Abstract
The exploratory borehole Megiddo-Jezre’el 1 (MJ1) was drilled in Israel, in the Bet She’an Valley which branches out from the Central Jordan Rift. It reached the depth of 5060 m and bottomed within the Upper Triassic Mohilla Fm. Following the increase of groundwater exploitation, the Cl- concentrations increased and ionic ratios changed indicating inflow of Ca2+-Cl- brines, the origins of which were hitherto unknown. Data from the new MJ1 borehole revealed that rock porosities decrease with depth. Lowermost values of about 3% were interpreted from logs in Lower Jurassic and Triassic strata. The highest shut-in pressures were measured in the Upper Jurassic sequence raising the water much higher than the ground surface. Along the drilled section, there is a continuous downward increase in Cl- concentrations in the range of 12-186 g Cl-/l and a very clear stratification of brines. Data from the MJ1 borehole and from other exploration wells indicate that in the subsurface of the area, there are two definite source brines: Triassic brine and the Late Tertiary (so-called) Rift brine. Brines encountered in Jurassic and Cretaceous beds represent ancient mixtures of the two source brines involving various water-rock chemical transformations. Evidence of very high pressures in deep boreholes Devora 2A, Rosh Pinna 1, and MJ1 revealed the existence of a mechanism in which the deep brines are “piston-driven” upwards and possibly also laterally. The ongoing salinization of groundwater in the area is due to the inflow of the Late Tertiary Ca2+-Cl- Rift brines and not that of the Jurassic or Triassic brines. The hydrogeological and hydrochemical data from borehole MJ1 is of major importance for the management of groundwater resources in the Central Jordan Rift Valley and in the adjacent geologically connected areas.
1. Introduction
One of the major causes of the water crisis in Israel is the ongoing contamination of its groundwater resources by saline water bodies. The main deteriorating process occurring in groundwaters of the Jordan-Dead Sea Rift Valley and in adjacent connected areas is the upflow and migration of highly pressurized thermal Ca-chloride brines penetrating into fresh groundwater bodies [1]. In Israel, and particularly in the Rift and in branched-out areas, intensive exploitation of fresh groundwater has disturbed the natural equilibrium which prevailed between freshwater and saline water. The newly established groundwater flow regimes have facilitated the further migration of saline groundwater bodies, their participation in the active hydrological cycle, and the progressive contamination of fresh groundwater [2]. These processes were not anticipated by planners and water resources managers because there was no sufficient information about the occurrence and the physical regimes controlling the migration of the brines.
A common feature of all groundwater bodies flowing in the Jordan-Dead Sea Rift Valley is that these are mixtures of low salinity or freshwater bodies with a Ca2+-Cl- saline brine, otherwise known as the “Rift brine.” The origin of the low-salinity component may differ. South of the Dead Sea, it is usually the Lower Cretaceous paleowater from the so-called “Nubian sandstone aquifer” or recent floodwaters. Further northwards, the low-salinity component originates from the Cretaceous (Cenomanian-Turonian) and Tertiary (Eocene) groundwaters.
The saline Ca2+-Cl- component, i.e., the “Rift brine” [3], is identified by its high salinity and by typical ionic ratios such as [4]. Traditionally, Cl-/Br is given as a weight ratio, whereas all the other ratios are calculated from meq/l values. The brine is characterized by high pressures and elevated temperatures and is encountered mainly (but not exclusively) along the western margins of the Rift, from the region of Lake Tiberias in the north to the northern Arava in the south. The chemical evolution was characterized by the increase of the Ca2+ concentrations as indicated by the raise of the value to values of ≫1, decrease of the Na+/Cl- to values of <0.7, and Cl-/Br- decrease to values much lower than 286. All these chemical changes are clearly indicative of progressive inflow of pressurized and thermal Ca2+-Cl- brines which are encountered in the Jordan Rift Valley and in the adjacent areas which are in structural contact with the Rift [5].
Two types of brines were identified in the Rift [3]): (i)Predominantly phreatic and nonthermal saline groundwater with occurs along the eastern and southeastern shores of Lake Tiberias(ii)Pressurized, thermal, and saline groundwater with equivalent ratios of occurs mostly along the western margins of the Rift, between Lake Tiberias in the north and the Arava Valley in the far south. Freshwater-diluted Ca2+-Cl- brines mostly ascend near fault escarpments [1]
The study of groundwater salinization in the Rift started in the early 1960s [3]. Numerous researchers suggested a wide range of geological scenarios [6, 7].
Considering chemical and stable isotope data, it occurs that the water bodies contributing to saline groundwater in the Rift are as follows [6]: (i)Recent Dead Sea water which is a primary evaporation brine with high (ii)Neogene seawater evaporation brine in the flooded Rift as represented by the Ha’On saline water characterized by . This saline water flows out from springs and seepages along the eastern and southeastern shores of Lake Tiberias(iii)Neogene-Tertiary brines with encountered in Tiberias Hot Springs and beyond the Bet She’an basin(iv)Neogene to recent ablation brine with Na+/Cl- of about 1 encountered in the deep well Zemah 1 located at the southern tip of Lake Tiberias(v)Deep-seated brines of the (Ca, Mg)Cl2 type which are the subject of the present study
2. Objects of Investigations
As the result of the ongoing increase of groundwater exploitation in the Bet She’an and Harod Valleys since the early 1950s, the chemical composition of the exploited water changed due to variously diluted Ca2+-Cl- brines rising from depths and replacing the fresh groundwater. This process and its chemical details are similar for all water resources in the Rift, from Lake Kinneret in the north to the Arava in the south. In all these areas, scholars asked the same questions which remained equally unanswered because of a lack of information from horizons deeper than 800 m. These questions are as follows: (i)Is there only one upflowing brine or are there several brines of similar composition but generated in different rock units and on different occasions?(ii)What are the chemical compositions of the subsurface brine(s)?(iii)In which rock units were these brines generated?(iv)What is their updriving mechanism and at what rate?(v)At what depth do the different brine bodies occur?(vi)Assuming that answers are available to these questions, what would be the impact of upflowing brines on future regional plans for water exploitation?
