Flow Profile Determination from Inversion of Distributed Temperature Measurements
Conference Paper
Overview
Identity
Other
View All
Overview
abstract
AbstractDownhole temperature monitoring, either by production logging or fiber-optic sensors, has been used in many applications for production/injection/stimulation diagnosis and design optimization. To quantitatively generate flow profiles at downhole conditions, we need theoretical models to simulate fluid and heat flow, and inversion procedures to interpret measured temperature. Because of the large uncertainty, the inversion process strongly depends on the well structure, completion and also the initial estimate of the flow field for inversion. Initializing the inversion method without preliminary estimation of the flow profile often can lead to low inversion efficiency, non-convection or incorrect interpretation. In this paper, we present techniques that can be used to initialize the interpretation condition to obtain fast and correct interpretation.We summarize the available forward modeling and inversion approaches to interpret flow profiles from monitored temperature profiles. We also illustrate how to consider the effects of well completion and field operation as constrain when interpreting measured temperature. The uniqueness of temperature measurements and challenges in interpretation are discussed with field applications, including identifying water entry location and generating flow rate profiles for oil and gas producing wells, analyzing acid placement for matrix acidizing, and diagnosing production performance of multi-stage fractured horizontal wells. We discuss how to select methodologies and perform analysis for each particular case, and how to increase the interpretation efficiency and accuracy.The paper presents several field cases of using temperature to diagnose production, injection and stimulation performances. The detailed procedure is illustrating in the paper for each case, and possible mistakes in interpretation are summarized for each case. From the results, we show that because of the large number of parameters involved in temperature interpretation (reservoir condition, fluid properties, well and completion structures, and operating condition), and also because temperature is only measured along the wellbore, a non-unique solution with high uncertainty is a common problem. Logical and engineering judgement should be used to ensure a more reasonable interpretation. There is no "one size fits all" automatic procedure that generates a universal solution for the problems. A sensible initial guess of flow condition based on the local temperature gradient can lead to a fast convergence, reduce the uncertainties in the solutions, and lead to more accurate interpreted results. Locations where the slope of the temperature curve changes from positive to negative indicate possible gas entries, and magnitude of temperature changes at these locations is related (not linearly) to gas flow velocity. The same principle can also be applied to oil-producing wells.The paper summarizes the experiences and lessons learned from the past research to provide guidelines on how to more efficiently interpret distributed temperature data.
