Rock Typeand Permeability Prediction of a Heterogeneous Carbonate Reservoir Using Artificial Neural Networks Based on Flow Zone Index Approach

Abstract

The Cretaceous carbonates of Sarvak formation formed large hydrocarbon reservoirs in the South-west region of Iran. The studied field is a tight carbonate reservoir in which several exploration wells have been drilled, and is in the process of development. Since, only few wells have core data, therefore it was decided to integrate the available core and log data using new methods that describe the carbonates heterogeneity more precise. 3D modeling of permeability is an essential part of building robust dynamic model for proper reservoir management and making reliable predictions. A good definition of reservoir rock types (RRT) could relate somehow better geological modes to dynamic models. Rock typing by flow-zone-index (FZI) and rock-quality-index (RQI) values proved to be an effective technique to develop porosity-permeability transforms for RRTs in a reservoir model. RRTs were defined based on the core derived FZI through some mathematical and statistical approaches. Permeability estimation using artificial neural network approach (ANN) was then made through a two-step process. In the first step, FZI log was estimated from a trained neural network using the standard suite of logs as input (Gamma ray, Sonic, Density, Neutron porosity) and FZI-core as output in a subset of cored wells (Key wells). In the second step, individual trained neural networks implemented porosity-log and FZI-log from the first step to predict permeability-log for each RRT. Validation of the predictive capability of the method in two cored wells (Blind-test wells) that are located in the field proved the estimation technique to be robust and was found to be valuable to supplement core data in the prediction of log-permeability in the entire reservoir wells. For the sake of comparison between the result of this work and the work which was based on the integration of sedimentological, petrographical, and diagenetic study, the results were found to be in good agreement for most of the log interval. However, the predictions of the ANN approach in the regions where core data are not available are better and it follows the log property variation logically. Introduction Rock typing is a process for the classification of reservoir rocks into distinct units. These units are deposited under similar geological conditions and undergoes through similar diagenetic alterations. If the rocks are properly classified and defined, a given rock type is imprinted by a unique porosity/permeability correlation, capillary pressure profile, and set of relative permeability curves. In addition, the true dynamic characteristics of the reservoir will be captured in the reservoir simulation model as a more reliable permeability model is used1–3. On the classical plot, the relationship between permeability and porosity is not causal. Whereas porosity is generally independent of grain size, permeability is strongly dependent on grain size. Hence, traditional plot cannot be used reliably to estimate accurate permeability from porosity. Several investigators1–4 have noted the inadequacy of classical approach and have proposed alternative models for relating porosity to permeability specifically for using in carbonates reservoir. From the classical approach it can be concluded that for any given rock type, the different porosity/permeability relationships are evidence of the existence of different hydraulic units. In fact, several investigators5–7 had come to similar conclusions about porosity/ permeability relationships.

Determination of Reservoir Permeability Based on Irreducible Water Saturation and Porosity from Log Data and FZI (Flow Zone Indicator) from Core Data

“…9. The permeability of a reservoir rock is a function of the porosity and the irreducible water saturation.…”

Improved Permeability Prediction in Heterogeneous Carbonate Formations

“…The method is based on a modified Kozeny-Carmen equation and the concept of mean hydraulic radius. The derivation of the parameter, Reservoir Quality Index (RQI) and FZI are detailed by Kharrat (Kharrat, R. et al 2009). RQI is given by the following equation:…”

Predicting hydrofacies and hydraulic conductivity from direct-push data using a data-driven relevance vector machine approach: Motivations, algorithms, and application

“…A fundamentally different approach to model H‐G relations involves the application of statistical techniques, either simple parametric relations [e.g., Hyndman et al ., ] or more complex nonparametric methods [e.g., Mohaghegh et al ., ; Wong et al ., ; Lee and Datta‐Gupta , 1999; Chen et al ., ; Chen and Rubin , ; Paasche et al ., ; Shokir et al ., ; Dubois et al ., ; Al‐Anazi et al ., ; Elshafei and Hamada , ; Kharrat et al ., ; Al‐Anazi and Gates , 2010a, b; Dubreuil‐Boisclair et al ., ; Ruggeri et al ., ; Rumpf and Tronicke , ]. Unlike general theoretical or semiempirical models, statistical techniques are much more flexible and do not require prior knowledge about physical relations between various H‐G parameters or geological material.…”

“…Artificial neural networks (ANNs), which follow an empirical risk minimization of the training errors, are common form of learning machines that have been already considered. Different architectures of ANNs have been applied successfully for the prediction of lithofacies [ Chen and Rubin , ; Dubois et al ., ] or hydraulic properties in petroleum reservoirs [ Mohaghegh et al ., ; Wong et al ., ; Lee and Datta‐Gupta , ; Shokir et al ., ; Al‐Anazi et al ., ; Elshafei and Hamada , ; Kharrat et al ., ; Iturrarán‐Viveros and Parra , ] from cross hole or borehole geophysics data. However, despite their potential effectiveness, ANNs present some important drawbacks [ Camps‐Valls et al ., ]: (i) design and training often results in a complex, time‐consuming task, in which many parameters must be tuned; (ii) minimization of the training errors can lead to poor generalization performance; and (iii) performance can be degraded when working with small (sparse) data sets.…”