The recent drilling of the deep exploratory borehole Megiddo-Jezre’el 1 (MJ1) supplied answers to many of the abovementioned questions. The present paper summarizes the results and their contribution to the understanding of the genesis and movement of brines in the Rift and in the structurally connected regions.
3. Materials and Methodology
The present study is based on the geological evidence from the recently drilled deep (5060 m) exploratory borehole Megiddo-Jezre’el 1 (MJ1), located 800 m south of Kibbutz Sde Eliahu, in the Bet She’an Valley, at Israeli coordinates 248240/703782, and Kelly Bushing (KB) at -180.1 m bsl. Until then, subsurface geological information in the area was limited only to the western and southern margins of the Bet She’an Valley, where it was derived from water wells such as Revaya 1-7, Bardala 1-2, and Mehola 5 (Cretaceous Judea Group aquifer) (Figure 1) [8]. The depths of these water wells did not exceed 800 m. The northernmost well Shoqeq 1 exploits water from Tertiary Avedat Group limestones. There are no direct geological data from the central parts of the Valley. Gardosh and Brunner and Meiler et al. [5, 9] narrowed this information gap. They established the first subsurface structural model of the area based on the interpretation of seismic data. In borehole MJ1, the lithology and stratigraphy of the drilled section were established by careful examination of samples and on electrical logging. The major ions of all fluids encountered in the borehole during drilling operations were analyzed in the Geochemical Laboratory of the Geological Survey of Israel. The data are presented in Table 1. The accepted uncertainties of analyses were ±2% and, in few extreme and rare cases, up to ±3%. Information on porosities, temperatures, pressures, and Cl-concentrations were derived from logs [10], and Cl- concentrations were deciphered from resistivity logs and calibrated with chemical analyses of water from bottom hole samples and from swabbing. Swabbing tests were run at preselected horizons (Figure 2). The chemical composition of fluids collected during the present drilling operations was compared with hydrochemical data from relevant water resources in the area, documented since the early 1960s (data from the files of the Hydrological Service of Israel, Mekorot, National Water Co., and from [10]). The triangular plot is built by Microsoft Excel with XLSTAT. The dendrogram shows the hierarchical ordering of data of the output from hierarchical clustering using the open domain R Code Team [11].
4. Geological and Hydrogeological Setting
The Bet She’an-Harod basin (BSB) (Figure 1) is a deep morphotectonic depression located at the eastern boundary of the Levant basin. It branches out of the Central Jordan Rift Valley, separating the calcareous Gilboa and Fari’a-Beqaot mountains to the west and south from the basaltic plateau of Issakhar to the north in the Eastern Lower Galilee. BSB is one in a series of intracontinental basins along the Dead Sea Rift (DSR) in the Middle East region whose tectonic development is associated with the development of the Dead Sea Transform (DST) fault system which is a plate boundary separating the African and Arabian plates [12]. Within the study area, two major structures influence the tectonic development of the region: the Syrian Arc System (SAS) to which the Fari’a-Beqaot structure belongs and the DST strike-slip fault along which the convergence of Africa and Eurasia takes place. The coexistence of SAS and DST shapes the eastern part of the Levant basin and its margins to the east [9, 11, 13–17]. The superposition of SAS and DST caused an abrupt change of the regional tectonic regime since the Miocene. The Fari’a-Beqaot anticline plunges towards NNE beneath the southern part of BSB. The most significant effects of active tectonics are the fast subsidence of the area surrounding DST, large thickness of accumulated fluvial and lacustrine sediments, and abundance of faults and fractures of different dimensions and ages [18].
According to Derin [19], the Triassic beds in Israel are only exposed in the southern part of the country in the erosion cirques of Makhtesh Ramon and Har Arif and were penetrated by 36 deep boreholes spread mainly over the southern and central parts of the country. Only 7 boreholes, including MJ1, were drilled in northern Israel. The Triassic Ramon Group [20] is characterized by large-scale marine oscillations which are the response to a regression of global extent. Due to alternations of the normal shelf and intertidal hypersaline to brackish environments in the Upper Triassic sequence, thick beds of gypsum/anhydrite occur in Israel in the Mohilla Fm and in Jordan in the Abu Ruweis Fm [20, 21]. Halite was found in borehole Ramallah. Triassic halite was also found in the southern Palmyrides in Syria and Lebanon [22]. The Triassic sequences display a north and eastward thickening, from 1100 m in the Negev to over 2600 m in borehole Devora 2A [23] and probably in MJ1 (base not reached). According to Greitzer [24], the salinities of brines in the Negev (Permian) and Ramon (Triassic) Groups increase northeastwards. This is consistent with Bentor [25] in which the ongoing emergence of the Arabo-Nubian massif throughout the Paleozoic-Lower Cretaceous time span created a hydrological gradient northward to structurally lower-lying areas.
