The Santos Basin is located in the southeastern waters of South America, adjacent to the Campos Basin to the north and the Pelotas Basin to the south, covering an area of 32×10⁴ km². There are two major exploration systems in the basin, namely pre-salt and post-salt, with the pre- salt system being the main focus. The pre-salt structure, generally NE-SW striking, shows a tectonic pattern of “two uplifts and three depressions”. From west to east, the western depression, the Aram-Uirapuru uplift, the central depression, the Lula-Sugar uplift and the eastern depression are developed successively (Fig. 1a).

The tectonic evolution of the Santos Basin is related to the disintegration of the Gondwana continent and the expansion of the South Atlantic since the Mesozoic, with three stages of rifting, transition and drift. Three ultra-thick sedimentary sequences were developed correspondingly, i.e., the rifting mega-sequence, transitional mega-sequence and drift mega-sequence ^[17]. The rifting mega-sequence includes the Picarras Formation, Itapema Formation and Barra Velha Formation. The Picarras Formation is mainly composed of deep lacustrine mudstone and shale, and the lacustrine shale is one of the main hydrocarbon source rocks in the Santos Basin. The Itapema Formation is dominated by lacustrine shale, marl and coquina. The coquina stratum is one of the main pre-salt reservoirs. The Barra Velha Formation is mainly composed of stromatolites and spherulites which is the main pre-salt reservoirs. The Ariri Formation is formed within the transitional mega-sequence, which is under an evaporative environment with mainly a set of thick evaporite deposits (halite and anhydrite) to form a regional cap rock. The drifting mega-sequence is dominated by marine deposition (Fig. 1b).

The oil and gas exploration activities of the Santos Basin began in the 1980s. Early exploration mainly aimed at structural traps in the shallow-water area. The discoveries of hydrocarbon were mainly small- to medium-sized gas reservoirs and minor oil reservoirs. At the beginning of 21th century, pre-salt exploration commenced in the basin until the Lula oilfield was discovered in 2006, which confirmed the existence of an effective pre-salt petroleum system and opened the exploration prelude of deep-water pre-salt hydrocarbon. The history of oil and gas exploration can be divided into 3 phases.

Six principles for tectonic unit division have been proposed based on the tectonic evolution and sedimentary sequences of the Santos Basin: namely, determining the units based on the tectonic characteristics of rift basin, dividing the units based on the structural pattern of uplift-depression at the late rift stage, determining the uplift-depression boundary based on the primary faults and sedimentary characteristics of the late rift stage, determining the bulge-sag boundary based on the secondary faults and sedimentary characteristics of the late rift stage, determining the segment boundary based on the slip-strike faults of the early rift stage, and combining geological and geographical information to name the units. The six principles have laid the foundation for the division of tectonic units in the Santos Basin.

Based on gravity anomaly and seismic data, an accurate post-salt stratigraphic distribution was obtained and its gravity anomaly was stripped by forward modeling. The gravity anomaly of the salt layer was obtained by the method of minimum curvature separation, and then the accurate pre-salt gravity anomaly was obtained. Combined with the pre-salt fault system inferred by the NVDR-THDR (normalized vertical derivative of the total horizontal derivative) method, the pre-salt structural pattern of the Santos Basin was restored under the constraints of seismic, drilling and ODP (ocean drilling program) data. A new tectonic unit division scheme was proposed. Under the guidance of the above ideas, seven criteria for tectonic unit division were determined: high-precision Bouguer gravity anomaly is the basic data for determining the unit division of the basin; the cross-basin seismic profile is the key data for determining the structural pattern and unit division; high-precision three-dimensional seismic data is the main basis for determining the unit boundary; logging and core data are important basis for determining paleogeography; stratigraphic thickness based on two-dimensional seismic data is the reference for determining the unit boundary; magnetic anomaly is the auxiliary data for identifying the basin boundary and structural pattern; and regional tectonic information is auxiliary for determining the control factors of tectonic evolution. The determination of the seven criteria has indicated the implementation method for the tectonic unit division of the Santos Basin.

