A FUZZY LOGIC APPROACH TO ESTIMATING HYDRAULIC FLOW UNITS FROM WELL LOG DATA: A CASE STUDY FROM THE AHWAZ OILFIELD, SOUTH IRAN

Abstract

Porosity‐permeability relationships in the framework of hydraulic flow units can be used to characterize heterogeneous reservoir rocks. Porosity is a volumetric parameter whereas permeability is a measure of a rock's flow properties and depends on pore distribution and connectivity. Thus zonation of a reservoir using flow zone indicators and the identification of flow units can be used to evaluate reservoir quality based on porosity‐permeability relationships. In the present study, we attempt to make a quantitative correlation between flow units and well log responses using fuzzy logic in the mixed carbonate‐clastic Asmari Formation at the Ahwaz oilfield, South Iran. A hybrid neuro‐fuzzy approach was used to verify the results of fuzzy modelling. For this purpose, well log and core data from three wells at Ahwaz were used to make an intelligent formulation between core‐derived flow units and well log responses. Data from a separate well was used for evaluation and validation of the results. The results of this study demonstrate that there is a good agreement between core‐derived and fuzzy‐logic derived flow units. Fuzzy logic was successful in modelling flow units from well logs at well locations for which no core data was available.

Prospects of Intelligent Oil and Gas Fields

“…These problems include the intellectualization of analysis of large amounts of data collected from oil and gas fields, the intellectualization of drilling process, the forecast of reserves and optimization of oil and gas production, the optimization of the location and management of oil and gas fields, etc. For solving these and other problems, artificial neural networks [15][16][17][18][19], fuzzy logic [20][21][22][23][24], expert systems [25][26][27][28], machine learning methods [29], intelligent agents [30,31], genetic algorithms [32][33][34][35], data extracting methods [36,37], case-based reasoning -CBR [38][39][40], etc.…”

“…Fuzzy logic is used in the areas associated with oil and gas production technology. These areas include physics of oil reservoir [20,21], determination of oil and gas reserves [22], increasing oil and gas production [23], decision-making process [26], etc.…”

Improved reservoir characterization by employing hydraulic flow unit classification in one of Iranian carbonate reservoirs

“…Interpretation of flow units based on the petrophysical properties, well log data, and stratigraphy are routinely done for reservoir characterization. As reported by several researchers, this technique provides a better input data set for numerical flow simulation, when compared to those provided by lithological or depositional facies (Chen et al, 1950;Davies et al, 1996;Soto et al, 2001;Bagci et al, 2007;Schatz et al, 2007;Kadkhodaie-Ilkhchi et al, 2009;Nooruddin et al, 2011;Izadi et al, 2012Izadi et al, , 2013Mirzaei-Paiaman et al, 2015). Different methods can be applied for interpreting the flow units and their corresponding petrophysical properties (Bagci et al, 2007;Banga et al, 2007;Schatz et al, 2007;Brigaud et al, 2014;Goda et al, 2014;Pyrcz et al, 2014;Rebelle, 2014).…”

“…searches have been conducted on the evaluation of different flow unit identification techniques (Ali-Nandalal et al, 2003;Hamon, 2003;Frank et al, 2005;Guo et al, 2005;Salman et al, 2009;Hollis et al, 2010;Kale et al, 2010;Rebelle, 2014;Mirzaei-Paiaman et al, 2018). However, a comparative study that includes applications of different methods in reservoir description is lacking (Kadkhodaie-Ilkhchi et al, 2009;Izadi et al, 2013;Mirzaei-Paiaman et al, 2015;Chen et al, 2017).…”

Supervised Intelligence Committee Machine to Evaluate Field Performance of Photocatalytic Asphalt Pavement for Ambient Air Purification

“…With multiple AI models for NO x concentration prediction, one may achieve the optimal performance, reap the benefits of all AI models, and avoid bias to a single AI model. A combination of multiple AI models to estimate NO x concentration under different environmental and operating variables with an ICM method was suggested (20,21). The ICM linearly combines the outputs of individual AI models through a set of weights.…”