Data Transformation and User Modes in Property Modelling

Distributions Tab

To add detail to the facies model

Porosity & SW

Variograms show the relationship between residual values at different separation distances

Simulation

The maximum dissimilarity

To transform input data into a Gaussian distribution

The degree of dissimilarity at zero distance

Mean of 0 and Standard Deviation of 1

Allows grouping of zones in a multizone grid

Stochastic simulation

To provide a good enough estimate for quick look volumetrics

To quantify the spatial continuity and variability of the data

To condition the petrophysical model using a seismic parameter

Correcting the residuals

To calculate the 2D probability density function

The strength of the relationship between seismic data and parameters

To define the 3D spatial distribution of parameters

User mode

To account for clustering of data in certain areas

Truncate data and real.

To define a change in petrophysical property within an individual facies body

User mode

Basic Transformations

Logarithm

Depositional depth trend

The purpose of petrophysical modelling is to determine the distribution of reservoir properties, such as porosity and water saturation (SW), which are used for volumetric calculations and as input for reservoir simulation.

The facies model provides a geological framework and characterizes the quality of the reservoir. The petrophysical model then adds detail to the facies model by determining the distribution of reservoir properties.

The Variograms tab is used to define the spatial correlation and variability of the reservoir properties. It allows for the estimation of variogram models, which are used in the simulation of the reservoir properties.

A variogram model is used to determine the likely value of a particular property by specifying the continuity and variability of the parameter based on the separation distance between observations.

Variance ($\sigma^2$) represents the degree of dissimilarity between points, while sill is the maximum dissimilarity. In the absence of trends, sill is equal to the variance. However, in the presence of trends, there is usually no sill.

The nugget term represents the degree of dissimilarity at zero distance, which is typically attributed to microstructure or instrument noise/measurement error in sampling.

Stochastic simulation is used to provide more realistic reservoir simulation results in the dynamic modelling phase.

The steps involved in the workflow for modelling a petrophysical parameter are: 1. Analyze well logs and use geological concepts to define trends in the data. 2. Remove trends as part of the data transformation process. 3. Define correlations between pairs of data. 4. Define variogram models to quantify spatial continuity and variability. 5. Distribute the residuals using input data and transform back to original units.

A facies model allows geological concepts to be imparted in 3D space and makes the subsequent distribution of petrophysical properties more robust.

Seismic cosimulation involves using a seismic parameter or other parameter indicating a relationship with the modelled parameter to condition the petrophysical model.

Prediction (or kriging) and Simulation.

Prediction (or kriging) is a mean value description that relies on spatial continuity encapsulated in a variogram. It is locally accurate and smooth in appearance. Simulation, on the other hand, aims at a more realistic global representation of geological heterogeneity. It is stochastic in nature, globally accurate, and has many equally probable outcomes.

Pros include the ability to group zones together, use DA from one zone in another zone, set up analogue data as input, perform variogram analysis, and include a DA object generated from raw log data. Cons include the potential for confusion with project workflow/documentation and slightly longer processing time.

The main function of the Distributions tab is to transform the input data into a Gaussian or 'Normal' distribution prior to simulation, and then transform the Gaussian distribution back to the input data after simulation.

The Cloud transform is used to calculate the 2D (bivariate) probability density function (pdf) for the 3D parameter and the well log parameter. It is used to establish the relationship between the parameters and is an easy-to-use solution.

The Transformation Sequence Workflow is used to transform data in a specific sequence. It incorporates various steps such as truncating the input data, incorporating geological trends, incorporating skewness, and correcting the residuals to give a mean of zero.

The Variograms tab is used to define the 3D spatial distribution of the parameters to be modelled. It helps in understanding the correlation between different variables and in estimating the spatial variability of the parameters.

The Correlation tab is used to specify the linear correlation between two residual distributions after transformation. It can be used to cosimulate petrophysical parameters or seismic data, and helps in understanding the relationship between different variables.

The six transformation categories in petrophysical modelling are: 1) Truncate data 2) Truncate realisation 3) Truncate data and real 4) Mean (Constant) 5) Logarithm 6) Skewness

The purpose of declustering in petrophysical modelling is to account for the clustering of data and to attach a weighting to each data point.

The Kriging info button provides details about the type of kriging used in the software, which is selected automatically based on the user mode and whether trends and/or seismic data are used.

The three user modes in the RMS software are: 1) User mode: The selection of the user mode should depend on the extent of the input data and the complexity of the reservoir. 2) Advanced mode: This mode provides the full range of options for property modelling. 3) Constant mode: This mode allows the user to specify a constant value for a facies. The selection of the user mode should be based on the level of detail and geological realism of the input facies model.

The two types of depth trends in petrophysical modelling are: 1) Compactional depth trend: This trend represents the change in petrophysical properties with depth due to compaction and cementation processes. 2) Depositional depth trend: This trend represents the trends resulting from deposition and is related to stratigraphic depth.