Producing from liquids-rich, ultra-low-permeability reservoirs requires long, horizontal wells with multiple fractures—a situation that demands a better understanding of well-completion practices in relation to reservoir dynamics to maximize benefit. This paper attempts to augment that understanding through a stochastic reservoir modeling approach. A reservoir simulation model for a typical condensate well in the Eagle Ford liquids-rich area was used in a decision-under-uncertainty framework to identify optimal completion and production strategies. The important factors considered are for fractures (length, conductivity, conductivity endurance, and spacing), reservoir (matrix permeability), fluid (saturation pressure and condensate-gas ratio), and well constraints (bottomhole pressure and rates). The effect of these factors, grouped into decision and uncertainty variables, on well productivity were examined to identify the optimal combination of values for each decision variable, considering the impact of uncertainty variables represented by a statistical metric.
Completion techniques and proppant selection that maximize well productivity in conventional or even tight formations by maximizing fracture conductivity are not necessarily optimal for ultra-low-permeability reservoirs. The marginal benefit of higher fracture conductivity diminishes rapidly in such reservoirs, and lower-grade proppants can be used. The optimal completion strategy consists of balancing the effects of decision variables based on a clear objective of maximizing reserves or accelerating production or a specific combination thereof. This is because the variables interact; for example, longer fractures both accelerate production and add reserves (bigger drainage volumes), whereas if drainage volumes interfere, closer fracture spacing can accelerate production without increasing reserves. The rapid falloff in production rates for wells in ultra-low-permeability reservoirs encourages operators to establish high initial rates. In liquids-rich wells, such a strategy can leave a large quantity of unproduced liquids in the fractures that also impedes production rates. At a very low drawdown, however, the well may not even produce. Hence, an optimal production strategy maximizing the liquid yield at the surface should be planned and employed.
During the fracture-treatment design process, large uncertainties that affect fracture geometry and properties are often ignored, leading to designs that are suboptimal for well productivity in the field. This study considers decision and uncertainty variables related to both completion and production. Insights developed with respect to the interaction of various factors from the study allow for a fuller understanding and provide practical guidelines for completion and production practices. The dynamic behavior of condensate banks in the presence of hydraulic fractures as it relates to production practices is also examined—this has not been discussed in detail in the literature.