Transient Modelling of Gas-in-Riser Unloading and Investigation of Supersonic Flow Conditions and Pressure Build-Up at Surface

Abstract

Gas-in-riser has always been one of the major risks in oil & gas drilling operations. A gas kick entering the riser can lead to riser unloading. Recently, there has been an increased awareness of the gas behaviour when it enters the riser during drilling. Managing the flow rate and pressure is of critical importance in case of such an event. When left unchecked, this event can lead to devastating outcomes such as a blowout, which not only causes loss of life and infrastructure but also takes a huge toll on the environment. Operational procedures like the Riser Gas Handling Guidelines from the International Association of Drilling Contractors (IADC) have been developed to mitigate this. However, there is still a need for in-depth research on this phenomenon.

This thesis aims to inspect the effects of gas-in-riser unloading on the pressure at the outlet, assessing different scenarios in which this can lead to complications. If no backpressure is applied at the top of the riser, it is common to assume that the pressure at the outlet is atmospheric. This is true if the two-phase flow is subsonic, meaning when the liquid velocity is lower than the sound velocity. However, as will be shown in this thesis, the flow during riser unloading can become supersonic under certain conditions, and one can no longer assume the pressure to be atmospheric at the outlet. A temporary pressure build-up will take place at the riser outlet.

To quantify this pressure build-up, a transient flow model based on the drift flux model and the advection upwind splitting method version V (AUSMV) scheme was modified to handle the transition from subsonic to supersonic flow. An extensive simulation study varying the riser depth, mud weight, and gas kick volumes was performed to predict the pressure build-up at the outlet under different conditions. The summary of results is presented in tables that can be found in Chapter 4.3.

For the simulation study, a 5-inch drillpipe in a riser with an internal diameter of 19- inch is considered. Water-based mud (WBM) was assumed as the liquid in the riser. The gas is supposed to be methane, and the gas kick density is calculated using a density model based on the ideal gas law. A no-slip condition is applied making the gas kick concentrated in the mud. The temperature is assumed to increase linearly from 4°C to 25°C from the bottom to the top of the riser. The gas kick was circulated out slowly with a flowrate of 1500 lpm. Discretization of the riser into a large number of cells (200) in combination with the slope limiter technique was used to minimize the effect of numerical diffusion. These assumptions are conservative for predicting maximum rates at the surface and are used to investigate worst-case riser unloading scenarios, as observed in oil-based mud (OBM). The simulations were performed for riser depths of 2000 m, 2500 m, and 3000 m. Mud weights of 1.2 sg, 1.5 sg, and 1.8 sg were considered for each riser depth. Initial kick volumes of 1 m3, 10 m3, 25 m3, 50 m3, and 100 m3 were introduced at the bottom of each riser for all three mud weights.

The simulation results demonstrate that during unloading, supersonic flow occurs at the riser outlet for all the mud weights in the riser with a kick volume of 10 m3 or more. This supersonic flow happens as the sound velocity is seen to be significantly low for a two-phase mixture near the riser outlet during unloading. These supersonic flow conditions result in a temporary pressure build-up for around 40 seconds at the riser outlet. This is the first attempt to examine this phenomenon using the drift flux model; thus, further investigations are recommended.