The Monte Alpi (southern Italy) is well known as an outcropping analogue of the Val d'Agri subsurface units which host a number of oil fields of economic importance. This paper presents an example of a detailed study which integrates stratigraphy, structural geology, tectonic analysis and fracture network study. The new structural model presented is based on a detailed 1:12,500 field survey of the area, whose results are illustrated in a comprehensive outcrop map, geological map and structural cross-sections, and stereographic projections and palaeostress inversion calculations of a mesostructural data set. The Monte Alpi Unit is characterised by a 2000-m-thick sequence of Jurassic-Cretaceous platform carbonates, conformably covered by Middle-Upper Miocene calcarenites. This sequence is unconformably covered by Upper Messinian terrigenous clastics, mainly along a number of major growth faults. Our detailed mapping revealed that the Monte Alpi Unit is tectonically overlain by the Lagonegro basinal Unit (thin siliceous carbonates and remnants of pelagic pelitic and carbonate deposits, all highly deformed) and by the Monte la Spina Unit (cataclastic Mesozoic carbonates and dolomites). This solves earlier raised questions concerning the palaeogeographic provenance of the carbonates outcropping in the area, which can thus: without any doubt be correlated with the external carbonate domains known from the subsurface. The present-day structure shows a structural inversion along the major Late Miocene growth faults which acted as sinistral transpressive faults. The latter faults are part of a regional set dissecting the southern Apennines, active up till at least Middle Pleistocene times. A fracture study was performed on the Mesozoic carbonates of the area in order to collect data regarding length distribution, orientation, density and spatial heterogeneity of the fracture network to be used for considerations regarding exploration and development of the Val d'Agri reservoirs. The data were collected on 37 horizontal (strata-parallel) and vertical outcrop surfaces through photographic restitution and digitising of fracture lineament traces, and measurements of orientations of all visible fractures. Furthermore, three major outcropping rock walls along the flanks of Monte Alpi and Santa Croce were photographed and fracture and fault zone lineaments were digitised. All data were inserted and calibrated in a three-dimensional (3D) Geographic Information Systems and Computer Aided Design (GISCAD) environment consisting of a database and a 3D geographical framework comprising topographic map, Digital Terrain Model (DTM), geological map, aerial photograph lineament map and a 3D reconstruction of major fault surfaces. Making use of the potential of the 3D working environment, outcrop surfaces and rock walls were reconstructed at their correct 3D position and orientation. Data analysis comprised stereographic contour diagrams and rosette diagrams of 1800 fracture orientations in order to recognise main fracture sets. Three corrections were performed on these data: length weighting, Terzaghi correction (removal of sample orientation bias) and tilt removal. From a comparison between orientation data on various scales combined with field evidence, it can be deduced that fault and fracture sets are part of the same structural pattern, and that fracture sets were generated prior to Plio-Pleistocene tectonics and passively rotated. Fracture and fault length distribution was analysed by the generation of diagrams which plot cumulative frequency versus length in log-log space, normalising for area of observation. These diagrams, composed of about 30,000 measurements and constituting the most complete data set presented in the literature up till now, show that small-scale and large-scale features form part of one scale invariant system with fractal properties. No scale gap is present in our data set which comprises rock wall data in the 'subseismic' scale range. It appears that the 3D fracture network can best be described by a fractal dimension of 2.98 and a coefficient of 0.25. The outcome of the study places important constraints on fracture network simulation exercises in hydrocarbon reservoir simulation studies, which use numerical input parameters in order to make predictions of the tectonic texture present in the subsurface. (C) 2001 Elsevier Science B.V. All rights reserved.
