Search

Summary Some crude oils contain naturally occurring surfactants that avoid hydrate agglomeration. Natural hydrate antiagglomeration has been linked to different crude oil fractions, including asphaltenes. Asphaltenes can promote the formation of stable water-in-oil (W/O) emulsions due to their amphiphilic properties. The surfactant-like behavior of asphaltenes is related to their aggregation state. Asphaltenes are strong emulsifying agents when in an aggregated state but weak emulsifying agents when either precipitated or well solubilized in the bulk oil phase. The asphaltene aggregation state may be artificially modified, changing its interfacial activity, by mixing crude oil with heptane–toluene mixtures. This work investigated the influence of the asphaltene aggregation state on gas hydrate agglomeration. Results show that the natural hydrate antiagglomerant properties of crude oils can be highly dependent on the artificially induced asphaltene aggregation state. For instance, if asphaltenes were induced to be solubilized into the bulk oil phase, the natural hydrate antiagglomerant behavior was diminished. However, when asphaltene aggregation was induced, gas hydrate agglomeration was avoided. These new findings could have significant implications for the implementation of novel hydrate management strategies that can reduce or eliminate the need for external interventions and hence minimize capital and operational expenditures by taking advantage of the intrinsic natural antiagglomerant properties of some crude oils.

High-pressure and low-temperature conditions during subsea pipeline transportation favor the formation of gas hydrates and may create challenging flow assurance problems. Once gas hydrates have plu...

Summary Hydrates present a major flow-assurance problem because of the relatively fast timescales at which they can form, grow, agglomerate, and plug a flowline. Thermodynamic inhibitors (THIs) are a common chemical strategy for preventing hydrate formation. However, THIs can be cost prohibitive because the water content (WC) in the flowline and the seawater depth increase. Among the alternative hydrate-management strategies, there is the use of anti-agglomerants (AAs), which are active at considerably lower concentrations than THIs. AAs prevent agglomeration and allow hydrates to flow as a noninteracting slurry in subsea oil flowlines. The selection of the correct AA dosage is critical for safe and economical AA application. This work uses one model oil and two crude oils to study hydrate agglomeration in the presence of a commercial AA in a laboratory-scale autoclave equipped with particle-characterization devices. Despite known limitations, the particle-sizing probes used here provided unique insight into the hydrate-agglomeration process occurring at high pressure in crude oil systems, enabling the identification of different characteristics of a hydrate slurry based on the AA concentration and the oil phase used. For example, the droplet size before hydrate formation was a poor indicator of AA performance for the AA and oils tested. Systems with similar droplet size before hydrate formation yielded significantly different hydrate-agglomerate sizes. Similarly, increased or decreased emulsion stability did not show any correlation to AA performance. The effective AA dosage was observed to be lower for model oil systems compared with crude oil systems, which provided insight into using model oils to evaluate AA performance vs. crude oils. Results also indicated that when an effective AA dosage is applied, hydrate-agglomerate size was similar for all oil phases tested. Further parameters need to be taken into account such as impeller geometry, pressure, temperature, and WC, which may influence AA performance.

Hydrate bedding is defined as the gravitational segregation of hydrate particles, leading to their accumulation at the bottom of the pipe. Previous research has shown that hydrate bedding is a phys...

Wax deposition on inner surfaces of pipelines is a costly problem for the petroleum industry. This flow assurance problem is of particular interest during the production and transportation of waxy oils in cold environments. An understanding of known mechanisms and available thermodynamic models will be useful for the management and planning of mitigation strategies for wax deposition. This paper presents a critical review of wax prediction models used for estimation of wax deposition based on chemical hydrocarbon compositions and thermobaric condition. The comparative analysis is applied to highlight the effective mechanisms guiding the wax deposition, and how this knowledge can be used to model and provide solutions to reducing wax deposition issues. One group of thermodynamic models assume that the precipitated wax is a solid solution. These models are divided into two categories: ideal (Erickson and Pedersen models) and non-ideal solutions (Won and Coutinho models). In the other group of models, the wax phase consists of many solid phases (Lira-Galeana model). The authors summarized the limitations of the models, evaluated, and identified ways to represent the overview of existing thermodynamical models for predicting wax precipitation. Within the strong demand from industry, the results of this manuscript can aid to aspire engineers and researcher.

Removing stagnant liquid in a loaded horizontal gas well remains an unsolved challenge. Current practices for horizontal well deliquification are limited in terms of reliability and continuity, resulting on increased OPEX and CAPEX, behind down time and additional equipment installation. Experimental evaluation of a proposed artificial lift method for horizontal well deliquification, showed average removal efficiencies of 75% of the stagnant liquid volume. The experimental facility consisted of an experimental flow loop, that replicates conditions of liquid-loaded horizontal wells, with a horizontal section of 40 feet and a vertical section of 40 feet. The method is based on the chamber lift principles, using intermittent injection of gas at high pressure and low volumetric flow rates to the horizontal section of the well. Removal efficiency increased by 12% by using saccharidic additives and sodium chloride, to increase the surface tension between the injected gas (compressed air) and the liquid (water). This work presents a mechanistic model of the proposed artificial lift method, based on the momentum balance of the gas and the liquid slug flowing along the horizontal and vertical sections of the system, including numerical regressions for the prediction of the surface tension and viscosity of the liquid mixture as a function of temperature and the concentration of the tested additives. The model is used to determine the required available injection pressure at surface, and the location of the valve mandrel, as same as to estimate the removed liquid volume, discharge volumetric rate, and discharge pressure of the liquid slug at the surface facilities. The model is validated against experimental data obtained from the experimental flow loop.

Gas hydrates constitute a serious flow assurance problem. Over the last decades, industry has faced this problem by using avoidance methods (e.g. injection of thermodynamic hydrate inhibitors) and management strategies (e.g. addition of hydrate anti-agglomerants). In the former, hydrates are completely avoided by shifting the hydrate boundary towards higher pressure and lower temperatures; in the latter, hydrates are allowed to form but their tendency to agglomerate is reduced. It should be noted that some crude oils are naturally able to avoid hydrate agglomeration, this non-plugging tendency may originate from the surfactant-like behavior of fractions like asphaltenes and acids. Recent works have shown that the natural non-plugging potential of certain oils can be affected by the addition of polar molecules like alcohols. There is another strategy for managing hydrate that consist of the addition of THIs at a concentration lower that the one required to full hydrate inhibition. In this case, hydrates are under-inhibited. Studies carried out on hydrate agglomerating systems have shown that under-inhibition might prevent hydrate agglomeration only in a specific range of THI concentrations and sub-cooling; however, work on non-plugging oils is scarce. In this paper, the hydrate agglomeration of two crude oils under-inhibited with methanol and MEG was evaluated through a visual rocking cell apparatus and a high-pressure rheometer. Results showed that THIs and the crude oil's natural surfactants were capable of acting synergistically in reducing hydrate agglomeration and improving the system flowability.

