Abstract

In the Busko-Zdrój and Solec-Zdrój region (Poland), curative waters with different concentrations of sulphur(II) compounds are extracted. In this paper, waters from 14 intakes were considered. The waters analysed are chloride-sodium, sulphide, iodide (Cl–Na, S, I) or chloride-sodium, sulphide (Cl–Na, S) and are especially associated with deep circulation systems (Jurassic limestones) or shallow circulation systems mostly connected with Cretaceous formations (Senonian marls and limestones and Cenomanian sands and sandstones). The aim of this research is to evaluate the similarities or differences between samples of curative water taken from different geological formations and locations. Principal component analysis was used to assess the similarities between samples of curative waters collected from the intakes being examined. Two principal components were extracted. The analysis indicated that there is a difference between the chemical composition of waters taken from different stratigraphies. Water samples from Cretaceous formations of Cenomanian and Senonian age are similar. There is a small difference observed for waters abstracted from the Neogene and Cretaceous. However, these differences mostly resulted from SO42− and Ca2+ concentrations.

INTRODUCTION

In European Union law, there are no regulations referring directly to curative water. Curative waters are handled as medicinal products, that is why they are subjected to the requirements included in Directive 2001/83/EC of the European Parliament and of the Council of 6 November 2001 on the Community code relating to medicinal products for human use (Directive 2001; Wątor et al. 2016).

In Poland there are two legal documents regarding curative waters – the Geological and Mining Law of 9 June 2011 (Journal of Laws [Dz. U.] No. 163/2011 item 981 — GML 2011) and the Regulation of the Minister of Health of 13 April 2006 on the scope of the studies required to determine the medicinal properties of natural medicinal resources and the medicinal properties of climate, the criteria for their evaluation and a specimen certificate confirming these properties (Journal of Laws [Dz. U.] No. 80/2006 item 565 — RMH 2006). These documents provide a definition of curative water: groundwaters uncontaminated by chemical and microbiological agents, which should exhibit a natural variability of physicochemical parameters and meet at least one of the conditions set forth in the Law concerning minimum concentrations of the specific components that determine the medicinal properties of these waters (such as sulphur(II) compounds, iodide, fluoride, iron(II) ions and others).

There are many various statistical methods for environmental data analysis. The most popular are geostatistical techniques or geochemical modeling as well as techniques of data analysis that can be effectively used to classify and analyse the similarities between features such as intakes of curative water. The results of this analysis can help to identify if water from all of the locations considered have the same origin or come from the same aquifer despite the different lithology or stratigraphy. However, it is difficult to compare interesting research items (water intakes) with each other, especially when we have a large number of variables describing them. Usually, some of the variables considered are strongly correlated with each other (redundancy of information). For these purposes, the most commonly used methods are factor analysis (especially principal component analysis – PCA) and cluster analysis (CA). Such techniques can also be applied during the analysis of the spatial variability of the chemical composition of water. At the same time, they allow one to separate items (intakes – locations of sample collections) which are similar to each other in terms of the features analysed (including concentrations of physicochemical indicators of waters).

The aim of this research is to evaluate the similarities or differences between samples of curative water taken from different geological formations and locations based on the example of selected intakes in the Świętokrzyskie Province. There is still a lack of knowledge about the origin of the curative water exploited in the region examined. The complicated geological block-fold structure suggests that waters from different intakes are separated and can originate from different geological processes but there is no clear evidence for this consideration. The results of PCA analysis can help to answer the question of whether waters from deeper or shallower levels are mixing as well as enabling one to state whether they constitute several separate or one single hydraulic level.

MATERIALS AND METHODS

Principal component analysis was used to assess the similarities between samples of curative waters collected from the 14 intakes located in six cities: Busko-Zdrój (B-4b, B-8b, B-13, B-16a, B-17 and C-1 intakes), Dobrowoda (G-1 intake), Las Winiarski (LW-1 and LW-2 intakes), Piestrzec (Dar Natury intake), Solec-Zdrój (Solec 2B, Solec 2 and Szyb Solecki intakes) and Wełnin (Wełnin intake). All these towns are located in the Świętokrzyskie Province (Figure 1).

