## Summary

At the base of a steam-assisted-gravity-drainage (SAGD) steam chamber, a liquid pool is developed, which is a key component for bitumen production. A producer is placed in the liquid pool, and its production is mainly controlled by its liquid level and the temperature gradient across its depth. A “subcool control” or “thermodynamic steam-trap control” is a typical operating strategy to control steam coning to the producer. Part I of this study (Irani 2018) presented a methodology to evaluate the production rate for a given pressure drawdown and subcool in a SAGD liquid pool; and Part III (Irani and Gates 2018) modified such a formulation for a stability analysis of the Nsolv™ (Nenniger and Nenniger 2000, 2001) process that contained a large fraction of liquid butane. In this study, first, the effect of localized hot spots on well control is formulated as a virtual skin factor in the liquid-pool deliverability equation. The results of this work suggest that a longer hot spot will yield to lower differential pressure and make it more challenging to control the steam breakthrough by choking the well at a given rate. Another key finding is that the steam coning becomes less controllable for higher-permeability reservoirs.

Flow-control devices (FCDs) have been used extensively in horizontal wells for conventional oil and gas production to prevent early water breakthrough or gas coning. Although FCDs are commonly installed to prevent steam coning after steam breakthrough and to manage hot spots as retrofit completions by SAGD operators, in recent years, FCDs have been often installed to improve SAGD well-pair performance as part of the initial completion. The benefits associated with this technology in the SAGD industry have been studied with reservoir simulations and validated with field experience, but a theoretical study that discusses the main factors for a correct FCD selection on the basis of operational conditions and reservoir heterogeneity is required. In this study, the liner-deployed FCD and liquid-pool systems are coupled, and two criteria are suggested for a design of liner-deployed FCDs on the basis of the pressure-drop ratio of the FCD relative to the liquid pool ($ΔPFCD/ΔPpool$) and the coefficient of variation ($CoV$) of inflow for the liner-deployed-FCD wellbore ($CoVFCD$). The results of this study show that in higher-permeability reservoirs, the ideal FCD design should have more ports to reduce the differential pressure to flow response. While FCDs will improve inflow conformance relative to completions without FCDs, the effect of permeability in this improvement is minimal. This improvement is larger in applications operating at lower target subcool values. Reducing the target-wellbore subcool value can improve well deliverability twofold: First, FCD-completed wells produce more at lower subcools and, second, reducing the subcool value helps to improve inflow uniformity along the length of the lateral. By effectively removing the fluids available to the producer, the growth of the steam chamber can be maximized through accelerated injection rates.