conditions which allow salt transference along grain boundaries and thus the establishment of
110
P. G. F O O K E S .
W. J. F R E N C H
& S. M. M. R I C E
permanent grain-surface layers capable of transmitting soluble ions.
Flowing groundwater
Reaction, precipitation and solution rates are greatly enhanced by flow of groundwater and hence are sensitive to the permeability and porosity of soils. Soil permeability is influenced by the percentage of silt sizes, degree of cementation, and crystal growth within the pores. A value of approximately 10-5 m/sec has been found to be typical for the fine sands which are common along arid shorelines--in sabkhas and in salinas. If the topography is flat the gradient of the phreatic surface will almost invariably be less than 1%. The true velocity of groundwater flow is unlikely then to exceed 10 -6 m/sec (allowing for typical void-ratios and tortuosity of pores). Thus high velocities of groundwater flow do not occur naturally in the arid environments under consideration but can be associated with works such as cofferdam construction or temporary dewatering for foundations or basements. In these cases the local hydraulic gradient may approach unity. There is extensive literature on damage to dams caused by water loss through sulphate-rich rocks under the high head applied by the impounded water (e.g. Calcano & Alzura 1967; Brune 1965). Particularly serious effects can occur in fissures where there are high local flow rates. Solution widens the fissures until water losses are unacceptable or the structure is undermined. This may be accentuated by mechanical piping in the terminal stages. Similar failures can also occur where the soluble salts are dispersed in permeable strata though here the total amount of soluble material may be higher before significant damage occurs. James & Lupton (1978) have published experimental data and suggested methods of calculation which enable dissolution rates of gypsum and anhydrite to be estimated for interstitial flow and flow through fissures. For a given flow regime and disposition of soluble minerals, the rate of solution depends on the temperature and chemistry of the percolating water. For interstitial flow, James & Lupton (op.cit.) discuss the concept of a solution front which moves in the direction of flow. The front is a surface delimiting the rocks from which the soluble components have been removed.The width of the gradational zone at this boundary is usually small and in rocks of uniform composition and containing gypsum, the sulphate concentration downstream of the front is unaffected while upstream practically all the sulphate has been removed. The rate of advance of the solution front into the rocks containing the soluble phase can be calculated from
where U is the rate of advance of the solution front, k is the permeability, i is the hydraulic gradient and C is a constant depending on amount and disposition of sulphate particles and solubility in the percolating water. James & Lupton's method takes no account of the effects of the chemical recombination or precipitation of new minerals within the flow regime. Laboratory work carried out by the authors has shown that this can be a significant factor even under high hydraulic gradients. For instance, laboratory tests on samples of sabkha sand containing gypsum and anhydrite (hydraulic gradient about unity) have shown that the establishment of a solution front is prevented, in the short term at least, by conversion of the anhydrite to gypsum. This can lead to an abrupt decrease in permeability due to the increase in volume associated with the reaction. Work by Akili & Torrance (1981) has included laboratory investigations of the increase in penetration resistance of a sand when calcium salts are precipitated in the voids. It is interesting to note that they were unable to simulate field reactions which occur in carbonate soils, presumably because of the rates of the reactions involved, and used a chemical precipitation method to deposit salts in the sand. A knowledge of reaction and solution rates is clearly of great importance from the engineering point of view but it is apparent that the most important factors relate to the nature of the deposits and the nature and rate of the solvent flow and these must be studied in a site investigation of the special arid environments.
Consequences for civil engineering Ground investigation

