Injectivity Profiling in Horizontal Wells via Distributed Temperature Monitoring

Abstract

Abstract Downhole permanent monitoring sensors offer unique options for continuous evaluating flow regimes, completion integrity, formation properties, flow assurance effectiveness and optimal production/injection operations. The use of temperature data for vertical injection wells has been widely applied in practice and has been thoroughly discussed in literature. The problems occur for Distributed Temperature Sensors (DTS) installed in horizontal well due to small variations of geothermal temperature along the wellbore.Unlike the conventional gauges, DTS data are acquired in both temporal and spatial dimensions that exploit the great potential for real-time operation optimization. We present an interpretation technique for analysis of temperature profiles recorded with fiber optic distributed temperature sensor (DTS) in horizontal water injectors. Simple semi analytical models were developed for simulating mass and heat transfer in horizontal well and collector during injection and warm back. Two inversion procedures based on these models were developed allowing calculation of injectivity profile from temperature profiles recorded during injection and warm-back. These procedures were successfully validated on CFD numerical model and tested on field data. Introduction Permanent temperature monitoring of horizontal injection wells contains information on injectivity profile due to dependence of heat transfer on injected fluid flow rate. Three regimes can be used for an inversion procedure:quasi steady heat transfer fluid flow during fluid injection;temperature recovery during warmback in shut-in well;transient heat transfer in well during re-injection after shut-in. Development of corresponding inversion procedures in this work is based on development of semi analytical models of heat and mass transfer process connected with fluid injection into the reservoir. In this paper inversion procedures are presented for regimes (1) and (2). They were validated on 2D numerical model and tested on field case.