Desalination and Water Treatment www.deswater.com

doi: 10.5004/dwt.2020.25505

Strategic water reserve using aquifer recharge with desalinated water in Abu Dhabi

Mohamed A. Dawoud

Water Resources Department, Environment Agency - Abu Dhabi, UAE. P. O. Box 45553, email: [email protected]

abstract United Arab (UAE) as well as other in arid mainly relay on desalinated water for domestic water supply. The challenges facing UAE, however, is the vulnerability of desalination plants to pollution and emergency conditions. There should be available alternatives for reserving freshwater sources for emergency and peak demand conditions. Aquifer storage and recovery (ASR) technique has been proposed as a cost-effective large water storage alternative that can help to meet the needs of domestic sector in crisis/emergency situations. ASR technique is used to store excess desalination water during non-peak hours for recovery during emergency or peak hours. In an attempt to capture, store and redistribute desalinated water and to regulate the quality, quantity, timing and distribution of water flows, Abu Dhabi Emirate started in 2002 to study and evaluate the construction of ASR pilot project in western region. The initial target injection of pilot ASR facilities is to store 2.5 MIGD using five injection wells and one infiltration basin. Desalinated water is infiltrated into a desert dune sand aquifer using “sand-covered gravel-bed” recharge basins. In this study, we evaluate the hydrogeological and hydrogeochemical stratification of the (sub)oxic target aquifer, and water quality changes of DSW during trial infiltration runs. A three dimensional model was developed to assess the impact of hydrologic and operational parameters and factors on the recovery efficiency of the pilot ASR experiment. The model was integrated with geodatabase tool to produce an easily updateable model. Results from this study demonstrate the interaction between hydrologic and operational parameters on the predictions of recovery efficiency. The pilot observations and modeling results demonstrate that in scenario A recovered water quality still complies with Abu Dhabi’s drinking water standards (even up to 85% recovery). Keywords: Three-dimensional modeling; Geodatabase; Visualization; Artificial recharge; Water supply; Abu Dhabi

1. General background and improved technologies are all needed to explore new approaches that are more cost-effective to desalinate water. Aquifer storage and recovery (ASR) involves the injec- It has been argued that the best long-term solution for the tion of freshwater in an aquifer through wells or infiltra- water crises in the domestic sector is to build a network tion basins for the purpose of creating a subsurface water of large-scale desalination plants. The problem faced by supply that is recovered at a later time, to meet seasonal, the GCC countries, however, is the vulnerability of desali- long-term, emergency, natural crises or other demands nation plants to pollution, natural crises, and emergency (Pyne, 2007). United Arab Emirates (UAE) as well as other conditions. The possible alternatives for reserving fresh Gulf Cooperation Council (GCC) Countries mainly relay water sources for emergency and peak demand condi- on desalinated water as the main source of fresh water for tions are: (1) to increase ground reservoirs and distribution domestic, sustainable development and security of their network storage capacity or (2) using a groundwater ASR communities. Moreover, desalination is still considered as system. It has been proved that increasing the capacity of very expensive source of water compared with other nat- the ground reservoirs and network is very expensive and ural resources. Research and development, new ideas, not environmentally friendly. One solution is to store this

Presented at the 13th Gulf Water Conference – Water in the GCC: Challenges and Innovative Solutions. 12–14 2019, Kuwait