The Triassic sequences display a trend of northward thickening, from about 1100 m in the south to more than 1700 m in boreholes Ramallah 1 and over 2600 m in Devora 2A (base not reached in these wells) [19]. Data from deep boreholes [19] indicate several hundred meters of thickening from the Coastal Plain eastwards. This thickened Triassic section was deposited in a basin defined as the “Judea-Galilee Low.”
In large parts of the country, the Ramon Group contains five formations (from bottom-upwards): Yamin, Zafir, Ra’af, Saharonim, and Mohilla (of Scythian-Carnian age) built of marine carbonates, shales, and gypsum/anhydrite [19]. The depositional environment of the present study is the Mohilla Fm built of dolomite, anhydrite, and shales indicating a tidal environment ranging from a lagoon to a sabkha [19]. The thickness of the Mohilla Fm is remarkably variable. In the Negev and Dead Sea area, it is in the 20-200 m range, whereas in the north (in borehole Devora 2A), it reaches 870 m. Thick sequences of the Asher Volcanics occur in numerous boreholes between the Triassic and Jurassic sediments [19]. In borehole Atlit 1, this unit is 3.5 km thick, whereas in Devora 2A (closest to MJ1), it is 272 m. According to Derin [19] and personal comm. (2019), the Triassic sequence in central and northern Israel is identical to the Germano-Trias type spread throughout southern Europe, between Morocco in the west and Turkey in the east.
All Jurassic Fms are included in the Arad Group [26]. It stands out by the large variety of facies reflected in the occurrence of different lithologies. The environments of deposition range from shallow clastic shelf to deep open marine. The transition from the Triassic to the Jurassic is defined by a major angular unconformity causing erosional truncation and karstification of the underlying Triassic beds. These events created karst pockets filled with variegated clays—Mishor Fm—penetrating deeply into the Upper Triassic beds. Further flushing and displacement of previously formed brines were either flushed west- and seawards or downwards to deeper stratigraphic levels in the subsurface. Hence, pre-Jurassic brines could have been preserved in few natural reservoirs locally protected by geological conditions from flushing and from total evacuation [12, 27]. The maximal cumulative thickness of the Jurassic sequence including the offshore attains about 4700 m [26]. The Jurassic period stands also out by its basaltic and ultrabasic activity as evidenced in borehole Devora 2A (1025-1192 m) [19, 28].
At the beginning of the Lower Cretaceous (Kurnub Group), tectonic and erosive activities continue, and the whole area was uplifted [28–30]. A massive paleokarst surface developed along an unconformity in Upper Jurassic beds [31]. Considering lithostratigraphic evidence, it stands to reason that the Jurassic sequence was permeated through the karstic zone by large masses of water which flushed out the older bodies of brine. Intense faulting facilitated interaquifer connection and mixing of various water bodies. Interactions of seawater with basalts (Tayasir Volcanics) have generated Ca2+-Cl- brines.
During the Cenomanian, marine conditions prevailed over large parts of the Middle East and a thick sequence of carbonate sediments (Judea Group) was laid down. At the end of the Turonian, the folding of the Syrian Arch began causing the formation of basins and highs [32]. The fluids that permeated the Judea Group were probably seawater which might have locally undergone chemical changes due to water/rock interaction and gases from volcanic activity. These fluids were flushed out of the emerging land built of pre-Oligocene formations and, since the Quaternary, were replaced by natural recharge. At present, in the Judea Group, the groundwater is characterized by Cl- concentrations which do not exceed 21 g/l. The values of Na+/Cl- and Cl-/Br- resemble those of seawater (0.86 and 286, respectively), and values are not higher than 3.
During the Mio-Pliocene, brines were mainly generated by repeated pre-Messinian ingressions of seawater. Following a deep penetration of the Tethys seawater, an inland lagoon was created in the nascent Rift. Evaporites precipitated in this lagoon, and their subsequent dissolution contributed to a new generation of brines. The Tertiary Hazeva and the Quaternary Dead Sea Groups are restricted to the Jordan-Dead Sea-Arava Rift Valley and are composed of up to 3000 m thick clastic, evaporite, and limnic sediments.
The regional exposed stratigraphic column of the Gilboa, Fari’a-Beqaot, and Issakhar mountains surrounding the Bet She’an basin includes the main rock units forming the regional aquifers that drain towards the Valley. These aquifers are 600-800 m thick Judea Group (Middle-Upper Cretaceous) built of karstic, highly permeable aquiferous limestones and dolomites. The approximately 400 m thick Eocene Avedat Group is characterized by chalks and limestones. The Neogene to Quaternary Tiberias and Dead Sea Groups comprising volcanic, lacustrine, marine, and continental rock units are also exposed in the area. Groundwater originating from Judea and Avedat aquifers exploited by boreholes is usually of the Ca2+-HCO3- type and of low salinity (<400 mg Cl-/l). Groundwater flowing in Neogene to Quaternary volcanic rocks and from interbedded lacustrine and continental strata are usually of higher salinity. Natural replenishment of the three regional aquifers occurs upon the exposed outcrops of the corresponding groups [33].
