**3. First Estimates of Theoretical CO2 Geological Storage Capacity in Gas Fields and Deep Saline Aquifers**

Three different types of storage objects were found as prospective for geological CO2 storage in the Adriatic offshore. Firstly, the Pliocene and Pleistocene sands/sandstones that have favourable petrophysical properties and are documented to be gas-tight. The second option is seen in Miocene sandstones locally present in offshore foreland basins like the Dugi otok basin, and the third is found in Upper Cretaceous limestones with primary and secondary porosity covered with impermeable Miocene or Pliocene sediments. In petroleum geological exploration terminology, these three exploration targets would be called "plays". By analogy, what is described in the following text are the three "geological CO2 storage plays."

To evaluate the geological CO2 storage potential of these plays, we firstly conducted regional-scale mapping of the Top carbonate complex horizon and then delineated areas that were more favourable from others. Actually, these areas are the preliminary mapped structural uplifts, which would have qualified them for the "structurally defined deep saline aquifers" if their local geological models were confirmed. These initial estimates were done based on the results of previous petroleum geological exploration, integrating them into the concept of the "theoretical storage capacity assessment," meaning that the most important properties are mapped: subsurface extension and depth range of the most important porous and permeable rock formations, thickness of their impermeable cover and zones of seismic activity that should be avoided. With regional estimates of these properties and areal extension of favourable zones, it becomes possible to make numerical estimates of storage capacity on a basin scale. This is usually called "theoretical capacity" and its only purpose is planning. It can be at first planning of land use, due to potential conflicts of interest, but the most important is planning of future targeted exploration in prospective areas. This is the way it has always been done with mineral resources, to gradually come from regional assessments to local geological models of the subsurface on locations where the exploitation (storage in this case) projects might be developed. Large capital investments in such operations dictate this procedure, which is mirror of the one used in the upstream petroleum industry, and consequently, has been proven to be the best way to substantiate the investment decisions. In that sense, a techno-economic pyramid depicting different levels of estimates of CO2 storage capacities were developed based on the concept of energy resource pyramid introduced by McCabe [43]. Assessment of theoretical storage capacity means to make a numerical estimate of the total resource; with additional works some of it will become "effective," meaning that this is the capacity that might really be used since the uncertainties have been sufficiently reduced, and in the end the third conceived level would be the "viable" capacity that also includes economical aspects, and is by analogy equal to the "balanced reserves." In this paper the estimates of theoretical capacity are presented, based on the publicly available data. It should be noted that not all units/exploration targets are at the same level of assessment and are, therefore, described separately, in the following subchapters.

## *3.1. Potential Storage Objects in Depleted Gas Reservoirs*

Theoretical capacity estimates were firstly performed for three gas fields in the Northern Adriatic offshore—the Ida, Ika and Marica Fields (locations in Figure 1). The capacity was calculated based on the total recoverable volume of gas under reservoir conditions, considering that CO2 could replace the volume that was previously occupied by the gas in the reservoirs. All three assessed gas fields have multiple reservoirs of Pliocene and Pleistocene sands/sandstones, and in addition, the Ika gas field contains one reservoir in Upper Cretaceous limestones [10]. Presently, the reservoirs are not depleted. Their total potential storage capacity is estimated, and they can be converted to storage objects by making use of existing offshore installations (network of pipelines shown in Figure 1).

This theoretical storage capacity estimate has been performed based on publicly accessible data on recoverable reserves [44,45] by using the 1:1 replacement principle—the amount of CO2 that can be stored underground into a depleted oil or gas field is equal to total oil or gas (that will be) produced:

$$
\rho\_{\rm CO\_2} = \rm LIR \times \rho\_{\rm CO\_2} \times B \tag{1}
$$

where *B* is the gas or oil formation volume factor (ratio of volume of fluid in reservoir versus volume in standard conditions); *m*CO2 is the mass (kg) of CO2 that can be stored; ρCO2 is the CO2 density at reservoir conditions; and *UR* is the total volume of oil or gas produced; i.e., the proven ultimately recoverable recoverable oil or gas.

For calculation of CO2 density for geosequestration, the real gas equation of state was used [46]. Formation volume factor for oil is very accurate because it was measured in laboratory. For the gas fields, assuming that the real gas volume correction, i.e., the compressibility factor Z of the gas at the surface, is 1, volume factor *Bg* can be expressed as:

$$B\_{\mathcal{S}} = 0.0034632 \times (T\_{\mathcal{I}}/p\_{\mathcal{I}}) \times Z \tag{2}$$

where *Tr*: reservoir Temperature (K); *pr*: reservoir pressure (bar).

Table 1 shows a summary of parameters used in calculation of the CO2 storage capacity of the three gas-fields in the Northern Adriatic. The total estimated CO2 storage capacity for these gas-fields amounts to 32.112 Mt. Notably, most of the reservoirs of these three fields are still in production and will not be available for at least a decade.

