GNGTS 2024 - Atti del 42° Convegno Nazionale

Session 3.2 GNGTS 2024 coarse-grained high-permeable ( K > 10 −2 m/s) deeper one (sand and gravel). The dam was operatng with a maximum water level well below the originally designed working conditons (256 m a.s.l.), since seepage phenomena occurring at the right abutment of the reservoir were confrmed by piezometric measurements that showed an increase of groundwater level downstream, which occurred when the reservoir level was higher than 250 m a.s.l. (Fig. 2b). Therefore, assessing the thickness and lateral extension of the coarse-graded highly permeable unit is pivotal for a correct design of the planned rehabilitaton interventon (cut-of wall). Figure 2. Penne case study (Central Italy): (a) aerial view with locaton of the ERT/SRT line and of the three boreholes; (b) hydraulic head logged as a functon of the reservoir level (afer Cardarelli, Cercato and De Donno, 2018). Both ERT e SRT were performed along a 142 m alignment (Fig. 2a), with equally 2 m spaced sensor points (electrodes and geophones locatons). ERT profle was acquired using the IRIS Instruments Syscal Pro resistvity-meter with 48 stainless steel electrodes in a pole-dipole array confguraton, while P-wave seismic data were recorded employing a 48-channel system of 8 Hz vertcal geophones by running a shot every two geophones using an 8-gauge Minibang shotgun. Three boreholes spaced 40 m apart along the geophysical line were drilled down the alignment for validatng the geophysical surface investgatons (Fig.2a). Results from individual inversion (Fig.3a) exhibit a three-layer geometry: low resistvity and P-wave velocity values (10-40 Ωm, 400-800 m/s) at shallow depths indicate the presence of the fne-grained alluvial deposits, followed by a less conductve formaton (ρ = 100-300 Ωm, v P = 1000-1800 m/s) that can be linked to the coarse- graded alluvial material and by a deep low-resistvity stfer medium (ρ = 5-10 Ωm, v P > 1800 m/s) revealing the presence of the fysch unit. Resistvity model clearly shows the thickness of the sand and gravel layer (about 7-10 m), but its lateral extent cannot be accurately determined in an area of lower resoluton, even if it seems to vanish in the second part of secton. Boreholes clarify the geometry of this layer: it has a thickness of 7 m (from 253 to 246 m a.s.l.) in BH1 (x = 23.5 m) and 8 m (at the same elevaton) in BH2 (x = 63.5 m) confrming it has a sub-horizontal trend, but it was not detected in BH3 (x = 103.5 m), so we can deduce it ends between 63.5 and 103.5 m. The jointly inverted resistvity model (Fig.3b) displays a similar thickness of the coarse-graded material but its lateral extent is signifcantly reduced, ending around 80-85 m in good agreement with borehole data. Conversely, the P-wave velocity model remains nearly unchanged. The NCG index (Fig.3c) shows a decrease in the higher values at the central zone of the secton, where the joint approach conspicuously impacts in beter describing the terminaton of the sand and gravel unit. Finally, the quanttatve integraton of geophysical models through the results of cluster analysis is able to well