Soils and rocks containing evaporites can vary greatly over short distances; variations in conditions can markedly affect the equilibrium chemistry and physical properties of the sediment. For this reason it is often preferable to carry out the relevant site investigation work in several stages. The preliminary investigation provides information which helps in the planning of subsequent stages so that work is carried out effectively and economically. Areas of upward leaching and lateral flow are potentially the most hazardous and require particular attention during both the investigation and the construction. (Fookes & Higginbottom 1980; Fookes & French 1977). A typical plan of investigation for an area of sabkha might be as follows: (i) Walk/drive/fly-over survey and desk study. (ii) Mapping the area and its surrounds at a standard justified by the project. The themes recorded might indicate geology, geomorphology, hydrogeology as well as topography,
u-
k.i (1 + c)
(1)
GROUNDWATER
GEOCHEMISTRY
AND CONSTRUCTION
IN M I D D L E
EAST
111
surface phenomena such as gradients, water courses, outcrops of solid rock, and ground moisture zones. The standard could range from a single sketch plan to detailed geomorphological mapping identifying the different land forms present and the processes by which they were formed and are being altered at the present. Brunsden et al. (1979) and Doornkamp et al. (1979) give examples of geomorphological mapping; reference can also be made to the Geological Society Working Party report on land surface evaluation for engineering purposes (1982) and to the excellent text by Cooke et al. (1982). (iii) Widely spaced static penetration tests to define broad variations. (iv) Boreholes and sampling, taking particular care to look for and recover potentially troublesome materials. "Fill in" penetration test pattern. (v) Laboratory testing. The combination of static penetration tests and boreholes often proves useful in sandy materials where it is difficult to obtain undisturbed samples. A preliminary classification of the area into simple geomorphological units during the initial desk study will assist in siting the test and boreholes so as to obtain the optimum information. When considering the behaviour of granular materials, it is useful to distinguish between mechanically and chemically stable soils. The former derive their strength predominantly from aggregate interlock while the latter are dependent to a significant degree on chemical cementation which is often due to the presence of chlorides, sulphates or carbonates (Fookes & Higginbottom 1980). The degree of cementation may vary seasonally and may be subject to leaching or softening leading to strength loss in wet conditions. Construction development can change local conditions markedly. Roads may intersect seasonal water courses which are almost undetectable during dry parts of the year. Paving or filling areas may reduce evaporation from the ground surface. This, and discharge of waste-water in urban development, may lead to changes in equilibrium groundwater level of a metre or more. Chemically stable sandy soils are particularly sensitive to changes in groundwater chemistry, but even indurated deposits will be affected, albeit more slowly. The mechanical properties and chemical composition of soils can change rapidly, particularly in the vertical direction. Common examples are windblown sands overlying older gravels and high levels of evaporite in the capillary fringe bounded above and below by relatively 'clean' material. Variations can be investigated by boreholes with continuous sampling or by sampling at close vertical intervals from trial pit
sides. A sampling interval of about 300 mm would probably be appropriate for trial pits through the capillary rise in beach sand. For further discussion, see Fookes & Collis (1975b).
Testing
Investigation stages (i) to (iv) given above should provide enough information to identify the geology of the site and surrounding area, the dynamic processes at work and to assess the probable impact of proposed development on these processes. Laboratory testing may then be carried out to provide more information on material properties and, together with testing during the works, as control on fill placement and compaction. The results of standard tests can be misleading. This is because of the frequent occurrence of crystals which contain water of crystallization and are susceptible to crushing, abrasion or solution during tests, the destruction of cement during sampling and remoulding for laboratory testing and by solution or abrasion during the tests, and the presence of shells and shell fragments (common constituents of these soils) which are susceptible to crushing. Tables 2 and 3 summarize possible errors in standard test results from the above causes and suggest suitable courses of action. The information is not exhaustive and the tables were developed to illustrate the most common sources of error which can give rise to erroneous engineering decisions. Chemical testing of soils in the areas under consideration poses difficulties not commonly experienced in less saline soils. Serious errors can derive from failure to detect all the sulphate or chloride because there is too much present for the extraction procedures. It may be necessary either to reduce the amount of sample taken or to make several separate extractions. Both water and acid extractions should be made for cross-correlation and as a check on the total amount of chemical likely to be available with time. A second problem arises from sample size, grading and frequency. Significant local high levels can easily be missed unless samples sizes are large and sampling points are numerous. The collection schedule must be determined during the field study and a very high density of sample points will be indicated if, say, gypsum or anhydrite are identified in layers or as dispersed crystals. The presence of sulphates, chlorides and smectite clays in abundance also lead to errors in the estimation of loss on drying because of combined and absorbed water. The discrepancy between moisture contents determined by oven drying (which can dehydrate chemicals) and the Speedy moisture tester can easily exceed twenty-five per cent of the measured value (Fig. 10(a)). The moisture problem also affects sieve analysis (Fig. 10). BS 1377 (1975) details two test methods. Q. J. eng. Geol. London, 1985, Vol. 18
112
"I-,-~BLE 2. Influence o f local conditions on chemical, moisture content and grading test results Test Total Sulphate Water Soluble Sulphate Method BS 1377, Test 9 BS 1377, Test 10 Possible errors If large amounts of sulphate present, precipitate may be excessive. Variety of anions present may preclude use of the ion exchange column. Engineering significance Sulphates under-estimated. Advice Reduce weight of test specimen. See BS 1377, Test 9, Note 3. Modify test method. See BS 1377, Test 10, Note 2 which recommends Test 9 above after water extract made. Correlate with total sulphate. Correlate with total chloride. Discuss with chemist if problems suspected. Correlate with methods in ASTM C25-72 and VOGEL.
Only partial extraction of available sulphate, Water Soluble Chloride Total Chloride Carbonate Content BS 812, Part 4 or Quantab BS 1881, Part 6, Test 9 ASTM D3042 Possibly only partial extraction of available chloride, Interference from other ions (especially halides). Carbonate content by difference---other acid soluble material will be recorded as carbonate. Hydrated salts (e.g. CaSO42H20) may lose water of hydration (i.e. chemical water) on heating.
Under-estimate of sulphate available with time. Underestimate of chloride available with time. Possible errors in results (usually over-estimate). Possible significant overestimate of carbonate.
Moisture Content
BS 1377, Test 1A
(lO5OC)
Too high a moisture content Use lower drying temperaindicated. This would vary ture or use Speedy. depending on amount and Note: the gauge reading gives type of salts present, moisture content as % wet sample weight and correction factor should be applied. Long drying time needed. Run trials to define suitable drying temperature. Correlate with Speedy and keep checking sulphate and chloride contents. Keep bottle shaded. Calibrate regularly by comparing with BS 1377 method at intervals. Also check calibration of equipment by testing inert material of known moisture content. Compare with dry method if salts (especially NaC1) present. Reduce drying temperature.
As above, but lower drying temperature
Dehydration temperature of gypsum may be 42~ or less, in presence of other salts. Bottle overheats in direct sunlight. Low readings clayey soils. Mix sample thoroughly in bottle to get full reading.
Speedy Moisture Tester
High result.
Grading
BS 1377, Test 7A (wet sieving)preferred method
Loss of salts by solution.
Generally too fine a grading.
Oven drying involved, therefore possible loss of water of crystallization. Mechanical abrasion (not as severe as for Test 7B below). Too fine a grading.
Examine sample and make a judgement on amount of breakdown. Examine sample.
BS 1377, Test Weakly cemented material, 7B (dry sievsalt crystals and shells ing)--as a break down. check on Test 7A Fines may stick together. above
Too fine a grading.
Too coarse a g r a d i n g . )
G R O U N D W A T E R GEOCHEMISTRY AND CONSTRUCTION IN MIDDLE EAST TABLE 3. Influence of local conditions on classification and earthworks control test results Test Specific Gravity Method BS 1377, Test 6 Possible errors Oven drying involved therefore possible loss of water of crystallization. Loss of salts by solution. Engineering significance Too high an SG. Advice
113
Reduce drying temperature.
Too high an SG (see below).
No remedy possible unless paraffin or white spirit used in place of water. See BS 1377, Test 6(B). Note 4.
Specific Gravity and Water Absorption
BS 812, Test 2. Loss of salts by solution.
Too high an SG as soluble salts generally light. Error probably worse than for BS 1377 method. Too low an absorption. Reduce drying temperature.
Oven drying involved therefore possible loss of water of crystallization. Atterberg Limits BS 1377, Tests 2to 5 Salts may disintegrate, dissolve in pore water, or lose water of crystallization in oven drying.
Moisture content and grading in error.
Examine sample and make judgement on amount of breakdown. Reduce drying temperature and/or use Speedy method for comparison. In extreme cases, use natural groundwater for mixing.
Natural groundwater may be Index results altered, but highly saline but distilled effect depends on clay water is used in mixing type and is unlikely to be specimens, significant. Laboratory Compaction BS 1377, Tests 12, 13 or 14 Salts may disintegrate, weakly cemented material and shells may break up. See notes on moisture content testing. Oversize material removed from sample may contain large proportion of shells. CBR (disturbed) BS 1377, Test 16. ASTM D1883--73 Salts may disintegrate. Packing fraction altered. Too fine a grading given, packing fraction altered and generally too great a density achieved.
Examine sample and make judgement on amount of break up.
Examine.
Too fine a grading given, packing fraction altered and generally too great a density achieved. Low results on top.
Examine sample and make judgement on amount of break up. Test top and bottom.
Gap graded soils may segregate giving low compaction on top of sample. CBR (soaked, disturbed) BS 1377, Test 16. ASTM D1883-73 Salts may go into solution.
Dramatic decrease in CBR for some materials. Others increase.
Specify test if damp conditions or salts present. Soaking periods up to 10 days may be needed for full strength loss. Acceptance of material should be based on laboratory results. Use field test as index. Consider impact tester. Use Speedy which has been calibrated on salt free soil.
CBR (field test)
Difficulty in cleaning up test surface. Cannot usually test soaked material.
Generally different results from laboratory test.
Field dry density
BS 1377, Test 15
Moisture content in error.
Alters density, gives too high or low result.
114