The rapid formation of gas hydrates in subsea tiebacks is regarded as one of the most challenging problems in flow assurance. Due to economic reasons, the hydrate controlling strategy has moved from hydrate avoidance to hydrate management. Efficient design of hydrate management techniques requires a better understanding of hydrate formation kinetics, as well as a robust transient hydrate simulation tool. In this work, we use a transient hydrate simulation tool with the ability of predicting hydrate formation in oil- and water-dominated systems simultaneously to capture complicated multiphase flow scenarios expected in oil and gas flowlines. This modeling tool can be applied to estimate hydrate formation in flowlines exhibiting both water- and oil-dominated environments at different locations, which is especially useful in modeling transient shut-in/restart operations as well as high water cut scenarios. This simulation tool can track the concentration of multiple thermodynamic inhibitors and consider the injection of anti-agglomerants through the cohesive force change between hydrate particles. This paper presents modeling results of a real black oil subsea tieback system, considering its geometry, production data and fluid properties. The hydrate formation rates and plugging risks at different water cuts, gas-oil ratios and liquid holdups were estimated both during steady state operation and shut-in/restart conditions. A hydrate blockage formed in the real field at 49 vol% water cut was successfully predicted using the transient hydrate prediction model. This simulation tool can be applied to predict the hydrate formation rate and amount under the complicated multiphase flow scenarios in oil/gas fields, and could be useful for the design and optimization of hydrate management approaches in flowlines.

An optimally designed nozzle can be used as an efficient artificial lift system to deliquify gas wells. The objective of this work is to determine the most optimum nozzle parameters and geometry that would maximize critical pressure ratio while minimizing pressure drop across the nozzle under two-phase flow conditions. Six optimum nozzle geometries, evaluated under single-phase flow conditions in a previous study, are evaluated in this work. Two separate facilities were used to conduct the two-phase (air and water) experiments: a horizontal facility with 1.5 in. inner diameter (ID) PVC pipelines and a 30 ft long lateral pipeline section, and a vertical facility with 1.5 in. ID PVC pipelines and a 30 ft long vertical pipeline section. The horizontal facility was used to study the effect of introducing water into the system as a second phase, and the vertical facility was used to simulate conditions of a gas well with a nozzle installed as an artificial lift system. Horizontal flow experimental results showed that critical pressure ratio decreases for two-phase flow relative to single-phase flow. The distance of the annular churn flow pattern observed downstream the nozzle may be related to the nozzle performance; significantly lower performance was observed for the LJ nozzle in two phase horizontal testing relative to other nozzles – one characteristic this might be attributed to is the peculiarly longer distance of downstream annular churn flow pattern observed for this particular nozzle compared to other nozzles. Based on data analysis, the ASTAR nozzle group showed the most improved performance, similar to results obtained in single-phase horizontal testing. Based on vertical experimental results, the most optimum nozzle geometry to deliquify a loaded gas well was identified as the ASTAR nozzle 2, a parabolic nozzle geometry. A significant temperature drop downstream the nozzle was observed during these experiments. It is important to account for temperature change while designing a nozzle in order to avoid flow assurance problems, such as gas hydrate formation. The trends observed in relative performance between different nozzle geometries (rankings based on critical pressure ratio and pressure drop) for two-phase vertical experiments were similar to those observed for single-phase horizontal experiments. Therefore, single-phase experiments could be sufficient to evaluate relative nozzle geometry performance based on critical pressure ratio, for the conditions used in this study.

Summary Gas hydrates are considered a major flow-assurance challenge in subsea flowlines. They agglomerate rapidly and form hydrate blockages. During transient operations [shut-in and restart (RS)], risk of blockage formation owing to hydrates can be greater compared to that during the continuous operations. In particular, hydrate formation during an unplanned shut-in and subsequent restart could lead to increased operational hazards. In this work, flow-loop tests were conducted under both continuous-pumping (CP) and RS conditions, using Conroe crude oil with three different water fractions (30, 50, 90 vol%) at 5 wt% salinity, over a range of mixture velocities (from 2.4 to 9.4 ft/sec). It was determined that RS operations resulted in an earlier onset of hydrate particle bedding—twice as fast as those in CP tests—from the interpretation of pressure-drop and mass-flow-rate (MFR) measurements. Droplet imaging using a particle vision and measurement (PVM) probe suggested larger water droplets (100–300 µm) during the shut-in, as compared to the CP tests (=40 µm) at 50 and 90 vol% water cuts (WCs). For the tests performed using a demulsifier at 200 ppm, PVM images suggested larger water droplets (mean droplet size = 94 µm), as compared with the test with no demulsifier (mean droplet size = 21 µm). The test using a demulsifier resulted in higher pressure drops and lower MFRs compared with the test with no demulsifier, indicating poor hydrate transportability when water was partially dispersed in the oil phase. The current study indicated that partially dispersed systems present greater risks of hydrate plugging as compared with the fully dispersed systems in the range of water volume fractions from 50 to 95 vol% WC, which was the phase inversion point of the water-in-crude-oil (Conroe14 crude) system. The flow-loop-test analyses presented in this work can potentially aid in an improved mechanistic understanding of RS operations, involving unplanned shut-ins and restarts.

The production from horizontal tight gas wells is affected negatively by the accumulation of liquids in the bottom of the well (i.e., liquid loading), to the point that additional hydrocarbon production is not possible. In this work we propose an artificial lift method for liquid-loaded horizontal wells, in tight gas shale reservoirs, based on intermittent gas lift principles. An experimental loop was built to test a prototype of the system and explore the operating envelope that determines the optimal liquid removal results. Experimental results are presented in terms of the liquid removal capacity of the device, as a function of the gas injection rate and the Eotvos (Eo) number. Liquid removal values (produced liquid volume / stagnant liquid volume) of up to 75%, and a volumetric efficiency (volume of liquid produced / volume of gas injected) of up to 78% was obtained for the lowest Eotvos number tested (12) at the lowest gas injection rate (0.25 cf/min). Both the liquid removal and the volumetric efficiency behaved inversely proportional to the Eo number and the gas injection flowrate. Reducing the ID of the line, proved to be an efficient approach to reduce the Eotvos number. Using this option resulted in better sweep, but in an increase of the frictional pressure drop. The improved liquid sweep effect of the Taylor bubble at lower Eo is related to a larger radius of the Taylor bubble and a change on the shape of its nose.