Based on the above principles and criteria, the Santos Basin is divided into five first-order tectonic units, including two uplifts (Lula-Sugar uplift and Aram-Uirapuru uplift) and three depressions (central depression, western depression and eastern depression) (Fig. 1a). In the previous structural pattern, the Aram-Uirapuru uplift was considered to be part of the central depression ^[8-9] (due to its low topography, it was considered to have reservoir risks, resulting in few international petroleum companies bidding). In the new structural pattern, the Aram-Uirapuru uplift is considered to be a paleo-uplift adjacent to the central depression, and in the paleogeographical location, it was separated from the Lula- Sugar uplift (known mature hydrocarbon area) across the lake, so that the two should have similar hydrocarbon accumulation conditions. The great discovery and drilling results made in the Alarm Block in 2021 have fully confirmed the correctness of the understandings above.

The pre-salt source rocks in the Santos Basin mainly consist of lacustrine black shale of Picarras and Itapema formations of Lower Cretaceous, which are characterized by wide distribution, large thickness, good organic matter types, high abundance and great generation potential ^[1]. Drilling data reveals that this set of source rocks is 100 to 300 m thick and deposited in the lacustrine environment during the syn-rift period, especially the source rocks in the three structural units, namely the central depression, the western depression and the eastern depression, are of good quality with TOC of 0.51%-7.32% (the average value of 1.67%), HI of 292.13-998.00 mg/g (the average value of 721.61 mg/g), the average value of generation potential (the sum of S₁ and S₂) of 18.2 mg/g, and primarily Type I and Type II₁ kerogen. The hydrocarbon-generating sag in Santos Basin is over 100 000 km², which has great hydrocarbon-generating potential and resources.

The maturity of the pre-salt source rocks in Santos Basin is characterized by the zonation of "mature in the central and east, over-mature in the west", and the central depression provides the main oil sources. The thick salt rock in the central depression distributes continuously, which delays the pre-salt source rocks evolution. The pre-salt source rocks are at the peak stage of oil generation with R_o of 0.7%-1.3%, which shows that oil generation is dominant. Due to the large scale and strong hydrocarbon supply capacity of hydrocarbon generation kitchens in the central depression, sufficient oil sources are provided for the pre-salt oil and gas fields. The eastern depression also develops continuous thick salt layer, which mainly generates oil. However, the bulge belt and sag belt of the eastern depression are arranged alternately, and the sag distribution is discontinuous, leading to the great changes in the sedimentary thickness of source rocks. Only several small hydrocarbon source kitchens are developed locally, with a relatively insufficient hydrocarbon supply capacity. The salt rock is thin and discrete in the western depression, and the R_o values of most source rocks are over 1.7%, which is at the peak stage of gas generation. Therefore, the western depression generates gas due to the limited scale of hydrocarbon generation kitchens, and the hydrocarbon supply capacity is far less than that in the central depression.

Based on basin modeling, it is predicted that the pre-salt oil and gas resources in the Santos Basin are about 3899.0×10⁸ t, of which the oil and gas resources of Itapema Formation and Picarras Formation are 2446.3× 10⁸ t and 1452.7×10⁸ t, accounting for 62.74% and 37.26% of the total resources, respectively. Since the end of the Barra Velha Formation deposition, the Santos Basin has experienced multiple phases of hydrocarbon generation and expulsion, which is mainly manifested by the migration and accumulation of oil and gas from the hydrocarbon source kitchens along faults or structural ridges to the structural traps in the uplift zone. The central depression, as the primary hydrocarbon source kitchens of pre-salt reservoirs, has supplied more than 95% of the total discovered PIIP in the discovered oil and gas fields. The probable PIIP in Buzios oilfield on the eastern Lula-Sugar uplift belt exceeds 48×10⁸ t oil equivalent, and that of the Lula oilfield is about 35×10⁸ t. The Aram oil field, on the western Aram-Uirapuru uplift belt, has a huge amount of PIIP as well, and the structural traps on the uplift have a high degree of oil and gas filling, demonstrating a very strong oil and gas supply capacity of the source rocks in the central depression.