(b) Effect of drying duration and temperature Fine carbonate sand, high gypsum content . . . . . . moderate . . . . 25 . . . . . . . small .... ......... Carbonate conglomerate, little gypsum, some clay infill to joints. 20(>50%) (~20~ (<10~ (<10%)
(a) Comparison of moisture content measured by oven drying and Speedy Tester. + Dredged carbonate sandstone - little gypsum 25:~ Sandy wadi gravel x "Sabka sand" 20-
o~
1510f 50
020
. . . . . . .
.*
"O 0J r
~ o ~ ~ 1
oven dry
e s s than
o~
10-
J~
"~
5 +
80 ~
=E
_11110~ -I -I
,5 lb 1~ 2'0 MC by oven drying ( % ; II0~
2'5
lb
2'0
3'0
4'0
5b
8'0
Drying time (h)
(c) Effect of specimen preparation on Index properties of two specimens ($1,$2) of marine carbonate silt with some clay. $1 $2 | r V &
..,.,.+
20-
Specimen prepared ot E] natural MC Specimen air dried Specimen oven dried
~, 1o+m
v
,-I
'~e
,.:.,, m
Open symbols- prepared with sea water Shaded symbols- prepared with distilled water.
~- 0
50
Liquid limit (O/o)
6'o
~o
(d) Comparison of B.S. wet and dry sieving results on shelly carbonate sand Dry sieving .... Wet sieving
100"
///
==
A
//
'06
0:6 Particle size(mm )
6;0
FI6. 10. Some examples of effects of salts on standard engineering testing methods.
GROUNDWATER
G E O C H E M I S T R Y AND C O N S T R U C T I O N IN M I D D L E E A S T
115
Using the method given in Test 7A (wet sieving) the soil is subjected to: (i) oven drying, (ii) soaking with dispersing agent, (iii) wet sieving through two sieves, (iv) oven drying and (v) dry sieving through a complete set of sieves. With some soils this can promote breakdown of shells by mechanical abrasion while sulphate and chloride crystals or cements may be affected by mechanical abrasion or solution. The mechanical abrasion caused by Test 7B (dry sieving) is greater than that in Test 7A. In addition, the oven drying can cause agglomerations of soil particles to form. These can be difficult to break down by hand as required in the British Standard, especially if the soil contains carbonate or significant amounts of silt or clay. Steps (ii) and (iii) in the wet sieving test limit the formation of these agglomerations. Comparative tests performed on fine carbonate shelly sand have indicated seven per cent silt content (dry sieving) rising to nearly thirty per cent if wet sieving is used (Fig. 10(b)). The ASTM test method for sieve analysis is similar to the B.S. method. A completely wet test which only involves drying the final sieved fractions has been adopted by the National Coal Board for testing colliery shale. This last method reduces abrasion and cementing to a minimum but is very time consuming. The results of index tests can be very sensitive to specimen preparation. For example, tests on a marine carbonate silt of high plasticity have shown a doubling of the plasticity index when oven-dried specimens were used in place of specimens prepared from their natural moisture content. Similar variations were observed between specimens prepared with distilled water and sea water, the former giving the higher plasticity index (Fig. 10(c)). The in situ California Bearing Ratio (Field) tests provide useful field checks but their value is severely limited by the inability to vary the test moisture content and by difficulties in obtaining a sample afterwards and the location of any oversize stones noted in relation to the plunger. Cementing of the top 10 to 20 mm may occur rapidly in 'salty' fill and will give high CBR results especially on drying. Looser underlying layers should be tested separately for their CBR value. Many of the practical difficulties of obtaining and interpreting field CBR results can be overcome by using lightweight impact testers such as the Clegg impact value tester. The main advantages of this type of equipment are: (i) it is portable and can be operated by one man; (ii) it does not require a reaction kentledge; (iii) tests are performed very rapidly before drying out and a large number of results can be obtained in the time taken for one CBR test. A more accurate picture of the standard of workmanship is obtained since a large number of results are available and unusually high or low values are easily identified.
However, if CBR values are required for quality control or design purposes, it is necessary to calibrate the impact test results against CBR test results. The correlation usually varies with the CBR/Impact value result and separate trials are needed for each material involved. The authors have little experience of testing cohesive desert soils but it appears possible that the impact value result would be sensitive to the soil saturation since the strength and stiffness of clays usually depends on the rate of load.
Influence of evaporite phases on design and construction

Urban developments in the Middle East situated in or near sabkhas or salinas and consequently in areas of high water table are difficult to drain. Discharge of domestic waste water through soakways causes a rise in water levels, perhaps with localized ponding. This presents a public health hazard, increases the aggressiveness of the salty environment, accelerates deterioration of building materials and, in flat terrain, the area over which the upper limit of capillary rise reaches the ground surface may increase markedly. Conversely, the installation of piped waste water disposal systems, together with relatively small scale land drainage of the worst affected areas, can produce a noticeable improvement in conditions. Urban sabkha flats have been seen to revert to a more 'dune like' condition within about a year of installing drainage. Roads are particularly susceptible to salt attack since routes of communication frequently follow the coast and the potential limit of capillary rise is often above the top surface of the road. Deterioration can be minimized by ensuring that the wearing course is impermeable to water vapour (a minimum of 40 mm of dense bituminous wearing course is suggested), or by placing the road on an embankment such that the pavement layers are above the level of capillary rise (Fookes & French 1977). Use of gravel layers as capillary breaks is suspect since salt crystallization may cause comminution of the gravel, but laboratory tests on manufactured filter fabrics such as 'Filtram' suggest they may have great promise (French et al. 1982). Figure 11 shows a notional harbour and related industrial development and illustrates some effects which may be observed in a region with carbonate rocks and carbonate superficial deposits. Table 4 summarizes the causes and the time scales involved and suggests appropriate action though the list is not exhaustive. Not mentioned in Table 4 is the observation that gypsum and halite, in common with many other materials, can undergo extreme creep deformations if subjected to suitable stresses. Depending on the stresses the rate of creep may decrease with time, remain more or less constant, or increase towards a failure condition. The stresses required to cause creep
Q. ). eng. Geol. London, 1985, Vol. 18
116
P. G. F O O K E S , W. J. F R E N C H & S. M. M. RICE
Storm Beach
Coastal Dunes
Sabhka
~ . . . . .
MLW.
9 . , . _ _ 9 1 . o , .
~,j,,. ~ ~
: ....:
~ . ~ " -
Variably cemented carbonate rock with evaporites and perhaps clays. NATURAL COASTLINE WITH AGGRADING BEACH Concrete walling
"--.
-:
. ' . . . . . . , ,.. , . . . . . .~ ~ : ~ .. . .. . . ., ~. .
|174 ",' , -~ . :.' '.' "." '."'.'5
M.H.W
M.~.w ~ (2 _ ~ . - , " -" ." . - .

