Summary Increasing flooding-solution viscosity with polymers provides a favorable mobility ratio compared with brine flooding and hence improves volumetric sweep efficiency. Flooding with a polymer solution exhibiting elastic properties has been reported to increase displacement efficiency, resulting in a sustained doubling of the recovery enhancement compared with the use of conventional viscous-polymer flooding (Wang et al. 2011). Flooding with viscoelastic-polymer solutions is claimed also to increase recovery more than expected from changes in capillary number alone (Wang et al. 2010). This increase in displacement efficiency by viscoelastic polymers is reported to occur because of changes in the steady-state-flow profile and enhancements in oil stripping and thread formation. However, within the industry there are doubts that a genuine effect is observed, or that improvements in displacement efficiency occur with field-applicable flow regimes (Vermolen et al. 2014). In this study, we demonstrate that flooding with viscoelastic-polymer solutions can indeed increase recovery more than expected from changes in capillary number. We show a mechanism of fluctuations in flow at low Reynolds number by which viscoelastic-polymer solutions provide improvements in displacement efficiency. The mechanism, known as elastic turbulence, is an effect previously unrecognized in this context. We demonstrate that the effect may be obtained at field-relevant flow rates. Furthermore, this underlying mechanism explains both the enhanced capillary-desaturation curves and the observation of apparent flow thickening (Delshad et al. 2008; Seright et al. 2011) for these viscoelastic solutions in porous media. The work contrasts experiments on flow and recovery by use of viscous and viscoelastic-polymer solutions. The circumstances under which viscoelasticity is beneficial are demonstrated. The findings are applicable to the design of formulations for enhanced oil recovery (EOR) by polymer flooding. A combination of coreflooding, micromodel flow, and rheometric studies is presented. The results include single-phase and multiphase floods in sandstone cores. Polymer solutions are viscoelastic [partially hydrolyzed polyacrylamide (HPAM)] or viscous (xanthan). The effects of molecular weight, flow rate, and concentration of the HPAMs are described. The data lead us to suggest a mechanism that may be used to explain the observations of improved displacement efficiency and why the improvement is not seen for all viscoelastic-polymer floods.

Abstract This work describes two types of laboratory chemical enhanced oil recovery (EOR) flooding experiment. Results are presented for micromodels and core floods (Bentheimer sandstone). Saturation behaviour is followed directly by optical microscopy in the micromodels and these images enable interpretation of the saturation development in the sandstone core as determined using spatially resolved nuclear magnetic resonance (NMR). The oil is a simple alkane (decane) but the surfactant combination is designed for EOR. The work is supported by surfactant phase behaviour, oil-water interfacial tension (spinning drop), and effluent analysis (surfactant concentration). The data and analysis presented here confirms much of that previously inferred general behaviour, using modern direct observation techniques (NMR and microfluidics). These micromodel data directly illustrate local behaviour of middle phases within the porous network, the mobilisation of ganglia, and the formation of middle phase from ganglia. The correlation of micromodel data with spatially resolved core data is striking, and allows identification of behaviours within the core. An aqueous formulation of anionic surfactants and butan-2-ol is injected into decane-saturated porous media. The formulations were selected to give rise to equilibrium L1 and L3 phase behaviours at different NaCl concentrations. The aqueous formulation contains 1% of each of a C12,13 alcohol-propoxy-sulfate and a C20-24 internal olefin sulfonate (supplied by Shell Chemicals) and 8% butan-2-ol. This formulation with 2% and 4% NaCl give rise to L1 and L3 phase behaviour and oil-water (measured) interfacial tensions of 0.10 and 0.006 mN/m, respectively, which at the flow rates used give corresponding capillary numbers of 3.5×10-5 and 6.0×10-4, compared to a surfactant-free flood value of ~ 1×10-7. Both surfactant formulations provide improved oil displacement from the core. Adding surfactant in “L1-based” formulations reduces interfacial tension. However, “L3-based” formulations give ultra-low interfacial tensions and hence better recovery, exhibiting complex behaviour consistent with (1) the formation of, and then (2) the displacement of, microemulsion phases.

Single-phase flows of viscoelastic polymer solutions in both microfluidic devices and rock cores exhibit apparent flow thickening. We demonstrate that this thickening occurs above a critical Deborah number corresponding to the onset of spatio-temporal fluctuations. These fluctuations are observed to occur over a broad range of spatial and temporal scales consistent with elastic turbulence. The fluctuations provide a previously unreported mechanism for enhancing the displacement of a second, capillary trapped, immiscible phase.

