Hydrochemistry, Vulnerability and Protection of the Torres Vedras Aquifer. Abstract

Transcription

1 Hydrochemistry, Vulnerability and Protection of the Torres Vedras Aquifer Ana Carina Ferreira Veríssimo Dissertation for Masters Degree in Geological and Mining Engineering Instituto Superior Técnico, Universidade Técnica de Lisboa Abstract The present paper is intended to be an hydrogeological and hydrochemical contribution to the understanding and characterization of the Torres Vedras aquifer. First, the case study is characterized, including a geological, climatic, hydrogeological and hydrochemical characterization. The predominant facies in the aquifer system are sodium chloride and calcium bicarbonate. The temporal analysis of quality monitoring stations analysis identified the occurrence of contamination in groundwater extraction wells, particularly manganese, oil, barium and mercury. For the last period analyzed, the groundwater in this study didn t have enough quality to produce water for human consumption in accordance with Decree-Law No. 236/98 of 1 August. The determination of vulnerability to pollution of the aquifer by the method IS (Ribeiro, 2005), indicates a susceptibility to pollution medium-high (dominant classes 25 to 45). The occurrence of higher classes is associated with agricultural areas, pig farms and the landfill. Given the absence of perimeters of protection for public water supply wells in the area, according to Decree-Law No. 382/99 of September 22, it s proposed defining them in the current study for the intermediate and expanded zone for the municipal wells AC22 AC23 by numerical modeling, comparing the results with the fixed radius method. The last one resulted to be insufficient. The stochastic modeling, which considers the field heterogeneity, allows the generation of 10 equiprobable scenarios of occurrence of contamination in the extraction well after 3500 days, which allows the elaboration of a map of risk of contamination, which is intended as a tool for decision support and future planning. Key-words: aquifer, Torres Vedras, hydrochemistry, index of susceptibility to pollution, protection perimeters, stochastic modeling. 1 Introduction The underground water is a valuable natural resource and as such should be protected from deterioration and chemical pollution. This is particularly important when using the groundwater for human consumption. The Water Framework Directive (2000/60/EC) contains general provisions regarding the protection and conservation of groundwater. As provided in Article 17 of this Directive, measures should be adopted to prevent and control groundwater pollution, including criteria for assessing good chemical status and for the identification of significant and sustained increase in pollutants concentrations. To achieve the goals assigned by the legislation, there is a need to provide ourselves with appropriate methodologies and techniques for assessing the vulnerability of aquifers and the risk of groundwater pollution, as well as implementation of protective perimeters for wells bound for human consumption. Portuguese legislation include the delineation of protection perimeters for wells for public supply, in accordance with Decree-Law No. 382/99 of 22 September. In short, to protect groundwater there is a need to implement measures for their protection and to provide decision-makers in management tools and appropriate planning, which require detailed knowledge of the hydrogeological and chemical conditions of the areas in question. The recent development of Torres Vedras region, mainly due to the geographical proximity to Lisbon and with good road access, makes it important a further study in order to increase geological, hydrogeological and hydrochemical knowledge The current dependence on the company Águas do Oeste (ADO) as the main public water supply origin, which in case of failure will be done through the reserved municipal wells, which mostly exploit the aquifer of Torres Vedras, makes it of great importance to characterize groundwater quality and development of tools for decision and support in planning and management of these groundwater, 1