1944-3994/1944-3986 © 2020 Desalination Publications. All rights reserved. 124 13th Gulf Water Conference Proceedings / Desalination and Water Treatment 176 (2020) water in groundwater aquifers using aquifer storage and 1985; Missimer et al., 2002; Reese, 2002). For example, aqui- recovery technique. For example, the maximum stored fer transmissivity, the product of permeability and aquifer water in the ground reservoirs and distribution network is thickness, must be high enough to permit injection under enough only for 24 h, except in Saudi Arabia and Kuwait, reasonable pressures and to allow for high flow recoveries. where it is 3 and 5 d, respectively (Fig. 1; Dawoud 2008). However, if the transmissivity is too high, injected water Thus, in any crisis or emergency, the stored water will not can migrate excessively (sometimes beyond the capture zone be enough to cover the demand. ASR has been explored of the ASR well) and in turn reduce recovery efficiencies. in diverse settings worldwide (Bichara 1974; Khanal 1980; In Abu Dhabi, the pumping test evaluations revealed Bouwer et al. 1990; Calleguas 2004; Artimo et al. 2008). a range of the transmissivity from some 100 to 3,000 m²/d, Aquifer storage and recovery is used to overcome the with an average of 1,065 m²/d and a median of 950 m²/d. groundwater depletion and unavailability of strategic fresh The lower values mostly refer to wells, in which the deeper water reserve (Heilweil 2005), store and recover ground­ section of the aquifer was tested. The calculated geohy- water (Lowe 2005), provide seasonal and long-term storage, draulic conductivity varies between 2 and 106 m/d with and regulate and improve water quality (Topper et al. 2004). an average of 27 m/d and a median of 28 m/d as shown in It is particularly useful as a purification method for surface Table 1. Therefore, the optimum transmissivity is within a and wastewaters (Rüetschi and Wülser 1999; Brissaud 2003; limited range depending on desired pumping rates and Al-Katheeri 2007). Artificially recharged water can be intro- recovery efficiencies (Missimer et al., 2002). In most hydro- duced into aquifers in various geological settings, such as logic settings, the simplistic “underground storage tank” river basins (Lowe et al. 2003), fractured sandstone bedrock conceptual model applied to ASR system may have lim- (Heilweil 2005), and esker aquifers (Artimo et al. 2003). ited utility. Multiple processes including physical, chemical In an arid such as the Emirate of Abu Dhabi with and biological, which can degrade water quality and quan- no permanently existing natural surface water and limited tity, control the ASR efficiency. Groundwater models can groundwater resources, artificial recharge and storage of play an important role in the assessment of ASR scheme. surplus desalinated water in aquifers can play a major role Development of good calibrated modeling tool enables a in the management of water resources. Due to the absence variety of approaches concerning characterization of aquifer of large storage reservoirs, desalination plants producing that will host artificially recharged groundwater. Collection fresh water for urban supply are forced to operate at sub-­ and management of primary hydrologic data are key factors optimal conditions. Thus, artificial recharge of groundwater to successful characterization of any ASR system. Optimum and storage of freshwater is deemed necessary and prom- injection and recovery rates are a function of site-specific ising technology for meeting seasonal peak demand and conditions. The applications of developing geodatabases offset periods of water deficit due to long-term emergency combined with geographic information systems (GIS) and and natural crisis conditions. Also, it will help to manage the three dimensional (3-D) modeling tools have been applied daily and seasonal fluctuations in desalination water pro- in models of regional scale (Kolm 1996; Russell et al. 1996; duction and consumption as the production of desalination Kassenaar et al. 2004; Ross et al. 2005; Thorleifson et al. plants is constant and the demand is not constant. The excess 2005; Cools et al. 2006) and at local scale (Shah 2004). Three- amount of produced desalinated water during the non-peak dimensional modeling tool is needed to build consistent hours could be stored in aquifers. conceptual model for groundwater flow (Shafer et al. 2006) The success of the ASR scheme is normally measured in terms of recovery efficiency, which is defined as the per- centage of water injected into a system in an ASR site that Table 1 fulfills the targeted water quality when recovered. The Values of transmissivity and geohydraulic conductivity for the recovery efficiency is controlled by a wide variety of fac- study area (derived from pumping test evaluation) tors including ambient hydraulic gradient; aquifer perme- ability, porosity, heterogeneity, thickness, and confinement; Well No. Transmissivity (T) Geohydraulic ambient groundwater density and quality; injected water conductivity (kf) density and quality; ASR operation (Bear, 1979; Merritt, (m²/d) (m²/s) (m/d) (m/s) GWA-141 1,050 1.2E-02 19 2.2E-04 6 GWA-144 250 2.9E-03 6 7.1E-05 5 GWA-148B 500 5.8E-03 13 1.5E-04

