An Integrated IR4.0 Software Enables Autonomous Casing Crossflow Rate Calculation

Abstract

This paper will present an alternative calculation technique to predict wellbore crossflow rate in a water injection well resulting from a casing leak. The method provides a self-governing process for wellbore related calculations inspired by the fourth industrial revolution technologies. In an earlier work, calculations techniques were presented which do not require the conventional use of downhole flowmeter (spinner) to obtain the flow rate. Rather, continuous surface injection data prior to crossflow development and shut-in well are used to estimate the rate. In this alternative methodology, surface injection data post crossflow development are factored in to calculate the rate with the same accuracy. To illustrate the process an example water injector well is used. To quantify the casing leak crossflow rate, the following calculation methodology was applied:Generate a well performance model using pre-crossflow injection data. Normal modeling techniques are applied in this step to obtain an accurate model for the injection well as a baseline case.Generate an imaginary injection well model: An injection well mimicking the flow characteristics and properties of the water injector is envisioned to simulate crossflow at flowing (injecting) conditions. In this step, we simulate an injector that has total depth up to the crossflow location only and not the total depth of the example water well.Generate the performance model for the secondary formation using post crossflow data: The total injection rate measured at surface has two portions: one portion goes into the shallower secondary formation and another goes into the deeper (primary) formation. The modeling inputs from the first two steps will be used here to obtain the rate for the downhole formation at crossflow conditions.Generate an imaginary production well model: The normal model for the water injector will be inversed to obtain a production model instead. The inputs from previous steps will be incorporated in the inverse modeling.Obtaining the crossflow rate at shut-in conditions: Performance curves generated from step 3 & 4 will be plotted together to obtain an intersection that corresponds to the crossflow rate at shut-in conditions. This numerical methodology was analytically derived and the prediction results were verified on syntactic field data with very high accuracy. The application of this model will benefit oil operators by avoiding wireline logging costs and associated safety risks with mechanical intervention.