2 such as maps of vulnerability, using the method of Susceptibility Index (Ribeiro, 2005) and definition of protection perimeters using numerical modeling and stochastic simulation, comparing the results with the suggested method for the delineation of protection perimeters in Decree-Law No. 382/99 of September 22nd (calculated fixed radius method). 2 Caracterization of the Case Study The aquifer of Torres Vedras occupies an area of 80 km 2 and is located in the district of Lisbon, covering the counties of Torres Vedras, Cadaval and Alenquer (Figure 2.1). Figure Geographical location of the aquifer system of Torres Vedras (Adapted from SNIRH) The precipitation in the study area is around 600 to 1000 mm/year, evapotranspiration values from 500 to 700 mm/year. The recharge, which should be around 10% to 15% of precipitation (Almeida et al., 2000), has values of approximately 75 to 110 mm/year. The aquifer is part of the Ribeiras do Oeste Hydrographic Basin, and the main water lines that cross it are Rio Sizandro and Rio Alcabrichel. The topographical elevations in the study area are between 19 to 213 m. 2.1 Geology The most important water-bearing formation in the area is the formation of Torres Vedras Sandstone (Sandstone with plant fossils, Torres Vedras and Cercal), dated as belonging to the Early Cretaceous. The Early Cretaceous of Torres Vedras is part of the West Mesozoic Basin of Portugal. The "Gres de Torres Vedras" consists of feldspathic sandstones and kaolins with variable size, generally poorly calibrated, with abundant intercalations of clay lenticules, silts and some conglomeratic levels. Figure 2.2 Geology of the Torres Vedras aquifer (adapted from Almeida, et al, 2000) The geological map of the study area can be seen in Figure 2.2. According to Neves Palma e Vieira da Silva (1982), the sandstone and sandstone-clay facies, culmination of the Upper Jurassic, are also part of the regional aquifer system. The presence of numerous lenticule clays gives a confined nature to the aquifer system, and a multilayer, porous type. Tectonically, the aquifer system belongs to the anticline of Torres Vedras. 2.2 Hydraulic Parameters and Productivity In SNIRH are available the values of piezometric levels for the period of in four monitoring stations of the Network of quantity (Figure 2.3), interpolated for the entire study area by ordinary kriging. 2

3 Figure 2.3 Two-dimensional representation for hydraulic heads, interpolated by ordinary kriging, data from the Network of Quantity (SNIRH). Blue arrows indicate the groundwater flow direction. As shown by Figure 2.3, in the aquifer system of Torres Vedras, the water point with lower value of piezometric surface it is the 374/130. The flow direction (blue arrows) also converges to this site (NE- SW and NW-SE). The transmissivity values are between 2.5 and 400 m 2 /day (Almeida et al, 2000), although Neves Palma and Vieira Silva assign values in the range of 10 to 200 m 2 /day. The great variability of this parameter is related to the heterogeneity of the system, which is multilayered, with abundant clay lenticules. The known flow rates (Q - L/s), produced by the wells that capture the system don t have a high productivity, with flow rates between 1,1 e 20 L/s, with a median of 3,9 L/s. 3 Hydrochemistry The values for determining the hydrochemical facies and water quality in the aquifer of Torres Vedras were obtained from SNIRH within the monitoring network quality. The analysis period is The analysis frequency is mostly within six months (in March/April and September/October each year), however, not all elements are analyzed at the same periods, and there are stations with very little data. The analyzed values of sodium bicarbonate and potassium are very low in number. The descriptive statistics of the analyzed parameters is shown in Table 3.1. Table Values of basic statistical sampling carried out in the main Quality network of the Torres Vedras aquifer between March 2001 and September 2009, values in mg/l, unless indicated. Parameters n Mean Minimum Q1 Median Q3 Maximum Variance Conductivity ,12 200,00 368,25 554,50 758,00 962, ,68 (µs/cm) ph 94 6,47 3,80 5,60 6,75 7,30 8,50 1,24 Total Hardness ,68 18,86 76,07 119,77 254,32 356, ,50 (mg/l CaCO 3) K ,95 1,10 3,40 5,50 9,00 13,50 13,22 Ca ,47 6,30 12,43 24,00 76,40 110, ,66 Mg ,66 0,40 9,13 13,00 18,30 36,10 56,17 HCO 3 - SO ,43 12,00 26,00 121,00 245,75 293, , ,71 5,30 21,00 34,00 79,00 154, ,72 Cl ,73 27,20 72,00 80,00 115,00 169, ,42 NO ,01 0,08 0,82 1,15 3,50 61,98 89,44 Na ,33 34,00 49,25 68,00 76,20 144, ,11 Total Fe 66 1,77 0,05 0,08 0,35 1,89 34,75 21,73 Through analysis of Table 3.1, we highlight the following key aspects: The mineralization of the water samples is medium to high, with a median of electrical conductivity of mg / l.; The ph is essentially neutral/acid, but with large variations, occurring very acidic (AC22 capture and 374/130) and other neutral / basic (maximum 8.5 and minimum of 3.8), and the interquartile range between 5.6 and 7.3; There is a wide range of water hardness (18.86 minimum and maximum of mg / l CaCO3); 3

