**3. Geology and Hydrogeological Characteristics of the Study Area**

The Palaeozoic formations, 300 to 400 m thick, are the geological basement of the broader area. The Palaeozoic is represented by (a) clastic materials, shales and sandstones, which alternate with greywackes and conglomerates; (b) basic igneous volcanic rocks; and (c) lenticular intercalations of thin bedded carbonate units. Weak metamorphosis, successive folds and schistosity are also observed. This background is overlain by Mesozoic sediments such as: (a) phyllites and sandstone, (b) meta-pyroclastic and meta-volcanic rocks hornstones and tuffs; (c) Triassic limestones, dolomitic limestones and dolomites; and (d) Cretaceous limestones. The Cenozoic is represented by: (a) Paleocene flysch; (b) Neogene deposits of marls with lignite intercalations in places, sandstone, marly limestone; and (c) Pleistocene (clay, sands, gravels, torrential fans of loosely and cohesive conglomerates) and Late Quaternary alluvial deposits (clay, loams, sands and gravels) at the top [28–30]. The geological settings are illustrated in Figure 2.

**Figure 2.** The geology of the study area.

The hydrogeology of the broader area has been extremely influenced by hydro-stratigraphic and tectonic factors as well as sea level changes [29–34]. The paleogeographic evolution, the aquifer geometry and thickness, the salinity controls, and the groundwater input are important features for the water source management. In past studies two main aquifer systems were described: (1) the Plio–Pleistocene sediments and (2) the Triassic limestone/dolomite and the Cretaceous limestone [18,28,35,36].

Recent detailed hydrogeological research studies [37,38] have recalculated the aquifer characteristics, providing a revised hydrogeological model of three distinct groundwater hydrostratigraphic units (including their subunits): (a) the Neogene-Quaternary sediments comprising (a.1) the Holocene -Upper Pleistocene aquifer system, generally unconfined and semi-confined consisted of clays, sands and gravels. It is recharged by rainfall and upward leakage from lower aquifers. The Holocene clays form local barriers to groundwater flow, causing the formation of marshes and swamps and/or upward leakage. The coastal clay strata prevent the lower standing aquifers from direct seawater intrusion where groundwater of good quality was found 90 m below sea level (m.b.s.l.) [37,38]; (a.2) the Pleistocene system, made up of marls with lignite layers in places, sands and marly limestones which is a multi-layered confined aquifer. There is Pleistocene sediments recharge and preferential lateral flow from the carbonate basement. The direction of the groundwater flow within the confined layers is SE, where a partly upward leakage onto the coastal zone occurs while, the main water volume is moving very slowly towards the Saronikos Gulf with hydraulic gradient 1–2‰. (b) The Cretaceous aquifer, west of the study area, is fractured and forms an unconfined aquifer of high productivity with not well-known thickness, but with hydraulic contact with the sea, and finally, (c) The Triassic aquifer, either confined or unconfined which has been affected by karstification and tectonism. The latter is comprised of limestones, dolomitic limestones and dolomites. The Cretaceous and Triassic aquifers flow to the southwest and discharge partly through coastal springs near the city of Eleusis and the entire North-Western coastline of the Eleusis Gulf and partly into Neogene–Quaternary deposits [39,40]. From this field work, the seasonal water level fluctuations in the unconfined aquifer were measured 0.4–0.5 m and within the confined aquifers 1.2–1.5 m. A conceptual model of the deposits in Figure 3 illustrates the multi-level aquifer conditions within the Pleistocene strata, where lenticular intercalation consists of a water-yielding formation [31].

**Figure 3.** Conceptual hydro-stratigraphic column of the Plio-Pleistocene.

#### **4. Materials and Methods**

The sampled area included all major potential polluting sources, from crude oil refineries and other related industries and the Military Airfield, as well as the city of Eleusis itself. Thirty-four (34) to forty-five (45) groundwater samples were collected, twice a year, in November and May, when the water level was at the lowest and the highest level, respectively. The whole investigation lasted three years, from 2003 to 2005. The results of the in-situ survey, including water level measurements, and lithology identification of each sampling location, are reported in Table 2. Sampling, storage, shipping and laboratory characterization of the target analytes were carried out according to USEPA Analytical Method 524.2 [41] in the Laboratory of Environmental Chemistry, Department of Chemistry, National and Kapodistrian University of Athens (NKUA). The samples were collected from the first five centimeters of the water surface of the wells or by pumping from the boreholes. The Solid Phase Microextraction (SPME) was applied to pre-concentrate the BTEX [42–44]. Gas Chromatography (GC) was applied to separate the compounds from the water samples, and Mass Spectrometry (MS) to detect them. The chromatography results were validated through the protocol procedures (linearity, precision, recover, standards). The Limits of Detection (LOD) achieved with the method applied were 0.095 ppb for benzene, 0.077 ppb for toluene, 0.099 ppb for ethylbenzene and 0.089 ppb for total xylenes [44]. The quantified results were illustrated on a GIS database to assess the distribution of the pollutants within the aquifers. The estimation of the groundwater spatial dispersion (spatial analysis) was accomplished using an ArcGIS software. The data layers were digitized and stored in the database, including sampling-well locations with the analytical data, main industries and drainage network, road network and towns/settlements. The concentrations were presented on concentration-contour maps. Using the Geostatistical analyst extension of ArcGIS, the initial values of each parameter (i.e., concentration BTEX compound) for a sampling season were interpolated in grid layers with a cell size of 20 × 20 m. The inverse squared distance (ISD) method was used to interpolate the data obtained. The contour maps provided, refer to the concentrations of benzene for all sampling seasons. The classification of the concentration values in the contour maps is different among the four target analytes, therefore when a target analyte has a wide range of concentrations (e.g., xylenes) there are more classes in the GIS map.


**Table 2.** Hydrogeological characteristics of the wells; (U), (C) and (L) stands for unconfined, confined and semi-confined aquifer respectively; hydraulic head.

<sup>1</sup> Refinery of Eleusis.
