It is not uncommon for uranium isotopes to be used as mechanistic or time scale tracers of natural processes [1
]. Uranium isotopes may be applied to determine the radiological suitability of groundwater for drinking purposes [3
], applied in ore bodies exploration [7
], or applied to assist in understanding historical climates [10
]. Of special interest is the study of the uranium redistribution in the sandstones and siltstones of aquifers of paleo-valleys, which are associated with the formation of increased concentrations of uranium in both rocks and in groundwater [13
]. A typical area with similar conditions is the Northern Dvina basin (NDB)—a paleo-valley in NW Russia (39°30′–41°57′ E and 64°06′–64°48′ N) (Figure 1
). The large industrial and urban agglomeration of Arkhangelsk–Severodvinsk–Novodvinsk is on this territory, and its water supply is currently affected using surface water. Groundwater has the best quality, but the rationale for the radiological safety of its use is required.
The aim of this paper is to carry out the paleoreconstruction of hydrodynamics of NDB with an assessment of the direction of the evolution of the uranium-isotope composition of rocks and groundwater during the Pleistocene-Holocene transition.
The following questions are of interest:
What trends are observed in the evolution of the uranium-isotopic composition of rocks?
Uranium almost completely transfers into the groundwater from water-bearing sediments via cracks and pores, through which water moves from recharge to discharge areas. Where are the maximum concentrations of uranium located in the water bearing rocks and how widely are they developed?
When the operation of water supply wells is carried out it can lead to upwelling water with higher radioactivity from the underlying horizons. On which sections of the NDB is it the most dangerous with respect to upwelling water?
The impact of uranium on water quality is determined by its concentration (chemical aspect) and isotopes activity (radiological aspect). How does the residence time (age) of groundwater in the aquifer influence on its radioactivity?
2. Materials and Methods
The object of this study is the Padun aquifer of the NDB. The Northern Dvina basin represents an onshore continuation of the Dvina Bay of the White Sea. It has been formed in the sequence of sediments with different ages, namely, the Middle Carboniferous carbonate-terrigenous (C2
), Upper Devonian-Lower Carboniferous terrigenous (D3
), and Vendian Padun (Vpd) and Mezen (Vmz) Formations (Figure 1
The NDB is 120 km long. Maximal dimension of its transverse section (along the coast of the Dvina Bay) is as follows: width of the top zone is 100–110 km, width of the bottom is 30–40 km. The bottom is located at 60 m below sea level (m.b.s.l) and the slopes are inclined at an angle of 2–3°.
The basin is largely filled with clayey sediments from the Mikulinian interglacial (mIIImk) Boreal sea. The clays are underlain by a 10 to 15 m-thick sequence of loams from the Moskovian moraine (gIIms). The underlying Vendian terrigenous sequence (600–700 m) is composed of alternating sandstones, siltstones, and mudstones. The Mikulinian clays are overlain by a 10 to 15 m-thick sequence of Valdaian glaciation (gIIIvd) sediments, which are largely represented by moraine boulder loams with local fluvioglacial and glaciolacustrine sands. The valley of the Northern Dvina River is filled with Upper Pleistocene and Holocene (QIII-IV) sandy clay sediments.
The upper part of the Vendian aquifer in the central part of the depression is characterized by inversion type of vertical hydrochemical zonality and decreasing groundwater mineralization from 29 to 12 g·L−1
). This pattern is primarily caused by the freshening of groundwater due to the long-term existence of terrestrial conditions during the Mesozoic and Cenozoic. The desalination of groundwater ceased after the Mikulino Sea transgression ~130 ka ago. Subsequently, the upper part of the Vendian aquifer began to receive saline interstitial water, which was squeezed out under the load of the Valdai glacier from Mikulinian interglacial marine clays. After the formation of the Northern Dvina River valley due to the erosional activity of water that flowed from the melting glacier, the valley cut the Mikulinian clayey sequence almost to its base. Saline water from the aquifer sediments of the terrigenous Padun Formation began to flow into the river valley and were later replaced by freshwater from the recharge area (Figure 1
Sediments of the Padun Formation of the study area were accumulated in the coastal marine environment. The Padun Formation of the Vend is 90–170 m thick and located above 150 m.b.s.l. It is composed of siltstones and sandstones with intercalations of mudstones. The rocks are characterized by reddish brown color on the NDB slope with light green lenses and spots the under central part of the NDB (at the NDB bottom).