As the reservoir pressure drops in the vicinities of horizontal gas wells drilled in tight reservoirs, the produced gas loses its ability to drag liquids to surface. This results in the accumulation of liquid in the bottom of the well. Since conventional artificial lift methods fail to remove the stagnant liquid in an efficient and reliable manner, the accumulated liquid limits the well production down to unprofitable levels. As an alternative solution to this problem, it is proposed a method called Backsweep, which relies on the use of intermittent gas injection in the horizontal section of the well. Results obtained in an experimental flow loop, using water and compressed air, show an improvement on the liquid removal efficiency (i.e., produced liquid volume / stagnant liquid volume) at Eotvos (Eo) numbers around 12, obtained by reducing the internal diameter (ID) of the flowlines, but with increased frictional pressure drops per unit length. This work focuses on the effect of changing the surface tension and viscosity of the stagnant liquid, to optimize the liquid removal efficiency, avoiding increases in frictional pressure drop from reducing the ID of the line. Sucrose, xanthan gum and sodium chloride at different concentrations, were used for this purpose. Xanthan gum at a concentration of 0.05% w/w delivered the best results of the three substances, with an increase of over 15% (up to 77% in average) when compared to only using water, as the Eo number was reduced in 5% and the viscosity number (Nvis) was increased in over 70 times. The improvement on the liquid removal efficiency of the device at lower Eo is associated to a shaper nose of the Taylor bubble and larger convergence angle of the flow lines in the liquid film next to the wall.

Highlights • Code comparisons build confidence in simulators to model interdependent processes. • International hydrate reservoir simulators are compared over five complex problems. • Geomechanical processes significantly impact response of gas hydrate reservoirs. • Simulators yielded comparable results, however many differences are noted. • Equivalent constitutive models are required to achieve agreement across simulators. Geologic reservoirs containing gas hydrate occur beneath permafrost environments and within marine continental slope sediments, representing a potentially vast natural gas source. Numerical simulators provide scientists and engineers with tools for understanding how production efficiency depends on the numerous, interdependent (coupled) processes associated with potential production strategies for these gas hydrate reservoirs. Confidence in the modeling and forecasting abilities of these gas hydrate reservoir simulators (GHRSs) grows with successful comparisons against laboratory and field test results, but such results are rare, particularly in natural settings. The hydrate community recognized another approach to building confidence in the GHRS: comparing simulation results between independently developed and executed computer codes on structured problems specifically tailored to the interdependent processes relevant for gas hydrate-bearing systems. The United States Department of Energy, National Energy Technology Laboratory, (DOE/NETL), sponsored the first international gas hydrate code comparison study, IGHCCS1, in the early 2000s. IGHCCS1 focused on coupled thermal and hydrologic processes associated with producing gas hydrates from geologic reservoirs via depressurization and thermal stimulation. Subsequently, GHRSs have advanced to model more complex production technologies and incorporate geomechanical processes into the existing framework of coupled thermal and hydrologic modeling. This paper contributes to the validation of these recent GHRS developments by providing results from a second GHRS code comparison study, IGHCCS2, also sponsored by DOE/NETL. IGHCCS2 includes participants from an international collection of universities, research institutes, industry, national laboratories, and national geologic surveys. Study participants developed a series of five benchmark problems principally involving gas hydrate processes with geomechanical components. The five problems range from simple geometries with analytical solutions to a representation of the world's first offshore production test of methane hydrates, which was conducted with the depressurization method off the coast of Japan. To identify strengths and limitations in the various GHRSs, study participants submitted solutions for the benchmark problems and discussed differing results via teleconferences. The GHRSs evolved over the course of IGHCCS2 as researchers modified their simulators to reflect new insights, lessons learned, and suggested performance enhancements. The five benchmark problems, final sample solutions, and lessons learned that are presented here document the study outcomes and serve as a reference guide for developing and testing gas hydrate reservoir simulators.

Gas hydrate formation in oil and gas flowlines can represent a safety concern and a cause for hindered production, resulting in economic losses. Hydrate risk mitigation can be attained through hydrate avoidance or management strategies. Hydrate avoidance methods aim to keep the flowline outside of the hydrate stability region through, for example, the use of thermodynamic hydrate inhibitors. Thermodynamic hydrate inhibitors (THIs) increasingly shift the hydrate boundary towards higher pressures and lower temperatures as a function of THI concentration in solution. Hydrate management strategies allow the flowline to operate inside the hydrate stability zone, without the risk of forming a plug by using kinetic hydrate inhibitors or commercial hydrate anti-agglomerants. Some crude oils, denoted non-plugging crude oils, have naturally occurring surfactants (e.g., asphaltenes) that can behave as a hydrate anti-agglomerant and allow the formation of a transportable non-agglomerating hydrate slurry. Recent work has suggested that the asphaltene-aggregation state is a parameter that may dictate the natural hydrate anti-agglomeration behavior of non-plugging crude oils. The ability of naturally occurring anti-agglomerants to prevent hydrate plugs is limited by the intrinsic amount found in the non-plugging oils and thus depends on the amount of hydrates formed in the flowline. Therefore, there is an opportunity to use thermodynamic hydrate under-inhibition (i.e., partial thermodynamic inhibition) in non-plugging crude oil systems that can no longer safely avoid a hydrate plug while relying on natural surfactants alone. If THI under-inhibition is used to partially reduce the amount of hydrates formed, the natural anti-agglomerants can prevent hydrate agglomeration of the remaining hydrates formed. The THI volume required for under-inhibition would be lower than that of complete THI inhibition, thereby reducing operational costs. In this work, alcohols with different hydrocarbon chains and mono-ethylene glycol were shown to have a key impact on the asphaltene-aggregation state, inferred through size measurements of solvent-extracted asphaltene particles, correlating with changes in emulsion stability, and the non-plugging potential of a crude oil as assessed by rocking cell tests. The effect of alcohols on the asphaltene-particle size was also shown to be highly sensitive to the presence of a free-water phase, likely due to observed alcohol partitioning into the water phase. Alcohols with intermediate-hydrocarbon chains prevented asphaltene aggregation more effectively and reduced their emulsification tendency compared to alcohols with shorter carbon chain lengths. Furthermore, short-chain alcohols or MEG showed no antagonism when used with the non-plugging oil tested, resulting in a partially inhibited system able to avoid hydrate agglomeration at a higher water cut compared to a non-inhibited system. On the other hand, alcohols with intermediate-chain length were found to be detrimental to the non-plugging potential of the specific crude oil tested, potentially due to its effect of reducing the asphaltene-particle size in solution. More experimental work is required to better understand these phenomena and determine if other non-plugging crude oils show a similar behavior.

