This project aimed to update the Geostatistical Model of a green oil field. It involved incorporating the geometric features of the clastic reservoirs by analyzing seismic attributes such as RMS-Amplitude, as well as pressure test results, etc. The objective was to create a new model to facilitate dynamic reservoir characterization, decision-making, and the field development planning.
The input data included the logs and petrophysical interpretation of the 4 wells drilled in the field, results of pressure tests, and the volume of the seismic attribute RMS-Amplitude (in the depth domain), among other key inputs.
The previous studies of this area using seismic attributes have identified anomalies that can help locate and outline sandy formations possibly holding hydrocarbons. They correspond to the anticipated fluvial-deltaic environment in this area (see image below). Hence, it was suitable to utilize this attribute to define and portray the facies shapes in this reservoir.
The table below summarizes how the seismic attribute and pressure test analyses, etc., were used to assign average values to the geometric parameters of the bodies in the assumed fluvial-deltaic environment.
Once the average values for the geometric parameters of the facies bodies were determined, the next step involved constructing synthetic facies logs for the 4 wells drilled in the reservoir. The synthetic facies logs must align with the already available petrophysical interpretation. See the image below for reference.
Completing the previous steps, a 3D facies modeling task was performed in the PETREL Geomodeling platform. The image below depicts the 3D facies distribution of the reservoir. This volume will be later populated with the porosities and water saturations.
After completing a probabilistic task with multiple realizations, using the available 3D facies modeling, the volume and maps for the Probability of SAND (PoSAND) Distribution, were generated. The image below displays the superposition of the PoSAND contours and the seismic attribute RMS-Amplitude on the top of one of the layers of interest. It's important to note the strong correspondence between these two significant reservoir features.
To ensure the project's success, it was crucial to have a robust and reliable petrophysical interpretation. This interpretation, among other important features, needed to explicitly consider the presence of multiple and different water levels in the reservoir column, also known as Free Water Levels (FWLs). This is extremely important for an accurate evaluation of oil volume in this clastic reservoir.
It was also important to cross-check, the results of facies modeling and seismic attribute analysis with the petrophysical interpretation to ensure they aligned well. This involved intersecting the trajectories of the four wells with the attribute volume to create corresponding seismic attribute logs, along with synthetic facies logs. The image above shows a cross-section of one of the target layers (A24) demonstrating a strong visual correlation supported by the quantitative correspondence shown for two wells in the cross plots in the image below.
The 3D facies model was then populated with petrophysical properties such as effective porosity (PHIE), and water saturation (SW); the FWLs and the seismic attribute RMS-Amplitude, were also included. The resulting Base Case was used to perform multiple realizations and uncertainty quantification analysis to evaluate reliable statistical summaries for inferences and predictions.
The map above displays the statistical summary of the Join-Probability for a specific layer. It was created by evaluating a sample of thousands of geostatistical realizations and focusing on the reservoir rock within the layer, specifically the Channel-Lobe facies with a porosity (PHIE) of 12% or higher and a water saturation (SW) of 45% or less.
This map, and several cross-sections, were used to fine-tune the trajectories of the new highly deviated wells planned to be drilled in the field. It's worth noting that the original trajectories, indicated on the map with yellow arrows, were developed using independent criteria; however, they closely align with the tendencies predicted by the geostatistical model.
Particularly useful to fine-tune the trajectories of the deviated proposed wells were the cross-sections showing the vertical and lateral distribution of SW and the FWLs in the layer targets. The image below illustrates the layer labeled A24. The SW 3D distribution corresponds to the statistical summary _p50 percentile, which was also evaluated from the sample of thousands of geostatistical realizations.
Some of the planned wells needed to reach the A18 layer, as illustrated in the image below. It is evident that the initial design did not fully hit the target along the path, and the geostatistical model helps the planners in optimizing it.
As a final note, it is important to mention that one of the proposed wells, with its trajectory optimized using this new model, was drilled and yielded as much as 4000 bbl of oil per day. The details are not provided here for the sake of confidentiality.
Hoping the provided example, which showcases a successful use case of a green hydrocarbon oil field in Mexico, will be informative and help streamline processes in the oil and gas industry.
Are you interested in learning more or conducting a one-on-one user test? Contact us through our Website or use the contact form on this page. Or please, visit our YouTube Channel and subscribe.
Comments