Paleo-structures refer to the positive paleogeomorphy formed mainly by extension, which can be divided into a rift stage and a depression stage during rift period in the Santos Basin. During the rift stage, the tectonic activities determined the embryonic form of the "two uplifts and three depressions" structural pattern, forming two carbonate platforms sandwiched between three deep lake zones, with the main lake located in the central depression. On the two platforms and many islands in lake zones and their periphery, the coquina, represented as freshwater lake basin carbonate rocks, was deposited with a thickness ranging from tens of meters to hundreds of meters, which is an important reservoir body for pre-salt hydrocarbon. During the depression stage, the tectonic activity was weakened and the structural pattern was inherited and developed. The Santos Basin evolved into a shallow saline lake basin, and spherulite and stromatolite facies were deposited with a thickness ranging from tens of meters to hundreds of meters (Fig. 2). It is the main reservoir body for pre-salt hydrocarbon distributed more widespread than the coquina reservoir.

Four types of positive paleogeomorphy, namely monoclinic fault block, fault-uplift, fault horst and volcanic high, control the thickness and scale of carbonate rock, thus controlling the distribution of favorable lacustrine carbonate reservoirs (Fig. 3). Monoclinic fault block and fault-uplift are the main positive paleogeomorphic units for the development of carbonate reservoir. The sedimentation is mainly characterized by accumulation and progradation. The mounded beach is commonly found with vertical superposition and lateral migration. The reservoir has a large thickness and wide distribution, which can be found in Lula-Sugar uplift, Aram-Uirapuru uplift and eastern depression. The total area of carbonate reservoir in the Lula-Sugar uplift exceeds 12 000 km². For example, in the Buzios oilfield whose tectonic background is a fault-uplift, coquina, spherulite and stromatolite reservoirs cover about 800 km², with an average thickness of about 200 m and a maximum thickness of 500 m. Fault horst and volcanic high are also important positive paleogeomorphic units for the development of carbonate reservoir, mainly distributed in the central depression and western depression, and characterized by independent structures. Carbonate sedimentation is characterized by accumulation. The area of carbonate reservoir is usually less than 200 km², and the thickness varies from tens of meters to 400 m, with large lateral variation.

In the pre-salt lacustrine carbonate intervals, three types of medium porosity and medium permeability reservoirs are developed, including coquinas, spherulite and stromatolite. The reservoir storage space is mainly composed of primary pores which are almost intergranular pores, intragranular pores and growth-framework pores. The average porosity is 8%-15%, and the maximum porosity may exceed 25%. The average permeability ranges from tens to hundreds of millidarcies (Fig. 4). The average net-to-gross ratio is about 62%, and the highest ratio may reach 95%. The distribution of the reservoirs is generally controlled by the sedimentary facies, but the burial diagenesis is relatively weak.

The salt rocks of the Lower Cretaceous Ariri Formation, which have large thickness and strong sealing capacity, are widely distributed in the Santos Basin. The salt rock is distributed in the NE-SW direction, about 650 km long from north to south, and 380 km wide from east to west. The thick salt rocks are mainly distributed in the central depression, eastern depression, the south and central sections of Lula-Sugar uplift and Aram-Uirapuru uplift, with a thickness of hundreds of meters to 4500 m and continuous lateral spread, without the development of salt windows. The thick salt rocks directly cover the carbonate rocks and become efficient regional cap rocks. Oil and gas are effectively preserved and sealed in the pre-salt traps. At present, all the large oil and gas fields discovered under the salt interval are covered by thick salt rocks, such as Lula oilfield. However, in the western depression and the northern section of the Aram-Uirapuru uplift, the salt rocks are obviously thinner or missing, with many salt windows developed. Oil and gas migrate to the post-salt interval through the salt windows and salt-related faults to form reservoirs, such as Mexilh oil and gas field.

