Introduction

Seismic inversion is a technique that facilitates the retrieval of underlying models of the physical properties of rocks and fluids from seismic and well data (Bashir et al. 2019). The integration of well-log properties into seismic data is crucial for enhancing data bandwidth and resolution in seismic inversion (Babasafari et al. 2020, 2021). The well logs are calibrated by considering intervals instead of boundaries in these data. Throughout this process, the wavelet effect is removed, and the lithology, porefill, and porosity are identified, making interpretation easier and more accurate. Seismic inversion uses seismic data to generate acoustic impedance, interval properties generated by the seismic inversion are used to define the geological interpretation of the rock thickness. The impedance generated is a function of sonic velocity and bulk density for petrophysical properties (Mahgoub et al. 2022; Shan et al. 2023; Dally et al. 2023; Sharifi et al. 2023). Seismic inversion is a commonly employed technique within the oil and gas sector for the purpose of detecting oil and/or gas deposits located beneath the earth’s surface (Abdolahi et al. 2022; El-Nikhely et al. 2022; Riahi and Fakhari 2022). Seismic inversion is highly valued by geophysicists due to its ability to simplify the interpretation of lithological and fluid-related impacts. Additionally, it provides clarity on stratigraphic relationships (Simm and Bacon 2014).

Several models have been published and described for sedimentary and carbonate rocks. These models rely on the propagation of elastic waves in two phases with restricted porosity (Tatham 1982). Geological models that incorporate pore geometry and aspect ratio are subject to limitations in their ability to accommodate the highest possible porosity values (Shoukat et al. 2023). The aforementioned constraint arises from the premise that there is no interaction between individual pores or fractures. Several studies and research works (Mavko 1979; Winkler and Nur 1979; Slater 2007; Grana and Della Rossa 2010) have considered the impact of fluid saturation on seismic wave attenuation while assessing the aspect ratio and physical shape of cracks. The factors mentioned above hold significance with respect to fluid motion and potentially permeability. However, they exceed the boundaries of the present case study.

Several theoretical frameworks have been proposed to explain the propagation of elastic waves in sedimentary rocks with finite porosity. The model of Biot (Biot 1962a, b), of which a portion has been conveniently summarised and organised by Geertsma and Smit, is perhaps the most successful of these (Geertsma and Smit 1961). The current model incorporates pore and crack geometry into the discussion, but its applicability is limited due to a weakness in Biot’s theory. The determination of gas reserves heavily relies on two fundamental parameters of reservoirs, namely porosity and water saturation (Iqbal and Rezaee 2020; Ali et al. 2021; Bashir et al. 2022).

Seismic interpretation is an approach that involves the identification of horizons and the utilisation of seismic attributes such as envelope and sweetness to identify potential areas of gas accumulation. The primary focus of this process is on the accurate picking of horizons and the application of these attributes to highlight areas of interest. The research utilised the model-based inversion (MBI), which is a deterministic inversion technique. MBI is a widely used inversion approach that relies on convolution (Russell 1988). The technique uses iterative forward modelling to calculate synthetic seismic data as part of the inversion algorithm (Jafari et al. 2017). This procedure requires a starting/initial model, which is then perturbed and compared to the seismic data. The initial model can be an interpolation of well data or a general trend model based on geological data (Simm and Bacon 2014). MBI is used in this study in conjunction with well data to estimate porosity in the gas reservoir at X field, Central Luconia, offshore Sarawak. Based on a 3D seismic survey, MBI is used to develop a relationship between AI and porosity for reservoir characterisation (Bashir et al. 2019). The geographical location of central Luconia province in the Sarawak Basin is depicted in Fig. 1. This region is bounded by North Luconia, West Luconia, the West Baram Delta, and the Balingian region (Janjuhah et al. 2018). It is situated in close proximity to Borneo and Sarawak state in Malaysia. The selected study area is located in Central Luconia, a region characterised by significant carbonate build-ups and platforms (Ali et al. 2021; Shoukat et al. 2023). The focus of the study is to investigate the porosity and hydrocarbon occurrence in one of the carbonate platforms within this area.