The natural outlets of all regional aquifers are two groups of springs. The Gilboa Springs emerge along the northeastern margins of Mt. Gilboa which is outlined by the Gilboa fault. The Mid-Valley Springs (MVS) emerge in the center of the BS Valley along a major morphotectonic lineament caused by subsurface faults branching out northwestwards from the Rift system [34]. The salinity of the MVS is at present <2000 mg Cl-/l. When systematic quality monitoring began in the early 1950s, salinities were of the order of <500-800 mg Cl-/l. It increased gradually in clear and proven dependence from regional freshwater exploitation by wells.
5. Results
5.1. The Exploratory Borehole MJ1
Based on geophysical evidence, borehole MJ1 was located on the northern plunge of the Fari’a-Beqaot anticline buried beneath the fill of the Bet She’an Valley (Figure 1). In the subsurface, the borehole is located near the intersection of longitudinal faults outlining the western rim of the Rift and faults branching archwise northwestwards out from the Rift system creating the Gilboa fault [9].
The total depth of the borehole is 5060 m, and the thicknesses of the penetrated rock sequences are as follows (Figure 2):
1020 m: Tertiary and Quaternary beds
483 m: Cretaceous Judea and Kurnub Group formations (incomplete due to structural reasons)
3123 m: Jurassic Arad Group formations
+430 m: Triassic Ramon Group formations
At the depth of 1020 m, there begins a major lacuna of the whole Avedat, Mt. Scopus, and the upper part of the Judea Group. The calculated lacuna is about 1000-1250 m. There are two possible reasons for this lacuna—either a major fault or a major discordance extending over the Fari’a structure [27, 43]. There is quite strong evidence for the existence of a major fault which facilitates upflow of hot (72°C) and saline (4500 mg Cl-/l) water at the depth of 1280 m. These unusual phenomena disappeared immediately after lowering the protective casing and isolating the inferred fault zone. A similar stratigraphic phenomenon was described from a nearby borehole Mehola 5.
Considering porosities interpreted from electrical logs run in the drilled section, the following depth intervals were outlined (Figure 2):
Quaternary-Miocene interval. In this interval, the porosities attain the highest values (aver. 32%). However, from acquired experience, it occurs that in the 0-865 m interval, permeabilities are very low and cannot be regarded as aquiferous.
Judea and Kurnub Groups. This depth range includes rock units related to the Judea Group (karstic carbonates) and to the Kurnub Group (sands and sandstones) (average porosity of 26%). These rock units could potentially be very good aquifers.
Arad Group, Haifa Fm. The thick calcareous Jurassic sequence is characterized by very low porosities (aver. 4%). The 3267-3310 m sequence of Rosh Pinna shales and the 3469-3523 m intervals stand out by its high content of shales. Fluids would only flow through faults and/or fissures.
Triassic Ramon Group. In the Mohilla formation (anhydrite/gypsum, dolomite, and limestone), porosities are even lower (aver. 3%) than in the overlying beds.
Two measurements of shut-in pressures are on record (Figure 2). At depth of 3414 m, the water level rose 2962 m to -452 m below KB (-632 m below MSL). Another measurement was made at depth of 1748 m, where the water level rose 1618 m to the level of 130 m below KB (-310 m MSL). This data is of the highest regional hydrological importance.
The recorded temperatures increase in borehole MJ1 (Figure 3).
From the surface down to a depth of about 100 m (down to the Cover Basalt), the temperature gradient is high (about 45°C/1000 m). Some temperature fluctuations were noticed which could be due to cooling effects by active groundwater flow from Mt. Gilboa and the Fari’a-Beqaot anticline in the west and southwest towards the center of the Bet She’an Valley. The presence of young volcanic bodies may explain the high temperatures at shallow depths. The Quaternary and Lower Cretaceous sections are characterized by low gradients of about ~10°C/1000 m. This depth zone contains mostly clastic rocks, and as observed in nearby well Mehola 5, fresh groundwater flow reduces the thermal gradient.
In the prevalent carbonate parts of the Jurassic and Triassic sequences, the temperature gradient is higher (27-28°C/1000 m). A certain increase in the gradient was noticed below the impermeable Rosh Pinna shales that presumably act as a sort of a hydrological barrier. In the Upper Jurassic down to the Rosh Pinna shales, the gradient rises to about 27°C/1000 m, whereas beneath the Rosh Pinna shales, the gradient changes to about 28°C/1000 m.
The water body in the upper parts of the Arad Group units cannot be regarded as brine because its Cl- concentration is lower than that in seawater. Na+/Cl- and values differ from the values characterizing Ca2+-Cl- brines. Unlike the brine in Lower Jurassic and Triassic beds, this saline water body contains bicarbonate which indicates hydrological contact with active replenishment. δ18O values were of -3.81 and -4.69. The data are presented in Table 1.
5.2. Changes in Porosity and Cl- Concentrations
Measurements of water salinity along the drilled section of borehole MJ1 revealed a continuous increase in Cl- concentrations with depth (Table 1). Four different zones of groundwater salinity could be outlined (from top-downwards, Figure 2):
Zone A. This fits almost perfectly with the sequence of post-Jurassic (Arad Group) strata which include Lower and Upper Cretaceous, Miocene, and Pliocene rock units. In this zone, porosities are particularly high in the aquiferous Kurnub Group.
Zones B and C. Within the very thick (3123 m) sequence of the Jurassic calcareous Haifa Fm, it is separated into two members by the Rosh Pinna shales (3270-3310 m): the Middle to Upper Jurassic member which is prevalently built of limestone with high porosity and low Cl- concentration of its water and the lower member with enhanced contents of shale reflected by its lower porosity and higher salinity. The interbedded Rosh Pinna shales consist of gray to black fossiliferous detrital shales and marl interbedded with limestone and dolomite. The Upper Member overlying these shales has a much higher porosity, and the Cl- concentration of its water is much lower. However, the very high pressures measured in the Jurassic beds underlying these shales could have caused the upward movement of these brines through the shales and into the overlying calcareous sequence.
