Reservoir simulation

Learnings Applied to Reservoir Simulation of Unconventional Plays

In the complete paper, the authors revisit fundamental concepts of reservoir simulation in unconventional reservoirs and summarize several examples that form part of an archive of lessons learned.


Reservoir simulation is valuable in understanding dynamics of unconventional reservoirs. Applications include estimating long-term production behavior, enhancing well-spacing and pad-modeling efficiency, optimizing completion and stimulation of horizontal wells, and understanding production drivers that cause differences in productivity between wells. In the complete paper, the authors revisit fundamental concepts of reservoir simulation in unconventional reservoirs and summarize several examples that form part of an archive of lessons learned.

Reservoir Simulation Applications in Unconventionals

Reservoir simulation plays an important role in many stages of unconventional reservoir field development. In the initial pilot phase, a calibrated reservoir simulation model can provide an indication of estimated ultimate recovery (EUR), which is advantageous given few wells with short production time. EUR evaluation with incorporated uncertainties is a key variable in asset evaluation and investment decisions.

During the ramp and development phases, reservoir simulation aids completion optimization by providing a link between completion parameters, such as cluster spacing and stimulation pumping parameters, and a well’s production. Extending this idea further, the integrated nature of reservoir simulation enables identification of production drivers and understanding of why wells are producing differently.

During the mature stages of unconventional reservoir development, well spacing and hydraulic fracture and pressure interference, together with refracturing, require advanced modeling techniques that can include a combination of hydraulic fracture simulation, 3D stress simulations, and advanced reservoir simulation.

Reservoir Simulation Modeling in Unconventionals

Unconventional reservoirs are heterogeneous across multiple scales and exhibit variations in production performance across wells. The production variation is influenced by a large variation in production drivers that simultaneously affect the production. Examples of these drivers include reservoir properties such as hydrocarbon-filled porosity, moveable hydrocarbons, matrix permeability, relative permeability and pressure/volume/temperature, completion properties such as stress-state and natural fractures, and stimulation (pumping) parameters. Reservoir simulation constitutes one of the best tools to understand the dominant production drivers because of the multidisciplinary data required to build simulation models for unconventional reservoirs.

Because of the orders of magnitude of difference in matrix permeability between conventional and ­unconventional reservoirs, pressure gradients from the sandface to the formation (pressure drawdowns) are larger in unconventional reservoirs and drainage radii are more likely to be smaller. Therefore, in the case of shale reservoirs, single-well models usually suffice except for the case of multiwell pads. Very fine grid cells (on the order of inches to a few feet) are used close to, and immediately surrounding, the perforation clusters, whereas in conventional reservoirs, coarser grid cells are used, especially in full-field models, even with local grid refinements.

Given that shale reservoirs have hydrocarbons produced at source, and most have kerogen-hosted pore systems, water saturation often is very low and water production from the formation shale matrix generally is not common. In shale simulation models, defining capillary pressure as zero and thus without transition zone, with all water as irreducible, will suffice in many instances, although notable exceptions exist.

Unconventional reservoirs cannot produce commercially without massive hydraulic fracture treatments that seek to maximize surface area for hydrocarbons to flow. This process often forms complex hydraulic fractures. The permeability-enhanced area (PEA) is often thought of as a simplification of a multiporous system consisting of propped tensile fractures and an associated fracture network that includes the newly broken shear fractures or the reopened and extended healed natural fractures resulting from shear failure of weakness joints or natural fractures. Hydraulic fracture complexity is enhanced (less planar) with lower stress anisotropy and a larger density of (sealed) natural fractures and planes of weakness. The hydraulic fracture complexity is incorporated in shale reservoir simulations by PEAs that have orders of magnitude higher permeability than that of the rock matrix. This calls for finer gridding inside and close to the hydraulic fractures, compared with the far-field matrix regions.

As a result of the large hydraulic fracture treatments, the initial flowback production period is often characterized with high water production (and water cut). This aspect is captured in shale reservoir simulations by defining high initial water saturation in the PEA and requires small timesteps to improve model stability.

In shale gas reservoirs, the effect of additional gas production from Langmuir gas desorption at kerogen sites may be important, especially for long-term production forecasts.

