Identification of River Valley Areas Threatening the Chemical Status of Groundwater, in the Example of the Upper Course of the Ner River Basin, Central Poland

Samples of river sediments in the E, C and H profiles were collected from the river bed using a sludge pipe. Sampling sites were selected on the basis of earlier gradientmeter measurements of water flow directions in river alluvium. Places representing zones of river water infiltration into the bottom of the channel, as well as places of its drainage after mixing with groundwater, were selected (Fig. 1). Determination of chemical elements in river sediments was carried out using the aqua regia extraction method. The content of elements in the obtained filtrate was determined by the ISO 11885/ISO 17294-2 method in the Wessling laboratory and by ICP-OES emission spectrometry using a Perkin-Elmer device in the Laboratory of Geomicrobiology and Environmental Geochemistry of the Faculty of Geology, University of Warsaw.

Cation exchange capacity (CEC) was also determined in the sediments collected from the Ner river bed. The CEC value was estimated by the Breeuwsma formula (Appelo, Postma 1993), based on the percentage of particles <2 m and the percentage of organic matter. The percentage content of particles <2 m was determined based on the analysis of the granulometric composition of the samples, while the content of organic matter was determined using the loss on ignition method. In addition, in all samples, the CEC was determined by the Kappen method. The research was carried out in the Applied Geology Laboratory of the Faculty of Geology, University of Warsaw.

The conductivities of the individual layers were subject to changes at the calibration stage. The initial values were adopted on the basis of own calculations from test pumpings, data from borehole charts (CBDH) and result maps of the modelled Main Groundwater Reservoir No. 401 on a regional scale (Rodzoch et al. 2013). The isotropy of the medium in the directions (x and y) was assumed; for the direction (z), the values were one order lower.

The values of infiltration recharge were mapped using the II type boundary condition (Q = const.). The amount of recharge in the range from 0 mm · year⁻¹ in the drainage zones to 120 mm · year⁻¹ was assumed. With a shallow groundwater table, mainly in the drainage zones, the amount of groundwater evaporation was also determined. The maximum annual values were assumed to be 50 mm · year⁻¹ in the case of the presence of a groundwater table on the ground surface, and the so-called depth of extinction of the impact of this process at 1.0 m (Dowgiałło et al. 2002).

In the study area, the groundwater withdrawal was mapped as a type II point condition with known outflow rate, adopted on the basis of data obtained from the POBORY database, Polish Geological Institute-National Research Institute (average value over many years).

The III type boundary condition represented drainage through watercourses and the lateral inflow through the model boundaries (Fig. 7). The III type RIVER condition was set for model cells simulating rivers characterised both as boundaries of the model and within the model. For this condition, the width of the watercourses was introduced, the layer of water in the watercourses was adopted on the basis of field observations, supplemented with data from the Map of the Hydrographic Division of Poland in the scale of 1:200,000 (Czarnecka 2006). The DRAIN condition was set at the drainage ditch courses visible in the image of the WMS orthophotomap service (GUGiK 2023).

Fig. 7.

The condition of the III type – General Head Boundary (GHB) – was used to map the vertical boundaries of the model along a separate fragment of the system. After the hydrogeological analysis, it was found that the boundaries of the modelled system are open to water exchange with the surroundings from the south and east for the first layer and from all sides for deeper layers.

The calibration process consisted in finding a solution to the inverse problem by the method of successive approximations. The solution found for the condition of the aquifer system was confirmed by calibrating the heads at 180 points, including aquifers marked as Q1, Q2– Ng, Cr3_1 and Cr3_2. In the calibration process, the properties of the filtration field were refined by increasing or decreasing the conductivity. The obtained range of values was slightly lower than the initial values: from 0.8 m · day⁻¹ to 5 m · day⁻¹ for the subsurface aquifer (model layer no. 1) and 5–10 m · day⁻¹ for the Quaternary aquifer (model layer no. 3). The conductivity of the top part of the Upper Cretaceous was determined at 4–7 m · day⁻¹, the middle part at 3 m · day⁻¹ and the bottom part at 1 m · day⁻¹. The conductivity of aquitard formations was set to the value of water permeability in the range of 0.00005–0.05 m · day⁻¹.