Hydrates are crystalline water compounds that may form in deepsea pipelines and have the ability to quickly and efficiently plug flow. Given this issue, gas hydrate formation tests have been conducted and reported in a number of published articles. However, most published data has focused on continuous "steady-state" operations and a gap exists for the more complicated, yet interesting transient shut-in/restart scenarios. Motivation to study such systems are further exacerbated given that industrial fields can operate outside the gas hydrate formation region during continuous production. Hydrates become a major concern especially during unplanned shutdowns and restarts. Therefore, this work focuses only on hydrate transient results. Industrially relevant scenarios are investigated by looking into gas hydrate formation and its effect on fluid transportability through visualization and pressure drop during shut-in and restart in a laboratory-scale flowloop. The effects of parameters such as water content and dispersion state, shear, chemical injection, and oil properties on hydrate formation under transient conditions are reviewed and general agreement is shown between the studies. Overall, the general consensus is that lower water contents, dispersed water, high shear, thermodynamic inhibitor injection, dosing of anti-agglomerants or kinetic hydrate inhibitors, higher oil viscosities, and oils with natural chemical surfactant properties can be favorable for ensuring successful restart of hydrate forming systems. Flowloop tests in a gas-dominated system show that film growth over and around the liquid pool at the bottom of the pipe could lead to plugging in a significant portion of the flowline. This could continue to build up over time (with the addition of fresh hydrate forming materials in an actual system) and inhibit gas flow. Further tests on an oil-dominated system with an intermediate water content demonstrate how shut-in conditions could potentially increase the risk of plugging compared to dispersed lower water content systems. In the former case, water accumulation to low spots in the flowloop allowed for localized areas of high water concentration, where rapid hydrate bed formation in partially dispersed oil-water sections controlled the overall fluid transportability.

The high pressure and low temperature operational conditions of deepwater subsea facilities often result in the formation of gas hydrates, which is one of the most challenging flow assurance issues. The variant water cuts and flowline geometries could result in complicated flow patterns leading to different hydrate formation mechanisms. In previous work, we have developed hydrate formation models both for oil- and water-dominated systems. The model for oil-dominated systems considers the water phase being dispersed as droplets in an oil continuous layer, and uses either a kinetics or a transport model to calculate hydrate formation at the interface of water droplets. On the other hand, water-dominated systems contain small amounts of oil, and a mass transfer-based hydrate growth model is used. This paper presents a new transient hydrate formation model that predicts hydrate growth and transportability in oil- and water-dominated environments, which might be present simultaneously in oil and gas pipelines due to changes in fluid distribution in complex multiphase flow systems. Simulations of high-pressure pilot-scale flowloop experiments at different water cuts are used to verify this modeling approach. Furthermore, the proposed hydrate simulation tool is applied to analyze the flow dynamics of a subsea tieback involving hydrate formation at varying water cuts after flowloop verification. This simulation model is targeted to be widely used to simulate oil-dominated pipelines, as well as high water cut systems leading to phase inversions, and represents a step further towards the development of more comprehensive hydrate formation models.

The current hydrate kinetics model implemented in the multiphase flow simulator OLGA treats hydrate growth in oil-continuous systems by considering the solidification of emulsified water droplets to form a hydrate-in-oil slurry that is assumed to be stable. To date, the validity of this model has not been established for gas-dominant systems, where gas void fractions can exceed 90 vol %. Here, six experimental data sets, collected using a 40-m single-pass gas-dominant flowloop operating in the annular-flow regime, were compared with predictions made using the current hydrate kinetics model. The comparison identified discrepancies in the predicted flow regime and the gas–water interfacial area that significantly affect kinetic hydrate-growth-rate calculations; these discrepancies might be due, in part, to differences in dynamic similarity between flowloop experiments and industrial-scale simulations. By adjusting only the kinetic rate scaling factor, it was not possible to match the pressure drop observed ...

That the majority of the pore sizes of nanoporous unconventional reservoirs are on the order of the molecular dimensions of large hydrocarbons of crude oils implies the possibility of molecular sieving during the flow of crude oil in these formations. The objective of this study is to provide evidence for the filtration of hydrocarbon mixtures in nanoporous media and to support the hindered transport models for flow in nanoporous, unconventional reservoirs. The hindered transport of hydrocarbons in nanoporous media may be contributed by a variety of mechanisms, such as molecular sieving, size exclusion, selective adsorption, and retention in dead-end pores. Core experiments conducted by Zhu et al. (2019) intended to examine filtration and hindered transport in its entirety. The experimental setup used in this research, on the other hand, was designed to focus on the molecular sieving component of filtration. Therefore, although hindered transport in porous media resembles depth filtration, where adsorption and retention in dead-end pores and tight pore-throats may be significant contributors of filtration, in this study, thin polymeric membranes simulated surface filtration due to molecular sieving only. Two versions of the experimental setup have been used: The first setup was a regular dead-end cell to study the effects of temperature, pressure, and CO₂ on filtration. The second setup adopted a stirred cell to alleviate concentration polarization and eliminate accumulation of heavy hydrocarbons on the filter surface. The results of both experiments have been compared and interpreted. The filtration of hydrocarbon molecules by polymeric membranes was verified. The retention efficiency of a hydrocarbon is proportional to its molecular weight. The effect of concentration polarization was significant in extended period tests using regular dead-end cell. It reduced the retention of large hydrocarbon molecules due to increased higher chemical potential from higher concentration. The effect of concentration polarization was alleviated significantly using stirred cell. The retention of large hydrocarbon molecules was higher in this scenario.CO₂ treatment of hydrocarbon mixture increased the total flux and reduced retention of large hydrocarbon molecules using regular dead-end cell. However, during a stirred cell test, CO₂ treatment of hydrocarbon mixture did not reduce retention of heavy or medium hydrocarbons and only affected the retention and flux of light hydrocarbons. The retention of large hydrocarbon molecules was proportional to filtration pressure and inversely proportional to temperature. That is, in general, the higher the pressure and the lower the temperature, the higher the retention of large molecules.An improved understanding of the factors affecting the retention of large hydrocarbon molecules in nanoporous media will facilitate developing models for the filtration behavior and shed light on potential methods to reduce retention of large hydrocarbon molecules in shale reservoirs.

Current oil prices have shifted the offshore flow assurance philosophy from "hydrate avoidance" to "hydrate management". In order to manage hydrate formation in the pipeline, not only do we need to have a good understanding of the hydrate formation process, but also a comprehensive hydrate formation predictive tool is necessary. In this work, a transient hydrate simulation tool coupled with a multiphase flow simulator, which predicts the hydrate formation rate and amount to determine hydrate slurry transportability, is applied to assess the hydrate formation risk in an offshore gas condensate subsea tieback under design stage. A simulation model is developed using the geometry, fluid properties and predicted production data from this field. This gas condensate field has an offshore separator that connects to a subsea production line. After separation, gas and liquids are transported by export lines. In this study, the hydrate formation in the production and the liquid export lines was estimated using the transient hydrate simulation tool. Simulation studies were performed to assess the hydrate plugging risk at both steady state and transient operations considering different water cuts. This simulation tool has been demonstrated to be useful in modeling hydrate management during subsea pipeline design and optimization, and can provide guidelines for safe and cost-effective hydrate management in the field.

