A new three-level well-life system is helping cement-job designers to diagnose problems in advance to build cement sheathes that maintain integrity throughout the life of a well.

Comparison between results of a physical experiment and the 3-D simulator after 20 seconds. The actual fluid profile from the experiment is shown on the right and the simulated fluid profiles from the 3-D simulator are shown on the left.

Two advances in software development have had positive impacts on well-life systems designed and developed to furnish effective zonal isolation throughout the life of producing wells. One software package presents diagnostic routines that estimate the stress levels likely to be placed on well-cement sheaths. The second is an advanced computational fluid dynamics (CFD) simulator that models, in three dimensions, multiple aspects of mud displacement during cementing.

These software systems empower oil and gas well cement-job designers to visualize accurately the damage mechanisms that can destroy the integrity
of set cement sheaths. Conventional emphasis has been placed on cement slurry properties; events that occur during the life of the well dictate redirection of the emphasis.

Background

The software discussed in this article supports an “intelligent” three-level well-life system (WLS) that “defends” the integrity of the cement sheath throughout the life of the well.

A condition known as sustained casing pressure (SCP) presents a constant challenge to the oil and gas industry; gas leaked through the cement sheath travels to the surface and keeps measurable pressure on the casing. Further, loss of pressure containment in localized areas of the well may also pose significant problems even if such loss does not result in flow to the surface. Such damage mechanisms can result in production loss and reduced effectiveness of stimulation treatments.

Level 1. The first level of defense requires analysis, design, and delivery of a cementing system that has been engineered to withstand the forces exerted on it by the various well construction, stimulation, and production operations. This level should include determination of cement properties under triaxial conditions and finite element analysis (FEA) of the casing, cement, and surrounding formation. Delivery should include cementing best practices to maximize mud removal and minimize formation of gas channels through the unset cement.

Level 2. The uncertainty of future well events dictates a cement installation that will react and respond (RAR) to cement sheath-failure events. RAR systems include the addition of swelling compounds to the slurry; such additives will swell in the presence of flowing gas or crude oil to fill cracks, voids, or channels resulting from cement sheaths debonding from the casing or well bore.

Level 3. Use of swellable-element tools is the third component of the “intelligent” zonal isolation system. A swellable element is installed on the outside diameter of the casing before it is run in; one or more of these devices can be used in a well. These tools use in situ temperature, wellbore fluids, and flowing reservoir materials as the fuel to swell the element. When required by wellbore conditions, the elements swell to fill in uncemented areas, incongruities, or void space. The rubber in these tools can swell to more than twice its original size to fill significant volumes. The swelling element will remain dormant, ready to react in the event of failure.

Design software

The “intelligent” WLS is supported by a software program that accounts for damage from phase to phase such as drilling, cementing, and displacement. Outputs are the stress state for the wellbore components and a comparison to failure criteria — tensile, compressive, and debonding. WLS is a process for improving safety and economics of oil and gas wells by supporting these functions:
• Model cement sheath properties that can withstand stresses from well operations during the entire life cycle of the well;
• Develop cement formulations to meet the properties;
• Design placement procedures and deliver the engineered cement system; and
• Evaluate and monitor the cement job performance through logs and reservoir performance over the life of the well.

3-D displacement model

The Problem. To avoid costly remedial work and/or production delays, operators and engineers seek to remove drilling fluid prior to final displacement of the cement slurry. To help optimize mud removal, usually the drilling mud is circulated for extended periods, followed by a spacer fluid that creates a beneficial fluid-to-fluid interface to enhance mud displacement.

Over time, fluid intermingling may inhibit the capability of a fluid to perform its intended purpose, for example, the intermixing of drilling fluid with cement slurry may lead to contamination of the cement. Such contamination may cause a failure of the cement to set, a lower than anticipated top of cement, and a cost increase due to extra waiting time or remedial repair.

Only one simulator in the industry predicts cement location and integrity in a fully 3-D wellbore environment. The simulator provides cement-job designers the capability of three-dimensional viewing of mud displacement success during a proposed cementing operation. The effectiveness of the simulation has been verified by visually comparing simulator solutions to the results obtained from physical experiments (flow device in Figure 1) bearing the same parameters that were used as input to the simulator.

In a validation exercise, the 3-D simulator was modeled with the same parameters as in the physical experiment (Figure 2). The fluid being displaced (e.g., drilling fluid) was characterized by the color blue and the displacing fluid (e.g., spacer fluid or cement slurry) was characterized by the color red. The color contours are explained in the legend. The figure presents a comparison between the 3-D simulator and an experiment as a function of elapsed time. They show the varying responses of annular flow as a function of offset on both the wide gap and narrow gap of the annulus. The figures also show that the 3-D simulator is comparable with the physical experiments over the processing time tested. Understanding the displacement phenomenon and having the ability to simulate it provides insight into estimating the extent of contamination and intermixing over time to help predict potential causes of stringing or channeling. Visualization using a 3-D simulator can help improve the reliability of cement operations.

Displacement tests with different fluid combinations were conducted over a range of pump rates and annular geometries. Visualization images of the movement and evolution of the fluid-to-fluid interface were used to validate the 3-D simulator.

A Solution. The fluid interface model presented here allows operators to predict causes of cement slurry contamination and mud channeling, thereby helping operators avoid these occurrences in a well. The simulator is also formulated on a general curvilinear coordinate mesh system whose boundaries can conform to highly eccentric annuli or localized washouts. This capability enables prediction of required material volume, potentially reducing the cost of an operation.

The associated proprietary 3-D visualization tool kit with interactive movie playback controls enables operators and engineers, for the first time, to effectively monitor how the mud, spacer, and cement interfaces move and evolve over time. These complex mixing interfaces are also characterized by a “best fit” rheological optimizer that can dynamically and accurately characterize the fluid viscosities within the wellbore environment. The optimizer allows for determination of the “true” top of cement, the effect that fluid intermingling may have on downhole pressures, and the extent of mud channeling.

To enable the simulator, the planner furnishes well depth, borehole diameter, casing diameter, casing standoff, drilling fluid (mud) properties, spacer fluid properties, cement slurry properties, and pump rate schedule. No assumptions of these parameters are made, so simulator solutions are not influenced by user tendencies or desire to put a “slant” on the outcome.