As a result of the calibration, errors characterising the compliance of the model calculation results with the input data at the calibration points were determined (Fig. 8). The maximum difference between the measured and calculated heads was -11 m; the absolute mean was 2.81 m. The estimation error of 0.27 m and the absolute mean error of 3.68 m were considered low enough to complete the calibration process and to recognise the credibility of the obtained model. During the calibration, special attention was paid to adjusting the modelling results to the groundwater levels at the Ignacew wells, assuming the position of the dynamic water table caused by the withdrawal (concordance approx. 1 m).

Fig. 8.

The Ignacew wells’ (A1, A2, A3) work was mapped by assigning appropriate discharge rate. During field works, it was found that the actual value of withdrawal is similar to the value approved in the water law permit and amounts to approximately 90% (Table 3). To assess the impact on the Quaternary aquifer and the surface water of the Ner River, remaining in hydraulic contact with the Upper Cretaceous aquifer, the heads of all aquifers were determined in quasi-natural conditions with inactive wells’ simulation and in the conditions of the active wells.

The numerical model of the filtration field made it possible to determine the 25-year time (t = 9125 days) of the lateral inflow to the Ignacew wells (A1, A2, A3), owing to the recognition of the directions and filtration velocity. For this purpose, the MODPATH module included with Visual Modflow Flex 7.0 was used. This module enables ‘tracking’ of water particles flowing in the three-dimensional space of the model.

Model tests have shown that when pumping the Ignacew well with the assumed actual rate, the water inflow comes from different directions: for the A1 well, from the north and south ; for the A2 well, from the east and south and for the A3 well, from the south and south-east (Fig. 9). The inflow to all wells is through the Ner river valley in the area of C and E research profiles and hydrogeological sections, where the presence of hydrogeological windows between the Quaternary and Cretaceous aquifers was found. The range of influence of the well on the river valley is the smallest for the A2 well and the largest for the A3 well. The inflow time ranges from 1 year to 25 years, and at the intersection with the Ner river bed, it is from 6 years to 15 years. The work of A1, A2 and A3 wells with assumed rate causes a local reversal of piezometric pressures (Fig. 10). In the Ner valley, the Upper Cretaceous aquifer groundwater table is lower than the Quaternary aquifer, which makes it possible for surface water to infiltrate to the main usable aquifer. This depression ranges from 0.04 m to 1.99 m (the largest on the northern border of the valley), while in natural conditions, the level of the Upper Cretaceous table was higher than the Quaternary with a value of 0.01–0.39 m.

Fig. 9.

Fig. 10.

Since the previous studies, including model studies, showed the possibility of seepage of water from the vadose zone and leakage of water from the overlying layers to the Upper Cretaceous aquifer, the next stage of work was to determine the potential migration risk of pollutants accumulated in the overlying layers. For this purpose, sorption parameters of river sediments, accumulated pollutants in them and the chemical composition of surface water and groundwater were determined.

The results of laboratory analyses of the element content in the collected samples of the Ner alluvium (Table 4) were compared with the archival data contained in the Hydrogeochemical Atlas of the Łódź agglomeration (Lis, Pasieczna 1998). The contents of most elements in the alluvium are small and definitely lower than the average values determined for the entire agglomeration area. The exceptions are cadmium, chromium, copper, mercury, strontium, silver and zinc. Chromium content higher than average was found in all tested samples. According to Lis and Pasieczna (1998), the main potential source point of contamination with this element are discharges of sewage into the tributaries of the Ner, i.e. the Łódka and Dobrzynka rivers. The estuary of the Łódka is located 1.8 km from the farthest upstream Ner, the H cross-section (Fig. 5). In 1998, the Łódka alluvial had the highest average chromium content in the entire area of the agglomeration, amounting to 102 ppm. The highest content of chromium in a single measurement (977 ppm) was determined in the Dobrzynka alluvium, whose estuary to the Ner is located from the H-section at a distance of about 6 km upstream. In the tested samples, the highest concentrations of chromium were found in the H-section (69.60 ppm at H2). The decrease in the content of this element in individual research profiles follows the direction of the flow of the Ner, up to the value of 18.7 ppm at point C2 located in the most downstream C section (Table 4).

