Tar mats are encountered in many Middle East carbonate reservoirs. In cases in which tar acts as a flow barrier between water and light oil, water injectors work most efficiently when placed horizontally immediately above the tar mat to maintain pressure support during production. As this strategy is pursued, a real-time tar-detection method is needed while drilling. The low mobility caused by tar can be detected with a formation-pressure-while-drilling (FPWD) tester, while a nuclear-magnetic-resonance (NMR) tool provides oil-viscosity estimation.
Introduction
The Ratawi limestone reservoir was discovered in 1963 in the offshore Neutral-Zone Concession area between Saudi Arabia and Kuwait. The production of light sour crude oil commenced in 1966. After an increase in gas/oil ratio and a sharp drop in reservoir pressure occurred, dump-water injection was started in 1975. Until recently, peripheral dump-water injection was used.
Geologically, the reservoir is divided into three units, Units A, B, and C from top to base, respectively. Of these, Unit C is the main producing unit. The top of Unit C is considered as a possible sequence boundary/unconformity. Units A and B represent initial flooding over Unit C. The two units are composed of alternating porous limestone and less-porous or tight limestone, and show overall transgressive-system tract sequence.
The heavy-oil zone was recognized above the oil/water contact. Additional drillstem tests and investigation of the separation between shallow and deep resistivity indicated fieldwide distribution of the heavy-oil zone. Wireline NMR and pressure data indicated presence of a tar/asphaltene zone in the heavy-oil zone.
The operator’s strategy was to carry out peripheral water injection, placing the injectors above and as close as possible to the heavy-oil/tar zones, enabling better sweep and volumetric displacement of light oil toward potential producers. The reservoir characterization around the injector locations is critical for assessing carbonate heterogeneity, diagenetic uncertainty, asphaltene presence, and a transition zone of heavy-to-light oil. Assessment of risk-evaluating conventional-wireline-log responses, seismic attributes, and nearby well dynamic information resulted in good prognosis of the area for deciding the entry point or landing point of the injectors and was crucial in well placement.
Offshore operations did not allow evaluation of the tar/asphaltene zone in each well because of higher operating costs from drilling a pilot hole. In recent wells, the placement was achieved on the basis of mobility steering as close as possible to the heavy-oil/tar zone.
Methodology
The drilling bottomhole assembly consisted of conventional logging-while-drilling (LWD) sensors such as gamma ray, density, neutron porosity, and resistivity. In addition, other LWD mobility-indicating technologies were used, such as FPWD and NMR. Real-time transmission of LWD-NMR and FPWD data was important to identify low-mobility layers (tar) and thus enabled the operator to make quick decisions regarding the well plan, if necessary.
For a detailed discussion of the LWD-NMR tool and its use, please see the complete paper.
FPWD. The FPWD was used to measure formation pore pressure and to calculate mobility in real time. FPWD is widely used now for detecting heavy/immobile fluids in combination with LWD NMR in Middle East carbonates. As the heavy-oil layer is approached, it can be observed that mobilities decrease to the range of 0.1–0.2 md/cp. Zones containing tar are generally characterized by lost seals or supercharged pressures. On occasion, the hole size is also seen to be slightly enlarged, on the basis of the caliper log.
Reservoir Navigation. As indicated previously, the aim was to place the wells above the heavy oil and as close to it as practicable. The heavy oil is located within the transition zone, characterized by a move from low to high water saturation (Sw). This can be seen by a reduction of resistivity that is not wholly uniform among nearby offset wells (Fig. 1 above).
A geological model is created on the basis of the offset-well data. Reservoir-navigation services use the resistivity profile in conjunction with formation dip and dip azimuth to forward model the resistivity response along the planned well trajectory. Other formation-evaluation curves are modeled along the well trajectory, but these are geometrically modeled rather than forward modeled. Before drilling, numerous well trajectories can be modeled. This enables the user to predict LWD-curve response in different scenarios such as entering the heavy oil or exiting the roof. Before drilling, no NMR data were available from offset wells.
Each well in the drilling campaign was landed at a safe distance above the transition zone. Because the heavy-oil contact is undulating, the well was planned to be drilled at 90°. The TVD of the well was increased in steps until either the NMR or FPWD tool indicated that the well was entering the heavy-oil zone. The geological model was used to correlate the real-time formation-evaluation curves with the model curves, to ascertain the well’s stratigraphic position within the reservoir. Real-time bound-water (BW) and bulk-volume-movable-fluid (BVMF) data from the NMR tool, together with the formation-pressure and mobility data from the FPWD tool, are also shown in the geological model. This enables the user to visualize the changing mobility and BVMF at different TVDs and measured depths (MDs) along the well path.