Zone D. This extends between 4550 m and +5060 m and contains brines with Cl- concentration over 171 g/l. The top of this salinity zone is higher than the lithological boundary of the Triassic Mohilla Fm (at 4626 m) which is built mainly of massive dolomite and layers of anhydrite. Here again, the high pressure could have caused the upward movement of the brine into the overlying Lower Jurassic beds.
5.3. Comparison of MJ1 with Brines in the Vicinity
Following the definition of the hydrochemical identities of water bodies encountered in the study area by means of their characteristic ionic ratios, an attempt was made to verify—by means of cross plots between total dissolved equivalents (TDE) and various ionic ratios—whether these waters can be assembled into larger groups and the possible relations between them.
The TDE vs. Na+/Cl- plot (Figure 4; Table 2) reveals that trend 1 includes brines from the Lower Jurassic, i.e., ER4-25 and DST of borehole MJ1, Rosh Pinna 1, Devora 2A, and Tiberias Hot Springs (THS), members of the diluted Ha’On/Tiberias brine family. The deepest Triassic samples (DST and ER2) follow trend 2.
The TDE vs. K+/Cl- plot suggests mixing of Upper Triassic and Lower Jurassic brines from borehole MJ1 but with a significant change in K+/Cl- (Figure 5(a)). The Devora 2A sample and the deep sample from borehole Rosh Pinna 1 plot near this trend. Different trends are shown by Upper Jurassic and Lower Cretaceous brines from MJ1. The red trend indicates the association of Rosh Pinna and THS brines.
(a)
(b)
(c)
(d)
The TDE vs. Na+/K+ plot reveals four trends of the Lower Jurassic, THS/shallow and Rosh Pinna (red lines), DST/Devora 2A, and Upper Jurassic brines (Figure 5(b)). Most Upper Jurassic and Lower Cretaceous brines cluster at both low TDE and Na+/K+ values.
The TDE vs. Mg2+/Ca2+ plot shows a cluster of Lower Jurassic brines including deep Rosh Pinna (Figure 5(c)). All other groups of brines scatter widely. The red trend shows the combination of Rosh Pinna and THS.
The K+/Cl- vs. Na+/Cl- plot (in meq/l) reveals various trends in MJ1 (Figure 5(d)). The Upper Jurassic brines indicate mixing with basaltic water. The DST brines are lower in 1000 K+/Cl- than in seawater (18.6). The Lower Jurassic brine plots at low Na+/Cl-. The Triassic brines from the environment and the Lower Cretaceous brines scatter widely. HTS and the Rosh Pinna samples yield a mixing line.
6. Discussion
6.1. General Geochemical Considerations
All brines in the northern Rift basin and in the adjacent and structurally related areas represent evaporated seawater, the composition of which is altered by a number of chemical processes which are time-independent but depend on temperature and slightly on pressure. Mixing of brines and waters of different origins develop specific trends.
In the Na+ vs. Cl- plot, all data follow a common trend with a slope of about 1 suggesting that halite dissolution played an important role (Figure 6(a)). These basinal brines are of the Ca2+-Cl- type and are generated by reactions as follows:
(a)
(b)
(c)
(d)
Due to these reactions, Na+, K+, Ca2+, Mg2+, SO42-, and HCO3- vary. Although diluted by freshwater, all brines in Table 2 are saturated with respect to carbonates (Figure 6(a)).
The slope of -1 in the exchange of Na+ vs. Ca2+ in fluids proves albitization of plagioclase due to brine contact with basaltic matter (Figure 6(b)). The intercepts with the -axis at 1 and 0.9 indicate halite fluids and seawater, respectively. In borehole MJ1, the Upper Jurassic brines, the brines from borehole Devora 2A, and some Lower Cretaceous Mehola 5 represent halite dissolution brines, whereas all the others are based on Na+/Cl- of about 0.9, i.e., originate from seawater or are mixtures of both sources.
6.2. Involvement in Dolomitization
Involvement in dolomitization is demonstrated in Figure 6(c). The value of Mg2+/Ca2+ of 0.17 is in the range of dolomite formation. The drop in Mg2+ concentrations is mainly due to dilution with freshwater or basaltic water. The increase of Ca2+ by dolomitization facilitates elsewhere dedolomitization and precipitation of calcite and gypsum/anhydrite. Due to high Ca2+ in Triassic brines, their SO42- and HCO3- are low (Table 1).
Reactions (1)–(4) do not occur in the same geological environment. Reaction (1) only occurs in the presence of basalts and tuffs, reaction (2) in limestones, and reactions (3) and (4) in unsaturated solutions but in the presence of gypsum/anhydrite and calcite. Thus, the final equilibria of these reactions are randomly achieved in the rocks from which the samples for analysis are collected. Mixing with either higher or lower salinity water is common.
6.3. Sulfate
The proportional increase of sulfate and Ca in Figure 6(d) is best explained by the dissolution of gypsum/anhydrite.