Discharges of municipal and industrial wastewater are also a potential source point of cadmium, copper and zinc contamination of the alluvium of the Ner, with the content of these elements higher than the average for the entire Łódź agglomeration in the H and E profiles (Table 4). Significant mercury content reaching a maximum of 0.74 ppm (point H2) was also found in the H and E profiles, and the origin can be seen mainly in the discharge of sewage into the Olechówka and Dobrzynka rivers flowing into the Ner (Lis, Pasieczna 1998).

When analysing the elemental composition of Ner alluvium, one should also pay attention to bismuth. The maximum contents of this element reaching as much as 640 ppm were found in the H profile at point H2. This element was not determined during the research on the Geochemical Atlas of the Łódź agglomeration (Lis, Pasieczna 1998). According to Kabata-Pendias and Pendias (1999), in the zone of hypergenesis, bismuth is not very mobile and is part of insoluble compounds retained by iron hydroxides and manganese and organic matter. Bismuth contents in soils reaching several 100 ppm are found near copper smelters and in metal mining areas (Pasieczna 2012). In the case of the Ner alluvium, such high contents of this element can be associated with pollution supplied to the river with sewage discharged by an unidentified metallurgical plant. In the H2 sample, containing the largest amounts of bismuth, the presence of macroscopically visible fragments of easily crumbling sinters was found. Such a focus on pollution may also be indicated by high values of the coefficient of determination describing the dependence of bismuth content in relation to zinc (R² = 0.83), tin (R² = 0.92), lead (R² = 0.93), copper (R² = 0.93) and titanium (R² = 0.98).

Contamination of alluvial sediments with metals causes their toxic impact on aquatic organisms. According to the criteria used by the Chief Inspectorate for Environmental Protection (GIOŚ), using the values of Threshold Effect Concentration (TEC), Probable Effect Concentration (PEC) and Midpoint Effects Concentrations (MEC) (MacDonald et al. 2000, WT-732 2003), the content in all the samples of arsenic, iron, manganese and nickel allowed them to be included in the first level, i.e. uncontaminated sediments that do not have a negative impact on aquatic organisms. In the case of silver, lead, cadmium, chromium, copper and zinc in some samples, and in the case of mercury in all samples in which this element was determined, second level or third was found, indicating sediment contamination, which may have a negative impact on aquatic organisms. The classification was carried out in terms of elements determined as part of the monitoring of river sediments carried out by the Chief Inspectorate of Environmental Protection (Table 4).

The results of the determination of the CEC of the alluvium accumulated in the Ner river bed, by the Breeuwsma formula and the Kappen method, are similar and indicate a low capacity of these formations to store absorbed substances. This is related to the lithology of river sediments, among which the occurrence of fine-grained sands and sands without a fraction of less than 2 m and with a low content of organic carbon, reaching a maximum of 1.36%, was found (Table 5). The results of the Kappen method indicate that the sorption complex is saturated with acidic cations (e.g. H⁺, Al³⁺). The exception is the sample taken at the H2 site, in which the exchange capacity for basic cations (BEC) was two orders of magnitude greater than the electrolytic acidity. It is a sample in which the fragments of easily crumbling sinters were found macroscopically. It is assumed that such a high BEC value in the H2 sample may be the result of dissolution in hydrochloric acid (HCl) of the metals contained in the sediments (e.g. bismuth), and not displacement of basic cations from the sorption complex. It is worth mentioning that due to the lack of the smallest fractions and organic matter, alluvial sediments do not constitute a typical colmation layer here that would retain pollution from the infiltrating river water as known in other studies (Wang et al. 2007).

In terms of the content of macroelements, the Ner water is clearly distinct from both waters taken from river sediments and groundwater (Figs 11 and 12). With their hydrochemical characteristics, the multi-ionic nature of water types (HCO₃–Cl–Ca–Na, Cl–HCO₃–Na–Ca), with a significant share or even dominance of chloride and sodium ions is noteworthy, which indicates their strong anthropogenic transformation. The poor quality of the water of the Ner River is also represented by the values of electrolytic conductivity exceeding 1200 μS · cm⁻¹, low content of dissolved oxygen (below 4.04 mg · dm⁻³) and concentrations of nutrients: nitrate nitrogen (over 4.93 mg N-NO₃ · dm⁻³), ammonium nitrogen (above 1.57 mg N-NH₄ · dm⁻³) and phosphate phosphorus (above 0.88 mg P-PO₄ · dm⁻³). The concentrations of the above-mentioned elements in the Ner significantly exceed the threshold values of good chemical status of surface water set out in the Regulation of the Minister of Infrastructure of 25 June 2021 (Journal of Laws 2021, item 1475).

