According to the US Geological Survey Fact Sheet 70-03, "The estimated volume of technically recoverable heavy oil (434 billion bbl) and natural bitumen (651 billion bbl) in known accumulations is about equal to the Earth's remaining conventional (light) oil reserves." Many producing companies have avoided heavy oil due to the expense and technical challenges associated with producing it. However, due to the increasing global demand for oil and the current economic climate, more operating companies are beginning to enter the heavy oil arena.
In heavy oil zones, conventional wireline logs (i.e., resistivity, acoustic, density, neutron) may indicate the presence of oil in the reservoir. At best, analysis of the conventional logs provides the quantity of heavy oil. However, the conventional log responses are insensitive to oil viscosity, the key property that controls the producibility of heavy oil. Moreover, many heavy oil reservoirs are shallow and contain fresh to variable-salinity connate water. This causes uncertainty in the conventional saturation models.
Previously, nuclear magnetic resonance (NMR) techniques have had limited success in heavy oil fields due to a lack of contrast between heavy oil and bound water. However, a combination of the latest-generation NMR logging instrument, new acquisition techniques optimized for heavy oil and innovative new interpretation techniques have proven NMR to be a powerful tool in heavy oil reservoirs. NMR logging can reduce saturation uncertainties in heavy oil reservoirs relative to conventional, resistivity-based saturation analysis. Moreover, NMR is unique in the ability to provide an estimate of in-situ viscosity.
Until recently, most borehole NMR measurements and interpretation techniques have been focused on the transverse relaxation time, T2, of the reservoir fluid. The T2 of oil is known to vary inversely with viscosity; as the viscosity of oil increases, its T2 decreases. As illustrated, heavy oil is characterized by very short T2 relaxation times, as are capillary-bound water (BVI) and clay-bound water (CBW). If there is a lack of T2 contrast between two fluids, such as heavy oil and CBW, T2 alone is insufficient to characterize and quantify the fluids. Figure 1 shows that ultra-heavy oil, such as bitumen or tar, is characterized by such a short T2 relaxation time that it may not be measured by NMR logging tools.
Historically, simple analysis techniques have been used to identify heavy oil by combining conventional measurements with NMR measurements. Because heavy oil is characterized by very short T2 times, it is commonly included in the NMR clay-bound water volume. If the NMR log indicates large quantities of clay-bound water in a reservoir interval where the conventional logs indicate that the reservoir is clean, the interpretation is that the reservoir fluid is heavy oil that is affecting the NMR clay-bound water volume. In ultra-heavy oil reservoirs, if the T2 of the oil is so short that it isn't measured by the NMR logging tool, the result is a porosity deficit between conventional porosity measurements that include the ultra-heavy oil volume and NMR porosity that does not include the ultra-heavy oil volume. The porosity deficit may be interpreted as the volume of tar or bitumen in the reservoir.
More advanced NMR acquisition and analysis techniques have been developed which aid in differentiating between water and oil in the reservoir. The NMR acquisition techniques involve acquiring data with multiple inter-echo times (TE). This enhances the contrast between the diffusivity of water and oil. The analysis techniques then identify the fluids based on their diffusion contrast. Just as the T2 of oil varies inversely with viscosity, the diffusivity (D) of oil also varies inversely with viscosity. Therefore, there is a large diffusivity contrast between heavy oil and bound water. However, bound water and heavy oil are both relatively insensitive to diffusion-based measurement techniques; the dominant decay mechanism for heavy oil is bulk relaxation, while the decay of bound water is dominated by surface relaxation.
The MR Explorer (MREX) instrument acquires NMR data using multiple magnetic field gradients. Because the diffusion-sensitive NMR decay mechanism is dependent on the combination of TE and D, the magnetic field gradient, the tool is suited to enhance the diffusivity contrast between heavy oil and bound water. Baker Atlas developed a new acquisition sequence specifically optimized to increase the sensitivity of bound water and heavy oil to the diffusion decay mechanism. This allows the direct differentiation between heavy oil and bound water with NMR-based measurements. The logging tool and the acquisition mode have recently been successful at quantifying heavy oil in Venezuela and Kazakhstan.

