Summary Performance of fields under thermal recovery processes has been improved considerably by experimenting with new techniques and innovative operational solutions in heavy-oil and extra heavy oil fields (Butler 1994; Singhal et al. 1998). By far, steam-assisted-gravity-drainage (SAGD) and cyclic-steam-stimulation (CSS) methods have been the most compelling and have shown a successful history in these oil fields (Butler et al. 2000; Butler 2001). At the Celtic heavy-oil field, Husky Energy experienced three phases of optimization over the course of 10 years of recovery. The Celtic field consists of 28 SAGD pairs located in the Lloydminster heavy-oil region that have been in service since 2001. Among these pairs, there are three SAGD pairs (F Pad East) with early productivity challenges in addition to downtime caused by sand production. Following new geological realization and reservoir studies, the injector wells were converted to production wells performing a horizontal CSS process. In 2008, and in an effort to maintain production, a detailed numerical-simulation study was conducted to compare different steam support configurations and the result was justified for field execution. The outcome resulted in four directional steam-injection wells (G Pad) permanently started to support the current horizontal producers. In 2009, the new configuration commenced and is still in operation today. Communication with horizontal wells has been confirmed on one side of the pattern, and it has been extending gradually to the other side as it was investigated in other work (Edmunds 1991; Al-Salhi et al. 1997). Recently, acquired field data suggest that the process is exceeding the expected oil profile and its steam consumption is less than anticipated. The geological review, history of the F Pad East pairs, results of numerical-simulation study, and recommendations are presented in this paper.
Summary The Lloydminster, Saskatchewan, area Pikes Peak steam project has been onproduction since 1982. A key part of the development strategy was use of anundeveloped pressure-isolation wall to allow the two halves of the pool to bedeveloped with different time schedules and exploitation processes. Use of thepressure-isolation wall allowed different exploitation options to be used, butduring late project life the oil left in the wall needed to be recovered toenhance project economics. A significant challenge in recovering the oil in thewall was the lack of pressure containment or oil saturation on both flanks,resulting in the need for a gentle recovery process to avoid pushing the oiloutside the wall. On the basis of numerical-simulation results, the recoveryprocess selected for the wall was use of two cyclic-steam-stimulation (CSS)horizontal wells operated in a gentle manner. After the horizontal wells weredrilled, however, temperature logs showed the reservoir surrounding one of thehorizontal wells had already been heated by offsetting steam injection, andthat well has been continually produced at good rates without the need forsteam stimulation. The second new horizontal wall well was initially operatedsuccessfully using the CSS process designed using numerical simulation, butthat well is now also being operated in an external drive mode. The results ofthis field study show that the oil in a wall used to pressure isolate sectionsof a steam-project pool can be recovered economically without pushing the oilinto the adjacent depleted areas by use of processes designed to account forlocal wall conditions.
The Husky-CNOOC Madura Limited (HCML) MDA-4 exploration well (2011) in the Madura Strait region targeted Globigerina limestones in the Mundu Sequence (3.8 Ma) and the Paciran Sequence (2.0 Ma). The MDA Field is covered by Merpati 3D Seismic (2005). Seismic features observed from the 3D volume include phase change or polarity reversal at the top of gas filled reservoirs of the MDA structure and DHI flat-spot approximating to the gas-water contact (GWC). The reservoirs are primarily planktonic foraminifera grainstones, packstones and wackestones that have been deposited as pelagic rains and were subsequently redistributed by sea floor bottom currents. Differentiating the Mundu and Paciran Sequences relies heavily on biostratigraphy and chronostratigraphy, as there are no significant lithological features that can be observed between the sequences. This article introduces a method to construct detailed well correlations of the two sequences based on Mundu–Paciran Nannofossil Zones (MPNZ), using high resolution biostratigraphy events. The methodology uses varying nannofossil abundances in the interval NN18 (Late Pliocene) to NN11 (Late Miocene). The best reservoir performance in the study area may occur in the MPNZ-7 and MPNZ-6, which were deposited at the late stage of the depositional cycles.