Recently the concept of "no external gas hydrate control measures" has been proposed, whereby gas hydrate formation can occur in oil and gas subsea pipelines during steady state and transient operations, with the operational window defined by predictive analytic tools. Flow assurance engineers routinely use computer programs, including transient multiphase flow simulators coupled to a gas hydrate kinetics model to simulate gas hydrate formation and transportability. Given the complexity in multiphase flow modeling, modern machine learning technologies, especially artificial intelligence, could be applied to solve high-level, non-linear problems, such as evaluating gas hydrate risk based on measurable process parameters. In this work, several machine learning techniques, such as regression, classification, feature learning with an algorithm/framework like support vector machine (SVM) and neural networks (NN), are applied to analyze the data sets on: 1) hydrate tests conducted at pilot-scale flowloop facilities (4,500 data points), as well as 2) transient operation field data. The classification/regression model based on flowloop test data uses several independent input variables (features), such as water cut, gas-oil ratio, hydrate particle cohesive force, fluid velocity, oil viscosity, specific gravity, interfacial tension, and time in the hydrate stable zone, to output the hydrate fraction and probability of hydrate plugging in the pipeline. The semi-supervised learning model was applied based on the field data use as input, including water cut, shut-down time (where applicable), and gas-oil ratio to determine the level of hydrate resistance to flow during restart or dead oil displacement after production shut-down. The flowloop based machine learning model exhibited good prediction accuracies in test and validation processes, and was used to assess the hydrate risks in an actual field. The field data based machine learning model demonstrated the ability to construct field risk maps. The machine learning technique could be potentially applied in hydrate management to evaluate hydrate risks in subsea oil/gas pipelines. As a complement to more complex transient multiphase flow simulations, this machine learning approach can aid in the development of advanced hydrate management strategies.

The rapid formation of gas hydrates is one of the major concerns in offshore oil and gas transportation due to the possible plugging of unmanaged hydrate agglomerates. Within the oil and gas industry, hydrate management guidelines have shifted over the last decade from hydrate avoidance to hydrate management. Allowing hydrate formation in pipelines without blockages requires understanding of the hydrate formation and plugging mechanisms under different operational conditions. Furthermore, a predictive tool which can be coupled with the multiphase flow conditions encountered in the field is highly desired in order to ensure successful transportation of hydrate slurries. In the realm of modeling, one such hydrate plugging mechanism that remains unaddressed is the bedding of hydrate particles, which may occur when hydrate particles aggregate within the pipeline. In this work, a bedding model based on force balances has been developed and coupled with a transient multiphase flow simulator. It is proposed that agglomeration controls bedding and bedding is a function of the agglomerate size, which is assumed to follow a log-normal distribution based on an autoclave experimental study. The bedding model has been tested and validated against large-scale flowloop experiments at different water cuts and liquid velocities. Pressure drop comparison between the simulations and experiments shows an improvement after implementing the bedding model for simulations with low liquid velocity. This work may contribute to the understanding of the hydrate plugging mechanisms and be applied to optimize hydrate management guidelines in offshore oil and gas production.

Fractured enhanced geothermal system (EGS) reservoirs are typically sensitive to thermal and mechanical change induced by cold water injection. It has been observed that the permeability at the cold water injector is significantly enhanced. The physical thermal-hydrologic-mechanic (THM) process behind this phenomenon is that, the injection of cold water decreases the temperature of the reservoir rock and causes the matrix block to shrink, resulting in an increase of the fracture aperture and fracture permeability. Therefore, it is of great importance to quantify the effect of thermally induced fracture aperture change to better predict the behavior/performance of EGS reservoirs. In this work, we develop a novel correlation of the thermal-induced normal change of fracture aperture. The new correlation is based on the analytical solution of the governing displacement equations. Compared to the existing empirical correlations, the new correlation can better describe the physical processes by including the thermal effect on the matrix-fracture deformation. We have verified this correlation with respect to refined simulation results and implemented this correlation in a fully coupled massively parallel geothermal simulator, THM-EGS. We have applied this correlation to study field scale problems with certain parameters from Habanero Field in Copper Basin, Australia. Our results demonstrate that the fracture permeability near the cold water injector could be enhanced 7 times.

As industry moves toward the production of oil and gas resources in deep offshore environments, the prospective formation of natural gas hydrates under low temperature, high-pressure conditions poses an increasing risk of pipeline blockage. Successful management of this risk requires a precise forecast of the hydrate growth rate. Such predictions should incorporate quantitative descriptions of both deterministic and probabilistic hydrate phenomena. In this work, we present a first step towards such a quantitative risk assessment for systems that form hydrate slurries, based on a consideration of the stochastic nature of hydrate formation. Our framework introduces a new approach to risk assessment, by coupling a laboratory-derived probabilistic nucleation model with existing deterministic calculations for hydrate slurry viscosification. This new approach is used to extend a previously-described hydrate risk algorithm, the Hydrate Flow Assurance Simulation Tool (HyFAST), which enables the rapid assessment of hydrate slurry viscosification using the most advanced models available. Importantly, while the previous version of HyFAST was constrained to calculations in flowloop and autoclave geometries, for which it is sufficient to consider a single volume element, the new version of the HyFAST algorithm allows calculations for flowlines wherein a series of multiple volume elements are considered. The advanced hydrate formation models within this algorithm allow identification of critically important hydrate formation scenarios, and may serve as a screening tool for cases that then require study within rigorous hydrodynamic packages such as OLGA® or LedaFlow® to examine behaviour during key transient events. By coupling the outputs of our predictive algorithm and quantitative risk framework, we demonstrate a new capability to assess what constitutes an acceptable level of hydrate formation. We use this algorithm in a series of case studies that compare the effectiveness of common hydrate mitigation strategies over the life of a reservoir, including the optimization of a wellhead choke opening for both production rate and hydrate blockage risk.

Gas hydrates forming in deep water subsea tiebacks can stop production of offshore wells by blocking the pipeline. This is considered one of the most challenging problems in flow assurance. In this work, we present a transient hydrate simulation tool with the ability of predicting hydrate formation in oil-dominated systems and the purpose of capturing the complicated multiphase flow scenarios that occur in oil and gas flowlines. This modeling tool can be applied to estimate hydrate formation in flowlines with different water cuts, which is especially useful in modeling transient shut-in/restart operations. A field case is used as a benchmark to verify the prediction performance of the proposed hydrate simulation model. The geometry, fluid properties and production data from an offshore well is reproduced using the hydrate formation model, quantifying hydrate formation and estimating plugging risk at different liquid holdup, water cuts and produced gas-oil ratios. Hydrate anti-agglomerant chemicals were used in this field case, and are considered in the simulations of steady state and shut-in restart operations. The proposed transient hydrate prediction model successfully predicted the formation of hydrate plugs as observed in the real field. This hydrate formation simulation tool can be applied for the design and optimization of prevention, management and remediation procedures of hydrates in flowlines.

