**5. Results and Discussion**

The nonspatial analysis which involves the results of the target pollutants laboratory determination, is provided with the concentration values presented in Supplementary Material (Table S1 for benzene, Table S2 for toluene, Table S3 for ethylbenzene and in Table S4 for the xylenes). All values are expressed in parts per billion (ppb). The sample id "22" was considered unsuitable for chromatographic determination, because its oily texture and odor appeared likely to provokedamage to the instrumentation. The produced results of the first year indicated the need to broaden/expand the sampling area.

The BTEX are not commonly found in nature, as for instance the metals, and since they are hazardous for human health, it could be considered that every concentration over a non-detectable level may be regarded as "pollution" or "contamination". Regarding benzene values, we remark that seventeen (17) out of forty-five (45) samples exceeded the EU MCL in all autumn monitoring seasons, and another five (5) once or twice again in autumn. Overall, benzene pollution might be alleged as not-at-a high-risk, though significant. More specifically, in November 2003, the concentrations of benzene range from the detection limit up to 29 ppb, in May 2004 the water level had an average rise of 0.45 m and the maximum concentration was 14 ppb, whereas the concentrations of all other samples were significantly lower than the ones in November 2003. In November 2004, new samples were collected from the Refinery of Eleusis (id Petrola "10300", Petrola "10400") and the industrial area of Pyrcal (id "10500", "10600", at the western boundary of the study area). Exactly at the same area, the maximum concentration (57 ppb) was found. Between May and November 2004, the average fall of the water level was 0.24 m. In May 2005 concentrations ranged from the detection limit up to 16 ppb, and the average rise of the water level was 0.39 m. Finally, in November 2005, the concentrations ranged from the detection limit up to 20 ppb and it seems that they were slightly lower than in the previous autumn sampling periods. Moreover, the sample id "22" was a mix of water and oil, which raised the question of its origin (a "backdoor" discharge?). This sample was not chromatographically characterized at all, to avert damage to the analytical instrumentation but has served as a "guide" for new sampling in the surrounding area.

As for the concentrations of toluene, ethylbenzene and xylenes, their values were much lower than their MCLs (Supplementary Material). More specifically, the pollution of toluene can be assumed as negligible to low, since its concentration values are much lower than the MCL. The maximum concentration of 45 ppb (still, lower than the MCL) was reported in the installations of the "Pyrkal" industry (id "10500" and "10600"). For ethylbenzene, almost all values are close to the detection limit. For the xylenes, the pollution should also be considered as negligible to low, compared to its MCL. Moreover, it is remarkable that, after the winter rainfall, the BTEX concentration values in May are significantly reduced, reaching very low to non-detectable levels. This may be attributed to the significant dilution of the target compounds by the recharge of the aquifer, which at the end, discharges to the sea through submarine water springs at the western boundary of the study area, mainly.

Concerning the spatial distribution depicted on the GIS contour maps, four separate locations of benzene pollution are developed: (a) northeast of the Hellenic Refinery of Aspropyrgos (ELPE), (b) at the western and the eastern boundaries of the Military Airfield of Eleusis, (c) south-east of the town of Aspropyrgos and (d) next to the Refinery of Aspropyrgos (ELPE). Figures 4–8 illustrate the estimated groundwater distribution. The spatial distribution of toluene, ethylbenzene and xylenes is the same as benzene; therefore, the results are not presented and commented for each compound individually.

The standard deviation is very high only for benzene for all monitoring seasons. As for the other compounds, the variability is potentially attributed to, either the existence of a non-permanent pollution input or, an eventually differentiated natural attenuation. The examination of the (B + T)/(E + X) ratio could provide some useful information on the original time of gasoline release [44,45]. Due to the difference in their mobility, concentration ratios of the individual BTEX compounds (e.g., B/E, T/X, etc.) in fuel-contaminated groundwater tend to change uniformly with time. The relative content of BTEX compounds in manufactured gasoline has varied with time; initial ratio values are often unavailable, which limits wide application of these ratios as time indicators. Although, in the study area we cannot

identify numerous oil industries or fuel stations that have undertaken historically serious changes, we could still use the BTEX ratio: R<sup>b</sup> = (B + T)/(E + X) in an attempt to better comprehend some environmental changes in dissolved plumes, if any.

**Figure 4.** The estimated groundwater distribution of benzene in November 2003.

**Figure 5.** The estimated groundwater distribution of benzene in May 2004.

**Figure 6.** The estimated groundwater distribution of benzene in November 2004.

**Figure 7.** The estimated groundwater distribution of benzene in May 2005.

**Figure 8.** The estimated groundwater distribution of benzene in November 2005.

Kaplan's multiple investigations [45,46], mostly empirical, have shown that near the source and immediately after a gasoline spill, R<sup>b</sup> reaches values between 1.5 and 6, which indicates a recent release (typically <5 years). In the absence of NAPL, R<sup>b</sup> in a dissolved gasoline plume is close to that of the original gasoline (0.8 < R<sup>b</sup> < 1.1). The ratio then decreases as a function of time, and values less than 0.5 usually reflecting gasoline residence time longer than 10 years. R<sup>b</sup> = 6.6 confirms the presence of a thick gasoline rich layer. In general, a two-fold decrease of R<sup>b</sup> occurs in 2.3 years.