ays ) 4 GWA-151 900 1.0E-02 22 2.6E-04 (D

e 3 GWA-153 830 9.6E-03 16 1.8E-04 GWA-156 800 9.3E-03 28 3.3E-04

ese rv 2 R GWA-164 245 2.8E-03 6 6.5E-05 1 GWA-172 140 1.6E-03 5 5.9E-05 0 Kuwait QatarBahrain Sudia United Arab GWA-177 1,200 1.4E-02 45 5.2E-04 Arabia Emirates GWA-178 1,500 1.7E-02 38 4.4E-04 GWA-215B 2,500 2.9E-02 51 5.9E-04 Fig. 1. Storage capacity for emergency water in GCC countries GWA-240 950 1.1E-02 28 3.2E-04 (Dawoud, 2008). 13th Gulf Water Conference Proceedings / Desalination and Water Treatment 176 (2020) 125 to simulate the movement of artificially recharged water • The geological settings and relative remoteness of the through the unsaturated zone building the fresh water bub- northern Liwa area offer a vast natural storage capacity ble and the interaction of native water in the aquifer system and an excellent protection of naturally and artificially with fresh injected water. In Abu Dhabi Feflow package was recharged water resources from environmental influ- used for simulating the groundwater flow in the study area. ences at the surface. In comparison, man-made storage facilities with a comparable capacity would be almost 2. Sit selection impossible to build and to maintain; • A deep-seated underground reservoir utilizing natural The success of the pilot project is mainly dependent on sand formations as storage with a possible extension of the selection of the most suitable site. To select the most up to 400 km2 and with sufficient aquifer thickness and suitable location for the ASR pilot experiment, Abu Dhabi unsaturated depth to groundwater table is hardly vul- Emirate (Fig. 2) was divided into 10 × 10 km squares. A nerable as a whole against environmental hazards and site selection suitability index was developed including all vandalism. factors affecting the success of the ASR. GIS was used to • The mixture of the artificially recharged desalinated overlay various layers and calculate the site suitability index seawater and the native existing fresh groundwater will and evaluate and propose an initial array of potential ASR improve quality of the resource and presumably favor- site locations. The suitability index was based on the prem- able hydrochemical conditions. Huge volumes of the ise of maximizing ASR effectiveness while minimizing any existing fresh groundwater (salinity of up to 1,000 ppm) attendant impacts (Fig. 3). Multiple planning factors were could then meet the international WHO-Drinking Water used for the evaluation of ASR feasibility in Abu Dhabi Standard. Emirate including the availability of recharge water in terms • Such a groundwater enhancement project in the north- of quantity and quality, topography, cost of required surface ern Liwa area would be an essential back-up water sup- facilities and infrastructures, aquifer native groundwater ply for the Emirate of Abu Dhabi. The proposed system quality, unsaturated aquifer thickness, aquifer extent and can cope with seasonal consumption peaks as well as boundary conditions, and aquifer hydraulic parameters. emergency situations easily From this analysis, it was found that northern Liwa area is the most suitable site for the pilot ASR project (Fig. 4). The current investigations mainly focus on the water sup- The following advantages for the water supply system ply of the of Abu Dhabi. Yet, any provision scheme for of the Emirate of Abu Dhabi are obvious in case of the other urban or agricultural areas is possible, depending on the northern Liwa area would be developed and utilized in the water balance the supplier wants to achieve through abstrac- way described above: tion and injection. The artificially recharged groundwater

Fig. 2. General location map for Abu Dhabi Emirate. 126 13th Gulf Water Conference Proceedings / Desalination and Water Treatment 176 (2020)

S

0 0 0 h

a

rj

Suitability of Areas for Large-Scale ASR-Project 0 0 0 0 0 41a 40 Arzana Dubai h Cluster excluded if three or more Criteria not fullled 0 0 13 20 23 0 49 49 45 0 Evaluation Criteria, min. Requirements and max. Points: Available Thickness of Vadose Zone>=10 m max: 25 points 0 14 16 22 26 29 31 52 0 43 45 41 57 Available Aquifer System Thickness >=10 m max. 10 points 0 0 0 17 23 28 26 29 40 31 37 54 0 Available Aquifer Thickness >=10m max: 10 points Abu Dhabi 0 Native Groundwater Salinity <=25,000 ppmmax. 20 points 0 0 0 15 20 23 23 24 25 29 27 37 47 51 0 Natural Geohydraulic Gradient <=5 ‰ max. 15 points 0 0 Arabian Gulf Distance to next Border / Shore >=5 km max. 10 points0 0 0 0 0 18 24 27 34 30 33 34 40 34 42 48 0 Number of existing Wells <=500 max. 10 points