Over the past decade, the paradigm of gas hydrate research has transitioned from avoidance via thermodynamic inhibition toward management, where a limited amount of hydrate may be allowed to form in the flowline. This new paradigm enables new field development concepts, including longer-distance subsea tie-backs and limited chemical inhibition. This presentation reviews research datasets from flowloops on hydrate deposition phenomena and includes an initial analysis on the impact of shear stresses on hydrate deposits. This study integrates experimental data from unique high-pressure laboratory flowloops, including single-pass, gas-dominant and recirculating, liquid-phase flowloops. The flowloops have visual observation ports that allow video imaging to be performed throughout the high pressure tests. These apparatuses together are estimated to provide coverage over 1-200 Pa of flowing shear stress at the wall, over a range of total liquid inventories and water holdup. For this work, conceptual diagrams of deposition in oil, water, and gas-dominant systems are presented. Film growth rates are derived at the Colorado School of Mines (CSM) from visual observation in a deposition loop. Hydrate particle deposition rates are determined from a gas-dominant flowloop in Western Australia by fitting experimental data obtained from gas-dominant systems. The results demonstrate that the absolute shear stress required to prevent hydrate particles from depositing at the wall, or to remove/slough deposited hydrate from the wall, can vary with the primary fluid phase. Initial shear stress calculations from a specific test estimate that less than 5 Pa can be required to remove/slough hydrates from the wall in the aqueous phase, while more than 100 Pa may be required in a gas-continuous pipeline. This review suggests that deposition and flowing shear stress may be critical considerations in the design and operation of hydrocarbon systems, to help prevent hydrate blockage in industrial operations.

Summary This work presents a new simple algorithm for the rapid screening of hydrate plug formation risk, using experimental models of gas hydrate plug formation in oil-dominant systems. The algorithm is based on hydrate formation from an emulsified water phase, where resultant hydrate particles may interact to form large aggregates that increase slurry viscosity and pressure drop. Predictions of pressure drop were compared with a hydrate-forming industrial flow loop, resulting in average absolute deviations between model and experiment of less than 5 psi for liquid-phase Reynolds numbers of less than 75,000 and water content below 70 vol% of all liquid.

The presence of gas hydrates in subsea production is identified as a major flow assurance challenge due to the rapid formation of hydrates and their risk of causing flowline blockages. The risk of hydrate-associated blockages could be higher for transient (shut-in/restart) operations compared to continuous production. Often in offshore production, there are unplanned shut-ins that could lead to increased operational hazards. Despite the practical importance, few studies have addressed the hydrate plugging mechanisms during transient operation. To investigate hydrate transportability in transient operation, tests were performed using an industrial-scale flowloop (3.8 inch in internal diameter and 295 feet in length) with crude oil for a range of water fractions (30–90 vol.%), at 5 wt.% salinity, and across a range of mixture velocities (2.4 -9.4 ft/s). For the tests performed at 50 vol.% water cut (WC) at 5.7 ft/s, the pressure drop and mass flow rate measurements suggest that the transient (or restart, RS) test could result in an earlier onset of hydrate bedding, occurring approximately twice as fast, and may lead to a higher relative pressure drop when compared to the continuous pumping (CP) test. These flowloop experiments suggest that for the RS tests, rapid hydrate formation and agglomeration followed by the water-in-oil (W/O) emulsion destabilization could cause an early increase in pressure drop (and higher operational risks) compared with the CP tests. The tests using an anti-agglomerant (AA) additive (at 1 and 2 vol.%) were performed at 50 vol.% water fraction for all mixture velocities. The tests with an AA indicate that 2 vol. % AA (at all velocities) could prevent hydrate plugging and maintain a slurry flow at 50 vol. % WC for both RS and CP tests. The tests with 1 vol.% AA (at 5.7 ft/s for 50 vol.% WC) show larger fluctuations in pressure drop in the RS test compared to CP tests performed under similar conditions. It is suggested that an improved understanding of hydrate plugging mechanisms during transient operations could help in the development of advanced strategies to manage hydrate transportation in typical subsea operations, particularly involving unplanned shut-ins and subsequent restarts.

Gas hydrates can form in subsea oil and gas flowlines, where the depths of seawater and ocean conditions provide the thermodynamic environment for hydrate stability. Hydrates present a major flow assurance problem due to the relatively fast timescales at which they can form, grow/agglomerate, and plug a flowline. The common strategy for preventing hydrate formation uses thermodynamic inhibitors (THIs). However, THIs can be cost prohibitive or impractical as the water content in the flowline and its seawater depth increases. Therefore, there is growing interest in the use of alternative hydrate management strategies, such as the injection of low dosage hydrate inhibitors (LDHIs), which are active at considerably lower concentrations than THIs (e.g. 2 vol.% of LDHI versus 50 vol.% of THI). Anti-agglomerants (AAs) are a type of LDHI that prevent agglomeration and allow hydrates to flow as a slurry in oil and gas subsea flowlines. Before field deployment, AAs are screened and selected using laboratory set-ups, mimicking field conditions, in order to evaluate their performance and determine the effective dosage. Current hydrate agglomeration characterization methods implemented in the industry are non-uniform and qualitative, which can lead to conservative recommendations. In this work, the possibility of quantifying hydrate agglomeration in the presence of AAs is investigated, along with studies of the mechanisms via which AAs may operate. One mineral oil and two crude oils were used with a commercial AA in a high pressure stirred autoclave, equipped with particle imaging probes. Motor current input at a fixed RPM was monitored throughout the experiments and serves as an indicator of relative viscosity of the hydrate slurry. This investigation enabled the development of a comprehensive AA performance evaluation. Hydrate agglomeration was detected and quantified by simultaneous increases in the relative motor current and chord length distribution.

