AFA Documentation
Breadcrumbs

Scaling in Reservoir Engineering

According to Blasingame[n.d.], production data analyses and pressure transient analyses “see“ the reservoir as a volume-averaged set of properties (see image below). In AFA terminology, the terms “Equivalent Radial Homogenous Reservoir” is often used. As new analytical models or solutions are developed, they will have the same view of the reservoir.

Below, the images, taken from Halderson, show conceptual scales associated with porous media mediate averages (left), and pressure-buildup “view“ of relatively large-scale internal heterogenities.

image-20250508-081504.png

Micro = Scale of pores and grains

Macro = Scale of conventional core plugs

Mega = Large grid blocks in simulation models

Giga = Total formation or regional scale.



image-20240301-030355.png

This problem is independent of the quality or type of PTA/RTA model; in other words, new solutions will also have this view of the reservoir. When using various transient and/or PSS models, one must understand the scale of the reservoir features. AFA often uses the term “Equivalent Homogeneous Reservoir“ model to describe this phenomena.

Jordan (2009) performed a number of simulated examples with variable permeability and produced the conceptual plots below, which showed

  • As a general rule, the effective permeability derived from PTA started high, and reduced towards a constant as production (or time) approach PSS.

image-20240501-032027.png
image-20240501-032217.png


Internal Boundary Example

In this example, a reservoir was simulated with an internal no-flow boundary (a 72 Acre region with zero permeability) within a larger 640 Acre reservoir of 20 md, and 5.5 Bcf of gas. Flowing material balance analysis produced an equivalent reservoir of 16.5 md, with a OGIP of 5.4 Bcf. The net area calculated (using the volumetric parameters such as net pay, porosity, and more) was 567 Acre (which closely approximates 568 = 640-72). Production based analysis produced the equivalent reservoir which:

  • Will produce reliable results for physics based forecasting honouring effective productivity index, and material balance.

  • Not necessarily great for reservoir characterization.


image-20240501-010542.png
FMB Analysis of a Reservoir with an Internal Boundary (Jordan 2009)
image-20240501-012325.png
Production forecast of Base Model and Equivalent Homogeneous Reservoir Model (Jordan, 2009)

Numerical Simulation Example

In this example, we generate a numerical model with the following criteria:

Parameter

Single Phase Vertical Well

Area (Square)

10,000 x 10,000 ft

Permeability.

0.05 - 21 md (Average = 5.31 md)

Permeability field was generated using a uniform distribution generator

Net Pay

0.1 - 18 m (Average = 10.1 m, or 32.8 ft)

Net pay field was generated using a uniform distribution generator

Porosity

10%

Initial Pressure

5,000 psia

Water Saturation

0

Gas Gravity

0.7 Dim

OGIP

87.2 Bcf (numerically calculated)

Graphical representations of the random permeability and net pay field are shown below:

image-20240501-013625.png
Random Permeability Distribution (Jordan, 2009)
image-20240501-013715.png
Random Net Pay Distribution (Jordan, 2009)

Using the numerical model, a production forecast was generated with a slightly varying BHP, ranging from nearly 2,500 psia to a somewhat stable value of about 1,900 psia - a relatively large drawdown compared to the initial pressure of 5,000 psia.


image-20240501-014044.png
Production Forecast for Numerical Simulated Example

The Flowing Material Balance (or FMB) for the simulated example is shown below. As can be seen, the FMB produced the following results:

Parameter

FMB

Simulator

Average Permeability

8.141 md

0.05 - 21 md (Average = 5.31 md)

OGIP (or IGIP)

86.6 Bcf

87.2 Bcf

Drainage Area

2504 Acres

2295 Acres (10,000 x 10,000 ft)

Given that the FMB assumes a constant net pay of 30 ft, the total effective drainage area will not be similar to the numerical model - but nonetheless, it is a close approximation. We do know that the FMB will provide the correct productivity index; thus comparing the production forecasts (from transient to PSS condition), we find excellent correlation

image-20240501-014516.png
FMB of Numerically Simulated Example (Jordan, 2009)
image-20240501-015511.png
FMB Model vs Simulated Model (Jordan, 2009)

Numerical Simulation Example: Introduction of No-Flow Boundary

Following on the previous discussion, an arbitrary shaped no-flow boundary was introduced into the simulator (all other parameters remained the same). Also shown in the pressure contour for the system (at an arbitrary time) we have a none-radial, none symmetrical flow pattern.


image-20240501-022646.png
Random Permeability Distribution (Jordan, 2009), Including Internal No-Flow Boundary
image-20240501-022855.png
3D Pressure Profile for Numerical Model (Jordan, 2009), Including Internal No-Flow Boundary.

The FMB, as well as a traditional PTA analysis shown below.

  • The FMB suggests an effective permeability of 4.8 md with an effective drainage area of 2570 Acres, and OGIP of 88.8 Bcf.

  • The PTA suggests an average reservoir permeability of 8.1 md, with late-time PSS shown past “apparent“ radial flow.

  • A production forecast using both FMB/PTA results shows radically different results.


FMB

PTA Analysis


image-20240501-023149.png
FMB of Simulated Example





image-20240501-023320.png
PTA of Simulated Example



image-20240501-023934.png
Comparison of RTA (or PDA) model and PTA Model


References:

  • T. A. Blasingame “Drowning In Data — The Future of Reservoir Performance Analysis

  • D. Ilk, and T. A. Blasingame “Forecasting: Forecasting Future Performance for Various Production/ Completion/ Development Scenarios”, PETE 612, Unconventional Oil and Reservoirs, 2018.

  • Colin Lyle Jordan, Robert Jackson, Short Course: Complex Well Analyses & Reservoir Engineering Methods for Unconventional Gas, 11th Annual Unconventional Gas Conference, November 20, 2009.

  • Abraham Sageev, Pressure Transient Analysis of Reservoirs With Linear or Internal Circular Boundaries, Stanford Geothermal Program Interdisciplinary Research, SGP-TR-65.

  • Helge H. Haldorsen, Simulator Parameter Assignment and the Problem of Scale in Reservoir Engineering, Reservoir Characterization Journal, 1986