Tight oil or gas sandstones fall into the unconventional reservoir category because of their very low permeabilities, but, unlike shale reservoirs, these hydraulic fractures are believed to be planar. Consequently, no PEA is used in tight sandstone simulation models. Furthermore, because their permeabilities tend to be at least a couple of orders of magnitude higher than those of shale reservoirs, pressure interference between wells may be much more significant depending on the interwell spacing. Additionally, tight sandstones may have zones of formation water production that need to be accounted for in the simulation.

Reservoir Simulation Model Types

The two main model types in unconventional reservoirs are structured and unstructured grid. However, regardless of the model type used, it is due diligence in data integration that will ultimately define success. The essence of a work flow for single-well shale reservoir simulation is to gain an understanding of the different petrophysical, geological, geomechanical, stimulation, and operational drivers that affect the production. Through this data integration and understanding, successful reservoir simulation models are created. Consistency in modeling approach, which ensures that reservoir simulation results from different wells can be compared with one another, is also critical.

To illustrate integration, consider data acquired during the drilling phase, such as borehole images and lost circulation of drilling fluid, which could imply the presence of natural fractures or planes of weakness. Further observations could be made from prefracture diagnostic pressure tests. This could be consistent with a very high observed productivity early in the well’s life that could imply large PEA thickness (or a larger surface area of the hydraulic fracture system as a result of the energy from the hydraulic fractures opening up sealed natural fractures and planes of weakness).

As part of the reservoir simulation effort, parallel processes such as rate-transient analysis (RTA) exist that can provide value, but their results should be used with care. For example, direct application of fracture half-length and permeability from RTA may not be appropriate for completion evaluation and optimization. RTA results should also be used with caution when used directly in reservoir simulation models, although they can be used as an indicator or as additional information to help in the interpretation.

The complete paper illustrates and discusses structured and unstructured reservoir grid models. One of the most important steps in the use of structured reservoir grid models is to use information from hydraulic fracture modeling and operations to constrain the initial geometry and permeability of the PEA in the reservoir simulation model.

Unconventional Reservoir Simulation Challenges

Because unconventional reservoirs can be challenging and time-­consuming to simulate, understanding beforehand which of the four reservoir simulation applications will be the focus of the work will facilitate the task at hand. The complete paper discusses 10 of these challenges.

Applied Learnings in Reservoir Simulation of Unconventionals

This section of the complete paper shows through representative graphical examples a few practical learnings captured from different parts of the world in shale reservoir simulation. Topics include modeling shale wells on a per-stage basis; well shut-in, choke changes, and flowback operations; cluster spacing, PEA thickness, and grid size; supercharging effects; and supercharging.


  • The four main applications of reservoir simulation in unconventional reservoirs are determination of EUR, optimization of completion and stimulation parameters, understanding production drivers (between wells), and optimizing well spacing. Defining the principal objective will help better manage the challenges and the modeling approach.
  • Production heterogeneity of unconventional wells reflects multiple production drivers acting simultaneously. A consistent modeling approach rich in data integration is key.
  • Reducing the horizontal well simulation model to one stage can be desirable in some cases; this allows reduction in simulation runtime and finer gridding if required.
  • Well shut-ins (and the possibility of large changes in choke size) could be detrimental to hydraulic fractures, causing an apparent decrease in well productivity. This occurrence has been identified in numerous reservoir simulation cases where the well productivity decreases immediately after a well opens up after a shut-in.
  • When performing reservoir simulations on multiple wells, a consistent modeling approach is required that includes the same grid configuration. This allows for better comparisons between wells. Small perforation cluster spacing requires finer gridding to better capture the transient flow behavior. Careful consideration needs to be made for consistent gridding requirements across wells with different cluster spacings.
  • Supercharging effects call for the definition of higher pressure around perforation clusters; this is important if the objective of the study is focused on short-term production behavior.

This article, written by JPT Technology Editor Judy Feder, contains highlights of paper SPE 199164, “Applied Learnings in Reservoir Simulation of Unconventional Plays,” by Raphael Altman, SPE, Roberto Tineo, SPE, and Anup Viswanathan, SPE, Schlumberger, et al., prepared for the 2020 SPE Latin American and Caribbean Petroleum Engineering Conference, 27–31 July, Virtual. The paper has not been peer reviewed.