**Table 1.** Summary of parameters used for CO2 storage capacity estimation of the Northern Adriatic gas-fields.


*3.2. Potential CO2 Storage Objects in Deep Saline Aquifers*

The past exploration results that were available and used in this part of the study originate from the 1971–1985 period when only few regional deep wells were drilled. These wildcat wells were not totally dry—there were gas shows in most of them and traces of heavy oil in two. With maximal depth exceeding 6000 m, this drilling campaign brought up information on the lithostratigraphy of Triassic, Jurassic, Cretaceous, Paleogene, Neogene and Quaternary sediments. Additionally, a set of correlation horizons was established then, based on the lithostratigraphy from wells and sequence boundaries observed on reflection seismics (Figure 5):


The transversal correlation section A–B shown in Figure 2. (for location see Figure 4) is given here as an attempt to illustrate a possible reconstruction of the subsurface geology based exclusively on the vintage deep regional wells.

**Figure 5.** Schematic correlation of section 1–1 (location in Figures 3 and 4).

The oldest E horizon delineated as the base of the carbonates was not drilled in the area of Dugi otok basin, so the thickness can only be interpreted based on the seismic data. The oldest penetrated unit is the one below the D horizon (Figure 5), composed of Triasssic dolomites and dolomitic limestones, characterized by frequent occurrence of stylolites, moldic and fenestral porosity (predominantly developed in the basal and middle part of the unit).

Jurassic and Cretaceous limestones together, build up the D–C interval. Their drilled thickness is from 900 to more than 4000 m, depending on structural position of analysed wells. The basal part of this unit is of grey to greenish dolomitic limestones with chert lenses and nodules, also with sporadic black marl intercalations. Porosity is markedly variable—from several up to even 20% based on well log interpretation [28]. The central part of this unit consists of Lower Cretaceous white limestones that have joints filled with anhydrite; limestones with stylolites; limestones with chert; and bituminous limestones and overlying dolomites. Sporadically, these sediments are characterized by increased porosities within the zones encompassing several tens of meters, interpreted to be caused by brittle tectonics. Overlying Upper Cretaceous layers are composed of dense limestones (occasionally with chert or bitumen), and bioclastic limestones (chalk), white limestones with chert and rudist limestones. Joints are not common in this unit and the existing ones are filled with organic matter—either bitumen or heavy oil, and the same goes for the stylolites. There are zones in this sub-unit where secondary porosity can be expected, but without any information about frequency and orientation of predominate joint sets. Porosities and permeabilities of Jurassic and Cretaceous limestones are strongly controlled by diagenetic and postdiagenetic processes, including dolomitization, recrystallization, dissolution, leaching, erosion and weathering. Defining the intensity and distribution of these processes in rocks would require a thorough reinterpretation of well and seismic data, which is beyond the scope of this paper.

Early to Middle Eocene limestones are the dominant lithology in the C–B interval, where basal parts of this unit also include chalk limestones and carbonate breccia. Less developed lithologies are dense laminated limestones and calcitic marls with chert nodules. Striations are observed in cores and stylolite joints as well. The youngest part of Eocene (close to B horizon) exhibits coarser carbonate bioclastics—calcarenites and calcrudites. The drilled thickness of the Eocene unit in the area is usually between 120 and 800 m, although there are wells (like Well KM-2 in Figure 5) where it is totally missing. In such cases the B horizon corresponds with the "top carbonate complex horizon".

The composite unit of the B–A interval includes clastic and carbonate sediments of Upper Eocene, Oligocene and Miocene age. Differently subsiding tectonic blocks within the Dugi otok basin are not reflected only in orientation and size of structures, but also in changes of accommodation space. Here, the Eocene-Miocene unit is found at depths between 300 and 1330 m with a variable thickness from 100 m to more than 3000 m in the central part of this structural depression. Basal part of this interval is made of Eocene flysch-like deposits—mainly marls, marly calcarenites and limestones. They are covered by Oligocene sandstones with intercalations of calcitic marl and then by the Miocene flysch-like deposits again.

The youngest unit is comprised of deposits above A and bellow Q horizon, i.e., between the Top Miocene horizon and the seabed. That unit is composed of loose, silty-sandy sediments of Pliocene, Pleistocene and Holocene age. The depth of A horizon ranges from 300 m to over 1200 m, which at the same time gives approximately the thickness of this unit. Pliocene marls, sands and clays are distributed throughout the Dugi otok basin. Quaternary sediments are transgressive in the NE part and in conformity with Pliocene in the SW region. Transgression started in early the Pleistocene age with marine sedimentation that is still ongoing. The thickness of Quaternary sediments is also variable—from 300 m to more than 1200 m.

Based on the available knowledge of subsurface geology, there is potential for geological CO2 storage in the Adriatic off-shore in two regional units: Miocene sandstones (parts of the unit between A and B horizons, see Figure 5), and in the carbonate complex bounded on top by either C horizon or B horizon (Figure 5). The most important difference between these two units is in lithological compositions and porosity types. Miocene layers are a bit better explored and their depth range in combination with intergranular porosity appears to be more favourable, but they lack large structural closures. Underlying rocks of the carbonate complex have both the primary and secondary porosity (with locally increased permeability), and, in addition, there are numerous closed structures that are relatively easy to recognize on a structural contour map of the Top carbonate complex horizon (Figure 3). However, there is a lack of sufficient data to characterize and estimate areal distribution of reservoir properties in this unit.

The CO2 storage potential in these two large units must, therefore, be assessed in two different ways: Miocene sandstones are studied as a regional deep saline aquifer (DSA) named "Dugi otok", because they were mapped within the Dugi otok basin, while within the carbonate complex, several structural uplifts were identified and referred to as structurally defined aquifers.