Continuous thick salt layers have an influence on the thermal evolution of hydrocarbon source rocks and reservoir diagenesis. Firstly, due to the high thermal conductivity of salt rocks, the temperature below the thick salt layer is generally lower. When the salt layer reaches 2000 m thick, the temperature of the hydrocarbon source rock below the salt layer may decrease by 25.1 °C, and the R_o of the source rock may be lower than that of the area without salt by 0.6% ^[33], which delays the thermal evolution of the hydrocarbon source rock, and it is beneficial for late-stage oil generation and long-term oil charging. Secondly, due to the strong plasticity, strong sealing ability and high thermal conductivity of the ultra thick salt rock layer, the diagenesis of the pre-salt strata is significantly weakened, which enables primary reservoir pores to be retained ^[34]. At present, pre-salt carbonate reservoirs in the basin are in the early-middle diagenetic stage, with weak cementation and good reservoir physical properties.

Currently, several oil and gas fields with high-CO₂- content gas cap have been discovered in the pre-salt strata of the Santos Basin, such as the Jupiter reservoir, Abare West reservoir, Central Libra reservoir and Peroba reservoir, with CO₂ content ranging from 67% to 96%, δ¹³C_CO2 ranging from −8.11‰ to −5.35‰, and helium isotope R/Ra of 5.6-8.8. According to the discrimination diagram of geochemical parameters established by Dai et al. ^[35], the CO₂ in the Santos Basin is mainly of inorganic-mantle origin.

The distribution of CO₂ content in the Santos Basin is controlled by crustal thinning and through-going faults, and generally has the characteristics of “high in the east and low in the west” (Fig. 5). Reservoirs with high CO₂ content are all located near the large strike-slip pull-apart fault in the east of the basin, and the fault throw of its adjacent faults in the later stage is up to 1000-3000 m, indicating a strong late-stage tectonic activity. These faults are all located in the thinned crust area, providing channels for mantle CO₂, resulting in strong CO₂ charging and forming a gas cap. On the contrary, in the western part of the basin, the fault throw of the uplift-controlled fault is less than 1000 m, indicating that its activity was significantly weaker than that in the eastern uplift zone of the basin. The tectonic activity was dominated by the subsidence of the passive continental margin, with large crustal thickness and low CO₂ content.

A new understanding is proposed that the CO₂ charging is later than oil and gas charging. Due to the strong fluorescence of crude oil, which can shield the Raman spectrum signal of bubbles and make the bubble components of inclusions difficult to be analyzed and restrict the determination of CO₂ charging time in the Santos Basin. Based on the improved experimental process of detecting the gas phase of hydrocarbon inclusions by laser Raman spectroscopy, it was found that only CH₄ was detected in the early oil and gas inclusions, and CH₄ and CO₂ were detected in the late salt water inclusions. Based on this, it is proposed the new understanding of CO₂ charging later than oil and gas, which breaks the traditional understanding of "CO₂ first and then oil" based on basin modeling by the operators, and indicates that the early volcanic activity contributed limitedly to CO₂ charging. The discovered oil and gas fields also confirm that the correlation between CO₂ content and early volcanic activity is not close.

