Abstract
In this paper, we explore the potential of factor analysis in mapping and modeling fracture corridors using dynamic data in mature fields. Fracture corridors are sub-vertical tabular swarms of fault related conductive fractures which often intersect the entire reservoir, extend laterally for several hundred meters and play major role in reservoir dynamics. Most fracture corridors are too small to be captured by seismic maps. Therefore additional borehole data is needed to map fracture corridors deterministically.
The objective of factor analysis is to discover a few underlying variables which exert control over several observed variables. Measurements of dynamic data in wells, such as productivity Index (PI), gross rates, water cut, Khtest/Khcore ratio, Gas oil ratio (GOR), lost circulation etc. are visible manifestations of underlying factors such as porosity, permeability, mobility ratio, thickness etc. If one of the underlying variables can be identified as fracture corridors, it may be used to identify and map fracture corridors from dynamic data.
Factor analysis was applied in mature carbonate reservoir. The field has a large number of vertical wells with dynamic data and a small number of horizontal wells with borehole image logs. The analysis shows that two factors account for more than 90% of total variation in the observed data from vertical producers. The first factor is correlated with gross rates, water cut rate of rise, mud losses, stepping flow profiles, and injector-producer short cuts. This factor is interpreted as representing fracture corridors. Correlation of mud losses with image logs from horizontal wells confirmed the interpretation which eventually led to successful mapping of fracture corridors in the field from dynamic data.