Abstract A laboratory study on core plugs from a carbonate field confirmed the efficacy of an alkaline surfactant (AS) formulation at reservoir conditions. Continuous monitoring of remaining oil saturation (ROS) in short core plugs using spatially resolved nuclear magnetic resonance (NMR) can provide insights into the processes by which surfactants release post-water-flood remaining oil. In single short plugs, volumetric averages do not correctly quantify remaining oil due to capillary end effects (high remaining oil saturation post water flood observed at the outlet face of the plug), oil banks, and other flow heterogeneities. These effects are all quantifiable by NMR, allowing accurate monitoring of ROS. Here, consistent ROS values are obtained in the laboratory, in single-well pilots, and in single-well chemical-tracer (SWCT) tests. Laboratory core floods with nuclear magnetic resonance (NMR) monitoring at low magnetic fields therefore provide a valuable calibration of the NMR logs, in addition to the direct assessment of ROS. Corroboration across multiple length scales, spatial resolution, and correspondence with preferred logging measurements, all contribute to the value of NMR-monitored core-floods as a complement to field pilot studies.

Surfactant formulations that give rise to ultra-low interfacial tensions are used to enhance the recovery of crude oil by eliminating the trapping effects of capillary forces. The formulations, when mixed with oil, often give microemulsion-type phase behaviour. Depending on the surfactant affinity for oil and water, the microemulsion may exist as water-in-oil (L2), oil-in-water (L1), or bi-continuous (L3). Here, an aqueous formulation of anionic surfactants and butan-2-ol that would give rise to Ll and L3 phase behaviours at different NaC1 concentrations is injected into decane-filled porous media. The displacement processes in rock cores are followed by spatially resolved NMR and in microfluidic models by optical microscopy. Both surfactant formulation types enhance the oil displacement from the core but the L3-forming system shows more complex behaviour, consistent with first the formation and then the displacement of microemulsion phases, which leads to significant enhancements in oil recovery. (C) 2014 Elsevier B.V. All rights reserved.

We present a modified set of recommendations for the use of paramagnetic doping agents in nuclear magnetic resonance (NMR) piloting operations of oil recovery. On occasion, manganese has been used to provide transverse T-2 relaxation time contrast in NMR well logs. Aqueous fluid-phase doping agents may become important in future piloting operations, where accurate predictions of the doping agent behavior under adverse conditions are critical. Previously, chelated manganese-EDTA was recommended for use in clay-based drilling muds and sandstones to prevent surface adsorption, whereas (unchelated) manganese was recommended for use in clay-free carbonates due to the economic advantage of the simpler formulation and weak temperature dependence. Here we review the mechanisms of paramagnetic relaxation contrast and then present examples from a case study of laboratory-scale NMR measurements of crude-oil recovery from a microporous limestone. Based on our observations, we amend the previous recommendations. When the use of paramagnetic doping agents is not advisable, alternative methods for introducing fluid-phase contrast are available. Heavy-water substitution is a practical solution at the laboratory-scale and in single-well pilots.

Magnetic resonance imaging (MRI) is used to quantify in situ the recovery of crude oil from a strongly oil-wet microporous limestone core-plug. We demonstrate the capability of low-field MRI to continuously monitor oil saturation distribution by obtaining a series of spatially resolved transverse relaxation time (T2) distributions using the robust spin echo single point image (SESPI) profiling method to obtain T2 maps with a temporal resolution of 45min. These T2 maps are shown to provide comparable data to nuclear magnetic resonance (NMR) well-logs. The low injection rate of 1.4×10−3cm3s−1 (equivalent to an interstitial velocity of 1ftday−1 in the formation) allowed a large number of T2 maps to be acquired during the flood. Fluid-phase discrimination is achieved here in the T2 dimension; the brine relaxation time is reduced by addition of paramagnetic manganese. Some manganese is lost through adsorption on the limestone surface, but sufficient relaxation contrast is obtained to position an unambiguous oil/brine T2 cut-off. The spatial distributions of both the brine and oil are therefore determined simultaneously and independently. Capillary end effects are observed in the short core-plug due to the difference in wettability and permeability between the plug faces and the core-holder end-caps. The inclusion of the spatial dimension in the experiment allows a region of the plug, unaffected by end effects, to be considered representative of behavior in the reservoir. Overall, we highlight the importance of spatial resolution in laboratory-scale core analysis and demonstrate the capability of low-field MRI spectrometers to continuously monitor oil recovery experiments.