4 The median chloride (80 mg / l), sulfate (34 mg / l), sodium (68 mg / l) and iron (0.35 mg / l) exceed the maximum allowable (VMA) for human drinking water, under Decree-Law No. 236/98 (Annex IV); The predominant hydrochemical facies in the aquifer system of Torres Vedras are calcium bicarbonate and sodium chloride, as shown in the Piper diagram in Figure 3.1. Figure 3.1 Piper diagram of the Torres Vedras aquifer groundwater Water Quality for Human Consumption Using the same data set to characterize the hydrochemical facies of the aquifer system, the study of water quality for human consumption was proceeded. To evaluate the water quality for human consumption, for all elements analyzed, were detected which ones exceeded the values recommended by the Decree-Law 236/98, with the alterations of the Decree-Law 306/2007. There are only refered the most problematic and the parameters considered of greater importance. The number of tests per parameter is not constant, and there were parameters analyzed more frequently than others. In general, the values of the analysis conducted in relation to the legal limits of existing Portuguese legislation can be seen in Table 3.2 and Table 3.3. Tabela Water quality in accordance with Annex I - Category A1 Decree-Law No. 236/98 Number Annex I Category A1 (a) of Parameters analysis <VMR >VMR >VMA VMR VMA ph 94 56% 44% 6,5-8,5 - Temperature (ºC) 89 98% 2% 0% Conductivity( µs/cm) % Nitrates (mg/l NO 3) 93 95% 5% 1% Fluoride (mg/l F) 10 40% 100% 0% 0,7-1,0 1,5 Total Iron (mg/l) 66 29% 71% 50% 0,1 0,3 Manganese (mg/l) 62 37% 63% 0,05 - Total Copper (mg/l) 67 96% 4% 1% 0,02 0,05 Total Zinc (mg/l) 56 88% 13% 0% 0,5 3,0 Arsenic (mg/l) % 0% 0% 0,01 0,05 Total Cadmium (mg/l) % 0% 0% 0,001 0,005 Total lead (mg/l) 50 0% - 0,05 Total Mercury (mg/l) 23 91% 9% 4% 0,0005 0,0010 Barium (mg/l) 8 38% - 0,1 Sulfates (mg/l) 85 99% 1% 0% Chlorides (mg(l) 83 0% 0% Orthophosphate as P 2O 5 (mg/l) 77 99% 1% 0,4 - Total Hydrocarbons (mg/l) 17 35% - 0,05 Dissolved oxygen (%) 75 75% 25% 70 - Ammonia nitrogen (NH 4) 91 49% 51% 0,05-4