Currently, the petroleum reservoirs formed by CO₂ charging in the Santos Basin can be roughly divided into three types: oil reservoirs with dissolved CO₂, oil reservoirs with CO₂ gas-cap, and CO₂ reservoirs. Based on numerical simulation and physical experiments of CO₂ charging, it is found that the size of oil rim is negatively proportional to the amount of CO₂ charging, and positively correlated with GOR ^[36]. (1) In the early stage of CO₂ charging, CO₂ is mainly dissolved in oil to form CO₂-dissolved oil reservoirs, with a high GOR and CO₂ content of generally less than 58%. (2) In the middle stage of CO₂ charging, CO₂ extracts light components from oil to form a gas-cap oil reservoir with high-content CO₂. The volume of the oil rim decreases by 5.7%-6.0% when CO₂ increasing by 10%, and the CO₂ content is 58%-84%. (3) In the late stage, CO₂ almost occupies reservoir space and form CO₂ reservoirs. With more and more CO₂ filling, the volume of oil rim begins to decrease significantly. And with the decrease of light components in the oil and the extracting capacity of CO₂ in the gas cap reaching the limit, the exchange at the gas-liquid interface begins to be weak gradually, and the CO₂ content is generally higher than 84%.

Based on numerical simulation and physical experiment, it is predicted that the oil rim in the Central Libra block is 18%-20%, and that in the Peroba block is 4%-5%. The research results provide important technical supports for the exploration plan of the Santos Basin, so CNPC decisively abandoned the Peroba block and successfully acquired the risk exploration area of Aram block.

The salt domes of the Santos Basin are complex in shape and different in thickness. Multiple phases of gypsum rock were mixed in the process of salt rock deformation, resulting in large spatial differences in seismic velocity in the salt domes (the maximum difference up to 2000 m/s). In the velocity model for pre-stack depth migration, the salt layer always uses relatively uniform seismic velocity. The velocity error of the salt layer has an impact on the precision of the pre-salt structural map, leading to a depth error up to hundreds of meters. This poses challenges and operational risks to the evaluation of exploration targets, the deployment of well positions, and the forecasting of drilling engineering.

This study has innovated a seismic velocity modeling technology for the complex salt domes with the joint control of strata and facies. By utilizing the lithology information from mud logging data, numerical statistics of logging velocity, seismic facies identification, seismic interpretation and other data, various types of salt dome lithology, seismic facies and seismic velocity models were established. Based on the original seismic velocity model, the spatial distribution of the salt layer was constrained by seismic horizon interpretation, and the seismic velocity was constrained by the lithology-seismic facies-velocity model. The seismic velocity model reconstructed for the complex salt domes has effectively eliminated the influence of the velocity difference of thick salt layer.

This technology has drastically improved the precision of variable-velocity mapping of the deep-water pre-salt target layer in Brazil. The depth prediction error of the target layer has been reduced from 72-250 m to 4-35 m. It provides powerful technical supports for reserve evaluation and drilling engineering.