I' t ' "/
...._~.-..-.,
.
.....
- - ~ ~ ~
:.
f "
J.__. Seabed dredged
. . . . . Area re_graded,reclaimed with dredged fill and paved.
--t
WHARFAGE CONSTRUCTED & SEABED DREDGED

l M.H.W M.L.W ' 9 "
""
I-, L , - ~ " " '
"" " " 9 . . . . . .......... , ........ _~,.~.,,,.-
I . / 1 _Concrete " w a l l l n g ~ ,..r--
...
9 .-~.-. . .." .-. . ~ . -
"" ~"~ ~~ -,L L . L.I - I,
___
/ernporary dewatering DRY EXCAVATION OF INNER HARBOUR BASIN
M.H.W
i.:,,~
I " "
"
"
.....
~-,,;=,,,;,~,.]
f"'." ~**,j.u~t~i " ,.,: : .
INNER BASIN FLOODED
LEGEN._. D
I l I I I l l | I l l =
Top of capillary rise. Direction ol water flow. Arrow length is indicative of velocity.
POSSIBLE EFFECTS (~) Q (~ (~ Chemical sulphate attack on concrete (flowing water) Physical sulphate attack on concrete/black top (wet/dry) Rapid ingress of chlorides into concrete with corrosion of reinforcement Heave ( salt movement ) Massive dissolution of salts and evaporites, undermining of foundations.
FIG. 11. Notional harbour development showing some effects of ground water salt chemistry. For summary of effects see Table 4.
Q. ]. eng. Geol. London, 1985, Vol. 18
GROUNDWATER T A B L E 4.
GEOCHEMISTRY
AND CONSTRUCTION
IN MIDDLE
EAST
117
Some engineering consequences of salt-water systems (the location numbers on the left refer to specific sites of potential problems as shown on Fig. 11)
Techniques to investigate and confirm occurrence Observe locations of attack w.r,t. ground water level. Chemical testing of groundwater. Testing of concrete possible but complicated, Preventative and remedial measures
Location (see Fig. 11) 1.
Effect Chemical Sulphate attack on concrete,
Cause Calcium compounds in cement react with sulphates to give expansion and disintegration, Flowing water helps to remove products and expose fresh concrete to attack. Can also occur in wet conditions if there are high initial sulphate levels in mix. Years
Timescale
Significance Varies from disintegration of surface and loss of reinforcement cover to total disintegration,
Existing concreteAttack might be slowed down by keeping as dry as possible. New concrete-Ensure concrete is dense and waterproof. Provide tanking. Max. total acid soluble sulphates in concrete (as SO3)-4% by wt. of cement
Physical Sulphate attack on concrete masonry, rip-rap and other stone structures. Physical Sulphate attack on road pavements.
Expansive forces set up as salts crystallize in pores of concrete and stone, As above.
Half to many years,
Surface deterioration. Loss of reinforcement cover, Surface deterioration. Weakening and susceptibility to mechanical damage. Pot holes, Salt blisters. Mass concretepossible slight weakening of surface layers. Reinforced concrete-continuing corrosion of reinforcement.
Observe locations of attack and local environment.
New concrete-Ensure concrete is dense. Use clean sound stone and aggregates. Tanking. Use clean, sound aggregates. Surfacing impermeable to water vapour (min. 40 mm DBWC). Elevate pavement above capillary rise. New concrete-Ensure concrete is dense. Use sound aggregates. Use cover of at least 75 mm in aggressive (salty, wet/dry) conditions. Tanking. Max. total allowable acid soluble chloride in concrete (as NaCl)--varies 0.1% to 1% by wt of cement.* Use O.P.C. Minimize changes in ground water equilibrium caused by construction. Provide thick floor and pavement structures.
As above.
As above.
Chloride contamination of concrete,
Salts in mix constituents, Migration of chloride ions into concrete especially if exposed to wet/dry conditions, Ingress may be increased where concrete is cracked due to thermal or shrinkage movements. Accumulation of soluble salts from ground water.
Months for ingress. Months to years for signs of reinforcement corrosion.
Chloride testing of concrete samples,
Heave
Months to years.
Blisters in thin blacktop, Movement of lightly loaded floorslabs, May vary from small movements to complete destruction of structures, Settlement and perhaps failure of foundations,
Hydration of massive or dispersed salt deposits (e.g. anhydrite), 4. Dissolution of soluble deposits, Solution of salts, especially under conditions of rapid flow.
Indeterminate, but perhaps months to years,
Level surveys, observe local environment, e.g, ground water level and changes since construetion), Chemical testing of soil samples to check changes with time. Level surveys, observe water flow paths, Chemical testing of water and soil to check changes with time.
Depends on flow path length, concentration and solubility of salts.
Install cut-off, impermeable membrane, or grout to reduce water flows.
* Varies depending on cement type and whether concrete is mass or reinforced.
Q. ]. eng. Geol. London,
1985, Vol. 18
118
P. G. F O O K E S . W. J. F R E N C H & S. M. M. R I C E
of gypsum and halite are much lower than for most other commonly encountered engineering materials and are probably similar to the stresses under which the salts were first deposited. Therefore, it appears that heavily loaded foundations on gypsum beds could suffer from excessive settlements. Heave pressures ranging from about 70 kN/m 2 to 2100 kN/m e have been attributed to gypsum crystallization (Lutenegger et al. 1979). Concrete is the major structural material in the Middle East and has been used for extensive marine works as well as buildings. It has already been discussed widely (Fookes & Collis 1975a, b, 1976; Fookes et al. 1981; Kay et al. 1981; Pollock et al. 1981). The quality of the finished concrete depends on locally available materials, the environment and the workmanship on site. Without care taken in construction, reinforced concrete near the coast typically has a life expectancy about half of that in more temperate climates. Significant increases in durability can be achieved by careful design and detailing, together with good supervision on site.
Uplift pressures on the dock floor are relieved by an underfloor drainage blanket and a system of open-jointed drains. Inflow under the piers is intercepted by pressure relief wells spaced at about 10 m intervals around the perimeter of each dock. The mean hydraulic gradient between the seabed and the adjacent line of wells is about 0.4. However, there is a loss of head in a layer of silt on the seabed; this reduces the hydraulic gradient in the foundation strata to about 0.2. The discharge from the entire dock floor drainage system is about 0.5 m3/sec. Dock 3 was dewatered in December 1976. About a year later, routine surveys carried out for the installation of surface works, showed that part of pier D was settling on the seaward side. The area involved is shown hatched in Fig. 12. An extensive investigation into the nature and cause of the settlement was carried out. This included: (i) precise level surveys, (ii) chemical tests on seawater and water taken from pressure relief wells, (iii) measurement of flow from pressure relief wells, (iv) boreholes with chemical testing of rocks, permeability tests, caliper logs and petrographic studies, and (v) borehole extensometer tests to monitor variations of rock compression with depth. Figure 13 shows settlement results for the seaward side and the dock side of the settlement area on pier D. Figure 14 is plotted to the same time scale and compares the concentrations of selected ions in seawater with those in water sampled from pressure relief wells adjacent to the settlement area. Generally, each plotted point is the mean result for four samples (sea water), or two samples (well water). The settlement results show clearly that, although the amount of movement is similar on the seaward side and the dock side, they did not settle concurrently. Both curves show a decrease in rate of settlement during summer and early autumn. This is thought to be caused by the seasonal rise in sea temperature which is sufficient to account for an increase of a few millimetres in the caisson height. The concept of a 'solution front' progressing from seabed to relief wells is not incompatible with the settlement results. The actual movement recorded at the pier surface at any time would depend upon the position of the 'solation front', the amount of material removed, the degree by which the foundation is weakened and the interaction of the pier caissons with the foundation. Using the method discussed by James & Lupton (1978) for interstitial flow, an attempt to predict the rate of advance of the 'solution front' was made using the site investigation data to define values of permeability, hydraulic gradient and gypsum content in the strata. The value used for 'C' in equation (1) was derived by the method given by
Case histories
Dubai Dry Dock--flowing groundwater
Dubai Dry Dock is generally founded on relatively uniform carbonate sandstone (calcarenite) consisting of fine to medium sand grains together with less worn shell fragments. The calcarenite was considered to be deposited as dunes with minor fine lagoonal and intertidal marine cross-bedded often shelly deposits. These sediments have now been seen in numerous cores and excavations and are known to be highly varied in the vertical sequence and to show very localized facies variation. The majority of the rocks are weak but have a calcitic matrix. Some are shelly and have moulds left following solution of shells. Other layers are of very weak or uncemented silt, sand or carbonate elay. There are also discreet discontinuous layers which though thin are very rich in gypsum. In excavations some of the sediments show piped flow of water through channels which tend to occur at particular stratigraphic levels. They are relatively porous with typical values of 10 - 6 tO 10 - 4 m/sec for the coefficient of horizontal permeability. The layers which are macroporous could have locally higher permeability. Figure 12 outlines the layout of the dock. The dock piers and headwalls are formed by 127 precast concrete caissons. The dock pier caissons were floated into position and founded on the prepared seabed. After the piers were formed, bunds were placed across the seaward end of the docks, which were then dewatered to allow construction of the dock floors and gates.

