How to Improve Accuracy of a Kick Tolerance Model by Considering the Effects of Kick Classification, Frictional Losses, Pore Pressure Profile, and Influx Temperature

Abstract

Kick tolerance (KT) calculation is essential for a cost-effective well design and safe drilling operations. While most exploration and production operators have a similar definition of KT, the calculation is not consistent because of different assumptions that are made and the computational power of KT calculators. Dynamic multiphase drilling simulators usually provide KT estimates with a minimum number of assumptions. They are much more accessible nowadays for use in predicting the behavior of multiphase flow in drilling and well control operations. However, as far as we observed, the simulation services are mainly used for complex and marginal wells in which low KT may impose additional casing strings, unconventional costly drilling practices, or a high risk of major well control events. Thus, companies often use simplified steady-state models for relatively uncomplicated wells through their own KT calculation worksheets. This practice is usually justified by the misconception that simplified models are always conservative and give less KT than actual conditions. In contrast, some simplifications may lead to higher operational risks due to an overestimated KT, depending on well conditions and parameters. The primary objective of this work was to perform a quality assurance/quality control on KT calculation practices in Company P. Later on, based on our findings, we determined some solutions to improve accuracy in the simplified KT worksheets commonly used by engineers across the company. This became a driver for generating a new KT worksheet (Company Model), in particular for situations in which engineers do not have access to a kick simulator. However, it should not mislead readers about the requirements of the simulator for complex and low-KTwells. Quality assurance/quality control and subsequent investigations found that there are some important criteria and parameters that affect KT calculations, but they are missing in many simplified models or ignored by engineers because they are unaware of or lack adequate references. After reviewing relevant academic literature, common practices and assessing several off-the-shelf software programs, we generated a computer program using Visual Basic for applications to address KT sensitivity to different parameters in steady-state conditions. The newly developed program is based on a single gas bubble model that applies the effect of annular frictional losses, influx temperature, gas compressibility factor, well trajectory, and bottomhole assembly (BHA). Moreover, the program differentiates between swabbing and underbalanced conditions. A logical test is applied to determine the type of kick before computing the relevant influx volume. This kick classification concept is ignored in many KT models; this is a common mistake that leads to misleading results. The annular pressure loss (APL) parameter is sometimes assumed to be zero in KT spreadsheets, while as an additional stress load on the wellbore, it affects the kick budget and must be considered.