Forty-four rock samples were collected from 9 boreholes down to depths of 160 m, and 25 groundwater samples from 23 boreholes of depths up to 300 m (Figure 1
; Table 1
and Table 2
Water temperature, pH, Eh, DO, alkalinity, total dissolved solids (TDS), Fe concentrations, U concentrations (C8W
U activity ratio (ARt
C were determined in groundwater, as described by Malov [14
]. U content (238
U activity ratio, Fe content in rock samples were determined, as described by Malov et al. [15
A piston flow model was used to estimate the 14
C residence time of groundwater in the aquifer. The basic equation for groundwater dating is:
is the groundwater age,
is the 14
C decay constant,
is the 14
C after adjustment for the geochemical and physical processes in the aquifer (without radioactive decay), and
is the measured 14
C value of the total dissolved inorganic carbon (TDIC) [16
When analyzing the results of carbon isotope determinations, the following models were used to determine : (1) for δ13C < 0.5(δ13Cg + δ13Cs)—Mook model, taking into account the isotope exchange between soil CO2 and TDIC (indices ‘g’ and ‘s’ are soil CO2 and solid carbonate minerals, respectively); (2) for δ13C ˃ 0.5(δ13Cg + δ13Cs)—Han and Plummer model, taking into account the isotope exchange between solid carbonates and TDIC.
Further, a mass-balance model [14
] was used to estimate the 234
U residence time of groundwater in an aquifer under oxidizing conditions for U.
The main calculated equations are as follows:
is the groundwater residence time in an aquifer (age),
is the 234
U decay constant,
is the measured concentration of 238
U in solution at the point of sampling,
is the retardation factor,
is the measured 234
U activity ratio in the pore fluid at the point of sampling,
is the solid mass to fluid volume ratio,
is the concentration of 238
U in solid phase, and
is the recoil loss factor.
is the initial 234
U activity ratio in the pore fluid at the recharge area of the groundwater.
A model of the down-flow radioactive decay of 234
U excess in solution [18
] was used to estimate the 234
U residence time of groundwater in an aquifer under reducing conditions for U.
The main calculated equation is as follows [18
3. Results and Discussion
Groundwater on the NDB slope (Figure 2
b,d; Table 2
) represents fresh water. Their Eh ranges from −68 to 106 mV, pH from 7.7 to 9.1 and alkalinity from 1.0 to 5.2 meq/L. The residence time this water (age) ranged from 0.3 to 16.4 ka. Very low Fe concentration (4–203, average 41 ppb), medium ARt
(1.3–5.9, average 3.0) and high C8W
(0.3–12.5, average 5.4 ppb) are characteristic of this water. At the NDB bottom, near the redox barrier old salt water has ages ranging from 17 to 33 ka, TDS from 4 to 13 g/L, Eh from −38 to 2 mV, pH from 7.4 to 7.8 and alkalinity from 1.7 to 4.2 meq/L. Maximum C8W
(7.2–15.4, average 12 ppb), and high ARt
(4.8–7.2, average 5.9) and Fe (0.4–1.9, average 0.8 ppm) are observed here. Below the redox barrier exist the oldest (234
U age from 0.1 to 0.5 Ma) and the most salty (TDS from 9 to 22 g/L) waters. The Eh ranged from −23 to −151 mV, pH ranged from 7.6 to 9, alkalinity ranged predominantly from 0.2 to 0.9 meq/L. The minimum C8W
(0.1–1.4, average 0.4 ppb) and maximum Fe (0.7–8, average 3.7 ppm) are observed here. ARt
decreases from 7.16 ± 0.94 to 2.97 ± 0.45.