5 Tabela Water quality in accordance with Annex VI - Category A1 Decree-Law No. 236/98 Number of Annex VI (b) Parameters analysis <VMR >VMR >VMA VMR VMA Hardness (as mg/l CaCO 3 ) 41 (calculado) 0,0% Calcium (mg/l) 42 92,9% 7,1% Magnesium (mg/l) 42 97,6% 2,4% 0,0% Sodium (mg/l) 12 0,0% 100,0% 0,0% Potassium (mg/l) 17 94,1% 5,9% 5,9% Total nitrite (em NO 2 ) (mg/l) 94 97,9% 3,2% - 0,1 Oxidability (mg/l) 65 58,5% 41,5% 7,7% 2 5 Values not defined (a) (b) - Establishes limits for the quality of surface waters with an end to production for human consumption for Class A1 - physical treatment and disinfection; Sets the limits for drinking water. As can be seen by examining Table 3.2 and Table 3.3, among the major elements only the sodium and iron exceeded the VMR, nevertheless, 6% of the tests exceeded the VMA values for potassium. This parameter, however, has been examined with a low frequency in the study period and the excessive values may be derived from agricultural contamination, so this parameter should be monitored more regularly. The high levels of sodium and iron may be linked to the nature of the waters in the study, circulating in clay lithologies, rich of these elements, and not due to contamination. The presence of sodium concentrations can be caused by the low ph of these waters, which are aggressive to limestone, causing occurring cationic exchange of Na + by Ca 2+ levels in the clay. Regarding the limits imposed by Annex I of Decree-Law, for Class A1, which requires treatment and disinfection, the examination of Table III.3 shows problematic parameters: ph, fluoride, total iron, manganese, barium, total hydrocarbons, dissolved oxygen and ammoniacal nitrogen. The well 362/132 has no major problems regarding the parameters analyzed, except the low value of fluoride. The wells that are the most problematic are AC22, 362/130, 374/130 and JK14 (at the latest analysis, in 2003). 4 Vulnerability to Pollution Once identified the problems that contribute to poor quality of the groundwater aquifer under study for production of drinking water, it is urgent the need for its protection, regarding contamination. Therefore, a pollution focus inventory pollution in the area was made, and proposed an analysis of vulnerability to pollution of the aquifer system, using the Susceptibility Index (SI) method (Ribeiro, 2005), in order to propose a tool for decision support and planning future. The SI Index is calculated by the weighted sum of the values assigned to the four parameters used in DRASTIC: D, R, A and T, which are multiplied by 10 for better visualization of results, and a new parameter LU. The weighting of each parameter was another change made on the DRASTIC. To evaluate the relative importance of each of the five parameters used in this method, a panel was established in DELPHI Portuguese experts in hydrogeology, getting the final result shown in Table 4.1. Table Parameters and weighting factors of SI (in Ribeiro, 2005) Parameter D R A T LU Weigthtening Factor 0,186 0,212 0,259 0,121 0,222 5

6 Under this method, and as can be seen in Table 4.1, the most determinant in assessing the susceptibility of an aquifer to contamination is A (material of the aquifer), and the least influential is the parameter T (topography). The parameters R (recharge) and LU (land use) will have similar influences and the parameter D (depth of water table) a slightly higher impact (Ribeiro, 2005). The susceptibility index, since its creation in 2000, has been applied to assess the degree of susceptibility to agricultural contamination in several studies. The depth of water table was estimated by interpolation (IDW), using the available data of the Network of piezometers in SNIRH, using the mean value for the year Assigning the parameters mentioned in the method for the class of depth, results in the map in Figure 4.1. Figure 4.1 Carta de Profundidade do Nível freático, com as captações da rede de piezometria e respectivas profundidades do nível freático (m) com os quais foi feita a interpolação. The lighter areas indicated in Figure 4.1, correspond to the deeper water table, thus having lower scores. As already mentioned, the more superficial is the aquifer, the greater its susceptibility to pollution. Recharge takes place mainly in plane areas, with permeable soil. The mapping of preferential recharge areas (Figure 4.2), took into consideration the areas where the slope was less than 6%, excluding those impermeable areas. Figure 4.2 Mapping of the preferential recharge areas Based on a map of spatial distribution of rainfall, which considered the factor of topography, the effective recharge was calculated using 10% of the value indicated on the map above. The recharge values vary between 70 and 100 mm / year. 6

7 The aquifer material was classified according to the outcropping lithologies (Figure 4.3). It is considered that in clays the residence time high, receiving scores of 60 and sandstone and soil get the score 80, with residence time is relatively high. Figure 4.3 Map of the aquifer material As seen on Figure 4.3, the main lithologies are clay, therefore, more impermeable. The slopes on the study area were calculated from the Digital Terrain Model (DTM) (Figure 4.4). Figure 4.4- Slope Map The score was given depending on the existing classes, and lower slopes are more susceptible to pollution since there is preferably greater runoff in these areas and accumulation of lixiviate waters. The data of land use were mostly obtained by remote sensing data through the CORINE Land Cover (2000). A subsequent survey of sources of pollution carried out by the Municipality of Torres Vedras (Farinha et al., 2007), allowed the inclusion of major sources of pollution, particularly pig farms and quarries. The land use map with the scoring from the SI can be seen in Figure