Fig. 11.

Fig. 12.

In the alluvium water, among the macroelements, there is a clear dominance of hydrocarbonate and calcium ions. The sulphate, chloride and sodium ions appear as secondary in the water type (Fig. 11, Table 6). The values of the electrolytic conductivity (EC) are lower than those observed in river water and are within a wide range of 469 -1070 μS · cm⁻¹. Similarly, in alluvial water, the average content of nutrients is lower than in river water: 1.71 mg N-NO₃ · dm⁻³, 1.26 mg N-NH₄ · dm⁻³ and 0.29 mg P-PO₄ · dm⁻³. River alluviums are an extremely dynamic zone where surface waters mix with groundwater drained by the river. The shares of individual water streams are constantly changing depending on the temporary pressure systems. Changes in the flow of water result in changes in chemistry. The presented research results describe the temporary hydrochemical state of this zone. It can be assumed that the lower values of EC and the content of nutrients are, apart from the activity of living organisms, caused by the mixing of heavily polluted river water with groundwater in the alluvium, which, according to the conducted research, is of better quality. The EC reached a maximum of 676 μS · cm⁻¹, and the maximum content of nitrate nitrogen 1.36 mg N-NO₃ · dm⁻³, ammonium nitrogen 0.67 mg N-NH₄ · dm⁻³ and phosphate phosphorus 0.03 mg P-PO₄ · dm⁻³.

The collected groundwater samples represent the chemical composition of three aquifers (Fig. 11). The water of the Quaternary is characteris ed by mineralis ation ranging from 100 mg · dm⁻³ to 1237 mg · dm⁻³ with the predominance from 100 mg · dm⁻³ to 400 mg · dm⁻³ and domination in the composition of hydrocarbonate, calcium and rarely magnesium ion (HCO₃–Ca and HCO₃–Ca–Mg) (Fabianowski 2002, Rodzoch, Karwacka 2015, Małecki et al. 2017). At the same time, it should be emphasis ed that the higher values of water mineralis ation (above 750 mg · dm⁻³) were documented during pumping tests within the plateau, with the dominating value of about 500 mg · dm⁻³ (Murzynowski, Małecki 1982). The physicochemical properties and ionic composition of the groundwater studied today, covering the Quaternary aquifer, are similar to the results of archival research. A general mineralisation of about 500 mg · dm⁻³ is seen, with a clear dominance of HCO₃⁻ and Ca²⁺ ions (Table 6). However, significant concentrations of sulphate ions (100 mg · dm⁻³) are noteworthy, exceeding the background range determined by Fabianowski (2002) for this aquifer and allowing the tested water to be classified as HCO₃–SO₄–Ca.

The Neogene aquifer is exploited only locally due to variable hydrogeological parameters and significant iron and brown coal dust contents (Fabianowski 2002). For this reason, its hydrochemical recognition is insufficient for full characterisation. Based on the research carried out by the authors, it can be concluded that in terms of EC (676 μS · cm⁻¹) and mineralisation (525 mg · dm⁻³), the tested water is similar to the Quaternary water. Its composition is dominated by hydrocarbonate and calcium ions with a significant share of sodium ions. This is water of the HCO₃–Ca–Na type (Table 6).

The Upper Cretaceous aquifer is the most intensively exploited in the area covered by the study. The water of this aquifer is characterised by low electrolytic conductivity ranging from 100 μS · cm⁻¹ to 650 μS · cm⁻¹ and mineralisation of 200–500 mg · dm⁻³ (Murzynowski, Małecki 1982). These are simple water types of the HCO₃–Ca. The collected water sample falls within the background ranges determined for the Upper Cretaceous aquifer: EC = 376 μS · cm⁻¹, TDS = 309.8 mg · dm⁻³ and water type of HCO₃–Ca (Table 6).