Below the redox barrier, at the NDB bottom, the uranium becomes restored to U4+
and precipitates (Figure 2
b), its concentration in rocks reaches 20 ppm, and 234
U activity ratio in rocks decreases to 0.5–0.9 (Figure 2
a,c; Table 1
). However, complete precipitation does not occur because the process of recoil loss continues and both isotopes of U enter into the water. For every 1 atom of 234
U in rock, ~18,000 atoms of 238
U exist; therefore, recoil atoms inevitably encounter and knock other atoms from the crystal lattice, creating an area of disorder. The uranium atoms in the disordered zone are first transferred to water, resulting in a disturbance of the radioactive equilibrium in the water. Evidence that recoil atoms are not transferred into water alone but carry away a certain amount of 238
U atoms explains why ARt
under reducing conditions usually does not exceed 10–20. This is possible if 1000–2000 238
U atoms are transferred into the water with each recoil-atom 234
The maximum ARt
in groundwater directly close to the redox barrier is estimated to be 7.16 ± 0.94 (sample Mi2014
). In other samples, the activity ratio is lower, namely, from 6.72 ± 1.05 to 2.97 ± 0.45 (Figure 2
d; Table 2
), which suggests a greater age for the groundwater in these samples, in accordance with a simple model of the down-flow radioactive decay of excess 234
U in solution.
The U content in the red rock ranged from 0.11 ppm to 2.89 ppm, average 1.2 ppm (Table 1
). The U content in the green rock ranged from 1.52 ppm to 20.9 ppm, average 6.17 ppm. 234
U activity ratio in the red rock ranged from 0.87 to 1.51, average 1.16. 234
U activity ratio in the green rock ranged from 1.18 to 0.54, average 0.92 ppm. The average value of U content on the slope of the NDB is 1.36 ppm, 234
U activity ratio is 1.2; at the bottom of the NDB average values are 3.07 and 1.03, respectively (Table 1
; Figure 2
The average value of Fe content from 18 samples of red rocks on the slope of the NDB is 3.67%. In two samples from wells 184-1 on the NDB bottom Fe in the red rock is 2.33%, Fe in the green rock is 1.83%.
In Upper Vendian products of the rocks weathering were transferred to the study area from the nearby eastern tip of the Baltic Shield and were deposited together with buried organic matter. In subsequent geological periods (Upper Devonian–Lower Carboniferous) NDB was also in coastal marine and lake environments under a hot humid climate [19
]. In such anoxic environments, early diagenesis conditions favor the reduction of U6+
into low solubility U4+
, which decreases U concentrations in overlying waters and sediment pore-waters [21
]. This period was the most favorable for the supergene ore formation [22
]. During these periods apparently this was the main flow of uranium to the study area and its deposition as a result of hydrolysis, adsorption on natural sorbents promoted changes in the oxidizing conditions of the environment in reducing conditions.
