Introduction

This post is the first in a series that focuses on evaluating waterflood performance in a reservoir cross-section. The main objective of the series of posts is educational. There will be several posts, each describing a different aspect of waterflood performance evaluation in the cross-section. We will consider different tasks including building the reservoir simulation model, performing different sensitivity and economic analyses. For all these tasks, we will be using the same base simulation model or perform adjustments to the model to illustrate relevant ideas.

We will consider two different approaches for evaluating waterflood performance: analytical and reservoir simulation, and compare the results from both approaches. The two approaches will be considered in separate tutorial series. In this series, we will consider reservoir simulation for the waterflood performance evaluation.

Objective

As described previously, the focus of the series is educational. As a result, I have chosen to use a cross-section model (X-Z) for faster computation and post-processing; ease of illustrating and understanding the waterflooding concepts discussed; and ease of comparing analytical and reservoir simulation approaches.

Our main objective is to perform a waterflooding evaluation of the cross-section. We will work our way from reviewing the cross-section data to performing economic analyses using the production profiles obtained from the simulation runs.

Sample workflow

A sample workflow is illustrated in Figure 1. From the figure in the top left (then clockwise), we have the permeability vs. depth profile of the reservoir cross-section.

Figure 1: Sample workflow steps.

The next figure shows the well configuration we will adopt for waterflooding the cross-section. There are two types of well, a water injection well (injector) on the left side and an oil production well (producer) on the right side of the schematic. Injected water goes into the injector from the surface through the tubing, and into the reservoir (via the injector wellbore) where it displaces oil towards the producer. The oil follows a similar, but reversed path, from the producer wellbore into the tubing and then to the surface processing facilities.

Next, is a cross-section (X-Z) simulation model showing the X-direction permeability which is constant across each layer. The layer properties (thickness, PERMX, porosity, …) in the simulation model correspond to those from the cross-section. The flooding is modeled in two directions: X-direction (horizontal) and Z-direction (vertical). It is assumed that only two phases, oil and water, are present in the reservoir, the water phase in the reservoir is initially immobile, and there is no natural aquifer. Also, both wells are vertical and fully-penetrating, i.e, they are completed in all the grid blocks in the Z-direction. We perform a reservoir simulation run (using a simulator) to obtain the results from the waterflood simulation.

The next two figures (fourth and fifth) show example of simulation results that are obtained. The first of these figures is a water saturation map and shows the saturation fronts in each layer after 4.0 years of injection. The injected water moves from left side of model to the right towards the producer. The scenario modeled here is one where the layers are hydrodynamically isolated from each other, i.e., no communication between adjacent layers. Assuming continuous injection, the saturation map changes with time with subsequent maps showing higher water saturations.

The fifth figure shows the oil and water production rates obtained after the simulation run. The production rate profiles are typical of waterflooding operations. In this case, the oil production rate is constant for approximately 5 years (called plateau period) and then water breakthrough occurs, i.e., the injected water reaches the producer. This event is marked by a reduction in the oil production rate and an increase in the water production rate and water cut (fraction of water in total production stream). The bumps in the production curves corresponding to the successive flooding of the layers with the order of flooding depending on the layers’ flood velocities.

Finally, from the production rate profiles, we compute economic indicators for the particular scenario represented by the simulation model. The economic model takes as input the production profiles from the simulation run and pre-defined economic parameters including oil price, water disposal/treatment cost, water injection costs, etc.

The different components in described Figure 1 will be studied and described in detail later on, so there is no need to worry if something is not clear at this time.

Tasks

The different tasks will be addressed in separate posts to reduce mental load. They are grouped as follows:

• Data Analysis: Review cross-section data and assemble data required to build the simulation model

• Model Construction: Building the simulation model

• Sensitivity Analysis: Different sensitivity analyses to evaluate different factors that affect waterflood performance including relative permeability, mobility ratio, gravity, reservoir dip, offtake rates, etc. These set of tasks is a slight detour from our plan to move from the cross-section data to economic analyses. In practice, data is not known with certainty, and sensitivity analyses will be carried out to quantify uncertainties and risks to the project. The sensitivity analyses tasks have been included to show how the these factors can alter waterflood performance results.

• Economic Analysis - Compute economic indicators using the simulation results. Economic indicators that we will compute include payout time, net present value (NPV), Rate of Return (RoR). NPV and RoR (discounted) are used to rank projects and a project with higher NPV and RoR is usually better. These terms will be described in detail later when we get to the Economic Analyses. Additionally, we will perform some economic sensitivity analyses to investigate the effect of changes in our economic parameters, e.g., oil price, cost of water injection, cost of processing unwanted fluids, etc., on the economic indicators.

  In practice, computing a project’s economic indicators often require a three-dimensional, full-field simulation model containing multiple wells (producers and injectors) and the simulation model is much larger in size compared to the cross-section model considered in the series. However, I have decided to include economic calculations here for a more complete analyses and to demonstrate its important in waterflood project evaluation.

Theme of Series

The theme of the series is described as follows:

• Educational

  • Focus on learning.
  • Improve understanding of waterflooding evaluation and concepts for other colleagues (Managers, Geo-scientists, Drillers, Students, etc) and other stake holders.

• Simple

  • Keep the steps, descriptions and workflows reasonably simple. Different companies and operators have specific workflow adjustments and processes, e.g., choosing optimal well locations, well completions and controls for the wells. Also, the project would often (and should) involve multi-disciplinary teams who contribute different expertise to the overall project evaluation.

• Agnostic

  • Keep descriptions largely agnostic, i.e., without reference to specific company tools. However, I am using an automated workflow processing tool for all the analyses in this tutorial series.
  • Focus on the main ideas relevant to the topic of evaluating waterflood performance

Format

The different tasks involved in evaluating the waterflood performance in the cross-section will be described and analyzed in separate posts with each post adding different pieces as we move from the cross-section data to economic analyses. This (temporary) website will contain full details and analyses of the different tasks. All things being equal, I will make shorter versions available as LinkedIn posts.

Feedback and comments

Feedback and comments are welcome but the previously described theme (educational, simple, and agnostic) should be kept in mind. The feedback and comments should be constructive and should not lead to unnecessary arguments and/or detract from the objectives of the series.

I will do my best to respond to comments in a timely fashion. Also, I will gladly modify content to incorporate useful comments and feedback. However, please expect some delays if incorporating comments and/or feedback require workflow modification and/or code implementation.

What’s next

In the next post, we will describe the cross-section data and the reservoir simulation model that we will use for the waterflood performance evaluation.