K.L. Badgett, P.L. Hill, W.H. Mills, S.P. Mitchell, G.S. Vinson III, K.L. Wilkins
BP Exploration Inc.
Houston
An interdisciplinary team at BP Exploration Inc. is combining geoscience, reservoir, production, and drilling technology to target horizontal wells for MC 109 field in the Gulf of Mexico.
Horizontal wells have proven to be extremely helpful in management of MC 109, which lies in more than 1,000 ft of water beneath platform Amberjack, 18 miles from the mouth of the Mississippi River. The wells have provided access to reserves isolated by depositional features within the reservoir at a cost equal to or less than that of conventional drilling.
They afford an unconventional look at the reservoir, helping the team to better understand its complexities and the effects of those complexities on reservoir management.
In addition, horizontal drilling experience is contributing to an evolving model of how to plan, drill, and complete similar wells for other BP projects in the Gulf of Mexico.
The interdisciplinary team is responsible for field management, which encompasses reservoir description, reservoir development, and well productivity.
Rather than developing goals specific to the various disciplines, then compromising to produce a group goal, team members start from a field management plan that includes common goals and shared responsibilities. They quickly incorporate individual well results into each activity and use the resulting analysis to modify the overall plan.
With this approach, the team has overseen the drilling of 30 wells from the Amberjack platform in less than 2 1/2 years. The development has 38 completions and no dry holes.
RESERVOIR DESCRIPTION
MC 109 field is a combined structural and stratigraphic trap with two main reservoirs: the G and J sands.
The sands are middle Pliocene shelf-margin deltas bounded to the east by a major north-south trending fault system. The hydrocarbon accumulation is divided into three fault blocks by a series of east-west trending, down to the north, postdepositonal faults that cut the sequence but do not appear to have influenced sedimentation.
Structural dip on both reservoirs is to the southwest (Figs. 1, 2). Core and dipmeter data indicate that each reservoir is a stratigraphically complex series of delta mouth-bar progrades.
A generalized vertical sequence consists of prodelta silts overlain by a slumped mouth-bar facies of contorted sands and shales, which are capped by a prograding distal mouth-bar facies of flat-lying sands interbedded with muds and silts.
The prograding mouth-bar facies is visible on 3D seismic acoustic impedance (AI) data as a series of shingled trough reflectors separated by a low-amplitude zone.
These features - clinoforms - are believed to be associated with the contemporaneous slumping and progradation of deltaic mouth-bar facies. The low-amplitude zones between sandrich clinoform bodies have been mapped in detail (Fig, 3).
Initially, the composition of these breaks in seismic character were unknown and were considered potential barriers to flow. During the early development of the field, large oil - water contact changes - about 100 ft - and associated changes in fluid properties - API gravity and initial GOR - were observed within each fault block. This information indicated that some of the clinoform boundaries, or low-amplitude zones, acted as barriers in the reservoir.
Horizontal drilling and analog studies facilitated investigations into the nature of the clinoform boundaries. The A-5 horizontal well was targeted to penetrate three clinoform bodies, thereby cutting two clinoform boundaries.
Overlaying the measurement-while-drilling resistivity and interpreted clinoform boundaries on seismic extracted along the well bore shows that the seismic clinoform boundaries correlate with low-permeabihty siltstones (Fig. 4). Boundaries of this sort are thought to behave like baffles, impeding flow but not stopping it.
Comparison with the shelf-margin deltaic system of the Cretaceous Ferron sandstone in Muddy Creek Canyon, Utah, suggests that other boundaries may be syndepositional faults. These boundaries are more likely to act as flow barriers.
Collection and analysis of produced fluid data (API gravity, and GOR), pressure transient data, and varying oil-water contacts from the development wells confirm that each fault block is subdivided into smaller reservoirs (Fig. 5).
But not all clinoform boundaries are barriers to flow.
It is currently difficult to predict whether a clinoform boundary wig act as a barrier or baffle. In this environment, horizontal wells snow penetration and production from two or more clinoforms with one wellbore. This allows additional recovery from isolated clinoforms and increases production rates by accessing clinoforms separated by silt baffles.
MODEL NEEDED
Compartmentalization of the reservoirs within each of the three fault blocks in the field emphasized the need for a detailed reservoir model.
The central fault block, El (Fig. 3), was performance-modeled as a pilot study for waterflood and in preparation for later development of the full field model.
During construction of the detailed reservoir model, the team imported 3D seismic AI data to the geological modeling software. This allowed lateral (75 x 75 ft) and vertical (15 ft) resolution similar to the seismic data and thus provided significantly greater detail over conventional methods based on well control.
The ability to adequately map reservoir quality delta mouth-bar facies and identify areas of low permeability and reservoir separation on seismic AI data focused attention on the G Sand reservoir for early modeling.