Results
A number of wells were drilled using a combination of NMR-while-drilling and FPWD tools. In each well drilled, the NMR-while-drilling tool successfully identified the transition zone between free, movable oil and heavy oil. The Results subsection of the complete paper summarizes the findings of one of these wells. In each well drilled, the objective was to place the well as close to the heavy oil as possible to increase the volume of hydrocarbon available for production. To achieve this, the TVD was increased in steps until an indication of heavy oil was seen on the NMR or FPWD tool. In each well drilled, the inclination was decreased to 86° to confirm the presence of heavy oil in relation to the TVD of the wellbore.
Fig. 2 shows a geological model that contains both modeled and real-time formation-evaluation data. The plot contains a geological curtain section that illustrates the position of the well within the reservoir. The modeled stratigraphy is segregated into three zones within the reservoir:
- Zone A: An area of movable hydrocarbon (i.e., high BVMF and low BW).
- Zone B: A transition zone that corresponds to a decreasing trend of BVMF and an increase in BW.
- Zone C: A heavy-oil zone that has minimal BVMF and a large BW volume.
The expected range of formation pressures was between 2,700 and 2,800 psi, with expected mobilities of between 0.2 to 20 md/cp (less than 1 md/cp in the heavy-oil zone and greater than 10 md/cp in the light-oil zone).
The NMR and neutron porosity show good agreement within Zone A. Before Zones B and C were penetrated, 2,000 ft of good-quality reservoir was drilled in Zone A. The NMR tool indicates low BW. Porosities varied between 17 and 19 p.u., and both NMR and neutron porosities are consistent over this interval, although both formation pressure and fluid mobility do vary over this interval.
Formation pressure decreases as TVD increases along the wellbore. The formation pressure at the beginning of the lateral section is 3,026 psia and decreases to 2,732 psia at the base of Zone A, immediately above the transition zone. There are two further formation-pressure points within the transition zones of 2,716 psia. The current peripheral water injector is placed on the western flank of the reservoir. The gradual reduction in formation pressure from landing to toe seems to be a response to carbonate-reservoir lateral changes in heterogeneity. Better carbonate facies were encountered in midsegment, and toward the toe, before dropping the angle to test the heavy-oil zone. The well placement was marginally in an updip direction, and grainstone-type facies were observed in low-formation-pressure areas toward an apparently better area of the slope.
Fluid mobility, on the other hand, does not have a clear trend, and it varies in each zone along the well path. Before the use of NMR while drilling, well placement was based primarily upon resistivity data and formation-pressure and fluid-mobility measurements. It was assumed that low mobility equated to an increase in the hydrocarbon viscosity and subsequent reduction in fluid mobility. All wells drilled in the campaign illustrate that a low mobility reading does not necessarily equate to low BVMF or high BW volume. Likewise, fluid mobility is as high (or higher) in the transition zone as in the movable-hydrocarbon zone. This suggests that fluid mobility and formation pressure cannot be used in isolation for identifying the zone of transition from free, movable hydrocarbon to heavy oil or tar.
The LWD-NMR tool also provides a permeability index that was used as a qualitative indication of the presence of heavy oil. The permeability index has good agreement with the FPWD values. As the well enters the transition zone, the permeability decreases. This reflects the increasing viscosity of the hydrocarbon. The NMR readings were as expected within the reservoir with high BVMF (Zone A). There is excellent agreement between NMR and conventional porosity in this zone. No excess bound fluid and no supercharged pressures from FPWD indicated that Zone A is a good reservoir, with movable hydrocarbon with variable fluid mobility.
As the well moved into the transition zone (Zone B), a shift toward the left (toward faster relaxation times) was apparent. These were the first indications of heavy oil, but there were no unique tar indicators. Further indication of heavy oil comes from the excess bound fluid, reduction in BVMF, and a decrease in the permeability index.
As the well entered Zones B and C, few FPWD samples were obtained because the tool could not obtain a seal with the formation. This could be partly attributable to the presence of heavy oil and, in some instances, an overgauge hole. A further indication of the presence of heavy oil in some of the wells drilled was an increase in the caliper reading up to 0.7 in.
These indications proved to be very useful in making decisions in real time. The well plan was revised to navigate the well out of the heavy-oil zone. After the depth of the tar zone was identified, the rest of the well section was maintained above it. A total of 2,000-ft MD was drilled horizontally without penetrating the heavy oil. Gradually increasing the TVD of the wellbore meant that an additional 11 ft of reservoir was available to be flushed, maximizing hydrocarbon recovery.
This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 164282, “Hydrocarbon-Mobility Steering for Optimum Placement of a Power Water Injector Above Tar Mats—A Case Study From a Light-Oil Carbonate Reservoir in the Middle East,” by Craig Saint and Thomas Glowig, Baker Hughes, and Ashis S.S. Swain, Nasser A. Al-Khaldi, Mohd. H. Al-Otaibi, Abdul Aziz Al Ghareeb, and Ahmad Bader Al Bader, Khafji Joint Operations, prepared for the 2013 SPE Middle East Oil and Gas Show and Exhibition, Manama, Bahrain, 10–13 March. The paper has not been peer reviewed.