This work focuses on the development of specific methodologies to support managed-pressure-drilling (MPD) operations implemented on real-time diagnostic software. The developed methodology proposes as output ideal operating parameters, such as choke pressure, pump-flow rate, or adjustments related to drilling-fluid properties. The methodology considers the best approach to meet the restrictions imposed by the operational window. The real-time optimization procedures proposed are a further step to ensure reliability of MPD operations in challenging scenarios.

Real-Time MPD Optimization

Specific modules were developed from a drilling-problem diagnostic software to provide real-time support for MPD operations. The primary advance represented by these modules is the consolidation, in a computational tool, of MPD best practices. The tool, which uses real-time drilling data as input, will propose and enable immediate action in the event of anomalous situations or ones not predicted in project planning.

Dynamic Pore-Pressure Test (DPPT) and Dynamic Formation Integrity (DFIT) Procedures. These procedures allow a more-accurate estimate of values for pore and fracture pressures, respectively. These procedures are suited particularly for drilling in narrow operating window scenario, because they allow correction of geological-profile data obtained previously with current field geopressure. Thus, the probability of gain and loss events becomes smaller in an MPD operation when compared with a conventional drilling scenario.

During these tests, in each pressure step performed in the choke, the software calculates the volume balance related to the compression or decompression of the fluid in the well. The software calculates the volume that decreased from the active tank (DFIT) or increased in the active tank (DPPT). The software also calculates the volume balance related to the compression (DFIT) or decompression (DPPT) of the fluid in the well. These calculations determine whether these volume variations, both real and expected, follow the same magnitude. The difference between them is the criterion used to stop both procedures (DPPT and DFIT).

In Petrobras’ drilling operations, this proposed limit normally is set to 2 bbl. This preset threshold is a minimum acceptable value to consider that a gain or loss is occurring. If this difference surpasses the limit value, the module triggers a message warning the rig crew.

After these procedures, the simulator updates the operating window, using real downhole data [effective circulating density (ECD)] in a conservative way. The logic adopted for updating is to use the largest value between the DPPT and the value of the pore gradient present on the project. For the fracture gradient, inverse logic is adopted; the lowest value between the DFIT and the fracture gradient obtained on the project is used. Fig. 1 illustrates the pore-gradient update.

Criteria and Calculations of the MPD Optimization Module. Real-time optimization of MPD control is suggested by the software when any of the following conditions along the entire well are reached:

• ECD or equivalent-static-density (ESD) profiles approach a preset tolerance limit to the fracture or pore gradient
• A change in the fluid properties has taken place
• A new point has been collected for the operating window from DPPT or DFIT procedures

Once fed the operational data, the software can execute the optimization module. During the execution, the following steps are performed:

1. Calculate the ECD or ESD profiles using the geometry, fluids, and operating parameters inserted
2. Calculate area indices
3. Calculate new surface backpressure (SBP) and flow-rate values for the same anchor point
4. If necessary, calculate new set-point values and anchor-point position

Operational Parameters Methodology. Some limits are previously configured to ensure that the suggested results are in accordance with operational reality. The maximum and minimum limits of each parameter include

• Minimum and maximum opening of choke valve
• Maximum rotating-control-device pressure
• Maximum and minimum flow rate allowed by bottomhole-assembly equipment
• Minimum flow rate to reach desired minimum solids transport ratio

A range of project data is used to perform an analysis before the optimization to ensure that the inserted data obey the criteria imposed by the limits. Every 10 seconds, in real time, the software calculates ECD and ESD profiles, evaluating the differences from the operating-window limits. When any parameter appears outside these preset boundaries, optimization should be performed to redesign the operational parameters to meet actual limits. The optimization uses variables that do not vary along the well (flow rate and fluid weight) and variables that are constant in a specific position (choke pressure and anchor point).

Further manipulations described in the complete paper make possible the comparison between the stipulated operational limits with the variables involved in the calculations.

Calculations of ECD and ESD. The ECD and ESD calculations are a simple conversion from the pressure values (calculated by the hydrodynamic equations) to the equivalent density. These parameters are important because the fracture and pore gradients are also expressed in equivalent density. Thus, verification, if the conditions imposed on the operational parameters are within or outside the desired limits, becomes easier.

Anchor Point. This term refers to a given depth in the well at which a certain pressure set point is to be maintained during operations. In this way, drilling can be performed more safely, particularly when the operating window is very narrow. The control of this pressure is achieved by the manipulation of the choke pressure, taking into account the injection flow rate through the column and through the booster, and the different fluids that might be inside the well and their position. During the optimization process, the anchor-point calculation determines which flow-rate and choke-pressure values can set the desired pressure value at the anchor point. The calculation is discussed in the complete paper.

Area-Ratio Index. The main objectives of optimizing operational parameters are to find a set of values that meet the defined operational limits and to ensure that the results of calculated ECDs and ESDs keep a certain distance from the operating window. In this study, a method was selected that can achieve the latter objective. An equation relates the areas formed by the ECD and ESD profiles, and the evaluation limit is entered by the user. From this ratio between areas, a numerical optimization can guide how equations can be solved to obtain the smallest possible area ratio.

MPD Optimization Methodology. To execute the optimization routines, methods such as Broyden-Fletcher-Goldfarb-Shanno and sequential least-squares programming allow the restriction of bounded values, helping to obtain results that remain within the limits of the operational parameters. Within the optimization routine there is a sequence of parameters that will be used as primary variables, sorted according to the criteria used during the operation for the sequence of the modified parameters.

Results

Before running the optimization, determining whether the simulator truly represents the pressures along the well is critical. To that end, the anchor point is defined at the pressure-while-drilling sensor depth, and the set point is defined as the current ECD value. If the SBP is similar to the MPD-choke-sensor reading, the model is considered validated.

For the evaluation of module results, a synthetic scenario with values typically observed in Brazilian locations was considered. This scenario represents a drilling situation in which, at a given point, a circulation loss is observed owing to proximity to the fracture gradient. At this point, the software indicates the need to run an optimization to obtain better operating parameters to eliminate fluid loss. Ideal gradient positions suggested by optimization only could be reached by changing SBP. If the ideal SBP surpassed any equipment-operating limit, suggestions for flow-in, mud weight, or anchor-point depth, in that order, would be made.

Two optimization suggestions provided by the work flow are detailed in the complete paper. A breadth of operational parameter values can be used without compromising the MPD operation.

Conclusions

Experience obtained in several deep­water MPD drilling operations proved that a redesign of the drilling project must be performed in a range of situations, especially if an unexpected gain or loss event occurs. The MPD monitoring module was created in the context of a real-time drilling software that calculates hydraulic parameters almost instantaneously with real data. This methodology allows rapid project redesign and, consequently, the avoidance of nonproductive time, bringing more reliability and flexibility to MPD operations.