The successful discovery of petroleum exploration primarily depends on the understanding of the basin evolution and sedimentary filling though geological time. Well data also play a key role for reservoir presence and quality analysis; however, none of well fully penetrated the Oligocene Syn-rift sequence in the West Arthit area. Therefore, this study aims to overcome the challenge of limited well information by performing the Forward Stratigraphic Modeling (FSM) to determine basin evolution, depositional setting, and reservoir distribution in this area. The FSM model is constructed with the inputs of paleo-bathymetry, subsidence, sediment supply, water level, and climatic cycle. In addition, the stratigraphic sequence is reproduced based on field observations such as rock samples, seismic mapping, well-log responses, and publications from nearby areas. The main uncertainty of building the FSM model is the initial age of rifting phase due to a lack of well penetration that fully covered the Syn-rift sequence and the limited biostratigraphic data. Therefore, two different age scenarios are examined in this study analogue from the age model as it was published in the Malay Basin locating to the south of study area. Once the FSM model was built, the last step was to calibrate the prediction result with the actual well result and the conventional seismic data to achieve the best accuracy and to increase the confidence on using the model. The FSM model was successfully reproduced the stratigraphic successions of the Syn-rift sequence in West Arthit area. The base case model was chosen from the age scenario of 27.0-23.1 Ma which exhibited four major cyclicities and matched with seismic mapping. The study area had two depocenters, one in the northwest and another one in the southeast. The northern sub-basin was deepened earlier during the first rifting phase whereas the southern sub-basin was subsided later after the second rifting period. With the increase in sedimentation rate and subsidence rate during the third rifting phase, both depocenters were shallowed up and then become a shallow lake covering the whole study area. The last lifting phase coincided with the thermal subsidence that occurred and affected across the region; therefore, the regional extensive lacustrine accumulated in the study area. The results from this study provided a crucial information on petroleum system especially depositional architecture, reservoir distribution, and potential source rock identification, which were incorporated into the planning of future exploration targeting in this field. This study demonstrates the new innovative approach to determine the basin evolution and to understand the variation on depositional setting in the study area with limited well data. This approach also creates the project value by supporting the planning of future exploration and development wells. Furthermore, this technique can be applied to all projects to increase the discovery rate and to add the field reserves.

High Carbon dioxide (CO2) content presents a serious challenge in the development of Arthit Field. Accurate resource estimation, especially in the deep reservoir sections (Lower Miocene - Oligocene), depends on the accuracy of CO2 prediction. Formulated as a manual clustering approximation, conventional CO2 prediction requires intensive labor and fails to re-calibrate the model once the latest information is acquired. This paper introduces the application of machine learning concepts to the prediction of CO2. The proposed CO2 prediction methodology leverages machine learning techniques to enhance the understanding of a known field with existing CO2 concerns. Compared to the conventional manual clustering method, the machine learning model improves accuracy and reduces time and cost in the process of CO2 prediction and, in turn, resource estimation. Although our methodology is demonstrated specifically for CO2 in Arthit Field, it is equally applicable to other parameters and fields.

Reservoir top structure interpretation through seismic usually associated with uncertainty especially when sparse number of wells are available. This leads to uncertainty in gross rock volume and oil-in-place calculation. It is quite common with reservoir top interpretation picking varies significantly from optimistic to pessimistic case especially at the point distant away from wellbore. This paper integrates the geosteering distance-to-boundary (DTB) calculation from horizontal well drilling to minimize the uncertainty window of top interpretation from surface seismic to deliver more accurate remaining oil-in place. During the geosteering operation, seismic horizon interpretation of reservoir target sand with multiple scenarios in offshore field was compared with the distance to boundary calculation that map the reservoir top structure with azimuthal resistivity tool. Prior to geosteer horizontal well in the oil-rim reservoir after long production, pilot hole was drilled to update the current fluid contact. Reservoir top interpretation is revised by including horizon top structure tied to well data from the pilot hole and existing wells in the area. The revised top structure which includes the pessimistic and optimistic case generally has significant difference in remaining OIP. Upon geosteering horizontal well, DTB calculation from azimuthal resistivity tool was capable to map the top reservoir boundary in real-time with reference from the seismic top interpretation from heel to toe of the lateral with good quality mapping of entire top reservoir structure. Distance to boundary (DTB) with good quality of top reservoir structure mapping throughout the section tracked closely to optimistic case of seismic top interpretation of target sand. Geosteering with azimuthal resistivity with RSS directional tool mapped the top of the reservoir on the upside and OWC from the downside up to 4.8 m away from the tool. As the top is expected to dip down and intersected the trajectory at multiple scenarios, the mapped boundary provided insights of gentle dipping trend, suggesting the top of reservoir could extend laterally longer than pessimistic case. With the result closer to optimistic case, the calculated remaining OIP volume calculated was on the optimistic side with significant difference, larger volume, from pessimistic case of OIP calculation. Integration of seismic top interpretation with geosteering distance to boundary application in lateral well could minimize the uncertainty in reservoir top horizon interpretation. The workflow enabled to place the well in the sweet spot while managing the uncertainty of top structure. With production from horizontal well, the well production rate is expected to increase to 2900 BOPD for initial oil production compare to below 1000 BOPD prior to horizontal well drilling. This workflow could be applied for future lateral well with high uncertainty of the target to place the well in sweetspot for field life extension and production optimization.