0 0 0 0 0 0 0 0 0 16 21 25 29 31 31 37 40 49 48 29 30J e b e l

H

a f 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 17 20 23 27 29 32 33 36 49 50 50 29 21 i 0 0 As Sad t Al Khazna Zaroub 0 0 0 0 0 0 15 17 0 0 0 0 0 14 0 0 0 0 0 0 0 0 0 0 0 0 19 22 27 26 28 30 34 32 27 46 41 34 14Al 0 39 0 Ain 0 0 0 0 0 0 0 16 19 24 24 23 18 17 20 18 15 0 0 0 0 0 0 0 0 0 0 0 23 28 32 31 39 33 33 37 35 29 38 25 20 49 0

0 0 0 0 0 0 19 24 30 37 34 33 34 35 31 27 22 17 0 0 B 0 0 0 0 0 0 25 28 33 37 36 36 33 32 35 37 26 31 A B 0 0 0 0 0 0 23 28 37 46 40 44 44 43 47 43 44 30 26 24 26 23 21 0 21 27 30 31 38 40 39 38 35 33 28 31 34 36

0 0 0 0 0 25 36 41 46 60 57 56 57 57 60 56 48 47 39 36 35 31 28 30 35 35 41 42 41 41 45 35 35 28 29 35 37 37 L I H 0 0 0 25 28 38 45 54 58 61 70 68 62 68 66 65 59 48 44 43 39 43 44 41 A 42 47 50 45 44 45 40 37 32 30 32 32 28 S A S Az Zafrah 0 0 0 24 29 36 45 50 58 65 72 80 79 81 A 78 76 76 57 56 56 50 50 51 48 46 52 49 47 47 45 41 35 31 26 25 30 37 Suitability of Area for H

Large-Scale ASR-Project U 0 0 21 26 34 43 52 63 68 74 88 88 B 82 83 86 90 75 75 61 56 59 55 60 51 58 57 47 47 47 43 35 28 0 21 18 0 - 10 Al Faiha

ed 0 19 23 31 42 52 64 70 80 88 91 91 91 91 90 76 84 82 61 62 55 53 52 57 60 53 48 41 41 37 27 0 31 t

ui 10 - 20 B

ns A Bu Hamrah U 20 25 33 43 53 65 73 80 87 86 86 90 92 90 77 82 87 79 63 57 S 57 55 53 50 50 44 40 37 33 0 0 0 20 - 30 A Agrab 30 - 40 25 31 35 41 51 64 74 79 85 84 79 82 81 68 65 62 57 70 75 68 63 58 54 53 48 43 39 33 27 0 0 Al Qafa ed t

ui 40 - 50 34 38 42 52 65 71 74 77 71 65 67 46 42 49 47 29 33 52 56 57 58 57 53 50 47 43 31 0 0 S

d Bateen e t i 50 - 60 40 41 53 62 68 74 66 62 Huwail51 a 36 34 32 40 39 40 39 40 41 51 51 52 49 45 47 46 33 0 0 Sultanate m Ramlat Ar Rabbad i

L L I W A 60 - 70 55 58 65 67 65 58 40 33 29 31 0 0 0 40 41 32 33 49 39 37 38 42 41 0 0 25 of Oman H Haleiba A 70 - 80 H 33 0 0 0 S0 0 0 36 29 30 26 27 31 28 30 29 0 0 24 ASR-Pilot Project Area Mashhur ed Zarrarah it 80 - 90 u

S to 0 0 0 0 Qusahwir0 0a 0 0 0 0 25 0 90 - 100 No. of Criteria for which Kingdom of Saudi Arabia Mender Evaluation Minimum Requirement is 0 0 0 0 0 Points not fullled 0km25km 50km 75km 100km Fig. 3. Development of site suitability index.