Figure 1

Location of study area.

Figure 1

Location of study area.

The geological structure of the study area is characterised by plate-deposited layers (NW-SE direction) with a block-fold structure. Due to the varied geological structure, the area analysed is characterised by several water-bearing horizons (Zuber et al. 1997; Krawczyk et al. 1999; Krysiak 2000; Barbacki 2004; Lisik & Szczepański 2014). All the waters analysed are chloride-sodium, sulphide, iodide (Cl–Na, S, I) or chloride-sodium, sulphide waters (Cl–Na, S) (according to the Szczukariew-Prikłoński classification) and are associated with a deep circulation system (from Jurassic formations) or a shallow circulation system mostly connected with Cretaceous formations (Senonian and Cenomanian) (Table 1) (Lipiec 2009; Wątor 2013; Lipiec & Wiktorowicz 2015; Wątor et al. 2016).

Table 1

Basic information about the water intakes analysed (Krawczyk et al. 1999; Lisik & Szczepański 2014; Chowaniec et al. 2015) (Cr — Cretaceous, Crc — Cretaceous Cenomanian, Crst — Cretaceous Senonian, J3 — Jurassic, Ng — Neogene)

Intake nameGeological setting of the aquifer
Depth [m]Water type
LithologyStratigraphy
B − 4b Marls Crs 60.0 Cl–Na, S, I 
B − 8b Marls Crs 60.0 Cl–Na, S, I 
B − 13 Marls Crst 55.0 Cl–Na, S, I 
B − 16a Sandstones Crc 129.0 Cl–Na, S, I 
B − 17 Sandstones Crc 140.0 Cl–Na, S, I 
C − 1 Sands and sandstones Crc 663.0 Cl–Na, S, I 
LW − 1 Sands and sandstones Crc 163.0 Cl–Na, S 
LW − 2 Sands and sandstones Crc 165.0 Cl–Na, S 
Szyb Solecki Sandstones and limestones Cr 170.0 Cl–Na–SO4, S, I 
Solec 2 Marls with inserts of limestones and sandstones Cr 121.6 Cl–Na, S 
Solec 2B Marls with inserts of limestones and sandstones Cr 121.3 Cl–Na–SO4, S, I 
G − 1 Marls, limestones, conglomerates and sandstones Ng + Cr 300.0 Cl–SO4–Na, S, I 
Dar Natury Limestones Ng 90.0 SO4–HCO3–Ca, S 
Wełnin Limestones J3 170.0 Cl–Na, S, I 
Intake nameGeological setting of the aquifer
Depth [m]Water type
LithologyStratigraphy
B − 4b Marls Crs 60.0 Cl–Na, S, I 
B − 8b Marls Crs 60.0 Cl–Na, S, I 
B − 13 Marls Crst 55.0 Cl–Na, S, I 
B − 16a Sandstones Crc 129.0 Cl–Na, S, I 
B − 17 Sandstones Crc 140.0 Cl–Na, S, I 
C − 1 Sands and sandstones Crc 663.0 Cl–Na, S, I 
LW − 1 Sands and sandstones Crc 163.0 Cl–Na, S 
LW − 2 Sands and sandstones Crc 165.0 Cl–Na, S 
Szyb Solecki Sandstones and limestones Cr 170.0 Cl–Na–SO4, S, I 
Solec 2 Marls with inserts of limestones and sandstones Cr 121.6 Cl–Na, S 
Solec 2B Marls with inserts of limestones and sandstones Cr 121.3 Cl–Na–SO4, S, I 
G − 1 Marls, limestones, conglomerates and sandstones Ng + Cr 300.0 Cl–SO4–Na, S, I 
Dar Natury Limestones Ng 90.0 SO4–HCO3–Ca, S 
Wełnin Limestones J3 170.0 Cl–Na, S, I 

The aquifer lithology varies depending on the stratigraphy. Jurassic (J3) is mostly represented by marls and limestones. The most common Cretaceous (Cr) formations are sands and sandstones of the Cenomanian (Crc) and marls and limestones of the Senonian (Crs).