6.4. Brines from Triassic Beds
Figure 6(a) portrays Cl- concentrations in fluids of Upper Triassic formations collected from exploration boreholes. All fluids are of the Ca2+-Cl- type with high factors of up to 245 in borehole Gaash 2. In boreholes Devora 2A and Rosh Pinna 1, these brines are pressurized to 480 and 330 bar, respectively, and are thermal (162°C) [26]. The highest Cl- concentration in MJ1 (171.6 g Cl-/l) was encountered by bottom hole sampling at the depth of 4915 m (Table 2) in dolomite and anhydrite beds of the Triassic Mohilla Fm. This brine resembles those encountered in borehole Devora 2A and other boreholes penetrating the Mohilla Fm. The only difference is the very high Cl-/Br- values of 11438 in MJ1 indicating the presence of halite beds which should be regarded as a local occurrence in borehole MJ1 and was not found in Devora 2A. Such very high ratios were reported from Triassic sequences of gypsum interbedded with halite in Spain, Portugal, Switzerland, Germany, and Austria [35]. These brines were formed by massive ablation of halite and gypsum/anhydrite with significant influence of igneous processes [31]. The formation of such brines could have also been chemically linked to the very intensive Triassic volcanic activity in northern Israel, as evidenced from numerous boreholes (Asher, Yagur, Haifa, Atlit, and Devora 2A) (Fig. 2 in [29]). This is a typical Ca2+-Cl- brine with very high values of 132.4, , and (Table 2). The temperature of the brine was 160°C.
Bottom hole sampling at the depth of 5000 m revealed a brine that is of different composition. Its Cl- concentration is lower by 23 g/l, the concentration of K+ is much higher (+2.23 g/l), the Cl-/Br- value decreases from 11438 to 140 indicating the absence of halite, and finally, it has a much lowerfactor (88 instead of 132) (Table 2).
Triassic seawater is evaporated yielding increasing Cl- and decreasing SO42- concentrations. This latter trend is similar for the brines of boreholes Devora 2A, Rosh Pinna, and MJ1.
6.5. Brines from Jurassic Beds
Most saline Jurassic waters are compositionally comparable with seawater. The Jurassic seawater is diluted with freshwater causing decreasing SO42- and Cl- concentrations. The Jurassic Arad Group hosts at present two types of brines. In its lower part, the Cl- concentration is in the range of 39-105 g/l with low concentrations of sulfate (380-1000 mg/l). These are typical Ca2+-Cl- brines with , Na+/Cl- of 0.49-0.68, and low Cl-/Br- of 60-127. Such brines may have formed as the result of the evaporation of seawater. In the upper half of the Arad Group, the brines are of an entirely different type. These brines show Na+/Cl- values in the range of 0.78-1.18 and much higher Cl-/Br- values (111-248). values are much lower than in the deeper part of the group, i.e., in the 0.81-38 range. These brines could originate from seawater after dilution of evaporites. In borehole MJ1, no impervious layers separate between the Jurassic and Triassic sequences. In the Haifa Fm (1503-4626 m), two types of brines were encountered by swabbing in selected strata. The two brine bodies could be separated by the Rosh Pinna shales (Table 3).
During the tests, the shut-in pressure of the brine originating from Lower Jurassic beds rose to -632 m bsl and sometime later reached the level of -380 m bsl. Shut-in tests in Upper Jurassic beds indicated pressures of 12 atmospheres above KB, i.e., about -60 m bsl. This data is of primary importance for the management of the regional water resources of the area.
PHREEQC inverse modeling revealed that most brines in the Jurassic Fm seem to be derivatives of Triassic brines and basaltic water. Furthermore, it indicates that the brines encountered above the Rosh Pinna shales (sample ER51, 1775-1784 m, Table 1) could be related to the brines in Lower Jurassic beds (sample ER25 3310-3325 m) by dilution, dissolution, or precipitation of calcite (depending on the salinity of diluting solution), dissolution of anhydrite and silica, exsolution of CO2, and cation exchange on clay particles of shales. The low-salinity end-member in modeling is represented by groundwater from well Beqaot 2.
6.6. Brines from Lower Cretaceous Beds
The chemical compositions of fluids derived from different Lower Cretaceous Fms (Kurnub/Hatira Group) elsewhere in the country are very similar. The Cl- concentrations reach as high as 36 g/l. The values of Na+/Cl- and Cl-/Br- in these fluids (0.87-0.98 and 250-477, respectively) indicate that they could have evolved from seawater which did not reach the point of halite precipitation. The high could be the result of the involvement of both dolomitization and chemical reduction of sulfates.
In the exposed and buried parts of the Fari’a-Beqaot anticline and close to borehole MJ1, groundwater originating from Lower Cretaceous formations flow out from several springs (Hammam el Malih (HEM), Ein el Jamal, and El Hamma) and from boreholes Mehola 5 and Argaman (Figure 1). All these sources are concentrated in an area which is densely faulted. All springs are thermal and emerge along faults. HEM flows out from volcanic rocks exposed by faulting in the core of the anticline. In the MJ1 borehole, swabbing in Lower Cretaceous beds was carried out in the 1228-1232.5 m depth interval. By comparing the chemical parameters (Table 1), it appears that sources ER51-58 and Mehola 5 (at 265 m) relate to the same hydrochemical family which includes also the water of springs Hammam el Malih and Ein el Jamal and the water in all northern Argaman wells. As confirmed by PHREEQC inverse modeling, the chemistry of water from Lower Cretaceous formations could have been formed by natural mixing of Upper Jurassic brine with freshwater as exploited at present from well Beqaot 2.