Fig. 4. Selected site for ASR pilot project (Northern Liwa). 13th Gulf Water Conference Proceedings / Desalination and Water Treatment 176 (2020) 127 resource in the northern Liwa area could be considered as interdunal deposits prevail. They consist of caliche horizons the strategic drinking water reserve of the Emirate of Abu with traces of organic matter, siltstones and even marls that Dhabi for future generations. may be interpreted as playa lake sediments and give evi- dence of more frequent pluvial periods in the Pleistocene. The Tertiary unit can also be subdivided into an upper 3. Hydrogeological assessment unit, consisting of mudston layers and evaporites (gypsum, 3.1. Lithology and aquifer system geometry anhydrite, dolomite) of the Lower Fars Formation, and a lower subunit that is marked by the predominance of clas- To assess the aquifer geometry and the stratifications, tic sediments (sandstones, siltstones), that are intercalated geodatabase populated with the information collected from with layers of mudstones and anhydrite. Table 2 summa- 43 wells drilled within the study area and 140 existing wells rizes the upper stratigraphic sequence in the study area as around it. As the injection of fresh water will be in shal- revealed by the project boreholes. The approximate spa- low groundwater, the depth of most selected wells is less tial distribution of the stratigraphic units is shown in the than 150 m and is not penetrating the deep aquifer. However, hydrogeological cross-sections (Fig. 5). four boreholes were drilled deeper than 490 m. In conse- The boundary between the Quarternary upper unit quence, more detailed lithologic information is available for (aquifer) and the Quarternary lower unit (aquitard) is not the uppermost layers of the stratigraphic sequence of the clearly defined and cannot be easily correlated between region. Two main stratigraphic units have been encountered: boreholes. However, a significant increase of slightly cem­ ented interdunal deposits is generally noticed around 60 m + • Quarternary unit: Holocene and Pleistocene eolian fine MSL. Below this depth, the formation is still fully saturated, to medium sands and interdunal deposits. The thickness but does not contribute significant amounts of groundwa- of this unit varies between 100 and 150 m, depending on ter to wells, due to its relatively low permeability. This part the topographic height of the respective location within of the formation is thus considered as an aquitard. A very the study area. significant lithological and hydrogeological boundary is the • Tertiary unit: mudstones, evaporites and clastics of one between the Quarternary lower unit (aquitard) and the Miocene age. This unit has a thickness of over 350 m and Tertiary upper unit (aquiclude). This boundary is defined has not been completely penetrated by any project well. by the first occurrences of evaporites of the Lower Fars Formation and marks the bottom of the aquifer/aquitard The Quarternary unit may be divided into two subunits. system. As shown in the hydrogeological cross-sections, in The upper unit is characterised by the predominance of the study area, this interface is encountered at a depth of well-sorted, fairly loose eolian dune sands with occasional around 30 m-MSL. The static groundwater level is encoun- intercalations of fine-grained, slightly cemented interdunal tered between 104 m + MSL and 107 m + MSL. Hence, the deposits. In the lower subunit of the Quarternary, these average thickness of the main aquifer is about 40 to 50 m.

Table 2 Upper stratigraphic sequences in the study area

Unit Subunit Description Quaternary Unit Upper subunit Eolian, loose fine to medium sand Holocene + Pleistocene) Lower subunit Interdunal deposits (caliche, silt, marl and sand Tertiary Unit Upper subunit Mudstone and evaporites of the Lower Fars Formation Miocene Lower subunit Miocene clastics (sandstones, mudstones, anhydrite)

South North 225 Al Bateen Liwa Al Qafa / Bu Hasa <=1,000 ppm 225 200 Crescent >1,000 -1,500 ppm 200 175 GW-Shed >1,500 - 4,000 ppm 175 L) 150 >4,000 - 7,000 ppm 150 >7,000 - 10,000 ppm MS 125 125 >10,000 ppm lf (m 100 100 Gu n fresh n

io 75 75 ia at 50 brackish 50 ab ev 25 25 Ar El 0 saline 0 -25 -25 -50 Aquitard / Aquiclude 0 km 10 km 20 km 30 km 40 km 50 km -50 -75 -75 2530000 2540000 2550000 2560000 2570000 2580000 2590000 2600000 2610000 2620000 2630000 2640000 2650000 2660000 Vertical Exaggeration: 100-fold Latitude (UTM y40) Fig. 5. Hydrogeological cross-section for the selected area. 128 13th Gulf Water Conference Proceedings / Desalination and Water Treatment 176 (2020)