Subsea oil and gas flowlines can provide favorable conditions for gas hydrate formation, which can lead to flow assurance issues as hydrate particles agglomerate and accumulate in the flowline. Shut-in and restart operations are particularly critical for hydrate plug formation. Traditional strategies to mitigate hydrate plugging use total hydrate avoidance with thermodynamic inhibitors; however, thermodynamic inhibition can become cost-prohibitive as oil production moves towards harsher environments associated with deeper drilling. Hydrate management strategies using low dosage hydrate inhibitors (LDHI), such as anti-agglomerants (AA), are an attractive alternative to reduce operational and capital expenditures in offshore oil and gas production. In order to successfully deploy anti-agglomerants to mitigate hydrate plugging, a comprehensive understanding of the variables affecting the performance of these additives, such as oil composition and mixture velocity, is needed. Industrial-scale flowloop studies are valuable to investigate the influence of these variables on hydrate particle transportability when using AAs. These experimental setups could be also useful to assess AA performance during transient operations (i.e. shut-in and restart); however, large-scale flowloop data at these conditions is limited. High pressure industrial-scale flowloop tests were conducted using a non-dispersing oil at 50 vol. % water content and 70 vol. % liquid loading. The aqueous phase is a 3.5 wt. % NaCl solution and the gas phase comprises a natural gas favoring the formation of sII gas hydrates. The AA used in these tests is a quaternary ammonium salt. Both baseline (without AA injection) and AA dosed (2 vol. % AA) tests were conducted in order to compare the influence of mixture velocity on hydrate transportability using AAs with respect to systems without AA injection. Three different mixture velocities (2.3, 3.7 and 5.8 ft s) were employed. The experimental procedure included shut-in and restart operations. A combination of different data, such as temperature and pressure drop profiles, mass flow rate and droplet/particle size evolution was used to analyze the effects of AA injection at the different studied velocities. Additionally, water/oil dispersion tests were carried out in order to investigate the influence of the AA on the properties of the dispersion. Both hydrate growth rate and droplet/particle size were influenced by mixture velocity in baseline tests; however, experiments with 2 vol. % AA showed similar hydrate growth rates and droplet/particle sizes regardless of the mixture velocity. In addition, despite AAs reducing hydrate bedding at all mixture velocities with respect to baseline experiments, a certain velocity was needed to completely suppress any indication of hydrate bedding in these systems. Moreover, AA injection successfully inhibited hydrate particle size increase (agglomeration) under static conditions (shut-in), allowing solid material flow after restarting the system. Finally, dispersion tests showed that this particular AA formulation modifies the surface chemistry properties of the system and favors water-continuous dispersions at room conditions with respect to systems without addition of AA.

The formation and accumulation of natural gas hydrates in deep subsea flow lines are one of the most challenging flow assurance problems. Typical high pressure/low temperature operation conditions in deep subsea facilities promote rapid formation of gas hydrates. Recent observations in flow loop experiments in gas/water systems suggested a relationship between hydrate volume and flow regime on pressure drop. In the current work, a simple hydrodynamic slug flow model, based on fundamental multiphase flow concepts, is coupled with a transient hydrate kinetics model to study the effect of hydrate formation on slug flow in gas/water systems.

As oil/gas subsea fields mature, the amount of water produced increases, which results in an increased risk of gas hydrate plug formation in the flowlines. It is important to understand the mechanism of gas hydrate plug formation in high water cut systems in order to manage gas hydrate risk. In this work, we performed an extensive series of gas hydrate formation and dissociation experiments in a 4-inch diameter flowloop at the ExxonMobil research facility at Friendswood, TX. The flowloop was instrumented for pressure, temperature, density, and differential pressure measurements. The effect of mixture velocity (1–2.5 m/s) and liquid loading (50–90 vol.%) on gas hydrate plug formation was studied for 100 vol.% water cut (no oil present) systems, with methane as the gas hydrate former. The pressure drop across the pump due to flow did not substantially increase until a certain concentration of gas hydrates, defined as ϕ transition , formed; this transition, as measured by the rapid increase in pressure drop, can be used as an indication for onset of hydrate plug formation. ϕ transition was found to be unaffected by liquid loading and the presence of salt (3.5 wt.% NaCl) in water, while it increased with increasing mixture velocity. A gas hydrate plugging mechanism for 100 vol.% water cut systems is proposed, which involves a transition from homogeneous to heterogeneous suspensions of gas hydrate particles in water. We hypothesize that the large pressure drop observed after ϕ transition results from the formation of a gas hydrate bed and wall deposit. A correlation between ϕ transition and the mixture velocity is presented, this correlation allows the prediction of an onset of hydrate plug formation based on the fluids mixture velocity.

This paper studies the artificial lift methods available to deliquify loaded gas hydrate wells. When hydrate dissociates, a high amount of water will be produced. On a molecular basis, the gas-liquid ratio (GRL) is 1:5.75. A new hydrate reservoir classification system is proposed based on its future water production characteristics. In this paper we discuss the source of liquid loading for different reservoir types, required for the design of suitable artificial lift methods for the hydrate reservoir during its overall life. The challenges of artificial lift methods of hydrate wells are also investigated, such as insufficient reservoir energy, low GLR, low surface and borehole temperature, formation of ice and secondary hydrates, and sand production. Proper selection of artificial lift methods is critical to the long-term profitability of gas production from natural gas hydrate accumulations. The feasibility of different methods is the focus of this paper.

As oil and gas industry strive for better gas hydrate management methods, there is the need for better understanding of hydrate formation and plugging tendency in a multiphase flow. In this work, an industrial-scale high pressure flowloop was used to investigate gas hydrate formation and hydrate slurry conditions at different flow conditions; fully dispersed and partially dispersed systems. It has been shown that hydrate formation in partially dispersed system can be more problematic as compared to fully dispersed system. For hydrate formation in partially dispersed system, it was observed that there is rapid hydrate growth and rapid increase in pressure drop upon hydrate formation. This is in-contrast to fully dispersed system in which there is gradual increase in the pressure drop of the system. Further, for partially dispersed system, studies have shown that there is hydrate film growth at the pipe wall. This film growth increases the probability of hydrate particle agglomeration and bedding phenomenon, which that lead to flowline plugging. As there is different hydrate formation and plugging mechanism for fully and partially dispersed system, it is thus necessary to investigate and compare systematically the mechanism for both systems. In this project, all experiments were specifically designed to mimic the flow systems that can be found in actual oil and gas pipelines (full and partial dispersion) and understand the transportability of hydrate particles in both systems. Two variables were investigated in this work; amount of water (water cut, WC) and pump speed (fluid mixture velocity). Three different water cuts were investigated; 30, 50 and 90 vol.% water cut. Similarly, three different pump speeds were investigated; 0.91, 1.89, 2.99 m/s. The results from these meausrements were analyzed in terms of relative pressure drop (?Prel) and hydrate volume fraction (fhyd). It was observed that for all water cuts investigated in this work, the ?Prel decreases with increasing pump speed, at a similar hydrate volume fraction. Flowloop plugging occurred for tests with 50 vol.% water cut and pump speeds lower than 1.89 m/s, and for tests with 90 vol.% water cut at a pump speed of 0.91 m/s. Additionally, in all 90 vol.% water cut tests, emulsion breaking where the two phases (oil and aqueous water) separated was observed upon hydrate formation.

Deep subsea facilities with high pressure and low temperature operation, encounter formation of gas hydrates as the most challenging problem in flow assurance. CSMHyK is a transient gas hydrate model specially designed for oil-dominated systems that predicts the formation and transportability of gas hydrates in flowlines. This paper presents a description of three sub-models included in the current version of CSMHyK: kinetics model, transport model and cold flow model; the product of intense efforts over a decade of hydrate research, involving over 20 students at the Center for Hydrate Research at the Colorado School of Mines. A set of conceptual pictures is also presented to describe physical phenomena of gas hydrate formation in water-dominated and gas-dominated systems, as the initial step of a development process that aims to extend CSMHyK towards a comprehensive model, to predict where and when hydrate plugs will form.