The water encountered in the deeper parts of the Kurnub Group (sample ER58) was more saline (7.38 g Cl-/l) than the water in Upper Jurassic beds (4.07 g Cl-/l) encountered at a 500 m deeper level (ER51) (Table 4). This “inverse phenomenon” could have been caused by the close vicinity to the major fault which caused the previously mentioned stratigraphic lacuna. This major disturbance could have created a conduit facilitating the interaquifer connection. The high pressures were probably the driving force raising the deep hot brine (72°C) to 1280 m. This geothermal anomaly disappeared at a later stage when the casing was lowered closing the interconnecting fault zone. The chemical parameters indicate that the saline water (7.38 g Cl-/l) encountered in the Kurnub Group beds was the product of dilution of the Upper Jurassic brine sampled on 7.11.2018 (27.9-50.9 g Cl-/l). The Cl- concentrations are similar to those measured in springs and water wells in the area.
The Ha’On brine with which is encountered along the eastern and southern shores of Lake Kinneret represents seawater which invaded the Rift and underwent about 37% evaporation (derived from data according to McCaffrey et al. [36]). In spite of different Ca2+/Mg2+ values, the similarity of Na+/Cl-, Cl-/Br-, and (Ca2++Mg2+)/Na+ values suggest that the Ha’On brine or any of its diluted equivalents was converted into the Tiberias brines outflowing along the western rims of the Rift. Under the given petrologic conditions with alkaline olivine basalt flows covering the Tertiary limestones south of Tiberias, it is reasonable to infer that the chemical evolution of Ha’On into Tiberias brines is the result of dolomitization of limestones and alteration of plagioclase to albite, illite, and chlorite. Halite and gypsum are enhanced either by leaching evaporites or by mixing with an evaporite dissolution brine of corresponding composition [34]. A third trend leads to Ha’On and THS brines occurring close to Lake Tiberias. Their compositions indicate a lesser degree of evaporation of Tertiary seawater and dilution by freshwater than in the case of the Triassic brines.
The subsurface geological structure of the Bet She’an Valley area is dominated by an intricate system of major faults and fissures related to the Rift system. Notwithstanding the dense network of faults in the subsurface of the Bet She’an Valley, there are no indications for upflow of deep-seated Triassic-Lower Jurassic brines into the overlying Lower Cretaceous-Lower Miocene sequence. In view of this evidence, it is clear that along the MJ1 borehole and in the surrounding areas, there are several well-defined, separate, and easily identifiable bodies of groundwater in the salinity range of brine to brackish. The thermal gradient in the Lower Cretaceous-Lower Miocene sequence is considerably lower than that in the deeper (Jurassic-Triassic) formations. This is due to the cooling effect caused by contemporary recharge by precipitation on exposed replenishment areas on Mt. Gilboa and on the Fari’a-Beqaot anticline.
6.7. Geochemical Families
In the Rift and in adjoining areas, there are two definite source brines: the Triassic brine and the Late Tertiary one, also known as the Rift brine [3].
The Tertiary Ha’On-Tiberias family is neither a derivative of Jurassic brines nor a derivative of Triassic brines but is a separate entity [34]. The brines encountered in Jurassic and Cretaceous beds represent ancient mixtures of Triassic and Tertiary brines involving various water-rock chemical transformations The arrows on the triplot point to compositional changes due to dissolution of halite, albitization of plagioclase, dolomitization of limestone, dissolution of dolomite, and mixing of Triassic brine and THS (Figure 7).
The deep brine in borehole Devora 2A (4836-4864 m; 123.4 g Cl-/l) plots together with the 5000 m brines in borehole MJ1 suggesting common Triassic origin (Figure 5). The deep Rosh Pinna brine has similar genetic roots like the Triassic and Lower Jurassic brines encountered in MJ1 suggesting common Triassic origin. Hence, the deep brines encountered in boreholes MJ1, Devora 2A, and Rosh Pinna 1 form a tightly connected regional hydrochemical family. This is also backed up by the dendrogram (Figure 8). Though of similar origin, the brines of Devora 2A and Rosh Pinna had further histories of their own. Whereas the Rosh Pinna brine was diluted by THS water, such a process was not observed in the brine of Devora 2A. The Lower Jurassic samples of MJ1 and shallow and low TDE Rosh Pinna brine plot on the same dilution line of the Triassic and the probably freshwater-diluted Ca2+-Cl- THS brine of the Ha’On/Tiberias brine family. The dilution by these brines could have occurred during the Pleistocene, and therefore, the dilution of the shallow Rosh Pinna brine is also a very young process (Figure 4). All diagrams in Figure 5 indicate by red lines a close association and a fixed relationship between the Tiberias Hot Springs (THS) brines and the diluted Rosh Pinna brine.
6.8. Implications on Regional Water Management
The Ca2+-Cl- brine penetrating into the pristine (currently replenished) groundwater in springs and water wells causing their salinization is the Miocene Rift Brine and not the older (Jurassic and Triassic) brines which were discovered and thoroughly documented in borehole MJ1. This most important conclusion is valid not only for the area of the Bet She’an and Harod Valleys but also for the whole Central Jordan Valley including Lake Kinneret with its severe salinization problems. The factor common to all brines in the region is the very high hydraulic pressures. To the present date, there are no clear answers to the question on the factors generating the very high hydraulic pressures (up to several hundred atmospheres) which mobilize the brines originating in deep-seated formations. We may infer heavy lithostatic pressures or structural pressures in the immediate vicinity of the transfer. Very high pressures have been previously reported in wildcat Devora 2A and in other wildcats and were now clearly confirmed for wildcat MJ1. In wildcat Rosh Pinna 1, the pressure release tests were not brought to completion and maximal pressure was not attained. However, in the partial test, the water rose from the depth of 3864 m to 566 m (-242 m below MSL), i.e., under the pressure of 330 atm (personal communication by N. Schlein to Rosenthal 1988). Hence, the data from these three deep boreholes indicates a mechanism in which the deep brines are piston-driven upwards and possibly also laterally.