The entire aquifer/aquiclude system exhibits a total thick- locally; its effect on the large-scale groundwater flow is ness of around 130 m. still negligible.

3.2. Groundwater flow 3.3. Aquifer hydraulic parameters Generally, in geohydraulic terms, the Western Region of To calculate the aquifer hydraulic properties within the the Emirate of Abu Dhabi is separated by a major ground- study area, 23 pumping tests were carried out between Octo­ water divide, running approximately from east to west. ber 2001 and February 2002. Analyzing the results of these It passes the greater Liwa area some 20 to 30 km north of the pumping tests gave more detailed information regarding: Liwa Crescent. All shallow to medium deep-seated ground- water north of it flows to the north, towards the Arabian • Geohydraulic parameters (such as transmissivity, geo- Gulf as the receiving body, while all groundwater south hydraulic conductivity and storage coefficient of the of it flows southerly to Saudi Arabia. There, huge Sabkha aquifer) areas, much lower in elevation than the central part of the • Well performance characteristics (well capacity, well Western Region, function as a discharge area, where tre- efficiency) mendous groundwater volumes constantly evaporate. The • Groundwater quality (general hydrochemical composition, study area is located exactly along this major groundwa- vertical salinity profile and possible dependency of ter divide. The measured groundwater level in December the quality on the discharge rate). 2001 was between 103.6 m + MSL and 107.2 m + MSL, with an aerial average of 106 m + MSL (Fig. 6). From the high- The calculated geohydraulic conductivity varies between est geohydraulic head in the central part of the eastern 2 and 60 m/d with an average of 27 m/d and a median of half of the study area, groundwater flows naturally radial 28 m/d and the calculated transmissivity ranges from 100 to to the adjacent areas under a maximum geohydraulic gra- 3,000 m²/d, with an average of 1,065 m²/d and a median of dient of 0.5‰. The calculated natural velocity of ground- 950 m²/d. The lower values mostly refer to wells, in which water movement ranges from 1 to 10 m/a. Although the the deeper section of the aquifer was tested. The storage geohydraulic gradient is comparatively low, the groundwa- coefficient was obtained only for those pumping tests, where ter flow system is a dynamic one, which cannot be main- additional piezometers in the vicinity of the tested well could tained without a driving force. Groundwater abstraction by be monitored. However, the determined values strongly discharging wells modifies the current flow pattern only depend on the test duration. For the 2-d lasting test, an