It is worth noting that the study by Małecki et al. (2017) and Ziułkiewicz (2003), conducted on the scale of the entire Łódź agglomeration, documented the stability of the chemical composition of the waters of this aquifer during its long-term exploitation. However, on a local scale, a potential inflow of polluted water into the Upper Cretaceous aquifer cannot be omitted, especially in places where hydraulic windows occur between the Quaternary and Cretaceous aquifers.

The impact of anthropopressure on the water chemistry is noticeable not only in their chemical composition, but also in the distribution of speciation of water macroelements. In the case of alkaline elements (Na and K) and alkaline earth metals (e.g. Ca and Mg), the dominance of simple ionic speciation is observed. It is characteristic for water occurring in natural conditions. On the other hand, in surface water, water taken from alluvial water and water of the Quaternary aquifer, the significant share of sulphate and polyelemental speciation is clearly visible, which proves the anthropogenic transformation of their chemistry. In the water from the Upper Cretaceous aquifer, the assessment of sodium and potassium speciation did not reveal the presence of sulphate, and in the case of calcium and magnesium, these speciation are secondary to hydrocarbonate speciation (CaHCO₃⁺ and MgHCO₃⁺) (Table 7). The observed trends in the speciation distribution indicate the high degree of their anthropogenic transformations and are consistent with the results of research conducted by Jóźwiak and Krogulec (2006) in the area of the Warsaw agglomeration.

The characteristics of the content of microelements in water and sediment are also a sensitive indicator of anthropogenic impacts. Studies of the alluvial deposits of the Ner documented the presence of chromium, zinc, copper, mercury and cadmium and, in individual cases, lead and silver in concentrations considered harmful to living organisms. The next step in such an assessment is to determine whether the above-mentioned elements accumulated in river sediments are activated and are also present in water in concentrations that pose a threat to living organisms.

In the tested water samples, the concentration of silver did not exceed the detection limit of 0.005 mg · dm⁻³ and the limit value for drinking water (0.010 mg · dm⁻³). The same is true for zinc and copper. The maximum contents of Zn (0.09 mg · dm⁻³) and Cu (0.005 mg · dm⁻³) do not exceed the contents that may have a negative impact on living organisms. In the case of copper, they also do not exceed the limit values for drinking water (Table 6).

Concentrations of chromium are considered in surface water as harmful to biological activity with content exceeding 0.1 mg · dm⁻³ (Kabata-Pendias, Pendias 1999), and in drinking water, the content of this element should not exceed 0.05 mg · dm⁻³ (Journal of Laws of 2017, item 2294). In the tested water samples, the chromium content was lower than the limit values given above. The highest concentrations, reaching 0.010 mg · dm⁻³, were found in the water taken from the alluvium of the Ner River. The results of speciation modelling indicate that this element occurs mainly in the sixth oxidation state in the form of the easily soluble CrO₄^2- form (Tables 6 and 7). This is worrying because forms of hexavalent chromium are mobile in the soil and water environment and are subject to intensive bioaccumulation (Kabata-Pendias, Pendias 1999).

The contents of cadmium, mercury, arsenic and lead in water from the alluvium of Ner, in individual samples, exceed the limit values for drinking water, which are: 0.005 mg Cd · dm⁻³, 0.01 mg Pb · dm⁻³, 0.01 As · dm⁻³ and 0.001 mg Hg · dm⁻³ (Journal of Laws of 2017, item 2294). In addition, mercury concentrations exceed the values allowed for drinking water also in the sample taken from the Quaternary aquifer (0.00132 mg · dm⁻³). It should be emphasised that all forms of mercury are toxic. Mercury easily passes from adsorbed forms to water (Kabata-Pendias, Pendias 1999). Extreme mercury content in the water taken from the Ner alluvium significantly exceeds the Hg content found both in the analysed groundwater and surface water, which may indicate the release of mercury accumulated in the alluvium into the solution. The same trend was observed for cadmium, lead and bismuth, and for zinc, the course of the cumulative curve for surface water and alluvial water is similar, lower values occur in groundwater (Figs 12 and 13).

Fig. 13.

The analysis of microelements concentrations in the tested waters in the case of their mixing indicates a negative impact of water from alluvial sediments on the chemical status of waters of the Quaternary and Upper Cretaceous horizons.