Until recently, there were two main obstacles to developing and managing fresh groundwater resources in areas structurally related to the Rift. These obstacles were a lack of information on the possible presence of brines in the subsurface and of their hydraulic heads. The recently acquired information (on subsurface structures, hydrochemical stratigraphy, and pressures) supports water planners with precious information.
Another important aspect is the massive development of water desalination plants intended to replace groundwater, the salinity of which deteriorated during the last decades due to penetration of brines. Wildcat MJ1 revealed the existence of very great volumes of brackish groundwater (<10 g Cl-/l) which could be used for desalination. Unfortunately, the low permeabilities characterizing Lower Jurassic and Triassic beds do not facilitate disposal of the industrial brine produced during the process of desalination. Moreover, the high artesian pressures which were discovered in the subsurface will push up and out any injected liquid, thus causing major environmental damage. Therefore, the main efforts should be concentrated on the removal of desalination brine, whereas the availability of brackish brine serving as a raw material for desalination does not present any difficulty.
7. Conclusions
The hydrogeological and hydrochemical data collected during the drilling of wildcat MJ1 significantly improved the understanding of groundwater salinization in the Rift and in the structurally connected areas.
For the first time, there is a very clear stratigraphic picture of permeable and nonpermeable strata (aquifers and aquicludes) in the subsurface of the area which are well separated by impervious strata of clays or shales. The various tests proved that the saline groundwater bodies along the MJ1 borehole are characterized by specific and well-defined hydraulic heads and are not fused into a unique and big mass of brine. The difference in the thermal gradient between the Jurassic and Cretaceous formations is a clear indication of isolation and separation between the upper and younger stratigraphic section and the deeper Jurassic-Triassic beds.
In view of this evidence, it is clear that along the MJ1 borehole and in the surrounding areas, there are several well-defined, separate, and easily identifiable bodies of groundwater in the salinity range of brine to brackish. Their distribution along the stratigraphic column is due to their specific gravity, and there is no evidence of interformational mixing. The hydrochemical evidence proves unequivocally that in the subsurface of the drilled area, there are two source brines and two brines which evolved as the result of their mixing and rock-water interaction. Each of these well-separated brines can be easily identified by their chemical composition.
The two source brines are as follows: (i)The evaporative brine formed during the Triassic and now stored in the Upper Triassic Mohilla formation(ii)The Late Tertiary (so-called) Rift brine
The two brines which derived from their mixing and chemical transformations and may be identified by their characteristic and different chemistries and hydraulic heads are as follows: (i)The marine Lower Jurassic brine occurring below the Rosh Pinna shales(ii)The Upper Jurassic brine is clearly separated from the previous (deeper) one by the Rosh Pinna shales(iii)The Jurassic saline waters are dilutions of seawater (no evaporation) with the dissolution of small amounts of sulfates. The brines in boreholes Devora 2A and Rosh Pinna may have the same geochemical source as in borehole MJ1, i.e., evaporated Triassic seawater with the dissolution of large amounts of carbonates
The Mehola Group represents diluted seawater. The water emerging from Ha’On and THS are derivatives of evaporated Tertiary seawater and dilution with freshwater. The increase of Ca2+ in THS is due to the dolomitization of limestones.
The results of the present study bear on issues of regional water management. It occurs that salinization of regional pristine groundwater in springs and wells is caused only by the Miocene Rift brines which were encountered in the upper horizons of the regional hydrochemical stratigraphy. The older (Jurassic or Triassic) brines are hitherto not involved in such a process. The very high pressures preclude the development of desalination projects because they exclude the disposal of brines by injection into deep-seated strata.
Data Availability
All data used to support the findings of this study are included within the article in the relevant tables. For any data or explanation of data, please write to the senior author.
Conflicts of Interest
I herewith declare that during the research activities and while preparing our manuscript 9812597 for publication, between the authors or between them and any other parties, there were no conflicts of interests in the domains of finances, affiliations, and intellectual properties and of any personal, ideological, academic, or other aspects.
Acknowledgments
The authors are deeply indebted to Zion Oil and Gas, Inc., for their help, consideration, and deep interest in the regional problems of salinization. Many thanks are due to Afikey Maim-Emeq Ha’Maayanot (Bet She’an Water Supply Co.), its director Mr. Ran Bin-Nun, and Mr. Gadi Gal for supporting, financing, and helping in all activities related to elucidating the hydrochemistry and hydrology of brines encountered during the drilling of wildcat MJ1. They became great friends and spared no effort to facilitate our work. Special thanks are due to the team of the Geochemical Laboratory of the Geological Survey of Israel in Jerusalem and to its head Dr. Naomi Porat for their excellent analytical work, special advice, and friendly attitude. Many thanks and deep appreciation to Dr. Boaz Arnon who interpreted water salinity from the logs.