Fig. 6. Groundwater levels contour map (before injection, 2001). 13th Gulf Water Conference Proceedings / Desalination and Water Treatment 176 (2020) 129 average storage coefficient of S = 0.16 was calculated, while table of about 5 m. Here, less than one mm/a of evaporation the 10-d lasting test finally gave the higher value of S = 0.32. has to be considered. If the capillary fringe does not reach The projected short to medium-term yield of the tested wells the ground surface, evaporation occurs mostly as transport is comparatively high and ranging from 40 to 250 m³/h with of water vapour. In order to obtain long-term groundwater an average of about 130 m³/h (GTZ/DCO 2002). evaporation data, two holes have been drilled with a diam- eter of 26″ down to 50′, and a bottom sealed 21½″ steel con- 3.4. Field infiltration tests ductor pipe was installed. An impermeability test of the 21½″ steel casing was conducted followed by the installation of In order to obtain information on infiltration rates and a 2“ pipe inside the monitoring well. The steel casing was vertical geohydraulic conductivity, one well infiltration test filled up with sand and afterwards water was pumped from and four tank infiltration tests were conducted. The results a tank through the 2″ pipe into the well until the sand was of the infiltration test indicated that the infiltration rate per saturated and the water level reached the top of the 21½″ m2 ranges from 0.77 to 0.49 m3/h. The evaluated vertical casing. The impact of temperature, humidity and wind hydraulic conductivity in the vadose zone ranged from 1.6 speed was correlated with the evaporation and possible 10–4 to 2.5 10–4 m/s, indicating an approximately 40% lower recharge in the capillary zone. The test results and obtained conductivity value than that of the horizontal hydraulic data indicated that significant evaporation from the Liwa conductivity. Aquifer could only occur in zones of very shallow ground- water table (<1 m). Literature evaporation data support these 3.5. Evaporation column tests preliminary results and observations. Evaporation has to be considered as an essential part in 3.6. Groundwater quality the artificial recharge project and needs to be surveyed pre- cisely. Evaporation from bare soils occurs by capillary rise of Complete hydrochemical analyses were carried out soil water up to the evaporation front, at certain soil depth, for 400 wells. The evaluation of the hydrochemical and followed by molecular diffusive water vapour transport hydroisotopical composition of the groundwater within the through the dry soil layer above. For bare soil conditions and study area is based on 40 groundwater samples, 37 sam- certain depths of groundwater table, the following evapora- ples of which have been collected from wells, mostly at tion rates are reported (Table 3). the end of the long-term constant discharge pumping test. This indicated that direct evaporation from the ground- Furthermore, three samples of the very shallow groundwater water body is negligible beyond a depth to the groundwater were taken from excavations. The analyzed total dissolved

Table 3 Evaporation rates of sands (after Kontny 1993)

Depth of groundwater Mean depth of groundwater Groundwater evaporation (mm/a) table (m bgl)* table (m bgl)* Coarse sand Fine sand 0–5 2.5 2.18 6.91 5–10 7.5 0.53 0.79 10–15 12.5 0.28 0.38 15–20 17.5 0.20 0.24

*m bgl = Meter below ground level.