Fig. 1
figure 1

Geographical map of the Luconia province (Borneo). The field of study is located in Central Luconia

Full size image

Geology and stratigraphy of the study area

The formation of the Central Luconia carbonate platform occurred on a regional structural high that is bounded by a fault. This is due to the area being structurally isolated from the supply of clastic materials. The Luconia platform is delimited by the current shelf edge to the north, and by the geological provinces of Tatau Horst and Graben, Balingian, and Baram Delta to the west, south, and east, respectively (Madon 1999). The Middle to Late Miocene witnessed a major period of carbonate deposition that shows a relatively low degree of deformation, indicating a stable basement and weak tectonic activity (Janjuhah et al. 2018; Imran et al. 2021). The growth of carbonate platform with the reefal body got initiated on pre-existing structural highs formed during Late Oligocene rifting as the area was free from the clastic influx (Ali et al. 2021; Bashir et al. 2021). The subsidence effects of both post-Miocene tectonic activities and eustatic sea-level rise have resulted in the ultimate demise of carbonate deposition in the Central Luconia platform included in the study area. The analysis of seismic data revealed that the cessation of carbonate development in the investigated area occurred very early. Carbonate build-ups on the Luconia platform are short-lived in the south compared to the north, which can be explained by clastic sediment inflow in the basin and its provenance. In general, high-energy carbonates with excellent reservoir characteristics tend to accumulate in the outer regions of carbonate build-ups. Conversely, the central portions of these build-ups typically contain carbonates with inferior reservoir quality that were deposited under low-energy conditions.

Off the coast of Sarawak, Malaysia, the Central Luconia province comprises a shelf that is 200–300 km in width. Regional extension and seafloor spreading in the northern South China Sea during the Oligocene shaped the province, which is submerged between 60 and 140 m of water (Koša 2015; Rankey et al. 2019). The region of Central Luconia is characterised by the prevalence of siliciclastic sediments. The clastic strata in this area are relatively thinner and flatter and are not subject to significant gravity-driven deformation. Also, the sediments of the area are dominated by prograding clinoforms, which are developed from the initial subdivision of stratigraphy into the “regressive cycles separated by major transgressions” (Doust 1981; Hageman 1987). Each cycle is initiated with a transgressive basal layer that follows the regressive unit. The mentioned nonconformity segment of the geological cycle subsequently undergoes a transformation into the superimposed foundational layer of sand (Mat-Zin and Tucker 1999). These cycles, numbered I through VIII, span the time period from the Eocene to the present and can be further subdivided into higher-frequency cycles at different scales, as noted by Koša (2015). Studies by Krebs and Van Vliet (2009) show that carbonates are present in all levels of the basin’s stratigraphy, with the Middle Miocene reportedly experiencing major deposition of carbonates as a result of regional extensional events that occurred during cycle III.

The carbonate strata demonstrate high-frequency variations across multiple scales. Typically, these variations consist of argillaceous transgressive carbonate facies that are influenced by echinoids, fragments, rhodolites, and lime mud. These facies are then overlain by cleaner highstand carbonate facies, which are primarily composed of coral fragments and are capped by an exposure surface (Koša 2015). Primary-facies variations that are characteristic of Cenozoic tropical carbonate build-ups are detected horizontally within the cycles. The build-ups are composed of fringing grainstones and coral framestones. Rudstones are prevalent in the high-energy edges, while wackestone-mudstone facies are found in the interiors. These facies are deposited within protected lagoons.

Exploration history of the study area

Well-1 was first discovered in the early 1970s in the vicinity of Bintulu, located in the Sarawak Basin. The geological formation of the field is characterised by a carbonate build-up platform. The gas reservoir covers an area of approximately 22 square kilometres. The geological formation observed in the field is primarily composed of limestone deposits belonging to Cycle IV/V, which are afterwards covered by clastic sediments of Cycle VI. The production of the field commenced in the early 1980s, with an original hydrocarbon generation interval of 1100 ft thickness. The initial reservoir pressure was 2445 psia at datum at 4500 ft TVDss. By the end of 2014, the field experienced a depletion of around 2000 PSI. The average porosity in the field is 23%, and a well test revealed a permeability range of 100–400 mD.

The field’s initial interpretation and subsequent reservoir modelling studies were based on 2D seismic data. A 660 square kilometre 3D dataset, with a full-fold coverage, was acquired between 2001 and 2002. The dataset covers the Well-1 fields and other fields, as well as some potential exploration prospects. This study (seismic interpretation and inversion) was based on the 3D seismic data that was processed in 2002. The quality of the data is generally high, with the exception of the data points located around and below the platform. The 3D seismic data was utilised to update the two-way time horizons. A total of five seismic surveys were carried out in the field, comprising four sets of 2D seismic acquisition and a 3D dataset. Figure 2 displays the stratigraphic composition of the Sarawak shelf and certain tectonic occurrences in the surrounding area.