The results of analyses carried out in 2010–2016 were used for further calculations. Samples from the intakes examined were collected by qualified samplers (including the Authors) in accordance with the guidelines described in the ISO 5667-11 standard, with different frequencies and in a different time period dependent on water exploitation and intake usage and to monitor seasonal variations and trends. Unstable parameters such as pH, Eh, electrical conductivity and temperature were measured in the field using on-line systems. Analysis of the chemical composition was performed in a certified laboratory using standardised and appropriately verified analytical methods fit for the purpose of curative water analysis. Concentrations of major ions – Ca2+, Mg2+, Na+, K+ and SO42− — were determined using inductively coupled plasma optical emission spectrometry (ICP-OES, Optima 7300DV spectrometer, Perkin Elmer, USA). For hydrocarbonate and chloride analysis, titration methods were used. The concentrations of sulphur (II) compounds were determined using the thiomercurymetric and iodometric titration methods. The number of results used in the statistical analysis varied depending on the intake (Table 2). In total, the results of 189 physicochemical analyses were used.

Table 2

Mean values for the results of the analyses performed in 2010–2016

Intake nameNumber of analysispHTDS [mg/L]Na+ [mg/L]K+ [mg/L]Mg2+ [mg/L]Ca2+ [mg/L]Cl [mg/L]SO42 [mg/L]HCO3 [mg/L]S(II) [mg/L]
B − 4b 7.03 13,820 4,050 90.48 244.2 398.8 6,695 1,841 398.4 28.56 
B − 8b 7.07 13,100 4,078 102.7 223.0 356.8 6,153 1,798 428.4 38.40 
B − 13 7.14 12,467 3,849 93.48 214.5 305.1 5,776 1,702 440.2 43.02 
B − 16a 7.19 13,433 4,107 102.8 236.1 387.1 6,292 1,901 433.9 37.41 
B − 17 7.05 13,704 4,238 93.31 242.0 398.7 6,384 1,934 449.0 47.32 
C − 1 41 7.20 11,609 3,887 87.07 151.9 230.0 5,249 1,570 422.7 36.95 
LW − 1 36 7.04 12,349 4,004 109.8 207.9 334.9 5,549 1,678 447.8 50.29 
LW − 2 17 7.06 12,614 3,997 100.7 199.3 332.6 5,764 1,783 429.8 50.62 
Szyb Solecki 6.94 18,398 5,166 115.3 381.1 873.9 8,161 3,160 424.2 109.5 
Solec 2 7.26 21,006 6,156 117.7 523.3 750.5 9,982 3,132 210.2 33.18 
Solec 2B 7.16 16,748 4,881 82.97 377.7 503.3 6,820 3,383 576.3 196.7 
G − 1 15 6.96 14,899 4,350 107.1 336.3 524.3 6,784 2,252 426.7 53.01 
Dar Natury 7.27 2,321 34.5 2.42 40.15 561.2 21.5 1,300 339.3 4.85 
Wełnin 26 6.68 33,955 9,928 215.1 957.6 1 132 17,552 2,444 962.5 806.3 
Intake nameNumber of analysispHTDS [mg/L]Na+ [mg/L]K+ [mg/L]Mg2+ [mg/L]Ca2+ [mg/L]Cl [mg/L]SO42 [mg/L]HCO3 [mg/L]S(II) [mg/L]
B − 4b 7.03 13,820 4,050 90.48 244.2 398.8 6,695 1,841 398.4 28.56 
B − 8b 7.07 13,100 4,078 102.7 223.0 356.8 6,153 1,798 428.4 38.40 
B − 13 7.14 12,467 3,849 93.48 214.5 305.1 5,776 1,702 440.2 43.02 
B − 16a 7.19 13,433 4,107 102.8 236.1 387.1 6,292 1,901 433.9 37.41 
B − 17 7.05 13,704 4,238 93.31 242.0 398.7 6,384 1,934 449.0 47.32 
C − 1 41 7.20 11,609 3,887 87.07 151.9 230.0 5,249 1,570 422.7 36.95 
LW − 1 36 7.04 12,349 4,004 109.8 207.9 334.9 5,549 1,678 447.8 50.29 
LW − 2 17 7.06 12,614 3,997 100.7 199.3 332.6 5,764 1,783 429.8 50.62 
Szyb Solecki 6.94 18,398 5,166 115.3 381.1 873.9 8,161 3,160 424.2 109.5 
Solec 2 7.26 21,006 6,156 117.7 523.3 750.5 9,982 3,132 210.2 33.18 
Solec 2B 7.16 16,748 4,881 82.97 377.7 503.3 6,820 3,383 576.3 196.7 
G − 1 15 6.96 14,899 4,350 107.1 336.3 524.3 6,784 2,252 426.7 53.01 
Dar Natury 7.27 2,321 34.5 2.42 40.15 561.2 21.5 1,300 339.3 4.85 
Wełnin 26 6.68 33,955 9,928 215.1 957.6 1 132 17,552 2,444 962.5 806.3 

