Network Constrained Production Optimization by Means of Multiple Shooting (MS)

TABLE OF CONTENTS

List of Tables

List of Figures

1 INTRODUCTION

1.1 PROBLEM STATEMENT

1.2 METHODS OF SOLUTION

1.3 LITERATURE REVIEW OF SOLUTION

1.4 COMPARISONS OF ALTERNATIVE METHODS OF SOLUTION

2 METHODOLOGY (THEORY AND CALCULATIONS)

3 SCIENTIFIC DISCUSSION

3.1 ADVANTAGES OF MS SIMULATION METHOD

3.2 DRAWBACKS OF MS SIMULATION METHOD

4 CONCLUSION AND RECOMMENDATIONS

5 REFERENCE

List of Tables

Table 1: Comparison of Alternative methods of Solution for Optimization

List of Figures

Figure 1: Reservoir-Production Layer Hierarchy

Figure 2: Reservoir Model Sample (source: http://parspk.co/services/)

Figure 3b: Sample ESP Well schematic view

Figure 4: Sample Surface network Diagram

 

1.1         PROBLEM STATEMENT

The main objective for the research was to solve the following problems;

The ability to integrate surface production gathering network constraint with the reservoir performance to achieve long term production optimization.

Ability to model fluid flow from the reservoir to surface network system in steady state condition, understanding that in the reservoir the fluid flow is usually in transient state conditions due to  the dynamic changes in the reservoir pressure.

Waterflooding optimization problem that deals with delaying water breakthrough in producing wells and improving long term oil recovery factor in Waterflooding EOR application.

1.2         METHODS OF SOLUTION

 

For this research study, the chosen method of solution for solving the various problems as state in the Problem statement (2.1) will be the Multiple Shooting (MS) numerical simulation method. The MS method is reviewed more in detail in the literature review section (2.3) of this report.

 

1.3         LITERATURE REVIEW OF SOLUTION

 

Multiple Shooting (MS) method is a well-known optimization tool in process control. It is designed to handle processes with several output constraints. Multiple shooting numerical analysis using interval analysis is typically used to find upper and lower bounds for the solution of two-point boundary value problems in ordinary differential equations.  In MS method, the investigation horizon (time period) is split into number of time frames or intervals instead of a single time frame as we have in Single shooting (SS) method. This allows for the various time frame to be given it’s own independent initial conditions and control variables, and thus the output state and auxiliary variables become available explicitly in the optimizer at each interval boundaries across the investigation horizon (total time period). As a result, output constraints can be applied in MS formulation by imposing constraints to the optimization variables that become available at the boundaries of the individual shooting intervals. Further, the MS formulation is attractive because the shooting intervals can be simulated in parallel creating opportunities to consider output constraints in a much broader scale than we can do currently.

                                                                            Figure 1: Reservoir-Production Layer Hierarchy

 

1.4         COMPARISONS OF ALTERNATIVE METHODS OF SOLUTION

Table 1: Comparison of Alternative methods of Solution for Optimization

Single Shooting (SS) Method

Multiple Shooting (MS) Method

 The SS method

The paper focuses on the challenge of optimizing constrained production network using the multiple shooting method. The paper presented the reservoir model, wellbore system and the surface production network to be used in the analysis and applied in solving the optimization problem. The author of the paper applied the following methodology was used in the presentation of the paper;

The multiple shooting method was reviewed and its application in solving the optimization challenge of network production system was highlighted. The governing equations for the MS method are presented as;

……………… eq. 1

……………………. Eq 2

………………….. Eq 3

The reservoir model used in carrying out the simulation in MRST was described appropriately.

Figure 2: Reservoir Model Sample (source: http://parspk.co/services/)

The ESP system in the production well profile was appropriately described and the equations governing the operational limits were clearly expressed as follows;

These stated equations are used to determine the safe operational limits of the ESP systems which can be captured in the figure below (Figure 3)

                                                                    Figure 3a: ESP Operating Curves

Figure 3b: Sample ESP Well schematic view

It is important to

The surface network production model used for the analysis was appropriately described and presented in the paper.

Figure 4: Sample Surface network Diagram

The key performance index used in the analysis to show the optimization solution obtained was the “Net Present Value (NPV)” and the governing equation is presented as follows;

 

Two case studies were used to assess the implementation the MS method solution approach to a production network constrained optimization problem to see the viability and practicality of the solution method

From the results of the simulation carried out with MS method, we can see how the integration of constrained inputs and constrained state output are optimized to deliver the intended increase in NPV results. The simulation highlighted some key advantages and drawbacks that we can see in this optimization approach, which we intend to review in the following subsection.

3.1         ADVANTAGES OF MS SIMULATION METHOD

From the results of the paper we can deduce the following advantages.  MS Optimization solution is an innovative approach for Production Optimization. The advantages are as follows;

The MS method is an effective computational solution for simulating full field grid reservoir models with applied surface production network constraints to address the challenge of reservoir and production optimization.

The MS method shows the conceptual possibilities of carrying out an integrated optimization solution for a field from reservoir to storage tank while addressing all constraints and limitation.

 

3.2         DRAWBACKS OF MS SIMULATION METHOD

Reviewing the paper, we observed some drawbacks. The drawbacks identified are stated as follows;

The MS method RT tool used for the simulating is Mathematically complex for non-technical personnel since MATLAB is not often commonly used for reservoir simulation within the academia.

The main optimization solution was centred on the ensuring the operational limit of the ESP was maintained during operation, but the problem statement of the paper does not capture the challenge of operating outside the ESP limit as per

The key metric (NPV) for measuring the success of the optimization solution did not consider some key component in the cost structure used in the governing equations.

There wasn’t a clear description of the effect of the injection in production, only the after effect of the water cut in the production.

The single shooting (single control) method applied for the comparison of  NPVs of different strategies in the MS method simulation, should be expatiated to describe the strategies used in the single shooting method.

Based on the paper review carried out, our concluding statements are as follows;

The comparison of the

The key metric

The recommendations suggested based on the drawbacks of the MS method observed, are as follows;

The NPV governing equation needs to be improved to capture the initial investments and full operational cost for running the surface production network and facilities.

The study of the application of the MS method for optimization of reservoir to production network constraint should be extended to other artificial lift methods such as Gaslift and alternative EOR methods like CO2 injection, WAG injection, etc.

Further study should be carried out to account for two phase flow and saturated reservoirs to see if the method is still conceptually possible and computationally effective.

5           References

Amao, Matthew. Artificial Lift Methods and Surface Operations . 30 September 2013. October 2018. <https://fac.ksu.edu.sa/sites/default/files/3-artificialliftsystems.pdf>.

Canadian Oilwell Systems Company Ltd. Basic Artificial Lift. 2018. 15 October 2018. <http://www.coscoesp.com/esp/basic%20artificial%20lift%20tech%20paper/Basic%20Artificial%20Lift.pdf>.

Grand View Research. Artificial Lift Systems Market Analysis By Product (Rod Lifts, Electrical Submersible Pumps, Progressive Cavity Pumps, Gas Lift, Hydraulic Pumps), By Region, And Segment Forecasts, 2018 – 2025. May 2017. October 2018. <https://www.grandviewresearch.com/industry-analysis/artificial-lift-systems-market>.

RIGZONE. How does Artificial Lift Works? 2018. 4 November 2018. <https://www.rigzone.com/training/insight.asp?insight_id=315>.