Fluidic Diode Autonomous ICD Range 2A Single-Phase Testing

Abstract

Evenly distributed production along the length of the wellbore is important for maximizing the oil recoverables over the life of the well. Traditional, passive in-flow control devices (ICDs) perform well at balancing the completion pressure differential with the reservoir pressure differential so that an even influx across production zones is maintained. This helps to delay unwanted fluid break through. When unwanted fluids, typically of lower viscosity, do finally break through, they can take over the well, significantly reducing the production of oil. Autonomous Inflow Control Devices (AICDs) are a new generation of ICDs. When oil is producing from all zones, the AICD will behave as a passive ICD, balancing flow. However, when lower viscosity (undesired) fluids break through, the AICD chokes them, significantly reducing flow from the zone producing these fluids. This autonomous function enables the well to drain the oil producing zones faster than the undesirable fluid zones, thereby maximizing total oil production. The AICD creates this change in behavior without control lines, moving parts, or electronics.The paper describes the laboratory testing performed to evaluate the performance of the fluidic diode type AICD Range 2A in field-like conditions and compares flow performance curves to a traditional nozzle type ICD. The AICD Range 2A utilizes similar fluid vectoring as the Range 3B (Least et al, 2013), but includes more of an autonomous on/off type switching function instead of a gradual change in performance. The range 2A is currently best suited for oil viscosities of 1.5-10 cP. Results from single-phase experimental flow testing with model oil, water, and nitrogen are presented and discussed.The test results demonstrated that the AICD could restrict flow rates of undesirable fluids. The discussion further shows that if technology such as the new AICD is applied to new well completion designs, total oil recovery can be enhanced, by increasing the life of the well and reducing production of undesirable fluids.

Application of mathematical and machine learning models to predict differential pressure of autonomous downhole inflow control devices

“…However, the range 2 device includes a more autonomous on/off type switching function whereas the Range 3 operates in response to gradual changes in performance. Tests show that the AICD range 2 is most effective with oil viscosities in the 1.5-10 cP range (Zhao et al, 2014).…”

“…AICD are a new technology that has stimulated industry interest due to their low costs, simplicity, precision, autonomous operation, and, because they do not require an energy source (Aakre et al, 2013). The performance of these devices is controlled by viscosity differences of fluids and fluid mechanics principles, including Bernoulli's principle, and static and dynamic fluid pressure variations through the valve, making them powerful and efficient operating devices (Aakre et al, 2014;Zhao et al, 2014;Jovanov, 2016;Eltaher, 2017). Also, unlike ICDs, a large number of them can be used in one well (Halvorsen et al, 2012).…”

“…The viscosity of the flowing fluid is measured using a laboratory viscometer, set up in accordance with the ASTM D-445 standard, and the density of the fluid is determined based on the ASTM D-4052 standard, in field-like conditions (Least et al, 2012) . Zhao et al (2014) also performed laboratory tests to measure the performance of the range 2 Equiflow AICD simulating field like conditions. Liquid testing required a two-phase-flow facility, specifically designed to assess oil and water mixtures.…”

“…Therefore, in all models the viscosity, density, flow rate and range of the valves are the input variables considered and the differential pressure across the device is the output / dependent variable. 90 data points are compiled from published evaluations of ranges 2, 3 and 4 Equiflow AICD (Least et al, 2012;Zhao et al, 2014). The data set is divided into two parts, in which 80 % of overall data are used to develop the model (training subset), and the remaining 20 % are used to test the developed model (testing subset).…”

“…Further increasing the viscosity to 0.229, Table 2. Equiflow AICDs performance data (Least et al, 2012;Zhao et al, 2014) (Least et al, 2012;Zhao et al, 2014).…”