Fig. 2
figure 2

Summary for the stratigraphy of the Sarawak continental shelf basin (Madon 1999)

Full size image

Methodology and data optimisation

Seismic interpretation

The process of seismic interpretation is straightforward, but it can become intricate when subsurface complexity, such as the presence of carbonate buildup, is involved. In such scenarios, enhancing and augmenting the precision of subsurface geology is a skill that requires expertise and ability. The reliability of an interpreter can be enhanced by utilising a conceptual model, in addition to the interpreter’s skill. Prior to commencing the interpretation, an overview of the regional geology of the study area was conducted. Horizons are geological boundaries that exhibit variations in rock properties between two adjacent rock layers located beneath the Earth’s surface. These variations primarily manifest as changes in acoustic properties such as density and velocity. Seismic horizon interpretation is the process of identifying and monitoring seismic reflectors that are horizontally consistent. The selected horizons for this project are the upper and lower boundaries of the carbonate layer, which can be identified by a negative (red/brown) loop on the cropped seismic data.

Seismic inversion

Seismic inversion is an approach that aims to convert a seismic volume into a volume of acoustic impedance, which essentially involves transforming seismic data back into geological information (Babasafari et al. 2020, 2021). The seismic inversion deal with involving the generation of acoustic impedance contrast is illustrated in Fig. 3. This is achieved by convolving a seismic trace with a wavelet and reflection coefficient (RC) using a recursive formula or iterative modelling scheme. The wavelet utilised in our methodology is produced through vibriosis (Land) or Air gun (Marine), while the reflectivity pertains to the earth layer. By convolving the wave and reflectivity, the seismic trace for acoustic impedance can be obtained, as illustrated in Fig. 3. The modelling and inversion processes are reversible, with modelling proceeding from left to right and inversion proceeding from right to left. The input and output remain unchanged throughout.

Fig. 3
figure 3

The seismic trace modelling to inversion and vice versa

Full size image

In order to ensure the effectiveness of the MBI method, it is necessary to filter the log impedance curve to exclusively preserve low frequencies. This is essential for the recovery of information that may have been lost during the seismic survey, particularly at very low frequencies. The model-based algorithm commences by utilising the well’s impedance subsequent to its low-pass filtration (Russell 2016). The MBI technique relies on the convolution theory, which posits that a seismic trace is the outcome of a wavelet convolved with reflectivity and noise (Maurya and Sarkar 2016).

$$ S = W * R + {\text{Noise}} $$
(1)

Convolution of the reflectivity with the source wavelet results in the elimination of a significant portion of the high-frequency information. If the noise in the data is uncorrelated with the seismic signal, it is possible to solve the trace using the earth reflectivity function. The determination of the zero-offset RC is based on the nonlinear equation, which takes into account the variations in acoustic impedance between layers.

$$ {\text{Acoustic impedance}}\;\left( {Z_{p} } \right) = {\text{density}} * {\text{sonic}}\;{\text{velocity}} $$
$$ R_{i} = \frac{{z_{i + 1} - z_{i} }}{{z_{i + 1} + z_{{\dot{i}}} }} $$
(2)

where \({R}_{i}\) is the zero-offset P-wave RC at the i-th interface and \({z}_{i}\) is the seismic impedance of the i-th layer.

Workflow

The workflow comprises of five major steps as illustrated in Fig. 4.

  1. a)

    Data import (related to wells, seismic information, and horizons).

  2. b)

    Well to seismic correlation and wavelet extraction.

  3. c)

    Generation of the initial model.

  4. d)

    Inversion analysis.

  5. e)

    Perform inversion and apply to volume.

  6. f)

    Porosity and lithology volume.

Fig. 4
figure 4

Model-base inversion (MBI) workflow used in the study

Full size image

The methodology of seismic inversion involves the conversion of a singular 3D seismic data volume into acoustic impedance. This is achieved through the integration of various input data, including seismic data, well data, and a fundamental stratigraphic interpretation (horizon), as depicted in Fig. 4. Seismic inversion offers various benefits, such as acoustic impedance, wavelet effect reduction, tuning, and side lobe suppression. It enables the correlation of layer properties with reservoir properties and impedance with porosity.