Principal component analysis was carried out using PS IMAGO© software provided by Predictive Solutions. IBM® SPSS Statistics® is the analytical engine of PS IMAGO©. The multivariate data consist of 10 variables and 189 observations – the results of the physicochemical analyses of curative water from the Busko-Zdrój and Solec-Zdrój region. The methodology of PCA analysis on the examples of different items (including water samples) is widely discussed by Miller & Miller (2015), Otto (2016), Härdle & Hlávka (2007) or McKillup & Dyar (2010).

All data were normalised and the Kaiser–Mayer–Olkin and Bartlett's tests were done to appraise correlations between data and the adequacy of the input data for the PCA concept. The Kaiser–Mayer–Olkin measure of sampling accuracy is equal to 0.9 and indicates a good selection of data.

RESULTS AND DISCUSSION

The concentrations of the major ion (Na+, K+, Mg2+, Ca2+, Cl, SO42−, HCO3), S(II) compounds as a specific component of these waters as well as pH and total dissolved solids (TDS) values were analysed in detail. Table 2 presents the results obtained.

The pH values of the waters that were measured ranged from 6.68 (Wełnin) to 7.27 (Dar Natury). The lowest amount of TDS was observed in water from Dar Natury (2,321 mg/L) intake and the highest results were measured in Wełnin (33,955 mg/L), and three intakes located in Solec-Zdrój: Solec 2 (16,748 mg/L), Szyb Solecki (18,398 mg/L) and Solec 2B (16,748 mg/L). A similar situation was observed for major ion concentrations – the lowest values were measured in Dar Natury intake and the highest in Wełnin. The intake results for S(II) ranged from 4.95 mg/L in Dar Natury to 806.3 mg/L in Wełnin.

The main goal was to choose the principal component numbers. The major criterion for determining the coefficient of linear combinations for the main components is to determine a new variable with the maximum variance. The main components explain the maximum possible variability contained in the original data. Usually, the first component explains the largest variability. The selection of the number of significant components is based on three criteria: total variance explained (in environmental sciences 60% of total explained variance is required), eigenvalue (usually components with eigenvalues above 1 are selected) and screen plot (the graph shows a clear gap between the steep inclination of important factors and the gradually decreasing slope of others). In the example presented, we have established two components. They explain almost 90% of total variance. The first component is mostly dependent on pH, TDS, Na+, K+, Ca2+, Mg2+, Cl, HCO3 and S(II) values whereas the second component depends strongly on SO42− and also Ca2+ concentration and can be joined with gypsum and anhydrite deposits, which occur locally in the Busko-Zdrój and Solec-Zdrój region forming discontinuous layers.

The chart of correlation between the extracted principal components lets us assess the similarity of the intakes analysed. As Felipe-Sotello et al. (2015) showed, the PCA method is a very good statistical tool for the identification of similarities and differences between water samples. Items which are nearest to the centre of the graph are the most similar to each other (Figure 2).

Figure 2

The distribution of the curative water samples analysed depending on the principal components values.

Figure 2

The distribution of the curative water samples analysed depending on the principal components values.