Summary A produced-hydrocarbon stream from a wellhead encounters formation of solid gas-hydrate deposits, which plug flowlines and which are one of the most challenging problems in deep subsea facilities. This paper describes a gas-hydrate model for oil-dominated systems, which can be used for the design and optimization of facilities focusing on the prevention, management, and remediation of hydrates in flowlines. Using a typical geometry and fluid properties of an offshore well from the Caratinga field located in the Campos basin in Brazil, the gas-hydrate model is applied to study the hydrate-plugging risk at three different periods of the well life. Additionally, the gas-hydrate model is applied to study the performance of the injection of ethanol as a thermodynamic hydrate inhibitor in steady-state flow and transient shut-in/restart operations. The application of the transient gas-hydrate model proved to be useful in determining the optimal ethanol concentration that minimized the hydrate-plugging risk.

A review of surface chemistry concepts is presented, with the principal objective of identifying interfacial phenomena and surface chemistry interactions involved in gas hydrate formation and agglomeration in oil and gas pipelines. There are five types of interfaces where gas hydrates may form and aggregate: gas/liquid, liquid/liquid, gas/solid, liquid/solid, and solid/solid; where the gas is the hydrocarbon gas, liquid is either oil, water, or condensate, and solid is either gas hydrate or the pipe wall surface. A review of fundamental interfacial concepts can help create a better understanding of phenomena at these interfaces, and can help industry move from hydrate prevention to risk management. Two areas of surface chemistry have been selected to illustrate the concepts and mechanisms associated with these systems: surfactants and emulsions. Examples from the literature pertaining to gas hydrates are presented for each system.

Se ha demostrado en experimentos de laboratorio y experiencias de campo que el método de recuperación mejorada de petróleo por inyección de álcali, surfactante y polímero (ASP) es efectivo en la reducción de la saturación residual de petróleo, a través de la reducción de tensión interfacial y la relación de movilidad entre las fases acuosa y oleica. Dos aspectos críticos para el éxito de un proyecto de inyección de ASP son: i) la optimización de la formulación de ASP en laboratorio; y ii) la optimización del esquema de inyección a utilizar a escala de campo. Este trabajo presenta un enfoque eficiente para la solución del segundo aspecto. El enfoque se basa en la construcción de modelos de superficie de respuesta cuadrática a partir de la salida de un simulador de yacimientos y diseño de experimentos tipo D-óptimo. Esta metodología permite establecer de forma efectiva y eficiente el esquema de inyección de ASP óptimo, y fue utilizada para determinar la tasa de inyección, tamaño del tapón y fecha inicial de inyección de un caso de estudio a escala de proyecto piloto. El esquema de inyección óptimo reduce sustancialmente la cantidad de químicos utilizados al comparar con los sugeridos por el diseño de laboratorio.The EOR method so called alkaline-surfactant-polymer (ASP) flooding has proved to be effective in reducing the oil residual saturation in laboratory experiments and field projects through the reduction of interfacial tension and mobility ratio between oil and water phases. Two issues are critical for a successful ASP flooding project: i) addressing issues related to optimization of the laboratory design, and ii) establishing an optimal injection scheme for the field scale flooding process. This paper presents an efficient solution approach for the latter issue. The approach is based on the construction of quadratic response surface models (surrogates) of reservoir simulator outputs and three-level D-optimal design of experiments. It allows to effectively and efficiently establish the optimum ASP injection scheme, and was applied to determine the optimal values of injection rates, slug size and initial date for injection of an case study at pilot project level. The optimum injection scheme resulted in substantial savings in chemicals used when compared to the laboratory design.

After conventional waterflooding processes the residual oil in the reservoir remains as a discontinuous phase in the form of oil drops trapped by capillary forces and is likely to be around 70% of the original oil in place (OOIP). The EOR method so-called Alkaline–Surfactant–Polymer (ASP) flooding has been proved to be effective in reducing the oil residual saturation in laboratory experiments and field projects through reduction of interfacial tension and mobility ratio between oil and water phases. A critical step for the optimal design and control of ASP recovery processes is to find the relative contributions of design variables such as, slug size and chemical concentrations, in the variability of given performance measures (e.g., net present value, cumulative oil recovery), considering a heterogeneous and multiphase petroleum reservoir (sensitivity analysis). Previously reported works using reservoir numerical simulation have been limited to local sensitivity analyses because a global sensitivity analysis may require hundreds or even thousands of computationally expensive evaluations (field scale numerical simulations). To overcome this issue, a surrogate-based approach is suggested. Surrogate-based analysis/optimization makes reference to the idea of constructing an alternative fast model (surrogate) from numerical simulation data and using it for analysis/optimization purposes. This paper presents an efficient global sensitivity approach based on Sobol's method and multiple surrogates (i.e., Polynomial Regression, Kriging, Radial Base Functions and a Weighed Adaptive Model), with the multiple surrogates used to address the uncertainty in the analysis derived from plausible alternative surrogate-modeling schemes. The proposed approach was evaluated in the context of the global sensitivity analysis of a field scale Alkali–Surfactant–Polymer flooding process. The design variables and the performance measure in the ASP process were selected as slug size/concentration of chemical agents, and cumulative oil recovery, respectively. The results show the effectiveness and efficiency of the proposed approach since it allows establishing the relative contribution of the design variables (main factors and interactions) to the performance measure variability using a limited number of computationally expensive reservoir simulations.

After conventional waterflood processes the residual oil in the reservoir remains as a discontinuous phase in the form of oil drops trapped by capillary forces and is likely to be around 70% of the original oil in place (OOIP). The EOR method socalled alkaline–surfactant–polymer (ASP) flooding has proved to be effective in reducing the oil residual saturation in laboratory experiments and field projects through the reduction of interfacial tension and mobility ratio between oil and water phases. A critical step to make ASP floodings more effective is to find the optimal values of design variables that will maximize a given performance measure (e.g., net present value, cumulative oil recovery) considering a heterogeneous and multiphase petroleum reservoir. Previously reported works using reservoir numerical simulation have been limited to sensitivity analyses at core and field scale levels because the formal optimization problem includes computationally expensive objective function evaluations (field scale numerical simulations). This work presents a surrogate-based optimization methodology to overcome this shortcoming. The proposed approach estimates the optimal values for a set of design variables (e.g., slug size and concentration of the chemical agents) to maximize the cumulative oil recovery from a heterogeneous and multiphase petroleum reservoir subject to an ASP flooding. The surrogate-based optimization approach has been shown to be useful in the optimization of computationally expensive simulation-based models in the aerospace, automotive, and oil industries. In this work, we improve upon this approach along two directions: (i) using multiple surrogates for optimization, and (ii) incorporating an adaptive weighted average model of the individual surrogates. The cited approach involves the coupled execution of a global optimization algorithm and fast surrogates (i.e., based on Polynomial Regression, Kriging, Radial Basis Functions and a Weighted Average Model) constructed from field scale