Data input

The present research (seismic inversion) is based on high-quality 3D seismic data, excluding those located around and beneath the platform. The inversion data comprises of 460 inlines and 401 crosslines, with a trace spacing of 25 m in the inline direction and 12.5 m in the cross-line direction. The sample rate used is 3 ms. The study utilised two interpreted horizons, namely Carb_T (top of carbonate) and Carb_B (base of carbonate). The geophysical logs, including calliper, density, gamma-ray, resistivity, neutron porosity, and P-wave, are utilised with satisfactory overall quality (Table 1).

Table 1 Wavelet parameters used in generating wavelet
Full size table

Results and discussion

Seismic interpretation

The main objective of seismic interpretation is to identify the upper and lower boundaries of the carbonate reflector or horizon. This boundary signifies the transition between distinct rock formations. Thorough investigation of stratigraphy, structure, and depositional setting is a prerequisite for any interpretation. A map of the carbonate platform was generated using seismic reflection data. The map indicated the presence of Carb_top and Carb_base at the top and base of the platform, respectively (refer to Fig. 5a). In the study area, the green-coloured horizon denotes the uppermost layer, while the yellow-coloured horizon signifies the carbonate base. Depositional features were identified during the study, as depicted in Fig. 5b. These features include geological characteristics observed on the seismic reflection, such as unconformity, onlap, and toplap.

Fig. 5
figure 5

a Interpreted horizons on top of carbonate (green) and at the base of carbonate (yellow). These two reflectors/horizons are the input for seismic inversion and b Carbonate platform margin (top & base) and its seismic stratigraphic features (onlap & off-lap)

Full size image

The amplitude is a geophysical response that corresponds to various reservoir properties during prospect evaluation and geophysical analysis. The study has demonstrated that amplitude and amplitude versus offset (AVO) are effective in specific geological and petrophysical scenarios, particularly in cases where the rocks are relatively young in geologic age (Tertiary) and unconsolidated. The awareness of data polarity is a crucial aspect that interpreters must consider. In general, wet sands or sands containing hydrocarbons exhibit a softer consistency compared to the adjacent shale, resulting in a pronounced negative or trough response. This phenomenon is commonly referred to as a bright spot or direct hydrocarbon indicator (DHI). Regrettably, there exist numerous geologic and lithologic scenarios that result in the adverse outcome described below.

  • Hard/soft shale.

  • Coal layers.

  • High-quality brine sand.

The issue mentioned is able to be resolved by examining the AVO domain, as depicted in Fig. 6a. Under optimal conditions, when dealing with reservoir sand, it is expected to observe a phenomenon known as “rising AVO”. This refers to a negative soft impedance that becomes increasingly negative and softer with offset. It is important to note that this is in contrast to the expected falling AVO that is typically observed in lithology cases.

Fig. 6
figure 6

a A direct hydrocarbon indicator (DHI) or flat-spot in the reservoir, b envelope cross-line (a highlighted area shows amplitude anomalies—possible HC) indicating the abrupt termination on AVO envelope attribute in a gas reservoir in Sarawak basin, c seismic time slice at − 1250 having amplitude anomaly due to gas effect (left: original seismic & right: sweetness), d RMS amplitude inline (a highlighted area shows amplitude anomalies—possible HC), and e RMS amplitude applied to the carbonate top (high-amplitude anomalies shown in green–red potential hydrocarbon bodies)

Full size image

Multi-attribute analysis techniques, such as incorporating seismic amplitude attributes like envelop, sweetness, and root-mean-square (RMS), have been demonstrated to aid in interpretation and enhance other methods like 3D visualisation interpretation. Various specialised processing techniques, such as AVO and seismic inversion, are beneficial in distinguishing hydrocarbon fluids and their accumulations. The implementation of this technology in Malaysia has yielded several achievements, particularly in the detection of stratigraphic traps and the outlining of channel features for the purpose of discovering new hydrocarbon deposits. The deep water exploration basin has identified various features, including floor fan deposits, that assist in defining channel extent and reducing geological risks in most prospects. This is similar to the shallow water exploration basin.

