Integrating geological, petrophysical groups and MICP cluster approaches: Methodology unifying RRT scheme, prediction and modeling workflows

  • Sara Hasrat Khan

Student thesis: Master's Thesis


Oil and gas ventures throughout the world seek to produce hydrocarbons from the subsurface with maximal recovery and low cost to benefit ratios. This requires a thorough understanding of the depositional and diagenetic parameters that control the reservoir quality. The dilemma exists is connecting descriptive geological factors to numerical measurements of reservoir quality. The concept of Reservoir Rock Typing helps resolve this by clustering groups of rocks with similar depositional fabrics, which have undergone similar diagenetic processes, have similar petrophysical relationships and are continuous over a large area. These could be geological (Dunham texture, Flugel microfacies), petrophysical groups (Winland, FZI) or based on MICP clusters. The overall RRT workflow is divided into 3 major parts - RRT Scheme that characterises the final poro-perm clusters based on the best adopted methodology, RRT Prediction which aims to increase the data density in uncored wells using neural network approaches and RRT modelling that controls the spatial and vertical distribution of these units in a 3D framework. The main downfall of this approach is that each type of RRT is defined using different approaches, datasets of varying scales and by different disciplines- there is no methodology to reconcile all three. Since eventual models are numerical, the petrophysical groups and MICP clusters are relied upon, which commonly cut through different geological groups. The aim of this thesis is to establish a methodology to cluster the numerical measures of reservoir quality (petrophysical groups and MICP clusters) using geological parameters. The Arab C reservoir from Arzanah field was selected due to its heterogeneity. The reservoir is deposited on a low angle ramp in the Upper Jurassic. The depositional profile ranges from supra tidal to intertidal. There are 10 cycles of alternate thin massive anhydrite beds and reservoir facies. The reservoir fabrics are pervasively dolomitized as per the seepage reflux mechanism. The reservoir facies range from dolomudstones to dolorudstones. The top of each reservoir zone is capped by the presence of bioclastic rudstones-grainstones which have higher reservoir quality. Using conventional published techniques of reservoir rock typing does not favour the clustering. The final RRT's are established by linking the depositional fabrics and diagenesis to sequence stratigraphy. Each systems tract forms a unique RRT resulting in 4 clusters with unique poro-perm behaviours. RRT 1 (best reservoir quality) is composed of high energy bioclastic rudstonegrainstones, which belong to the HST, that have undergone leaching of aragonite allochems, as they were subject to subaerial exposure. RRT 2 is composed of RRT1, where the pore space has been filled with rhombic dolomite cement. RRT 3 is composed of mud dominated packstones, rudstones and floatstones, where the mud matrix has been pervasively dolomitized thereby enhancing reservoir quality. RRT 4 is the worst in terms of the reservoir quality, composed of mudstones and dolograinstones. Since only 22 wells contain RRT information, well log cut offs were used to establish electro-RRT's, thereby increasing overall data density. The structural model is built using sequence boundaries as markers, which is picked from Gamma Ray logs. Thus the zonation scheme follows the systems tracts that helps the spatial distribution of the RRTs. The standard conventional RRT scheme has been compared to the proposed RRT scheme, defined by sequence stratigraphic observations, using blind tests as validation. The proposed RRT scheme shows higher degree of match. The primary findings of the study is that the current methodologies of RRT do not respect the geology. The only way to allow geological RRT to guide the petrophysical groups and MICP is to link the deposition and diagenesis to sequence stratigraphy instead of using pie charts and correspondence analysis.
Date of Award2016
Original languageAmerican English
SupervisorJorge Salgado Gomes (Supervisor)


  • Applied sciences
  • Petrophysical
  • Rrt scheme
  • Static reservoir
  • Petroleum engineering
  • 0765:Petroleum engineering

Cite this