The transgressive period, in Middle Carboniferous-Permian led to the formation of a cover of the terrigenous-carbonate deposits (see Figure 1
), however, during the long continental environmental interspace in the Mesozoic–Pliocene was formed the NDB palaeo valley. Its depth could reach 250–300 m. The depth of the valley and its Pliocene age are confirmed by the data of other paleovalleys of the East European platform [24
]. Within its boundaries, most of the Paleozoic sediments were destroyed, and Vendian deposits were brought to the surface (see Figure 1
The processes of chemical weathering of Vendian deposits led to the formation of a strong oxidation zone, developed above 250 m.b.s.l. The greater increase in the proportion of Fe2+
iron is typically the presence of bitumen, organic carbon, pyrite, and rock which maintain the gray-green color [27
]. This period probably provided the main redistribution of uranium accumulated during the Paleozoic in the Padun aquifer of the NDB. The inverse correlation between the concentrations of uranium and iron is typical for Padun rock (see above). It is a result of removal of U from NDB slopes in oxidizing conditions and accumulation of U at the bottom of the NDB in reducing conditions, and accumulation of Fe on the slopes and removal from bottom. As a result, a significant part of the equilibrium U on the slopes of the NDB had been replaced by a newly formed “hydrogenic” U (precipitated from groundwater), with an initial AR0
of modern fresh groundwater = 3, and the initial U content of the rocks (U0
) ≈ U content of modern siltstones of the region (2.6 ppm). The ending of the period of co-precipitation of hydrogenic uranium with iron hydroxide on NDD slopes can be estimated from the equation [17
= 2.8263 × 10−6
is average 234
U activity ratio of the red rock = 1.16 (see Table 1
What is obtained is: t1
= 0.9 Ma, which should roughly correspond to the period of a sharp cold snap in the region and filling of the paleo-valley by clay material [30
]. The duration of the subsequent removal of hydrogenic uranium from the NDD slopes can be estimated from the equation [14
—concentration of U that was passed from the red siltstones to the water for time t
= 2.6 ppm (see above) and 238
U is the red rock average 238
U content (1.2 ppm) (see Table 1
). Consequently С8W
= (2.6 − 1.2) = 1.4 ppm; Ms
—solid mass to fluid unit volume ratio = 9.2; С8R
is average concentration of U in solid phase for a time t
U): 2 = (2.6 + 1.2): 2 = 1.9 ppm; Rd
—dissolution rate for U; R
= 24; and average Rd
= 3.6 × 10−6
We get: t2
= 0.5 Ma. The difference (t1
) should roughly correspond to the duration of glaciations and marine transgressions in the past 0.9 Ma, when the movement of groundwater in the Padun aquifer was significantly delayed or even absent [34
]. The removal of uranium was not apparent, but radioactive decay continued.
Lower values of average 234
U activity ratio in the green siltstones (0.92) (see Table 1
) can be explained by the fact that these deposits have reached a steady state of the 234
U activity ratio that depends only on their size (the average grain size dp
≈ 30 µm) [38
], because they were under reducing conditions over 1 Ma. A significantly higher content of uranium in them compared to red siltstones shows a considerable variability in the permeability values of the aquifer, whereby they were away from the paths of groundwater filtration and have retained uranium. A similar situation is typical for the preserved iodine water lens here (sample Bb2003
), the source of which is the iodine-containing seaweed from the Mikulinian interglacial Boreal sea [39
The processes of chemical weathering of Vendian deposits led to the formation of a strong oxidation zone, developed above 250 m.b.s.l. The inverse correlation between the concentrations of uranium and iron is a result of removal of U from NDB slopes in oxidizing conditions and accumulation of U at the bottom of the NDB in reducing conditions, and accumulation of Fe on the slopes and removal from bottom. Almost all the U on the slopes of the paleo-valley could be replaced by a newly formed hydrogenic U with a higher 234U/238U activity ratio.
Afterwards, dissolution and desorption of hydrogenic U occurred from the slopes of the paleo-valley during periods with no glaciations and marine transgressions. Elevated concentrations of U are preserved in reduced lenses at the NDB bottom. In these areas, the most dangerous is the upwelling of groundwater from the underlying horizons, since during the operation of water supply wells it can lead to the creation of local zones of oxidizing conditions in the perforated screens zone and the transition of uranium into the aqua solution.
For groundwater under oxidizing conditions, an increase in the concentration of uranium is characteristic of an increase in the residence time (age) of water in the aquifer. Also, the 234U/238U activity ratio increases, increasing the radioactivity of groundwater. Therefore, the most rational conclusion would be to aim at using the groundwater from the slopes of the NDB for drinking water supply.