Hamaca heavy oil reservoir
The primary clay in the Hamaca reservoir, kaolinite, is characterized by a low specific surface area and small CEC. These features cause the clay-bound water in the Hamaca reservoir to relax at longer T2 times than is typical for clay-bound water. The T2 cutoff typically used for identifying clay-bound water is 3 microseconds (ms), while the clay-bound water in the Hamaca reservoir is characterized by T2 relaxation times of approximately 4 ms to 6 ms. The heavy oil in the Hamaca reservoir is characterized by T2 times in the range of approximatey 1 ms to 6 ms. Where there is overlap between the heavy oil and clay-bound water in the T2 spectrum, 2-D NMR techniques are used to differentiate between the fluids based on diffusivity.
An additional feature of the 2-D NMR images is the capability to determine something about the quality of the heavy oil. In the petroleum industry, oil is characterized by a single value of viscosity or gravity. These numbers do not provide any insight into the different constituent components of the oil. However, the NMR T2 spectrum is capable of providing such information.
The four 2-D NMR images in Figure 2 come from the heavy oil-bearing sands in the Hamaca reservoir. The heavy oil signals in maps (a) and (c) appear to have a broader distribution than in maps (b) and (d). However, the heavy oil in map (a) is different than the heavy oil in map (c). Map (a) shows more intensity below a T2 time of 2 ms than above it.
Conversely, map (c) shows more intensity at approximately 6 ms. This indicates that the heavy oil in map (a) contains heavier components with lesser amounts of lighter components, while the heavy oil in map (c) contains lighter components with lesser amounts of heavier components. The heavy oil signal in map (b) has a narrower distribution with all of the signal at or below 2 ms. This indicates that the heavy oil in map (b) is composed dominantly of heavy components.
The data yields 2-D NMR images as a continuous log. The T2 relaxation time of the heavy oil, determined from the 2-D NMR images, may be used to estimate the in-situ viscosity of the heavy oil. Because the images are acquired as a continuous log, the in-situ viscosity of the heavy oil can be presented as a continuous profile. The heaviest components of the heavy oil from the Hamaca reservoir correspond to a T2 time of 1.3 ms, or an in-situ viscosity of 2,200 centipoise (cp). Also, because the images are acquired on a continuous basis, they are useful for identifying fluid contacts (Figure 3).

Buzachi heavy oil field
Even with the advances in NMR tools and acquisition techniques, there are still reservoir conditions under which it may be difficult to distinguish between heavy oil and extremely fast-decaying bound-water components due to insufficient diffusivity contrast. In these cases, integrating the NMR log with conventional logs can improve the interpretation quantitatively.
The NMR data was analyzed using two different techniques, SIMET, a forward modeling-based multiple inversion technique, and 2-D NMR imaging. The heavy oil components which could be identified with both techniques, without integrating the NMR log with conventional logs, were deemed to be the lighter heavy oil components and were quantified as VHO,L. In order to quantify the heavier components of the heavy oil, VHO,H, a non-NMR clay-bound water constraint, is applied.
The non-NMR clay-bound water constraint is determined by quantifying the shale volume from conventional logs. Unlike earlier methods that relied on sole use of the gamma ray log to quantify shale volume, the clay-bound water constraint is determined using a combination of gamma ray, spontaneous potential, a combination of neutron and density porosity, and resistivity logs. Using the conventional log shale volume determination as a constraint, the difference between the NMR-based clay-bound water volume and the conventional log shale volume is deemed to be the heavier components of the heavy oil.
One of the primary reasons for logging the MREX service in Kazakhstan was to determine permeability. Because heavy oil is characterized by very short T2 relaxation times, it will be included in the clay-bound or capillary-bound water volumes rather than in the movable fluid volume. As a result, the NMR permeability estimate will be too low. The permeability can be recomputed, including the heavy oil in the volume of movable fluid. This results in a more representative permeability estimate.
Combining the latest-generation NMR logging instrument with new acquisition techniques specifically optimized for heavy oil has improved the success of NMR logging in heavy oil reservoirs. In addition to providing saturations and permeability as continuous log profiles, NMR remains the only technology sensitive to the property that controls the producibility of the oil, viscosity. Additionally, the 2-D NMR images provide insight into a distribution of the constituent components of the oil rather than using a single value such as viscosity or gravity to characterize the oil.

Comments