Table 3 Salinity-related classification of groundwater samples

TDS (ppm) Classification Samples ≤1,000 Freshwater (World Health Organization WHO) 28 >1,000–1,500 Freshwater 7 ≤1,500 Freshwater (Local UAE-Standard 35 >1,500–4,000 Slightly brackish 3 >4,000–7,000 Medium brackish 1 >7,000–10,000 Strongly brackish 0 >10,000–25,000 Slightly saline 1 >25,000–50,000 Medium saline 0 >50,000–100,000 Strongly saline 0 >100,000 Brine 0 130 13th Gulf Water Conference Proceedings / Desalination and Water Treatment 176 (2020) solids (TDS) of the 40 groundwater samples ranged from Topper R, Barkmann PE, Bird DA, Sares MA (2004) Artificial 348 to 12,314 ppm. The number of samples according to recharge of ground water in Colorado-a statewide assessment. the applied groundwater salinity classification is shown in Environmental Geology 13. Colorado Geological Survey Division of Minerals and Geology Department of Natural Resources, Table 3. Denver, Colorado. In terms of mineralization, as much as 70% of the ana- Brissaud F (2003) Groundwater recharge with recycled municipal lyzed samples meet the limit of the WHO-Drinking Water wastewater: criteria for health related guidelines. In: Aertgeerts Standard and almost 90% of the analyzed samples are fresh, R, Angelakis A (eds) State of the art report health risks in according to the local standard (TDS: ≤1,500 ppm). Deeper aquifer recharge using reclaimed water. Water, Sanitation and Health Protection and the Human Environment World Health screened wells produce groundwater of higher salinity: three Organization Geneva and WHO Regional Office for Europe samples are slightly brackish (>1,500 to 4,000 ppm), one is Copenhagen, Denmark 2:10–15. medium brackish (>4,000 to 7,000 ppm) and another one Rüetschi D, Wülser R (1999) Die Künstliche Grundwasseran­ is slightly saline (>10,000 to 25,000 ppm). As sodium is the reicherung der Wasserversorgung Basel. Kapitel 3. AWBR dominant cation, and chloride is the prevailing anion, hydro- (Arbeitsgemeinschaft der Wasserwerke Bodensee-Rhein) Jahres­ bericht 1999, pp 71–84. chemically, the tapped groundwater can be characterised as Al-Katheeri ES (2007) Towards the establishment of water manag­ “alkaline water, predominantly chloridic”. The calculated ement in Abu Dhabi Emirate. Water Resources Management content of sodiumchloride (salt) as the major dissolved min- 22:205–215. eral phase ranges from 77 to 7,169 ppm with a median value Artimo A, Mäkinen J, Berg RC, Abert CC, Salonen V-P (2003) of 431 ppm. Three-dimensional geologic modeling and visualization of the Virttaankangas aquifer, southwestern Finland. Hydrogeol J 11:378–386. 4. Conclusions Pyne D, 2007, Overview of Aquifer Storage Recovery in the USA and the World, Conference on Aquifer Storage Recovery and In Abu Dhabi emirate, two recharge schemes using Artificial Recharge. injection wells and infiltration basins were used for inject- Bear J, 1979. Hydraulics of Groundwater. McGraw-Hill, New York, p. 276–299. ing the desalinated water into the shallow groundwater Merritt ML, 1985. Subsurface storage of freshwater in south aquifer system. Recovery cycles results indicated that under lorida: a digital model analysis of recoverability. US Geological the given conditions the recovery ratios ranged between Survey Water-Supply Paper 2261, 44 pp. 85% and 90% were physically recovered. At the end of 250 Missimer TM, Guo W, Walker CW, Maliva RG, 2002. Hydraulic d lasting period of constant recharge, the lateral migration and density considerations in the design of aquifer storage and recovery systems. Florida Water Resources Journal, pp: 30–35. of the outer injected freshwater body was only 0.2 m/d. Reese R, 2002. Inventory and review of aquifer storage and recovery For 75% recovery ratio, the recovered water salinity will be in Southern Florida. US Geological Survey, Water- Resources up to 430 ppm, and for 85% recovery ratio, the recovered Investigations Report 02-4036, pp: 55. water salinity will be up to 485 ppm. Both schemes proved Artimo A, Sarapera S, Ylander I, 2008, Methods for Integrating and to function perfectly. However, in contrast to the infiltra- Extensive Geodatabase with 3-D Modeling and Data Manage­ ment Tools for the Virttaankan Artificial Recharge Project, tion basin scheme, for the dual-purpose wells of the well Southwestern Finland, Water Resources Management, Vol. (22), gallery scheme there are indications of reducing injection pp. 1723–1739. and abstraction capacity over time due to clogging effects. Bichara AF, 1974, An Experimental Study of Long-term Artificial Moreover, considering the local hydrogeological conditions, Recharge of Groundwater into Confined Aquifers, Master the infiltration basin conception is advantageous as it is Thesis, University of Strathclyde. Khanal NN, 1980, Advanced Water Supply Alternative for Upper easier to operate and maintain. It was recommended to use East Coast Planning Area; part 1-Fesability of Cyclic Storage of recharge basin in the full scheme project. Fresh Water in a Brackish Aquifer, and Part 2- Desalinization Alternative. South Florida Water Management Technical Publication No. 80–6, pp. 75. References Bouwer H, Pyne G, Goodrich J. 1990. “Recharging Groundwater”. Dawoud M (2008) Strategic Water reserve in Abu Dhabi Emirate. Civil Engineering, June 1990, pp. 63–66. Lowe M (2005) An overview of aquifer storage and recovery projects Calleguas Municipal Water District. 2004. Las Poses Basin Aquifer in Utah. Geological Society of America Abstracts with Programs, Storage and Recovery Project, Project Summary Memo, Calleguas vol 37, no. 7, p 164. Water District, Thousand Oaks, California, 4 p. Heilweil VM (2005) Artificial recharge to fractured bedrock: the GTZ/DCO, 2002, Artificial Recharge and Utilization of the Navajo Sandstone of Southwestern Utah. Geological Society of Groundwater Resource in Liwa Area, Final Feasibility Study America Abstracts with Programs, vol. 37, no 7, p 165. Report, Abu Dhabi, UAE, p. 101.