Seabed dredged to - 1 1 . 5 m Dock Gates
119
'\
a
0 .I
; ./~
~
E
i
/J / / f/ /-/ /#, / / /-j
SEA
Area reclaimed with dredged sand and sandstone.
App~x.5OOm to original ~shoreline
0
I
100
I metres PLAN
200
I
M.S.L. v ,1.0m
I///////~tn
[///_////J
9 .//~~//../////,/~
f.m.carbonate sandstone
0
V./. pjer D.,./. I
Dock floor with underlying drainage blanket
-_12.3m |
The chemical tests show that, in the relief well water, calcium and sulphate were enriched and bicarbonate depleted, relative to sea water. Measured rates of flow tended to oscillate about a relatively constant value and the amount of suspended matter was small. With the advance of a solution front the difference in composition between the sea and relief well water would be more or less constant until the solution front had traversed the whole width of the pier. However, the interpretation of the chemistry of removal of gypsum from the sediments below the dry dock is a complex of many factors including the flow rate, the composition of the sea water, the size, shape and distribution of the gypsum, the kinetics of solution, and the other factors affecting solubility. The most important factor is the geological nature of the sediments. For the solution of gypsum to be effective its loss must allow settlement to take place. The rocks must therefore lose strength and become capable of compression. The most extreme condition is obviously where a single, perhaps thin, bed of gypsum is progressively removed. Equally serious could be the removal of gypsum crystals from an uncemented silt or sand, or from a silt or sand cemented only by gypsum. Conversely the removal of gypsum from the rocks which are otherwise strongly cemented--perhaps by calcite-might lead to little or no settlement. Both lateral and vertical detailed variations in the sequence of sediments can be of great significance in determining whether (and how) settlement takes place. The rates of flow could increase, as material is dissolved, if fissures are opened or porosity is increased. Conversely the porosity and permeability of the ground are affected by settlement of the pier, and the solution mechanism could be modified by chemical reaction. One possible such reaction might be CaCO3 + 8042- = CaSO4 + CO32-. Readings from borehole extensometers indicated that on the seaward side of the pier, compression was almost wholly confined to the strata less than 20 m below datum. On the dock side, compression appeared to be occurring between levels of about - 1 6 and - 2 3 m, though borehole caliper logs did not show any consistent evidence of less competent strata at these levels. Figure 15 summarizes the sulphate content of cores from boreholes drilled in 1978 and early 1979, i.e. during the period of settlement on the seaward side but before the dock side settlement commenced. The results show that sulphate has been removed from the strata at about 18 m below datum: (i) on the seaward side prior to 1978. (ii) under the pier centre between early 1978 and early 1979, (iii) on the dock side after 1979.
".~.'. 1 7 - 2 5 m L l ~((' v - - - . . . . ~'~.Under floor drainage ~.~," and perimeter pressure o relief wells.
0
dolomitic oo limestone o o conglomerate o o

