Tuning of Computer Model Parameters in Managed-Pressure Drilling Applications Using an Unscented-Kalman-Filter Technique

Abstract

Summary To manage the annular pressure profile during managed-pressure drilling (MPD) operations, simulations performed with advanced computer models are needed. To obtain a high degree of accuracy in these simulations, it is crucial that all parameters describing the system are as correct as possible. A new methodology for real-time updating of key parameters in a well-flow model by taking into account real-time measurements, including measuring uncertainty, is presented. Key model parameters are tuned using a recently developed estimation technique based on the traditional Kalman filter. The presented methodology leads to a more-accurate prediction of well-flow scenarios. Although the present study is motivated by applications in MPD, the idea of tuning model parameters should be of great importance in a wide area of applications. The performance of the filter is studied, using both synthetic data and real measurements from a North Sea high-pressure/high-temperature (HP/HT) drilling operation. Benefits by this approach are seen in more-accurate downhole-pressure predictions, which are of major importance for safety and economic reasons during MPD operations.

Real-time reservoir model updating in thermal recovery: Application of analytical proxies and Kalman filtering

“…The work by Zhai et al (2009) on forecasting inter-wall connectivity in waterflooding, or the work by Gravdal et al (2010) for tuning computer models in drilling are among the limited examples. The latest generation of the KF is the ensemble Kalman filter (EnKF) (Evensen, 1994) which is suitable for nonlinear systems with large number of variables.…”

Pressure regulation with kick attenuation in a managed pressure drilling system

“…In [1], [2] a sophisticated two-phase flow simulator is used to predict the downhole pressure based on the drift-flux formulation. In [3], an unscented Kalman filter exploiting downhole measurements is used to tune the predicted pressure loss due to friction in both the drill string and the annulus. To facilitate model-based observer and controller design, low order models for the downhole pressure have recently been developed [4].…”

Adaptive observer design for the bottomhole pressure of a managed pressure drilling system

“…This tuning can be done by an experienced operator or by using automatic tuning methods such as parameter estimation algorithms. In [13] an unscented Kalman filter is used to update the friction estimate in both the drill string and the annulus. The scheme uses a measurement of the bottomhole pressure to update the parameters every 30 seconds.…”

“…qpump is the volume flow through the mud pump. qbit is the volume flow through 47th IEEE CDC, Cancun, Mexico, Dec. [9][10][11]2008 WeB13.5 (14) == qbit + IlPp, (13) and (6) can be written as in Pa is small as Md is greater than M a . Treating AI as a known constant considerably reduces observer complexity.…”

“…In this section convergence properties of Pbit == Pbit -Pbit will be proved. Motivated by (13) an estimate of Pbit is…”