Fast Proxy Models for Probabilistic Evaluation of Downhole Fluid Sampling Operations

SPE Annual Technical Conference and Exhibition

Gisolf

et al. 2017

Acquisition of fluid samples using wireline formation testers (WFT) is an integral part of reservoir evaluation and fluid characterization. Recent developments in formation tester hardware have enabled wireline-based fluid sampling in a wide range of downhole conditions. However, accurate quantification of oil-based mud (OBM) filtrate contamination using data from downhole fluid analysis (DFA) sensors alone remains challenging, especially in difficult sampling environments and for advanced sampling tools which have complex inflow geometries and active guarding of filtrate flows. Such tools and conditions lead to contamination behaviors that do not follow simple power-law models which are commonly assumed in OBM contamination monitoring (OCM) algorithms. In this paper, we introduce a new OCM algorithm based on inversion of DFA data using a full 3D numerical flow model of the contamination cleanup process. Based on formation and fluid properties and operational tool settings, the model predicts the evolution of filtrate contamination as a function of time and pumped volume and can thus be used to forward-model the DFA sensor responses. Sensor data are then inverted in real time to provide contamination predictions. Real-time computation is enabled through fast, high-fidelity proxy models for the cleanup process. The proxy models are trained on and thoroughly vetted against a large number of full-scale numerical simulations. Compared to existing algorithms, the new OCM method is now applicable for all types of sampling hardware and a wider set of operating conditions. By directly relying on a model of the cleanup process, the physical properties of the formation and fluids (such as porosity, permeability, viscosity, and depth of filtrate invasion) are estimated during the inversion, thus providing additional valuable information for formation evaluation. The new method is demonstrated by practical application in both synthetic and field examples of oil sampling in OBM. The synthetic examples demonstrate the robustness of the algorithm and show that the true formation and fluid properties can be recovered from noise-corrupted sensor data. The field example presented demonstrates that contamination predictions are in good agreement with results from laboratory analysis, and the inverted formation properties are consistent with estimates based on openhole logs and pressure measurements.

Real-Time Formation Evaluation and Contamination Prediction Through Inversion of Downhole Fluid Sampling Measurements

SPE Annual Technical Conference and Exhibition

Gisolf

et al. 2017

Real-Time Formation Evaluation and Contamination Prediction Through Inversion of Downhole Fluid-Sampling Measurements

SPE Reservoir Evaluation &Amp; Engineering

Gisolf

et al. 2018

Summary Acquisition of fluid samples using wireline-formation testers (WFTs) is an integral part of reservoir evaluation and fluid characterization. Recent developments in formation-tester hardware have enabled wireline-based fluid sampling in a wide range of downhole conditions. However, accurate quantification of oil-based-mud (OBM) filtrate contamination using data from downhole-fluid-analysis (DFA) sensors alone remains challenging, especially in difficult sampling environments and for advanced sampling tools that have complex inflow geometries and active guarding of filtrate flows. Such tools and conditions lead to contamination behaviors that do not follow simple power-law models that are commonly assumed in OBM-contamination-monitoring (OCM) algorithms. In this paper, we introduce a new OCM algorithm derived from an inversion of DFA data using a full 3D numerical flow model of the contamination-cleanup process. Using formation and fluid properties and operational tool settings, the model predicts the evolution of filtrate contamination as a function of time and pumped volume, and can thus be used to forward model the DFA sensor responses. Sensor data are then inverted in real time to provide contamination predictions. Real-time computation is enabled through fast, high-fidelity proxy models for the cleanup process. The proxy models are trained on and thoroughly vetted against a large number of full-scale numerical simulations. Compared with existing algorithms, the new OCM method is now applicable for all types of sampling hardware and a wider set of operating conditions. By directly relying on a model of the cleanup process, the physical properties of the formation and fluids (such as porosity, permeability, viscosity, and depth of filtrate invasion) are estimated during the inversion, thus providing additional valuable information for formation evaluation. The new method is demonstrated by practical application in both synthetic and field examples of oil sampling in OBM. The synthetic examples demonstrate the robustness of the algorithm and show that the true formation and fluid properties can be recovered from noise-corrupted sensor data. The field example presented demonstrates that contamination predictions are in good agreement with results from laboratory analysis, and the inverted formation properties are consistent with estimates derived from openhole logs and pressure measurements.

Optimization of Downhole Fluid Sampling Using Focused Probes

Machado

SPE Europec Featured at 80th EAGE Conference and Exhibition

2018

Acquisition of fluid samples using wireline formation testers is an integral part of reservoir evaluation and fluid characterization. The use of focused probe technology, in which the flow of mud filtrate to the main sampling port is shielded by one or more guard ports, has proved to be particularly effective in acquiring clean formation fluid samples with short pump-out times. To both improve operational efficiency for the current focused probes and optimize the design of next-generation tools, it is necessary to understand how to optimally operate the tools in different sampling environments. The objective of this study is to investigate optimal sampling strategies for focused tools in presence of formation and fluid property uncertainty. In particular, the pump rates for the sample and guard ports in a focused tool can be manipulated, allowing for optimization of pump rate profiles to maximize overall sampling efficiency. We use a numerical forward model of the filtrate cleanup process coupled with optimization. We study the problem of pump rate profile optimization in different sampling environments and compare the results against two operating strategies commonly applied in the field. The optimization results show that significant sampling time savings are possible (for each sampling station) compared with a default fixed-rate strategy, especially in environments characterized by a high viscosity contrast between the formation fluid and mud filtrate. These savings translate directly into rig time savings for the operator. In general, the results in the paper provide guidance on optimal focused-sampling operation in different environments.