0
0 I
20 I metres
40 I
SECTION
AT
X - X
FIG. 12. Layout of Dubai Dry Dock. James & Lupton and in this case is 348 (mks units). Table 5 shows a comparison between predicted and observed results. Considering the difficulties in assigning to the active zone representative values of sulphate content and permeability and the complications introduced by foundation-structure interaction, the agreement could be considered reasonable.
120
0
r'. G .
FOOKES,
W.
J.
FRENCH
& S. M .
M.
RICE
5O
100
'
'
'
'
~ 1979
'
'
, 1980
'
'
'
v 1981
1978
FIG. 13. Dubai Dry Dock--level results from Pier D settlement area. Small amounts of settlement have occurred elsewhere on the dock and it is likely that the basic mechanism for these settlements is the same as that described above, with the magnitude and rate of settlement depending on local variations in sediment chemistry and water flow rates. For settlement to occur as a result of undoubted gypsum solution, it is most likely to relate to the layers which are richest in gypsum, have the lowest degree of cementation, and through which water is likely to flow at the greatest rate. Such layers have been found, both from excavations and borehole records, to occur at particular levels. For TABLE 5. Comparison of observed behaviour at Dubai Dry Dock with solution front model GROUND CONDITIONS Soil/rock type Hydraulic gradient (i) Horizontal permeability, m/sec (from borehole tests) (k) Permeant Proportion of gypsum in soil/rock Variably cemented gypsiferous carbonate sandstone 0.2
10 - 4
Sea water 60%
RESULTS OBTAINED CALCULATED rate of advance of solution front mm/day (after James & Lupton 1978) OBSERVED rate of advance of solution front mm/day (from surface settlements)
About 15
these layers to affect the settlement observed they must extend across the caissons. Potential settlement could be enhanced if silt or clay grade material is also removed in suspension. This might happen where water is piped through restricted conduits but is less likely where the flow is through small disseminated pores. The conclusion therefore, based on the evidence available to date, is that settlement of the pier caissons was due to solution of sulphates by the groundwater regime. Solution occurred first at the upstream end of the flowpath and progressed downstream at the rate of about 15mm per day. However, the precise mechanism for settlement remains ambiguous. Petrographic examination of the core materials taken at various stages of the study of the settlement shows that gypsum is present as a minor amount of matrix material filling microscopic voids between nearly spherical grains. This gypsum is removed during the solution process, but the removal leaves space which shows no evidence of collapse, the integrity of the rock being maintained by the micrite (calcium carbonate) matrix. Hence it must be postulated that unless the compression of the sediment is elastic, removal of this gypsum matrix cannot bring about settlement of the order of magnitude required. The chemical data also show that the original sulphate levels varied erratically through the boreholes with maxima indicating the presence of layers particularly rich in gypsum. These maxima are not present in later sequences of analyses. It seems possible therefore that the removal or thinning of layers rather richer in sulphate has occurred and this is more likely to have led to the measured settlement. Different facies, and lateral
GROUNDWATER
1000"
GEOCHEMISTRY
A N D C O N S T R U C T I O N IN M I D D L E E A S T
121
E
Q.
0 2000-
1978
1979
1980
1981
E
v O~
1000 1978 2001979 1980
1981
E
Q,
v Oq
0 to ,1-
"-.-......,
100 1978
i
_ _ .
/---"--~
!
if,,
,,,/-
1979
1980
1981
4000-
E e~
v
/~---
....,~
I ~
''~"
--
....
---~
" - ' - ...,.
3000
A
E e~
e~
~
1978 10-
1979
1980
1981
"o
c
, ,
' l ' l
1978
1979
1980
1981
KEY
Well water
Sea water
. . . .
FIG. 14. Dubai Dry D o c k - - r e s u l t s of chemical tests on water.
122
P. G . F O O K E S ,
W. J. F R E N C H
& S. M. M, R I C E
Caisson
I
. . . .
Dock Floor
',
-,f
>o~ 2~
I1
sandstone
, C'i'C'
Relief Well
sojo/o g
-34
o SOzP/o) g
Ir
o so--
o so3 /o)
Key.
~-~
Drilled Jan/Feb1978 Drilled June 1978(envelopes of Drilled March 1979
results plotted)
FIG. 15. Dubai Dry Dock---summary of total sulphate contents from boreholes in settlement area. variations in the geology could lead to different settlement patterns. conglomerate. They were constructed in dry trenches ahead of the main excavations. As well as a large conventional site investigation, a substantial programme of geochemical monitoring was carried out by sampling from quaywall excavations, trial pits and piezometers installed in boreholes. After consideration of the initial results it was concluded that the most efficient way of assessing long-term changes in geochemistry was to monitor sulphate and chloride levels in groundwater. It was felt that this would identify local or regional changes of significance to the harbour structures in time to undertake more detailed investigation. In addition, changes in total sulphate and chloride content of fill were monitored. In general, sulphate levels in ground water were found to be fairly consistent at about 0.5%, while chloride levels in groundwater showed wider variation from about 3% to 15%. Figure 17 illustrates the regional variations before, during and after construction. An example of local changes in groundwater chemistry caused by flooding the quay wall excavations is plotted in Fig. 18 which shows a typical monitoring arrangement. During the period concerned the inner basin was still dewatered, which is the reason for the hydraulic gradient of about 1 in 60 apparent in
Mina Jebel Aft---capillary rise conditions