The analysis of the distribution of the curative water analysed showed that the chemical composition of the water in each intake is stable in time (the points for every single intake are close to each other). As another researcher indicated, the PCA method can be used for the identification of temporal (also seasonal) and spatial variability of the chemical composition of different waters (Villegas et al. 2013; Palma et al. 2014; Kim et al. 2015; Cortes et al. 2016; Ayed et al. 2017; Islam et al. 2017; Sahu et al. 2018). Water from the Wełnin intake shows the greatest difference from the others – the results of the analyses of this water created a separate cluster. In the second group, which was created by the rest of the study waters (from Busko-Zdroj, Solec-Zdroj, Las Winiarski, Dobrowoda and Piestrzec), the intakes located in Solec-Zdrój differ slightly (Figure 2). The second component, related to gypsum and anhydrite deposits, caused this variability. To check if these differences are connected with the aquifer stratigraphy (water-bearing bed), the second variant of the graphical dependencies between the values of component 1 and component 2 was drawn (Figure 3).

Figure 3

The distribution of intakes from different water-bearing beds depending on the principal components values.

Figure 3

The distribution of intakes from different water-bearing beds depending on the principal components values.

The analysis indicated that there is a difference between the chemical composition of water taken from different water-bearing beds (Figure 3). Water from Jurassic limestones is characterised by the highest value of component 1. Water samples from Cretaceous formations of the Cenomanian and Senonian are similar and do not change much from water exploited from the Neogene. There is a small difference observed for water intakes from the Neogene and Cretaceous. However, these differences mostly result from the component 2 value and, as a consequence, Ca2+ and SO42− concentrations.

CONCLUSIONS

According to Polish regulations, curative waters are groundwaters uncontaminated by chemical and microbiological agents, which should exhibit a natural variability of physicochemical parameters and meet at least one of the conditions set forth in the Law concerning minimum concentrations of the specific components that determine the medicinal properties of these waters (such as sulphur(II) compounds, iodide, fluoride, iron(II) ions and others). In the region of Busko-Zdrój and Solec-Zdrój examined, curative waters with a high concentration of sulphur(II) compounds are presented. These waters are extracted from different aquifer formations associated with the lithology of the Jurassic, Cretaceous or Neogene. There is still a lack of information about the origin of curative water exploited in the study region. The complicated geological block-fold structure suggests that water from the different intakes tested are separated and can originate from different geological processes. Principal component analysis was used to divide the intakes analysed into groups. The analysis indicated a wide difference between water from a Jurassic intake and the others. Between intakes exploited from the Cretaceous and mixed aquifer lithology, a small difference can be seen. This mostly results from the Ca2+ and SO42− concentration and can be related to discontinuous gypsum and anhydrite formations occurring in the area examined. These suggest that there is the possibility of mixing water from deeper or shallower levels of the Neogene and Cretaceous (Cenomanian and Senonian) formations. On the basis of the results obtained, it can be stated that the Cretaceous intakes considered constitute one curative water reservoir and probably are in hydraulic contact with each other. Taking into account the fact that the resources of the study waters are non-renewable, the growing exploitation of water from one of the existing intakes can cause a lowering of the water table in the other one. Sustainable management of the sulphide waters in the region discussed is therefore extremely important.

ACKNOWLEDGEMENT

The work was partially supported by AGH 11.11.140.797.