Quantitative in situ monitoring of oil recovery from sedimentary rock is demonstrated for the first time using advanced two-dimensional (2D) nuclear magnetic resonance (NMR) correlation measurements on a low field spectrometer. The laboratory-scale NMR system was chosen to provide a common physics of measurement with NMR well-logging tools. The NMR protocols are used to monitor recovery of a heavy Middle East crude oil from high permeability sandstone plugs using a brine (water) flood followed by chemical enhanced oil recovery agents: polymer and alkaline–surfactant–polymer solutions. 2D correlations between relaxation time (T 1, T 2) and apparent self-diffusion coefficient D app are used to obtain simultaneously a volumetric determination of the oil and aqueous fluid-phase saturations present in the porous material. The T 1 − T 2 and D app − T 2 correlations are bulk measurements of the entire rock core-plug; excellent agreement is shown between the measures of remaining oil (from NMR) and recovered oil (from gravimetric assay of the effluent). Furthermore, we introduce the capability to measure spatially resolved T 2 distributions on a low field spectrometer using a rapid frequency-encoded y − T 2 map. A non-uniform distribution of remaining oil is observed due to viscous instabilities in the flowing liquids; the final oil saturation ranges from \({S_{\rm o}^{\rm{(final)}} \approx 0}\) to 20 % along the direction of flow. These results highlight the quantitative nature of the NMR data obtainable in low field NMR core analysis and also the importance of spatially resolved measurements when studying short core-plugs.

Before possible multi-well pilots and full field-scale implementation, chemical enhanced oil recovery (EOR) projects usually follow an incremental screening procedure from the laboratory through single-well pilots. Laboratory-scale core floods at reservoir conditions validate surfactant performance before committing to field trials. However, in single short plugs, volumetric averages do not correctly quantify remaining oil due to capillary end effects, oil banks, and other flow heterogeneities. Magnetic resonance imaging (MRI) protocols extend traditional core floods and allow quantitative resolution of non-uniform saturation on millimetre length scales. At low magnetic field strengths, MRI data correspond directly to the measurement physics of nuclear magnetic resonance (NMR) logging tools. NMR and dielectric dispersion tools are ideal for quantifying remaining oil in single-well field trials, being independent of Archie parameters which are uncertain under complex salinity and wettability distributions. We report a laboratory study on short core plugs from a North Oman carbonate field, confirming the efficacy of an alkaline surfactant (AS) formulation at reservoir conditions. Good injectivity of surfactant was seen in selected reservoir material, suggesting that surfactant formulation was not responsible for inconsistent injectivity seen in single well field tests using a logdrill-inject-log pilot protocol. The low-field MRI results corroborate oil saturation changes seen by wireline logs on a decimetre length scale in the single-well field test. Remaining oil determinations can thus be compared at various length scales ranging from millimetres (MRI), to decimetres (log-drill-inject-log pilots), and finally to field scale trials using single well chemical tracer tests.

Abstract Laboratory and single well pilot nuclear magnetic resonance (NMR) logging results are obtained for an enhanced oil recovery (EOR) project using a common physics of measurement at both scales. Screening for chemical EOR efficacy usually begins in the laboratory before moving to single or multiple well field pilots. Laboratory experiments were conducted on a low-field bench-top NMR magnet with fluid injection protocols that matched the log-inject-log concept of the single-well in situ EOR evaluation (Arora et al. 2010). In situ measurements of the oil and brine saturations during the flood, by "diffusion-editing" protocols and by relaxation measurements alone, are shown to be in quantitative agreement with gravimetric assays of recovered oil. Spatially resolved T2 analysis (the laboratory equivalent of a standard NMR well log) revealed non-uniform oil saturation during the EOR process in short core plugs. The final remaining oil distribution is confirmed by a reservoir simulation in a geometry identical to the NMR-compatible core holder. In a region of the core-plug not influenced by end effects, complete recovery of the oil was observed, consistent with the the single-well in situ EOR evaluation. The quantitative estimates of remaining oil, using a tool-equivalent NMR protocol, demonstrate the potential for NMR logging of remaining oil in monitoring wells completed with NMR-transparent casing.

Oil and water fractions have been identified in fluid saturated carbonate rock cores using a novel T2-T1-d pulse sequence. The inclusion of the chemical shift dimension d allows T2-T1 plots to be generated independently for the oil and water. The T2-T1-d pulse sequence utilises a “double-shot” T1 measurement that provides free induction decays (FIDs) as a function of both relaxation times for suitably broad line samples. The T2-T1-d data set is acquired in the same experimental time as a conventional T1-T2 measurement (without chemical resolution) of equivalent data density. Here we demonstrate that different behaviour can be observed between the oil and water fractions in water wet and preferentially oil wet cores, and that saturation states can be determined. This technique could provide a quantitative NMR measure of wettability.