The pre-salt lacustrine carbonate reservoirs in the Santos Basin have the characteristics of deep burial, complex lithology, multi-phase migration, strong heterogeneity, and distinct seismic responses, making it difficult to accurately predict by using a simple conventional reservoir prediction method. This study has integrated a series of lacustrine carbonate reservoir prediction technologies based on the combination of seismic facies, highlights and pre-stack elastic inversion. (1) Lacustrine carbonate reservoir prediction based on seismic facies. There is a certain correspondence between seismic facies units and geological facies units. As lithology changes, some physical parameters of reflected waves will vary. Seismic facies reflects the characteristics of lithology to a certain extent, thus unifying the units with similar seismic-stratigraphic parameters in the same seismic sequence into the same seismic facies (Table 2). First, by analyzing the lithology and seismic facies, a facies model is established. Then the mound morphology of grainstone and coquinas is identified and characterized by attribute fusion, texture attributes, and clustering analysis. By comparing with similar geological bodies, the associated attributes of the geological body are confirmed, and the seismic recognition model of different geological bodies is established. This eliminates the ambiguity of seismic phenomena, improves interpretation accuracy and effectively recognizes the carbonate shoal and its distribution ^[20]. (2) Carbonate reservoir prediction based on highlights. Highlights are mainly used to identify the abnormal amplitude of the reflected wave at dominant frequency, and reflect the difference between the peak and the average amplitude in an effective frequency band. After comprehensive analysis of the seismic characteristics of the carbonate shoal, the highlight technology combining amplitude and frequency was adopted (Fig. 6). This further improves the distinction between limestone and tight limestone reservoirs. It enables the accurate delineation of the target shoal boundaries and multi-episode features ^{[19,21 -22]}. (3) Reservoir prediction based on multi-parameter pre-stack inversion. Rock physics analysis reveals that a single elastic parameter can differentiate igneous rock to a certain extent, but it is hard to effectively distinguish carbonate reservoir from non-reservoir. Multi-parameter cross-plot analysis indicates that P-wave and S-wave impedances, density and GR values can be utilized to better differentiate reservoirs. Consequently, pre-stack elastic inversion, waveform indication inversion, pre-stack Bayesian inversion and other technologies are applied for multi-elastic-parameter quantitative prediction of carbonate reservoirs. Especially based on a FWI velocity model, wave impedance inversion has remarkable improvement on the deep high-impedance layer and the low-impedance layer under igneous rock, and wave impedance continuity of limestone reservoir. The effect of igneous rock has been effectively eliminated, thus improving the accuracy of reservoir prediction. The pre-salt carbonate reservoirs in Libra and Aram blocks have been successfully predicted, and the prediction accuracy in complex lithology zones has been increased from 60% to 85%, providing an effective approach for well location and reserve evaluation ^{[10,12,26⇓ -28]}.

Two types of pre-salt igneous rocks, intrusive rock (diabase) and effusive rock (basalt), are developed in the Santos Basin ^[37]. The intrusive rock shows strong amplitude, low frequency and parallel continuous reflections on seismic sections, and high velocity and high impedance on logging curves. The intrusive rock is different from carbonate reservoirs, but it has similar lithological characteristics as tight mudstone, calcium/siliceous mudstone. This leads to multiple solutions to recognition and prediction. Through deep analysis of amplitude, frequency and wave impedance characteristics, an intrusive rock identification combination technology featured by strong amplitude, low frequency and high impedance was developed to achieve accurate prediction of intrusive rock (Fig. 7). The prediction coincidence rate is currently 100% and the thickness error is less than 3 m. Because of multi-phase eruption, small velocity difference from carbonate rock, various types of seismic facies and complex spatial distribution, seismic prediction is difficult for effusive rock. In this study, a seismic facies combination identification model and pre-stack elastic inversion technology under the guidance of underwater volcanic mechanism model was developed. It effectively improves effusive rock prediction by increasing the prediction coincidence rate up to 86%. This technology provides a powerful technical support for distribution prediction, optimization of well locations and reserve evaluation of pre-salt igneous reservoirs in Brazil.

