Here we report on in-depth water diversion using sodium silicate to increase oil recovery at the Snorre field, offshore Norway. A comprehensive qualification program revealed that the onset of gelation can be controlled; this was demonstrated in realistic core flood experiments as well as in a single well injection pilot. This paper highlights key design, response measurement plan and operational experiences from a large scale interwell field pilot of sodium silicate injection in a reservoir segment at the Snorre field on the Norwegian Continental Shelf. The operation of injecting 113 000 m 3 preflush, 240 000 m 3 sodium silicate and 49 000 m 3 postflush was performed from June to October 2013. The goal is to create an in-depth restriction between a subsea water injection well and a platform oil producer with approximately 2 000 m well spacing, and thereby improve the reservoir sweep by water injection.To perform the field pilot a 35 000 ton shuttle tanker was converted to a well stimulation vessel with the necessary equipment to accommodate a higher number of people, a desalination plant, storage and mixing equipment and high pressure pumps. The vessel was connected directly to a subsea water injection well and injected during a period of 5 months. The chosen design was to (a) soften the formation water by a KCl preflush, (b) control the gelation kinetics using HCl acid as activator, mixed into the diluted silicate solution and (c) displace the silicate solution by a KCl postflush followed by seawater injection.The pilot injection operation is completed, and the displacement of the sodium silicate with following seawater injection to the planned position for in-depth plugging is still on-going. The operational success criteria of proving ability to perform large scale transport, mixing and injection of sodium silicate from a shuttle tanker directly into a subsea well with no near wellbore plugging is met. Future production response will reveal if the success criteria of in-depth plugging, improved reservoir sweep and decrease in water cut will be met.
Immiscible water-alternating-gas (IWAG) experiments performed on equilibrated fluids are summarised together with the corresponding two-phase gas-oil and water-oil displacements. Experimental studies at reservoir condition and also mechanistic experiments over many years have shown accelerated oil production and higher core flood oil recovery as a result of three-phase flow. The three-phase effects that are included and analysed are; trapped gas, and mobility for secondary processes (ex. water after gas injection). The oil recovery from the different oil recovery processes represented by; gas, water, and WAG core displacements are also compared. The oil recovery has been related to the trapped gas saturation, and the efficiency of the trapped gas on oil recovery is found to be varying with core wettability. Experimental results have shown that both gas and water relative permeability generally is reduced during three-phase flow. Multivariate analysis has been used to investigate relations between variables like Sgt = f(k, , Sgi, krw e), Sorm = f(k, , WI, Sorw, Sgt) and Sorg = f(k, , WI, Sorw, krg e). The paper tries to address the question of what three-phase parameters influence oil recovery, and how these parameters are related. This is an important question for modelling and optimising the WAG process.
Usually extended two-phase capillary pressures are used in three-phase simulations, because three-phase capillary pressures are not possible or hard to measure. In this work three-phase capillary pressure surfaces are created by at pore network model. The input parameters to this network model are found by matching two-phase capillary pressure curves. This matching is done with a slightly modified EnKF routine. Tables with three-phase capillary pressures are created and used as input to flow simulations.