REFERENCES

REFERENCES
Ayed
B.
,
Jmal
I.
,
Sahal
S.
,
Mokadem
N.
,
Saidi
S.
,
Boughariou
E.
&
Bouri
S.
2017
Hydrochemical characterization of groundwater using multivariate statistical analysis: the Maritime Djeffara shallow aquifer (Southeastern Tunisia)
.
Environmental Earth Sciences
76
(
24
),
821
.
Barbacki
A. P.
2004
Zbiorniki wód geotermalnych niecki miechowskiej i środkowej części zapadliska przedkarpackiego. Wydawn. Inst. Gospodarki Surowcami Mineralnymi i Energią PAN (In Polish).
Chowaniec
J.
,
Gągulski
T.
&
Gorczyca
G.
2015
Chemizm niezagospodarowanych wód mineralnych rejonu Buska-Zdroju i Solca-Zdroju
.
Przegląd Geologiczny
63
(
10/1
),
645
651
(In Polish)
.
Cortes
J. E.
,
Muñoz
L. F.
,
Gonzalez
C. A.
,
Niño
J. E.
,
Polo
A.
,
Suspes
A.
,
Siachoque
S. C.
,
Hernández
A.
&
Trujillo
H.
2016
Hydrogeochemistry of the formation waters in the San Francisco field, UMV basin, Colombia–A multivariate statistical approach
.
Journal of Hydrology
539
,
113
124
.
Directive 2001/83/EC of the European Parliament and of the Council of 6 November 2001 on the Community code relating to medicinal products for human use (OJ L 311, 28.11.2001).
Felipe-Sotelo
M.
,
Henshall-Bell
E. R.
,
Evans
N. D. M.
&
Read
D.
2015
Comparison of the chemical composition of British and Continental European bottled waters by multivariate analysis
.
Journal of Food Composition and Analysis
39
,
33
42
.
GML 2011 Geological and Mining Law of 9 June 2011 (Journal of Laws [Dz. U.] No. 163/2011 item 981).
Härdle
W.
&
Hlávka
Z.
2007
Multivariate Statistics (p. 79)
.
Springer
,
New York
,
NY, USA
.
ISO 5667-11:2009 Water quality -- Sampling -- Part 11: Guidance on sampling of groundwaters. ISO, Geneva, Switzerland.
Krawczyk
J.
,
Mateńko
T.
,
Mądry
J.
&
Porwisz
B.
1999
Wody lecznicze Buska Zdroju w świetle dotychczasowych badań
.
W: Współczesne Problemy Hydrogeologii. T.
9
,
159
164
.
Państw. Inst. Geol., Warszawa (In Polish)
.
Krysiak
Z.
2000
Tectonic evolution of the Carpathian Foredeep and its influence on Miocene sedimentation
.
Geol. Quart
44
(
2
),
137
156
.
Lipiec
I.
2009
Wstępna analiza stabilności składników chemicznych ocena trendów zmian jakości wód leczniczych Solca-Zdroju
.
Biuletyn Państwowego Instytutu Geologicznego
(436, z. 9/2)
,
323
328
(In Polish).
Lipiec
I.
&
Wiktorowicz
B.
2015
Charakterystyka hydrogeochemiczna wód siarczkowych rejonu Solca-Zdroju
.
Przegląd Geologiczny
63
(
10/2
),
908
911
(In Polish).
Lisik
R.
&
Szczepański
A.
2014
Siarczkowe Wody Lecznicze W Części Zapadliska Przedkarpackiego
.
Fundacja Posteris
(In Polish).
McKillup
S.
&
Dyar
M. D.
2010
Geostatistics Explained: an Introductory Guide for Earth Scientists
.
Cambridge University Press
,
Cambridge, UK
.
Miller
J. N.
&
Miller
J. C.
2015
Statistics and Chemometrics for Analytical Chemistry
.
Pearson Education
,
Harlow, UK
.
Otto
M.
2016
Chemometrics: Statistics and Computer Application in Analytical Chemistry
.
John Wiley & Sons Hoboken
,
NJ, USA
.
Palma
P.
,
Ledo
L.
,
Soares
S.
,
Barbosa
I. R.
&
Alvarenga
P.
2014
Spatial and temporal variability of the water and sediments quality in the Alqueva reservoir (Guadiana Basin; southern Portugal)
.
Science of the Total Environment
470
,
780
790
.
RMH
2006
Regulation of the Minister of Health of 13 April 2006 on the scope of the studies required to determine the medicinal properties of natural medicinal resources and medicinal properties of climate, the criteria for their evaluation and a specimen certificate confirming these properties (Journal of Laws [Dz. U.] No. 80/2006 item 565 – RMH 2006) (In Polish)
.
Villegas
P.
,
Paredes
V.
,
Betancur
T.
&
Ribeiro
L.
2013
Assessing the hydrochemistry of the Urabá Aquifer, Colombia by principal component analysis
.
Journal of Geochemical Exploration
134
,
120
129
.
Wątor
K.
2013
Przestrzenno-czasowa analiza zmienności składu chemicznego wód leczniczych rejonu Buska-Zdroju. Rozprawa doktorska. Akademia Górniczo-Hutnicza im. Stanisława Staszica w Krakowie. Wydział Geologii, Geofizyki i Ochrony Środowiska (In Polish, unpublished)
.