8 Figure 4.5 Map with SI scoring, depending on the land use The score for forests is zero, so we can see from Figure 4.5 that more than half of the aquifer is occupied by activities not considered at risk. By analyzing the same figure, it appears that the most critical activities are located SW of the aquifer. The calculation of IS on the aquifer study is performed using ArcGIS, turning the maps made in vector data (features) into raster data. Each raster obtained is multiplied by the corresponding weight, resulting in the mapping of susceptibility index as shown in Figure 4.6. Figure 4.6 Susceptibility Index Map and histogram. The calculated values of SI (Figure 4.6) lie in the range 20 to The analysis of this map shows that there are no classes of extreme values. Analyzing the histogram in Figure 4.6, it appears that the dominant classes are the susceptibility of medium-low (between 25 and 45). The occurrence of upper-middle classes (55 to 65.5) is by 6%. 5 Protection Perimeters Since the publication of Decree-Law No. 382/99 of 22 September, it became necessary to develop hydrogeological studies to calculate the perimeters of protected areas of public wells that draw more than 1,000 m 3 /day or serving more than 500 inhabitants. The wells chose to protection perimeters delimitation were the AC22 and AC23. 8

9 There are several approaches: geometric (for definition of arbitrary fixed radius, as is the case the radius of the immediate protection area proposed by Decree-Law No. 382/99), analytical and numerical, proposed by several authors, among which are in the present work the analytical method proposed by this Decree, the method of the Calculated Fixed Radius. These are compared with results obtained by numerical simulation performed. The risk map was made by binarizing all matrices obtained by simulation, considering square cells of 10 m, by assigning 0 and 1 for each cell, where 0 implies no occurrence of contamination and the occurrence of a contamination in the well after 3500 days. The choice of the size of cells to discretize sought to obtain a good definition of the areas obtained. The sum of the 10 matrices obtained will give a range from 0 (no occurrence of contamination) to 10 (the occurrence of contamination), which can be turned into probabilities by dividing by the number of simulations. The risk map is presented in Figure 5.1. Figure 5.1 Risk map of contamination in the wells AC22 and AC23, after 3500 days, comparison with the perimeter obtained considering an homogenized transmissivity (on the right); global view on the left. Conforme se verifica na Figure 5.1, o perímetro obtido considerando a transmissividade homogénea a toda a área tem tendência a subvalorizar certas áreas e a sobrevalorizar outras, ainda que considere a direcção do fluxo de água subterrânea. This methodology can be of great importance during the decision-making regarding the limitations of land use that the Decree-Law imposes on these protection zones. For the intermediate zone of protection, compared with the results obtained during the numerical modeling (r = 50 m) and taking into account that the generation that was coming from a fairly circular shape, the values are in themselves very similar (r = 40 m for the calculated fixed radius). Comparing the results obtained for the protection zone extended to the results obtained by stochastic modeling (Figure 5.2): Figure 5.2 Comparison of the results obtained for the protection zone extended to the results obtained by stochastic modeling As can be seen through the analysis of Figure 5.3, the perimeters defined with ASM and the calculated fixed radius method are not consistent. The risk of contamination in the well at the end of 3500 days, decreases as we move away from them, taking into account the direction opposite to the direction of groundwater flow. The perimeter generated by stochastic modeling is asymmetric and irregular. 9