Good correlation of wire line AI to seismic AI provided the basis for developing reservoir attributes. Quantitative rock property correlations, derived from interpretation of cores, logs, seismic data, maps, and fluid data, were applied to the seismic AI data set to predict reservoir properties. These properties included porosity, permeability, water saturation, and rock type (Fig. 6).
The model-building process used geological modeling and visualization software at each step to ensure proper calibration with well results. Seismic rock properties from the model agreed well with those from individual wells.
The resulting reservoir model has been run with VIP reservoir simulation software to aid in waterflood visualization and reservoir management issues.
Results of the pilot study have improved the team's understanding of the reservoir architecture within the field and helped in near-term waterflood planning.
Work is under way to incorporate stratigraphic features observed in well logs, which are below seismic resolution, in the model. Field performance data and initial simulation results suggest that these features have considerable effect on reservoir performance.
The team is considering a Combination of deterministic and stochastic methods to provide subseismic detail for future modeling.
RESERVOIR DEVELOPMENT
With new and existing technology, BP has successfully drilled and completed horizontal wells in relatively thin pays in unconsolidated formations. It has been able to keep horizontal well bores in target cylinders with radii as small as 20 ft.
The basic drilling plan for BP Gulf of Mexico horizontal wells calls for drilling a 12 1/4 in. borehole in the intermediate section to the top of the pay horizon and setting a 9 5/8 in. casing. An 8 1/2 in. production interval through the reservoir follows. Prepacked screens are installed through the completion interval. The wellbore is allowed to collapse around the screens as the well is drawn down and produced (Fig. 7).
The MC 109 team selected drilling and completion fluids carefully, conducting lab tests of several drilling fluids for use in horizontal sections. The drilling fluid had to minimize damage, clean up easily, produce no environmental harm, and meet drilling requirements with respect to lubricity and hole cleaning.
Return permeability tests under various simulated confining stresses clearly favored a sized salt drilling system (SSDS). Field results exhibited excellent carrying capacity in the horizontal section as the hole remained free of excess torque or drag,
Well planning has become one of the most critical aspects of the MC 109 project. It provides consistency in development planning and establishes the framework for detailed, well-by-well planning.
BP has successfully drilled and completed 8 horizontal wells in the Gulf of Mexico in the past 16 months. Trajectories have varied from short horizontal sections (1,000 ft) at 89 to long inverted sections (2,800 ft) at 95.
Although all of the wells varied with respect to geologic objectives, most were similar with respect to drilling and completion.
WELL PRODUCTIVITY
Each horizontal well has generally presented a new set of reservoir issues for consideration. Although the completion type has been similar in all the wells, the completion method has varied to accommodate the reservoir architecture and drilling results.
The initial horizontal well, A-5, was drilled where tight constraints on the reservoir were provided by previously drilled conventional wells.
The clinoform boundaries were thought to be potential barriers, but seismic data did not clearly indicate that the clinoforms were isolated. Later wells targeted clearly isolated clinoforms and were drilled without the benefit of control points from nearby wells.
In the process, the wells have spawned changes in designing the bottomhole configuration, modifed the placement of directional measurements, and challenged limitations to effectively complete wells where the borehole dropped into the water column.
The main application of horizontal wells at MC 109 is to access reserves that are isolated by multiple clinoform boundaries. However, additional production benefits are gained through standoff from the oil-water contact in thin oil columns and more efficient well completions.
The net result is additional recoverable reserves, fewer wells, and accelerated production rates (Fig. 8).
BIBLIOGRAPHY
Johnston, R.A., Porter, D.A., Hill, P.L., Smolen, B.C., "Planning, Drilling, and Completing a Record Horizontal Well in the Gulf of Mexico," OTC-7351, Offshore Technology Conference, 1993, pp. 853-860.
Lindsay, J.F., Prior, D.B., Coleman, J.M, "Distributary-Mouth Bar Development and the Role of Submarine Landslides in Delta Growth, South Pass, Mississippi Delta," AAPG Bull., Vol. 68, No. 11, 1984, pp. 17321743.
Mayall, M.J., Yeilding, C.A., Oldroyd, J.D., Pulham, A.J., Sakurai, S., "Facies in a Shelf-Edge Delta-An Example from the Subsurface of the Gulf of Mexico, Middle Pliocene, Mississippi Canyon, Block 109," AAPG Bull., Vol. 76, No. 4, 1992, pp. 435-448.
Mills, W.H., Mitchell, S.P., Rigler, J.L., Vinson, G.S. III, Wilkins, K.L., "Mississippi Canyon 109: Reservoir Characterization of a Shelf-Edge Delta," unpublished.
Copyright 1994 Oil & Gas Journal. All Rights Reserved.