Considering the difficulty of hydrocarbon detection in carbonate reservoirs, a series of hydrocarbon detection technologies for lacustrine carbonates have been integrated based on spectral decomposition, AVO and pre-stack dual parameters inversion. (1) Hydrocarbon detection based on spectral decomposition. Research on fluid substitution and forward modeling in carbonate reservoirs with different fluids and saturations has revealed that with the increase of oil saturation, the reflection amplitude at the top of the oil layer would be enhanced. Well-seismic analysis has shown that the top of the oil layer is characterized by strong amplitude and a large attenuation gradient. The hydrocarbon-rich reservoir follows a dynamic principle of “low-frequency resonance and high-frequency attenuation”. After analyzing and comparing the seismic amplitudes of different frequency components by spectral decomposition to optimize the dominant frequency component for seismic hydrocarbon detection (Fig. 8), this study has demonstrated that the hydrocarbon responses of high-frequency and low-frequency components are stronger in the hydrocarbon-bearing zone ^[24,29]. (2) Hydrocarbon detection based on AVO gradient. The AVO characteristics of different fluids in carbonate reservoirs are similar, making it challenging to differentiate between them. Exploring and analyzing the AVO attribute of oil-bearing and water-bearing reservoirs reveals that the gradients of different fluids are similar, but the absolute AVO intercept is greater when the reservoir contains oil. Through fluid substitution and forward modeling analysis of well data, a fluid identification chart based on absolute AVO intercept is established. The intercept sensitive to oil layers is used for hydrocarbon detection in carbonate reservoirs ^[30]. (3) Hydrocarbon detection based on pre-stack elastic dual parameters. This method involves the analysis of petrophysics and seismic forward modeling to optimize the P-wave impedance and density which can identify different fluids in carbonate reservoirs. In this study, a dual-parameters fluid identification chart was established, and pre-stack P-wave impedance and density dual-parameters inversion was effectively used for hydrocarbon detection. The prediction accuracy of this method can be 85%. (4) To overcome the difficulty of seismic identification of supercritical CO₂ fluid in carbonate reservoirs, pre-stack and post-stack seismic prediction technologies for supercritical CO₂ fluid have been developed. Research on pre-stack frequency dispersion AVO gradient attributes has shown that the CO₂ gradient of pre-stack frequency dispersion AVO is larger, while the oil and water gradient of pre-stack frequency dispersion AVO is relatively small. By considering peak frequency and absorption attenuation, this technology can qualitatively predict the reservoir fluids in different structural parts. It has been successfully used to predict the CO₂ caps and oil rims in the central part of the Libra block and the Peroba block in Brazil, as well as the CO₂ content in the Aram block. It provides important technical supports for target evaluation, well deployment, and exploration strategy ^[31].

In view of the complex lithology of carbonate rock, intrusive rock and effusive rock in Brazil, and the difficulty of logging identification, a complex lithology identification technology for lacustrine carbonate rock has been developed based on "conventional logging data+ element capture logging data + artificial intelligence" through calibrating logging, whole core, sidewall core, thinsection data in three steps: (1) Classify carbonate rock and igneous rock by element capture logging data; (2) Identify basalt and diabase in igneous rock by element capture logging data and gamma spectrometry logging data ^[38-39], and identify igneous rock by crossplot and new lithology index parameters; (3) Select the sensitive lithology curve to identify carbonate rock in different intervals through intelligent fuzzy clustering identification. The coincidence rate of identifying igneous rock and carbonate rock is up to 95%, the coincidence rate of identifying basalt and diabase in igneous rock is up to 86%, and the coincidence rate of identifying lithology in carbonate rock is up to 70%. This technology lays a solid foundation for logging evaluation of pre-salt carbonate reservoirs and investigating igneous rock in Brazil.

The coexistence of different fluids (i.e., oil, gas, water and CO₂) in the pre-salt carbonate reservoirs makes it difficult for logging identification. Therefore, a semi- quantitative evaluation and fluid identification technology for supercritical CO₂ was developed based on conventional logging data, NMR while drilling data (NMRWD), wireline NMR logging data and formation test data. It makes full use of the physical properties of CO₂ without hydrogen atoms and its density in supercritical state is comparable to that of oil ^[40]. A semi-quantitative CO₂ identification chart was established by drawing a neutron-density crossplot and analyzing the difference between total wireline NMR porosity and total NMRWD porosity. First the semi-quantitative CO₂ evaluation limits were set to be 0, 40%, 60%, 80%, 100% (the blue dotted line in Fig. 9). Then the fluid type was further evaluated by CO₂ content ^[41]. Based on the above techniques, this paper carried out a semi-quantitative evaluation of CO₂ and fluid type identification based on 15 wells and 42 intervals in the Santos Basin. Comparing the identified results with the measured data, it was found that the coincidence rates of semi-quantitative evaluation of CO₂ and fluid identification reached 73% and 80% by using the neutron-density crossplot. This technique provides a new method for logging fluid identification, especially for supercritical CO₂ identification and evaluation of pre-salt carbonate reservoirs in the Santos Basin, Brazil.