Applying the R<sup>b</sup> values onto our reported concentrations, we may conclude that for the Refinery of Eleusis (id "10300", "10400", "10500" and "10600"), located at the western boundaries of the contour maps, where the highest concentration values were reported, the calculated R<sup>b</sup> < 0.5 indicates a long permanent BTEX occurrence. For the second most polluted area surrounding the Military Airfield, (sample id "98", "5000" at west, "3000", "44" at south, "54", "55" at east), the calculated 1.5 < R<sup>b</sup> < 6 reveals a recent release of BTEX, consistent to the nature of the pollution source, including heavy local atmospheric pollution from the fuel used in planes and the washing out of the atmosphere by rain. The fact that the groundwater from the southern boundaries of the Airfield (id "102", "104", "106") seems not to be affected by BTEX release due to the confined nature of the groundwater system, might be partly attributed to geomorphological and hydrological specifications not yet fully understood. As for the concentrations near the Refinery of Aspropyrgos (id "8"), the R<sup>b</sup> < 0.5 indicates a long permanent BTEX source. Apparently, the city of Eleusis itself seems not to be polluted from BTEX, though it is located at the furthest downgradient part of the aquifer flow lines. Theoretically, the

pollution should "track" the groundwater flow into the Gulf of Eleusis and, thus pollution should have visibly affected the southern areas, near the coastline. It is likely that the benzene has not reached the gulf because of biodegradation along the flow paths. Biodegradation is an important process attenuating benzene and other BTEX constituents in groundwater in many areas.

Finally, it is worthwhile to examine the occurrence of 0.6 ppb and 2.4 ppb at the boreholes with id "1", "3", located within the Triassic carbonates (limestone/dolomites). There, the groundwater is pumped within the Mesozoic limestones from a depth of 78 up to 160 m, with no evident pollution source nearby. If we exclude an ad hoc illegal pollution, benzene contamination may be explained by some hydraulic communication between the Pleistocene sediments and the carbonate ones.

Overall, the local character of the BTEX occurrence may be documented by the confined, multi-layered hydro-stratigraphic system as well as, the existence of impermeable aquifer barriers. Additionally, the very low concentrations of the spring sampling periods reveal both a good aeration and a dynamic seasonal enrichment/dilution of the groundwater system.

#### **6. Conclusions and Suggestions**

The laboratory determination and the spatial analysis and study of BTEX, in the Thriassion Plain lead to some useful general conclusions: Some high concentrations from BTEX were identified in autumn when the water table is at the furthest downgradient level. Benzene concentration values are much more elevated than those of the other BTEX compounds. The spring concentrations of all compounds are very low. The considerably reduced rainfall of the summer months coupled with enhanced abstraction through groundwater pumping could explain the difference between the autumn and spring benzene concentrations. The high values of benzene as well as the empirical Kaplan's values, indicate that the pollution originates mainly from permanent sources namely oil refinery and handling sources, located relatively close to the sampling points. The spatial distribution of BTEX in groundwaters show that they were concentrated mainly in four, rather restricted locations. Three of them were in the close vicinity of evident pollution sources (a military airfield and two crude oil refineries), whereas, the other one corresponds to an abandoned site with no outstanding pollution sources where wells exist, eventually used occasionally for illegal dumping of oily wastes. The town of Eleusis seems not to be polluted from BTEX, despite the fact that it is located at the lowest part of the aquifer flow lines. This could be attributed to impermeable strata working as water and pollution barriers. The hydro-stratigraphic particularities, along with general hydrogeochemical conditions play an important role to the BTEX fate and attenuation rate.

Recent developments in the expansion of the works of the refineries, including changes in the operations of the Military airport, make it necessary, in our point of view, to proceed to a new and deeper BTEX pollution investigation, to revisit our approached attenuation after 15 years of operation. In case that further studies and monitoring demonstrate that the pollution persists, the National and Local Authorities should take urgently active and effective measures to prevent further deterioration and environmental damage. The designation of the city of Eleusis as cultural capital of Europe 2021, though somehow disturbed by COVID-19, may offer an opportunity for also refocusing on environmental pollution and the groundwater as a valuable natural resource that requires further attention including by keeping it free from hydrocarbon pollution.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2077-1312/8/12/1018/s1, Table S1. Concentrations of benzene, Table S2. Concentrations of toluene, Table S3. Concentrations of ethylbenzene; and Table S4. Concentrations of xylenes.

**Author Contributions:** Conceptualization, P.M.; methodology, P.M.; statistical analysis (GIS), H.D.S., G.D.B.; validation, P.M., E.S., D.H.; formal analysis, P.M.; investigation, P.M., E.S., D.H, G.K., M.S.; resources, P.M., D.H.; data curation, P.M., E.S.; writing—original draft preparation, P.M.; writing—review and editing, P.M., G.K., E.S., D.H., S.D.Z., A.A.; visualization, P.M.; supervision, M.S., A.A.; project administration M.S., A.A.; funding acquisition, M.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was co-funded by the General Secretariat of Research and Technology and the National and the Kapodistrian University of Athens (NKUA) in the context of the EU 3rd Framework Programme, Action 8.3.1. (01 ED 150).

**Conflicts of Interest:** The authors declare no conflict of interest.