Seismic attributes

Familiarity with the sequence stratigraphic framework is necessary to comprehend seismic characteristics. Depositional geometries can be identified by visually analysing maps of attributes. Additionally, attributes for series can be extracted by evaluating seismic data. This approach involves utilising seismic properties to identify oil or gas reservoirs through two distinct techniques. In order to begin, a visual examination of the seismic properties is performed, which entails mapping the attributes and searching for patterns that correspond to our knowledge of the depositional strata. A qualitative understanding of seismic property anomalies can be formulated at this point. The alternative method involves forward modelling of the seismic properties by referring to well-log data and then calibrating the attributes to the geology thicknesses in the model. The method puts a numerical cap on how abnormalities in attributes should be interpreted.

Envelop, sweetness, and RMS were among the characteristics we looked at to make sure the findings were accurate. Tracking down the trace as can be seen in Fig. 6b, the amplitude and oscillatory function of the data are represented by the envelop attribute, which is also known as the Instantaneous amplitude attribute. The circle in Fig. 6b depicts how the data are distributed based on amplitude, with higher amplitudes having a higher probability of hydrocarbon generation. It is possible to tell a bright spot from a gas accumulation or an unconformity from a lithological shift or a spatial correlation in porosity thanks to this feature. The envelope seismic attribute (namely, reflection strength) displays bright spots on both negative and positive events and is the most popular trace attribute (Koson et al. 2014). It is useful for highlighting discontinuities, changes in lithology, tuning effects, and sequence boundaries (Marfurt, Kirlin and Gersztenkorn 1999). The property does not depend on the reflection’s phase or polarity, which could otherwise alter its apparent intensity. The envelope seismic attribute is useful for pinpointing regions of high and low acoustic impedance, which can be indicative of gas accumulation and/or bright spots. Strong envelope values, as shown in Fig. 6b, are likely indicative of gas accumulation near the top of the carbonate platform.

As can be seen in Fig. 6c, the reservoir layer is highlighted by the sweetness seismic attribute, which is a time slice at a reservoir depth of 1250 ms, to emphasise the purity of the hydrocarbon accumulation. The original seismic survey (left in Fig. 6c), which reflects the interface properties within the reservoir, and the sweetness attribute results (right in Fig. 6c), where it reduces the contribution of the higher amplitude and lower frequency events to such data as the sandy interval with the highest value, are shown below. The yellow colour represents the higher frequency and the blue the lower frequency in the data.

The calculation of this attribute involves the division of either the instantaneous amplitude or the amplitude envelope. The detection of thick, clean reservoirs and their hydrocarbon content is facilitated by the utilisation of sweetness. According to some individuals, the calculation of Sweetness involves dividing the instantaneous amplitude by the square root of the instantaneous frequency (Ismail et al. 2019). High sweetness values can be found in regions with low amplitude and high frequency, and low sweetness values are found in regions with high amplitude and low frequency. Also, the sweetness attribute is applied to the seismic time slice at − 1250 (top of the carbonate platform). The original seismic survey (left) shows high amplitude in the carbonate platform while its sweetness attribute (right) displays amplitude anomalies which can thereby be interpreted as a result of the gas effect. These filed data are proven a gas field, which verifies the method accurately in determining the bright spot of a gas reservoir.

The RMS amplitude is an indicator used to quantify reflectivity within a specific time frame. It is calculated by taking the square root of the sum of squared amplitudes. The application of RMS amplitude is utilised to expose direct hydrocarbon indicators (DHIs) such as bright spots and amplitude anomalies in seismic data (Hamidi et al. 2018a, 2018b). It also serves to indicate the energy content of seismic data, as well as the porosity, and lithologic and stratigraphic variations. RMS amplitude attribute is applied to the post-stack data to compute the square root of the sum of squared amplitudes divided by the number of samples within the specified window as shown in Figs. 6d and 6e. Figure 6d is the inline cross section of the reservoir which shows a higher amplitude in the 2D section. Furthermore, the verification of the attribute is done by a time slice at the reservoir depth (Fig. 6e), which shows a higher RMS amplitude in green to red colour with a higher possibility of hydrocarbon generation.

Seismic inversion

The seismic inversion algorithm is utilised which involves the integration of seismic data and well logging. This is achieved through inverse modelling of the logs using seismic data (Bosch et al. 2009). The precision of seismic data inversion outcomes is reliant on various factors, including the seismic data and well-log resolution, the geologic interpretation/horizon (Satinder Chopra 2007), and the signal-to-noise ratio. In addition, a seismic wavelet is a source signature that is needed for the process of inverting seismic data, and a good wavelet is the core of inversion (Karim et al. 2016). In order to extract wavelets, seismic and well-log data are required. Seismic inversion of carbonate rock’s workflow is depicted in Fig. 4 of this study.

