Methane hydrates were produced in an isochoric high pressure cell in the presence of an ultralow concentration of methanol as an additive. Methanol concentrations examined were in the range 1.5-20 ppm by weight. The real gas equation and the nucleation probability distribution have been used to understand the effect of ultralow concentrations of methanol on the amount of hydrate formation, the rate of nucleation, and the range of probability distribution function in which random nucleation of methane hydrates occur. Comparisons with the pure water baseline experiment showed that an ultralow concentration of methanol exhibits dual effect, both as an inhibitor and as a promoter on structure I methane hydrate formation.

The effect of the cooling rate and gas composition on the kinetics of hydrate formation and the stochastic nature of nucleation has been examined by forming structure II hydrates from two different synthetic natural gases: one with two components (SNG2) and the other with seven components (SNG7). The hydrate equilibrium properties of SNG2 and SNG7 were comparable, and all experiments were initiated at the same temperature and pressure conditions in an autoclave cell. The initial degree of subcooling and other parameters that could affect the kinetics were, thus, approximately the same during all nucleation experiments. From the experimental results, SNG2 showed an increasing nucleation rate, while SNG7 showed a decreasing effect as the cooling rate increases in steps of 2 °C/h. In addition, it is observed that the rate of nucleation is dependent upon gas compositions, and different gas compositions respond differently on the same cooling rate. In an attempt to understand the experimental results, the classical nucleation theory of multicomponent systems, the probability distribution function, and the principle of irreversible thermodynamics have been employed. The observed effects are, thus, related to the dependence of the critical size upon the temperature and gas composition, chemical oscillations as a result of the change in solubility of the individual gas components, and coupled mass and heat fluxes during cooling. Such findings on system responses are of paramount importance for a reliable evaluation of the effect of additives on hydrate formation processes.

Kinetics of hydrate formation is of paramount importance for hydrate prevention in pipelines and gas storage in a hydrate state. The effect of gas composition on the kinetics of structure II (sII) hydrate growth has been examined by using two different synthetic natural gases, one with two components (SNG2) and the other with seven components (SNG7). The hydrate equilibrium properties of SNG2 and SNG7 were comparable and the initial degree of subcooling was thus approximately the same during all growth experiments. The same PT conditions and stirring rate were also set for both SNGs enabling us to examine the effect of gas composition on the growth behavior of the same crystal structure. The amount of gas consumed and the structure of the growth curves have been investigated to understand the effect of gas composition on the kinetics of sII hydrate formation. In addition, several chemicals (MeOH, PVP, PVCap) have been tested for their effects on the amount of hydrate formation and structure of the growth curves. The results showed that the amount of sII hydrate formed from SNG7 was doubled as compared to SNG2 with and without the additives. This could be explained by the higher amount of large cavity preferring gas components in SNG7 as compared to SNG2. On the other hand, the growth rate of sII hydrate made from SNG7 first increased toward a maximum value after hydrate onset, and then decreased toward zero at the end of experiment. This growth behavior was not observed for sII hydrate structure produced from SNG2. The SNG2 growth rate showed a maximum at onset and then decreased toward zero eventually. The addition of the chemicals did not change the characteristic growth behavior produced by the individual gas mixtures. The chemicals had an effect in shifting the initial growth rates either up or down based on concentration and type of the chemicals used while the structure of growth rate versus time was kept the same. This study confirmed that gas composition alone is an important parameter for sII hydrate growth kinetics, and the same hydrate structure produced from different gas compositions does not necessarily show the same growth behavior.

Summary In this paper we describe the analysis, test, and design work to deliver an optimal lower completion for a trilateral well by integrating passive and autonomous inflow-control devices (ICDs) (AICDs) at the Alvheim Field offshore Norway. In 2015, both passive ICDs and AICDs were tested in the laboratory with Alvheim fluids at reservoir conditions. The experimental flow testing demonstrated that the AICD chokes gas more efficiently than the passive ICD. The experimental results enabled correct modeling of AICDs in both the reservoir-simulation model and the simpler steady-state inflow model. The following lower-completion strategy was established for the new well: Where the well was close to the overlying gas cap, AICDs should be used, whereas passive ICDs with variable strength were to be used elsewhere to optimize the inflow. During the drilling phase, the steady-state model was updated with the as-drilled information; the lower-completion design for each branch focused on obtaining what was estimated to be an optimal inflow depending on the oil volume per drainage area. A key uncertainty in the design work was whether shaly zones along the wellbore would creep/collapse with time and act effectively as packers. The lower completion covered 7 km of reservoir penetration in the three branches, and 15 unique oil tracers were installed to evaluate the cleanup and the inflow profile along the well. The well started producing in May 2016 and a successful cleanup was confirmed by oil-tracer responses. In August 2016, a restart-tracer-sampling campaign was performed after a 12-day shut-in, and this formed the basis for a “chemical production log.” The tracer-based inflow interpretation was compared quantitatively with the model-predicted inflow and qualitatively to the tracer responses seen during the cleanup. The comparison confirmed that the lower completion works as initially planned. The interpretation further indicated that the upper zone has a lower degree of pressure support than the lower zone, and that the larger shaly sections have creeped/collapsed and act as packers. The well has exceeded predrill production expectations, with an average oil rate of 3375 std m3/d (21,240 STB/D) during the first production year. A large part of exceeding the predrill expectations is attributed to the lower-completion design, where the focus has been to optimize such that the whole well contributes, from the heel to all toes.

Optimizing production from a multilateral well requires understanding and characterization of the well and completion performance in the flow path from the reservoir to the separator. Situations when crucial information may be gathered includes transient flow phases such as cleanup and restart after shut-ins. Understanding the inflow from the near-wellbore into each lateral and monitoring the functionality of completion components are important tools in establishing completion effectivness and wellbore performance. This paper describes new applications of chemical tracers in multilateral horizontal wells and utilizes flowbacks in understanding cleanup efficiency, confirming inflow from the toe-section of long horizontal wells and monitoring the functionality of ICV components based on a case study on four multilateral wells in the Alvheim field, Norway. A new generation of chemical tracers were embedded in a polymer matrix and installed in ICD screens on four long horizontal multilateral wells. In total, 27 unique oil tracers were used for 12 months of oil marking period designed for monitoring early production. The oil tracers were released when the tracers were in contact with oil and mobilized to topside sampling point with the produced fluids when the wells were opened for cleanup. The samples were analyzed in a laboratory and the tracer responses were used to study cleanup efficiency and to monitor ICV during sequential cleanup of laterals. The tracers were also used to understand the inflow contribution from the toe section of each lateral. The case study showed that intelligent well tracers provide a direct proof of contribution from the toe-sections of the multilateral wells and enabled monitoring of ICV functionality and synchronization.