An unexpected and unwanted influx of gas or "kick" into the wellbore during hydrocarbon drilling can cause catastrophic blowout incidents, resulting in human casualties, ecological damage, and asset losses. The ability of the oil and gas industry to control gas kick depends on our ability to accurately detect and monitor gas migration in a borehole in real-time. This study demonstrates the application of optical fiber-based Distributed Acoustic Sensors (DAS) for early detection and monitoring of gas in wellbore. Multiphase flow experiments conducted in a 5000 ft. deep test-well are analyzed for different injection, circulation, and pressure conditions. In each case, the low-frequency component of DAS demonstrates a superior capability to detect gas signatures both inside the tubing and the annulus of the well, even at small gas volumes. In comparison, the highfrequency DAS data seems limited in detail. The gas influx velocity was calculated using the frequency-wavenumber analysis of the gradient of the low-frequency DAS phase with respect to time, which shows good agreement with theoretical velocity estimates using flow models and surface gauge measurements. This study demonstrates a novel workflow to analyze low-frequency DAS to qualitatively and quantitatively map gas influx in a wellbore.

Early detection of a gas kick is crucial for preventing uncontrolled blowout that could cause loss of life, loss of assets, and environmental damage. Multiphase flow experiments conducted in this research demonstrate the capability of downhole fiber optic sensors to detect a potential gas influx in real-time in a 5000 ft deep wellbore. Gas rise velocities estimated independently using fiber optic distributed acoustic sensor (DAS), distributed temperature sensor (DTS), downhole gauges, surface measurements, and multiphase flow correlations show good agreement in each case, demonstrating reliability in the assessment. Real-time data visualization was implemented on a secure cloud-based platform to improve computational efficiency. This study provides novel insights on the effect of circulation rates, gas kick volumes, backpressure, and injection methods on gas rise dynamics in a full-scale wellbore.

The newly developed workflow is based on a 4-D sector modeling study on a west flank of Field A, offshore Saudi Arabia. The approach involves the use of time-lapse monitoring of reservoir saturation profile for injectors and producers as well as 4-D water flood front movement in the reservoir to understand dynamic reservoir challenges. These challenges are zones with future water breakthrough, dynamic injection/production interaction, impact of reservoir drive mechanism on well injectivity or productivity, initial and future reservoir conditions. With this understanding, rate optimization process is performed on the open hole injectors and producers. Thereafter, completion is optimized to achieve uniform influx profile, flow restriction and zonal isolation where required along the horizontal section. Four different completion strategies were evaluated by alternating between Inflow Control Devices (Nozzle based) completion and open hole in injectors and producers. Even though the sector model used in this study is subject to further calibrations as more geological, petrophysical and production data become available; the study outcome based on the novel workflow demonstrated the challenges involved in making a completion strategy with ICD's in injectors and producers. The results showed an incremental cumulative oil gain of (5%) with a one year delay in water breakthrough over a period of 20 years. It also indicated that further improvement in oil recovery can be achieved by ICD completion above that achieved in the rates optimization process of injectors and producers. The study showed a positive indication that ICD completion is beneficial in challenging reservoirs with mobility ratios considerably greater than one. The industry should consider a novel 4-D well – reservoir integrated modeling approach to making completion strategy and evaluating ICD completion design prior to deploying them in a field wide campaign.

Long horizontal wells in naturally fractured carbonate reservoirs often exhibit very high water-cut within months of production because of the early arrival of water from natural fractures. Passive inflow control devices (P-ICDs) have been used globally to balance influx, delay water or gas breakthrough to prolong well life. However, some wells have continued to experience high water-cut despite the control measures. Image log review has revealed the uncertainty is in the identification of fractures and its conductivity networks. Two additional zonal control technologies are presented in this paper: on/off ICDs and intelligent (IC) or smart completions in comparison. A software-based 3D reservoir model was built to represent a horizontal oil-producer in a fractured carbonate reservoir penetrating a thin oil rim. The first model simulated well production performance in a well with on/off ICD. Intervention was replicated in time (i.e., taking longer) to shut-off ICDs. The second model evaluated production forecast over the same period for the same well, this time equipped with an IC in the open hole (OH). Actions in this case were taken right away from the surface (i.e. without downhole intervention) to identify and restrict or shut-off intervals with water breakthrough. Time-lapsed 3D reservoir model calibration is possible with ICs as they provide real-time downhole pressure and temperature across each interval. The timely control of zonal valves from surface actuation reduced production of water or gas. On/off ICDs, on the other hand, necessitated scheduling a production log (PL) to confirm the interval of water or gas breakthrough and performing coiled-tubing (CT) intervention to shut-off the problematic zone. Intervention comes at cost of interrupting well production and reducing net oil recovery. A simplified cost-benefit analysis of both cases showed that despite a higher initial capital investment in ICs, well operating costs were substantially lower with higher oil recovery. In IC solution, costs for running production logs and intervention tools were eliminated and so was the risk of losing these tools in the hole and the loss in production during the intervention period. Continuous monitoring of downhole pressure data helped reservoir characterization and prediction of reservoir production behavior without compromising production on-stream time. A comparison of different reservoir flow control devices suggests that ICs are the optimal choice in some fractured carbonate reservoir conditions. They provide real-time monitoring of each producing zone and surface control of the flow control valve (FCV) settings in real-time as reservoir performance changes. They enable production testing evaluation—without production logging and interventive shifting with CT, i.e. to determine the source of water entry and optimization of multi-zone production without downhole intervention.

Early detection and quantification of gas kicks during drilling and completions is essential to proper well control and in the prevention of blowouts. The utilization of distributed sensing techniques, acoustic (DAS) and temperature (DTS), enables real-time elucidation of these multiphase flow events. Identifying and validating event signatures (fingerprinting) in these sensing technologies is crucial to informing operators of how to interpret these data streams. Performing full-scale analysis allows these events to be properly characterized, given the complexities in the fluid mechanics and gas dynamics. This project utilizes a 9-5/8 inch and 5200 foot deep wellbore at the LSU PERTT Laboratory retrofit with distributed fiber optics (DAS and DTS) and 4 permanent pressure-temperature gauges to sense and visualize gas kick dynamics downhole in real time. Several experiments were performed involving the injection of nitrogen kicks through a chemical injection line and also by bullheading down 2-7/8 inch tubing in both stagnant and circulating water. Variations in flow rate, kick size, and backpressure are investigated including gas migration during shut-in. DTS and DAS data are collected downhole, along with gauge pressure and temperature at four depths along the wellbore. Data is consolidated with the rig recorded surface data to create a complete picture of the experiments. Several observations are possible with this new methodology. First, the gas kick is immediately visible (audible) entering the wellbore by the sensors and the gas front was traceable in real time as it rose to the surface, allowing for detection of a kick, improved estimation of kick size, and easy calculation of rise velocity. Second, the distribution of gas axially in the wellbore was visible and provided insights into the duration of the event. Third, the compressibility dynamics can be visualized with the DAS thus elucidating details of bubble and slugging sizes and dynamics and when discrete gas has completely circulated out of the wellbore. The frequency ban filtering of the data further augments the fidelity of gas bubble sizes and dynamics. These initial results provide a proof of concept for using downhole sensing for real time riser gas dynamic detection and characterization.