Mina Jebel All is a 70-berth harbour constructed on the desert foreshore about 39 km southwest of Dubai. The main excavation and filling operations were as follows: (i) a 17 km dredged offshore approach channel, (ii) a dredged entrance channel protected by breakwaters and reclaimed areas, (iii) a dredged outer harbour basin constructed immediately inland of the coastal dune ridge, (iv) a dry excavated inner harbour basin constructed in the sabkha area, (v) about 3500 hectares of reclamation. The total excavation was 110 million cubic metres. The sandy superficial deposits are underlain by carbonate sandstones of varying thickness (sometimes absent) and carbonate conglomerates. Figure 16 shows the layout of the site and the use of the different fill materials available. Most of the quay walls were made of mass concrete blockwork and were founded on the
123
SHORELINE
Dredged horbourbasin
[['-l:(:-.'~tl ~.~ii:.'~"i~
Dry excavated horbour basin
To Abu Dhabi
km
Oman
~ A r e a filled by mixed dredged sand, sandstone and conglomerate. Area filled by dry excavated sand,sandstone and conglomerate. Area filled by dry excavated sand. ~ ] r F r ~ Area filled by dry excavated cong)omerote (blasted and unblosted). Main geochemistry monitoring locations:
FA'] Trio I pits
( ~ Boreholes/p iezometers
I~G. 16. Mina Jebel All--general plan showing use of fill. Fig. 18. This produced a slow flow of fresh seawater into the salt-rich strata. Despite this, the salt contents recorded in groundwater analyses are remarkably uniform--at least in the short term. Low values recorded in early 1979 are thought to be caused by drilling water used during the installation of the piezometer. Similar results recorded elsewhere, together with a theoretical assessment of salt movement by diffusion through groundwater, suggests that many decades could pass before regional changes of engineering significance became apparent. In contrast to the discussion immediately above,
124
Sand and sandstone
P. G. F O O K E S ,
W. J. F R E N C H
& S. M. M. R I C E
11
0
~)e
Conglomerate 1] ~
0
X
0 A. Percentage sulphate (as SO3) in ground water sampled Sand and sandstone 15
1~
x
9A~
O O
X
z3
z3
Legend
Conglomerate 15]
0
0 "
O
chloride (as NaCI) in ground water
estimated from conductivity. Sampled 1979- prior to flooding harbour; Sampled December 1981after flooding harbour.
Sampled 1977-chloride
B. Percentage
Outer Basin I 1
sampled
._~ ="
Inner Basin = I 2 I 3 Distance from coast (km) I 4 I 5
FIG. 17. Mina Jebal Ali--summary of sulphate and chloride contents of groundwater. See Fig. 16 for locations.
changes in the chemical environment of soil deposition and fill can occur rapidly. Chemically clean fill can become contaminated within weeks or months of placing. One example has been shown in Fig. 7. Table 6 summarizes observations made at five locations around Mina Jebel Ali. There is evidence of salts being deposited in all cases but the rate of deposition and the level above groundwater varies according to the type of the natural ground, the type of fill (composition, grading, placed dry or wet) and the history of water level changes.
Conclusions
Factors of relevance to civil engineering works occur predominantly from the interaction between climate and geology and are reflected in geochemical effects. In summary they are: (i) In hot drylands, where the rates of evaporation are considerably greater than in wet or temperate lands, significant changes can occur at and immediately below the ground
GROUNDWATER
GEOCHEMISTRY
AND
CONSTRUCTION
IN MIDDLE
EAST
125
Dat__~UmOJexcavation - L ~ewatered- I
Quay wall
Quay wall excavation flooded
I I
=, |
! -a-~
c,
, ,
/
......
....
..............................
--.-..,-"
...-
-6'i
.1.0'
/r
I -
'
-I
1979
'
1980
'
1981
o.5.
- - - - .'7.."--" ...........................
...-"
,...'
.--
-5O
0
9.0-
Note. Readings taken approximately monthly. Minor fluctuations in readings not plotted. pH t e s t s discontinued
7.0| | v._ - = - ~ -
:;
....... ~ , .
. . . . . . 9 "7..'7.w.~'..'%. ~ .-. ~'..,--. ~'. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
F - - - ...q.---'-" ................
, ~ O~
9 :- _~_O~
10
uz
3
15
]
-- ~ ..................................................... - .--r..--.."~.--, ,'~-~--
=~
;5 v .~
O I ,"
........
Piez.
Symbol
:
t ,,." j ,,. ; , v , , 1979 , , , I ~ ' ' ; i , 1980 ,
,,1
B2(2) B3(2) , , , I
--.... ' ' ' ' ' I
1981
i-Dredged wall
|
+4"0
M . S . L + 1.1 v
/
I Quay
::;:.::.:ii!:iiii:::::.....
sandstone B1 _ ~- + 4 . 5
selected
fill 112 (1) (2) ~ + 5
r General
fill B3 (1) (2)~-+6.5
,' J l
l::::i
' ....
0.0
"1
9 . 9 . ".
..'.
"
"
,-,.o
v_lO 0
-14.5 v ir
l__
fiilJ /iiil 1!~1 84m
I ! I 113m Typical Section (Not to scale)
o oo o oo o oo ~ o ConglomerateO~ jOO_o_ o~ o~ I~ o oo o oo
o c
~ (
""
"148m
"
I-
FIG. 18. M i n a J e b e l A l i - - - q u a y w a l l c h e m i s t r y m o n i t o r i n g
arrangement
and typical results. 1 9 8 5 , V o l . 18
Q. J. eng. Geol. London,
126
P. G. F O O K E S ,
W. J. F R E N C H
& S. M. M. R I C E
TABLE 6. Mina Jebel Ali--Observations of capillary rise and salt deposition Location (see Fig. 16) TP1 Original ground Dune sand Reclamation material Water level changes Pits excavated March 1979 Dec. 1979 Aug. 1980 Comments Capillary rise steady at 2-3 m. Gypsum deposited about 1.5 m above water level. Rate of deposition slowing in 1980. (See Fig. 7). No consistent data. Water level below base of pit and fill is coarse. Average moisture content increasing. Some evidence of halite being deposited close to ground surface. Moisture contents low. Substantial deposition of gypsum for about 2 m above estimated post 1980 water level. Water level about 2.5 m below ground level. Capillary rise extends to surface. Gypsum deposited at surface. Capillary rise about 0.7 m. Gypsum deposited throughout capillary rise zone.
2.5 m dredged In area dewatered for quay sand and sandwall construction. Rewastone Placed tered Feb 1979. Water 1977. levels stabilized rapidly.
TP3
Sabkha
3.5 m dry excavated sand, sandstone and conglomerate. Placed 1978.
In area dewatered for quay wall construction. Rewatered July 1979. (See Fig. 18 for illustration of water level changes).
March 1979 Dec. 1979 Aug. 1980
TP5
Dune sand
3.5 m dredged Fluctuations in water level sand and sandcontinued to early 1980-stone. Placed caused by disposal of 1978. dredged fill in adjacent areas. Area of cut. Completed in 1978. Tidal fluctuations (range about 2 m).
Jan. 1980 Aug. 1980
TP6
Dune sand
Aug. 1980
TPA
Sabkha
2 m dry excavated conglomerate and sand. Placed 1978.
Not affected by construction.
June 1979
surface during the life of any engineering structure. In unfavourable circumstances, these changes can present a risk to the performance of the structure. Salts, small shells and other material present in samples taken for standard engineering tests can significantly affect the result of the test. (ii) The requirements for significant interaction between geochemistry and structures are typically a high water table and the presence of evaporite salts. Changes occur through salts being dissolved or precipitated in juxtaposition to the engineering structure, or through reaction between the salts and structural materials such as concrete. (iii) These changes can be simplified under three headings: (a) settlement, (b) heave, (c) chemical reaction. (iv) Settlements over and above those expected from normal geotechnical processes can occur due to solution of salts such as carbonates,
sulphates and chlorides in flowing water, and can cause severe disruption to structures within months or a few years. The rate and magnitude of settlement can be approximately assessed by existing geochemical methods. (v) Heave can be due to crystallization from groundwater or reactions between soils and groundwater. The rate and magnitude of heave can again be assessed from existing geochemical techniques. Disruption of lightly loaded structures may occur within months to a few years. (vi) Chemically aggressive environments for substructures and material in contact with the ground can occur extensively. Typical examples are physical and chemical sulphate attack on cementitious materials and physical salt weathering of road surfacing. Disruption can occur with months to a few years. (vii) Made ground in areas where water flow or capillary rise is significant can be altered by chemical changes occurring within a few weeks to a few years. For example, sulphate
GROUNDWATER GEOCHEMISTRY AND CONSTRUCTION IN MIDDLE EAST or chloride may be deposited giving rise to aggressive conditions in initially clean, inert fill. (viii) To combat the above conditions, site investigations in areas of high water table, where evaporation is significant and upward leaching can occur, should include an appraisal of the geology, ground chemistry and groundwater chemistry. Engineering sampling and testing methods require a little modification of those normally employed in temperate areas. (ix) Although many of the mechanisms involved are not fully understood, it is probable sufficient knowledge exists to evaluate the likely risks in any given situation and to design and construct safe serviceable structures.
ACKNOWLEDGEMENTS. The authors record their appreciation of assistance from Sir William Halcrow and Partners and Halcrow International Partnership and their permission to publish this paper. In particular, the help of Dr E. A. Kay and the contributions to site monitoring of Dr N. Ghosh, Mr I. Iacovou and Mr G. Pallai are acknowledged.
127
References
AKILI, V. & TORRANCE, J. E. 1981. The development and geotechnical problems of sabkha, with preliminary experiments on the static penetration resistance of cemented sands. Q. J. eng. Geol. London, 14, 59-73. BRITISH STANDARDS INSTITUTION. 1975. Method of test for soils for civil engineering purposes. BS 1377. BRUNE, G. 1965. Anhydrite and gypsum problems m engineering geology. Assoc. eng. GeoL 2, 26-33. BRUNSDEN, D., DOORNKAMP,J. C. & JONES, D. K. C. 1979. The Bahrain surface materials resources survey. Geogr. J. 145, (1), 1-35. BUSH, P. 1973. Some aspects of the diagenetic history of the sabkha in Abu Dhabi, Persian Gulf. In PURSER, V. H. (ed.) The Persian Gulf. Springer-Verlag, Berlin 395407. CALCANO, C. E. & ALZURA, P. R. 1967. Problems of dissolution of gypsum in some dam sites. Bull. Venezuelan Soc. Soil Mech., Fdn Eng., July-Sept., 1967. COOKE, R. U. 1981. Salt weathering in deserts. Proc. Geol. Assoc. 92, 1-16. , BRUNSDEN, D., DOORNKAMP,J. C. & JONES, D. K. C.
1982. Urban Geomorphology in Drylands. Oxford University Press, Oxford. DOORNKAMP, J. C., BRUNSDEN, D., JONES, D. K. C., COOKE, R. U. & BUSH, P. R. 1979. Rapid geomorphological assessments for engineering. Q. J. eng. Geol. London, 12, 189-204. EVANS, G., SCHMIDT, V., BUSH, P. & NELSON, H. 1969. Stratigraphy and geologic history of the sabkha; Abu Dhabi. Sedimentology, 12, 145-59. FOOKES, P. G. COLLIS, L. 1975a. Problems in the Middle East. Concrete, 9, (7) 12-17. & 1975b. Aggregates and the Middle East. Concrete, 9, (11) 14-19. &~ 1976. Cracking and the Middle East. Concrete, 10, (2) 14-19. & FRENCH, W. J. 1977. Soluble salt damage to surfaced roads in the Middle East. Highway Eng. 24, 10-20. & HIGGtNBOTrOM, I. 1980. Some problems of construction aggregates in desert areas, with particular reference to the Arabian peninsula. 1. Occurrence and special characteristics. 2. Investigation, production and quality control. Proc. ICE, 68, Part 1, 39-90. ~, POLLOCK, D. J. & KAY, E. A. 1981. Middle East Concrete. Rates of deterioration. Concrete, 15, (9) 12-19. FRENCH, W. J., POOLE, A. B., RAVENSCROFT, P. & KHIABANI, M. 1982. Results of preliminary experiments on the influence of fabrics on the migration of ground water and water-soluble minerals in the capillary fringe. Q. J. eng. Geol. London, 15, 187-199. GEOLOGICALSOCIETY. 1982. Working Party Report on land surface evaluation for engineering purposes. Q. J. eng. Geol. London, 15, 265-316. HOFFMAN, A. 1972. Chromatographic theory of infiltration metasomatism and its application to felspars. Am. J. Sci. 272, 69-90. JAMES, A. N. & LUPTON, A. R. R. 1978. Gypsum and anhydrite in foundations of hydraulic structures. Geotechnique, 28, 249-72. KAY, E. A., FOOKES, P. G. & POLLOCK, D. J. 1981. Middle East Concrete. Deterioration related to chloride ingress. Concrete, 15, (11) 22-8. LANE, K. S. & WASHBURN, D. E. 1946. Capillary test by Capillarimeter and by soil filled tubes. Proc. Highw. Res. Ed. 26, 460-73. LU'I~NEGGER, A. J., WOLLENHAUPT,N. C. & HANDY, R. L. 1979. Laboratory simulation of shale expansion by induced gypsum growth. Can. Geotech. J. 16, 405-409. MEYBOOM, P. 1967. Hydrogeology; ground water in Canada. Rep. geol. Surv. Canada. Econ. Geol. 24. POLLOCK, D. J., KaY, E. A. & FOOKES, P. G. 1981. Crack mapping for investigation of Middle East Concrete. Concrete, 15, (5) 12-18.