The paper reviews recent development in modelling of immiscible WAG processes (IWAG). Several different approaches have been made to model three-phase relative permeability in particular. In most cases capillary pressure has been neglected in application of these models in numerical simulations of IWAG. The argument behind eliminating capillary pressure is to simplify the model, and the assumption that capillary pressure is of less importance for the problem analysed or because there are no experimental data available. This study also shows the consequence of neglecting capillary pressure. A novel approach to include three-phase capillary pressure is discussed, and also a procedure for estimation of three-phase capillary pressure from more available two-phase capillary pressure data. A practical case study of history matching IWAG core floods is also presented. Introduction Numerical simulation of any EOR process is a key to the prediction of incremental oil recovery. Water -alternating -gas (WAG) injection has found an increasing application in both clastic and carbonate reservoirs and at both miscible and immiscible gas conditions 1,2. The addition oil recovery is typically in the range of 5–10 percent of the original oil in place. WAG injection is an oil recovery method initially aimed to improve sweep during gas injection. Possible improved microscopic efficiency in three-phase zones of the reservoir may come as an added benefit from the WAG injection. Today the WAG process (both miscible and immiscible) is considered for a number of new fields in the North Sea. WAG is defined as any injection of both water and gas into the same reservoir. This including definition covers MWAG (miscible), IWAG (immiscible), HWAG (hybrid), SWAG (simultaneous), and also tertiary gas injection or tertiary water flooding. When miscibility is developed along the gas slug, as gas displaces oil, it is referred to as a miscible WAG. In this case the main purpose of the water slug is to increase the volumetric sweep since the residual oil saturation will be low after the miscible front has passed. Immiscible WAG (IWAG) is pulses of gas and water injected, where the gas cannot develop miscibility with oil. Still some compositional exchanges between gas and oil may be important for the fluid characterization and oil recovery. Hybrid WAG is when a large slug of gas is injected followed by a number of small slugs of water and gas. Simultaneous injection has also been performed, but is reported to give reduced injectivity in some cases. In this paper immiscible WAG processes will be discussed. This process involves both drainage and imbibition processes, three phase flow modelling approach, and hysteresis in relative permeabilities and capillary pressure. The complexity of the WAG process is further complicated by mass exchanges (swelling and stripping of the oil by the injected gas). The focus is here on fluid flow functions and the mass exchanges are here ignored. Simulation studies of IWAG3–7 have shown that analytical models like Stone 8 and Jenkins 9 may strongly underestimate the extent of the three-phase zone. The extent of the three-phase zone is influenced by phase trapping and three-phase relative permeabilities, see Figure 1. The residual oil saturation in the three-phase zone may also be reduced compared to two-phase flow, due to effect of trapped gas on residual oil and also three-phase relative permeabilities 10–12. Experimental studies showed accelerated oil production and higher core flood oil recovery as a result of three-phase flow 10–13. The oil recovery has been related to the trapped gas saturation 11–13, and the influence of the effect of trapped gas is found to be varying with core wettability 13. Experimental results have also shown that both gas and water relative permeability may be reduced during three-phase flow. Simulation results have shown that the two-phase relative permeability hysteresis models were unable to describe the three phase experiments. The match of WAG experiments was improved by the three-phase relative permeability approach 14,15.
The effect of capillary pressure related to immiscible WAG (Water Alternate Gas) is studied by use of a numerical simulator. The capillary pressure is found to have a significant effect on the pressure gradient and the total oil production both in two-phase and three-phase flow situations. When the capillary pressure is included in the simulation the total oil production is considerably lower than when the capillary pressure is neglected. Experimentally measured two-phase capillary pressure was used as input to the numerical simulator. The two-phase capillary pressure was further used to estimate three-phase flow, related to WAG processes. A network model was applied to generate a consistent set of two-phase and three-phase capillary pressure. The network model was anchored to measured two-phase data, and threephase capillary pressure was constructed. The gas-oil and mercury capillary pressure anchored the pore structure parameters, while water-oil capillary pressure anchors the wettability parameters in the network model. The network model quantifies the difference between three-phase and two-phase capillary pressure, and in the cases studied the difference between two-phase and three-phase capillary pressure was significant. Results and discussion Experiment: Data from a North Sea reservoir was used in this work. Several flow experiments were performed on a core at reservoir conditions. One experiment started with gas injection and was followed by water injection, G1W2. Figure 1 shows the oil-, water-and gasproduction from this injection sequence. The other injection sequence started with water injection and was followed by gas injection, another water injection and finally a gas injection period, W1G2W3G4. Note that the oil production in the G2 period increases in two steps. This could be due to double displacement in the early phase of the G2 injection period and direct displacement in the later phase. Figure 2 shows the production data for this experiment. Some key numbers from the experiment is listed in table 1. Capillary pressure and wettability were measured on a core plug with similar qualities as the composite core used in the flow experiments. The wettability was mixed-wet large; large pores oil-wet and small pores water-wet. The capillary pressure curves were matched with a correlation function 6 and scaled to match the endpoints from the flow experiments. The capillary pressure curves used as input to the simulation is shown in figures 3 and 4. Mercury data was also available, but for a different core than the one used for the capillary pressure and wettability measurements.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.