First, the wavelets are estimated using Well-1 as shown in Fig. 7 because the vertical well is measured by P-Sonic logging. The wavelet estimated from the traces around the well, and the time gate used are extracted from 1150 to 1800 ms. The RC at the location of the well is obtained using Eq. (2). The extraction of wavelets from the well and the synthetic seismogram is the most important process in the seismic inversion as shown in Fig. 7c. The first step in the seismic inversion was to correlate the well logs and the seismic data, then to calibrate them by generating a synthetic seismogram (Fig. 7). In this case, a synthetic seismogram was generated by first converting the acoustic impedance from depth to time using check shot data, then convolving the reflectivity with the appropriate wavelet. The appropriate wavelet for a well is obtained iteratively, by extracting different wavelets and selecting a wavelet that makes a strong correlation between the real and synthetic traces. The extracted wavelet must be a good approximation of the seismic wavelet to achieve a good correlation between the synthetic seismogram of the well logs and the seismic traces at a given well location. This method is distinguished by its utilisation of a low-frequency model of P as a starting point, which is then progressively disseminated until an optimal match is achieved between the seismic data and the synthetic trace. The process is iterated multiple times until a satisfactory correlation coefficient is achieved between the synthetic and actual seismograms at the Well-1 site. A seismic well tie is depicted as an example in Fig. 7c. In this case, a synthetic seismogram was generated using the wavelet properties depicted in Fig. 7a and b.

Fig. 7
figure 7

Wavelet extraction using well log which is zero-phase. a Wavelet time response; b wavelet amplitude and phase response, and c well-to-seismic correlation window

Full size image

Initial model

Next to the correlation of seismic data and well logging through the generation of a synthetic seismogram, an equation is applied to compute a low-frequency model of P-impedance. This results in the acquisition of two values, denoted as [1] and [2]. Zabihi Naeini and Exley (2017) have stated that the low-frequency model holds significant importance in seismic inversion, particularly in model-based inversion. Seismic data usually do not contain low-frequency data but well logs contain both low- and high-frequency data. The low-frequencies are important in seismic inversion to recover the absolute impedance. To determine the low-frequency cutoff point, several band-pass filters are applied to the seismic data in order to accurately estimate the absent frequency range (Karim et al. 2016). The initial model shows the missing lower and higher frequency contents in the seismic data later can be obtained when the well-log data are involved (Karim et al. 2016). Figure 8 displays the model generated for Well-1 in a seismic section. The low-frequency initial model is derived by applying a low pass filter to the acoustic impedance logs obtained from the wells. This model is then extrapolated along with the seismic data at each event, as illustrated in Fig. 8a. The model exhibits a clear acoustic impedance contrast, which distinguishes sedimentary and rock units. Specifically, unit 1 spans from 1023 to 1220 ms, unit 2 spans from 1220 to 1440 ms, unit 3 spans from 1440 to 1585 ms, and unit 4 spans from 1600 to 1833 ms. The carbonate platform (reservoir) falls into unit 2 and the acoustic impedance values range from 5600 to 7300 ((m/s)*(g/cc)).

Fig. 8
figure 8

a Initial model on the seismic section in the location of Well-1. The black well-log curve is the sonic log and yellow-green shows the top and base carbonate respectively and b inversion analysis. From left to right, column 1 (error in black, real log impedance in blue, and inversion impedance in red), column 3 (synthetic traces), column 4 (original seismic trace), and column 5 (error)

Full size image

Inversion analysis

The first step involves conducting inversion at the well site to ensure quality control and to enhance the inversion parameters. Figure 8b displays the results of the inversion analysis, indicating an overall error of 564.025. The correlation between the synthetic and original traces is 0.996, while the error, defined as the difference between the traces, is 0.081.

The MBI analysis results (Fig. 9a, b) show that the impedance in the zone of interest is in the range of 5500–9200 ((m/s)*(g/cc)), which has been proven with the well-log placement on the impedance section as a blind well test. Higher resolution impedance at lower frequencies has been derived from inverted seismic data, and all lithologies match. The section depicted in Fig. 9b can be utilised to make additional predictions regarding small reservoirs and to provide wellpoint recommendations to relevant industries. This is because there exists another reservoir between the carbonate top and base that falls outside the range of seismic reflection data interpretation. Impedance rises with depth, as predicted by the model. Porosity tends to decrease with depth in most cases. Figure 10 shows the well-log trace and the inverted trace plotted side by side on track 12 for verification purposes, and the negligible difference between the two traces demonstrates that the inversion results are accurate.