Fookes Et Al 1985 QJEG Dubai Etc Evaporites

Uploaded by

Copyright:

Available Formats

Fookes Et Al 1985 QJEG Dubai Etc Evaporites

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fookes Et Al 1985 QJEG Dubai Etc Evaporites

Uploaded by

Copyright:

Available Formats

Q. J. eng. Geol. London, 1985 Vol. 18, pp. 101-128.

Printed in Northern Ireland

Areas of upward leaching

I Prevailing wind Oceanic salt

Limestone~_.j~-'G r~undw.ater ~/~~',.".:',:.V:.":/.

Aeolian transport of salt Groundwater movement

Salty surface run off Salt deposits

Relevant features of the dryland environment

GROUNDWATER GEOCHEMISTRY AND CONSTRUCTION IN MIDDLE EAST

The coastal environment--Sabkha

t,. G. FOOKES, W. J. FRENCH & S. M. M. RICE

Q. J. eng. Geol. London, 1985, Vol. 18

Inland saline environments~Playa and Salina

Sample location Parts per million of: Ca Mg Na K SOn CI HCO3

Sabkha 1250 4000 30 000 1300 9950 56 600 150

The influence of the carrier soils--chromatography

The engineering aspects of this environment Mechanical behaviour of sabkha soils

V = flow velocity of groundwater

/e----- Sabkha sand

60 (ram) i Fine J M ISAND e d u im J C o a r s e FineJ IGRAVEL M e d u ir n J C o a r s e

FIG. 5. Influence of groundwater velocity on growth rate of spherical calcite concretions.

FIG. 6. Typical grading of sabkha sand and aeolian dune sand.

P. G. FOOKES, W. J. FRENCH & S. M. M. RICE

Capillary rise and engineering works

(c) Relation of chloride and moisture contents

(d) Relation of sulphate and moisture contents

Chloride (as NaCl) x 100 ....

160 240 Sulphate as (SO3) x 100 MC(%)

Measurements immediately after

dewatering stopped (about 24

Measurements about 33 months __.__ Measurements about 40 months after reclamation

Q. J. eng. Geol. London, 1985, Vol. 18

Apparent height of capillary rise above ground water level (m) 4 |

O.1 __1 C.Silt _1_

1.0 D10mm _l_ M.Sand _[_C.Sand

Capillary rise Symbol: estimated by: |

Material Coarse sand ) to silt /

FIG. 9. Estimates of capillary rise by various methods.

No.of results 101 a

~ Sabkha Not Sabkha

permanent grain-surface layers capable of transmitting soluble ions.

Consequences for civil engineering Ground investigation

Q. J. eng. Geol. London, 1985, Vol. 18

P. G. FOOKES, W. J. FRENCH & S. M. M. RICE

As above, but lower drying temperature

Speedy Moisture Tester

BS 1377, Test 7A (wet sieving)preferred method

Loss of salts by solution.

Generally too fine a grading.

Examine sample and make a judgement on amount of breakdown. Examine sample.

Too fine a grading.

Q. J. eng. Geol. London, 1985, Vol. 18

Reduce drying temperature.

Too high an SG (see below).

Specific Gravity and Water Absorption

BS 812, Test 2. Loss of salts by solution.

Moisture content and grading in error.

Examine sample and make judgement on amount of break up.

Dramatic decrease in CBR for some materials. Others increase.

CBR (field test)

Measurements about 33 months . Measurements about 40 months after reclamation