Experimental setup at Texas A&M Dual Gradient Drilling Lab aims to study riser gas migration in controlled mud level drilling system
Scaled model of marine riser allows for analysis of gas behavior; next phase of project will include quantification of the escaping gas rate
By Omer Kaldirim and Jerome J. Schubert, Texas A&M University
Growing demand for energy is pushing operators to drill in ever-increasing ocean depths. Deepwater offshore reservoirs often exist where the required pressure from the drilling mud for well control is very near the pressure at which overlying layers can break down, resulting in loss of circulation of drilling mud to thief zones, inviting dangerous kicks of high-pressure gas and oil. Such narrow drilling pressure windows leave razor-thin margins for error.
Mud-pressure constraints on well control and formation breakdowns, combined with the usual unknown drilling parameters, require continued research to expand the range of deepwater reservoirs that can be technically and economically exploited. The tremendous expense of deep offshore drilling creates challenges to perform research on actual wells. This article describes a study on riser gas migration, a critical safety and technical challenge in deepwater drilling. It discusses aspects of research needed to address the issue.
Managed pressure drilling (MPD) methods have been useful in mitigating some difficulties. However, issues still exist that need physical descriptions and solutions. One deepwater drilling issue is the migration of the gas that enters the well. The volume, density and pressure are easy to quantify using current methods. Yet, at great depths, the added pressure causes gas to condense and enter the drilling mud as a liquid. Such entrained gas is difficult or even impossible to detect until it vaporizes again in the riser pipe. This high-pressure flash of gas jeopardizes the strength integrity of the riser pipe. Depending on the mud pressure and flow rate in the riser, gas can suddenly come out of solution with massive increases in volume. Research is needed to develop technology to manage this.
At the Dual Gradient Drilling Laboratory at Texas A&M University, systems are being developed to experimentally study riser gas migration. The objective is to analyze gas behavior through experiments using a scaled model of a marine riser. The model is based on controlled mud level (CML) drilling, a variation of dual-gradient drilling where the mud is pumped out of the riser at a pre-determined depth to reduce the drilling mud level and to create a section at top of the riser evacuated to air and open to the atmosphere. This creates a dual-density fluid-column with air and drilling mud.
With the CML method, riser gas migration becomes an even more important subject. Migrating gas can collect at the top section in the riser, creating an explosive mixture with air at low pressure. Further, the sudden gas expansion in the riser takes place earlier. This expansion can create a force great enough to push mud to the top of the riser and create pressure spikes, which can lead to the loss of integrity of the well. Additionally, the integrity of the riser becomes compromised due to the imbalance of the external and internal pressures.
The riser model at Texas A&M simulates the pumped riser section of the CML riser pipe to study gas migration and methods to mitigate this issue.
The model is scaled down from a 19.5-in. riser to a 6-in. clear PVC pipe with a 2-in. PVC pipe located at the center to circulate mud. The clear PVC allows the fluid level to be monitored and for gas migration and behavior to be observed. A 6-in. clear PVC pipe tee is placed 7 ft from the base to divert the flow from the annulus to a pump in order to simulate the pumped riser section. These pumps are controlled using variable frequency drives and a software called LabView. Two pressure transducers are located on the riser model at the base and at the center of the tee to measure pressure change. Flow rate is measured at the inlet and outlet for the mud entering and leaving the system. A gas injection port was installed at the base. The system is monitored and controlled using the aforementioned software, as well as Data Acquisition Tools, a device used to transfer data from transducers and flow meters to the computer. Additionally, a top fill system, a 1-in. PVC pipe system to feed mud to the top of the riser model, has been installed to fill the annulus at the top.
In the initial phase of the study, tests were performed using water as the mud phase and air as the gas phase. The main purposes of the study were to (1) calibrate the measurement instruments and controllers, and (2) to establish corresponding fluid levels for various pump rates. Once the initial step was completed, experiments began by injecting gas at various rates to test the gas-handling capability of the discharge pump. It was determined that the pumps are capable of handling the gas concentration range of interest.
In the next phase, observational tests were performed, and data was recorded for further analysis. The mud level was set to predetermined heights, and gas was injected to create the desired gas-liquid concentrations. Rates between 80 gal/min and 170 gal/min and a gas injection rate of 0.5 standard cu ft/min were pumped. A top fill rate of 1.8 gal/min was used to create a downward flow to slow the gas migration. The experiments were recorded using two video cameras, one located at the discharge point and the second located just above the discharge point.
Initial observations showed smaller gas bubbles were easier to eliminate from the system. As the bubble size increased, diversion became more difficult. Keeping the gas injection rate constant, improved dispersion of smaller bubbles was observed. With low circulation rates, the gas phase transitioned from small dispersed bubbles to larger independent, rapidly rising bubbles. With the larger gas bubbles, only small trailing gas bubbles of negligible volume were diverted. From observing the discharge point, it was noticed that a higher velocity flow was created by reducing diameter from 6 in. to 2 in. As the gas bubbles migrated to the discharge point, the migration rate increased. At the discharge point, the flow direction for bubbles changed from vertical to horizontal to exit the system at an increasing velocity. Any bubbles that escaped and migrated past the discharge point were slower due to the downward suction. It was also observed that the top fill had no effect on gas migration due to the low flow rate.
In the next phases of study, the plan is to perform tests using engineered drilling mud. The experimental setup also will be redesigned to provide measurements of the gas eliminated from the system. This will allow the gas rate escaping to the voided section of the riser to be quantified. The discharge pump also will be optimized and automated, depending on the desired fluid height and circulation rate. Pump capacities and top-fill capacity will be reevaluated and redesigned to provide greater flow rates. The number of pressure measurement points will be increased to study the change in pressure as the gas bubble migrates along the riser. This measurement will provide insight into riser gas migration rate as the gas bubbles move up the riser.
These experiments will need to be verified and scaled up in order to better simulate an actual riser gas migration case. At that point, simulation software and computational fluid dynamics will be used. These simulations will initially be designed to simulate the drilling laboratory and redesigned to scale up the model to be used in an actual case with engineered mud and methane. The ultimate goal is to be able to physically describe riser gas migration for both extreme cases and benign cases. DC
This article is based on a presentation at the 2017 IADC/SPE Managed Pressure Drilling and Underbalanced Operations Conference, 28-29 March, Rio de Janeiro, Brazil.