Fig. 9
figure 9

a Model-based inversion (MBI) impedance. The resolution of the data is increased as the impedance increases with the depth, and b QC blind well inserts of colour computed impedance (compared with the model the correlation is quite good)

Full size image
Fig. 10
figure 10

Export of impedance traces at well location (filtered well impedance, inversion impedance result exported from seismic track 19 & unfiltered impedance from seismic track 12)

Full size image

Porosity estimation

After identifying the potential locations of porosity, the seismic data is utilised to approximate additional petrophysical parameters such as permeability and lithology based on the well logs. The significance of seismic attributes in the analysis of reservoir properties has been highlighted by various researchers (Hampson et al. 2001; Tebo and Hart 2005; Khoshdel and Riahi 2007). Seismic attributes are defined through mathematical processing of seismic trace data, as is the case with any such processing. Seismic data can be utilised to model and characterise reservoirs, provided that there exists a credible correlation between seismic attributes and petrophysical properties. The current research utilised well-logging porosity, raw seismic data, and acoustic impedance obtained from inversion to forecast porosity. The following is an analysis of the data:

  • To quality check seismic and well logging data to select the appropriate attributes.

  • Using linear or nonlinear algorithms, extract a relationship between seismic attributes and well logs.

  • To create a volume of the log properties by applying the extracting relationship to a seismic volume.

The seismic sections underwent a comparative analysis of each step, and subsequently, the most optimal method was chosen for the purpose of porosity prediction. An inverse relation between porosity and impedance with depth is shown in Fig. 11a and b. In the plot, the highlighted rectangular area shows our area of interest with an impedance ranging between 5500 and 9200 ((m/s)*(g/cc)) which contains the porosity values ranging from 15 to 35% in total. Large variations in acoustic impedance from low to high value, related to the deposition, diagenetic and petrophysical characteristics of carbonate rocks, can be considered as an indication of variations in carbonate mineralogy. The predictive capability for petrophysical parameter extraction can be enhanced through the concurrent utilisation of a set of seismic attributes. (Khoshdel and Riahi 2007). Finding the best predictors of a target parameter can be done quickly and easily with the help of a multi-regression analysis (Marfurt et al. 1999).

Fig. 11
figure 11

a Cross-plot of porosity & P-impedance showing a characteristic decrease of porosity & increase of impedance with depth and b cross-plot of porosity & P-impedance showing porosity in the region highlighted in the range of 15–35%

Full size image

The red rectangle delineates regions proximal to the top of the carbonate platform, which may potentially serve as a reservoir. The porosity estimation within the highlighted zone, as indicated by the rectangle, varies between 10 and 35%. The identified zone exhibits significant promise for the presence of hydrocarbon reservoirs, making it a favourable location for oil and gas exploration. The research conducted will facilitate the development of fields for future drilling and reservoir modelling.

Conclusions

The research establishes a cohesive approach by integrating semi-quantitative and petrophysical analysis techniques and verifying them with seismic attributes studies. This approach allows for the prediction of porosity and P-impedance in field development. In addition, the current study utilised MBI seismic inversion methodology to improve reservoir modelling. The primary research findings are outlined below.

  • The lithology prediction was achieved through the utilisation of seismic inversion, attribute analysis, seismic interpretation, and well-log analysis.

  • The successful execution of this process necessitates the incorporation of a properly extracted wavelet, a stable low-frequency model, a precise depiction of the acoustic impedance, and the impedance-porosity relationship.

  • According to the analysis, it has been observed that there is a direct relationship between the depth below the reservoir and both impedance and porosity. Specifically, as the depth increases, the impedance also increases while the porosity decreases.

  • Lastly, based on research outcome, it is suggested that the zone characterised by impedance and porosity has the potential to function as a reservoir for hydrocarbons.

After analysing field production data, it has been decided that areas with a noticeable variation in impedance and a strong amplitude indicate a zone that is completely saturated with gas hydrate. Based on the data, it can be concluded that the porosity of these zones ranges from 10 to 35%.