10 The perimeter determined by the calculated fixed radius does not consider the direction of groundwater flow. Thus, it appears that for the shorter distances, corresponding to the intermediate zone there is less discrepancy between the two types of perimeters. However, for larger distances (enlarged zone), the discrepancy is more apparent. These differences result from the lack of hydrogeological data (excluding porosity) when doing the delimitation of the perimeter protection with the fixed radius method. 6 Conclusões This study, which covered different, but interlinked areas, as the hydrochemical study, the characterization of susceptibility to pollution and the subsequent delineation of perimeters of protection for the aquifer of Torres Vedras, allowed to conclude: The Torres Vedras aquifer is a multilayered aquifer and confined, with the presence of numerous clay lenticules that confers it heterogeneity; On the surface outcrop mainly clay lithologies, which increase the residence time of pollutants, acting as a groundwater protection at the level of pollution, however, they need protection because the risk of contamination despite average, exists and needs be considered; The water distributed to the public network, whose management company is the SMAS, is mostly composed of water purchased to EPAL, and the wells themselves contribute to a significant extent. The scheme aims to ensure the operational exploitation of the supply system in case of failure of the conduct of the ADO. However, water from the catchment itself, the aquifer under study is currently unfit for production of drinking water; The hydrochemical facies predominant in the aquifer system of Torres Vedras are sodium chloride and calcium bicarbonate, mineralization medium to high, ph-neutral to acid; The temporal analysis performed for the parameters analyzed the quality of groundwater for each monitoring station indicated past problems of contamination in some wells, particularly oil, manganese, barium, mercury and nitrogen ammoniacal; In all monitoring stations were analyzed and 2009, the amount of fluoride was insufficient for the required by law; The determination of the susceptibility index (Ribeiro, 2005) concerning the vulnerability to pollution of the aquifer classifies the aquifer of Torres Vedras as being of medium to high susceptibility (dominant classes 25th 45), not occurring classes of extreme values, being the highest class of 55 to 65.5 (which occurs in 6% of the total area of the aquifer); The delimitation of the protection perimeter by numerical modeling for the intermediate zone of the wells AC22 and AC23 proved to be very similar between the method of stochastic modeling and method fixed radius, however, the results for the protection perimeter enlarged zone with these two methods were different from each other. The fixed radius method does not consider the direction of groundwater flow, or other hydrogeological parameters (excluding the value of porosity). 7 Bibliographic References ALMEIDA, C., MENDONÇA, J.J.L., JESUS, M.R., GOMES, A.J. (2000) Sistemas Aquíferos de Portugal Continental Torres Vedras O25, Centro de Geologia da Faculdade de Ciências da Universidade de Lisboa; Decreto-Lei n.º 236/98 de 1 de Agosto, Normas Critérios e Objectivos com a finalidade de proteger o meio aquático e melhorar a qualidade das águas em função dos seus principais usos, Diário da República nº 176, Série I Parte A, Ministério do Ambiente, Lisboa; Decreto-Lei n.º 382/99 de 22 de Setembro, Normas e Critérios para para a delimitação dos perímetros de protecção de captações para abastecimento público, Diário da República nº 222, Série I Parte A, Ministério do Ambiente, Lisboa; Directiva 2000/60/CE, de 23 de Outubro de 2000, Parlamento Europeu e do Conselho, Jornal Oficial da União Europeia; 10

11 FARINHA, J., FERREIRA, J.C., RODRIGUES, E., QUARESMA, C., AMORIM, J., CARVALHO, P., FERREIRA, F., SOUSA, M. (2007) - Plano Municipal de Recursos Naturais, Volumes I e II, Câmara Municipal de Torres Vedras e Faculdade de Ciências e Tecnologia da Universidade Técnica de Lisboa; NEVES PALMA e VIEIRA DA SILVA, A. (1982) Estudo Hidrogeológico do Sistema Aquífero Cretácico de Torres Vedras, Barcelona; RIBEIRO L. (2005) Desenvolvimento e aplicação de um novo índice de susceptibilidade dos aquíferos à contaminação de origem agrícola - in Actas do 7º Simpósio de Hidráulica e Recursos Hídricos dos Países de Língua Oficial Portuguesa, ed. CDROM, APRH, Évora, Portugal; 11